Pt. 60, App. A 40 CFR Ch. I (7–1–98 Edition)
stainless steel, or other nonporous ma- Method 2D—Measurement of gas volumetric terial of suitable thickness. flow rates in small pipes and ducts (c) Each owner or operator seeking to Method 2E—Determination of landfill gas; comply with § 60.752(b)(2)(i)(A) shall gas production flow rate Method 3—Gas analysis for carbon dioxide, convey the landfill gas to a control sys- oxygen, excess air, and dry molecular tem in compliance with § 60.752(b)(2)(iii) weight through the collection header pipe(s). Method 3A—Determination of Oxygen and The gas mover equipment shall be sized Carbon Dioxide Concentrations in Emis- to handle the maximum gas generation sions From Stationary Sources (Instru- flow rate expected over the intended mental Analyzer Procedure) use period of the gas moving equipment Method 3B—Gas analysis for the determina- using the following procedures: tion of emission rate correction factor or (1) For existing collection systems, excess air Method 3C—Determination of carbon diox- the flow data shall be used to project ide, methane, nitrogen, and oxygen from the maximum flow rate. If no flow data stationary sources exists, the procedures in paragraph Method 4—Determination of moisture con- (c)(2) of this section shall be used. tent in stack gases (2) For new collection systems, the Method 5—Determination of particulate maximum flow rate shall be in accord- emissions from stationary sources ance with § 60.755(a)(1). Method 5A—Determination of particulate emissions from the asphalt processing [61 FR 9919, Mar. 12, 1996, as amended at 63 and asphalt roofing industry FR 32753, June 16, 1998] Method 5B—Determination of nonsulfuric EFFECTIVE DATE NOTE: At 63 FR 32753, June acid particulate matter from stationary 16, 1998, in § 60.759, paragraph (a)(3)(iii) was sources amended by revising the first and second Method 5C [Reserved] sentences, effective Aug. 17, 1998. For the Method 5D—Determination of particulate convenience of the user, the superseded text emissions from positive pressure fabric is set forth as follows: filters Method 5E—Determination of particulate § 60.759 Specifications for active collection emissions from the wool fiberglass insu- systems. lation manufacturing industry (a)* * * Method 5F—Determination of nonsulfate (3) * * * particulate matter from stationary (i) * * * sources (ii) * * * Method 5G—Determination of particulate (iii) The values for k, Lo, and CNM OC deter- emissions from wood heaters from a dilu- mined in field testing shall be used, if field tion tunnel sampling location testing has been performed in determining Method 5H—Determination of particulate the NMOC emission rate or the radii of influ- emissions from wood heaters from a ence. If field testing has not been performed, stack location the default values for k, Lo and CNM OC pro- Method 6—Determination of sulfur dioxide vided in § 60.754(a)(1) shall be used. emissions from stationary sources Method 6A—Determination of sulfur dioxide, * * * * * moisture, and carbon dioxide emissions from fossil fuel combustion sources APPENDIX A TO PART 60—TEST METHODS Method 6B—Determination of sulfur dioxide and carbon dioxide daily average emis- Method 1—Sample and velocity traverses for sions from fossil fuel combustion sources stationary sources Method 6C—Determination of Sulfur Dioxide Method 1A—Sample and velocity traverses Emissions From Stationary Sources (In- for stationary sources with small stacks strumental Analyzer Procedure) or ducts Method 7—Determination of nitrogen oxide Method 2—Determination of stack gas veloc- emissions from stationary sources ity and volumetric flow rate (Type S Method 7A—Determination of nitrogen oxide pitot tube) emissions from stationary sources—Ion Method 2A—Direct measurement of gas vol- chromatographic method ume through pipes and small ducts Method 7B—Determination of nitrogen oxide Method 2B—Determination of exhaust gas emissions from stationary sources (Ul- volume flow rate from gasoline vapor in- traviolet spectrophotometry) cinerators Method 7C—Determination of nitrogen oxide Method 2C—Determination of stack gas ve- emissions from stationary sources—Al- locity and volumetric flow rate in small kaline-permanganate/colorimetric meth- stacks or ducts (standard pitot tube) od
574
VerDate 11
Method 7D—Determination of nitrogen oxide Method 20—Determination of nitrogen ox- emissions from stationary sources—Al- ides, sulfur dioxide, and diluent emis- kaline-permanganate/ion sions from stationary gas turbines chromatographic method Method 21—Determination of volatile or- Method 7E—Determination of Nitrogen Ox- ganic compound leaks ides Emissions From Stationary Sources Method 22—Visual determination of fugitive (Instrumental Analyzer Procedure) emissions from material sources and Method 8—Determination of sulfuric acid smoke emissions from flares mist and sulfur dioxide emissions from Method 23—Determination of Poly- stationary sources chlorinated Dibenzo-p-Dioxins and Poly- Method 9—Visual determination of the opac- chlorinated Dibenzofurans From Station- ity of emissions from stationary sources ary Sources Alternate method 1—Determination of the Method 24—Determination of volatile matter opacity of emissions from stationary content, water content, density, volume sources remotely by lidar solids, and weight solids of surface coat- Method 10—Determination of carbon mon- ings oxide emissions from stationary sources Method 24A—Determination of volatile mat- Method 10A—Determination of carbon mon- ter content and density of printing inks oxide emissions in certifying continuous and related coatings emission monitoring systems at petro- Method 25—Determination of total gaseous leum refineries nonmethane organic emissions as carbon Method 10B—Determination of carbon mon- Method 25A—Determination of total gaseous oxide emissions from stationary sources organic concentration using a flame ion- Method 11—Determination of hydrogen sul- ization analyzer fide content of fuel gas streams in petro- Method 25B—Determination of total gaseous leum refineries organic concentration using a nondisper- Method 12—Determination of inorganic lead sive infrared analyzer emissions from stationary sources Method 25C—Determination of nonmethane organic compounds (NMOC) in MSW Method 13A—Determination of total fluoride landfill gases emissions from stationary sources— SPADNS zirconium lake method Method 25D—Determination of the Volatile Organic Concentration of Waste Samples Method 13B—Determination of total fluoride Method 25E—Determination of Vapor Phase emissions from stationary sources—Spe- Organic Concentration in Waste Samples cific ion electrode method Method 26—Determination of Hydrogen Chlo- Method 14—Determination of fluoride emis- ride Emissions From Stationary Sources sions from potroom roof monitors for pri- Method 27—Determination of vapor tightness mary aluminum plants of gasoline delivery tank using pressure- Method 14A— Determination of Total Fluo- vacuum test ride Emissions from Selected Sources at Method 28—Certification and auditing of Primary Aluminum Production Facili- wood heaters ties Method 28A—Measurement of air to fuel Method 15—Determination of hydrogen sul- ratio and minimum achievable burn fide, carbonyl sulfide, and carbon disul- rates for wood-fired appliances fide emissions from stationary sources Method 29—Determination of metals emis- Method 15A—Determination of total reduced sions from stationary sources sulfur emissions from sulfur recovery The test methods in this appendix are re- plants in petroleum refineries ferred to in § 60.8 (Performance Tests) and Method 16—Semicontinuous determination § 60.11 (Compliance With Standards and of sulfur emissions from stationary Maintenance Requirements) of 40 CFR part sources 60, subpart A (General Provisions). Specific Method 16A—Determination of total reduced uses of these test methods are described in sulfur emissions from stationary sources the standards of performance contained in (impinger technique) the subparts, beginning with Subpart D. Method 16B—Determination of total reduced Within each standard of performance, a sulfur emissions from stationary sources section title ‘‘Test Methods and Procedures’’ Method 17—Determination of particulate is provided to: (1) Identify the test methods emissions from stationary sources (in- to be used as reference methods to the facil- stack filtration method) ity subject to the respective standard and (2) Method 18—Measurement of gaseous organic identify any special instructions or condi- compound emissions by gas chroma- tions to be followed when applying a method tography to the respective facility. Such instructions Method 19—Determination of sulfur dioxide (for example, establish sampling rates, vol- removal efficiency and particulate, sul- umes, or temperatures) are to be used either fur dioxide and nitrogen oxides emission in addition to, or as a substitute for proce- rates dures in a test method. Similarly, for
575
VerDate 11
sources subject to emission monitoring re- cluding a written description of the alter- quirements, specific instructions pertaining native method in the test report (the written to any use of a test method as a reference method must be clear and must be capable of method are provided in the subpart or in Ap- being performed without additional instruc- pendix B. tion, and the the degree of detail should be Inclusion of methods in this appendix is similar to the detail contained in the test not intended as an endorsement or denial of methods); and (3) providing any rationale or their applicability to sources that are not supporting data necessary to show the valid- subject to standards of performance. The ity of the alternative in the particular appli- methods are potentially applicable to other cation. Failure to meet these requirements sources; however, applicability should be can result in the Administrator’s disapproval confirmed by careful and appropriate evalua- of the alternative. tion of the conditions prevalent at such sources. METHOD 1—SAMPLE AND VELOCITY The approach followed in the formulation TRAVERSES FOR STATIONARY SOURCES of the test methods involves specifications for equipment, procedures, and performance. 1. Principle and Applicability In concept, a performance specification ap- 1.1 Principle. To aid in the representative proach would be preferable in all methods measurement of pollutant emissions and/or because this allows the greatest flexibility total volumetric flow rate from a stationary to the user. In practice, however, this ap- source, a measurement site where the efflu- proach is impractical in most cases because ent stream is flowing in a known direction is performance specifications cannot be estab- selected, and the cross-section of the stack is lished. Most of the methods described herein, divided into a number of equal areas. A tra- therefore, involve specific equipment speci- verse point is then located within each of fications and procedures, and only a few these equal areas. methods in this appendix rely on perform- 1.2 Applicability. This method is applica- ance criteria. ble to flowing gas streams in ducts, stacks, Minor changes in the test methods should and flues. The method cannot be used when: not necessarily affect the validity of the re- (1) flow is cyclonic or swirling (see Section sults and it is recognized that alternative 2.4), (2) a stack is smaller than about 0.30 and equivalent methods exist. Section 60.8 meter (12 in.) in diameter, or 0.071 m2(113 provides authority for the Administrator to in.2) cross-sectional area, or (3) the measure- specify or approve (1) equivalent methods, (2) ment site is less than two stack or duct di- alternative methods, and (3) minor changes ameters downstream or less than a half di- in the methodology of the test methods. It ameter upstream from a flow disturbance. should be clearly understood that unless oth- The requirements of this method must be erwise identified all such methods and considered before construction of a new fa- changes must have prior approval of the Ad- cility from which emissions will be meas- ministrator. An owner employing such meth- ured; failure to do so may require subsequent ods or deviations from the test methods alterations to the stack or deviation from without obtaining prior approval does so at the standard procedure. Cases involving the risk of subsequent disapproval and re- variants are subject to approval by the Ad- testing with approved methods. ministrator, U.S. Environmental Protection Within the test methods, certain specific Agency. equipment or procedures are recognized as 2. Procedure being acceptable or potentially acceptable and are specifically identified in the meth- 2.1 Selection of Measurement Site. Sam- ods. The items identified as acceptable op- pling or velocity measurement is performed tions may be used without approval but at a site located at least eight stack or duct must be identified in the test report. The po- diameters downstream and two diameters tentially approvable options are cited as upstream from any flow disturbance such as ‘‘subject to the approval of the Adminis- a bend, expansion, or contraction in the trator’’ or as ‘‘or equivalent.’’ Such poten- stack, or from a visible flame. If necessary, tially approvable techniques or alternatives an alternative location may be selected, at a may be used at the discretion of the owner position at least two stack or duct diameters without prior approval. However, detailed downstream and a half diameter upstream descriptions for applying these potentially from any flow disturbance. For a rectangular approvable techniques or alternatives are cross section, an equivalent diameter (De) not provided in the test methods. Also, the shall be calculated from the following equa- potentially approvable options are not nec- tion, to determine the upstream and down- essarily acceptable in all applications. stream distances: Therefore, an owner electing to use such po- 2LW tentially approvable techniques or alter- D = natives is responsible for: (1) assuring that e ()+ the techniques or alternatives are in fact ap- LW plicable and are properly executed; (2) in- where L=length and W=width.
576
VerDate 11
An alternative procedure is available for paring the results with acceptability cri- determining the acceptability of a measure- teria, is described in Section 2.5. ment location not meeting the criteria 2.2 Determining the Number of Traverse above. This procedure, determination of gas Points. flow angles at the sampling points and com-
2.2.1 Particulate Traverses. When the corresponds: (1) to the number of duct diame- eight- and two-diameter criterion can be ters upstream; and (2) to the number of di- met, the minimum number of traverse points ameters downstream. Select the higher of shall be: (1) twelve, for circular or rectangu- the two minimum numbers of traverse lar stacks with diameters (or equivalent di- points, or a greater value, so that for cir- ameters) greater than 0.61 meter (24 in.); (2) cular stacks the number is a multiple of 4, eight, for circular stacks with diameters be- and for rectangular stacks, the number is tween 0.30 and 0.61 meter (12–24 in.); (3) nine, one of those shown in Table 1–1. for rectangular stacks with equivalent diam- eters between 0.30 and 0.61 meter (12–24 in.). TABLE 1±1. CROSS-SECTION LAYOUT FOR When the eight- and two-diameter cri- RECTANGULAR STACKS terion cannot be met, the minimum number of traverse points is determined from Figure Number of traverse points Matrix layout 1–1. Before referring to the figure, however, 9 ...... 3x3 determine the distances from the chosen 12 ...... 4x3 measurement site to the nearest upstream 16 ...... 4x4 and downstream disturbances, and divide 20 ...... 5x4 each distance by the stack diameter or 25 ...... 5x5 equivalent diameter, to determine the dis- 30 ...... 6x5 tance in terms of the number of duct diame- 36 ...... 6x6 ters. Then, determine from Figure 1–1 the 42 ...... 7x6 49 ...... 7x7 minimum number of traverse points that
577
VerDate 11
2.2.2 Velocity (Non-Particulate) Tra- In addition for stacks having diameters verses. When velocity or volumetric flow greater than 0.61 m (24 in.) no traverse points rate is to be determined (but not particulate shall be located within 2.5 centimeters (1.00 matter), the same procedure as that for par- in.) of the stack walls; and for stack diame- ticulate traverses (Section 2.2.1) is followed, ters equal to or less than 0.61 m (24 in.), no except that Figure 1–2 may be used instead traverse points shall be located within 1.3 cm of Figure 1–1. (0.50 in.) of the stack walls. To meet these 2.3 Cross-sectional Layout and Location criteria, observe the procedures given below. of Traverse Points. 2.3.1.1 Stacks With Diameters Greater 2.3.1 Circular Stacks. Locate the traverse Than 0.61 m (24 in.). When any of the traverse points on two perpendicular diameters ac- points as located in Section 2.3.1 fall within cording to Table 1–2 and the example shown 2.5 cm (1.00 in.) of the stack walls, relocate in Figure 1–3. Any equation (for examples, see Citations 2 and 3 in the Bibliography) them away from the stack walls to: (1) a dis- that gives the same values as those in Table tance of 2.5 cm (1.00 in.); or (2) a distance 1–2 may be used in lieu of Table 1–2. equal to the nozzle inside diameter, which- For particulate traverses, one of the diam- ever is larger. These relocated traverse eters must be in a plane containing the points (on each end of a diameter) shall be greatest expected concentration variation, the ‘‘adjusted’’ traverse points. e.g., after bends, one diameter shall be in the Whenever two successive traverse points plane of the bend. This requirement becomes are combined to form a single adjusted tra- less critical as the distance from the disturb- verse point, treat the adjusted point as two ance increases; therefore, other diameter lo- separate traverse points, both in the sam- cations may be used, subject to approval of pling (or velocity measurement) procedure, the Administrator. and in recording the data.
578
VerDate 11
TABLE 1±2. LOCATION OF TRAVERSE POINTS IN CIRCULAR STACKS [Percent of stack diameter from inside wall to traverse point]
Number of traverse points on a diameterÐ Traverse point number on a diameter 2 4 6 8 10 12 14 16 18 20 22 24
1 ...... 14.6 6.7 4.4 3.2 2.6 2.1 1.8 1.6 1.4 1.3 1.1 1.1 2 ...... 85.4 25.0 14.6 10.5 8.2 6.7 5.7 4.9 4.4 3.9 3.5 3.2 3 ...... 75.0 29.6 19.4 14.6 11.8 9.9 8.5 7.5 6.7 6.0 5.5 4 ...... 93.3 70.4 32.3 22.6 17.7 14.6 12.5 10.9 9.7 8.7 7.9 5 ...... 85.4 67.7 34.2 25.0 20.1 16.9 14.6 12.9 11.6 10.5 6 ...... 95.6 80.6 65.8 35.6 26.9 22.0 18.8 16.5 14.6 13.2 7 ...... 89.5 77.4 64.4 36.6 28.3 23.6 20.4 18.0 16.1 8 ...... 96.8 85.4 75.0 63.4 37.5 29.6 25.0 21.8 19.4 9 ...... 91.8 82.3 73.1 62.5 38.2 30.6 26.2 23.0 10 ...... 97.4 88.2 79.9 71.7 61.8 38.8 31.5 27.2 11 ...... 93.3 85.4 78.0 70.4 61.2 39.3 32.3 12 ...... 97.9 90.1 83.1 76.4 69.4 60.7 39.8 13 ...... 94.3 87.5 81.2 75.0 68.5 60.2 14 ...... 98.2 91.5 85.4 79.6 73.8 67.7 15 ...... 95.1 89.1 83.5 78.2 72.8 16 ...... 98.4 92.5 87.1 82.0 77.0 17 ...... 95.6 90.3 85.4 80.6 18 ...... 98.6 93.3 88.4 83.9 19 ...... 96.1 91.3 86.8 20 ...... 98.7 94.0 89.5 21 ...... 96.5 92.1 22 ...... 98.9 94.5 23 ...... 96.8 24 ...... 98.9
2.3.1.2 Stacks With Diameters Equal to or at the centroid of each equal area according Less Than 0.61 m (24 in.). Follow the proce- to the example in Figure 1–4. dure in Section 2.3.1.1, noting only that any If the tester desires to use more than the ‘‘adjusted’’ points should be relocated away minimum number of traverse points, expand from the stack walls to: (1) a distance of 1.3 the ‘‘minimum number of traverse points’’ cm (0.50 in.); or (2) a distance equal to the matrix (see Table 1–1) by adding the extra nozzle inside diameter, whichever is larger. traverse points along one or the other or 2.3.2 Rectangular Stacks. Determine the both legs of the matrix; the final matrix number of traverse points as explained in need not be balanced. For example, if a 4x3 Sections 2.1 and 2.2 of this method. From ‘‘minimum number of points’’ matrix were Table 1–1, determine the grid configuration. expanded to 36 points, the final matrix could Divide the stack cross-section into as many be 9x4 or 12x3, and would not necessarily equal rectangular elemental areas as tra- have to be 6x6. After constructing the final verse points, and then locate a traverse point matrix, divide the stack cross-section into as many equal rectangular, elemental areas as
579
VerDate 11
traverse points, and locate a traverse point The alternative procedure described in at the centroid of each equal area. Section 2.5 may be used to determine the ro- The situation of traverse points being too tation angles in lieu of the procedure de- close to the stack walls is not expected to scribed above. arise with rectangular stacks. If this prob- 2.5 Alternative Measurement Site Selec- lem should ever arise, the Administrator tion Procedure. This alternative applies to must be contacted for resolution of the mat- sources where measurement locations are ter. less than 2 equivalent stack or duct diame- 1 2.4 Verification of Absence of Cyclonic ters downstream or less than ⁄2 duct diame- Flow. In most stationary sources, the direc- ter upstream from a flow disturbance. The tion of stack gas flow is essentially parallel alternative should be limited to ducts larger than 24 in. in diameter where blockage and to the stack walls. However, cyclonic flow wall effects are minimal. A directional flow- may exist (1) after such devices as cyclones sensing probe is used to measure pitch and and inertial demisters following venturi yaw angles of the gas flow at 40 or more tra- scrubbers, or (2) in stacks having tangential verse points; the resultant angle is cal- inlets or other duct configurations which culated and compared with acceptable cri- tend to induce swirling; in these instances, teria for mean and standard deviation. the presence or absence of cyclonic flow at the sampling location must be determined. NOTE: Both the pitch and yaw angles are The following techniques are acceptable for measured from a line passing through the traverse point and parallel to the stack axis. this determination. The pitch angle is the angle of the gas flow component in the plane that INCLUDES the traverse line and is parallel to the stack axis. The yaw angle is the angle of the gas flow component in the plane PERPENDICU- LAR to the traverse line at the traverse point and is measured from the line passing through the traverse point and parallel to the stack axis. 2.5.1 Apparatus. 2.5.1.1 Directional Probe. Any directional probe, such as United Sensor Type DA Three- Dimensional Directional Probe, capable of measuring both the pitch and yaw angles of gas flows is acceptable. (NOTE: Mention of trade name or specific products does not con- stitute endorsement by the U.S. Environ- Level and zero the manometer. Connect a mental Protection Agency.) Assign an iden- Type S pitot tube to the manometer. Posi- tification number to the directional probe, tion the Type S pitot tube at each traverse and permanently mark or engrave the num- point, in succession, so that the planes of the ber on the body of the probe. The pressure face openings of the pitot tube are per- holes of directional probes are susceptible to pendicular to the stack cross-sectional plugging when used in particulate-laden gas plane; when the Type S pitot tube is in this streams. Therefore, a system for cleaning position, it is at ‘‘0° reference.’’ Note the dif- the pressure holes by ‘‘back-purging’’ with ferential pressure (∆p) reading at each tra- pressurized air is required. 2.5.1.2 Differential Pressure Gauges. In- verse point. If a null (zero) pitot reading is clined manometers, U-tube manometers, or obtained at 0° reference at a given traverse other differential pressure gauges (e.g., point, an acceptable flow condition exists at magnehelic gauges) that meet the specifica- that point. If the pitot reading is not zero at tions described in Method 2, section 2.2. 0° reference, rotate the pitot tube (up to ±90° yaw angle), until a null reading is obtained. NOTE: If the differential pressure gauge Carefully determine and record the value of produces both negative and positive read- the rotation angle (α) to the nearest degree. ings, then both negative and positive pres- After the null technique has been applied at sure readings shall be calibrated at a mini- each traverse point, calculate the average of mum of three points as specified in Method 2, the absolute values of α; assign αvalues of 0° section 2.2. to those points for which no rotation was re- 2.5.2 Traverse Points. Use a minimum of quired, and include these in the overall aver- 40 traverse points for circular ducts and 42 age. If the average value of αis greater than points for rectangular ducts for the gas flow 20°, the overall flow condition in the stack is angle determinations. Follow section 2.3 and unacceptable and alternative methodology, Table 1–1 or 1–2 for the location and layout subject to the approval of the Administrator, of the traverse points. If the measurement must be used to perform accurate sample and location is determined to be acceptable ac- velocity traverses. cording to the criteria in this alternative
580
VerDate 11
procedure, use the same traverse point num- ber and locations for sampling and velocity n measurements. ∑()RR− 2 2.5.3 Measurement Procedure. i 2.5.3.1 Prepare the directional probe and S = i=1 differential pressure gauges as recommended d − by the manufacturer. Capillary tubing or ()n 1 surge tanks may be used to dampen pressure Eq. 1–4 fluctuations. It is recommended, but not re- quired, that a pretest leak check be con- Where:
ducted. To perform a leak check, pressurize Sd=Standard deviation, degree. or use suction on the impact opening until a 2.5.5 The measurement location is accept- reading of at least 7.6 cm (3 in.) H O registers 2 able if R¯ ≤20° and S ≤10°. on the differential pressure gauge, then plug d the impact opening. The pressure of a leak- 2.5.6 Calibration. Use a flow system as de- free system will remain stable for at least 15 scribed in Sections 4.1.2.1 and 4.1.2.2 of Meth- seconds. od 2. In addition, the flow system shall have 2.5.3.2 Level and zero the manometers. the capacity to generate two test-section ve- Since the manometer level and zero may locities: one between 365 and 730 m/min (1200 drift because of vibrations and temperature and 2400 ft/min) and one between 730 and 1100 changes, periodically check the level and m/min (2400 and 3600 ft/min). zero during the traverse. 2.5.6.1 Cut two entry ports in the test sec- 2.5.3.3 Position the probe at the appro- tion. The axis through the entry ports shall priate locations in the gas stream, and ro- be perpendicular to each other and intersect tate until zero deflection is indicated for the in the centroid of the test section. The ports yaw angle pressure gauge. Determine and should be elongated slots parallel to the axis record the yaw angle. Record the pressure of the test section and of sufficient length to gauge readings for the pitch angle, and de- allow measurement of pitch angles while termine the pitch angle from the calibration maintaining the pitot head position at the curve. Repeat this procedure for each tra- test-section centroid. To facilitate align- verse point. Complete a ‘‘back-purge’’ of the ment of the directional probe during calibra- pressure lines and the impact openings prior tion, the test section should be constructed to measurements of each traverse point. of plexiglass or some other transparent ma- A post-test check as described in section terial. All calibration measurements should 2.5.3.1 is required. If the criteria for a leak- be made at the same point in the test sec- free system are not met, repair the equip- tion, preferably at the centroid of the test ment, and repeat the flow angle measure- section. ments. 2.5.6.2 To ensure that the gas flow is par- 2.5.4 Calculate the resultant angle at each allel to the central axis of the test section, traverse point, the average resultant angle, follow the procedure in Section 2.4 for cy- and the standard deviation using the follow- clonic flow determination to measure the ing equations. Complete the calculations re- gas flow angles at the centroid of the test taining at least one extra significant figure section from two test ports located 90°apart. beyond that of the acquired data. Round the The gas flow angle measured in each port values after the final calculations. must be ±2° of 0°. Straightening vanes should 2.5.4.1 Calculate the resultant angle at be installed, if necessary, to meet this cri- each traverse point: terion. Ri=arc cosine [(cosine Yi)(cosine Pi)] 2.5.6.3 Pitch Angle Calibration. Perform a calibration traverse according to the manu- Eq. 1–2 facturer’s recommended protocol in 5° incre- Where: ments for angles from ¥60° to +60° at one ve- Ri=Resultant angle at traverse point i, de- locity in each of the two ranges specified gree. above. Average the pressure ratio values ob- Yi=Yaw angle at traverse point i, degree. tained for each angle in the two flow ranges, Pi=Pitch angle at traverse point i, degree. and plot a calibration curve with the average 2.5.4.2 Calculate the average resultant for values of the pressure ratio (or other suit- the measurements: able measurement factor as recommended by the manufacturer) versus the pitch angle. ∑ R Draw a smooth line through the data points. = i Plot also the data values for each traverse R Eq. 13 - point. Determine the differences between the n measured data values and the angle from the Where: calibration curve at the same pressure ratio. ¯ R=Average resultant angle, degree. The difference at each comparison must be n=Total number of traverse points. within 2° for angles between 0° and 40° and 2.5.4.3 Calculate the standard deviations: within 3°for angles between 40 ° and 60°.
581 Pt. 60, App. A, Meth. 1A 40 CFR Ch. I (7–1–98 Edition)
2.5.6.4 Yaw Angle Calibration. Mark the sion Standards and Engineering Division. three-dimensional probe to allow the deter- U.S. Environmental Protection Agency, Re- mination of the yaw position of the probe. search Triangle Park, NC. 27711. July 31, 1980. This is usually a line extending the length of 12 p. the probe and aligned with the impact open- 11. Hawksley, P.G.W., S. Badzioch, and J.H. ing. To determine the accuracy of measure- Blackett. Measurement of Solids in Flue ments of the yaw angle, only the zero or null Gases. Leatherhead, England, The British position need be calilbrated as follows. Place Coal Utilisation Research Association, 1961. the directional probe in the test section, and p. 129–133. rotate the probe until the zero position is 12. Knapp, K.T. The Number of Sampling found. With a protractor or other angle Points Needed for Representative Source measuring device, measure the angle indi- Sampling. In: Proceedings of the Fourth Na- cated by the yaw angle indicator on the tional Conference on Energy and the Envi- three-dimensional probe. This should be ronment, Theodore, L., et al. (ed.). Dayton, ° ° within 2 of 0 . Repeat this measurement for Dayton Section of the American Institute of any other points along the length of the Chemical Engineers. October 3–7, 1976. p. 563– pitot where yaw angle measurements could 568. be read in order to account for variations in 13. Smith, W.S. and D.J. Grove. A Proposed the pitot markings used to indicate pitot Extension of EPA Method 1 Criteria. ‘‘Pollu- head positions. tion Engineering.’’ XV (8):36-37. August 1983. 3. Bibliography 14. Gerhart, P.M. and M.J. Dorsey. Inves- 1. Determining Dust Concentration in a tigation of Field Test Procedures for Large Gas Stream, ASME. Performance Test Code Fans. University of Akron. Akron, OH. No. 27. New York, 1957. (EPRI Contract CS–1651). Final Report (RP– 2. Devorkin, Howard, et al. Air Pollution 1649–5) December 1980. Source Testing Manual. Air Pollution Con- 15. Smith, W.S. and D.J. Grove. A New trol District. Los Angeles, CA November Look at Isokinetic Sampling—Theory and 1963. Applications. ‘‘Source Evaluation Society 3. Methods for Determination of Velocity, Newsletter.’’ VIII (3):19–24. August 1983. Volume, Dust and Mist Content of Gases. Western Precipitation Division of Joy Manu- METHOD 1A—SAMPLE AND VELOCITY TRA- facturing Co. Los Angeles, CA. Bulletin WP– VERSES FOR STATIONARY SOURCES WITH 50. 1968. SMALL STACKS OR DUCTS 4. Standard Method for Sampling Stacks for Particulate Matter. In: 1971 Book of 1. Applicability and Principle ASTM Standards, Part 23. ASTM Designa- tion D–2928–71. Philadelphia, PA 1971. 1.1 The applicability and principle of this 5. Hanson, H.A., et al. Particulate Sam- method are identical to Method 1, except pling Strategies for Large Power Plants In- this method’s applicability is limited to cluding Nonuniform Flow. USEPA, ORD, stacks or ducts less than about 0.30 meter (12 ESRL, Research Triangle Park, NC. EPA–600/ in.) in diameter or 0.071 m2 (113 in.2) in cross- 2–76–170, June 1976. sectional area, but equal to or greater than 6. Entropy Environmentalists, Inc. Deter- about 0.10 meter (4 in.) in diameter or 0.0081 mination of the Optimum Number of Sam- m2 (12.57 in.2) in cross-sectional area. pling Points: An Analysis of Method 1 Cri- 1.2 In these small diameter stacks or teria. Environmental Protection Agency, Re- ducts, the conventional Method 5 stack as- search Triangle Park, NC. EPA Contract No. sembly (consisting of a Type S pitot tube at- 68–01–3172, Task 7. tached to a sampling probe, equipped with a 7. Hanson, H.A., R.J. Davini, J.K. Morgan, nozzle and thermocouple) blocks a signifi- and A.A. Iversen. Particulate Sampling cant portion of the cross section of the duct Strategies for Large Power Plants Including and causes inaccurate measurements. There- Nonuniform Flow. U.S. Environmental Pro- fore, for particulate matter (PM) sampling in tection Agency. Research Triangle Park, NC. small stacks or ducts, the gas velocity is Publication No. EPA–600/2–76–170. June 1976. measured using a standard pitot tube down- 350 p. stream of the actual emission sampling site. 8. Brooks, E.F., and R.L. Williams. Flow The straight run of duct between the PM and Gas Sampling Manual. U.S. Environ- sampling and velocity measurement sites al- mental Protection Agency. Research Tri- lows the flow profile, temporarily disturbed angle Park, NC. Publication No. EPA–600/2– by the presence of the sampling probe, to re- 76–203. July 1976. 93 p. develop and stabilize. 9. Entropy Environmentalists, Inc. Tra- 1.3 The cross-sectional layout and loca- verse Point Study. EPA Contract No. 68–02– tion of traverse points and the verification of 3172. June 1977. 19 p. the absence of cyclonic flow are the same as 10. Brown, J. and K. Yu. Test Report: Par- in Method 1, Sections 2.3 and 2.4, respec- ticulate Sampling Strategy in Circular tively. Differences from Method 1, except as Ducts. Emission Measurement Branch, Emis- noted, are given below.
582
VerDate 11
2. Procedure stream of the PM sampling site. See Figure 2.1 Selection of Sampling and Measure- 1A–1. If such locations are not available, se- ment Sites. lect an alternative PM sampling site that is 2.1.1 PM Measurements. Select a PM sam- at least 2 equivalent stack or duct diameters pling site located preferably at least 8 equiv- downstream and 21⁄2 diameters upstream alent stack or duct diameters downstream from any flow disturbance. Then, locate the and 10 equivalent diameters upstream from velocity measurement site 2 equivalent di- any flow disturbances such as bends, expan- ameters downstream from the PM sampling sions, or contractions in the stack, or from a site. Follow Section 2.1 of Method 1 for cal- visible flame. Next, locate the velocity meas- culating equivalent diameters for a rectan- urement site 8 equivalent diameters down- gular cross section.
583
VerDate 11
2.1.2 PM Sampling (Steady Flow) or only Method 1 may be followed, with the PM sam- Velocity Measurements. For PM sampling pling and velocity measurement performed when the volumetric flow rate in a duct is at one location. To demonstrate that the constant with respect to time, Section 2.1 of flow rate is constant (within 10 percent)
584
VerDate 11
when PM measurements are made, perform 3. Bibliography complete velocity traverses before and after 1. Same as in Method 1, Section 3, Cita- the PM sampling run, and calculate the devi- tions 1 through 6. ation of the flow rate derived after the PM 2. Vollaro, Robert F. Recommended Proce- sampling run from the one derived before the dure for Sample Traverses in Ducts Smaller PM sampling run. The PM sampling run is Than 12 Inches in Diameter. U.S. Environ- acceptable if the deviation does not exceed 10 mental Protection Agency, Emission Meas- percent. urement Branch, Research Triangle Park, 2.2 Determining the Number of Traverse NC. January 1977. Points. 2.2.1 PM Sampling. Use Figure 1–1 of METHOD 2—DETERMINATION OF STACK GAS VE- Method 1 to determine the number of tra- LOCITY AND VOLUMETRIC FLOW RATE (TYPE S PITOT TUBE) verse points to use at both the velocity measurement and PM sampling locations. 1. Principle and Applicability Before referring to the figure, however, de- 1.1 Principle. The average gas velocity in termine the distances between both the ve- a stack is determined from the gas density locity measurement and PM sampling sites and from measurement of the average veloc- to the nearest upstream and downstream dis- ity head with a Type S (Stausscheibe or re- turbances. Then divide each distance by the verse type) pitot tube. stack diameter or equivalent diameter to ex- 1.2 Applicability. This method is applica- press the distances in terms of the number of ble for measurement of the average velocity duct diameters. Next, determine the number of a gas stream and for quantifying gas flow. of traverse points from Figure 1–1 of Method This procedure is not applicable at meas- 1 corresponding to each of these four dis- urement sites which fail to meet the criteria tances. Choose the highest of the four num- of Method 1, Section 2.1. Also, the method bers of traverse points (or a greater number) cannot be used for direct measurement in cy- so that, for circular ducts, the number is a clonic or swirling gas streams; Section 2.4 of Method 1 shows how to determine cyclonic multiple of four, and for rectangular ducts, or swirling flow conditions. When unaccept- the number is one of those shown in Table 1– able conditions exist, alternative procedures, 1 of Method 1. When the optimum duct diam- subject to the approval of the Administrator, eter location criteria can be satisfied, the U.S. Environmental Protection Agency, minimum number of traverse points required must be employed to make accurate flow is eight for circular ducts and nine for rec- rate determinations; examples of such alter- tangular ducts. native procedures are: (1) to install straight- 2.2.2 PM Sampling (Steady Flow) or Ve- ening vanes; (2) to calculate the total volu- locity Measurements. Use Figure 1–2 of metric flow rate stoichiometrically, or (3) to Method 1 to determine the number of tra- move to another measurement site at which verse points, following the same procedure the flow is acceptable. used for PM sampling traverses as described 2. Apparatus in Section 2.2.1 of Method 1. When the opti- Specifications for the apparatus are given mum duct diameter location criteria can be below. Any other apparatus that has been satisfied, the minimum number of traverse demonstrated (subject to approval of the Ad- points required is eight for circular ducts ministrator) to be capable of meeting the and nine for rectangular ducts. specifications will be considered acceptable.
585
VerDate 11
2.1 Type S Pitot Tube. The Type S pitot The face openings of the pitot tube shall, tube (Figure 2–1) shall be made of metal tub- preferably, be aligned as shown in Figure 2– ing (e.g., stainless steel). It is recommended 2; however, slight misalignments of the open- that the external tubing diameter (dimen- ings are permissible (see Figure 2–3). sion Dt Figure 2–2b) be between 0.48 and 0.95 The Type S pitot tube shall have a known 3 3 centimeter ( ⁄16 and ⁄8 inch). There shall be coefficient, determined as outlined in Sec- an equal distance from the base of each leg tion 4. An identification number shall be as- of the pitot tube to its face-opening plane signed to the pitot tube; this number shall be (dimensions P and P Figure 2–2b); it is rec- A B permanently marked or engraved on the ommended that this distance be between 1.05 and 1.50 times the external tubing diameter. body of the tube.
586
VerDate 11
Figure 2-2. Properly constructed Type S pitot tube, shown in: (a) end view; face opening planes perpendicular to transverse axis; (b) top view; face opening planes parallel to longitudinal axis; (c) side view; both legs of equal length and centerlines coincident, when viewed from both sides. Baseline coefficient values of 0.84 may be assigned to pitot tubes constructed this way.
587
VerDate 11
Figure 2-3. Types of face-opening misalignment that can result from field use or improper construction of Type S pitot tubes. These will not affect the baseline value of Cp(s) so long as α1 and α2 ≤ 10°, β1 and β2 ≤ 5°, z ≤ 0.32 cm (1/8 in.) and w ≤ 0.08 cm (1/32 in.) (citation 11 in Bibliography).
A standard pitot tube may be used instead static holes of the standard pitot tube by of a Type S, provided that it meets the speci- ‘‘back-purging’’ with pressurized air, and fications of Sections 2.7 and 4.2; note, how- then taking another ∆p reading. If the ∆p ever, that the static and impact pressure readings made before and after the air purge holes of standard pitot tubes are susceptible are the same (±5 percent), the traverse is ac- to plugging in particulate-laden gas streams. ceptable. Otherwise, reject the run. Note Therefore, whenever a standard pitot tube is that if ∆p at the final traverse point is un- used to perform a traverse, adequate proof suitably low, another point may be selected. must be furnished that the openings of the If ‘‘back-purging’’ at regular intervals is part pitot tube have not plugged up during the of the procedure, then comparative ∆p read- traverse period; this can be done by taking a ings shall be taken, as above, for the last two velocity head (∆p) reading at the final tra- back purges at which suitably high ∆p read- verse point, cleaning out the impact and ings are observed.
588
VerDate 11
2.2 Differential Pressure Gauge. An in- thermometer, mercury-in-glass thermom- clined manometer or equivalent device is eter, or other gauge, capable of measuring used. Most sampling trains are equipped with temperature to within 1.5 percent of the min- a 10-in. (water column) inclined-vertical ma- imum absolute stack temperature shall be nometer, having 0.01-in. H2O divisions on the used. The temperature gauge shall be at- 0-to 1-in. inclined scale, and 0.1-in. H2O divi- tached to the pitot tube such that the sensor sions on the 1- to 10-in. vertical scale. This tip does not touch any metal; the gauge shall type of manometer (or other gauge of equiva- be in an interference-free arrangement with lent sensitivity) is satisfactory for the meas- respect to the pitot tube face openings (see urement of ∆p values as low as 1.3 mm (0.05 Figure 2–1 and also Figure 2–7 in Section 4). in.) H2O. However, a differential pressure Alternative positions may be used if the gauge of greater sensitivity shall be used pitot tube-temperature gauge system is cali- (subject to the approval of the Adminis- brated according to the procedure of Section trator), if any of the following is found to be 4. Provided that a difference of not more true: (1) the arithmetic average of all ∆p than 1 percent in the average velocity meas- readings at the traverse points in the stack urement is introduced, the temperature is less than 1.3 mm (0.05 in.) H2O; (2) for tra- gauge need not be attached to the pitot tube; verses of 12 or more points, more than 10 per- this alternative is subject to the approval of cent of the individual ∆p readings are below the Administrator. 1.3 mm (0.05 in.) H2O; (3) for traverses of 2.4 Pressure Probe and Gauge. A piezom- ∆ fewer than 12 points, more than one p read- eter tube and mercury- or water-filled U- ing is below 1.3 mm (0.05 in.) H2O. Citation 18 tube manometer capable of measuring stack in Bibliography describes commercially pressure to within 2.5 mm (0.1 in.) Hg is used. available instrumentation for the measure- The static tap of a standard type pitot tube ment of low-range gas velocities. or one leg of a Type S pitot tube with the As an alternative to criteria (1) through (3) face opening planes positioned parallel to above, the following calculation may be per- the gas flow may also be used as the pressure formed to determine the necessity of using a probe. more sensitive differential pressure gauge: 2.5 Barometer. A mercury, aneroid, or other barometer capable of measuring at- mospheric pressure to within 2.5 mm Hg (0.1 in. Hg) may be used. In many cases, the baro- metric reading may be obtained from a near- by National Weather Service station, in which case the station value (which is the Where: absolute barometric pressure) shall be re- quested and an adjustment for elevation dif- ∆pi=Individual velocity head reading at a ferences between the weather station and the traverse point, mm H2O (in. H2O). n=Total number of traverse points. sampling point shall be applied at a rate of minus 2.5 mm (0.1 in.) Hg per 30-meter (100 K=0.13 mm H2O when metric units are used foot) elevation increase or vice-versa for ele- and 0.005 in. H2O when English units are used. vation decrease. 2.6 Gas Density Determination Equip- If T is greater than 1.05, the velocity head ment. Method 3 equipment, if needed (see data are unacceptable and a more sensitive Section 3.6), to determine the stack gas dry differential pressure gauge must be used. molecular weight, and Reference Method 4 or NOTE: If differential pressure gauges other Method 5 equipment for moisture content de- than inclined manometers are used (e.g., termination; other methods may be used magnehelic gauges), their calibration must subject to approval of the Administrator. be checked after each test series. To check 2.7 Calibration Pitot Tube. When calibra- the calibration of a differential pressure tion of the Type S pitot tube is necessary gauge, compare ∆p readings of the gauge (see Section 4), a standard pitot tube is used with those of a gauge-oil manometer at a as a reference. The standard pitot tube shall, minimum of three points, approximately preferably, have a known coefficient, ob- representing the range of ∆p values in the tained either (1) directly from the National stack. If, at each point, the values of ∆p as Bureau of Standards, Route 270, Quince Or- read by the differential pressure gauge and chard Road, Gaithersburg, Maryland, or (2) gauge-oil manometer agree to within 5 per- by calibration against another standard cent, the differential pressure gauge shall be pitot tube with an NBS-traceable coefficient. considered to be in proper calibration. Other- Alternatively, a standard pitot tube designed wise, the test series shall either be voided, or according to the criteria given in 2.7.1 procedures to adjust the measured ∆p values through 2.7.5 below and illustrated in Figure and final results shall be used subject to the 2–4 (see also Citations 7, 8, and 17 in Bibliog- approval of the Administrator. raphy) may be used. Pitot tubes designed ac- 2.3 Temperature Gauge. A thermocouple, cording to these specifications will have liquid-filled bulb thermometer, bimetallic baseline coefficients of about 0.99±0.01.
589
VerDate 11
2.7.1 Hemispherical (shown in Figure 2–4), 2.8 Differential Pressure Gauge for Type S ellipsoidal, or conical tip. Pitot Tube Calibration. An inclined manom- 2.7.2 A minimum of six diameters straight eter or equivalent is used. If the single-veloc- run (based upon D, the external diameter of ity calibration technique is employed (see the tube) between the tip and the static pres- Section 4.1.2.3), the calibration differential sure holes. pressure gauge shall be readable to the near- 2.7.3 A minimum of eight diameters est 0.13 mm H2O (0.005 in. H2O). For multi- velocity calibrations, the gauge shall be straight run between the static pressure readable to the nearest 0.13 mm H O (0.005 in. holes and the centerline of the external tube, 2 H O) for ∆p values between 1.3 and 25 mm H O following the 90 degree bend. 2 2 (0.05 and 1.0 in. H2O), and to the nearest 1.3 2.7.4 Static pressure holes of equal size mm H2O (0.05 in. H2O) for ∆p values above 25 (approximately 0.1 D), equally spaced in a pi- mm H2O (1.0 in. H2O). A special, more sen- ezometer ring configuration. sitive gauge will be required to read ∆p val- 2.7.5 Ninety degree bend, with curved or ues below 1.3 mm H2O [0.05 in. H2O] (see Cita- mitered junction. tion 18 in Bibliography).
590
VerDate 11
3. Procedure tube may be used to dampen ∆p fluctuations. 3.1 Set up the apparatus as shown in Fig- It is recommended, but not required, that a ure 2–1. Capillary tubing or surge tanks in- pretest leak-check be conducted, as follows: stalled between the manometer and pitot (1) blow through the pitot impact opening
591
VerDate 11
until at least 7.6 cm (3 in.) H2O velocity pres- 3.3 Measure the velocity head and tem- sure registers on the manometer; then, close perature at the traverse points specified by off the impact opening. The pressure shall Method 1. Ensure that the proper differential remain stable for at least 15 seconds; (2) do pressure gauge is being used for the range of the same for the static pressure side, except ∆p values encountered (see Section 2.2). If it using suction to obtain the minimum of 7.6 is necessary to change to a more sensitive cm (3 in.) H2O. Other leak-check procedures, gauge, do so, and remeasure the ∆p and tem- subject to the approval of the Administrator, perature readings at each traverse point. may be used. Conduct a post-test leak-check (mandatory), 3.2 Level and zero the manometer. Be- cause the manometer level and zero may as described in Section 3.1 above, to validate drift due to vibrations and temperature the traverse run. changes, make periodic checks during the 3.4 Measure the static pressure in the traverse. Record all necessary data as shown stack. One reading is usually adequate. in the example data sheet (Figure 2–5). 3.5 Determine the atmospheric pressure.
592
VerDate 11
3.6 Determine the stack gas dry molecular sentially air, an analysis need not be con- weight. For combustion processes or proc- ducted; use a dry molecular weight of 29.0. esses that emit essentially CO2, O2, CO, and For other processes, other methods, subject N2, use Method 3. For processes emitting es-
593
VerDate 11
to the approval of the Administrator, must Type S pitot tube is not always used; in be used. many instances, the pitot tube is used in 3.7 Obtain the moisture content from Ref- combination with other source-sampling erence Method 4 (or equivalent) or from components (thermocouple, sampling probe, Method 5. nozzle) as part of an ‘‘assembly.’’ The pres- 3.8 Determine the cross-sectional area of ence of other sampling components can the stack or duct at the sampling location. sometimes affect the baseline value of the Whenever possible, physically measure the Type S pitot tube coefficient (Citation 9 in stack dimensions rather than using blue- Bibliography); therefore an assigned (or oth- prints. erwise known) baseline coefficient value may 4. Calibration or may not be valid for a given assembly. 4.1 Type S Pitot Tube. Before its initial The baseline and assembly coefficient values use, carefully examine the Type S pitot tube will be identical only when the relative in top, side, and end views to verify that the placement of the components in the assem- face openings of the tube are aligned within bly is such that aerodynamic interference ef- the specifications illustrated in Figure 2–2 or fects are eliminated. Figures 2–6 through 2–8 2–3. The pitot tube shall not be used if it fails illustrate interference-free component ar- to meet these alignment specifications. rangements for Type S pitot tubes having ex- After verifying the face opening align- ternal tubing diameters between 0.48 and 0.95 ment, measure and record the following di- cm (3⁄16 and 3⁄8 in.). Type S pitot tube assem- mensions of the pitot tube: (a) the external blies that fail to meet any or all of the speci- tubing diameter (dimension Dt, Figure 2–2b); and (b) the base-to-opening plane distances fications of Figures 2–6 through 2–8 shall be calibrated according to the procedure out- (dimensions PA and PB, Figure 2–2b). If Dt is between 0.48 and 0.95 cm (3⁄16 and 3⁄8 in.) and lined in Sections 4.1.2 through 4.1.5 below, if PA and PB are equal and between 1.05 and and prior to calibration, the values of the 1.50 Dt, there are two possible options: (1) the intercomponent spacings (pitot-nozzle, pitot- pitot tube may be calibrated according to thermocouple, pitot-probe sheath) shall be the procedure outlined in Sections 4.1.2 measured and recorded. through 4.1.5 below, or (2) a baseline (isolated NOTE: Do not use any Type S pitot tube as- tube) coefficient value of 0.84 may be as- sembly which is constructed such that the signed to the pitot tube. Note, however, that impact pressure opening plane of the pitot if the pitot tube is part of an assembly, cali- tube is below the entry plane of the nozzle bration may still be required, despite knowl- (see Figure 2–6b). edge of the baseline coefficient value (see Section 4.1.1). 4.1.2 Calibration Setup. If the Type S If Dt, PA, and PB are outside the specified pitot tube is to be calibrated, one leg of the limits, the pitot tube must be calibrated as tube shall be permanently marked A, and the outlined in 4.1.2 through 4.1.5 below. other, B. Calibration shall be done in a flow 4.1.1 Type S Pitot Tube Assemblies. During system having the following essential design sample and velocity traverses, the isolated features:
594
VerDate 11
595
VerDate 11
4.1.2.1 The flowing gas stream must be 4.1.2.2 The cross-sectional area of the cali- confined to a duct of definite cross-sectional bration duct must be constant over a dis- area, either circular or rectangular. For cir- tance of 10 or more duct diameters. For a cular cross-sections, the minimum duct di- rectangular cross-section, use an equivalent ameter shall be 30.5 cm (12 in.); for rectangu- diameter, calculated from the following lar cross-sections, the width (shorter side) equation, to determine the number of duct shall be at least 25.4 cm (10 in.). diameters:
596
VerDate 11
be used without first referring to the special = 2LW considerations presented in Section 4.1.5. De Eq.- 21 Note also that this procedure applies only to (L+W) single-velocity calibration. To obtain cali- Where: bration data for the A and B sides of the De=Equivalent diameter Type S pitot tube, proceed as follows: L=Length 4.1.3.1 Make sure that the manometer is W=Width properly filled and that the oil is free from To ensure the presence of stable, fully de- contamination and is of the proper density. veloped flow patterns at the calibration site, Inspect and leak-check all pitot lines; repair or ‘‘test section,’’ the site must be located at or replace if necessary. least eight diameters downstream and two 4.1.3.2 Level and zero the manometer. diameters upstream from the nearest dis- Turn on the fan and allow the flow to sta- turbances. bilize. Seal the Type S entry port. NOTE: The eight- and two-diameter criteria 4.1.3.3 Ensure that the manometer is level are not absolute; other test section locations and zeroed. Position the standard pitot tube may be used (subject to approval of the Ad- at the calibration point (determined as out- ministrator), provided that the flow at the lined in Section 4.1.5.1), and align the tube so test site is stable and demonstrably parallel that its tip is pointed directly into the flow. to the duct axis. Particular care should be taken in aligning 4.1.2.3 The flow system shall have the ca- the tube to avoid yaw and pitch angles. pacity to generate a test-section velocity Make sure that the entry port surrounding around 915 m/min (3,000 ft/min). This velocity the tube is properly sealed.
must be constant with time to guarantee 4.1.3.4 Read ∆ pstd and record its value in a steady flow during calibration. Note that data table similar to the one shown in Fig- Type S pitot tube coefficients obtained by ure 2–9. Remove the standard pitot tube from single-velocity calibration at 915 m/min the duct and disconnect it from the manom- (3,000 ft/min) will generally be valid to with- eter. Seal the standard entry port. in ±3 percent for the measurement of veloci- 4.1.3.5 Connect the Type S pitot tube to ties above 305 m/min (1,000 ft/min) and to the manometer. Open the Type S entry port. within ±5 to 6 percent for the measurement Check the manometer level and zero. Insert of velocities between 180 and 305 m/min (600 and align the Type S pitot tube so that its A and 1,000 ft/min). If a more precise correla- side impact opening is at the same point as tion between Cp and velocity is desired, the was the standard pitot tube and is pointed flow system shall have the capacity to gen- directly into the flow. Make sure that the erate at least four distinct, time-invariant entry port surrounding the tube is properly test-section velocities covering the velocity sealed. range from 180 to 1,525 m/min (600 to 5,000 ft/ 4.1.3.6 Read ∆p8 and enter its value in the min), and calibration data shall be taken at data table. Remove the Type S pitot tube regular velocity intervals over this range from the duct and disconnect it from the ma- (see Citations 9 and 14 in Bibliography for de- nometer. tails). 4.1.2.4 Two entry ports, one each for the 4.1.3.7 Repeat steps 4.1.3.3 through 4.1.3.6 ∆ standard and Type S pitot tubes, shall be cut above until three pairs of p readings have in the test section; the standard pitot entry been obtained. port shall be located slightly downstream of 4.1.3.8 Repeat steps 4.1.3.3 through 4.1.3.7 the Type S port, so that the standard and above for the B side of the Type S pitot tube. Type S impact openings will lie in the same 4.1.3.9 Perform calculations, as described cross-sectional plane during calibration. To in Section 4.1.4 below. facilitate alignment of the pitot tubes dur- 4.1.4 Calculations. ing calibration, it is advisable that the test 4.1.4.1 For each of the six pairs of ∆p read- section be constructed of plexiglas or some ings (i.e., three from side A and three from other transparent material. side B) obtained in Section 4.1.3 above, cal- 4.1.3 Calibration Procedure. Note that culate the value of the Type S pitot tube co- this procedure is a general one and must not efficient as follows:
597
VerDate 11
Cp(std)=Standard pitot tube coefficient; use 0.99 if the coefficient is unknown and the tube is designed according to the criteria of Sections 2.7.1 to 2.7.5 of this method.
∆pstd=Velocity head measured by the stand- Where: ard pitot tube, cm H2O (in. H2O)
Cp(s)=Type S pitot tube coefficient 598
VerDate 11
∆ps=Velocity head measured by the Type S coefficients so obtained will be valid so long pitot tube, cm H2O (in. H2O) as the pitot tube-thermocouple combination is used by itself or with other components in 4.1.4.2 Calculate C¯ (side A), the mean A- p an interference-free arrangement (Figures 2– side coefficient, and C¯ (side B), the mean B- p 6 and 2–8). side coefficient: calculate the difference be- 4.1.5.1.3 For assemblies with sample tween these two average values. probes, the calibration point should be lo- 4.1.4.3 Calculate the deviation of each of ¯ cated at or near the center of the duct; how- the three A-side values of Cp(s) from Cp (side ever, insertion of a probe sheath into a small A), and the deviation of each B-side value of ¯ duct may cause significant cross-sectional Cp(s) from Cp (side B). Use the following equa- area blockage and yield incorrect coefficient tion: values (Citation 9 in Bibliography). There- fore, to minimize the blockage effect, the calibration point may be a few inches off- center if necessary. The actual blockage ef- fect will be negligible when the theoretical σ 4.1.4.4 Calculate , the average deviation blockage, as determined by a projected-area from the mean, for both the A and B sides of model of the probe sheath, is 2 percent or the pitot tube. Use the following equation: less of the duct cross-sectional area for as- semblies without external sheaths (Figure 2– 10a), and 3 percent or less for assemblies with external sheaths (Figure 2–10b). 4.1.5.2 For those probe assemblies in which pitot tube-nozzle interference is a fac- tor (i.e., those in which the pitot-nozzle sepa- ration distance fails to meet the specifica- 4.1.4.5 Use the Type S pitot tube only if tion illustrated in Figure 2–6a), the value of σ σ the values of (side A) and (side B) are less Cp(s) depends upon the amount of free-space than or equal to 0.01 and if the absolute between the tube and nozzle, and therefore is ¯ ¯ value of the difference between Cp (A) and Cp a function of nozzle size. In these instances, (B) is 0.01 or less. separate calibrations shall be performed 4.1.5 Special considerations. with each of the commonly used nozzle sizes 4.1.5.1 Selection of calibration point. in place. Note that the single-velocity cali- 4.1.5.1.1 When an isolated Type S pitot bration technique is acceptable for this pur- tube is calibrated, select a calibration point pose, even though the larger nozzle sizes at or near the center of the duct, and follow (>0.635 cm or 1⁄4 in.) are not ordinarily used the procedures outlined in Sections 4.1.3 and for isokinetic sampling at velocities around 4.1.4 above. The Type S pitot coefficients so 915 m/min (3,000 ft/min), which is the calibra- ¯ ¯ obtained, i.e., Cp (side A) and Cp (side B), will tion velocity; note also that it is not nec- be valid, so long as either: (1) the isolated essary to draw an isokinetic sample during pitot tube is used; or (2) the pitot tube is calibration (see Citation 19 in Section 6). used with other components (nozzle, thermo- 4.1.5.3 For a probe assembly constructed couple, sample probe) in an arrangement such that its pitot tube is always used in the that is free from aerodynamic interference same orientation, only one side of the pitot effects (see Figures 2–6 through 2–8). tube need be calibrated (the side which will 4.1.5.1.2 For Type S pitot tube-thermo- face the flow). The pitot tube must still meet couple combinations (without sample probe), the alignment specifications of Figure 2–2 or select a calibration point at or near the cen- 2–3, however, and must have an average devi- ter of the duct, and follow the procedures ation (σ) value of 0.01 or less (see Section outlined in Sections 4.1.3 and 4.1.4 above. The 4.1.4.4).
599
VerDate 11
4.1.6 Field Use and Recalibration. the appropriate coefficient value (whether 4.1.6.1 Field Use. assigned or obtained by calibration) shall be 4.1.6.1.1 When a Type S pitot tube (iso- used to perform velocity calculations. For lated tube or assembly) is used in the field,
600
VerDate 11
calibrated Type S pitot tubes, the A side co- metric fixed points, e.g., ice bath and boiling efficient shall be used when the A side of the water (corrected for barometric pressure) tube faces the flow, and the B side coefficient may be used. For temperatures above 405° C shall be used when the B side faces the flow; (761° F), use an NBS-calibrated reference alternatively, the arithmetic average of the thermocouple-potentiometer system or an A and B side coefficient values may be used, alternate reference, subject to the approval irrespective of which side faces the flow. of the Administrator. 4.1.6.1.2 When a probe assembly is used to If, during calibration, the absolute tem- sample a small duct (12 to 36 in. in diameter), peratures measured with the gauge being the probe sheath sometimes blocks a signifi- calibrated and the reference gauge agree cant part of the duct cross-section, causing a within 1.5 percent, the temperature data reduction in the effective value of Cp(s). Con- taken in the field shall be considered valid. sult Citation 9 in Bibliography for details. Otherwise, the pollutant emission test shall Conventional pitot-sampling probe assem- either be considered invalid or adjustments blies are not recommended for use in ducts (if appropriate) of the test results shall be having inside diameters smaller than 12 made, subject to the approval of the Admin- inches (Citation 16 in Bibliography). istrator. 4.1.6.2 Recalibration. 4.4 Barometer. Calibrate the barometer 4.1.6.2.1 Isolated Pitot Tubes. After each used against a mercury barometer. field use, the pitot tube shall be carefully re- 5. Calculations examined in top, side, and end views. If the pitot face openings are still aligned within Carry out calculations, retaining at least the specifications illustrated in Figure 2–2 or one extra decimal figure beyond that of the 2–3, it can be assumed that the baseline coef- acquired data. Round off figures after final ficient of the pitot tube has not changed. If, calculation. however, the tube has been damaged to the 5.1 Nomenclature. extent that it no longer meets the specifica- A=Cross-sectional area of stack, m2(ft 2). tions of Figure 2–2 or 2–3, the damage shall Bws=Water vapor in the gas stream (from either be repaired to restore proper align- Method 5 or Reference Method 4), propor- ment of the face openings or the tube shall tion by volume. be discarded. Cp=Pitot tube coefficient, dimensionless. 4.1.6.2.2 Pitot Tube Assemblies. After each Kp=Pitot tube constant, field use, check the face opening alignment of the pitot tube, as in Section 4.1.6.2.1; also, remeasure the intercomponent spacings of the assembly. If the intercomponent for the metric system and spacings have not changed and the face open- ing alignment is acceptable, it can be as- sumed that the coefficient of the assembly has not changed. If the face opening align- ment is no longer within the specifications for the English system. of Figures 2–2 or 2–3, either repair the dam- Md=Molecular weight of stack gas, dry basis age or replace the pitot tube (calibrating the (see Section 3.6) g/g-mole (lb/lb-mole). new assembly, if necessary). If the inter- Ms=Molecular weight of stack gas, wet basis, component spacings have changed, restore g/g-mole (lb/lb-mole).
the original spacings or recalibrate the as- =Md (1¥Bws) +18.0 Bws sembly. Eq. 2–5 4.2 Standard pitot tube (if applicable). If a standard pitot tube is used for the velocity Pbar=Barometric pressure at measurement traverse, the tube shall be constructed ac- site, mm Hg (in. Hg). cording to the criteria of Section 2.7 and Pg=Stack static pressure, mm Hg (in. Hg). shall be assigned a baseline coefficient value Ps=Absolute stack gas pressure, mm Hg (in. of 0.99. If the standard pitot tube is used as Hg). part of an assembly, the tube shall be in an =Pbar+Pg interference-free arrangement (subject to Eq. 2–6 the approval of the Administrator). Pstd=Standard absolute pressure, 760 mm Hg 4.3 Temperature Gauges. After each field (29.92 in. Hg). use, calibrate dial thermometers, liquid- Qsd=Dry volumetric stack gas flow rate cor- filled bulb thermometers, thermocouple-po- rected to standard conditions, dscm/hr tentiometer systems, and other gauges at a (dscf/hr). temperature within 10 percent of the average ts=Stack temperature, °C (°F). absolute stack temperature. For tempera- Ts=Absolute stack temperature, °K, (°R). tures up to 405° C (761° F), use an ASTM mer- =273+ts for metric. cury-in-glass reference thermometer, or Eq. 2–7 equivalent, as a reference; alternatively, ei- ther a reference thermocouple and poten- =460+ts for English. tiometer (calibrated by NBS) or thermo- Eq. 2–8
601
VerDate 11
Tstd=Standard absolute temperature, 293 °K 5.2 Average Stack Gas Velocity. (528° R).
vs=Average stack gas velocity, m/sec (ft/sec). ∆p=Velocity head of stack gas, mm H2O (in. H2O). 3,600=Conversion factor, sec/hr. 18.0=Molecular weight of water, g/g-mole (lb/ 5.3 Average Stack Gas Dry Volumetric lb-mole). Flow Rate.
= − Tstd Ps Qsd 3, 600() 1 Bws v s A Eq. 210 - Ts(avg) Pstd
To convert Qsd from dscm/hr (dscf/hr) to mental Protection Agency, Emission Meas- dscm/min (dscf/min), divide Qsd by 60. urement Branch, Research Triangle Park, 6. Bibliography NC. November 1976. 13. Vollaro, R. F. An Evaluation of Single- 1. Mark, L. S. Mechanical Engineers’ Hand- book. New York, McGraw-Hill Book Co., Inc. Velocity Calibration Technique as a Means 1951. of Determining Type S Pitot Tubes Coeffi- 2. Perry, J. H. Chemical Engineers’ Hand- cient. U.S. Environmental Protection Agen- book. New York. McGraw-Hill Book Co., Inc. cy, Emission Measurement Branch, Research 1960. Triangle Park, NC. August 1975. 3. Shigehara, R. T., W. F. Todd, and W. S. 14. Vollaro, R. F. The Use of Type S Pitot Smith. Significance of Errors in Stack Sam- Tubes for the Measurement of Low Veloci- pling Measurements. U.S. Environmental ties. U.S. Environmental Protection Agency, Protection Agency, Research Triangle Park, Emission Measurement Branch, Research NC (Presented at the Annual Meeting of the Triangle Park, NC. November 1976. Air Pollution Control Association, St. Louis, 15. Smith, Marvin L. Velocity Calibration MO, June 14–19, 1970.) of EPA Type Source Sampling Probe. United 4. Standard Method for Sampling Stacks Technologies Corporation, Pratt and Whit- for Particulate Matter. In: 1971 Book of ney Aircraft Division, East Hartford, CN. ASTM Standards, Part 23. Philadelphia, PA 1975. 1971. ASTM Designation D–2928–71. 5. Vennard, J. K. Elementary Fluid Me- 16. Vollaro, R. F. Recommended Procedure chanics. New York. John Wiley and Sons, for Sample Traverses in Ducts Smaller than Inc. 1947. 12 Inches in Diameter. U.S. Environmental 6. Fluid Meters—Their Theory and Applica- Protection Agency, Emission Measurement tion. American Society of Mechanical Engi- Branch, Research Triangle Park, NC. Novem- neers, New York, NY 1959. ber 1976. 7. ASHRAE Handbook of Fundamentals. 17. Ower, E. and R. C. Pankhurst. The 1972. p. 208. Measurement of Air Flow, 4th Ed., London, 8. Annual Book of ASTM Standards, Part Pergamon Press. 1966. 26. 1974. p. 648. 18. Vollaro, R. F. A Survey of Commer- 9. Vollaro, R. F. Guidelines for Type S cially Available Instrumentation for the Pitot Tube Calibration. U.S. Environmental Measurement of Low-Range Gas Velocities. Protection Agency. Research Triangle Park, U.S. Environmental Protection Agency, NC (Presented at 1st Annual Meeting, Source Emission Measurement Branch, Research Evaluation Society, Dayton, OH, September Triangle Park, NC. November 1976. (Unpub- 18, 1975.) 10. Vollaro, R. F. A Type S Pitot Tube Cali- lished Paper) bration Study. U.S. Environmental Protec- 19. Gnyp, A. W., C. C. St. Pierre, D. S. tion Agency, Emission Measurement Branch, Smith, D. Mozzon, and J. Steiner. An Experi- Research Triangle Park, NC July 1974. mental Investigation of the Effect of Pitot 11. Vollaro, R. F. The Effects of Impact Tube-Sampling Probe Configurations on the Opening Misalignment on the Value of the Magnitude of the S Type Pitot Tube Coeffi- Type S Pitot Tube Coefficient. U.S. Environ- cient for Commercially Available Source mental Protection Agency, Emission Meas- Sampling Probes. Prepared by the University urement Branch, Research Triangle Park, of Windsor for the Ministry of the Environ- NC. October 1976. ment, Toronto, Canada. February 1975. 12. Vollaro, R. F. Establishment of a Basline Coefficient Value for Properly Con- structed Type S Pitot Tubes. U.S. Environ-
602
VerDate 11
METHOD 2A—DIRECT MEASUREMENT OF meter elevation increase, or vice-versa for GAS VOLUME THROUGH PIPES AND elevation decrease. SMALL DUCTS 2.3 Stopwatch. Capable of measurement to within 1 second. 1. Applicability and Principle 3. Procedure 1.1 Applicability. This method applies to 3.1 Installation. As there are numerous the measurement of gas flow rates in pipes types of pipes and small ducts that may be and small ducts, either in-line or at exhaust subject to volume measurement, it would be positions, within the temperature range of 0 difficult to describe all possible installation to 50°C. schemes. In general, flange fittings should be 1.2 Principle. A gas volume meter is used used for all connections wherever possible. to measure gas volume directly. Tempera- Gaskets or other seal materials should be ture and pressure measurements are made to used to assure leak-tight connections. The correct the volume to standard conditions. volume meter should be located so as to 2. Apparatus avoid severe vibrations and other factors Specifications for the apparatus are given that may affect the meter calibration. below. Any other apparatus that has been 3.2 Leak Test. A volume meter installed demonstrated (subject to approval of the Ad- at a location under positive pressure may be ministrator) to be capable of meeting the leak-checked at the meter connections by specifications will be considered acceptable. using a liquid leak detector solution con- 2.1 Gas Volume Meter. A positive dis- taining a surfactant. Apply a small amount placement meter, turbine meter, or other di- of the solution to the connections. If a leak rect volume measuring device capable of exists, bubbles will form, and the leak must measuring volume to within 2 percent. The be corrected. meter shall be equipped with a temperature A volume meter installed at a location gauge (±2 percent of the minimum absolute under negative pressure is very difficult to temperature) and a pressure gauge (±2.5 mm test for leaks without blocking flow at the Hg). The manufacturer’s recommended ca- inlet of the line and watching for meter pacity of the meter shall be sufficient for the movement. If this procedure is not possible, expected maximum and minimum flow rates visually check all connections and assure at the sampling conditions. Temperature, tight seals. pressure, corrosive characteristics, and pipe 3.3 Volume Measurement. size are factors necessary to consider in 3.3.1 For sources with continuous, steady choosing a suitable gas meter. emission flow rates, record the initial meter 2.2 Barometer. A mercury, aneroid, or volume reading, meter temperature(s), meter other barometer capable of measuring at- pressure, and start the stopwatch. Through- mospheric pressure to within 2.5 mm Hg. In out the test period, record the meter tem- many cases, the barometric reading may be perature(s) and pressure so that average val- obtained from a nearby National Weather ues can be determined. At the end of the Service station, in which case the station test, stop the timer and record the elapsed value (which is the absolute barometric pres- time, the final volume reading, meter tem- sure) shall be requested, and an adjustment perature(s), and pressure. Record the baro- for elevation differences between the weath- metric pressure at the beginning and end of er station and the sampling point shall be the test run. Record the data on a table simi- applied at a rate of minus 2.5 mm Hg per 30- lar to Figure 2A–1.
603
VerDate 11
3.3.2 For sources with noncontinuous, the start and stop times corresponding to non-steady emission flow rates, use the pro- each process cyclical or noncontinuous cedure in 3.3.1 with the addition of the fol- event. lowing: Record all the meter parameters and 4. Calibration
604
VerDate 11
4.1 Volume Meter. The volume meter is (i.e., in relation to the flow moving device). calibrated against a standard reference Connect the temperature and pressure meter prior to its initial use in the field. The gauges as they are to be used in the field. reference meter is a spirometer or liquid dis- Connect the reference meter at the inlet of placement meter with a capacity consistent the flow line, if appropriate for the meter, with that of the test meter. and begin gas flow through the system to Alternatively, a calibrated, standard pitot condition the meters. During this condi- may be used as the reference meter in con- tioning operation, check the system for junction with a wind tunnel assembly. At- leaks. tach the test meter to the wind tunnel so The calibration shall be run over at least that the total flow passes through the test three different flow rates. The calibration meter. For each calibration run, conduct a 4- flow rates shall be about 0.3, 0.6, and 0.9 point traverse along one stack diameter at a times the test meter’s rated maximum flow position at least eight diameters of straight rate. tunnel downstream and two diameters up- For each calibration run, the data to be stream of any bend, inlet, or air mover. De- collected include: reference meter initial and termine the traverse point locations as spec- final volume readings, the test meter initial ified in Method 1. Calculate the reference and final volume reading, meter average volume using the velocity values following temperature and pressure, barometric pres- the procedure in Method 2, the wind tunnel sure, and run time. Repeat the runs at each cross-sectional area, and the run time. flow rate at least three times. Set up the test meter in a configuration Calculate the test meter calibration coeffi- similar to that used in the field installation cient, Ym, for each run as follows:
()()V− V t +273 P Y = rf ri r b Eq. 2 A - 1 m ()()− Vmf V mi t m +273 ()PPb+ g
Where: meter pressure set at the average value en- countered in the field test. Calculate the av- Ym=Test volume meter calibration coeffi- cient, dimensionless. erage value of the calibration factor. If the 3 calibration has changed by more than 5 per- Vr=Reference meter volume reading, m . 3 Vm=Test meter volume reading, m . cent, recalibrate the meter over the full tr=Reference meter average temperature, ° C. range of flow as described above. tm=Test meter average temperature, ° C. NOTE. If the volume meter calibration coef- Pb=Barometric pressure, mm Hg. ficient values obtained before and after a Pg=Test meter average static pressure, mm test series differ by more than 5 percent, the Hg. test series shall either be voided, or calcula- f=Final reading for run. tions for the test series shall be performed i=Initial reading for run. using whichever meter coefficient value (i.e., Compare the three Ym values at each of the before or after) gives the greater value of flow rates tested and determine the maxi- pollutant emission rate. mum and minimum values. The difference 4.2 Temperature Gauge. After each test between the maximum and minimum values series, check the temperature gauge at ambi- at each flow rate should be no greater than ent temperature. Use an American Society 0.030. Extra runs may be required to com- for Testing and Materials (ASTM) mercury- plete this requirement. If this specification in-glass reference thermometer, or equiva- cannot be met in six successive runs, the test lent, as a reference. If the gauge being meter is not suitable for use. In addition, the checked agrees within 2 percent (absolute meter coefficients should be between 0.95 and temperature) of the reference, the tempera- 1.05. If these specifications are met at all the ture data collected in the field shall be con- flow rates, average all the Y values from m sidered valid. Otherwise, the test data shall runs meeting the specifications to obtain an be considered invalid or adjustments of the average meter calibration coefficient, Y . m test results shall be made, subject to the ap- The procedure above shall be performed at proval of the Administrator. least once for each volume meter. There- 4.3 Barometer. Calibrate the barometer after, an abbreviated calibration check shall used against a mercury barometer prior to be completed following each field test. The the field test. calibration of the volume meter shall be checked by performing three calibration 5. Calculations runs at a single, intermediate flow rate Carry out the calculations, retaining at (based on the previous field test) with the least one extra decimal figure beyond that of
605
VerDate 11
3 the acquired data. Round off figures after the Vm=Meter volume reading, m . final calculation. Ym=Average meter calibration coefficient, 5.1 Nomenclature. dimensionless. Pb=Barometric pressure, mm Hg. f=Final reading for test period. Pg=Average static pressure in volume meter, i=Initial reading for test period. mm Hg. s=Standard conditions, 20° C and 760 mm Hg. Q =Gas flow rate, m3/min, standard condi- s Θ tions. =Elapsed test period time, min. 5.2 Volume. Tm=Average absolute meter temperature, °K.
()PP+ = − b g VYVVms0. 3853 m () mf mi Eq. 2 A - 2 Tm
5.3 Gas Flow Rate. 2.2 Organic Analyzers (2). Equipment de- scribed in Method 25A or 25B. 2.3 CO Analyzer. Equipment described in Vms Q = Eq. 2 A - 3 Method 10. s θ 2.4 CO2 Analyzer. A nondispersive infrared 6. Bibliography (NDIR) CO2 analyzer and supporting equip- ment with comparable specifications as CO 1. Rom, Jerome J. Maintenance, Calibra- analyzer described in Method 10. tion, and Operation of Isokinetic Source Sampling Equipment. U.S. Environmental 3. Procedure Protection Agency. Research Triangle Park, 3.1 Inlet Installation. Install a volume NC, Publication No. APTD–0576. March 1972. meter in the vapor line to incinerator inlet 2. Wortman, Martin, R. Vollaro, and P.R. according to the procedure in Method 2A. At Westlin. Dry Gas Volume Meter Calibra- the volume meter inlet, install a sample tions. Source Evaluation Society Newsletter. probe as described in Method 25A. Connect to Vol. 2, No. 2. May 1977. the probe a leak-tight, heated (if necessary 3. Westlin, P.R. and R.T. Shigehara. Pro- to prevent condensation) sample line (stain- cedure for Calibrating and Using Dry Gas less steel or equivalent) and an organic ana- Volume Meters as Calibration Standards. lyzer system as described in Method 25A or Source Evaluation Society Newsletter. Vol. 25B. 3, No. 1. February 1978. 3.2 Exhaust Installation. Three sample analyzers are required for the incinerator ex- METHOD 2B—DETERMINATION OF EXHAUST GAS haust: CO2, CO, and organic analyzers. A VOLUME FLOW RATE FROM GASOLINE VAPOR sample manifold with a single sample probe INCINERATORS may be used. Install a sample probe as de- scribed in Method 25A. Connect a leak-tight 1. Applicability and Principle heated sample line to the sample probe. Heat 1.1 Applicability. This method applies to the sample line sufficiently to prevent any the measurement of exhaust volume flow condensation. rate from incinerators that process gasoline 3.3 Recording Requirements. The output vapors consisting primarily of alkanes, of each analyzer must be permanently re- alkenes, and/or arenes (aromatic hydro- corded on an analog strip chart, digital re- carbons). It is assumed that the amount of corder, or other recording device. The chart auxiliary fuel is negligible. speed or number of readings per time unit 1.2 Principle. The incinerator exhaust must be similar for all analyzers so that data flow rate is determined by carbon balance. can be correlated. The minimum data re- Organic carbon concentration and volume cording requirement for each analyzer is one flow rate are measured at the incinerator measurement value per minute. 3.4 Preparation. Prepare and calibrate all inlet. Organic carbon, carbon dioxide (CO2), and carbon monoxide (CO) concentrations equipment and analyzers according to the are measured at the outlet. Then the ratio of procedures in the respective methods. For total carbon at the incinerator inlet and out- the CO2 analyzer, follow the procedures de- let is multiplied by the inlet volume to de- scribed in Method 10 for CO analysis sub- termine the exhaust volume and volume flow stituting CO2 calibration gas where the rate. method calls for CO calibration gas. The span value for the CO2 analyzer shall be 15 2. Apparatus percent by volume. All calibration gases 2.1 Volume Meter. Equipment described in must be introduced at the connection be- Method 2A. tween the probe and the sample line. If a
606
VerDate 11
manifold system is used for the exhaust ana- 4.1 Nomenclature.
lyzers, all the analyzers and sample pumps COe=Mean carbon monoxide concentration in must be operating when the calibrations are system exhaust, ppm.
done. Note: For the purposes of this test, CO2e=Mean carbon dioxide concentration in methane should not be used as an organic system exhaust, ppm. calibration gas. HCe=Mean organic concentration in system 3.5 Sampling. At the beginning of the test exhaust as defined by the calibration gas, period, record the initial parameters for the ppm. inlet volume meter according to the proce- HCi=Mean organic concentration in system dures in Method 2A and mark all of the re- inlet as defined by the calibration gas, corder strip charts to indicate the start of ppm. the test. Continue recording inlet organic K=Calibration gas factor and exhaust CO , CO, and organic concentra- 2 =2 for ethane calibration gas. tions throughout the test. During periods of process interruption and halting of gas flow, =3 for propane calibration gas. stop the timer and mark the recorder strip =4 for butane calibration gas. charts so that data from this interruption =Appropriate response factor for other are not included in the calculations. At the calibration gas. 3 end of the test period, record the final pa- Ves=Exhaust gas volume, m . 3 rameters for the inlet volume meter and Vis=Inlet gas volume, m . 3 mark the end on all of the recorder strip Qes=Exhaust gas volume flow rate, m /min. 3 charts. Qis=Inlet gas volume flow rate, m /min. 3.6 Post Test Calibrations. At the conclu- Θ=Sample run time, min. sion of the sampling period, introduce the s=Standard conditions: 20°C, 760 mm Hg. calibration gases as specified in the respec- 300=Estimated concentration of ambient CO2, tive reference methods. If an analyzer output ppm. (CO2 concentration in the ambient does not meet the specifications of the meth- air may be measured during the test pe- od, invalidate the test data for the period. riod using an NDIR). Alternatively, calculate the volume results 4.2 Concentrations. Determine mean con- using initial calibration data and using final centrations of inlet organics, outlet CO2, calibration data and report both resulting outlet CO, and outlet organics according to volumes. Then, for emissions calculations, the procedures in the respective methods and use the volume measurement resulting in the analyzers’ calibration curves, and for the the greatest emission rate or concentration. time intervals specified in the applicable 4. Calculations regulations. Concentrations should be deter- Carry out the calculations, retaining at mined on a parts per million by volume least one extra decimal figure beyond that of (ppm) basis. the acquired data. Round off figures after the 4.3 Exhaust Gas Volume. Calculate the final calculation. exhaust gas volume as follows:
K() HC VV= i Eq. 2 B - 1 es is + + − K() HCe CO2 e CO e 300
4.4 Exhaust Gas Volume Flow Rate. Cal- METHOD 2C—DETERMINATION OF STACK GAS culate the exhaust gas volume flow rate as VELOCITY AND VOLUMETRIC FLOW RATE IN follows: SMALL STACKS OR DUCTS (STANDARD PITOT
Qes=Ves/θ TUBE) Eq. 2B–2 1. Applicability and Principle 5. Bibliography 1.1 Applicability. 1. Measurement of Volatile Organic Com- 1.1.1 The applicability of this method is pounds. U.S. Environmental Protection identical to Method 2, except this method is Agency. Office of Air Quality Planning and limited to stationery source stacks or ducts Standards, Research Triangle Park, NC less than about 0.30 meter (12 in.) in diame- 27711. Publication No. EPA–450/2–78–041. Octo- ter or 0.071 m2 (113 in.2) in cross-sectional ber 1978. 55 p. area, but equal to or greater than about 0.10 meter (4 in.) in diameter or 0.0081 m2 (12.57 in.2) in cross-sectional area. 1.1.2 The apparatus, procedure, calibra- tion, calculations, and biliography are the
607
VerDate 11
same as in Method 2, Sections 2, 3, 4, 5, and a coefficient value of 0.99 unless it is cali- 6, except as noted in the following sections. brated against another standard pitot tube 1.2 Principle. The average gas velocity in with an NBS-traceable coefficient. a stack or duct is determined from the gas 2.2 Alternative Pitot Tube. A modified density and from measurement of velocity hemispherical-nosed pitot tube (see Figure heads with a standard pitot tube. 2C–1), which features a shortened stem and enlarged impact and static pressure holes, 2. Apparatus may be used. This pitot tube is useful in liq- uid drop-laden gas streams when a pitot 2.1 Standard Pitot Tube (instead of Type ‘‘back purge’’ is ineffective. Use a coefficient S). Use a standard pitot tube that meets the value of 0.99 unless the pitot is calibrated as specifications of Section 2.7 of Method 2. Use mentioned in Section 2.1 above.
608
VerDate 11
3. Procedure Method 1A. The static and impact pressure holes of standard pitot tubes are susceptible Follow the general procedures in Section 3 to plugging in PM–laden gas streams. There- of Method 2, except conduct the measure- fore, the tester must furnish adequate proof ments at the traverse points specified in
609
VerDate 11
that the openings of the pitot tube have not 2.3 Stopwatch. Capable of incremental plugged during the traverse period; this time measurement to within 1 second. proof can be obained by first recording the velocity head (∆p) reading at the final tra- 3. Procedure verse point, then cleaning out the impact 3.1 Installation. Use the procedure in and static holes of the standard pitot tube by Method 2A, Section 3.1. ‘‘back-purging’’ with pressurized air, and fi- 3.2 Leak Check. Use the procedure in nally by recording another ∆p reading at the Method 2A, Section 3.2. final traverse point. If the ∆p reading made 3.3 Flow Rate Measurement. after the air purge is within 5 percent of the 3.3.1 Continuous, Steady Flow. At least reading during the traverse, then the tra- once an hour, record the measuring device verse is acceptable. Otherwise, reject the flow rate reading, and the measuring device ∆ run. Note that if the p at the final traverse temperature and pressure. Make a minimum point is so low as to make this determina- of twelve equally spaced readings of each pa- tion too difficult, then another traverse rameter during the test period. Record the point may be selected. If ‘‘back purging’’ at barometric pressure at the beginning and regular intervals is part of the procedure, end of the test period. Record the data on a ∆ then take comparative p readings, as above, table similar to Figure 2D–1. for the last two back purges at which suit- ———————————————————————— ∆ able high p readings are observed. ———————————————————————— lllllllllllllll METHOD 2D—MEASUREMENT OF GAS VOLU- Plant lllllll llll METRIC FLOW RATES IN SMALL PIPES AND Date Run number lllllllllll DUCTS Sample location Barometric pressure, mm (in.) Hg 1. Applicability and Principle Startlll Finishlll Operators llllllllll 1.1 Applicability. This method applies to Measuring device numberlll Calibration the measurement of gas flow rates in small coefficientlll pipes and ducts, either before or after emis- Calibration gaslllll sion control devices. Last date calibratedllll 1.2 Principle. To measure flow rate or pressure drop, all the stack gas is directed Static Temperature through a rotameter, orifice plate or similar Time Flow rate pressure flow rate measuring device. The measuring reading mm (in.) °C (°F) °K (°R) device has been previously calibrated in a Hg manner that insures its proper calibration for the gas or gas mixture being measured. Absolute temperature and pressure measure- ments are also made to calculate volumetric flow rates at standard conditions. 2. Apparatus Specifications for the apparatus are given below. Any other apparatus that has been demonstrated (subject to approval of the Ad- ministrator) to be capable of meeting the specifications will be considered acceptable. 2.1 Flow Rate Measuring Device. A rotam- eter, orifice plate, or other flow rate measur- ing device capable of measuring all the stack flow rate to within 5 percent of its true Average. value. The measuring device shall be equipped with a temperature gauge accurate to within 2 percent of the minimum absolute stack temperature and a pressure gauge ac- Figure 2D–1. Flow rate measurement data. curate to within 5 mm Hg. The capacity of 3.3.2 Noncontinuous and Nonsteady the measuring device shall be sufficient for Flows. Use flow rate measuring devices with the expected maximum and minimum flow particular caution. Calibration will be af- rates at the stack gas conditions. The mag- fected by variation in stack gas temperature, nitude and variability of stack gas flow rate, pressure, compressibility, and molecular molecular weight, temperature, pressure, weight. Use the procedure in Section 3.3.1. compressibility, dew point, corrosiveness, Record all the measuring device parameters and pipe or duct size are all factors to con- on a time interval frequency sufficient to sider in choosing a suitable measuring de- adequately profile each process cyclical or vice. 2.2 Barometer. Same as in Method 2, Sec- noncontinuous event. A multichannel con- tion 2.5. tinuous recorder may be used.
610
VerDate 11
4. Calibration ments. John Wiley and Sons, Inc. New York, NY. 1969. 4.1 Flow Rate Measuring Device. Use the 3. Orifice Metering of Natural Gas. Amer- procedure in Method 2A, Section 4, and apply ican Gas Association. Arlington, VA. Report the same performance standards. Calibrate No. 3. March 1978. 88 p. the measuring device with the principal stack gas to be measured (e.g., air, nitrogen) METHOD 2E—DETERMINATION OF LANDFILL against a standard reference meter. A cali- GAS; GAS PRODUCTION FLOW RATE brated dry gas meter is an acceptable ref- erence meter. Ideally, calibrate the measur- 1. Applicability and Principle ing device in the field with the actual gas to 1.1 Applicability. This method applies to be measured. For measuring devices that the measurement of landfill gas (LFG) pro- have a volume rate readout, calculate the duction flow rate from municipal solid waste measuring device calibration coefficient, Y , m (MSW) landfills and is used to calculate the for each run as follows: flow rate of nonmethane organic compounds (NMOC) from landfills. This method also ap- ()()QTP = r r bar plies to calculating a site-specific k value as Ym Eq. 2 D - 1 provided in § 60.754(a)(4). It is unlikely that a ()()QTPPm m() bar+ g site-specific k value obtained through Meth- od 2E testing will lower the annual emission where: estimate below 50 Mg/yr NMOC unless the 3 Qr=reference meter flow rate reading, m / Tier 2 emission estimate is only slightly min (ft3/min). higher than 50 Mg/yr NMOC. Dry, arid re- 3 Qm=measuring device flow rate reading, m / gions may show a more significant difference min (ft3/min). between the default and calculated k values Tr=reference meter average absolute tem- than wet regions. perature, ° K (° R). 1.2 Principle. Extraction wells are in- Tm=measuring device average absolute stalled either in a cluster of three or at five temperature, ° K (° R). locations dispersed throughout the landfill. Pbar=barometric pressure, mm Hg (in. Hg). A blower is used to extract LFG from the Pg=measuring device average static pres- landfill. LFG composition, landfill pressures sure, mm Hg (in. Hg). near the extraction well, and volumetric For measuring devices that do not have a flow rate of LFG extracted from the wells readout as flow rate, refer to the manufac- are measured and the landfill gas production flow rate is calculated. turer’s instructions to calculate the Qm cor- responding to each Qr. 2. Apparatus 4.2 Temperature Gauge. Use the procedure and specifications in Method 2A, Section 4.2. 2.1 Well Drilling Rig. Capable of boring a Perform the calibration at a temperature 0.6 meters diameter hole into the landfill to that approximates field test conditions. a minimum of 75 percent of the landfill 4.3 Barometer. Calibrate the barometer to depth. The depth of the well shall not exceed be used in the field test with a mercury ba- the bottom of the landfill or the liquid level. rometer prior to the field test. 2.2 Gravel. No fines. Gravel diameter should be appreciably larger than perfora- 5. Gas Flow Rate Calculation tions stated in sections 2.10 and 3.2 of this method. Calculate the stack gas flow rate, Q , as s 2.3 Bentonite. follows: 2.4 Backfill Material. Clay, soil, and sandy loam have been found to be accept- + ()PPbar g able. = 2.5 Extraction Well Pipe. Polyvinyl chlo- QKYQs l m m Eq.-2 D 2 ride (PVC), high density polyethylene Tm (HDPE), fiberglass, stainless steel, or other where: suitable nonporous material capable of Kl = 0.3858 for international system of units transporting landfill gas with a minimum di- (SI); 17.64 for English units. ameter of 0.075 meters and suitable wall- thickness. 6. Bibliography 2.6 Wellhead Assembly. Valve capable of 1. Spink, L.K. Principles and Practice of adjusting gas flow at the wellhead and out- Flowmeter Engineering. The Foxboro Com- let, and a flow measuring device, such as an pany. Foxboro, MA. 1967. in-line orifice meter or pitot tube. A sche- 2. Benedict, Robert P. Fundamentals of matic of the wellhead assembly is shown in Temperature, Pressure, and Flow Measure- figure 1.
611
VerDate 11
2.7 Cap. PVC, HDPE, fiberglass, stainless 2.15 Differential Pressure Gauge. Water- steel, or other suitable nonporous material filled U-tube manometer or equivalent, capa- capable of transporting landfill gas with a ble of measuring within 0.02 mm Hg, for suitable wall-thickness. measuring the pressure of the pressure 2.8 Header Piping. PVC, HDPE, fiberglass, probes. stainless steel, or other suitable nonporous 3. Procedure material capable of transporting landfill gas with a suitable wall-thickness. 3.1 Placement of Extraction Wells. The 2.9 Auger. Capable of boring a 0.15 to 0.23 landfill owner or operator shall either install meters diameter hole to a depth equal to the a single cluster of three extraction wells in a top of the perforated section of the extrac- test area or space five wells over the landfill. tion well, for pressure probe installation. The cluster wells are recommended but may 2.10 Pressure Probe. PVC or stainless be used only if the composition, age of the steel (316), 0.025 meters. Schedule 40 pipe. solid waste, and the landfill depth of the test Perforate the bottom two thirds. A mini- area can be determined. CAUTION: Since mum requirement for perforations is slots or this method is complex, only experienced holes with an open area equivalent to four personnel should conduct the test. Landfill gas contains methane, therefore explosive 6.0 millimeter diameter holes spaced 90° mixtures may exist at or near the landfill. It apart every 0.15 meters. is advisable to take appropriate safety pre- 2.11 Blower and Flare Assembly. A water cautions when testing landfills, such as in- knockout, flare or incinerator, and an explo- stalling explosion-proof equipment and re- sion-proof blower, capable of extracting LFG fraining from smoking. at a flow rate of at least 8.5 cubic meters per 3.1.1 Cluster Wells. Consult landfill site minute. records for the age of the solid waste, depth, 2.12 Standard Pitot Tube and Differential and composition of various sections of the Pressure Gauge for Flow Rate Calibration landfill. Select an area near the perimeter of with Standard Pitot. Same as Method 2, sec- the landfill with a depth equal to or greater tions 2.1 and 2.8. than the average depth of the landfill and 2.13 Gas flow measuring device. Perma- with the average age of the solid waste be- nently mounted Type S pitot tube or an ori- tween 2 and 10 years old. Avoid areas known fice meter. to contain nondecomposable materials, such 2.14 Barometer. Same as Method 4, sec- as concrete and asbestos. Locate wells as tion 2.1.5. shown in figure 2.
612
VerDate 11
Because the age of the solid waste in a test where, area will not be uniform, calculate a weight- Aavg=average age of the solid waste tested, ed average to determine the average age of year the solid waste as follows. th fi=fraction of the solid waste in the i sec- n tion th = Ai=age of the i fraction, year Aavg∑ f i Ai i=1
3.1.2 Equal Volume Wells. This procedure traction well pipe. Perforations shall not be is used when the composition, age of solid closer than 6 meters from the cover. Perfora- waste, and landfill depth are not well known. tions shall be holes or slots with an open Divide the portion of the landfill that has area equivalent to 1.0 centimeter diameter had waste for at least 2 years into five areas holes spaced 90 degrees apart every 0.1 to 0.2 representing equal volumes. Locate an ex- meters. Place the extraction well in the cen- traction well near the center of each area. ter of the hole and backfill with 2.0 to 7.5 Avoid areas known to contain centimeters gravel to a level 0.3 meters nondecomposable materials, such as con- above the perforated section. Add a layer of crete and asbestos. backfill material 1.2 meters thick. Add a 3.2 Installation of Extraction Wells. Use a layer of bentonite 1.0 meter thick, and back- well drilling rig to dig a 0.6 meters diameter fill the remainder of the hole with cover ma- hole in the landfill to a minimum of 75 per- terial or material equal in permeability to cent of the landfill depth, not to exceed the the existing cover material. The specifica- bottom of the landfill or the water table. tions for extraction well installation are Perforate the bottom two thirds of the ex- shown in figure 3.
613
VerDate 11
3.3 Pressure Probes. Shallow pressure influence. Locate the deep pressure probes probes are used in the check for infiltration along three radial arms approximately 120 of air into the landfill, and deep pressure degrees apart at distances of 3, 15, 30, and 45 probes are used to determine the radius of meters from the extraction well. The tester
614
VerDate 11
has the option of locating additional pres- probes), shall extend to a depth equal to the sure probes at distances every 15 meters be- top of the perforated section of the extrac- yond 45 meters. Example placements of tion wells. Locate three shallow probes at a probes are shown in figure 4. distance of 3 m from the extraction well. The probes located 15, 30, and 45 meters Shallow probes shall extend to a depth equal from each well, and any additional probes lo- to half the depth of the deep probes. cated along the three radial arms (deep
615
VerDate 11
Use an auger to dig a hole, approximately holes or slots with an open area equivalent 0.15 to 0.23 meters in diameter, for each pres- to four 6.0 millimeter diameter holes spaced sure probe. Perforate the bottom two thirds 90 degrees apart every 0.15 meters. Place the of the pressure probe. Perforations shall be pressure probe in the center of the hole and
616
VerDate 11
backfill with gravel to a level 0.30 meters with cover material or material equal in per- above the perforated section. Add a layer of meability to the existing cover material. backfill material at least 1.2 meters thick. The specifications for pressure probe instal- Add a layer of bentonite at least 0.3 meters lation are shown in figure 5. thick, and backfill the remainder of the hole
617
VerDate 11
3.4 LFG Flow Rate Measurement. Deter- ure the LFG flow rate may be used with ap- mine the flow rate of LFG from the test proval of the Administrator. Locate the ori- wells continuously during testing with an fice meter as shown in figure 1. Attach the orifice meter. Alternative methods to meas- wells to the blower and flare assembly. The
618
VerDate 11
individual wells may be ducted to a common 3.7.1 Use the blower to extract LFG from header so that a single blower and flare as- a single well at a rate at least twice the stat- sembly and flow meter may be used. Use the ic flow rate of the respective well measured procedures in section 4.1 to calibrate the in section 3.6.2. If using a single blower and flow meter. flare assembly and a common header system, 3.5 Leak Check. A leak check of the above close the control valve on the wells not being ground system is required for accurate flow measured. Allow 24 hours for the system to rate measurements and for safety. Sample stabilize at this flow rate. LFG at the wellhead sample port and at a 3.7.2 Check for infiltration of air into the point downstream of the flow measuring de- landfill by measuring the temperature of the vice. Use Method 3C to determine nitrogen LFG at the wellhead, the gauge pressures of (N2) concentrations. Determine the dif- the shallow pressure probes, and the LFG N2 ference by using the formula below. concentration by using Method 3C. CAU- TION: Increased vacuum at the wellhead Difference=Co¥Cw may cause infiltration of air into the land- where, fill, which increases the possibility of a land- Co=concentration of N2 at the outlet, ppmv fill fire. Infiltration of air into the landfill Cw=concentration of N2 at the wellhead, may occur if any of the following conditions ppmv are met: the LFG N2 concentration is more The system passes the leak check if the than 20 percent, any of the shallow probes difference is less than 10,000 ppmv. If the sys- have a negative gauge pressure, or the tem- tem fails the leak check, make the appro- perature has increased above 55°C or the priate adjustments to the above ground sys- maximum established temperature during tem and repeat the leak check. static testing. If infiltration has not oc- 3.6 Static Testing. The purpose of the curred, increase the blower vacuum by 4 mm static testing is to determine the initial con- Hg, wait 24 hours, and repeat the infiltration ditions of the landfill. Close the control check. If at any time, the temperature valves on the wells so that there is no flow change exceeds the limit, stop the test until of landfill gas from the well. Measure the it is safe to proceed. Continue the above gauge pressure (Pg) at each deep pressure steps of increasing blower vacuum by 4 mm probe and the barometric pressure (Pbar) Hg, waiting 24 hours, and checking for infil- every 8 hours for 3 days. Convert the gauge tration until the concentration of N2 exceeds pressure of each deep pressure probe to abso- 20 percent or any of the shallow probes have lute pressure by using the following equa- a negative gauge pressure, at which time re- tion. Record as Pi. duce the vacuum at the wellhead so that the Pi=Pbar+Pg N2 concentration is less than 20 percent and where, the gauge pressures of the shallow probes are P =Atmospheric pressure, mm Hg positive. This is the maximum vacuum at bar which infiltration does not occur. Pg=Gauge pressure of the deep probes, mm Hg 3.7.3 At this maximum vacuum, measure Pbar every 8 hours for 24 hours and record the Pi=Initial absolute pressure of the deep probes during static testing, mm Hg LFG flow rate as Qs and the probe gauge 3.6.1 For each probe, average all of the 8 pressures for all of the probes as Pf. Convert hr deep pressure probe readings and record as the gauge pressures of the deep probes to ab- solute pressures for each 8-hour reading at Qs Pia. The Pia is used in section 3.7.6 to deter- mine the maximum radius of influence. as follows: 3.6.2 Measure the LFG temperature and P=Pbar+Pf the static flow rate of each well once during where,
static testing using a flow measurement de- Pbar=Atmospheric pressure, mm Hg vice, such as a Type S pitot tube and meas- Pf=Final absolute pressure of the deep probes ure the temperature of the landfill gas. The during short term testing, mm Hg flow measurements should be made either P=Pressure of the deep probes, mm Hg just before or just after the measurements of 3.7.4 For each probe, average the 8-hr deep the probe pressures and are used in deter- pressure probe readings and record as Pfa. mining the initial flow from the extraction 3.7.5 For each probe, compare the initial well during the short term testing. The tem- average pressure (Pia) from section 3.6.1 to perature measurement is used in the check the final average pressure (Pfa). Determine for infiltration. the furthermost point from the wellhead 3.7 Short Term Testing. The purpose of along each radial arm where Pfa ≤ Pia. This short term testing is to determine the maxi- distance is the maximum radius of influence mum vacuum that can be applied to the (ROI), which is the distance from the well af- wells without infiltration of air into the fected by the vacuum. Average these values landfill. The short term testing is done on to determine the average maximum radius of one well at a time. During the short term influence (Rma). testing, burn LFG with a flare or inciner- The average Rma may also be determined ator. by plotting on semi-log paper the pressure
619
VerDate 11
th differentials (Pfa-Pia) on the y-axis (abscissa) tvi=time of the i interval, hour (usually 8) versus the distances (3, 15, 30 and 45 meters) 3.9.2 Record the final stabilized flow rate from the wellhead on the x-axis (ordinate). as Qf. If, during the long term testing, the Use a linear regression analysis to determine flow rate does not stabilize, calculate Qf by the distance when the pressure differential is averaging the last 10 recorded flow rates. zero. Additional pressure probes may be used 3.9.3 For each deep probe, convert each to obtain more points on the semi-long plot gauge pressure to absolute pressure as in sec- of pressure differentials versus distances. tion 3.7.4. Average these values and record as 3.7.6 Calculate the depth (Dst) affected by Psa. For each probe, compare Pia to Psa. Deter- the extraction well during the short term mine the furthermost point from the well- test as follows. If the computed value of Dst head along each radial arm where Psa ≤ Pia. exceeds the depth of the landfill, set Dst This distance is the stabilized radius of influ- equal to the landfill depth. ence. Average these values to determine the 2 Dst=WD + Rma average stabilized radius of influence (Rsa). where, 3.10 Determine the NMOC mass emission rate using the procedures in section 5. Dst=depth, m 3.11 Deactivation of pressure probe holes. WD=well depth, m Upon completion of measurements, if pres- Rma=maximum radius of influence, m sure probes are removed, restore the integ- 3.7.7 Calculate the void volume for the ex- rity of the landfill cover by backfilling and traction well (V) as follows. sealing to prevent venting of LFG to the at- 2 V=0.40 Rma Dst mosphere or air infiltration. where, 4. Calibrations V=void volume of test well, m3 Rma=maximum radius of influence, m Gas Flow Measuring Device Calibration Dst=depth, m Procedure. Locate a standard pitot tube in 3.7.8 Repeat the procedures in section 3.7 line with a gas flow measuring device. Use for each well. the procedures in Method 2D, section 4, to 3.8 Calculate the total void volume of the calibrate the orifice meter. Method 3C may test wells (Vv) by summing the void volumes be used to determine the dry molecular (V) of each well. weight. It may be necessary to calibrate 3.9 Long Term Testing. The purpose of more than one gas flow measuring device to long term testing is to determine the meth- bracket the landfill gas flow rates. Construct ane generation rate constant, k. Use the a calibration curve by plotting the pressure blower to extract LFG from the wells. If a drops across the gas flow measuring device single blower and flare assembly and com- for each flow rate versus the average dry gas mon header system are used, open all control volumetric flow rate in cubic meters per valves and set the blower vacuum equal to minute of the gas. Use this calibration curve the highest stabilized blower vacuum dem- to determine the volumetric flow from the onstrated by any individual well in section wells during testing. 3.7. Every 8 hours, sample the LFG from the wellhead sample port, measure the gauge 5. Calculations pressures of the shallow pressure probes, the 5.1 Nomenclature. blower vacuum, the LFG flow rate, and use A =average age of the solid waste tested, the criteria for infiltration in section 3.7.2 avg year and Method 3C to check for infiltration. If infiltration is detected, do not reduce the Ai=age of solid waste in the ith fraction, year blower vacuum, but reduce the LFG flow A=age of landfill, year rate from the well by adjusting the control Ar=acceptance rate, megagrams per year valve on the wellhead. Adjust each affected CNMOC=NMOC concentration, ppmv as hexane well individually. Continue until the equiva- (CNMOC=Ct/6) lent of two total void volumes (V ) have been Ct=NMOC concentration, ppmv (carbon v equivalent) from Method 25C extracted, or until Vt=2 Vv. D = depth affected by the test wells, m 3.9.1 Calculate Vt, the total volume of LFG extracted from the wells, as follows. Dst=depth affected by the test wells in the short term test, m n DLF=landfill depth, m V = ∑60 Q t f = fraction of decomposable solid waste in t i vi the landfill i=1 th fi=fraction of the solid waste in the i sec- where, tion ¥1 Vt=total volume of LFG extracted from k=methane generation rate constant, year wells, m3 Lo=methane generation potential, cubic me- Qi=LFG flow rate measured at orifice meter ters per megagram th at the i interval, cubic meters per Lo=revised methane generation potential to minute account for the amount of
620
VerDate 11
nondecomposable material in the land- D=WD+Rsa fill, cubic meters per megagram 5.3 Use the following equation to cal- th Mi=mass of solid waste of the i section, culate the volume of solid waste affected by megagrams the test well. M =mass of decomposable solid waste af- r V =R 2 ² D fected by the test well, megagrams r sa M =number of wells 5.4 Use the following equation to cal- w culate the mass affected by the test well. Pbar=atmospheric pressure, mm Hg Pg=gauge pressure of the deep pressure Mr=Vrρ probes, mm Hg 5.5 Modify Lo to account for the Pi=initial absolute pressure of the deep pres- nondecomposable solid waste in the landfill. sure probes during static testing, mm Hg Lo′=f Lo Pia=average initial absolute pressure of the deep pressure probes during static test- 5.6 In the following equation, solve for k ing, mm Hg by iteration. A suggested procedure is to se- lect a value for k, calculate the left side of Pf=final absolute pressure of the deep pres- sure probes during short term testing, the equation, and if not equal to zero, select mm Hg another value for k. Continue this process until the left hand side of the equation Pfa=average final absolute pressure of the deep pressure probes during short term equals zero, #0.001. testing, mm Hg − Q Ps=final absolute pressure of the deep pres- k − × 5 f = sure probes during long term testing, ke Aavg ()5. 256 10 0 mm Hg 2 Lo© M r P =average final absolute pressure of the sa 5.7 Use the following equation to deter- deep pressure probes during long term testing, mm Hg mine landfill NMOC mass emission rate if Q =required blow flow rate, cubic meters per the yearly acceptance rate of solid waste has B ± minute been consistent ( 10 percent) over the life of the landfill. Qf=final stabilized flow rate, cubic meters ¥k 11 per minute Qt = 2 Lo′ Ar (1 ¥ e A) CNMOC / (5.256 × 10 ) Qi=LFG flow rate measured at orifice meter 5.8 Use the following equation to deter- during the ith interval, cubic meters per mine landfill NMOC mass emission rate if minute the acceptance rate has not been consistent Qs=maximum LFG flow rate at each well de- over the life of the landfill. termined by short term test, cubic me- ters per minute n 2 kLo© C NMOC −kt Qt=NMOC mass emission rate, cubic meters Q= ∑ M e i per minute t11 i ()5. 256× 10 = Rm=maximum radius of influence, m i 1 Rma=average maximum radius of influence, m 6. Bibliography Rs=stabilized radius of influence for an indi- vidual well, m 1. Same as Method 2, appendix A, 40 CFR Rsa=average stabilized radius of influence, m part 60. ti=age of section i, year 2. Emcon Associates, Methane Generation tt=total time of long term testing, year and Recovery from Landfills. Ann Arbor V=void volume of test well, m3 Science, 1982. Vr=volume of solid waste affected by the test 3. The Johns Hopkins University, Brown well, m3 Station Road Testing and Gas Recovery Pro- Vt=total volume of solid waste affected by jections. Laurel, Maryland: October 1982. the long term testing, m3 4. Mandeville and Associates, Procedure V =total void volume affected by test wells, v Manual for Landfill Gases Emission Testing. m3 WD=well depth, m 5. Letter and attachments from Briggum, ρ=solid waste density, m3 (Assume 0.64 S., Waste Management of North America, to megagrams per cubic meter if data are Thorneloe, S., EPA. Response to July 28, 1988 unavailable) request for additional information. August 5.2 Use the following equation to cal- 18,1988. culate the depth affected by the test well. If 6. Letter and attachments from Briggum, using cluster wells, use the average depth of S., Waste Management of North America, to the wells for WD. If the value of D is greater Wyatt, S., EPA. Response to December 7, than the depth of the landfill, set D equal to 1988 request for additional information. Jan- the landfill depth. uary 16, 1989.
621
VerDate 11
METHOD 3—GAS ANALYSIS FOR THE is analyzed for pecent CO2, percent O2, and if DETERMINATION OF DRY MOLECULAR WEIGHT necessary, for percent CO. For dry molecular weight determination, either an Orsat or a 1. APPLICABILITY AND PRINCIPLE Fyrite analyzer may be used for the analysis. 1.1 Applicability. 2. APPARATUS 1.1.1 This method is applicable for deter- As an alternative to the sampling appara- mining carbon dioxide (CO2) and oxygen (O2) concentrations and dry molecular weight of tus and systems described herein, other sam- a sample from a gas stream of a fossil-fuel pling systems (e.g., liquid displacement) may combustion process. The method may also be be used, provided such systems are capable of applicable to other processes where it has obtaining a representative sample and main- been determined that compounds other than taining a constant sampling rate, and are, otherwise, capable of yielding acceptable re- CO2, O2, carbon monoxide (CO), and nitrogen sults. Use of such systems is subject to the (N2) are not present in concentrations suffi- cient to affect the results. approval of the Administrator. 1.1.2 Other methods, as well as modifica- tions to the procedure described herein, are 2.1 Grab Sampling (Figure 3–1). also applicable for some or all of the above 2.1.1 Probe. Stainless steel or borosilicate determinations. Examples of specific meth- glass tubing equipped with an in-stack or ods and modifications include: (1) A multi- out-stack filter to remove particulate mat- point sampling method using an Orsat ana- ter (a plug of glass wool is satisfactory for lyzer to analyze individual grab samples ob- this purpose). Any other materials, inert to tained at each point; (2) a method using CO2 O2, CO2, CO, and N2 and resistant to tempera- or O2 and stoichiometric calculations to de- ture at sampling conditions, may be used for termine dry molecular weight; and (3) as- the probe. Examples of such materials are signing a value of 30.0 for dry molecular aluminum, copper, quartz glass, and Teflon. weight, in lieu of actual measurements, for 2.1.2 Pump. A one-way squeeze bulb, or processes burning natural gas, coal, or oil. equivalent, to transport the gas sample to These methods and modifications may be the analyzer. used, but are subject to the approval of the Administrator, U.S. Environmental Protec- 2.2 Integrated Sampling (Figure 3–2). tion Agency (EPA). 2.2.1 Probe. Same as in Section 2.1.1. 1.1.3 Note. Mention of trade names or spe- 2.2.2 Condenser. An air-cooled or water- cific products does not constitute endorse- cooled condenser, or other condenser no ments by EPA. greater than 250 ml that will not remove O2, 1.2 Principle. A gas sample is extracted CO2, CO, and N2, to remove excess moisture from a stack by one of the following meth- which would interfere with the operation of ods: (1) Single-point, grab sampling; (2) sin- the pump and flowmeter. gle-point, integrated sampling; or (3) multi- 2.2.3 Valve. A needle valve, to adjust sam- point, integrated sampling. The gas sample ple gas flow rate.
622
VerDate 11
2.2.4 Pump. A leaf-free, diaphragm-type rate. A flow rate range of 500 to 1000 cc/min pump, or equivalent, to transport sample gas is suggested. to the flexible bag. Install a small surge tank 2.2.6 Flexible Bag. Any leak-free plastic between the pump and rate meter to elimi- (e.g., Tedlar, Mylar, Teflon) or plastic-coated nate the pulsation effect of the diaphragm aluminum (e.g., aluminized Mylar) bag, or pump on the rotameter. equivalent, having a capacity consistent 2.2.5 Rate Meter. A rotameter, or equiva- with the selected flow rate and time length lent rate meter, capable of measuring flow of the test run. A capacity in the range of 55 rate to within 2 percent of the selected flow to 90 liters is suggested. To leak check the
623
VerDate 11
bag, connect it to a water manometer, and should remain stable for at least 0.5 minute. pressurize the bag to 5 to 10 cm H2O (2 to 4 Evacuate the flexible bag. Connect the in. H2O). Allow to stand for 10 minutes. Any probe, and place it in the stack, with the tip displacement in the water manometer indi- of the probe positioned at the sampling cates a leak. An alternative leak-check point; purge the sampling line. Next, connect method is to pressurize the bag to 5 to 10 cm the bag, and make sure that all connections (2 to 4 in.) H2O and allow to stand overnight. are tight. A deflated bag indicates a leak. 4.3 Sample at a constant rate. The sam- 2.2.7 Pressure Gauge. A water-filled U-tube pling run should be simultaneous with, and manometer, or equivalent, of about 30 cm (12 for the same total length of time as, the pol- in.), for the flexible bag leak check. lutant emission rate determination. Collec- 2.2.8 Vacuum Gauge. A mercury manom- tion of at least 30 liters (1.00 ft3) of sample eter, or equivalent, of at least 760 mm (30 in.) gas is recommended; however, smaller vol- Hg, for the sampling train leak check. umes may be collected, if desired. 2.3 Analysis. An Orsat or Fyrite type com- 4.4 Obtain one integrated flue gas sample bustion gas analyzer. For Orsat and Fyrite during each pollutant emission rate deter- analyzer maintenance and operation proce- mination. Within 8 hours after the sample is dures, follow the instructions recommended taken, analyze it for percent CO2 and percent by the manufacturer, unless otherwise speci- O2 using either an Orsat analyzer or a Fyrite fied herein. type combustion gas analyzer. If an Orsat analyzer is used, it is recommended that 3. SINGLE-POINT, GRAB SAMPLING AND Orsat leak check described in Section 6, be ANALYTICAL PROCEDURE performed before this determination; how- 3.1 The sampling point in the duct shall ei- ever, the check is optional. Determine the and CO by ther be at the centroid of the cross section or percentage of the gas that is N2 subtracting the sum of the percent CO and at a point no closer to the walls than 1.00 m 2 percent 0 from 100 percent. Calculate the dry (3.3 ft), unless otherwise specified by the Ad- molecular weight as indicated in Section 7.2. ministrator. 4.5 Repeat the analysis and calculation 3.2 Set up the equipment as shown in Fig- procedures until the individual dry molecu- ure 3–1, making sure all connections ahead of lar weights for any three analyses differ the analyzer are tight. If an Orsat analyzer is from their mean by no more than 0.3 g/g- used, it is recommended that the analyzer be mole (0.3 lb/lb-mole). Average these three leak checked by following the procedure in molecular weights, and report the results to Section 6; however, the leak check is op- the nearest 0.1 g/g-mole (0.1 lb/lb-mole). tional. 3.3 Place the probe in the stack, with the 5. MULTI-POINT, INTEGRATED SAMPLING AND tip of the probe positioned at the sampling ANALYTICAL PROCEDURE point; purge the sampling line long enough to allow at least five exchanges. Draw a sam- 5.1 Unless otherwise specified by the Ad- ple into the analyzer, and immediately ana- ministrator, a minimum of eight traverse points shall be used for circular stacks hav- lyze it for percent CO2 and percent O2. Deter- ing diameters less than 0.61 m (24 in.), a min- mine the percentage of the gas that is N2 and CO by subtracting the sum of the percent imum of nine shall be used for rectangular stacks having equivalent diameters less than CO2 and percent 02 O from 100 percent. Cal- culate the dry molecular weight as indicated 0.61 m (24 in.), and a minimum of 12 traverse in Section 7.2. points shall be used for all other cases. The 3.4 Repeat the sampling, analysis, and cal- traverse points shall be located according to culation procedures until the dry molecular Method 1. The use of fewer points is subject weights of any three grab samples differ to approval of the Administrator. from their mean by no more than 0.3 g/g- 5.2 Follow the procedures outlined in Sec- mole (0.3 lb/lb-mole). Average these three tions 4.2 through 4.5, except for the follow- molecular weights, and report the results to ing: Traverse all sampling points, and sam- the nearest 0.1 g/g-mole (0.1 lb/lb-mole). ple at each point for an equal length of time. Record sampling data as shown in Figure 3– 4. SINGLE-POINT, INTEGRATED SAMPLING AND 3. ANALYTICAL PROCEDURE Time Traverse pt. Q, liter/min % dev.a 4.1 The sampling point in the duct shall be located as specified in Section 3.1. 4.2 Leak check (optional) the flexible bag as in Section 2.2.6. Set up the equipment as shown in Figure 3–2. Just before sampling, leak check (optional) the train by placing a vacuum gauge at the condenser inlet, pulling a vacuum of at least 250 mm Hg (10 in. Hg), plugging the outlet at the quick disconnect, Average a and then turning off the pump. The vacuum % dev.=(QÐQavg)/Qavg X 100 (Must be ≤10%)
624
VerDate 11
Figure 3±3. Sampling rate data. 8. BIBLIOGRAPHY
6. LEAK-CHECK PROCEDURE FOR ORSAT 1. Altshuller, A.P. Storage of Gases and ANALYZER Vapors in Plastic Bags. International Jour- nal of Air and Water Pollution. 6:75–81. 1963. Moving an Orsat analyzer frequently 2. Conner, William D. and J.S. Nader. Air causes it to leak. Therefore, an Orsat ana- Sampling with Plastic Bags. Journal of the lyzer should be thoroughly leak checked on American Industrial Hygiene Association. site before the flue gas sample is introduced 25:292–297. 1964. into it. The procedure for leak checking an 3. Burrell Manual for Gas Analysts, Sev- Orsat analyzer is as follows: enth edition. Burrell Corporation, 2223 Fifth 6.1 Bring the liquid level in each pipette up Avenue, Pittsburgh, PA. 15219. 1951. to the reference mark on the capillary tub- 4. Mitchell, W.J. and M.R. Midgett, Field ing, and then close the pipette stopcock. Reliability of the Orsat Analyzer. Journal of 6.2 Raise the leveling bulb sufficiently to Air Pollution Control Association. 26:491–495. bring the confining liquid meniscus onto the May 1976. graduated portion of the burette, and then 5. Shigehara, R.T., R. M. Neulicht, and close the manifold stopcock. W.S. Smith. Validating Orsat Analysis Data from Fossil Fuel-Fired Units. Stack Sam- 6.3 Record the meniscus position. pling News. 4(2):21–26. August 1976. 6.4 Observe the menisus in the burette and the liquid level in the pipette for movement METHOD 3A—DETERMINATION OF OXYGEN AND over the next 4 minutes. CARBON DIOXIDE CONCENTRATIONS IN EMIS- 6.5 For the Orsat analyzer to pass the leak SIONS FROM STATIONARY SOURCES (INSTRU- check, two conditions must be met: MENTAL ANALYZER PROCEDURE) 6.5.1 The liquid level in each pipette must 1. Applicability and Principle not fall below the botton of the capillary 1.1 Applicability. This method is appli- tubing during this 4-minute interval. cable to the determination of oxygen (O2) 6.5.2 The menisus in the burette must not and carbon dioxide (CO2) concentrations in change by more than 0.2 ml during this 4- emissions from stationary sources only when minute interval. specified within the regulations. 6.6 If the anlyzer fails the leak-check pro- 1.2 Principle. A sample is continuously cedure, check all rubber connections and extracted from the effluent stream: a portion stopcocks to determine whether they might of the sample stream is conveyed to an in- be the cause of the leak. Disassemble, clean, strumental analyzer(s) for determination of and regrease leaking stopcocks. Replace O2 and CO2 concentration(s). Performance leaking rubber connections. After the ana- specifications and test procedures are pro- lyzer is reassembled, repeat the lead-check vided to ensure reliable data. procedure. 2. Range and Sensitivity 7. CALCULATIONS Same as Method 6C, Sections 2.1 and 2.2, except that the span of the monitoring sys- 7.1 Nomenclature tem shall be selected such that the average O2 or CO2 concentration is not less than 20 Md = Dry molecular weight, g/g-mole (1b/1b- mole). percent of the span. %CO2 = Percent CO2 by volume, dry basis. 3. Definitions %O2 = Percent O2 by volume, dry basis. 3.1 Measurement System. The total equip- %CO = Percent CO by volume, dry basis. ment required for the determination of the %N2 = Percent N2 by volume, dry basis. O2 or CO2 concentration. The measurement 0.280 = Molecular weight of N2 or CO, divided system consists of the same major sub- by 100. systems as defined in Method 6C, Sections 0.320 = Molecular wight of O2 divided by 100. 3.1.1, 3.1.2, and 3.1.3. 0.440 = Molecular weight of CO2 divided by 3.2 Span, Calibration Gas, Analyzer Cali- 100. bration Error, Sampling System Bias, Zero 7.2 Dry Molecular Weight. Use Equation 3– Drift, Calibration Drift, Response Time, and 1 to calculate the dry molecular weight of Calibration Curve. Same as Method 6C, Sec- the stack gas. tions 3.2 through 3.8, and 3.10. 3.3 Interference Response. The output re- M = 0.440(%CO ) + 0.320 (%O ) + 0.280(%N + d 2 2 2 sponse of the measurement system to a com- %CO) Eq. 3–1 ponent in the sample gas, other than the gas NOTE. The above equation does not con- component being measured. sider argon in air (about 0.9 percent, molecu- 4. Measurement System Performance Specifica- lar weight of 39.9). A negative error of about tions 0.4 percent is introduced. The tester may choose to include argon in the analysis using Same as Method 6C, Sections 4.1 through procedures subject to approval of the Admin- 4.4. istrator. 5. Apparatus and Reagents
625
VerDate 11
5.1 Measurement System. Any measure- alter the interference response (e.g., changes ment system for O2 or CO2 that meets the in the type of gas detector). Conduct the in- specifications of this method. A schematic of terference response in accordance with Sec- an acceptable measurement system is shown tion 5.4 of Method 20. in Figure 6C–1 of Method 6C. The essential 6.3 Measurement System Preparation, components of the measurement system are Analyzer Calibration Error, and Sampling described below: System Bias Check. Follow Sections 6.2 5.1.1 Sample Probe. A leak-free probe, of through 6.4 of Method 6C. sufficient length to traverse the sample 7. Emission Test Procedure points. 7.1 Selection of Sampling Site and Sam- 5.1.2 Sample Line. Tubing, to transport pling Points. Select a measurement site and the sample gas from the probe to the mois- sampling points using the same criteria that ture removal system. A heated sample line is are applicable to tests performed using not required for systems that measure the O 2 Method 3. or CO concentration on a dry basis, or trans- 2 7.2 Sample Collection. Position the sam- port dry gases. pling probe at the first measurement point, 5.1.3 Sample Transport Line, Calibration and begin sampling at the same rate as used Value Assembly, Moisture Removal System, during the sampling system bias check. Particulate Filter, Sample Pump, Sample Maintain constant rate sampling (i.e., ±10 Flow Rate Control, Sample Gas Manifold, percent) during the entire run. The sampling and Data Recorder. Same as Method 6C, Sec- time per run shall be the same as for tests tions 5.1.3 through 5.1.9, and 5.1.11, except conducted using Method 3 plus twice the sys- that the requirements to use stainless steel, tem response time. For each run, use only Teflon, and nonreactive glass filters do not those measurements obtained after twice the apply. response time of the measurement system 5.1.4 Gas Analyzer. An analyzer to deter- has elapsed to determine the average efflu- mine continuously the O or CO concentra- 2 2 ent concentration. tion in the sample gas stream. The analyzer 7.3 Zero and Calibration Drift Test. Fol- shall meet the applicable performance speci- low Section 7.4 of Method 6C. fications of Section 4. A means of controlling the analyzer flow rate and a device for deter- 8. Quality Control Procedures mining proper sample flow rate (e.g., preci- The following quality control procedures sion rotameter, pressure gauge downstream are recommended when the results of this of all flow controls, etc.) shall be provided at method are used for an emission rate correc- the analyzer. The requirements for measur- tion factor, or excess air determination. The ing and controlling the analyzer flow rate tester should select one of the following op- are not applicable if data are presented that tions for validating measurement results: demonstrate the analyzer is insensitive to 8.1 If both O2 and CO2 are measured using flow variations over the range encountered Method 3A, the procedures described in Sec- during the test. tion 4.4 of Method 3 should be followed to 5.2 Calibration Gases. The calibration validate the O2 and CO2 measurement re- gases for CO2 analyzers shall be CO2 in N2 or sults. CO2 in air. Alternatively, CO2/SO2, O2/SO2 , or 8.2 If only O2 is measured using Method O2/CO2/SO2 gas mixtures in N2 may be used. 3A, measurements of the sample stream CO2 Three calibration gases, as specified Section concentration should be obtained at the sam- 5.3.1 through 5.3.3 of Method 6C, shall be ple by-pass vent discharge using an Orsat or used. For O2 monitors that cannot analyze Fyrite analyzer, or equivalent. Duplicate zero gas, a calibration gas concentration samples should be obtained concurrent with equivalent to less than 10 percent of the span at least one run. Average the duplicate Orsat may be used in place of zero gas. or Fyrite analysis results for each run. Use 6. Measurement System Performance Test Proce- the average CO2 values for comparison with dures the O2 measurements in accordance with the procedures described in Section 4.4 of Meth- Perform the following procedures before od 3. measurement of emissions (Section 7). 8.3 If only CO is measured using Method 6.1 Calibration Concentration Verifica- 2 3A, concurrent measurements of the sample tion. Follow Section 6.1 of Method 6C, except stream CO concentration should be obtained if calibration gas analysis is required, use 2 using an Orsat or Fyrite analyzer as de- Method 3 and change the acceptance criteria scribed in Section 8.2. For each run, dif- for agreement among Method 3 results to 5 ferences greater than 0.5 percent between the percent (or 0.2 percent by volume, whichever Method 3A results and the average of the du- is greater). plicate Fyrite analysis should be inves- 6.2 Interference Response. Conduct an in- tigated. terference response test of the analyzer prior to its initial use in the field. Thereafter, re- 9. Emission Calculation check the measurement system if changes For all CO2 analyzers, and for O2 analyzers are made in the instrumentation that could that can be calibrated with zero gas, follow
626
VerDate 11
Section 8 of Method 6C, except express all the effluent gas concentration using Equa- concentrations as percent, rather than ppm. tion 3A–1. For O2 analyzers that use a low-level cali- bration gas in place of a zero gas, calculate
CC− C = ma oa ()C− C+ C Eq.-3 A 1 gas − m ma CCm o
Where: must be used for excess air or emission rate correction factor determinations. Cgas=Effluent gas concentration, dry basis, percent. 2. APPARATUS Cma=Actual concentration of the upscale calibration gas, percent. The alternative sampling systems are the Coa=Actual concentration of the low-level same as those mentioned in Section 2 of calibration gas, percent. Method 3. Cm=Average of initial and final system cali- 2.1 Grab Sampling and Integrated Sam- bration bias check responses for the pling. Same as in Sections 2.1 and 2.2, respec- upscale calibration gas, percent. tively, of Method 3. Co=Average of initial and final system cali- 2.2 Analysis. An Orsat analyzer only. For bration bias check responses for the low- low CO2 (less than 4.0 percent) or high O2 level gas, percent. (greater than 15.0 percent) concentrations, C¯ =Average gas concentration indicated by the measuring burette of the Orsat must the gas analyzer, dry basis, percent. have at least 0.1 percent subdivisions. For 10. Bibliography Orsat maintenance and operation proce- Same as bibliography of Method 6C. dures, follow the instructions recommended by the manufacturer, unless otherwise speci- METHOD 3B—GAS ANALYSIS FOR THE DETER- fied herein. MINATION OF EMISSION RATE CORRECTION 3. PROCEDURES FACTOR OR EXCESS AIR Each of the three procedures below shall be 1. APPLICABILITY AND PRINCIPLE used only when specified in an applicable 1.1 Applicability subpart of the standards. The use of these procedures for other purposes must have spe- 1.1.1 This method is applicable for deter- cific prior approval of the Adminsitrator. mining carbon dioxide (CO ), oxygen (O ), 2 2 NOTE .—A Fyrite-type combustion gas ana- and carbon monoxide (CO) concentrations of lyzer is not acceptable for excess air or emis- a sample from a gas stream of a fossil-fuel sion rate correction factor determinations, combustion provess for excess air or emis- unless approved by the Administrator. If sion rate correction factor calculations. both percent CO and percent O are meas- 1.1.2 Other methods, as well as modifica- 2 2 ured, the analytical results of any of the tions to the procedure described herein, are three procedures given below may be used for also applicable for all of the above deter- calculating the dry molecular weight (see minations. Examples of specific methods and Method 3). modifications include: (1) A multi-point sam- pling method using an Orsat analyzer to ana- 3.1 Single-Point, Grab Sampling and Analytical lyze individual grab samples obtained at Procedure. each point, and (2) a method using CO2 or O2 and stoichiometric calculations to determine 3.1.1 The sampling point in the duct shall excess air. These methods and modifications be as described in Section 3.1 of Method 3. may be used, but are subject to the approval 3.1.2 Set up the equipment as shown in Fig- of the Administrator, U.S. Environmental ure 3–1 of Method 3, making sure all connec- Protection Agency (FPA). tions ahead of the analyzer are tight. Leak 1.1.3 Note. Mention of trade names or spe- check the Orsat analyzer according to the cific products does not constitute endorse- procedure described in Section 6 of Method 3. ment by EPA. This leak check is mandatory. 1.2 Principle. A gas sample is extracted 3.1.3 Place the probe in the stack, with from a stack by one of the following meth- the tip of the probe positioned at the sam- ods: (1) Single-point, grab sampling; (2) sin- pling point; purge the sampling line long gle-point, integrated sampling; or (3) multi- enough to allow at least five exchanges. point, integrated sampling. The gas sample Draw a sample into the analyzer. For emis- is analyzed for percent CO2 percent O2, and, sion rate correction factor determinations, if necessary, percent CO. An Orsat analyzer immediately analyze the sample, as outlined
627
VerDate 11
in Sections 3.1.4 and 3.1.5, for percent CO2 or and CO; (2) determine the percentage of the percent O2. If excess air is desired, proceed as gas that is N2 by subtracting the sum of the follows: (1) immediately analyze the sample, percent CO2, percent O2, and percent CO from as in Sections 3.1.4 and 3.1.5, for percent CO2, 100 percent; and (3) calculate percent excess O2, and CO; (2) determine the percentage of air, as outlined in Section 4.2. the gas that is N2 by subtracting the sum of 3.2.5 To ensure complete absorption of the the percent CO2, percent O2, and percent CO CO2, O2, or if applicable, CO, follow the pro- from 100 percent, and (3) calculate percent cedure described in Section 3.1.4. excess air as outlined in Section 4.2. NOTE. —Although in most instances only 3.1.4 To ensure complete absorption of the CO or O is required, it is recommended that CO , O , or if applicable, CO, make repeated 2 2 2 2 both CO and O be measured, and that Sec- passes through each absorbing solution until 2 2 tion 3.4.1 be used to validate the analytical two consecutive readings are the same. Sev- data. eral passes (three or four) should be made be- tween readings. (If constant readings cannot 3.2.6 Repeat the analysis until the follow- be obtained after three consecutive readings, ing criteria are met: replace the absorbing solution.) 3.2.6.1 For percent CO2, repeat the analyt- ical procedure until the results of any three NOTE. —Since this single-point, grab sam- analyses differ by no more than (a) 0.3 per- pling and analytical procedure is normally cent by volume when CO is greater than 4.0 conducted in conjunction with a single- 2 percent or (b) 0.2 percent by volume when point, grab sampling and analytical proce- CO is less than or equal to 4.0 percent. Aver- dure for a pollutant, only one analysis is or- 2 age three acceptable values of percent CO , dinarily conducted. Therefore, great care 2 and report the results to the nearest 0.2 per- must be taken to obtain a valid sample and cent. analysis. Although in most cases, only CO 2 3.2.6.2 For percent O , repeat the analyt- or O is required, it is recommended that 2 2 ical procedure until the results of any three both CO and O be measured, and that Sec- 2 2 analyses differ by no more than (a) 0.3 per- tion 3.4 be used to validate the analytical cent by volume when O is less than 15.0 per- data. 2 cent or (b) 0.2 percent by volume when O2 is 3.1.5 After the analysis is completed, leak greater than or equal to 15.0 percent. Aver- check (mandatory) the Orsat analyzer once age the three acceptable values of percent again, as described in Section 6 of Method 3. O2, and report the results to the nearest 0.1 For the results of the analysis to be valid, percent. the Orsat analyzer must pass this leak test 3.2.6.3 For percent CO, repeat the analyt- before and after the analysis. ical procedure until the results of any three 3.2 Single-Point, Integrated Sampling and analyses differ by no more than 0.3 percent. Analytical Procedure. Average the three acceptable values of per- cent CO, and report the results to the near- 3.2.1 The sampling point in the duct shall est 0.1 percent. be located as specified in Section 3.1.1. 3.2.7 After the analysis is completed, leak 3.2.2 Leak check (mandatory) the flexible check (mandatory) the Orsat analyzer once bag as in Section 2.2.6 of Method 3. Set up again, as described in Section 6 of Method 3. the equipment as shown in Figure 3–2 of For the results of the analysis to be valid, Method 3. Just before sampling, leak check the Orsat analyzer must pass this leak test (mandatory) the train as described in Sec- before and after the analysis. tion 4.2 of Method 3. 3.2.3 Sample at a constant rate, or as 3.3 Multi-Point, Integrated Sampling and specified by the Administrator. The sam- Analytical Procedure. pling run must be simultaneous with, and for the same total length of time as, the pollut- 3.3.1 The sampling points shall be deter- ant emission rate determination. Collect at mined as specified in Section 5.3 of Method 3. least 30 liters (1.00 ft3) of sample gas. Smaller 3.3.2 Follow the procedures outlined in volumes may be collected, subject to ap- Sections 3.2.2 through 3.2.7, except for the proval of the Administrator. following: Traverse all sampling points, and 3.2.4 Obtain one integrated flue gas sam- sample at each point for an equal length of ple during each pollutant emission rate de- time. Record sampling data as shown in Fig- termination. For emission rate correction ure 3–3 of Method 3. factor determination, analyze the sample 3.4 Quality Control Procedure. within 4 hours after it is taken for percent CO2 or percent O2 (as outlined in Sections 3.4.1 Data Validation When Both CO2 and 3.2.5 through 3.2.7). The Orsat analyzer must O2 Are Measured. Although in most in- be leak checked (see Section 6 of Method 3) stances, only CO2 or O2 measurement is re- before the analysis. If excess air is desired, quired, it is recommended that both CO2 and procede as follows: (1) within 4 hours after O2 be measured to provide a check on the the sample is taken, analyze it (as in Sec- quality of the data. The following quality tions 3.2.5 through 3.2.7) for percent CO2, O2, control procedure is suggested. 628
VerDate 11
NOTE. —Since the method for validating tion 5.2.3. Then calculate the F0 factor as fol- the CO2 and O2 analyses is based on combus- lows: tion of organic and fossil fuels and dilution of the gas stream with air, this method does = 0. 209 Fd not apply to sources that (1) remove CO2 or F0 Eq. 3 B-2 O2 through processes other than combustion, Fc (2) add O2 (e.g., oxygen enrichment) and N2 in
proportions different from that of air, (3) add Fuel type F0 range CO2 (e.g., cement or lime kilns) or (4) have no fuel factor, Fo, values obtainable (e.g., ex- Coal: tremely variable waste mixtures). This Anthracite and lignite ...... 1.016±1.130 method validates the measured proportions Bituminous ...... 1.083±1.230 Oil: of CO2 and O2 for fuel type, but the method does not detect sample dilution resulting Distillate ...... 1.260±1.413 Residual ...... 1.210±1.370 from leaks during or after sample collection. Gas: The method is applicable for samples col- Natural ...... 1.600±1.836 lected downstream of most lime or limestone Propane ...... 1.434±1.586 flue-gas desulfurization units as the CO2 Butane ...... 1.405±1.553 added or removed from the gas stream is not Wood ...... 1.000±1.120 significant in relation to the total CO2 con- Wood bark ...... 1.003±1.130 centration. The CO2 concentrations from other types of scrubbers using only water or 3.4.1.3 Calculated F0 values, beyond the ac- basic slurry can be significantly affected and ceptable ranges shown in this table, should would render the Fo check minimally useful. be investigated before accepting the test re- sults. For example, the strength of the solu- 3.4.1.1 Calculate a fuel factor, F , using o tions in the gas analyzer and the analyzing the following equation: technique should be checked by sampling 20.% 9 − O and analyzing a known concentration, such F = 2 Eq.3 B -1 as air; the fuel factor should be reviewed and o verified. An acceptability range of ±12 per- %CO2 cent is appropriate for the F0 factor of mixed where: fuels with variable fuel ratios. The level of
%O2=Percent O2 by volume, dry basis. the emission rate relative to the compliance %CO2=Percent CO2 by volume, dry basis. level should be considered in determining if 20.9=Percent O2 by volume in ambient air. a retest is appropriate, i.e., if the measured If CO present in quantities measurable by emissions are much lower or much greater this method, adjust the O2 and CO2 values be- than the compliance limit, repetition of the fore performing the calculation for F0 as fol- test would not significantly change the com- lows: pliance status of the source and would be un- necessarily time consuming and costly. %CO2 (adj) = %CO2 + %CO %O2 (adj) = %O2 ¥ 0.5 %CO 4. CALCULATIONS where: %5CO = Percent CO by volume, dry basis. 4.1 Nomenclature. Same as Section 5 of Method 3 with the addition of the following: 3.4.1.2 Compare the calculated F0 factor with the expected F0 values. The following %EA = Percent excess air. table may be used in establishing acceptable 0.264 = Ratio of O2 to N2 in air, v/v. ranges for the expected F0 if the fuel being 4.2 Percent Excess Air. Calculate the per- burned is known. When fuels are burned in cent excess air (if applicable) by substituting combinations, calculate the combined fuel Fd the appropriate values of percent O2, CO, and and Fc factors (as defined in Method 19) ac- N2 (obtained from Section 3.1.3 or 3.2.4) into cording to the procedure in Method 19, Sec- Equation 3B–3.
%O− 0 . 5 % CO %EA = 2 ×100Eq. 3 B - 3 − − 0. 264 %N2()%. O 2 0 5 % CO
NOTE. The equation above assumes that and natural gas do not contain appreciable
ambient air is used as the source of O2 and amounts of N2) or when oxygen enrichment that the fuel does not contain appreciable is used, alternative methods, subject to ap- amounts of N2 (as do coke oven or blast fur- proval of the Administrator, are required. nace gases). For those cases when appre- ciable amounts of N2 are present (coal, oil 629
VerDate 11
5. BIBLIOGRAPHY 4.2 Recorder. Recorder with linear strip chart. Electronic integrator (optional) is rec- Same as Method 3. ommended. METHOD 3C—DETERMINATION OF CARBON DI- 4.3 Teflon Tubing. Diameter and length OXIDE, METHANE, NITROGEN, AND OXYGEN determined by connection requirements of FROM STATIONARY SOURCES cylinder regulators and the GC. 4.4 Regulators. To control gas cylinder 1. Applicability and Principle pressures and flow rates. 4.5 Adsorption Tubes. Applicable traps to 1.1 Applicability. This method applies to remove any O2 from the carrier gas. the analysis of carbon dioxide (CO2), meth- ane (CH4), nitrogen (N2), and oxygen (O2) in 5. Reagents samples from municipal solid waste landfills and other sources when specified in an appli- 5.1 Calibration and Linearity Gases. cable subpart. Standard cylinder gas mixtures for each 1.2 Principle. A portion of the sample is compound of interest with at least three con- injected into a gas chromatograph (GC) and centration levels spanning the range of sus- pected sample concentrations. The calibra- the CO2, CH4, N2, and O2 concentrations are determined by using a thermal conductivity tion gases shall be prepared in helium. detector (TCD) and integrator. 5.2 Carrier Gas. Helium, high-purity.
2. Range and Sensitivity 6. Analysis 2.1 Range. The range of this method de- 6.1 Sample Collection. Use the sample col- pends upon the concentration of samples. lection procedures described in Methods 3 or The analytical range of TCD’s is generally 25C to collect a sample of landfill gas (LFG). between approximately 10 ppmv and the 6.2 Preparation of GC. Before putting the upper percent range. GC analyzer into routine operation, optimize 2.2 Sensitivity. The sensitivity limit for a the operational conditions according to the compound is defined as the minimum detect- manufacturer’s specifications to provide able concentration of that compound, or the good resolution and minimum analysis time. concentration that produces a signal-to- Establish the appropriate carrier gas flow noise ratio of three to one. For CO2, CH4, N2, and set the detector sample and reference and O2, the sensitivity limit is in the low cell flow rates at exactly the same levels. ppmv range. Adjust the column and detector tempera- tures to the recommended levels. Allow suf- 3. Interferences ficient time for temperature stabilization. This may typically require 1 hour for each Since the TCD exhibits universal response change in temperature. and detects all gas components except the carrier, interferences may occur. Choosing 6.3 Analyzer Linearity Check and Calibra- the appropriate GC or shifting the retention tion. Perform this test before sample analy- times by changing the column flow rate may sis. Using the gas mixtures in section 5.1, help to eliminate resolution interferences. verify the detector linearity over the range of suspected sample concentrations with at To assure consistent detector response, he- least three points per compound of interest. lium is used to prepare calibration gases. This initial check may also serve as the ini- Frequent exposure to samples or carrier gas tial instrument calibration. All subsequent containing oxygen may gradually destroy calibrations may be performed using a sin- filaments. gle-point standard gas provided the calibra- 4. Apparatus tion point is within 20 percent of the sample component concentration. For each instru- 4.1 Gas Chromatograph. GC having at ment calibration, record the carrier and de- least the following components: tector flow rates, detector filament and 4.1.1 Separation Column. Appropriate col- block temperatures, attenuation factor, in- umn(s) to resolve CO2, CH4, N2, O2, and other jection time, chart speed, sample loop vol- gas components that may be present in the ume, and component concentrations. Plot a sample. linear regression of the standard concentra- 4.1.2 Sample Loop. Teflon or stainless tions versus area values to obtain the re- steel tubing of the appropriate diameter. sponse factor of each compound. Alter- NOTE: Mention of trade names or specific natively, response factors of uncorrected products does not constitute endorsement or component concentrations (wet basis) may recommendation by the U. S. Environmental be generated using instrumental integration. Protection Agency. NOTE: Peak height may be used instead of 4.1.3 Conditioning System. To maintain peak area throughout this method. the column and sample loop at constant tem- 6.4 Sample Analysis. Purge the sample perature. loop with sample, and allow to come to at- 4.1.4 Thermal Conductivity Detector. mospheric pressure before each injection.
630
VerDate 11
Analyze each sample in duplicate, and cal- 7.2 Concentration of Sample Components. culate the average sample area (A). The re- Calculate C for each compound using Equa- sults are acceptable when the peak areas for tions 3C–1 and 3C–2. Use the temperature and two consecutive injections agree within 5 barometric pressure at the sampling site to percent of their average. If they do not calculate Bw. If the sample was diluted with agree, run additional samples until consist- helium using the procedures in Method 25C, ent area data are obtained. Determine the use Equation 3C–3 to calculate the con- tank sample concentrations according to centration. section 7.2. P 7. Calculations = w Bw 3C - 1 Carry out calculations retaining at least Pbar one extra decimal figure beyond that of the acquired data. Round off results only after A the final calculation. C = 3C - 2 7.1 Nomenclature. RB()1− A = average sample area w Bw = moisture content in the sample, frac- tion Ptf C = component concentration in the sample, T A dry basis, ppmv C= tf 3C - 3 Ct = calculated NMOC concentration, ppmv C Pt Pti − equivalent − RB()1 w Ctm = measured NMOC concentration, ppmv C T T equivalent t ti Pbar = barometric pressure, mm Hg 8. Bibliography Pti = gas sample tank pressure after evacu- ation, mm Hg absolute 1. McNair, H.M., and E.J. Bonnelli. Basic Pt = gas sample tank pressure after sam- Gas Chromatography. Consolidated Printers, pling, but before pressurizing, mm Hg ab- Berkeley, CA. 1969. solute Ptf = final gas sample tank pressure after METHOD 4—DETERMINATION OF MOISTURE pressurizing, mm Hg absolute CONTENT IN STACK GASES Pw = vapor pressure of H2O (from table 3C–1), mm Hg 1. Principle and Applicability Tti = sample tank temperature before sam- 1.1 Principle. A gas sample is extracted at pling, °K a constant rate from the source; moisture is Tt = sample tank temperature at completion removed from the sample stream and deter- of sampling, °K mined either volumetrically or gravimetri- Ttf = sample tank temperature after pressur- cally. izing, °K 1.2 Applicability. This method is applica- r = total number of analyzer injections of ble for determining the moisture content of sample tank during analysis (where j = stack gas. injection number, 1 . . . r) Two procedures are given. The first is a R = Mean calibration response factor for spe- reference method, for accurate determina- cific sample component, area/ppmv tions of moisture content (such as are needed to calculate emission data). The second is an TABLE 3C±1.ÐMOISTURE CORRECTION approximation method, which provides esti- mates of percent moisture to aid in setting Vapor Pres- isokinetic sampling rates prior to a pollut- Temperature °C sure of H2O, mm ant emission measurement run. The approxi- Hg mation method described herein is only a 4 ...... 6.1 suggested approach; alternative means for 6 ...... 7.0 approximating the moisture content, e.g., 8 ...... 8.0 drying tubes, wet bulb-dry bulb techniques, 10 ...... 9.2 condensation techniques, stoichiometric cal- 12 ...... 10.5 culations, previous experience, etc., are also 14 ...... 12.0 acceptable. 16 ...... 13.6 18 ...... 15.5 The reference method is often conducted 20 ...... 17.5 simultaneously with a pollutant emission 22 ...... 19.8 measurement run; when it is, calculation of 24 ...... 22.4 percent isokinetic, pollutant emission rate, 26 ...... 25.2 etc., for the run shall be based upon the re- 28 ...... 28.3 sults of the reference method or its equiva- 30 ...... 31.8 lent; these calculations shall not be based
631
VerDate 11
upon the results of the approximation meth- temperature. Next, determine the moisture od, unless the approximation method is percentage, either by: (1) using a shown, to the satisfaction of the Adminis- psychrometric chart and making appropriate trator, U.S. Environmental Protection Agen- corrections if stack pressure is different cy, to be capable of yielding results within 1 from that of the chart, or (2) using satura- percent H2O of the reference method. tion vapor pressure tables. In cases where the pyschrometric chart or the saturation NOTE: The reference method may yield vapor pressure tables are not applicable questionable results when applied to satu- (based on evaluation of the process), alter- rated gas streams or to streams that contain native methods, subject to the approval of water droplets. Therefore, when these condi- the Administrator, shall be used. tions exist or are suspected, a second deter- mination of the moisture content shall be 2. Reference Method made simultaneously with the reference The procedure described in Method 5 for method, as follows: Assume that the gas determining moisture content is acceptable stream is saturated. Attach a temperature as a reference method. sensor [capable of measuring to ±1° C (2° F)] 2.1 Apparatus. A schematic of the sam- to the reference method probe. Measure the pling train used in this reference method is stack gas temperature at each traverse point shown in Figure 4–1. All components shall be (see Section 2.2.1) during the reference meth- maintained and calibrated according to the od traverse; calculate the average stack gas procedure outlined in Method 5.
632
VerDate 11
2.1.1 Probe. The probe is constructed of When stack conditions permit, other met- stainless steel or glass tubing, sufficiently als or plastic tubing may be used for the heated to prevent water condensation, and is probe, subject to the approval of the Admin- equipped with a filter, either in-stack (e.g., a istrator. plug of glass wool inserted into the end of 2.1.2 Condenser. The condenser consists of the probe) or heated out-stack (e.g., as de- four impingers connected in series with scribed in Method 5), to remove particular ground glass, leak-free fittings or any simi- matter. larly leak-free non-contaminating fittings. The first, third, and fourth impingers shall
633
VerDate 11
be of the Greenburg-Smith design, modified metric reading may be obtained from a near- by replacing the tip with a 1.3 centimeter (1⁄2 by National Weather Service station, in inch) ID glass tube extending to about 1.3 cm which case the station value (which is the (1⁄2 in.) from the bottom of the flask. The sec- absolute barometric pressure) shall be re- ond impinger shall be of the Greenburg- quested and an adjustment for elevation dif- Smith design with the standard tip. Modi- ferences between the weather station and the fications (e.g., using flexible connections be- sampling point shall be applied at a rate of tween the impingers, using materials other minus 2.5 mm Hg (0.1 in. Hg) per 30 m (100 ft) than glass, or using flexible vacuum lines to elevation increase or vice versa for elevation connect the filter holder to the condenser) decrease. may be used, subject to the approval of the 2.1.6 Graduated Cylinder and/or Balance. Administrator. These items are used to measure condensed The first two impingers shall contain water and moisture caught in the silica gel known volumes of water, the third shall be empty, and the fourth shall contain a known to within 1 ml or 0.5 g. Graduated cylinders weight of 6- to 16-mesh indicating type silica shall have subdivisions no greater than 2 ml. gel, or equivalent desiccant. If the silica gel Most laboratory balances are capable of has been previously used, dry at 175° C (350° weighing to the nearest 0.5 g or less. These F) for 2 hours. New silica gel may be used as balances are suitable for use here. received. A thermometer, capable of measur- 2.2 Procedure. The following procedure is ing temperature to within 1° C (2° F), shall be written for a condenser system (such as the placed at the outlet of the fourth impinger, impinger system described in Section 2.1.2) for monitoring purposes. incorporating volumetric analysis to meas- Alternatively, any system may be used ure the condensed moisture, and silica gel (subject to the approval of the Adminis- and gravimetric analysis to measure the trator) that cools the sample gas stream and moisture leaving the condenser. allows measurement of both the water that 2.2.1 Unless otherwise specified by the Ad- has been condensed and the moisture leaving ministrator, a minimum of eight traverse the condenser, each to within 1 ml or 1 g. Ac- points shall be used for circular stacks hav- ceptable means are to measure the con- ing diameters less than 0.61 m (24 in.), a min- densed water, either gravimetrically or imum of nine points shall be used for rectan- volumetrically, and to measure the moisture gular stacks having equivalent diameters leaving the condenser by: (1) monitoring the less than 0.61 m (24 in.), and a minimum of temperature and pressure at the exit of the twelve traverse points shall be used in all condenser and using Dalton’s law of partial other cases. The traverse points shall be lo- pressures, or (2) passing the sample gas cated according to Method 1. The use of stream through a tared silica gel (or equiva- fewer points is subject to the approval of the lent desiccant) trap, with exit gases kept Administrator. Select a suitable probe and ° ° below 20 C (68 F), and determining the probe length such that all traverse points weight gain. can be sampled. Consider sampling from op- If means other than silica gel are used to posite sides of the stack (four total sampling determine the amount of moisture leaving ports) for large stacks, to permit use of the condenser, it is recommended that silica shorter probe lengths. Mark the probe with gel (or equivalent) still be used between the heat resistant tape or by some other method condenser system and pump, to prevent to denote the proper distance into the stack moisture condensation in the pump and me- or duct for each sampling point. Place tering devices and to avoid the need to make known volumes of water in the first two corrections for moisture in the metered vol- impingers. Weigh and record the weight of ume. the silica gel to the nearest 0.5 g, and trans- 2.1.3 Cooling System. An ice bath con- tainer and crushed ice (or equivalent) are fer the silica gel to the fourth impinger; al- used to aid in condensing moisture. ternatively, the silica gel may first be trans- 2.1.4 Metering System. This system in- ferred to the impinger, and the weight of the cludes a vacuum gauge, leak-free pump, silica gel plus impinger recorded. thermometers capable of measuring tem- 2.2.2 Select a total sampling time such perature to within 3° C (5.4° F), dry gas meter that a minimum total gas volume of 0.60 scm capable of measuring volume to within 2 per- (21 scf) will be collected, at a rate no greater cent, and related equipment as shown in Fig- than 0.021 m3/min (0.75 cfm). When both mois- ure 4–1. Other metering systems, capable of ture content and pollutant emission rate are maintaining a constant sampling rate and to be determined, the moisture determina- determining sample gas volume, may be tion shall be simultaneous with, and for the used, subject to the approval of the Adminis- same total length of time as, the pollutant trator. emission rate run, unless otherwise specified 2.1.5 Barometer. Mercury, aneroid, or in an applicable subpart of the standards. other barometer capable of measuring at- 2.2.3 Set up the sampling train as shown mospheric pressure to within 2.5 mm Hg (0.1 in Figure 4–1. Turn on the probe heater and in. Hg) may be used. In many cases, the baro- (if applicable) the filter heating system to
634
VerDate 11
temperatures of about 120° C (248° F), to pre- the desired rate. Traverse the cross section, vent water condensation ahead of the con- sampling at each traverse point for an equal denser; allow time for the temperatures to length of time. Add more ice and, if nec- stabilize. Place crushed ice in the ice bath essary, salt to maintain a temperature of container. It is recommended, but not re- less 20° C (68° F) at the silica gel outlet. quired, that a leak check be done, as follows: 2.2.6 After collecting the sample, dis- Disconnect the probe from the first impinger connect the probe from the filter holder (or or (if applicable) from the filter holder. Plug from the first impinger) and conduct a leak the inlet to the first impinger (or filter hold- check (mandatory) as described in Section er) and pull a 380 mm (15 in.) Hg vacuum; a 2.2.3. Record the leak rate. If the leakage lower vacuum may be used, provided that it rate exceeds the allowable rate, the tester is not exceeded during the test. A leakage shall either reject the test results or shall rate in excess of 4 percent of the average correct the sample volume as in Section 6.3 sampling rate or 0.00057 m3/min (0.02 cfm), of Method 5. Next, measure the volume of whichever is less, is unacceptable. Following the moisture condensed to the nearest ml. the leak check, reconnect the probe to the Determine the increase in weight of the sili- sampling train. 2.2.4 During the sampling run, maintain a ca gel (or silica gel plus impinger) to the sampling rate within 10 percent of constant nearest 0.5 g. Record this information (see rate, or as specified by the Administrator. example data sheet, Figure 4–3) and calculate For each run, record the data required on the the moisture percentage, as described in 2.3 example data sheet shown in Figure 4–2. Be below. sure to record the dry gas meter reading at 2.2.7 A quality control check of the vol- the beginning and end of each sampling time ume metering system at the field site is sug- increment and whenever sampling is halted. gested before collecting the sample following Take other appropriate readings at each the procedure in Method 5, Section 4.4 sample point, at least once during each time 2.3 Calculations. Carry out the following increment. calculations, retaining at least one extra 2.2.5 To begin sampling, position the decimal figure beyond that of the acquired probe tip at the first traverse point. Imme- data. Round off figures after final calcula- diately start the pump and adjust the flow to tion.
635
VerDate 11
636
VerDate 11
FIGURE 4±3ÐANALYTICAL DATAÐREFERENCE Y =Dry gas meter calibration factor. ρ METHOD w=Density of water, 0.9982 g/ml (0.002201 lb/ ml). Impinger vol- Silica gel 2.3.2 Volume of Water Vapor Condensed. ume, ml weight, g
Final ...... − ρ ()Vf Vi w RTstd Initial ...... = Difference ...... Vwc(std) PMstd w 2.3.1 Nomenclature. =() − B ws=Proportion of water vapor, by volume, KVV1 f i Eq. 4-1 in the gas stream. Where: M w=Molecular weight of water, 18.0 g/g-mole 3 (18.0 lb/lb-mole). K1=0.001333 m /ml for metric units 3 P m=Absolute pressure (for this method, same =0.04707 ft /ml for English units as barometric pressure) at the dry gas 2.3.3 Volume of Water Vapor Collected in meter, mm Hg (in. Hg). Silica Gel. P std=Standard absolute pressure, 760 mm Hg (29.92 in. Hg). − R=Ideal gas constant, 0.06236 (mm Hg) (m3)/ ()Wf Wi RTstd (g-mole) (°K) for metric units and 21.85 Vwsg() std = (in. Hg) (ft3)/(lb-mole) (°R) for English PMstd w units. T =Absolute temperature at meter, °K (°R). − m = K2()Wf Wi Eq. 4 -2 T std=Standard absolute temperature, 293°K (528°R). Where: 3 V m=Dry gas volume measured by dry gas K 2=0.001335 m /g for metric units meter, dcm (dcf). =0.04715 ft3/g for English units ∆ V m=Incremental dry gas volume measured 2.3.4 Sample Gas Volume. by dry gas meter at each traverse point, dcm (dcf). ()P () T V m(std)=Dry gas volume measured by the dry V = V Y m std gas meter, corrected to standard condi- m() std m () () tions, dscm (dscf). Pstd Tm V wc(std)=Volume of water vapor condensed VP corrected to standard conditions, scm = K Y m m Eq. 4 - 3 (scf). 3 T V wsg(std)=Volume of water vapor collected in m silica gel corrected to standard condi- Where: tions, scm (scf). K 3=0.3858 °K/mm Hg for metric units V f=Final volume of condenser water, ml. =17.64 °R/in. Hg for English units V i=Initial volume, if any, of condenser water, ml. NOTE: If the post-test leak rate (Section 2.2.6) exceeds the allowable rate, correct the W f=Final weight of silica gel or silica gel plus impinger, g. value of V m in Equation 4–3, as described in Section 6.3 of Method 5. W i=Initial weight of silica gel or silica gel plus impinger, g. 2.3.5 Moisture Content.
+ VVwc()() std wsg std B = Eq. 4 - 4 ws + + VVVwc()()() std wsg std m std
NOTE: In saturated or moisture droplet- the ∆V m. Calculate the average. If the value laden gas streams, two calculations of the for any time increment differs from the aver- moisture content of the stack gas shall be age by more than 10 percent, reject the re- made, one using a value based upon the satu- sults and repeat the run. rated conditions (see Section 1.2), and an- 3. Approximation Method other based upon the results of the impinger The approximation method described analysis. The lower of these two values of below is presented only as a suggested meth- shall be considered correct. B ws od (see Section 1.2). 2.3.6 Verification of Constant Sampling 3.1 Apparatus. Rate. For each time increment, determine
637
VerDate 11
3.1.1 Probe. Stainless steel glass tubing, 3.1.8 Rate Meter. Rotameter, to measure sufficiently heated to prevent water con- the flow range from 0 to 3 lpm (0 to 0.11 cfm). densation and equipped with a filter (either 3.1.9 Graduated Cylinder. 25 ml. in-stack or heated out-stack) to remove par- 3.1.10 Barometer. Mercury, aneroid, or ticulate matter. A plug of glass wool, in- other barometer, as described in Section 2.1.5 serted into the end of the probe, is a satisfac- above. tory filter. 3.1.11 Vacuum Gauge. At least 760 mm Hg 3.1.2 Impingers. Two midget impingers, (30 in. Hg) gauge, to be used for the sampling each with 30 ml capacity, or equivalent. leak check. 3.1.3 Ice Bath. Container and ice, to aid in 3.2 Procedure. condensing moisture in impingers. 3.2.1 Place exactly 5 ml distilled water in 3.1.4 Drying Tube. Tube packed with new each impinger. or regenerated 6- to 16-mesh indicating-type silica gel (or equivalent desiccant), to dry Leak check the sampling train as follows: the sample gas and to protect the meter and Temporarily insert a vacuum gauge at or pump. near the probe inlet; then, plug the probe 3.1.5 Valve. Needle valve, to regulate the inlet and pull a vacuum of at least 250 mm sample gas flow rate. Hg (10 in. Hg). Note the time rate of change 3.1.6 Pump. Leak-free, diaphragm type, or of the dry gas meter dial; alternatively, a ro- equivalent, to pull the gas sample through tameter (0–40 cc/min) may be temporarily at- the train. tached to the dry gas meter outlet to deter- 3.1.7 Volume Meter. Dry gas meter, suffi- mine the leakage rate. A leak rate not in ex- ciently accurate to measure the sample vol- cess of 2 percent of the average sampling ume within 2%, and calibrated over the rate is acceptable. range of flow rates and conditions actually NOTE: Carefully release the probe inlet encountered during sampling. plug before turning off the pump.
638
VerDate 11
639
VerDate 11
FIGURE 4–5—FIELD MOISTURE Vm=Dry gas volume measured by dry gas DETERMINATION—APPROXIMATION METHOD meter, dcm (dcf).
Vm(std)=Dry gas volume measured by dry gas FIGURE 4±5ÐFIELD MOISTURE meter, corrected to standard conditions, DETERMINATIONÐAPPROXIMATION METHOD dscm (dscf).
Location ...... Comments: Vwc(std)=Volume of water vapor condensed, Test ...... corrected to standard conditions, scm Date ...... (scf). Operator ...... ρ =Density of water, 0.9982 g/ml (0.002201 lb/ Barometric pressure ...... w ml). Gas volume Rate meter Meter tem- Y=Dry gas meter calibration factor. Clock time through meter, setting m3/ perature, ° C 3 3 3 3.3.2 Volume of Water Vapor Collected. (Vm), m (ft ) min (ft /min) (° F) − ρ ()Vf Vi w RTstd Vwc = PMstd w − = K1() V f Vi Eq. 4 - 5 Where: 3 K1=0.001333 m /ml for metric units =0.04707 ft3/ml for English units. 3.3.3 Gas Volume. 3.2.2 Connect the probe, insert it into the stack, and sample at a constant rate of 2 lpm (0.071 cfm). Continue sampling until the dry gas meter registers about 30 liters (1.1 ft3) or until visible liquid droplets are carried over from the first impinger to the second. Record temperature, pressure, and dry gas meter readings as required by Figure 4–5. 3.2.3 After collecting the sample, combine where: ° the contents of the two impingers and meas- K2=0.3858 K/mm Hg for metric units ure the volume to the nearest 0.5 ml. =17.64 °R/in. Hg for English units 3.3 Calculations. The calculation method 3.3.4 Approximate Moisture Content. presented is designed to estimate the mois- ture in the stack gas; therefore, other data, which are only necessary for accurate mois- ture determinations, are not collected. The following equations adequately estimate the moisture content, for the purpose of deter- mining isokinetic sampling rate settings. 3.3.1 Nomenclature. 4. Calibration Bwm=Approximate proportion, by volume, of water vapor in the gas stream leaving 4.1 For the reference method, calibrate the second impinger, 0.025. equipment as specified in the following sec- Bws=Water vapor in the gas stream, propor- tions of Method 5: Section 5.3 (metering sys- tion by volume. tem); Section 5.5 (temperature gauges); and Mw=Molecular weight of water, 18.0 g/g-mole Section 5.7 (barometer). The recommended (18.0 lb/lb-mole). leak check of the metering system (Section Pm=Absolute pressure (for this method, same 5.6 of Method 5) also applies to the reference as barometric pressure) at the dry gas method. For the approximation method, use meter. the procedures outlined in Section 5.1.1 of Pstd=Standard absolute pressure, 760 mm Hg Method 6 to calibrate the metering system, (29.92 in. Hg). and the procedure of Method 5, Section 5.7 to R=Ideal gas constant, 0.06236 (mm Hg) (m3)/ calibrate the barometer. (g-mole) (°K) for metric units and 21.85 5. Bibliography (in. Hg) (ft3)/lb-mole) (°R) for English units. 1. Air Pollution Engineering Manual (Sec- Tm=Absolute temperature at meter, °K (°R). ond Edition). Danielson, J. A. (ed.). U.S. En- Tstd=Standard absolute temperature, 293°K vironmental Protection Agency, Office of Air (528° R). Quality Planning and Standards. Research Vf=Final volume of impinger contents, ml. Triangle Park, NC. Publication No. AP–40. Vi=Initial volume of impinger contents, ml. 1973.
640
VerDate 11
2. Devorkin, Howard. et al. Air Pollution termined gravimetrically after removal of Source Testing Manual. Air Pollution Con- uncombined water. trol District, Los Angeles, CA. November, 1.2 Applicability. This method is applica- 1963. ble for the determination of particulate 3. Methods for Determination of Velocity, emissions from stationary sources. Volume, Dust and Mist Content of Gases. 2. Apparatus Western Precipitation Division of Joy Manu- 2.1 Sampling Train. A schematic of the facturing Co., Los Angeles, CA. Bulletin WP– sampling train used in this method is shown 50. 1968. in Figure 5–1. Complete construction details METHOD 5—DETERMINATION OF PARTICULATE are given in APTD–0581 (Citation 2 in Bibli- EMISSIONS FROM STATIONARY SOURCES ography); commercial models of this train are also available. For changes from APTD– 1. Principle and Applicability 0581 and for allowable modifications of the 1.1 Principle. Particulate matter is with- train shown in Figure 5–1, see the following drawn isokinetically from the source and subsections. collected on a glass fiber filter maintained at The operating and maintenance procedures a temperature in the range of 120±14° C for the sampling train are described in (248±25° F) or such other temperature as APTD–0576 (Citation 3 in Bibliography). specified by an applicable subpart of the Since correct usage is important in obtain- standards or approved by Administrator, ing valid results, all users should read U.S. Environmental Protection Agency, for a APTD–0576 and adopt the operating and particular application. The particulate mass, maintenance procedures outlined in it, un- which includes any material that condenses less otherwise specified herein. The sampling at or above the filtration temperature, is de- train consists of the following components:
641
VerDate 11
2.1.1 Probe Nozzle. Stainless steel (316) or stant internal diameter. The probe nozzle glass with sharp, tapered leading edge. The shall be of the button-hook or elbow design, angle of taper shall be ≤30° and the taper unless otherwise specified by the Adminis- shall be on the outside to preserve a con- trator. If made of stainless steel, the nozzle
642
VerDate 11
shall be constructed from seamless tubing; 2.1.5 Filter Holder. Borosilicate glass, other materials of construction may be used, with a glass frit filter support and a silicone subject to the approval of the Administrator. rubber gasket. Other materials of construc- A range of nozzle sizes suitable for tion (e.g., stainless steel, Teflon, Viton) may isokinetic sampling should be available, e.g., be used, subject to approval of the Adminis- 0.32 to 1.27 cm (1⁄8 to 1⁄2 in.)—or larger if high- trator. The holder design shall provide a er volume sampling trains are used—inside positive seal against leakage from the out- diameter (ID) nozzles in increments of 0.16 side or around the filter. The holder shall be cm (1⁄16 in.). Each nozzle shall be calibrated attached immediately at the outlet of the according to the procedures outlined in Sec- probe (or cyclone, it used). tion 5. 2.1.6 Filter Heating System. Any heating 2.1.2 Probe Liner. Borosilicate or quartz system capable of maintaining a tempera- glass tubing with a heating system capable ture around the filter holder during sampling of maintaining a gas temperature at the exit of 120±14° C (248±25° F), or such other tem- end during sampling of 120±14° C (248±25° F), perature as specified by an applicable sub- or such other temperature as specified by an part of the standards or approved by the Ad- applicable subpart of the standards or ap- ministrator for a particular application. Al- proved by the Administrator for a particular ternatively, the tester may opt to operate application. (The tester may opt to operate the equipment at a temperature lower than the equipment at a temperature lower than that specified. A temperature gauge capable that specified.) Since the actual temperature of measuring temperature to within 3° C (5.4° at the outlet of the probe is not usually mon- F) shall be installed so that the temperature itored during sampling, probes constructed around the filter holder can be regulated and according to APTD–0581 and utilizing the monitored during sampling. Heating systems calibration curves of APTD–0576 (or cali- other than the one shown in APTD–0581 may brated according to the procedure outlined be used. in APTD–0576) will be considered acceptable. 2.1.7 Condenser. The following system Either borosilicate or quartz glass probe shall be used to determine the stack gas liners may be used for stack temperatures up moisture content: Four impingers connected to about 480° C (900° F); quartz liners shall be in series with leak-free ground glass fittings used for temperatures between 480 and 900° C or any similar leak-free non-contaminating (900 and 1,650° F). Both types of liners may be fittings. The first, third, and fourth used at higher temperatures than specified impingers shall be of the Greenburg-Smith for short periods of time, subject to the ap- design, modified by replacing the tip with 1.3 proval of the Administrator. The softening cm (1⁄2 in.) ID glass tube extending to about ° ° temperature for borosilicate is 820 C (1,508 1.3 cm (1⁄2 in.) from the bottom of the flask. F), and for quartz it is 1,500° C (2,732° F). The second impinger shall be of the Whenever practical, every effort should be Greenburg-Smith design with the standard made to use borosilicate or quartz glass tip. Modifications (e.g., using flexible con- probe liners. Alternatively, metal liners nections between the impingers, using mate- (e.g., 316 stainless steel, Incoloy 825,2or other rials other than glass, or using flexible vacu- corrosion resistant metals) made of seamless um lines to connect the filter holder to the tubing may be used, subject to the approval condenser) may be used, subject to the ap- of the Administrator. proval of the Administrator. The first and 2.1.3 Pitot Tube. Type S, as described in second impingers shall contain known quan- Section 2.1 of Method 2, or other device ap- tities of water (Section 4.1.3), the third shall proved by the Administrator. The pitot tube be empty, and the fourth shall contain a shall be attached to the probe (as shown in known weight of silica gel, or equivalent des- Figure 5–1) to allow constant monitoring of iccant. A thermometer, capable of measuring the stack gas velocity. The impact (high temperature to within 1° C (2° F) shall be pressure) opening plane of the pitot tube placed at the outlet of the fourth impinger shall be even with or above the nozzle entry for monitoring purposes. plane (see Method 2, Figure 2–6b) during sam- Alternatively, any system that cools the pling. The Type S pitot tube assembly shall sample gas stream and allows measurement have a known coefficient, determined as out- of the water condensed and moisture leaving lined in Section 4 of Method 2. the condenser, each to within 1 ml or 1 g 2.1.4 Differential Pressure Gauge. Inclined may be used, subject to the approval of the manometer or equivalent device (two), as de- Administrator. Acceptable means are to scribed in Section 2.2 of Method 2. One ma- measure the condensed water either gravi- ∆β nometer shall be used or velocity head ( ) metrically or volumetrically and to measure readings, and the other, for orifice differen- the moisture leaving the condenser by: (1) tial pressure readings. monitoring the temperature and pressure at the exit of the condenser and using Dalton’s 2 Mention of trade names or specific prod- law of partial pressures; or (2) passing the uct does not constitute endorsement by the sample has stream through a tared silica gel Environmental Protection Agency. (or equivalent desiccant) trap with exit gases
643
VerDate 11
kept below 20° C (68° F) and determining the ings (see Method 2, Figure 2–7). As a second weight gain. alternative, if a difference of not more than If means other than silica gel are used to 1 percent in the average velocity measure- determine the amount of moisture leaving ment is to be introduced, the temperature the condenser, it is recommended that silica gauge need not be attached to the probe or gel (or equivalent) still be used between the pitot tube. (This alternative is subject to the condenser system and pump to prevent mois- approval of the Administrator.) ture condensation in the pump and metering 2.2 Sample Recovery. The following items devices and to avoid the need to make cor- are needed. rections for moisture in the metered volume. 2.2.1 Probe-Liner and Probe-Nozzle Brush- NOTE: If a determination of the particulate es. Nylon bristle brushes with stainless steel matter collected in the impingers is desired wire handles. The probe brush shall have ex- in addition to moisture content, the im- tensions (at least as long as the probe) of pinger system described above shall be used, stainless steel, Nylon, Teflon, or similarly without modification. Individual States or inert material. The brushes shall be properly control agencies requiring this information sized and shaped to brush out the probe liner shall be contacted as to the sample recovery and nozzle. and analysis of the impinger contents. 2.2.2 Wash Bottles—Two. Glass wash bot- 2.1.8 Metering System. Vacuum gauge, tles are recommended; polyethylene wash leak-free pump, thermometers capable of bottles may be used at the option of the test- measuring temperature to within 3° C (5.4° er. It is recommended that acetone not be F), dry gas meter capable of measuring vol- stored in polyethylene bottles for longer ume to within 2 percent, and related equip- than a month. ment, as shown in Figure 5–1. Other metering 2.2.3 Glass Sample Storage Containers. systems capable of maintaining sampling Chemically resistant, borosilicate glass bot- rates within 10 percent of isokinetic and of tles, for acetone washes, 500 ml or 1000 ml. determining sample volumes to within 2 per- Screw cap liners shall either be rubber- cent may be used, subject to the approval of backed Teflon or shall be constructed so as the Administrator. When the metering sys- to be leak-free and resistant to chemical at- tem is used in conjunction with a pitot tube, tack by acetone. (Narrow mouth glass bot- the system shall enable checks of isokinetic tles have been found to be less prone to leak- rates. age.) Alternatively, polyethylene bottles Sampling trains utilizing metering sys- may be used. tems designed for higher flow rates than that 2.2.4 Petri Dishes. For filter samples, decribed in APTD–0581 or APDT–0576 may be glass or polyethylene, unless otherwise spec- used provided that the specifications of this ified by the Administrator. method are met. 2.2.5 Graduated Cylinder and/or Balance. 2.1.9 Barometer. Mercury aneroid, or To measure condensed water to within 1 ml other barometer capable of measuring at- or 1 g. Graduated cylinders shall have sub- mospheric pressure to within 2.5 mm Hg (0.1 divisions no greater than 2 ml. Most labora- in. Hg). In many cases the barometric read- tory balances are capable of weighing to the ing may be obtained from a nearby National nearest 0.5 g or less. Any of these balances is Weather Service station, in which case the suitable or use here and in Section 2.3.4. station value (which is the absolute baro- 2.2.6 Plastic Storage Containers. Air-tight metric pressure) shall be requested and an containers to store silica gel. adjustment for elevation differences between 2.2.7 Funnel and Rubber Policeman. To the weather station and sampling point shall aid in transfer of silica gel to container; not be applied at a rate of minus 2.5 mm Hg (0.1 necessary if silica gel is weighed in the field. in. Hg) per 30 m (100 ft) elevation increase or 2.2.8 Funnel. Glass or polyethylene, to aid vice versa for elevation decrease. in sample recovery. 2.1.10 Gas Density Determination Equip- 2.3 Analysis. For analysis, the following ment. Temperature sensor and pressure equipment is needed. gauge, as described in Sections 2.3 and 2.4 of 2.3.1 Glass Weighing Dishes. Method 2, and gas analyzer, if necessary, as 2.3.2 Desiccator. described in Method 3. The temperature sen- 2.3.3 Analytical Balance. To measure to sor shall, preferably, be permanently at- within 0.1 mg. tached to the pitot tube or sampling probe in 2.3.4 Balance. To measure to within 0.5 g. a fixed configuration, such that the tip of 2.3.5 Beakers. 250 ml. the sensor extends beyond the leading edge 2.3.6 Hygrometer. To measure the relative of the probe sheath and does not touch any humidity of the laboratory environment. metal. Alternatively, the sensor may be at- 2.3.7 Temperature Gauge. To measure the tached just prior to use in the field. Note, temperature of the laboratory environment. however, that if the temperature sensor is attached in the field, the sensor must be 3. Reagents placed in an interference-free arrangement 3.1 Sampling. The reagents used in sam- with respect to the Type S pitot tube open- pling are as follows:
644
VerDate 11
3.1.1 Filters. Glass fiber filters, without preweighed, but may be weighed directly in organic binder, exhibiting at least 99.95 per- the impinger or sampling holder just prior to cent efficiency (<0.05 percent penetration) on train assembly. 0.3-micron dioctyl phthalate smoke par- Check filters visually against light for ticles. The filter efficiency test shall be con- irregularities and flaws or pinhole leaks. ducted in accordance with ASTM Standard Label filters of the proper diameter on the Method D2986–71 (Reapproved 1978) (incor- back side near the edge using numbering ma- porated by reference—see § 60.17). Test data chine ink. As an alternative, label the ship- from the supplier’s quality control program ping containers (glass or plastic petri dishes) are sufficient for this purpose. In sources and keep the filters in these containers at all containing SO2 or SO3, the filter material times except during sampling and weighing. ± ° ± ° must be of a type that is unreactive to SO2 Desiccate the filters at 20 5.6 C (68 10 F) or SO3. Citation 10 in Bibliography, may be and ambient pressure for at least 24 hours used to select the appropriate filter. and weigh at intervals of at least 6 hours to 3.1.2 Silica Gel. Indicating type, 6 to 16 a constant weight, i.e., 0.5 mg change from mesh. If previously used, dry at 175° C (350° previous weighing; record results to the F) for 2 hours. New silica gel may be used as nearest 0.1 mg. During each weighing the fil- received. Alternatively, other types of ter must not be exposed to the laboratory at- desiccants (equivalent or better) may be mosphere for a period greater than 2 minutes used, subject to the approval of the Adminis- and a relative humidity above 50 percent. Al- trator. ternatively (unless otherwise specified by 3.1.3 Water. When analysis of the material the Administrator), the filters may be oven caught in the impingers is required, deion- dried at 105° C (220° F) for 2 to 3 hours, des- ized distilled water shall be used. Run blanks iccated for 2 hours, and weighed. Procedures prior to field use to eliminate a high blank other than those described, which account on test samples. for relative humidity effects, may be used, 3.1.4 Crushed Ice. subject to the approval of the Administrator. 3.1.5 Stopcock Grease. Acetone-insoluble, 4.1.2 Preliminary Determinations. Select heat-stable silicone grease. This is not nec- the sampling site and the minimum number essary if screw-on connectors with Teflon of sampling points according to Method 1 or sleeves, or similar, are used. Alternatively, as specified by the Administrator. Determine other types of stopcock grease may be used, the stack pressure, temperature, and the subject to the approval of the Administrator. range of velocity heads using Method 2; it is 3.2 Sample Recovery. Acetone-reagent recommended that a leak-check of the pitot grade, ≤0.001 percent residue, in glass bot- lines (see Method 2, Section 3.1) be per- tles—is required. Acetone from metal con- formed. Determine the moisture content tainers generally has a high residue blank using Approximation Method 4 or its alter- and should not be used. Sometimes, suppliers natives for the purpose of making isokinetic sampling rate settings. Determine the stack transfer acetone to glass bottles from metal gas dry molecular weight, as described in containers; thus, acetone blanks shall be run Method 2, Section 3.6; if integrated Method 3 prior to field use and only acetone with low sampling is used for molecular weight deter- blank values (≤0.001 percent) shall be used. In mination, the integrated bag sample shall be no case shall a blank value of greater than taken simultaneously with, and for the same 0.001 percent of the weight of acetone used be total length of time as, the particulate sam- subtracted from the sample weight. ple run. 3.3 Analysis. Two reagents are required Select a nozzle size based on the range of for the analysis: velocity heads, such that it is not necessary 3.3.1 Acetone. Same as 3.2. to change the nozzle size in order to main- 3.3.2 Desiccant. Anhydrous calcium sul- tain isokinetic sampling rates. During the fate, indicating type. Alternatively, other run, do not change the nozzle size. Ensure types of desiccants may be used, subject to that the proper differential pressure gauge is the approval of the Administrator. chosen for the range of velocity heads en- 4. Procedure countered (see Section 2.2 of Method 2). 4.1 Sampling. The complexity of this Select a suitable probe liner and probe method is such that, in order to obtain reli- length such that all traverse points can be able results, testers should be trained and sampled. For large stacks, consider sampling experienced with the test procedures. from opposite sides of the stack to reduce 4.1.1 Pretest Preparation. It is suggested the length of probes. that sampling equipment be maintained ac- Select a total sampling time greater than cording to the procedure described in APTD– or equal to the minimum total sampling 0576. time specified in the test procedures for the Weigh several 200 to 300 g portions of silica specific industry such that (1) the sampling gel in air-tight containers to the nearest 0.5 time per point is not less than 2 min (or g. Record the total weight of the silica gel some greater time interval as specified by plus container, on each container. As an al- the Administrator), and (2) the sample vol- ternative, the silica gel need not be ume taken (corrected to standard conditions)
645
VerDate 11
will exceed the required minimum total gas 4.1.4.1 Pretest Leak-Check. A pretest sample volume. The latter is based on an ap- leak-check is recommended, but not re- proximate average sampling rate. quired. If the tester opts to conduct the pre- It is recommended that the number of min- test leak-check, the following procedure utes sampled at each point be an integer or shall be used. an integer plus one-half minute, in order to After the sampling train has been assem- avoid timekeeping errors. The sampling time bled, turn on and set the filter and probe at each point shall be the same. heating systems at the desired operating In some circumstances, e.g., batch cycles, temperatures. Allow time for the tempera- it may be necessary to sample for shorter tures to stabilize. If a Viton A O-ring or times at the traverse points and to obtain other leak-free connection is used in assem- smaller gas sample volumes. In these cases, bling the probe nozzle to the probe liner, the Administrator’s approval must first be leak-check the train at the sampling site by obtained. plugging the nozzle and pulling a 380 mm Hg 4.1.3 Preparation of Collection Train. Dur- (15 in. Hg) vacuum. ing preparation and assembly of the sam- NOTE: A lower vacuum may be used, pro- pling train, keep all openings where con- vided that it is not exceeded during the test. tamination can occur covered until just If an asbestos string is used, do not con- prior to assembly or until sampling is about nect the probe to the train during the leak- to begin. check. Instead, leak-check the train by first Place 100 ml of water in each of the first plugging the inlet to the filter holder (cy- two impingers, leave the third impinger clone, if applicable) and pulling a 380 mm Hg empty, and transfer approximately 200 to 300 (15 in. Hg) vacuum (see Note immediately g of preweighed silica gel from its container above). Then connect the probe to the train to the fourth impinger. More silica gel may and leak-check at about 25 mm Hg (1 in. Hg) be used, but care should be taken to ensure vacuum; alternatively, the probe may be that it is not entrained and carried out from leak-checked with the rest of the sampling the impinger during sampling. Place the con- train, in one step, at 380 mm Hg (15 in. Hg) tainer in a clean place for later use in the vacuum. Leakage rates in excess of 4 percent 3 sample recovery. Alternatively, the weight of the average sampling rate or 0.00057 m / of the silica gel plus impinger may be deter- min (0.02 cfm), whichever is less, are unac- mined to the nearest 0.5 g and recorded. ceptable. Using a tweezer or clean disposable sur- The following leak-check instructions for the sampling train described in APTD–0576 gical gloves, place a labeled (identified) and and APTD–0581 may be helpful. Start the weighed filter in the filter holder. Be sure pump with bypass valve fully open and that the filter is properly centered and the coarse adjust valve, completely closed. Par- gasket properly placed so as to prevent the tially open the coarse adjust valve and slow- sample gas stream from circumventing the ly close the bypass valve until the desired filter. Check the filter for tears after assem- vacuum is reached. Do not reverse direction bly is completed. of bypass valve; this will cause water to back When glass liners are used, install the se- up into the filter holder. If the desired vacu- lected nozzle using a Viton A O-ring when um is exceeded, either leak-check at this ° ° stack temperatures are less than 260 C (500 higher vacuum or end the leak-check as F) and an asbestos string gasket when tem- shown below and start over. peratures are higher. See APTD–0576 for de- When the leak-check is completed, first tails. Other connecting systems using either slowly remove the plug from the inlet to the 316 stainless steel or Teflon ferrules may be probe, filter holder, or cyclone (if applicable) used. When metal liners are used, install the and immediately turn off the vacuum pump. nozzle as above or by a leak-free direct me- This prevents the water in the impingers chanical connection. Mark the probe with from being forced backward into the filter heat resistant tape or by some other method holder and silica gel from being entrained to denote the proper distance into the stack backward into the third impinger. or duct for each sampling point. 4.1.4.2 Leak-Checks During Sample Run. Set up the train as in Figure 5–1, using (if If, during the sampling run, a component necessary) a very light coat of silicone (e.g., filter assembly or impinger) change be- grease on all ground glass joints, greasing comes necessary, a leak-check shall be con- only the outer portion (see APTD–0576) to ducted immediately before the change is avoid possibility of contamination by the sil- made. The leak-check shall be done accord- icone grease. Subject to the approval of the ing to the procedure outlined in Section Administrator, a glass cyclone may be used 4.1.4.1 above, except that it shall be done at between the probe and filter holder when the a vacuum equal to or greater than the maxi- total particulate catch is expected to exceed mum value recorded up to that point in the 100 mg or when water droplets are present in test. If the leakage rate is found to be no the stack gas. greater than 0.00057 m3/min (0.02 cfm) or 4 Place crushed ice around the impingers. percent of the average sampling rate (which- 4.1.4 Leak-Check Procedures. ever is less), the results are acceptable, and
646
VerDate 11
no correction will need to be applied to the 5–2. Be sure to record the initial dry gas total volume of dry gas metered; if, however, meter reading. Record the dry gas meter a higher leakage rate is obtained, the tester readings at the beginning and end of each shall either record the leakage rate and plan sampling time increment, when changes in to correct the sample volume as shown in flow rates are made, before and after each Section 6.3 of this method, or shall void the leak-check, and when sampling is halted. sampling run. Take other readings required by Figure 5–2 Immediately after component changes, at least once at each sample point during leak-checks are optional; if such leak-checks each time increment and additional readings are done, the procedure outlined in Section when significant changes (20 percent vari- 4.1.4.1 above shall be used. ation in velocity head readings) necessitate 4.1.4.3 Post-test Leak-Check. A leak- additional adjustments in flow rate. Level check is mandatory at the conclusion of each and zero the manometer. Because the ma- sampling run. The leak-check shall be done nometer level and zero may drift due to vi- in accordance with the procedures outlined brations and temperature changes, make in Section 4.1.4.1, except that it shall be con- periodic checks during the traverse. ducted at a vacuum equal to or greater than Clean the portholes prior to the test run to the maximum value reached during the sam- minimize the chance of sampling deposited pling run. If the leakage rate is found to be material. To begin sampling, remove the no greater than 0.00057 m3/min (0.02 cfm) or 4 percent of the average sampling rate (which- nozzle cap, verify that the filter and probe ever is less), the results are acceptable, and heating systems are up to temperature, and no correction need be applied to the total that the pitot tube and probe are properly volume of dry gas metered. If, however, a positioned. Position the nozzle at the first higher leakage rate is obtained, the tester traverse point with the tip pointing directly shall either record the leakage rate and cor- into the gas stream. Immediately start the rect the sample volume as shown in Section pump and adjust the flow to isokinetic con- 6.3 of this method, or shall void the sampling ditions. Nomographs are available, which aid run. in the rapid adjustment of the isokinetic 4.1.5 Particulate Train Operation. During sampling rate without excessive computa- the sampling run, maintain an isokinetic tions. These nomographs are designed for use sampling rate (within 10 percent of true when the Type S pitot tube coefficient is isokinetic unless otherwise specified by the 0.85±0.02, and the stack gas equivalent den- Administrator) and a temperature around sity (dry molecular weight) is equal to 29±4. the filter of 120±14° C (248±25° F), or such APTD–0576 details the procedure for using other temperature as specified by an applica- the nomographs. If Cp and Md are outside the ble subpart of the standards or approved by above stated ranges do not use the the Administrator. nomographs unless appropriate steps (see Ci- For each run, record the data required on tation 7 in Bibliography) are taken to com- a data sheet such as the one shown in Figure pensate for the deviations.
647
VerDate 11
648
VerDate 11
When the stack is under significant nega- the lines must pass this leak-check, in order tive pressure (height of impinger stem), take to validate the velocity head data. care to close the coarse adjust valve before 4.1.6 Calculation of Percent Isokinetic. inserting the probe into the stack to prevent Calculate percent isokinetic (see Calcula- water from backing into the filter holder. If tions, Section 6) to determine whether the necessary, the pump may be turned on with run was valid or another test run should be the coarse adjust valve closed. made. If there was difficulty in maintaining When the probe is in position, block off the isokinetic rates due to source conditions, openings around the probe and porthole to consult with the Administrator for possible prevent unrepresentative dilution of the gas variance on the isokinetic rates. stream. 4.2 Sample Recovery. Proper cleanup pro- Traverse the stack cross-section, as re- cedure begins as soon as the probe is re- quired by Method 1 or as specified by the Ad- moved from the stack at the end of the sam- ministrator, being careful not to bump the pling period. Allow the probe to cool. probe nozzle into the stack walls when sam- When the probe can be safely handled, wipe pling near the walls or when removing or in- off all external particulate matter near the serting the probe through the portholes; this tip of the probe nozzle and place a cap over minimizes the chance of extracting deposited it to prevent losing or gaining particulate material. matter. Do not cap off the probe tip tightly During the test run, make periodic adjust- while the sampling train is cooling down as ments to keep the temperature around the this would create a vacuum in the filter filter holder at the proper level; add more ice holder, thus drawing water from the and, if necessary, salt to maintain a tem- impingers into the filter holder. perature of less than 20° C (68° F) at the con- Before moving the sample train to the denser/silica gel outlet. Also, periodically cleanup site, remove the probe from the sam- check the level and zero of the manometer. ple train, wipe off the silicone grease, and If the pressure drop across the filter be- cap the open outlet of the probe. Be careful comes too high, making isokinetic sampling not to lose any condensate that might be difficult to maintain, the filter may be re- present. Wipe off the silicone grease from the placed in the midst of a sample run. It is rec- filter inlet where the probe was fastened and ommended that another complete filter as- cap it. Remove the umbilical cord from the sembly be used rather than attempting to last impinger and cap the impinger. If a change the filter itself. Before a new filter flexible line is used between the first im- assembly is installed, conduct a leak-check pinger or condenser and the filter holder, dis- (see Section 4.1.4.2). The total particulate connect the line at the filter holder and let weight shall include the summation of all any condensed water or liquid drain into the filter assembly catches. impingers or condenser. After wiping off the A single train shall be used for the entire silicone grease, cap off the filter holder out- sample run, except in cases where simulta- let and impinger inlet. Either ground-glass neous sampling is required in two or more stoppers, plastic caps, or serum caps may be separate ducts or at two or more different lo- used to close these openings. cations within the same duct, or, in cases Transfer the probe and filter-impinger as- where equipment failure necessitates a sembly to the cleanup area. This area should change of trains. In all other situations, the be clean and protected from the wind so that use of two or more trains will be subject to the chances of contaminating or losing the the approval of the Administrator. sample will be minimized. Note that when two or more trains are Save a portion of the acetone used for used, separate analyses of the front-half and cleanup as a blank. Take 200 ml of this ace- (if applicable) impinger catches from each tone directly from the wash bottle being train shall be performed, unless identical used and place it in a glass sample container nozzle sizes were used on all trains, in which labeled ‘‘acetone blank.’’ case, the front-half catches from the individ- Inspect the train prior to and during dis- ual trains may be combined (as may the im- assembly and note any abnormal conditions. pinger catches) and one analysis of front-half Treat the samples as follows: catch and one analysis of impinger catch Container No. 1. Carefully remove the filter may be performed. Consult with the Admin- from the filter holder and place it in its iden- istrator for details concerning the calcula- tified petri dish container. Use a pair of tion of results when two or more trains are tweezers and/or clean disposable surgical used. gloves to handle the filter. If it is necessary At the end of the sample run, turn off the to fold the filter, do so such that the particu- coarse adjust valve, remove the probe and late cake is inside the fold. Carefully trans- nozzle from the stack, turn off the pump, fer to the petri dish any particulate matter record the final dry gas meter reading, and and/or filter fibers which adhere to the filter conduct a post-test leak-check, as outlined holder gasket, by using a dry Nylon bristle in Section 4.1.4.3. Also, leak-check the pitot brush and/or a sharp-edged blade. Seal the lines as described in Method 2, Section 3.1; container.
649
VerDate 11
Container No. 2. Taking care to see that After all acetone washings and particulate dust on the outside of the probe or other ex- matter have been collected in the sample terior surfaces does not get into the sample, container, tighten the lid on the sample con- quantitatively recover particulate matter or tainer so that acetone will not leak out when any condensate from the probe nozzle, probe it is shipped to the laboratory. Mark the fitting, probe liner, and front half of the fil- height of the fluid level to determine wheth- ter holder by washing these components with er or not leakage occured during transport. acetone and placing the wash in a glass con- Label the container to clearly identify its tainer. Distilled water may be used instead contents. of acetone when approved by the Adminis- Container No. 3. Note the color of the indi- trator and shall be used when specified by cating silica gel to determine if it has been the Administrator; in these cases, save a completely spent and make a notation of its water blank and follow the Administrator’s condition. Transfer the silica gel from the directions on analysis. Perform the acetone fourth impinger to its original container and rinses as follows: seal. A funnel may make it easier to pour Carefully remove the probe nozzle and the silica gel without spilling. A rubber po- clean the inside surface by rinsing with ace- liceman may be used as an aid in removing tone from a wash bottle and brushing with a the silica gel from the impinger. It is not Nylon bristle brush. Brush until the acetone necessary to remove the small amount of rinse shows no visible particles, after which dust particles that may adhere to the im- make a final rinse of the inside surface with pinger wall and are difficult to remove. acetone. Since the gain in weight is to be used for Brush and rinse the inside parts of the moisture calculations, do not use any water Swagelok fitting with acetone in a similar or other liquids to transfer the silica gel. If way until no visible particles remain. a balance is available in the field, follow the Rinse the probe liner with acetone by tilt- procedure for container No. 3 in Section 4.3. ing and rotating the probe while squirting Impinger Water. Treat the impingers as fol- acetone into its upper end so that all inside lows; Make a notation of any color or film in surfaces will be wetted with acetone. Let the the liquid catch. Measure the liquid which is acetone drain from the lower end into the in the first three impingers to within ±1 ml sample container. A funnel (glass or poly- by using a graduated cylinder or by weighing ethylene) may be used to aid on transferring it to within ±0.5 g by using a balance (if one liquid washes to the container. Follow the is available). Record the volume or weight of acetone rinse with a probe brush. Hold the liquid present. This information is required probe in an inclined position, squirt acetone to calculate the moisture content of the ef- into the upper end as the probe brush is fluent gas. being pushed with a twisting action through Discard the liquid after measuring and re- the probe; hold a sample container under- cording the volume or weight, unless analy- neath the lower end of the probe, and catch sis of the impinger catch is required (see any acetone and particulate matter which is Note, Section 2.1.7). brushed from the probe. Run the brush If a different type of condenser is used, through the probe three times or more until measure the amount of moisture condensed no visible particulate matter is carried out either volumetrically or gravimetrically. with the acetone or until none remains in Whenever possible, containers should be the probe liner on visual inspection. With shipped in such a way that they remain up- stainless steel or other metal probes, run the right at all times. brush through in the above prescribed man- 4.3 Analysis. Record the data required on ner at least six times since metal probes a sheet such as the one shown in Figure 5–3. have small crevices in which particulate Handle each sample container as follows: matter can be entrapped. Rinse the brush with acetone, and quantitatively collect FIGURE 5–3—ANALYTICAL DATA these washings in the sample container. Plant ———————————————————— After the brushing, make a final acetone Date ————————————————————— rinse of the probe as described above. Run No. ——————————————————— It is recommended that two people clean Filter No. —————————————————— the probe to minimize sample losses. Be- Amount liquid lost during transport ———— tween sampling runs, keep brushes clean and Acetone blank volume, ml ————————— protected from contaminations. Acetone wash volume, ml —————————— After ensuring that all joints have been Acetone blank concentration, mg/mg (Equa- wiped clean of silicone grease, clean the in- tion 5–4) —————————————————— side of the front half of the filter holder by Acetone wash blank, mg (Equation 5–5) —— rubbing the surfaces with a Nylon bristle brush and rinsing with acetone. Rinse each Container Weight of particulate collected, mg surface three times or more if needed to re- number move visible particulate. Make a final rinse Final weight Tare weight Weight gain of the brush and filter holder. Carefully rinse 1. out the glass cyclone, also (if applicable).
650
VerDate 11
g using a balance. This step may be con- Container Weight of particulate collected, mg number ducted in the field. Final weight Tare weight Weight gain ‘‘Acetone Blank’’ Container. Measure ace- 2. tone in this container either volumetrically or gravimetrically. Transfer the acetone to a Total. tared 250-ml beaker and evaporate to dryness at ambient temperature and pressure. Des- Less acetone blank. iccate for 24 hours and weigh to a constant weight. Report the results to the nearest 0.1 Weight of particulate matter. mg. NOTE: At the option of the tester, the con- Volume of liquid water col- lected tents of Container No. 2 as well as the ace- tone blank container may be evaporated at Impinger vol- Silica gel temperatures higher than ambient. If evapo- ume, ml weight, g ration is done at an elevated temperature, Final. the temperature must be below the boiling Initial. point of the solvent; also, to prevent ‘‘bump- Liquid collected. ing,’’ the evaporation process must be close- Total volume collected ...... g* ml ly supervised, and the contents of the beaker *Convert weight of water to volume by dividing total weight must be swirled occasionally to maintain an increase by density of water (1 g/ml). even temperature. Use extreme care, as ace- tone is highly flammable and has a low flash Increase, g = point. Volume water, ml 4.4 Quality Control Procedures. The fol- ()1 g/ ml lowing quality control procedures are sug- Container No. 1. Leave the contents in the gested to check the volume metering system shipping container or transfer the filter and calibration values at the field test site prior any loose particulate from the sample con- to sample collection. These procedures are tainer to a tared glass weighing dish. Des- optional for the tester. iccate for 24 hours in a desiccator containing 4.4.1 Meter Orifice Check. Using the cali- anhydrous calcium sulfate. Weigh to a con- bration data obtained during the calibration stant weight and report the results to the procedure described in Section 5.3, determine nearest 0.1 mg. For purposes of this Section, the ∆ H@ for the metering system orifice. 4.3, the term ‘‘constant weight’’ means a dif- The ∆ H@ is the orifice pressure differential ference of no more than 0.5 mg or 1 percent in units of in. H2O that correlates to 0.75 cfm of total weight less tare weight, whichever is of air at 528° R and 29.92 in. Hg. The ∆ H@ is greater, between two consecutive weighings, calculated as follows: with no less than 6 hours of desiccation time between weighings. T θ2 Alternatively, the sample may be oven ∆HH= 0.. 0319 ∆ m Eq 5 - 9 dried at 105° C (220° F) for 2 to 3 hours, cooled @ 2 2 P YV m in the desiccator, and weighed to a constant bar weight, unless otherwise specified by the Ad- Where: ministrator. The tester may also opt to oven ∆H=Average pressure differential across the dry the sample at 105° C (220° F) for 2 to 3 orifice meter, in. H2O. hours, weigh the sample, and use this weight Tm=Absolute average dry gas meter tempera- as a final weight. ture, ° R. Container No. 2. Note the level of liquid in P =Barometric pressure, in. Hg. the container and confirm on the analysis bar Θ=Total sampling time, min. sheet whether or not leakage occurred dur- ing transport. If a noticeable amount of Y=Dry gas meter calibration factor, leakage has occurred, either void the sample dimensionless. or use methods, subject to the approval of Vm=Volume of gas sample as measured by the Administrator, to correct the final re- dry gas meter, dcf. sults. Measure the liquid in this container ei- 0.0319=(0.0567 in. Hg/° R) x (0.75 cfm)2. ther volumetrically to ±1 ml or gravimetri- Before beginning the field test (a set of three cally to ±0.5 g. Transfer the contents to a runs usually constitutes a field test), operate tared 250-ml beaker and evaporate to dryness the metering system (i.e., pump, volume at ambient temperature and pressure. Des- meter, and orifice) at the ∆ H@ pressure dif- iccate for 24 hours and weigh to a constant ferential for 10 minutes. Record the volume weight. Report the results to the nearest 0.1 collected, the dry gas meter temperature, mg. and the barometric pressure. Calculate a dry Container No. 3. Weigh the spent silica gel gas meter calibration check value, Yc, as fol- (or silica gel plus impinger) to the nearest 0.5 lows:
651
VerDate 11
5.3 Metering System. 1 2 10 0. 0319 T 5.3.1 Calibration Prior to Use. Before its Y = m initial use in the field, the metering system c shall be calibrated as follows: Connect the VPm bar metering system inlet to the outlet of a wet Eq. 5–10 test meter that is accurate to within 1 per- Where: cent. Refer to Figure 5.5. The wet test meter should have a capacity of 30 liters/rev (1 ft3/ Yc=Dry gas meter calibration check value, dimensionless. rev). A spirometer of 400 liters (14 ft3) or 10=10 minutes of run time. more capacity, or equivalent, may be used for this calibration, although a wet test Compare the Yc value with the dry gas meter calibration factor Y to determine that: meter is usually more practical. The wet test meter should be periodically calibrated with 0.97Y< Y <1.03Y c a spirometer or a liquid displacement meter If the Yc value is not within this range, the to ensure the accuracy of the wet test meter. volume metering system should be inves- Spirometers or wet test meters of other sizes tigated before beginning the test. may be used, provided that the specified ac- 4.4.2 Calibrated Critical Orifice. A cali- curacies of the procedure are maintained. brated critical orifice, calibrated against a Run the metering system pump for about 15 wet test meter or spirometer and designed to minutes with the orifice manometer indicat- be inserted at the inlet of the sampling meter box may be used as a quality control ing a median reading as expected in field use check by following the procedure of Section to allow the pump to warm up and to permit 7.2. the interior surface of the wet test meter to be thoroughly wetted. Then, at each of a 5. Calibration minimum of three orifice manometer set- Maintain a laboratory log of all calibra- tings, pass an exact quantity of gas through tions. the wet test meter and note the gas volume 5.1 Probe Nozzle. Probe nozzles shall be indicated by the dry gas meter. Also note the calibrated before their initial use in the barometric pressure, and the temperatures of field. Using a micrometer, measure the in- the wet test meter, the inlet of the dry gas side diameter of the nozzle to the nearest meter, and the outlet of the dry gas meter. 0.025 mm (0.001 in.). Make three separate measurements using different diameters Select the highest and lowest orifice settings each time, and obtain the average of the to bracket the expected field operating range measurements. The difference between the of the orifice. Use a minimum volume of 0.15 high and low numbers shall not exceed 0.1 m3 (5 cf) at all orifice settings. Record all the mm (0.004 in.). When nozzles become nicked, data on a form similar to Figure 5.6, and cal- dented, or corroded, they shall be reshaped, culate Y, the dry gas meter calibration fac- sharpened, and recalibrated before use. Each tor, and ∆H@, the orifice calibration factor, nozzle shall be permanently and uniquely at each orifice setting as shown on Figure identified. 5.6. Allowable tolerances for individual Y and 5.2 Pitot Tube. The Type S pitot tube as- ∆H@, values are given in Figure 5.6. Use the sembly shall be calibrated according to the average of the Y values in the calculations in procedure outlined in Section 4 of Method 2. Section 6.
652
VerDate 11
653
VerDate 11
Before calibrating the metering system, it tion run at 0.00057m3/min (0.02 cfm); at the is suggested that a leak-check be conducted. end of the run, take the difference of the For metering systems having diaphragm measured wet test meter and dry gas meter pumps, the normal leak-check procedure will volumes; divide the difference by 10, to get not detect leakages within the pump. For the leak rate. The leak rate should not ex- these cases the following leak-check proce- ceed 0.00057 m3/min (0.02 cfm). dure is suggested: make a 10-minute calibra-
654
VerDate 11
5.3.2 Calibration After Use. After each calibrated if the calibration curves in APTD– field use, the calibration of the metering sys- 0576 are used. Also, probes with outlet tem- tem shall be checked by performing three perature monitoring capabilities do not re- calibration runs at a single, intermediate quire calibration. orifice setting (based on the previous field 5.5 Temperature Gauges. Use the proce- test), with the vacuum set at the maximum dure in Section 4.3 of Method 2 to calibrate value reached during the test series. To ad- in-stack temperature gauges. Dial thermom- just the vacuum, insert a valve between the eters, such as are used for the dry gas meter wet test meter and the inlet of the metering and condenser outlet, shall be calibrated system. Calculate the average value of the against mercury-in-glass thermometers. dry gas meter calibration factor. If the value 5.6 Leak Check of Metering System has changed by more than 5 percent, recali- Shown in Figure 5–1. That portion of the brate the meter over the full range of orifice settings, as previously detailed. sampling train from the pump to the orifice Alternative procedures, e.g., rechecking meter should be leak checked prior to initial the orifice meter coefficient may be used, use and after each shipment. Leakage after subject to the approval of the Administrator. the pump will result in less volume being re- 5.3.3 Acceptable Variation in Calibration. corded than is actually sampled. The follow- If the dry gas meter coefficient values ob- ing procedure is suggested (see Figure 5–4): tained before and after a test series differ by Close the main valve on the meter box. In- more than 5 percent, the test series shall ei- sert a one-hole rubber stopper with rubber ther be voided, or calculations for the test tubing attached into the orifice exhaust series shall be performed using whichever pipe. Disconnect and vent the low side of the meter coefficient value (i.e., before or after) orifice manometer. Close off the low side ori- gives the lower value of total sample vol- fice tap. Pressurize the system to 13 to 18 cm ume. (5 to 7 in.) water column by blowing into the 5.4 Probe Heater Calibration. The probe rubber tubing. Pinch off the tubing and ob- heating system shall be calibrated before its serve the manometer for one minute. A loss initial use in the field. of pressure on the manometer indicates a Use a heat source to generate air heated to leak in the meter box; leaks, if present, must selected temperatures that approximate be corrected. those expected to occur in the sources to be 5.7 Barometer. Calibrate against a mer- sampled. Pass this air through the probe at cury barometer. a typical simple flow rate while measuring 6. the probe inlet and outlet temperatures at Calculations various probe heater settings. For each air Carry out calculations, retaining at least temperature generated, construct a graph of one extra decimal figure beyond that of the probe heating system setting versus probe acquired data. Round off figures after the outlet temperature. The procedure outlined final calculation. Other forms of the equa- in APTD–0576 can also be used. Probes con- tions may be used as long as they give equiv- structed according to APTD–0581 need not be alent results.
655
VerDate 11
6.1 Nomenclature. Ca=Acetone blank residue concentration, mg/ 2 2 mg. An=Cross-sectional area of nozzle, m (ft ). Bws=Water vapor in the gas stream, propor- tion by volume.
656
VerDate 11
cs=Concentration of particulate matter in with the interval between the first and stack gas, dry basis, corrected to stand- second changes, min.
ard conditions, g/dscm (g/dscf). εp=Sampling time interval, from the final I=Percent of isokinetic sampling. (nth) component change until the end of La=Maximum acceptable leakage rate for ei- the sampling run, min. ther a pretest leak check or for a leak 13.6=Specific gravity of mercury. check following a component change; 60=Sec/min. equal to 0.00057 m3/min (0.02 cfm) or 4 per- cent of the average sampling rate, which- 100=Conversion to percent. ever is less. 6.2 Average Dry Gas Meter Temperature Li=Individual leakage rate observed during and Average Orifice Pressure Drop. See data the leak check conducted prior to the sheet (Figure 5–2). ‘‘ith’’ component change (i=1, 2, 3....n), 6.3 Dry Gas Volume. Correct the sample m3/min (cfm). volume measured by the dry gas meter to Lp=Leakage rate observed during the post- standard conditions (20° C, 760 mm Hg or 68° test leak check, m3/min (cfm). F, 29.92 in. Hg) by using Equation 5–1. ma=Mass of residue of acetone after evapo- ration, mg. mn=Total amount of particulate matter col- lected, mg. Mw=Molecular weight of water, 18.0 g/g-mole (18.0lb/lb-mole). Pbar=Barometric pressure at the sampling site, mm Hg (in. Hg). Ps=Absolute stack gas pressure, mm Hg (in. Hg). Pstd=Standard absolute pressure, 760 mm Hg Where; (29.92 in. Hg). K1=0.3858 °K/mm Hg for metric units R=Ideal gas constant, 0.06236 mm Hg-m3/°K-g- =17.64 °R/in. Hg for English units mole (21.85 in. Hg-ft3/°R-lb-mole). NOTE: Equation 5–1 can be used as written Tm=Absolute average dry gas meter tempera- ture (see Figure 5–2), °K (°R). unless the leakage rate observed during any of the mandatory leak checks (i.e., the post- Ts=Absolute average stack gas temperature (see Figure 5–2), °K (°R). test leak check or leak checks conducted Tstd=Standard absolute temperature, 293°K prior to component changes) exceeds La. If (528° R). Lp or i exceeds La, Equation 5–1 must be Va=Volume of acetone blank, ml. modified as follows: Vaw=Volume of acetone used in wash, ml. (a) Case I. No component changes made Vlc=Total volume of liquid collected in during sampling run. In this case, replace V impingers and silica gel (see Figure 5–3), m in Equation 5–1 with the expression: ml. θ Vm=Volume of gas sample as measured by [Vm—(Lp—La) ] dry gas meter, dcm (dscf). (b) Case II. One or more component Vm(std)=Volume of gas sample measured by changes made during the sampling run. In the dry gas meter, corrected to standard this case, replace Vm in Equation 5–1 by the conditions, dscm (dscf). expression: Vw(std)=Volume of water vapor in the gas sample, corrected to standard conditions, scm (scf). vs=Stack gas velocity, calculated by Method 2, Equation 2–9, using data obtained from Method 5, m/sec (ft/sec). W =Weight of residue in acetone wash, mg. a and substitute only for those leakage rates Y=Dry gas meter calibration factor. ∆H=Average pressure differential across the (Li or Lp) which exceed La. 6.4 Volume of Water Vapor. orifice meter (see Figure 5–2), mm H2O (in. H2O). φa=Density of acetone, mg/ml (see label on bottle). φw=Density of water, 0.9982 g/ml (0.002201 lb/ ml). ε=Total sampling time, min. ε1=Sampling time interval, from the begin- Where: 3 ning of a run until the first component K2=0.001333 m /ml for metric units change, min. =0.04707 ft3/ml for English units. ε i=Sampling time interval, between two suc- 6.5 Moisture Content. cessive component changes, beginning
657
VerDate 11
ρ Vw() std Wa = C a V aw a Eq. 5 - 5 Bws = Eq. 5 - 3 6.8 Total Particulate Weight. Determine Vm()() std +V w std the total particulate catch from the sum of the weights obtained from Containers 1 and NOTE: In saturated or water droplet-laden 2 less the acetone blank (see Figure 5–3). gas streams, two calculations of the mois- ture content of the stack gas shall be made, NOTE: Refer to Section 4.1.5 to assist in cal- one from the impinger analysis (Equation 5– culation of results involving two or more fil- 3), and a second from the assumption of satu- ter assemblies or two or more sampling rated conditions. The lower of the two values trains. of Bws shall be considered correct. The proce- 6.9 Particulate Concentration. dure for determining the moisture content c =(0.001 g/mg) (m /V (std)) based upon assumption of saturated condi- s n m tions is given in the Note of Section 1.2 of Eq. 5–6 Method 4. For the purposes of this method, 6.10 Conversion Factors: the average stack gas temperature from Fig- ure 5–2 may be used to make this determina- From To Multiply by tion, provided that the accuracy of the in- 3 stack temperature sensor is ±1° C (2° F). scf ...... m ...... 0.02832. g ...... mg ...... 0.001 6.6 Acetone Blank Concentration. g/ft3 ...... gr/ft3 ...... 15.43. g/ft3 ...... lb/ft3 ...... 2.205×10¥3. m g/ft3 ...... g/m3 ...... 35.31. C = a Eq. 5 - 4 a ρ Va a 6.11 Isokinetic Variation. 6.7 Acetone Wash Blank. 6.11.1 Calculation From Raw Data.
+ + ∆ 100TKVVY/TPH/s[] 31 c() m m() bar 13. 6 I = Eq. 5 - 7 θ 60 VsPA s n
Where: velocity and volumetric flow rate, if needed, 3 K3=0.003454 mm Hg¥m /ml¥°K for metric using data obtained in this method and the units. equations in Sections 5.2 and 5.3 of Method 2. =0.002669-in. Hg¥ft3/ml¥°R for English 7. Alternative Procedures units. 7.1 Dry Gas Meter as a Calibration Stand- 6.11.2 Calculation From Intermediate Val- ard. A dry gas meter may be used as a cali- ues. bration standard for volume measurements in place of the wet test meter specified in Section 5.3, provided that it is calibrated ini- tially and recalibrated periodically as fol- lows: 7.1.1 Standard Dry Gas Meter Calibration. 7.1.1.1 The dry gas meter to be calibrated and used as a secondary reference meter should be of high quality and have an appro- Where: priately sized capacity, e.g., 3 liters/rev (0.1 K4=4.320 for metric units ft 3/rev). A spirometer (400 liters or more ca- =0.09450 for English units. pacity), or equivalent, may be used for this 6.12 Acceptable Results. If 90 percent ≤ I ≤ calibration, although a wet test meter is 110 percent, the results are acceptable. If the particulate results are low in comparison to usually more practical. The wet test meter the standard, and I is over 110 percent or less should have a capacity of 30 liters/rev (1 than 90 percent, the Administrator may ac- ft 3/rev) and capable of measuring volume to cept the results. Citation 4 in the bibliog- within ±1.0 percent; wet test meters should raphy section can be used to make accept- be checked against a spirometer or a liquid ability judgments. If I is judged to be unac- displacement meter to ensure the accuracy ceptable, reject the particulate results and of the wet test meter. Spirometers or wet repeat the test. test meters of other sizes may be used, pro- 6.13 Stack Gas Velocity and Volumetric vided that the specified accuracies of the Flow Rate. Calculate the average stack gas procedure are maintained.
658
VerDate 11
7.1.1.2 Set up the components as shown in the manometer at the inlet side of the dry Figure 5.7. A spirometer, or equivalent, may gas meter should be minimized [no greater be used in place of the wet test meter in the than 100 mm H2O (4 in. H2O) at a flow rate of system. Run the pump for at least 5 minutes 30 liters/min (1 cfm)]. This can be accom- at a flow rate of about 10 liters/min (0.35 cfm) plished by using large diameter tubing con- to condition the interior surface of the wet nections and straight pipe fittings. test meter. The pressure drop indicated by
7.1.1.3 Collect the data as shown in the ex- range of flow rates should be between 10 and ample data sheet (see Figure 5–8). Make trip- 34 liters/min (0.35 and 1.2 cfm) or over the ex- licate runs at each of the flow rates and at pected operating range. no less than five different flow rates. The
659
VerDate 11
7.1.1.4 Calculate flow rate, Q, for each run coefficient, Yds, for each run. These calcula- using the wet test meter gas volume, Vw, and tions are as follows: the run time, θ. Calculate the dry gas meter
660
VerDate 11
7.1.2.2 As an alternative to full recalibra- P V tion, a two-point calibration check may be QK= bar w 1 + θ made. Follow the same procedure and equip- tw t std ment arrangement as for a full recalibration, ()+ but run the meter at only two flow rates Vw t ds t std Pbar [suggested rates are 14 and 28 liters/min (0.5 Y = and 1.0 cfm)]. Calculate the meter coeffi- ds V() t+ t ∆p ds w std P + cients for these two points, and compare the bar 13. 6 values with the meter calibration curve. If the two coefficients are within ±1.5 percent Where: of the calibration curve values at the same Kl=0.3858 for international system of units flow rates, the meter need not be recali- (SI); 17.64 for English units. brated until the next date for a recalibration 3 Vw=Wet test meter volume, liters (ft ). check. 3 Vds=Dry gas meter volume, liters (ft ). 7.2 Critical Orifices As Calibration Stand- tds=Average dry gas meter temperature, ° C ards. Critical orifices may be used as calibra- (° F). tion standards in place of the wet test meter tstd=273° C for SI units; 460° F for English specified in Section 5.3, provided that they units. are selected, calibrated, and used as follows: ° tw=Average wet test meter temperature, C 7.2.1 Section of Critical Orifices. ° ( F). 7.2.1.1 The procedure that follows de- P =Barometric pressure, mm Hg (in. Hg). bar scribes the use of hypodermic needles or ∆p=Dry gas meter inlet differential pressure, stainless steel needle tubings which have mm H O (in. H O). 2 2 been found suitable for use as critical ori- θ=Run time, min. fices. Other materials and critical orifice de- 7.1.1.5 Compare the three Y values at ds signs may be used provided the orifices act each of the flow rates and determine the as true critical orifices; i.e., a critical vacu- maximum and minimum values. The dif- ference between the maximum and minimum um can be obtained, as described in Section values at each flow rate should be no greater 7.2.2.2.3. Select five critical orifices that are than 0.030. Extra sets of triplicate runs may appropriately sized to cover the range of flow be made in order to complete this require- rates between 10 and 34 liters/min or the ex- ment. In addition, the meter coefficients pected operating range. Two of the critical should be between 0.95 and 1.05. If these spec- orifices should bracket the expected operat- ifications cannot be met in three sets of suc- ing range. cessive triplicate runs, the meter is not suit- A minimum of three critical orifices will able as a calibration standard and should not be needed to calibrate a Method 5 dry gas be used as such. If these specifications are meter (DGM); the other two critical orifices can serve as spares and provide better selec- met, average the three Yds values at each flow rate resulting in five average meter co- tion for bracketing the range of operating efficients, Yds. flow rates. The needle sizes and tubing 7.1.1.6 Prepare a curve of meter coeffi- lengths shown below give the following ap- cient, Yds, versus flow rate, Q, for the dry gas proximate flow rates: meter. This curve shall be used as a ref- erence when the meter is used to calibrate Flow rate (li- Flow rate (li- Gauge/cm ters/min) Gauge/cm ters/min) other dry gas meters and to determine whether recalibration is required. 12/7.6 32.56 14/2.5 19.54 7.1.2 Standard Dry Gas Meter Recalibra- 12/10.2 30.02 14/5.1 17.27 tion. 13/2.5 25.77 14/7.6 16.14 7.1.2.1 Recalibrate the standard dry gas 13/5.1 23.50 15/3.2 14.16 meter against a wet test meter or spirometer 13/7.6 22.37 15/7.6 11.61 annually or after every 200 hours of oper- 13/10.2 20.67 15/10.2 10.48 ation, whichever comes first. This require- ment is valid provided the standard dry gas 7.2.1.2 These needles can be adapted to a meter is kept in a laboratory and, if trans- Method 5 type sampling train as follows: In- ported, cared for as any other laboratory in- sert a serum bottle stopper, 13- by 20-mm strument. Abuse to the standard meter may sleeve type, into a 1⁄2-inch Swagelok quick cause a change in the calibration and will re- connect. Insert the needle into the stopper as quire more frequent recalibrations. shown in Figure 5–9.
661
VerDate 11
7.2.2 Critical Orifice Calibration. The pro- i.e., no detectable movement of the DGM cedure described in this section uses the dial shall be seen for 1 minute. Method 5 meter box configuration with a 7.2.2.1.2 Check also for leakages in that DGM as described in Section 2.1.8 to cali- portion of the sampling train between the brate the critical orifices. Other schemes pump and the orifice meter. See Section 5.6 may be used, subject to the approval of the for the procedure; make any corrections, if Administrator. necessary. If leakage is detected, check for 7.2.2.1 Calibration of Meter Box. The criti- cracked gaskets, loose fittings, worn O-rings, cal orifices must be calibrated in the same etc., and make the necessary repairs. configuration as they will be used; i.e., there 7.2.2.1.3 After determining that the meter should be no connections to the inlet of the box is leakless, calibrate the meter box ac- orifice. cording to the procedure given in Section 5.3. 7.2.2.1.1 Before calibrating the meter box, Make sure that the wet test meter meets the leak check the system as follows: Fully open requirements stated in Section 7.1.1.1. Check the coarse adjust valve, and completely close the water level in the wet test meter. Record the by-pass valve. Plug the inlet. Then trun the DGM calibration factor, Y. on the pump, and determine whether there is 7.2.2.2 Calibration of Critical Orifices. Set any leakage. The leakage rate shall be zero; up the apparatus as shown in Figure 5–10.
662
VerDate 11
7.2.2.2.1 Allow a warm-up time of 15 min- Orifices that do not reach a critical value utes. This step is important to equilibrate shall not be used. the temperature conditions through the 7.2.2.2.4 Obtain the barometric pressure DGM. using a barometer as described in Section 7.2.2.2.2 Leak check the system as in Sec- 2.1.9. Record the barometric pressure, Pbar, in tion 7.2.2.1.1. The leakage rate shall be zero. mm Hg (in. Hg). 7.2.2.2.3 Before calibrating the critical 7.2.2.2.5 Conduct duplicate runs at a vacu- orifice, determine its suitability and the ap- um of 25 to 50 mm Hg (1 to 2 in. Hg) above propriate operating vacuum as follows: Turn the critical vacuum. The runs shall be at on the pump, fully open the coarse adjust least 5 minutes each. The DGM volume read- valve, and adjust the by-pass valve to give a ings shall be in increments of 0.00283 m3 (0.1 vacuum reading corresponding to about half ft3) or in increments of complete revolutions of atmospheric pressure. Observe the meter of the DGM. As a guideline, the times should box orifice manometer reading, H. Slowly in- not differ by more than 3.0 seconds (this in- crease the vacuum reading until a stable cludes allowance for changes in the DGM reading is obtained on the meter box orifice temperatures) to achieve ± 0.5 percent in K′. manometer. Record the critical vacuum for Record the information listed in Figure 5–11. each orifice. 7.2.2.2.6 Calculate K′ using Equation 5–9.
663
VerDate 11
KVYPHT()+ ∆ /.13 6 K′ = 1 m bar amb Eq. 5 - 9 θ PTbar m
Where:
1 1 3o 3 o (m )( K ) 2 (ft )( R ) 2 = K′ Critical orifice coefficient, () mm. Hg ()min()in.Hg () min
Tamb=Absolute ambient temperature, °K (°R). Run number Average the K′ values. The individual K′ Dry gas meter values should not differ by more than ±0.5 1 2 percent from the average. min ...... 7.2.3 Using the Critical Orifices as Cali- Orifice man. rdg., >H mm (in.) H2O ...... bration Standards. Bar. pressure, Pbar ...... mm (in.) Hg ...... 7.2.3.1 Record the barometric pressure. Ambient temperature, °C (°F) ...... Date llll Train ID llll DGM cal. fac- tamb. tor llll Critical orifice ID llll Pump vacuum ...... mm (in.) Hg ...... K′ factor ...... Run number Average ...... Dry gas meter 1 2 Figure 5–11. Data sheet for determining K′ Final reading ...... m3 (ft3) ...... factor. Initial reading ...... m3 (ft3) ...... 7.2.3.2 Calibrate the metering system ac- 3 3 Difference, Vm ...... m (ft ) ...... cording to the procedure outlined in Sections Inlet/Outlet tempera- 7.2.2.2.1 to 7.2.2.2.5. Record the information tures: listed in Figure 5.12. ° ° Initial ...... C ( F) ...... / / 7.2.3.3 Calculate the standard volumes of Final ...... °C (°F) ...... / / air passed through the DGM and the critical Avg. Temperature, tm °C (°F) ...... Time, r ...... min/sec ...... / / orifices, and calculate the DGM calibration factor, Y, using the equations below:
+ ∆ = PHbar (/.)13 6 Vm() std K1 V m Eq. 510- Tm
θ tion factor, Y, at each of the flow rates = Pbar should not differ by more than ±2 percent Vcr ( std ) K ©Eq . 511- from the average. Tamb 7.2.3.5 To determine the need for recali- brating the critical orifices, compare the Vcr() std DGM Y factors obtained from two adjacent Y = Eq. 512- orifices each time a DGM is calibrated; for Vm() std example, when checking 13/2.5, use orifices 12/10.2 and 13/5.1. If any critical orifice yields where: a DGM Y factor differing by more than 2 per- Vcr(std)=Volume of gas sample passed through cent from the others, recalibrate the critical the critical orifice, corrected to standard orifice according to Section 7.2.2.2. 3 conditions, dsm (dscf). Date llll Train ID llll Critical ori- ° K1=0.3858 K/mm Hg for metric units fice ID llll Critical orifice K′ factor =17.64 °R/in. Hg for English units. llll 7.2.3.4 Average the DGM calibration val- ues for each of the flow rates. The calibra-
664
VerDate 11
Run number Environmental Protection Agency. Research Dry gas meter Triangle Park, NC 27711, Publication No. 1 2 EPA–600/7–77–060. June 1977. 83 p. Final reading ...... m3 (ft3) ...... 11. Westlin, P. R. and R. T. Shigehara. Pro- Initial reading ...... m3 (ft3) ...... cedure for Calibrating and Using Dry Gas 3 3 Difference, Vm ...... m (ft ) ...... Volume Meters as Calibration Standards. Inlet/outlet temperatures: Source Evaluation Society Newsletter. Initial ...... °C (°F) ...... / / 3(1):17–30. February 1978. ° ° 12. Lodge, J.P., Jr., J.B. Pate, B.E. Final ...... C ( F) ...... / / Ammons, and G.A. Swanson. The Use of Hypodermic Needles as Critical Orifices in Avg. Temperature, tm ...... °C (°F) ...... Time, r ...... min/sec ...... / / Air Sampling. J. Air Pollution Control Asso- ciation. 16:197–200. 1966. min ...... Orifice man. rdg., >H ...... mm (in.) ...... METHOD 5A—DETERMINATION OF PARTICULATE H2O. EMISSIONS FROM THE ASPHALT PROCESSING Bar. pressure, Pbar ...... mm (in.) Hg ...... AND ASPHALT ROOFING INDUSTRY Ambient temperature, tamb ..... °C (°F) ...... Pump vacuum ...... mm (in.) Hg ...... 1. Applicability and Principle 3 3 Vm(std) ...... m (ft ) ...... 1.1 Applicability. This method applies to 3 3 Vcr(std) ...... m (ft ) ...... the determination of particulate emissions DGM cal. factor, Y ...... from asphalt roofing industry process satu- Figure 5–12. Data sheet for determining rators, blowing stills, and other sources as DGM Y factor. specified in the regulations. 1.2 Principle. Particulate matter is with- 8. Bibliography drawn isokinetically from the source and 1. Addendum to Specifications for Inciner- collected on a glass filter fiber maintained at ator Testing at Federal Facilities. PHS, a temperature of 42°±10°C (108°±18°F). The NCAPC. Dec. 6, 1967. particulate mass, which includes any mate- 2. Martin, Robert M. Construction Details rial that condenses at or above the filtration of Isokinetic Source-Sampling Equipment. temperature, is determined gravimetrically Environmental Protection Agency. Research after removal of uncombined water. Triangle Park, NC. APTD–0581. April 1971. 2. Apparatus 3. Rom, Jerome J. Maintenance, Calibra- tion, and Operation of Isokinetic Source 2.1 Sampling Train. The sampling train Sampling Equipment. Environmental Pro- configuration is the same as shown in Figure tection Agency. Research Triangle Park, NC. 5–1 of Method 5. The sampling train consists APTD–0576. March, 1972. of the following components: 2.1.1 Probe Nozzle, Pitot Tube, Differen- 4. Smith, W. S., R. T. Shigehara, and W. F. tial Pressure Gauge, Filter Holder, Con- Todd. A Method of Interpreting Stack Sam- denser, Metering System, Barometer, and pling Data. Paper Presented at the 63d An- Gas Density Determination Equipment. nual Meeting of the Air Pollution Control Same as Method 5, Sections 2.1.1, 2.1.3 to Association, St. Louis, MO, June 14–19, 1970. 2.1.5, and 2.1.7 to 2.1.10, respectively. 5. Smith, W. S., et al. Stack Gas Sampling 2.1.2 Probe Liner. Same as in Method 5, Improved and Simplified With New Equip- Section 2.1.2, with the note that at high ment. APCA Paper No. 67–119. 1967. stack gas temperatures (greater than 250°C 6. Specifications for Incinerator Testing at (480°F)), water-cooled probes may be required Federal Facilities. PHS, NCAPC. 1967. to control the probe exit temperature to 7. Shigehara, R. T. Adjustments in the 42°±10°C (108±18°F). EPA Nomograph for Different Pitot Tube Co- 2.1.3 Precollector Cyclone. Borosilicate efficients and Dry Molecular Weights. Stack glass following the construction details Sampling News 2:4–11, October, 1974. shown in Air Pollution Technical Document– 8. Vollaro, R. F. A Survey of Commercially 0581, ‘‘Construction Details of Isokinetic Available Instrumentation For the Measure- Source-Sampling Equipment’’. ment of Low-Range Gas Velocities. U.S. En- vironmental Protection Agency, Emission NOTE: The tester shall use the cyclone Measurement Branch. Research Triangle when the stack gas moisture is greater than Park, NC. November, 1976 (unpublished 10 percent. The tester shall not use the paper). precollector cyclone under other, less severe 9. Annual Book of ASTM Standards. Part conditions. 26. Gaseous Fuels; Coal and Coke; Atmos- 2.1.4 Filter Heating System. Any heating pheric Analysis. American Society for Test- (or cooling) system capable of maintaining a ing and Materials. Philadelphia, PA. 1974. pp. sample gas temperature at the exit end of 617–622. the filter holder during sampling at 42°±10° C 10. Felix, L. G., G. I. Clinard, G. E. Lacey, (108°±18° F). Install a temperature gauge ca- and J. D. McCain. Inertial Cascade Impactor pable of measuring temperature within 3° C Substrate Media for Flue Gas Sampling. U.S. (5.4° F) at the exit side of the filter holder so
665
VerDate 11
that the sensing tip of the temperature 4.1.1 Pretest Preparation. Unless other- gauge is in direct contact with the sample wise specified, maintain and calibrate all gas, and the sample gas temperature can be components according to the procedure de- regulated and monitored during sampling. scribed in Air Pollution Technical Docu- The temperature gauge shall comply with ment-0576, ‘‘Maintenance, Calibration, and the calibration specifications defined in Sec- Operation of Isokinetic Source-Sampling tion 5. The tester may use systems other Equipment’’. than the one shown in APTD–0581. Prepare probe liners and sampling nozzles 2.2 Sample Recovery. The equipment re- as needed for use. Thoroughly clean each quired for sample recovery is as follows: component with soap and water followed by 2.2.1 Probe-Liner and Probe-Nozzle Brush- a minimum of three TCE rinses. Use the es, Graduated Cylinder and/or Balance, Plas- probe and nozzle brushes during at least one tic Storage Containers, and Funnel and Rub- of the TCE rinses (refer to Section 4.2 for ber Policeman. Same as Method 5, Sections rinsing techniques). Cap or seal the open 2.2.1, 2.2.5, 2.2.6, and 2.2.7, respectively. ends of the probe liners and nozzles to pre- 2.2.2 Wash Bottles. Glass. vent contamination during shipping. 2.2.3 Sample Storage Containers. Chemi- Prepare silica gel portions and glass filters cally resistant, borosilicate glass bottles, as specified in Method 5, Section 4.1.1. with rubber-backed Teflon screw cap liners 4.1.2 Preliminary Determinations. Select or caps that are constructed so as to be leak- the sampling site, probe nozzle, and probe free and resistant to chemical attack by length as specified in Method 5, Section 4.1.2. 1,1,1-trichloroethane (TCE), 500-ml or 1000- Select a total sampling time greater than ml. (Narrow mouth glass bottles have been or equal to the minimum total sampling found to be less prone to leakage.) time specified in the test procedures section 2.2.4 Petri Dishes. Glass, unless otherwise of the applicable regulation. Follow the specified by the Administrator. guidelines outlined in Method 5, Section 2.2.5 Funnel. Glass. 4.1.2, for sampling time per point and total 2.3 Analysis. For analysis, the following sample volume collected. equipment is needed: 4.1.3 Preparation of Collection Train. Pre- 2.3.1 Glass Weighing Dishes, Desiccator, pare the collection train as specified in Analytical Balance, Balance, Hygrometer, Method 5, Section 4.1.3, with the addition of and Temperature Gauge. Same as Method 5, the following: Sections 2.3.1 to 2.3.4, 2.3.6, and 2.3.7, respec- Set up the sampling train as shown in Fig- tively. ure 5–1 of Method 5 with the addition of the 2.3.2 Beakers. Glass, 250-ml and 500-ml. precollector cyclone, if used, between the 2.3.3 Separatory Funnel. 100-ml or great- probe and filter holder. The temperature of er. the precollector cyclone, if used, should be °± ° 3. Reagents about the same as for the filter, i.e., 42 10 C (108°±18°F). Use no stopcock grease on ground 3.1 Sampling. The reagents used in sam- glass joints unless the grease is insoluble in pling are as follows: TCE. 3.1.1. Filters, Silica Gel, and Crushed Ice. 4.1.4 Leak Check Procedures. Follow the Same as Method 5, Sections 3.1.1, 3.1.2, and procedures given in Method 5, Sections 4.1.4.1 3.1.4, respectively. (Pretest Leak Check), 4.1.4.2 (Leak Check 3.1.2 Stopcock Grease. TCE-insoluble, During Sample Run), and 4.1.4.3 (Post-Test heat-stable grease (if needed). This is not Leak Check). necessary if screw-on connectors with Teflon 4.1.5 Particulate Train Operation. Operate sleeves, or similar, are used. the sampling train as described in Method 5, 3.2 Sample Recovery. Reagent grade 1,1,1- Section 4.1.5, except maintain the gas tem- trichloroethane (TCE), 0.001 percent residue ™ perature exiting the filter at 42°±10°C and stored in glass bottles, is required. Run (108°±18°F). TCE blanks prior to field use and use only 4.1.6 Calculation of Percent Isokinetic. TCE with low blank values ( 0.001 percent). ™ Same as in Method 5, Section 4.1.6. The tester shall in no case subtract a blank 4.2 Sample Recovery. Using the proce- value of greater than 0.001 percent of the dures and techniques described in Method 5, weight of TCE used from the sample weight. Section 4.2, quantitatively recover any par- 3.3 Analysis. Two reagents are required ticulate matter into the following containers for the analysis: (additions and deviations to the stated pro- 3.3.1 TCE. Same as 3.2. cedures are as noted): 3.3.2 Desiccant. Same as Method 5, Sec- 4.2.1 Container No. 1 (Filter). Same in- tion 3.3.2. structions as Method 5, Section 4.2, ‘‘Con- 4. Procedure tainer No. 1.’’ If it is necessary to fold the 4.1 Sampling Train Operation. The com- filter, do so such that the film of oil is inside plexity of this method is such that in order the fold. to obtain reliable results, testers should be 4.2.2 Container No. 2 (Probe to Filter trained and experienced with Method 5 test Holder). Taking care to see that material on procedures. the outside of the probe or other exterior
666
VerDate 11
surfaces does not get into the sample, quan- ml; adjust the stack gas moisture content, if titatively recover particulate matter or any necessary (see Sections 6.4 and 6.5). Next, ex- condensate from the probe nozzle, probe fit- tract the water phase with several 25-ml por- ting, probe liner, precollector cyclone and tions of TCE until, by visual observation, the collector flask (if used), and front half of the TCE does not remove any additional organic filter holder by washing these components material. Evaporate the remaining water with TCE and placing the wash in a glass fraction to dryness at 93°C (200°F), desiccate container. Carefully measure the total for 24 hours, and weigh to the nearest 0.1 mg. amount of TCE used in the rinses. Perform Treat the total TCE fraction (including the TCE rinses as described in Method 5, Sec- TCE from the filter container rinse and tion 4.2, ‘‘Container No. 2,’’ using TCE in- water phase extractions) as follows: Transfer stead of acetone. the TCE and oil to a tared beaker and evapo- Brush and rinse the inside of the cyclone, rate at ambient temperature and pressure. cyclone collection flask, and the front half of The evaporation of TCE from the solution the filter holder. Brush and rinse each sur- may take several days. Do not desiccate the face three times or more, if necessary, to re- sample until the solution reaches an appar- move visible particulate. ent constant volume or until the odor of TCE 4.2.3 Container No. 3 (Silica Gel). Same is not detected. When it appears that the procedure as in Method 5, Section 4.2, ‘‘Con- TCE has evaporated, desiccate the sample tainer No. 3.’’ and weigh it at 24-hour intervals to obtain a 4.2.4 Impinger Water. Treat the impingers ‘‘constant weight’’ (as defined for Container as follows: Make a notation of any color or No. 1 above). The ‘‘total weight’’ for Con- film in the liquid catch. Follow the same tainer No. 2 is the sum of the evaporated par- procedure as in Method 5, Section 4.2, ‘‘Im- ticulate weight of the TCE-oil and water pinger Water.’’ phase fractions. Report the results to the 4.2.5 Blank. Save a portion of the TCE nearest 0.1 mg. used for cleanup as a blank. Take 200 ml of 4.3.3 Container No. 3 (Silica Gel). This this TCE directly from the wash bottle being step may be conducted in the field. Weigh used and place it in a glass sample container the spent silica gel (or silica gel plus im- labeled ‘‘TCE blank.’’ pinger) to the nearest 0.5 g using a balance. 4.3 Analysis. Record the data required on 4.3.4 ‘‘TCE Blank’’ Container. Measure a sheet such as the one shown in Figure 5A– TCE in this container either volumetrically 1. Handle each sample container as follows: or gravimetrically. Transfer the TCE to a 4.3.1 Container No. 1 (Filter). Transfer the tared 250-ml beaker and evaporate to dryness filter from the sample container to a tared at ambient temperature and pressure. Des- glass weighing dish and desiccate for 24 iccate for 24 hours and weigh to a constant hours in a desiccator containing anhydrous weight. Report the results to the nearest 0.1 calcium sulfate. Rinse Container No. 1 with mg. a measured amount of TCE and analyze this rinse with the contents of Container No. 2. NOTE: In order to facilitate the evapo- Weigh the filter to a constant weight. For ration of TCE liquid samples, these samples the purpose of Section 4.3, the term ‘‘con- may be dried in a controlled temperature ° ° stant weight’’ means a difference of no more oven at temperatures up to 38 C (100 F) until than 10 percent or 2 mg (whichever is great- the liquid is evaporated. er) between two consecutive weighings made 4.4 Quality Control Procedures. A quality 24 hours apart. Report the ‘‘final weight’’ to control (QC) check of the volume metering the nearest 0.1 mg as the average of these system at the field site is suggested before two values. collecting the sample. Use the procedure de- 4.3.2 Container No. 2 (Probe to Filter fined in Method 5, Section 4.4. Holder). Before adding the rinse from Con- 5. Calibration tainer No. 1 to Container No. 2, note the Calibrate the sampling train components level of liquid in the container and confirm according to the indicated sections of Meth- on the analysis sheet whether or not leakage od 5: Probe Nozzle (5.1), Pitot Tube Assembly occurred during transport. If noticeable (5.2), Metering System (5.3), Probe Heater leakage occurred, either void the sample or (5.4), Temperature Gauges (5.5), Leak Check take steps, subject to the approval of the Ad- of Metering System (5.6), and Barometer ministrator, to correct the final results. (5.7). Measure the liquid in this container either volumetrically to ±1 ml or gravimetrically 6. Calculations to ±0.5 g. Check to see if there is any appre- 6.1 Nomenclature. Same as in Method 5, ciable quantity of condensed water present Section 6.1, with the following additions: in the TCE rinse (look for a boundary layer Ct=TCE blank residue concentration, mg/mg. or phase separation). If the volume of con- mt=Mass of residue of TCE after evaporation, densed water appears larger than 5 ml, sepa- mg. rate the oil-TCE fraction from the water Vpc=Volume of water collected in precollec- fraction using a separatory funnel. Measure tor, ml. the volume of the water phase to the nearest Vt=Volume of TCE blank, ml.
667
VerDate 11
Vtw=Volume of TCE used in wash, ml. METHOD 5B—DETERMINATION OF NONSULFURIC Wt=Weight of residue in TCE wash, mg. ACID PARTICULATE MATTER FROM STATION- ρT=Density of TCE, mg/ml (see label on bot- ARY SOURCES tle). 1. Applicability and Principle. 6.2 Dry Gas Meter Temperature and Ori- 1.1 Applicability. This method is to be fice Pressure Drop. Using the data obtained used for determining nonsulfuric acid partic- in this test, calculate the average dry gas ulate matter from stationary sources. Use of meter temperature and average orifice pres- this method must be specified by an applica- sure drop (see Figure 5–2 of Method 5). ble subpart, or approved by the Adminis- 6.3 Dry Gas Volume. Using the data from trator, U.S. Environmental Protection Agen- this test, calculate Vm(std) by using Equation cy, for a particular application. 5–1 of Method 5. If necessary, adjust the vol- 1.2 Principle. Particulate matter is with- ume for leakages. drawn isokinetically from the source using 6.4 Volume of Water Vapor. the Method 5 train at 160 °C (320 °F). The col-
Vw(std)=Kl(Vlc+Vpc) lected sample is then heated in the oven at 160 °C (320 °F) for 6 hours to volatilize any Eq. 5A–1 condensed sulfuric acid that may have been Where: collected, and the nonsulfuric acid particu- 3 Kl=0.00133 m /ml for metric units. late mass is determined gravimetrically. =0.04707 ft 3/ml for English units. 2. Procedure. 6.5 Moisture Content. The procedure is identical to EPA Method 5 except for the following: B =V /[V +V ws w(std) m(std) w(std)] 2.1 Initial Filter Tare. Oven dry the filter Eq. 5A–2 at 160≤5 °C (320 ≤10 °F) for 2 to 3 hours, cool in a desiccator for 2 hours, and weigh. Des- NOTE: In saturated or water droplet-laden gas streams, two calculations of the mois- iccate to constant weight to obtain the ini- ture content of the stack gas shall be made, tial tare. Use the applicable specifications one from the impinger and precollector anal- and techniques of Section 4.1.1 of Method 5 ysis (Equations 5A–1 and 5A–2) and a second for this determination. 2.2 Probe and Filter Temperatures. Main- from the assumption of saturated conditions. tain the probe outlet and filter temperatures The lower of the two values of moisture con- at 160 14 °C (320 25 °F). tent shall be considered correct. The proce- ≤ ≤ 2.3 Analysis. Dry the probe sample at am- dure for determining the moisture content bient temperature. Then oven-dry the probe based upon assumption of saturated condi- and filter samples at a temperature of tions is given in the note of Section 1.2 of 160 5 °C (320 10 °F) for 6 hours. Cool in a des- Method 4. For the purpose of this method, ≤ ≤ iccator for 2 hours, and weigh to constant the average stack gas temperature from Fig- weight. Use the applicable specifications and ure 5–2 of Method 5 may be used to make this techniques of Section 4.3 of Method 5 for this determination, provided that the accuracy of determination. the in-stack temperature sensor is within ± ° ° 1 C (2 F). METHOD 5C—[RESERVED] 6.6 TCE Blank Concentration. METHOD 5D—DETERMINATION OF PARTICULATE Ct=mt/Vtρt MATTER EMISSIONS FROM POSITIVE PRES- Eq. 5A–3 SURE FABRIC FILTERS 6.7 TCE Wash Blank. 1. Applicability and Principle W =C V ρ t t tw t 1.1 Applicability. This method applies to Eq. 5A–4 the determination of particulate matter 6.8 Total Particulate Weight. Determine emissions from positive pressure fabric fil- the total particulate catch from the sum of ters. Emissions are determined in terms of the weights obtained from Containers 1, 2, concentration (mg/m3) and emission rate (kg/ and 3, less the TCE blank. h). 6.9 Particulate Concentration. The General Provisions of 40 CFR Part 60, § 60.8(e), require that the owner or operator cs=K2mn/Vm(std) of an affected facility shall provide perform- Eq. 5A–5 ance testing facilities. Such performance Where: testing facilities include sampling ports, safe sampling platforms, safe access to sampling K2=0.001 g/mg. 6.10 Isokinetic Variation and Acceptable sites, and utilities for testing. It is intended Results. Same as in Method 5, Sections 6.11 that affected facilities also provide sampling and 6.12, respectively. locations that meet the specification for ade- quate stack length and minimal flow dis- 7. Bibliography turbances as described in Method 1. Provi- The bibliography for Method 5A is the sions for testing are often overlooked factors same as that for Method 5. in designing fabric filters or are extremely
668
VerDate 11
costly. The purpose of this procedure is to 4.2 Determination of Number and Loca- identify appropriate alternative locations tion of Traverse Points. Locate the traverse and procedures for sampling the emissions points according to Method 1, Section 2.3. from positive pressure fabric filters. The re- Because a performance test consists of at quirements that the affected facility owner least three test runs and because of the var- or operator provide adequate access to per- ied configurations of positive pressure fabric formance testing facilities remain in effect. filters, there are several schemes by which 1.2 Principle. Particulate matter is with- the number of traverse points can be deter- drawn isokinetically from the source and mined and the three test runs can be con- collected on a glass fiber filter maintained at ducted. a temperature at or above the exhaust gas 4.2.1 Single Stacks Meeting Method 1 Cri- temperature up to a nominal 120 °C (120±14 °C teria. Select the number of traverse points or 248 ±25 °F). The particulate mass, which in- according to Method 1. Sample all traverse cludes any material that condenses at or points for each test run. above the filtration temperature, is deter- 4.2.2 Other Single Measurement Sites. For mined gravimetrically after removal of a roof monitor or monovent, single compart- uncombined water. ment housing, or other stack not meeting 2. Apparatus Method 1 criteria, use at least 24 traverse points. For example, for a rectangular meas- The equipment requirements for the sam- urement site, such as a monovent, use a bal- pling train, sample recovery, and analysis anced 5 x 5 traverse point matrix. Sample all are the same as specified in Sections 2.1, 2.2, traverse points for each test run. and 2.3, respectively, of Method 5 or Method 4.2.3 Multiple Measurement Sites. Sam- 17. pling from two or more stacks or measure- 3. Reagents ment sites may be combined for a test run, The reagents used in sampling, sample re- provided the following guidelines are met: covery, and analysis are the same as speci- (a) All measurement sites up to 12 must be fied in Sections 3.1, 3.2, and 3.3, respectively, sampled. For more than 12 measurement of Method 5 or Method 17. sites, conduct sampling on at least 12 sites or 4. Procedure 50 percent of the sites, whichever is greater. 4.1 Determination of Measurement Site. The measurement sites sampled should be The configurations of positive pressure fab- evenly, or nearly evenly, distributed among ric filter structures frequently are not ame- the available sites; if not, all sites are to be nable to emission testing according to the sampled. requirements of Method 1. Following are sev- (b) The same number of measurement sites eral alternatives for determining measure- must be sampled for each test run. ment sites for positive pressure fabric filters. (c) The minimum number of traverse 4.1.1 Stacks Meeting Method 1 Criteria. points per test run is 24. An exception to the Use a measurement site as specified in Meth- 24-point minimum would be a test combining od 1, Section 2.1. the sampling from two stacks meeting Meth- 4.1.2 Short Stacks Not Meeting Method 1 od 1 criteria for acceptable stack length, and Criteria. Use stack extensions and the proce- Method 1 specifies fewer than 12 points per dures in Method 1. Alternatively, use flow site. straightening vanes of the ‘‘egg-crate’’ type (d) As long as the 24 traverse points per (see Figure 5D–1). Locate the measurement test run criterion is met, the number of tra- site downstream of the straightening vanes verse points per measurement site may be at a distance equal to or greater than two reduced to eight. times the average equivalent diameter of the Alternatively, conduct a test run for each vane openings and at least one-half of the measurement site individually using the cri- overall stack diameter upstream of the stack teria in Section 4.2.1 or 4.2.2 for number of outlet. traverse points. Each test run shall count to- 4.1.3 Roof Monitor or Monovent. (See Fig- ward the total of three required for a per- ure 5D–2.) For a positive pressure fabric fil- formance test. If more than three measure- ter equipped with a peaked roof monitor, ment sites are sampled, the number of tra- ridge vent, or other type of monovent, use a verse points per measurement site may be measurement site at the base of the reduced to eight as long as at least 72 tra- monovent. Examples of such locations are verse points are sampled for all the tests. shown in Figure 5D–2. The measurement site The following examples demonstrate the must be upstream of any exhaust point (e.g., procedures for sampling multiple measure- louvered vent). ment sites. 4.1.4 Compartment Housing. Sample im- Example 1: A source with nine circular mediately downstream of the filter bags di- measurement sites of equal areas may be rectly above the tops of the bags as shown in tested as follows: For each test run, traverse the examples in Figure 5D–2. Depending on three measurement sites using four points the housing design, use sampling ports in the per diameter (eight points per measurement housing walls or locate the sampling equip- site). In this manner, test run number 1 will ment within the compartment housing. include sampling from sites 1, 2, and 3; run 2
669
VerDate 11
will include samples from sites 4, 5, and 6; taining isokinetic sampling rates. Note: All and run 3 will include sites 7, 8, and 9. Each sources of gas leakage, into or out of the fab- test area may consist of a separate test of ric filter housing between the inlet measure- each measurement site using eight points. ment site and the outlet measurement site Use the results from all nine tests in deter- must be blocked and made leak-tight. mining the emission average. Velocity determinations at measurement Example 2: A source with 30 rectangular sites with gas velocities within the range measurement sites of equal areas may be measurable with the type S pitot [i.e., veloc- tested as follows: For each of three test runs, ity head >1.3 mm H2O (0.05 in. H2O)] shall be traverse five measurement sites using a 3 x 3 conducted according to the procedures in matrix of traverse points for each site. In Method 2. order to distribute the sampling evenly over 4.4 Sampling. Follow the procedures spec- all the available measurement sites while ified in Section 4.1 of Method 5 or Method 17 sampling only 50 percent of the sites, number with the exceptions as noted above. the sites consecutively from 1 to 30 and sam- 4.5 Sample Recovery. Follow the proce- ple all the even numbered (or odd numbered) dures specified in Section 4.2 of Method 5 or sites. Alternatively, conduct a separate test Method 17. of each of 15 measurement sites using Sec- 4.6 Sample Analysis. Follow the proce- tion 4.2.1 or 4.2.2 to determine the number dures specified in Section 4.3 of Method 5 or and location of traverse points, as appro- Method 17. priate. Example 3: A source with two measure- 4.7 Quality Control Procedures. A QC ment sites of equal areas may be tested as check of the volume metering system at the follows: For each test of three test runs, tra- field site is suggested before collecting the verse both measurement sites using Section sample. Use the procedure defined in Section 4.2.3 in determining number of traverse 4.4 of Method 5. points. Alternatively, conduct two full emis- 5. Calibration sion test runs of each measurement site Follow the procedures as specified in Sec- using the criteria in Section 4.2.1 or 4.2.2 to tion 5 of Method 5 or Method 17. determine the number of traverse points. Other test schemes, such as random deter- 6. Calculations mination of traverse points for a large num- Follow the procedures as specified in Sec- ber of measurement sites, may be used with tion 6 of Method 5 or Method 17 with the ex- prior approval from the Administrator. ceptions as follows: 4.3 Velocity Determination. The velocities 6.1 Total volume flow rate may be deter- of exhaust gases from postitive pressure mined using inlet velocity measurements baghouses are often too low to measure accu- and stack dimensions. rately with the type S pitot specified in 6.2 Average Particulate Concentration. Method 2 [i.e., velocity head <1.3 mm H2O For multiple measurement sites, calculate (0.05 in. H2O)]. For these conditions, measure the average particulate concentration as fol- the gas flow rate at the fabric filter inlet fol- lows: lowing the procedures in Method 2. Calculate the average gas velocity at the measurement n site as follows: ∑ mi i=1 Q To C = Eq. 5 D - 2 v =i • Eq. 5 D - 1 n A T o i ∑ Voli Where: i=1 v¯ =Average gas velocity at the measurement Where: site(s), m/s (ft/s). m =The mass collected for run i of n, mg(gr). Q =Inlet gas volume flow rate, m3/s (ft3/s). i i Vol =The sample volume collected for run i A =Measurement site(s) total cross-sectional i o of n, sm3 (scf). area, m2 (ft2). C¯ =Average concentration of particulate for To=Temperature of gas at measurement site, 3 °K (°R) all n runs, mg/sm (gr/scf). Ti=Temperature of gas at inlet, °K (°R). 7. Bibliography Use the average velocity calculated for the The bibliography is the same as for Method measurement site in determining and main- 5.
670
VerDate 11
671
VerDate 11
METHOD 5E—DETERMINATION OF PARTICULATE emissions from wool fiberglass insulation EMISSIONS FROM THE WOOL FIBERGLASS IN- manufacturing sources. SULATION MANUFACTURING INDUSTRY 1.2 Principle. Particulate matter is with- drawn isokinetically from the source and 1. Applicability and Principle collected on a glass fiber filter maintained at 1.1 Applicability. This method is applica- a temperature in the range of 120 °±14 °C (248 ble for the determination of particulate °±25 °F) and in solutions of 0.1 N NaOH. The filtered particulate mass, which includes any
672
VerDate 11
material that condenses at or above the fil- 3.3.2 Hydrochloric Acid. HCl, concen- tration temperature, is determined gravi- trated, with a dropper. metrically after removal of uncombined 3.3.3 Organic Carbon Stock Solution. Dis- water. The condensed particulate material solve 2.1254 g of dried potassium biphthalate collected in the impinger solutions is deter- in CO2-free water and dilute to 1 liter in a mined as total organic carbon (TOC) using a volumetric flask. This solution contains 1,000 nondispersive infrared type of analyzer. The mg/l organic carbon. sum of the filtered particulate mass and the 3.3.4 Inorganic Carbon Stock Solution. condensed particulate matter is reported as Dissolve 4.404 g anhydrous sodium carbonate the total particulate mass. in about 500 ml of CO2-free water in a 1 liter 2. Apparatus volumetric flask. Add 3.497 g anhydrous so- dium bicarbonate to the flask and dilute to 2.1 Sampling Train. The equipment list 1 liter with CO -free water. This solution for the sampling train is the same as de- 2 contains 1,000 mg/l inorganic carbon. scribed in Section 2.1 of Method 5 except as 3.3.5 Oxygen Gas. CO -free. follows: 2 2.1.1 Probe Liner. Same as described in 4. Procedure Section 2.1.2 of Method 5 except use only 4.1 Sampling. The sampling procedures borosilicate or quartz glass liners. are the same as in Section 4.1 of Method 5 ex- 2.1.2 Filter Holder. Same as described in cept as follows: Section 2.1.5 of Method 5 with the addition of 4.1.1 Filtration Temperature. The tem- a leak-tight connection in the rear half of perature of the filtered gas stream, rather the filter holder designed for insertion of a than the filter compartment air tempera- thermocouple or other temperature gauge ture, is maintained at 120 °±14 °C (248 °±25°F). for measuring the sample gas exist tempera- 4.1.2 Impinger Solutions. 0.1 N NaOH is ture. used in place of water in the impingers. The 2.2 Sample Recovery. The equipment list volumes of the solutions are the same as in for sample recovery is the same as described Method 5. in Section 2.2 of Method 5 except three wash 4.2 Sample Recovery. The sample recov- bottles are needed instead of two and only ery procedure is as follows: glass storage bottles and funnels may be Water is used to rinse and clean the probe used. parts prior to the acetone rinse. Save por- 2.3 Analysis. The equipment list for anal- tions of the water, acetone, and 0.1 N NaOH ysis is the same as Section 2.3 of Method 5 used for cleanup as blanks following the pro- with the additional equipment for TOC anal- cedure as in Section 4.2 of Method 5. ysis as described below: NOTE: All parts of the sample collection 2.3.1 Sample Blender or Homogenizer. portion of the train (e.g., probe and nozzle, Waring type of ultrasonic. filter holder, impinger glassware) must be 2.3.2 Magnetic Stirrer. free of organic solvent residue before sample 2.3.3 Hypodermic Syringe. 0- to 100-µl ca- collection. It is necessary that all sampling pacity. apparatus that have been rinsed with ace- 2.3.4 Total Organic Carbon Analyzer. tone be flushed twice with water or dilute Beckman Model 915 with 215 B infrared ana- NaOH before the sample run. The rinse solu- lyzer or equivalent and a recorder. tions from this cleaning process should be 2.3.5 Beaker. 30 ml. discarded. If other solvents that are not 2.3.6 Water Bath. Temperature-controlled. readily soluble in water (e.g., TCE) are used, 2.3.7 Volumetric Flasks. 1,000 ml and 500 place the exposed sampling apparatus in a ml. drying oven at 105 °C for at least 30 minutes. 3. Reagents Container No. 1. The filter is removed and 3.1 Sampling. The reagents used in sam- stored in the same manner as in Section 4.2 pling are the same as used in Reference of Method 5. Method 5 with the addition of 0.1 N NaOH Container No. 2. Use water to rinse the sam- (dissolve 40 g of ACS reagent grade NaOH in ple nozzle, probe, and front half of the filter distilled water and dilute to 1 liter). holder three times in the manner described 3.2 Sample Recovery. The reagents used in Section 4.2 of Method 5 except that no in sample recovery are the same as used in brushing is done. Put all the wash water in Method 5 with the addition of distilled water one container, seal, and label. and 0.1 N NaOH as described in Section 3.1. Container No. 3. Rinse and brush the sample 3.3 Analysis. The reagents used in analy- nozzle, probe, and front half of the filter sis are the same as in Method 5 except as fol- holder with acetone as described for Con- lows: tainer No. 2 in Section 4.2 of Method 5. 3.3.1 Carbon Dioxide-Free Water. Distilled Container No. 4. Place the contents of the or deionized water that has been freshly silica gel impinger in its original container boiled for 15 minutes and cooled to room as described for Container No. 3 in Section temperature while preventing exposure to 4.2 of Method 5. ambient air with a cover vented with an Container No. 5. Measure the liquid in the ascarite tube. first three impingers and record the volume
673
VerDate 11
or weight as described for the Impinger As samples collected in 0.1 N NaOH often Water in Section 4.2 of Method 5. Do not dis- contain a high measure of inorganic carbon card this liquid, but place it in a sample con- that inhibits repeatable determinations of tainer using a glass funnel to aid in the TOC, sample pretreatment is necessary. transfer from the impingers or graduated Measure and record the liquid volume of cylinder (if used) to the sample container. each sample. If the sample contains solids or Rinse each impinger thoroughly with 0.1 N an immiscible liquid, homogenize the sample NaOH three times, as well as the graduated with a blender or ultrasonics until satisfac- cylinder (if used) and the funnel, and put tory repeatability is obtained. Transfer a these rinsings in the same sample container. representative portion of 10 to 15 ml to a 30- Seal the container and label to identify its ml beaker, acidify with about 2 drops of con- contents clearly. centrated HCl to a pH of 2 or less. Warm the 4.3 Analysis. The procedures for analysis acidified sample at 50 °C (120 °F) in a water are the same as in Section 4.3 of Method 5 bath for 15 minutes. While stirring the sam- with exceptions noted as follows: ple with a magnetic stirrer, withdraw a 20- µ Container No. 1. Determination of weight to 50- l sample from the beaker and inject it gain on the filter is the same as described for into the total carbon port of the analyzer. Container No. 1 in Section 4.3 of Method 5 ex- Measure the peak height. Repeat the injec- cept that the filters must be dried at 20 °±6 °C tions until three consecutive peaks are ob- ± (68 °F±10 °F) and at ambient pressure. tained within 10 percent of the average. Repeat the analyses for all the samples and Containers Nos. 2 and 3. Analyze the con- the 0.1 N NaOH blank. Prepare standard tents of Containers Nos. 2 and 3 as described curves for total carbon and for inorganic car- for Container No. 2 in Section 4.3 of Method bon of 10, 20, 30, 40, 50, 60, 80, and 100 mg/l by 5 except that evaporation of the samples diluting with CO -free water 10, 20, 30, 40, and must be at 20 °±6 °C (68 °±10 °F) and at ambient 2 50 ml of the two stock solutions to 1,000 ml pressure. and 30, 40, and 50 ml of the two stock solu- Container No. 4. Weigh the spent silica gel tions to 500 ml. Inject samples of these solu- as described for Container No. 3 in Section tions into the analyzer and record the peak 4.3 of Method 5. heights as described above. The acidification ‘‘Water and Acetone Blank’’ Containers. De- and warming steps are not necessary for termine the water and acetone blank values preparation of the standard curve. following the procedures for Acetone Blank Ascertain the sample concentrations for Container in Section 4.3 of Method 5. Evapo- the samples from the corrected peak heights rate the samples at ambient temperature for the samples by reference to the appro- [20 °±6 °C (68 °±10 ° F)] and pressure. priate standard curve. Calculate the cor- Container No. 5. For the determination of rected peak height for the standards and the total organic carbon, perform two analyses samples by deducting the blank correction as on successive identical samples, i.e., total follows: carbon and inorganic carbon. The desired Corrected peak height=A¥B quantity is the difference between the two Eq. 5E–1 values obtained. Both analyses are based on conversion of sample carbon into carbon di- Where: oxide for measurement by a nondispersive A=Peak height of standard or sample, mm or infrared analyzer. Results of analyses reg- other appropriate unit. ister as peaks on a strip chart recorder. B=Peak height of blank, mm or other appro- The principal differences between operat- priate unit. ing parameters for the two channels involve If samples must be diluted for analysis, the combustion tube packing material and apply an appropriate dilution factor. temperature. In the total carbon channel, a 5. Calibration high temperature [950 °C (1740 °F)] furnace Calibration of sampling and analysis heats a Hastelloy combustion tube packed equipment is the same as in Section 5 of with cobalt oxide-impregnated asbestos Method 5 with the addition of the calibration fiber. The oxygen in the carrier gas, the ele- of the TOC analyzer described in Section 4.3 vated temperature, and catalytic effect of of this method. the packing result in oxidation of both or- 6. Calculations ganic and inorganic carbonaceous material The calculations and nomenclature for the to CO2 and steam. In the inorganic carbon calculations are the same as described in ° ° channel, a low temperature [150 C (300 F)] Section 6 of Method 5 with the addition of furnace heats a glass tube containing quartz the following: chips wetted with 85 percent phosphoric acid. 6.1 Mass of Condensed Particulate Mate- The acid liberates CO2 and steam from inor- rial Collected. ganic carbonates. The operating temperature
is below that required to oxidize organic Mc= 0.001 Ctoc Vs matter. Follow the manufacturer’s instruc- Eq. 5E–2 tions for assembly, testing, calibration, and operation of the analyzer. Where:
674
VerDate 11
0.001=Liters per milliliter. 2.1 Analysis. mc=Mass of condensed particulate material 2.1.1 Erlenmeyer Flasks. 125-ml, with collected in the impingers measured as ground glass joints. TOC, mg. 2.1.2 Air Condenser. With ground glass Ctoc=Concentration of TOC in the liquid sam- joint compatible with the Erlenmeyer flasks. ple from TOC analysis in Section 4.3, mg/ 2.1.3 Beakers. 250-ml. l. 2.1.4 Volumetric Flasks. 1-liter, 500-ml Vs=Total volume of liquid sample, ml. (one for each sample), 200-ml, and 50-ml (one 6.2 Concentration of Condensed Particu- for each sample and standard). late Material. 2.1.5 Pipets. 5-ml (one for each sample and standard). = 2.1.6 Ion Chromatograph. The ion chro- Cc0./. 001 [] m c Vm() std Eq 5 E - 3 matograph should have at least the following components. Where: 2.1.6.1 Columns. An anion separation or 0.001=Grams per milligram. other column capable of resolving the sulfate Cc=Concentration of condensed particulate ion from other species present and a stand- matter in stack gas, dry basis, corrected ard anion suppressor column. Suppressor col- to standard condition, g/dscm. umns are produced as proprietary items; Vm(std)=Volume of gas sample measured by however, one can be produced in the labora- the dry gas meter, corrected to standard tory using the resin available from BioRad conditions, dscm, from Section 6.3 of Company, 32nd and Griffin Streets, Rich- Method 5. mond, California. Other systems which do 6.3 Total Particulate Concentration. not use suppressor columns may also be = + used. Ct C s C c Eq. 5 E - 4 2.1.6.2 Pump. Capable of maintaining a Where: steady flow as required by the system. 2.1.6.3 Flow Gauges. Capable of measuring Ct=Total particulate concentration, dry basis, corrected to standard conditions, the specified system flow rate. 2.1.6.4 Conductivity Detector. g/dscm. 2.1.6.5 Recorder. Compatible with the out- Cs=Concentration of filtered particulate mat- ter in stack gas, dry basis, corrected to put voltage range of the detector. 3. Reagents. standard conditions, g/dscm, from Equa- The reagents are the same as for Method 5 tion 5–6 of Method 5. with the following exceptions: 7. Bibliography 3.1 Sample Recovery. Water, deionized The bibliography is the same as in Method distilled to conform to American Society for 5 with the addition of the following: Testing and Materials Specification D1193–74, 1. American Public Health Association, Type 3, is needed. At the option of the ana- American Water Works Association, Water lyst, the KMnO4 test for oxidizable organic Pollution Control Federation. Standard matter may be omitted when high con- Methods for the Examination of Water and centrations of organic matter are not ex- Wastewater. Fifteenth Edition. Washington, pected to be present. DC 1980. 3.2 Analysis. The following are required: 3.2.1 Water. Same as in Section 3.1. METHOD 5F—DETERMINATION OF NONSULFATE 3.2.2 Stock Standard Solution, 1 mg PARTICULATE MATTER FROM STATIONARY (NH4)2SO4/ml. Dry an adequate amount of SOURCES primary standard grade ammonium sulfate 1. Applicability and Principle. at 105° to 110 °C for a minimum of 2 hours be- 1.1 Applicability. This method is to be fore preparing the standard solution. Then used for determining nonsulfate particulate dissolve exactly 1.000 g of dried (NH4)2SO4 in matter from stationary sources. Use of this water in a 1-liter volumetric flask, and di- method must be specified by an applicable lute to 1 liter. Mix well. subpart of the standards, or approved by the 3.2.3 Working Standard Solution, 25 µg Administrator, U.S. Environmental Protec- (NH4)2SO4/ml. Pipet 5 ml of the stock stand- tion Agency, for a particular application. ard solution into a 200-ml volumetric flask. 1.2 Principle. Particulate matter is with- Dilute to 200 ml with water. drawn isokinetically from the source using 3.2.4 Eluent Solution. Weigh 1.018 g of so- the Method 5 train at 160 °C (320 °F). The col- dium carbonate (Na2CO3) and 1.008 g of so- lected sample is then extracted with water. dium bicarbonate (NaHCO3), and dissolve in 4 A portion of the extract is analyzed for sul- liters of water. This solution is 0.0024 M fate content. The remainder is neutralized Na2CO3/0.003 M NaHCO3. Other eluents appro- with ammonium hydroxide before it is dried priate to the column type and capable of re- and weighed. solving sulfate ion from other species 2. Apparatus. present may be used. The apparatus is the same as Method 5 3.2.5 Ammonium Hydroxide. Concen- with the following additions. trated, 14.8 M.
675
VerDate 11
3.2.6 Phenolphthalein Indicator. 3,3-Bis(4- After the beakers have cooled, add five hydroxyphenyl)-1-(3H)-isobenzofuranone. drops of phenolphthalein indicator, and then Dissolve 0.05 g in 50 ml of ethanol and 50 ml add concentrated ammonium hydroxide until of water. the solution turns pink. Return the samples 4. Procedure. to the oven at 105 °C, and evaporate the sam- 4.1 Sampling. The sampling procedure is ples to dryness. Cool the samples in a des- the same as Method 5, Section 4.1, except iccator, and weigh the samples to constant that the probe outlet and filter temperatures weight. shall be maintained at 160°±14 °C (320°±25 °F). 4.4 Blanks. 4.2 Sample Recovery. The sample recov- 4.4.1 Filter Blank. Choose a clean filter ery procedure is the same as Method 5, Sec- from the same lot as those used in the test- tion 4.2, except that the recovery solvent ing. Treat the blank filter as a sample, and shall be water instead of acetone. analyze according to Sections 4.3.1 and 4.3.2. 4.3 Analysis. 4.4.2 Water. Transfer a measured volume 4.3.1 Sample Extraction. Cut the filter of water between 100 and 200 ml into a tared into small pieces, and place it in a 125-ml Er- 250-ml beaker. Treat the blank as a sample, lenmeyer flask with a ground glass joint and analyze according to Section 4.3.3. equipped with an air condenser. Rinse the 5. Calibration. shipping container with water, and pour the The calibration procedure is the same as rinse into the flask. Add additional water to Method 5, Section 5, with the following addi- the flask until it contains about 75 ml, and tions: place the flask on a hot plate. Gently reflux 5.1 Standard Calibration Curve. Prepare a the contents for 6 to 8 hours. Cool the solu- series of five standards by adding 1.0, 2.0, 4.0, tion, and transfer it to a 500-ml volumetric 6.0, and 10.0 ml of working standard solution flask. Rinse the Erlenmeyer flask with (25 µg/ml) to a series of five 50-ml volumetric water, and transfer the rinsings to the volu- flasks. (The standard masses will equal 25, metric flask including the pieces of filter. 50, 100, 150, and 250 µg.) Dilute each flask to Transfer the probe rinse to the same 500-ml volume with water, and mix well. Analyze volumetric flask with the filter sample. with the samples as described in Section 4.3. Rinse the sample bottle with water, and add Prepare or calculate a linear regression plot the rinsings to the volumetric flask. Dilute of the standard masses in µg (x-axis) versus the sample to exactly 500 ml with water. their responses (y-axis). (Take peak height 4.3.2 Sulfate (SO4) Analysis. Allow the measurements with symmetrical peaks; in sample to settle until all solid material is at all other cases, calculate peak areas.) From the bottom of the volumetric flask. If nec- this line, or equation, determine the slope, essary, centrifuge a portion of the sample. and calculate its reciprocal which is the cali- Pipet 5 ml of the sample into a 50-ml volu- bration factor, S. If any point deviates from metric flask, and dilute to 50 ml with water. the line by more than 7 percent of the con- Prepare a standard calibration curve accord- centration at that point, remake and reana- ing to Section 5.1. Analyze the set of stand- lyze that standard. This deviation can be de- ards followed by the set of samples using the termined by multiplying S times the re- same injection volume for both standards sponse for each standard. The resultant con- and samples. Repeat this analysis sequence centrations must not differ by more than 7 followed by a final analysis of the standard percent from each known standard mass (i.e., set. Average the results. The two sample val- 25, 50, 100, 150, and 250 µg). ues must agree within 5 percent of their 5.2 Conductivity Detector. Calibrate ac- mean for the analysis to be valid. Perform cording to manufacturer’s specifications this duplicate analysis sequence on the same prior to initial use. day. Dilute any sample and the blank with 6. Calculations. equal volumes of water if the concentration Calculations are the same as Method 5, exceeds that of the highest standard. Section 6, with the following additions: Document each sample chromatogram by 6.1 Nomenclature. listing the following analytical parameters: C =Water blank residue concentration, mg/ Injection point, injection volume, sulfate re- w ml. tention time, flow rate, detector sensitivity setting, and recorder chart speed. F=Dilution factor (required only if sample 4.3.3 Sample Residue. Transfer the re- dilution was needed to reduce the con- maining contents of the volumetric flask to centration into the range of calibration). 2 a tared 250-ml beaker. Rinse the volumetric Hs=Sample response, mm for height or mm flask, and add the rinsings to the tared beak- for area. er. Make certain that all particulate matter Hb=Filter blank response, mm for height or is transferred to the beaker. Evaporate the mm2 for area. water in an oven heated to 105 °C until only mb=Mass of beaker used to dry sample, mg. about 100 ml of water remains. Remove the mf=Mass of sample filter, mg. beakers from the oven, and allow them to mn=Mass of nonsulfate particulate matter, cool. mg.
676
VerDate 11
ms=Mass of ammonium sulfate in the sam- 7.1.3.1 Ion Exchange Column Preparation. ple, mg. Slurry the resin with 1 M HCl in a 250-ml mt=Mass of beaker, filter, and dried sample, beaker, and allow to stand overnight. Place mg. 2.5 cm (1 in.) of glass wool in the bottom of mw=Mass of residue after evaporation of the glass column. Rinse the slurried resin water blank, mg. twice with water. Resuspend the resin in S=Calibration factor, µg/mm. water, and pour sufficent resin into the col- Vb=Volume of water blank, ml. umn to make a bed 5.1 cm (2 in.) deep. Do not Vs=Volume of sample evaporated, 495 ml. allow air bubbles to become entrapped in the 6.2 Water Blank Concentration. resin or glass wool to avoid channeling, which may produce erratic results. If nec- essary, stir the resin with a glass rod to re- = mw Cw Eq. 5 F - 1 move air bubbles. after the column has been Vb prepared, never let the liquid level fall below the top of the upper glass wool plug. Place a 6.3 Mass of Ammonium Sulfate. 2.5-cm (1-in.) plug of glass wool on top of the resin. Rinse the column with water until the − ()99()HHs b eluate gives a pH of 5 or greater as measured = with pH paper. ms Eq. 5 F -2 ()1000 7.1.3.2 Sample Extraction. Follow the pro- cedure given in Section 4.3.1 except do not 6.4 Mass of Nonsulfate Particulate Mat- dilute the sample to 500 ml. ter. 7.1.3.3 Sample Residue. Place at least one mn=mt¥mb¥ms¥mf¥Vs Cw clean glass fiber filter for each sample in a Eq. 5F–3 Buchner funnel, and rinse the filters with water. Remove the filters from the funnel, 7. Alternative Procedures and dry them in an oven at 105 ± 5 °C; then cool in a desiccator. Weigh each filter to 7.1 The following procedure may be used constant weight according to the procedure as an alternative to the procedure in Section in Method 5, Section 4.3. Record the weight 4.3. of each filter to the nearest 0.1 mg. 7.1.1 Apparatus. Same as for Method 6, Assemble the vacuum filter apparatus, and Sections 2.3.3 to 2.3.6 with the following addi- place one of the clean, tared glass fiber fil- tions. ters in the Buchner funnel. Decant the liquid 7.1.1.1 Beakers. 250-ml, one for each sam- portion of the extracted sample (Section ple, and 600-ml. 7.1.3.2) through the tared glass fiber filter 7.1.1.2 Oven. Capable of maintaining tem- into a clean, dry, 500-ml filter flask. Rinse peratures of 75 ± 5 °C and 105 ±5 °C. all the particulate matter remaining in the 7.1.1.3 Buchner Funnel. volumetric flask onto the glass fiber filter 7.1.14 Glass Columns. 25-mm × 305-mm (1- with water. Rinse the particulate matter in. × 12-in.) with Teflon stopcock. with additional water. Transfer the filtrate 7.1.1.5 Volumetric Flasks. 50-ml and 500- to a 500-ml volumetric flask, and dilute to ml, one set for each sample, and 100-ml, 200- 500 ml with water. Dry the filter overnight at ml, and 1000-ml. 105 ± 5 °C, cool in a desiccator, and weigh to 7.1.1.6 Pipettes. Two 20-ml and one 200-ml, the nearest 0.1 mg. one set for each sample, and 5-ml. Dry a 250-ml beaker at 75 ± 5 °C, and cool 7.1.1.7 Filter Flasks. 500-ml. in a desiccator; then weigh to constant 7.1.1.8 Polyethylene Bottle. 500-ml, one weight to the nearest 0.1 mg. Pipette 200 ml for each sample. of the filtrate that was saved into a tared 7.1.2 Reagents. Same as Method 6, Sec- 250-ml beaker; add five drops of tions 3.3.2 to 3.3.5 with the following addi- phenolphtahalein indicator and sufficient tions: concentrated ammonium hydroxide to turn 7.1.2.1 Water, Ammonium Hydroxide, and the solution pink. Carefully evaporate the Phenolphthalein. Same as Sections 3.2.1, contents of the beaker to dryness at 75 ± 5 °C. 3.2.5, and 3.2.6 of this method, respectively. Check for dryness every 30 minutes. Do not 7.1.2.2 Filter. Glass fiber to fit Buchner continue to bake the sample once it has funnel. dried. Cool the sample in a desiccator, and 7.1.2.3 Hydrochloric Acid (HCl), 1 M. Add weigh to constant weight to the nearest 0.1 8.3 ml of concentrated HCl (12 M) to 50 ml of mg. water in a 100-ml volumetric flask. Dilute to 7.1.3.4 Sulfate Analysis. Adjust the flow 100 ml with water. rate through the ion exchange column to 3 7.1.2.4 Glass Wool. ml/min. Pipette a 20-ml aliquot of the fil- 7.1.2.5 Ion Exchange Resin. Strong cation trate onto the top of the ion exchange col- exchange resin, hydrogen form, analytical umn, and collect the eluate in a 50-ml volu- grade. metric flask. Rinse the column with two 15- 7.1.2.6 pH Paper. Range of 1 to 7. ml portions of water. Stop collection of the 7.1.3 Analysis. eluate when the volume in the flask reaches
677
VerDate 11
50-ml. Pipette a 20-ml aliquot of the eluate md = (me ¥ (Vf/Vd) mb) Eq. 5F–5 into a 250-ml Erlenmeyer flask, add 80 ml of 7.1.5.4 Mass of Ammonium Sulfate. 100 percent isopropanol and two to four drops of thorin indicator, and titrate to a pink end ()VVNWVV− point using 0.0100 N barium perchlorate. Re- = t c e f peat and average the titration volumes. Run ms Eq. 5 F- 6 a blank with each series of samples. Rep- VVa i licate titrations must agree within 1 percent or 0.2 ml, whichever is larger. Perform the 7.1.5.5 Mass of Nonsulfate Particulate Mat- ion exchange and titration procedures on du- ter. plicate portions of the filtrate. Results = ± − − should agree within 5 percent. Regenerate or mn m p m d m s m bk Eq. 5 F - 7 replace the ion exchange resin after 20 sam- 8. Bibliography. ple aliquotes have been analyzed or if the end point of the titration becomes unclear. 1. Mulik, J.D. and E. Sawicki. Ion Chromatographic Analysis of Environmental NOTE: Protect the 0.0100 N barium per- Pollutants. Ann Arbor, Ann Arbor Science chlorate solution from evaporation at all Publishers, Inc. Vol. 2. 1979. times. 2. Sawicki, E., J.D. Mulik, and E. 7.1.3.5 Blank Determination. Begin with a sample of water of the same volume as the Wittgenstein. Ion Chromatographic Analysis samples being processed and carry it through of Environmental Pollutants. Ann Arbor, the analysis steps described in Sections Ann Arbor Science Publishers, Inc. Vol. 1. 7.1.3.3 and 7.1.3.4. A blank value larger than 1978. 5 mg should not be subtracted from the final 3. Siemer, D.D. Separation of Chloride and particulate matter mass. Causes for large Bromide From Complex Matrices Prior to blank values should be investigated and any Ion Chromatographic Determination. Ana- problems resolved before proceeding with lytical Chemistry. 52(12):1874–1877. October further analyses. 1980. 7.1.4 Calibration. Calibrate the barium 4. Small, H., T.S. Stevens, and W.C. Bau- perchlorate solutions as in Method 6, Section man. Novel Ion Exchange Chromatographic 5.5. Method Using Conductimetric Determina- 7.1.5 Calculations. tion. Analytical Chemistry. 47(11):1801. 1975. 7.1.5.1 Nomenclature. Same as Section 6.1 with the following additions: METHOD 5G—DETERMINATION OF PARTICULATE
ma = Mass of clean analytical filter, mg. EMISSIONS FROM WOOD HEATERS FROM A DI- md = Mass of dissolved particulate matter, LUTION TUNNEL SAMPLING LOCATION mg. 1. Applicability and Principle me = Mass of beaker and dissolved particu- late matter after evaporation of filtrate, 1.1 Applicability. This method is applica- mg. ble for the determination of particulate mat- m = Mass of insoluble particulate matter, p ter emissions from wood heaters. mg. 1.2 Principle. Particulate matter is with- mr = Mass of analytical filter, sample filter, and insoluble particulate matter, mg. drawn proportionally at a single point from a total collection hood and sampling tunnel mbk = Mass of nonsulfate particulate matter in blank sample, mg. that combines the wood heater exhaust with ambient dilution air. The particulate matter N = Normality of Ba(Cl04)2 titrant, meq/ml. is collected on two glass fiber filters in se- Va = Volume of aliquot taken for titration, 20 ml. ries. The filters are maintained at a tem- ° ° Vc = Volume of titrant used for titration perature of no greater than 32 C (90 F). The blank, ml. particulate mass is determined gravimetri- Vd = Volume of filtrate evaporated, 200 ml. cally after removal of uncombined water. Ve = Volume of eluate collected, 50 ml. There are three sampling train approaches Vf = Volume of extracted sample, 500 ml. described in this method: (1) One dual-filter Vi = Volume of filtrate added to ion exchange dry sampling train operated at about 0.015 column, 20 ml. m3/min, (2) One dual-filter plus impingers Vt = Volume of Ba(Cl04)2 titrant, ml. sampling train operated at about 0.015 m3/ W = Equivalent weight of ammonium sulfate, min, and (3) two dual-filter dry sampling 66.07 mg/meq. trains operated simultaneously at any flow 7.1.5.2 Mass of Insoluble Particulate Mat- rate. Options (2) and (3) are referenced in ter. Section 7 of this method. The dual-filter m= m − m − m Eq. 5 F -4 sampling train equipment and operation, op- p r a f tion (1), are described in detail in this meth- 7.1.5.3 Mass of Dissolved Particulate Mat- od. ter.
678
VerDate 11
2. Apparatus of measuring temperature to within 1.5 per- cent of absolute temperature. 2.1 Sampling Train. The sampling train 2.2 Dilution Tunnel. The dilution tunnel configuration is shown in Figure 5G–1 and consists of the following components: apparatus is shown in Figure 5G–2 and con- 2.1.1 Probe. Stainless steel (e.g., 316 or sists of the following components: grade more corrosion resistant) or glass 2.2.1 Hood. Constructed of steel with a minimum diameter of 0.3 m (1 ft) on the about 95 mm (3⁄8 in.) I.D., 0.6 m (24 in.) in length. If made of stainless steel, the probe large end and a standard 0.15 to 0.3 m (0.5 to shall be constructed from seamless tubing. 1 ft) coupling capable of connecting to stand- 2.1.2 Pitot Tube. Type S, as described in ard 0.15 to 0.3 m (0.5 to 1 ft) stove pipe on the Section 2.1 of Method 2. The Type S pitot small end. tube assembly shall have a known coeffi- 2.2.2 90° Elbows. Steel 90° elbows, 0.15 to cient, determined as outlined in Method 2, 0.3 m (0.5 to 1 ft) in diameter for connecting Section 4. mixing duct, straight duct and damper (op- Alternatively, a standard pitot may be tional) assembly. There shall be at least two used as described in Method 2, Section 2.1. 90° elbows upstream of the sampling section 2.1.3 Differential Pressure Gauge. Inclined (see Figure 5G–2). manometer or equivalent device, as de- 2.2.3 Straight Duct. Steel, 0.15 to 0.3 m (0.5 scribed in Method 2, Section 2.2. One manom- to 1 ft) in diameter to provide the ducting for eter shall be used for velocity head (∆ p) the dilution apparatus upstream of the sam- readings and another (optional) for orifice pling section. Steel duct, 0.15 m (0.5 ft) in di- differential pressure readings (∆ H). ameter shall be used for the sampling sec- 2.1.4 Filter Holders. Two each made of tion. In the sampling section, at least 1.2 m borosilicate glass, stainless steel, or Teflon, (4 ft) downstream of the elbow, shall be two with a glass frit or stainless steel filter sup- holes (velocity traverse ports) at 90 ° to each port and a silicone rubber, Teflon, or Viton other of sufficient size to allow entry of the gasket. The holder design shall provide a pitot for traverse measurements. At least 1.2 positive seal against leakage from the out- m (4 ft) downstream of the velocity traverse side or around the filters. The filter holders ports, shall be one hole (sampling port) of shall be placed in series with the backup fil- sufficient size to allow entry of the sampling ter holder located 25 to 100 mm (1 to 4 in.) probe. Ducts of larger diameter may be used downstream from the primary filter holder. for the sampling section, provided the speci- The filter holder shall be capable of holding fications for minimum gas velocity and the a filter with a 100 mm (4 in.) diameter, ex- dilution rate range shown in Section 4 are cept as noted in Section 7. maintained. The length of duct from the NOTE: Mention of trade names or specific hood inlet to the sampling ports shall not ex- product does not constitute endorsement by ceed 9.1 m (30 ft). the Environmental Protection Agency. 2.2.4 Mixing Baffles. Steel semicircles ° 2.1.5 Filter Temperature Monitoring Sys- (two) attached at 90 to the duct axis on op- tem. A temperature gauge capable of meas- posite sides of the duct midway between the uring temperature to within 1.5 percent of two elbows upstream of sampling section. absolute temperature. The gauge shall be in- The space between the baffles shall be about stalled at the exit side of the front filter 0.3 m (12 in.). holder so that the sensing tip of the tem- 2.2.5 Blower. Squirrel cage or other fan perature gauge is in direct contact with the capable of extracting gas from the dilution sample gas or in a thermowell as shown in tunnel of sufficient flow to maintain the ve- Figure 5G–1. The temperature gauge shall locity and dilution rate specifications in comply with the calibration specifications in Section 4 and exhausting the gas to the at- Method 2, Section 4. Alternatively, the sens- mosphere. ing tip of the temperature gauge may be in- 2.3 Sample Recovery. Probe brushes, wash stalled at the inlet side of the front filter bottles, sample storage containers, petri holder. dishes, and a funnel as described in Method 5, 2.1.6 Dryer. Any system capable of remov- Section 2.2.1 through 2.2.4, and 2.2.8, respec- ing water from the sample gas to less than tively, are needed. 1.5 percent moisture (volume percent) prior 2.4 Analysis. Glass weighing dishes, des- to the metering system. System includes iccator, analytical balance, beakers (250 ml monitor for demonstrating that sample gas or smaller), hygrometer, and temperature temperature is less than 20 °C (68 °F). gauge as described in Method 5, Sections 2.3.1 2.1.7 Metering System. Same as Method 5, through 2.3.3 and 2.3.5 through 2.3.7, respec- Section 2.1.8. tively, are needed. 2.1.8 Barometer. Mercury, aneroid, or 3. Reagents other barometer capable of measuring at- mospheric pressure to within 2.5 mm Hg (0.1 3.1 Sampling. The reagents used in sam- in. Hg). pling are as follows: 2.1.9 Dilution Tunnel Gas Temperature 3.1.1 Filters. Glass fiber filters with a Measurement. A temperature gauge capable minimum diameter of 100 mm (4 in.), without
679
VerDate 11
organic binder, exhibiting at least 99.95 per- the duct area using the average of the two cent efficiency (<0.05 percent penetration) on diameters. A pretest leak-check of pitot 0.3-micron dioctyl phthalate smoke par- lines as in Method 2, Section 3.1, is rec- ticles. Gelman A/E 61631 has been found ac- ommended. Place the calibrated pitot tube ceptable for this purpose. at the centroid of the stack in either of the 3.1.2 Stopcock Grease. Same as Method 5, velocity traverse ports. Adjust the damper or Section 3.1.5. similar device on the blower inlet until the 3.2 Sample Recovery. Acetone-reagent velocity indicated by the pitot is approxi- grade, same as Method 5, Section 3.2. mately 220 m/min (715 fpm). Continue to read 3.3 Analysis. Two reagents are required the ∆ p and temperature until the velocity for the analysis: has remained constant (less than 5 percent 3.3.1 Acetone. As in Section 3.2. change) for 1 minute. Once a constant veloc- 3.3.2 Desiccant. Anhydrous calcium sul- ity is obtained at the centroid of the duct, fate, calcium chloride, or silica gel, indicat- perform a velocity traverse as outlined in ing type. Method 2, Section 3.3 using four points per 4. Procedure traverse as outlined in Method 1. Measure the ∆ p and tunnel temperature at each tra- 4.1 Dilution Tunnel. A schematic of a di- verse point and record the readings. Cal- lution tunnel is shown in Figure 5G–2. The culate the total gas flow rate using calcula- dilution tunnel dimensions and other fea- tions contained in Method 2, Section 5. Ver- tures are described in Section 2.2. Assemble ify that the flow rate is 4 ±0.45 sm3/min the dilution tunnel sealing joints and seams (140±14 scfm); if not, readjust the damper, to prevent air leakage. Clean the dilution and repeat the velocity traverse. The mois- tunnel with an appropriately sized, wire ture may be assumed to be 4 percent (100 per- chimney brush before each certification test. cent relative humidity at 85 °F). Direct mois- 4.1.1 Draft Determination. Prepare the ture measurements such as outlined in EPA wood heater as in Method 28, Section 6.2.1. Method 4 are also permissible. Locate the dilution tunnel hood centrally over the wood heater stack exhaust. Operate NOTE: If burn rates exceed 3 kg/hr (6.6 lb/ the dilution tunnel blower at the flow rate to hr), dilution tunnel duct flow rates greater be used during the test run. Measure the than 4 sm3/min (140 scfm) and sampling sec- draft imposed on the wood heater by the di- tion duct diameters larger than 150 mm (6 lution tunnel (i.e., the difference in draft in.) are allowed. If larger ducts or flow rates measured with and without the dilution tun- are used, the sampling section velocity shall nel operating) as described in Method 28, be at least 220 m/min (715 fpm). In order to Section 6.2.3. Adjust the distance between ensure measurable particulate mass catch, it the top of the wood heater stack exhaust and is recommended that the ratio of the average the dilution tunnel hood so that the dilution mass flow rate in the dilution tunnel to the tunnel induced draft is less than 1.25 Pa average fuel burn rate be less than 150:1 if (0.005 in. H2O). Have no fire in the wood heat- larger duct sizes or flow rates are used. er, close the wood heater doors, and open 4.2.2 Testing Velocity Measurements. fully the air supply controls during this After obtaining velocity traverse results check and adjustment. that meet the flow rate requirements, choose 4.1.2 Smoke Capture. During the pretest a point of average velocity and place the ignition period described in Method 28, Sec- pitot and thermocouple at that location in tion 6.3, operate the dilution tunnel and vis- the duct. Alternatively, locate the pitot and ually monitor the wood heater stack ex- thermocouple at the duct centroid and cal- haust. Operate the wood heater with the culate a velocity correction factor for the doors closed and determine that 100 percent centroidal position. Mount the pitot to en- of the exhaust gas is collected by the dilu- sure no movement during the test run and tion tunnel hood. If less than 100 percent of the wood heater exhaust gas is collected, ad- seal the port holes to prevent any air leak- just the distance between the wood heater age. Align the pitot to be parallel with the stack and the dilution tunnel hood until no duct axis, at the measurement point. Check visible exhaust gas is escaping. Stop the pre- that this condition is maintained during the test ignition period, and repeat the draft de- test run (about 30-minute interva1s). Mon- termination procedure described in Section itor the temperature and velocity during the 4.1.1. pretest ignition period to ensure the proper 4.2 Velocity Measurements. During the flow rate is maintained. Make adjustments pretest ignition period described in Method to the dilution tunnel flow rate as necessary. 28, Section 6.3, conduct a velocity traverse to 4.3 Sampling. identify the point of average velocity. This 4.3.1 Pretest Preparation. It is suggested single point shall be used for measuring ve- that sampling equipment be maintained and locity during the test run. calibrated according to the procedure de- 4.2.1 Velocity Traverse. Measure the di- scribed in APTD–0576. ameter of the duct at the velocity traverse Check and desiccate filters as described in port location through both ports. Calculate Method 5, Section 4.1.1.
680
VerDate 11
4.3.2 Preparation of Collection Train. Dur- on Figure 5G–3 at least once each 10 minutes ing preparation and assembly of the sam- during the test run. Since the manometer pling train, keep all openings where con- level and zero may drift because of vibra- tamination can occur covered until just tions and temperature changes, make peri- prior to assembly or until sampling is about odic checks during the test run. to begin. For the purposes of proportional sampling Using a tweezer or clean disposable sur- rate determinations, data from calibrated gical gloves, place one labeled (identified) flow rate devices, such as glass rotameters, and weighed filter in each of the filter hold- may be used in lieu of incremental dry gas ers. Be sure that each of the filters is prop- meter readings. Proportional rate calcula- erly centered and the gasket properly placed tion procedures must be revised, but accept- so as to prevent the sample gas stream from ability limits remain the same. circumventing the filter. Check each of the During the test run, make periodic adjust- filters for tears after assembly is completed. ments to keep the temperature between (or Mark the probe with heat resistant tape or upstream of) the filters at the proper level. by some other method to denote the proper Do not change sampling trains during the distance into the stack or duct. test run. Set up the train as in Figure 5G–1. At the end of the test run (see Method 28, 4.3.3 Leak-Check Procedures. Section 6.4.6), turn off the coarse adjust 4.3.3.1 Pretest Leak-Check. A pretest valve, remove the probe from the stack, turn leak-check is recommended, but not re- off the pump, record the final dry gas meter quired. If the tester opts to conduct the pre- reading, and conduct a post-test leak-check, test leak-check, conduct the leak-check as as outlined in Section 4.3.3. Also, leak-check described in Method 5, Section 4.1.4.1. A vac- the pitot lines as described in Method 2, Sec- uum 130 mm Hg (5 in. Hg) may be used in- tion 3.1; the lines must pass this leak-check stead of 380 mm Hg (15 in. Hg). in order to validate the velocity head data. 4.3.3.2 Post-Test Leak-Check. A leak- 4.3.6 Calculation of Proportional Sam- check is mandatory at the conclusion of each pling Rate. Calculate percent proportion- test run. The leak-check shall be done in ac- ality (see Calculations, Section 6) to deter- cordance with the procedures described in mine whether the run was valid or another Method 5, Section 4.1.4.1. A vacuum of 130 test run should be made. mm Hg (5 in. Hg) or the greatest vacuum 4.4 Sample Recovery. Begin recovery of measured during the test run, whichever is the probe and filter samples as described in greater, may be used instead of 380 mm Hg Method 5, Section 4.2, except that an acetone (15 in. Hg). blank volume of about 50 ml or more may be 4.3.4 Preliminary Determinations. Deter- used. mine the pressure, temperature and the aver- Treat the samples as follows: age velocity of the tunnel gases as in Section Container No. 1. Carefully remove the filter 4.2. Moisture content of diluted tunnel gases from the primary filter holder and place it in is assumed to be 4 percent for making flow its identified (labeled) petri dish container. rate calculations; the moisture content may Use a pair of tweezers and/or clean disposable be measured directly as in Method 4. surgical gloves to handle the filter. If it is 4.3.5 Sampling Train Operation. Position necessary to fold the filter, do so such that the probe inlet at the stack centroid, and the particulate cake is inside the fold. Care- block off the openings around the probe and fully transfer to the petri dish any particu- porthole to prevent unrepresentative dilu- late matter and/or filter fibers which adhere tion of the gas stream. Be careful not to to the filter holder gasket, by using a dry bump the probe into the stack wall when re- Nylon bristle brush and/or a sharp-edged moving or inserting the probe through the blade. Seal the container. porthole; this minimizes the chance of ex- Container No. 2. Remove the filter from the tracting deposited material. second filter holder using the same proce- Begin sampling at the start of the test run dures as described above. as defined in Method 28, Section 6.4.1. During the test run, maintain a sample flow rate NOTE: The two filters may be placed in the proportional to the dilution tunnel flow rate same container for desiccation and weighing. (within 10 percent of the initial proportion- Use the sum of the filter tare weights to de- ality ratio) and a filter holder temperature termine the sample mass collected. of no greater than 32 °C (90 °F). The initial Container No. 3. Taking care to see that sample flow rate shall be approximately 0.015 dust on the outside of the probe or other ex- m3/min (0.5 cfm). terior surfaces does not get into the sample, For each test run, record the data required quantitatively recover particulate matter or on a data sheet such as the one shown in Fig- any condensate from the probe and filter ure 5G–3. Be sure to record the initial dry gas holders by washing and brushing these com- meter reading. Record the dry gas meter ponents with acetone and placing the wash readings at the beginning and end of each in a labeled (No. 3) glass container. At least sampling time increment and when sampling three cycles of brushing and rinsing are nec- is halted. Take other readings as indicated essary.
681
VerDate 11
Between sampling runs, keep brushes clean inspected and cleaned, if necessary, prior to and protected from contamination. each certification test. After all acetone washings and particulate 5.2 Volume Metering System. matter have been collected in the sample 5.2.1 Initial and Periodic Calibration. Be- containers, tighten the lids on the sample fore its initial use and at least semiannually containers so that the acetone will not leak thereafter, calibrate the volume metering out when transferred to the laboratory system as described in Method 5, Section weighing area. Mark the height of the fluid 5.3.1, except that the wet test meter with a levels to determine whether leakage occurs capacity of 3.0 liters/rev (0.1 ft3/rev) may be during transport. Label the containers clear- used. Other liquid displacement systems ac- ly to identify contents. Requirements for curate to within 1 percent, may be used as capping and transport of sample containers calibration standards. are not applicable if sample recovery and Procedures and equipment specified in analysis occur in the same room. Method 5, Section 7, for alternative calibra- 4.5 Analysis. Record the data required on tion standards, including calibrated dry gas a sheet such as the one shown in Figure 5G– meters and critical orifices, are allowed for 4. Use the same analytical balance for deter- calibrating the dry gas meter in the sam- mining tare weight and final sample weights. pling train. A dry gas meter used as a cali- Handle each sample container as follows: bration standard shall be recalibrated at Containers No. 1 and 2. Leave the contents least once annually. in the sample containers or transfer the fil- 5.2.2 Calibration After Use. After each ters and loose particulate to tared glass certification or audit test (four or more test weighing dishes. Desiccate for no more than runs conducted on a wood heater at the four 36 hours before the initial weighing, weigh to burn rates specified in Method 28), check a constant weight, and report the results to calibration of the metering system by per- the nearest 0.1 mg. For purposes of this sec- forming three calibration runs at a single, tion, the term ‘‘constant weight’’ means a intermediate flow rate as described in Meth- difference of no more than 0.5 mg or 1 per- od 5, Section 5.3.2. cent of total sample weight (less tare Procedures and equipment specified in weight), whichever is greater, between two Method 5, Section 7, for alternative calibra- consecutive weighings, with no less than 2 tion standards are allowed for the post-test hours between weighings. dry gas meter calibration check. Container No. 3. Note the level of liquid in 5.2.3 Acceptable Variation in Calibration. the container, and confirm on the analysis If the dry gas meter coefficient values ob- sheet whether leakage occurred during tained before and after a certification test transport. If a noticeable amount of leakage differ by more than 5 percent, the certifi- has occurred, either void the sample or use cation test shall either be voided and re- methods, subject to the approval of the Ad- peated, or calculations for the certification ministrator, to correct the final results. De- test shall be performed using whichever termination of sample leakage is not appli- meter coefficient value (i.e., before or after) cable if sample recovery and analysis occur gives the lower value of total sample vol- in the same room. Measure the liquid in this ume. container either volumetrically to within 1 5.3 Temperature Gauges. Use the proce- ml or gravimetrically to within 0.5 g. Trans- dure in Method 2, Section 4.3, to calibrate fer the contents to a tared 250 ml or smaller temperature gauges before the first certifi- beaker and evaporate to dryness at ambient cation or audit test and at least semiannu- temperature and pressure. Desiccate and ally, thereafter. weigh to a constant weight. Report the re- 5.4 Leak-Check of Metering System sults to the nearest 0.1 mg. Shown in Figure 5G–1. That portion of the ‘‘Acetone Blank’’ Container. Measure ace- sampling train from the pump to the orifice tone in this container either volumetrically meter shall be leak-checked prior to initial or gravimetrically. Transfer the acetone to a use and after each certification or audit test. tared 250 ml or smaller beaker and evaporate Leakage after the pump will result in less to dryness at ambient temperature and pres- volume being recorded than is actually sam- sure. Desiccate and weigh to a constant pled. Use the procedure described in Method weight. Report the results to the nearest 0.1 5, Section 5.6. mg. Similar leak-checks shall be conducted for other types of metering systems (i.e., with- 5. Calibration out orifice meters). Maintain a laboratory record of all calibra- 5.5 Barometer. Calibrate against a mer- tions. cury barometer before the first certification 5.1 Pitot Tube. The Type S pitot tube as- test and at least semiannually, thereafter. If sembly shall be calibrated according to the a mercury barometer is used, no calibration procedure outlined in Method 2, Section 4, is necessary. Follow the manufacturer’s in- prior to the first certification test and structions for operation. checked semiannually, thereafter. A stand- 5.6 Analytical Balance. Perform a ard pitot need not be calibrated but shall be multipoint calibration (at least five points
682
VerDate 11
spanning the operational range) of the ana- Tm=Absolute average dry gas meter tempera- lytical balance before the first certification ture (see Figure 5G–3), o K (o R).
test and semiannually, thereafter. Before Tmi=Absolute average dry gas meter tem- each certification test, audit the balance by perature during each 10-minute interval, weighing at least one calibration weight i, of the test run, o K (o R). (class F) that corresponds to 50 to 150 percent Ts=Absolute average gas temperature in the of the weight of one filter. If the scale can- dilution tunnel (see Figure 5G 3), o K (o not reproduce the value of the calibration R). weight to within 0.1 mg, conduct the Tsi=Absolute average gas temperature in the multipoint calibration before use. dilution tunnel during each 10 minute in- 6. Calculations terval, i, of the test run, o K (o R). o Tstd=Standard absolute temperature, 293 K Carry out calculations, retaining at least (528 o R). one extra decimal figure beyond that of the V =Volume of acetone blank, ml. acquired data. Round off figures after the a V =Volume of acetone used in wash, ml. final calculation. Other forms of the equa- aw tions may be used as long as they give equiv- Vm=Volume of gas sample as measured by alent results. dry gas meter, dm3 (dcf). 6.1 Nomenclature. Vmi=Volume of gas sample as measured by B =Water vapor in the gas stream, propor- dry gas meter during each 10-minute in- ws 3 tion by volume (assumed to be 0.04). terval, i, of the test run, dm (dcf). Vm(std)=Volume of gas sample measured by cs=Concentration of particulate matter in stack gas, dry basis, corrected to stand- the dry gas meter, corrected to standard ard conditions, g/dsm3 (g/dscf). conditions, dsm3 (dscf). E=Particulate emission rate, g/hr. Vs=Average gas velocity in dilution tunnel, La=Maximum acceptable leakage rate for ei- calculated by Method 2, Equation 2–9, m/ ther a pretest or post-test leak-check, sec (ft/sec). The dilution tunnel dry gas equal to 0.00057 m3/min (0.02 cfm) or 4 per- molecular weight may be assumed to be cent of the average sampling rate, which- 29 g/g mole (lb/lb mole). ever is less. Vsi=Average gas velocity in dilution tunnel Lp= Leakage rate observed during the post- during each 10-minute interval, i, of the test leak-check, m3/min (cfm). test run, calculated by Method 2, Equa- ma=Mass of residue of acetone blank after tion 2–9, m/sec (ft/sec). evaporation, mg. Y=Dry gas meter calibration factor. maw=Mass of residue from acetone wash after ∆ H=Average pressure differential across the evaporation, mg. orifice meter, if used (see Figure 5G–2), mn=Total amount of particulate matter col- mm H O (in. H O). lected, mg. 2 2 Θ=Total sampling time, min. Mw=Molecular weight of water, 18.0 g/g-mole (18.0 lb/lb-mole). 10=10 minutes, length of first sampling pe- riod. Pbar=Barometric pressure at the sampling site, mm Hg (in. Hg). 13.6=Specific gravity of mercury. PR=Percent of proportional sampling rate. 100=Conversion to percent. Ps=Absolute gas pressure in dilution tunnel, 6.2 Dry Gas Volume. Correct the sample mm Hg (in. Hg). volume measured by the dry gas meter to Pstd=Standard absolute pressure, 760 mm Hg standard conditions (20 °C, 760 mm Hg or (29.92 in. Hg). 68°F, 29.92 in. Hg) by using Equation 5G–1. (If Qsd=Average gas flow rate in dilution tunnel, no orifice meter is used in sampling train, calculated as in Method 2, Equation 2–10, assume ∆ H=O or measure static pressure at dsm3/hr (dscf/hr). dry gas meter outlet.)
+ ()∆ + () ∆ = Tstd PHbar /./.13 6= PHbar 13 6 VVYm() std m KVY1 m Eq. 5 G- 1 Tm Pstd Tm
where; 6.3 Solvent Wash Blank. o Kl=0.3858 K/mm Hg for metric units. =17.64 o R/in. Hg for English units. m V NOTE: If L exceeds L , Equation 5G–1 must = a aw p a m aw Eq. 5 G- 2 be modified as follows: Replace Vm in Equa- tion 5G–1 with the expression: Va
[Vm¥(Lp¥La)Θ] 683
VerDate 11
6.4 Total Particulate Weight. Determine 7.2 Dual Sampling Trains. The tester may the total particulate catch, mn, from the sum operate two sampling trains simultaneously of the weights obtained from Containers 1, 2, at sample flow rates other than that speci- and 3, less the acetone blank (see Figure 5G– fied in Section 4.3.5 provided the following 4). specifications are met. 6.5 Particulate Concentration. 7.2.1 Sampling Train. The sampling train configuration shall be the same as specified cs=(0.001 g/mg) (mn/Vm(std)) in Section 2.1, except the probe, filter, and Eq. 5G–3 filter holder need not be the same sizes as 6.6 Particulate Emission Rate. specified in the applicable sections. Filter E=cs Qsd holders of plastic materials such as Nalgene Eq. 5G–4 or polycarbonate materials may be used (the Gelman 1119 filter holder has been found NOTE: Particulate emission rate results suitable for this purpose). With such mate- produced using the sampling train described rials, it is recommended not to use solvents in Section 2 and shown in Figure 5G–1 shall in sample recovery. The filter face velocity be adjusted for reporting purposes by the fol- shall not exceed 150 mm/sec (30 ft/min) dur- lowing methods adjustment factor: ing the test run. The dry gas meter shall be 0.83 Eadj=1.82 (E) calibrated for the same flow rate range as Eq. 5G–5 encountered during the test runs. Two sepa- rate, complete sampling trains are required 6.7 Proportional Rate Variation. Cal- for each test run. culate PR for each 10-minute interval, i, of 7.2.2 Probe Location. Locate the two the test run. probes in the dilution tunnel at the same level (see Section 2.2.3). Two sample ports θ ()Vmi v s T m T si are necessary. Locate the probe inlets within PR = ×100Eq. 5 G- 6 the 50 mm (2 in.) diameter centroidal area of 10()Vm v si T s T mi the dilution tunnel no closer than 25 mm (1 in.) apart. Alternate calculation procedures for pro- 7.2.3 Sampling Train Operation. Operate portional rate variation may be used if other the sampling trains as specified in Section sample flow rate data (e.g., orifice flow me- 4.3.5, maintaining proportional sampling ters or rotameters) are monitored to main- rates and starting and stopping the two sam- tain proportional sampling rates. The pro- pling trains simultaneously. The pitot values portional rate variations shall be calculated as described in Section 4.2.2 shall be used to for each 10-minute interval by comparing the adjust sampling rates in both sampling stack to nozzle velocity ratio for each 10- trains. minute interval to the average stack to noz- 7.2.4 Recovery and Analysis of Sample. zle velocity ratio for the test run. Propor- Recover and analyze the samples from the tional rate variation may be calculated for two sampling trains separately, as specified intervals shorter than 10 minutes with ap- in Sections 4.4 and 4.5. propriate revisions to Equation 5G–6. For this alternative procedure, the probe 6.8 Acceptable Results. If no more than 10 and filter holder assembly may be weighed percent of the PR values for all the intervals without sample recovery (use no solvents) exceed 90 percent ≤PR ≤110 percent, and if no described above in order to determine the PR value for any interval exceeds 80 percent sample weight gains. For this approach, ≤PR ≤120 percent, the results are acceptable. weigh the clean, dry probe and filter holder If the PR values for the test run are judged assembly upstream of the front filter (with- to be unacceptable, report the test run emis- out filters) to the nearest 0.1 mg to establish sion results, but do not include the results in the tare weights. The filter holder section calculating the weighted average emission between the front and second filter need not rate, and repeat the test run. be weighed. At the end of the test run, care- fully clean the outside of the probe, cap the 7. Alternative Sampling and Analysis Procedure ends, and identify the sample (label). Re- 7.1 Method 5H Sampling Train. The sam- move the filters from the filter holder as- pling and analysis train and procedures de- semblies as described for containers Nos. 1 scribed in Method 5H, Sections 2.1, 3.1, 3.2, and 2 above. Reassemble the filter holder as- 5.1, 5.2.3, 5.3, and 5.6 may be used in lieu of sembly, cap the ends, identify the sample similar sections in Method 5G. Operation of (label), and transfer all the samples to the the Method 5H sampling train in the dilution laboratory weighing area for final weighing. tunnel is as described in Section 4.3.5 of this Descriptions of capping and transport of method. Filter temperatures and condenser samples are not applicable if sample recov- conditions are as described in Method 5H. No ery and analysis occur in the same room. methods adjustment factor as described in For this alternative procedure, filters may Equation 5G–5, Section 6.6, is to be applied to be weighed directly without a petri dish. If the particulate emission rate data produced the probe and filter holder assemb1y are to by this alternative method. be weighed to determine the sample weight,
684
VerDate 11
rinse the probe with acetone to remove mois- age emission rate. Repeat the test run until ture before desiccating prior to the test run. acceptable results are achieved, report the Following the test run, transport the probe average emission rate for the acceptable test and fi1ter ho1der to the dessicator, and run, and use the average in calculating the uncap the openings of the probe and the fil- weighted average emission rate. ter holder assembly. Desiccate no more than 36 hours and weigh to a constant weight. Re- 8. Bibliography port the results to the nearest 0.l mg. 1. Same as for Method 5, citations 1 7.2.5 Calculations. Calculate an emission through 11, with the addition of the follow- rate (Section 6.6) for the sample from each ing: sampling train separately and determine the 2. Oregon Department of Environmental average emission rate for the two values. Quality Standard Method for Measuring the The two emission rates shall not differ by Emissions and Efficiencies of Woodstoves, more than 7.5 percent from the average emis- June 8, 1984. Pursuant to Oregon Administra- sion rate, or 7.5 percent of the weighted aver- tive Rules Chapter 340, Division 21. age emission rate limit in the applicable 3. American Society for Testing Materials. standard, whichever is greater. If this speci- Proposed Test Methods for Heating Perform- fication is not met, the results are unaccept- ance and Emissions of Residential Wood- able. Report the results, but do not include fired Closed Combustion-Chamber Heating the results in calculating the weighted aver- Appliances. E–6 Proposal P 180. August 1986.
685
VerDate 11
686
VerDate 11
Stove ———————————————————— Liquid lost during Date ————————————————————— transport, ml ———————————————— Run No. ——————————————————— Acetone blank volume, ml ————————— Filter Nos. ————————————————— Acetone wash volume, ml ——————————
687
VerDate 11
Acetone blank 2.1.1 Probe Nozzle. (Optional) Same as concentration, mg/mg ——————————— Method 5, Section 2.1.1. A straight sampling Acetone wash blank, mg —————————— probe without a nozzle is an acceptable alter- native. Weight of particulate col- 2.1.2 Probe Liner. Same as Method 5, Sec- lected, mg Container No. tion 2.1.2, except that the maximum length Final Tare Weight of the sample probe shall be 0.6 m (2 ft) and weight weight gain probe heating is optional. 2.1.3 Differential Pressure Gauge. Same as 1...... 2...... Method 5, Section 2.1.4. 3...... 2.1.4 Filter Holders. Two each of borosili- cate glass, with a glass frit or stainless steel Total ...... filter support and a silicone rubber, Teflon, Less acetone blank ...... or Viton gasket. The holder design shall pro- Weight of particulate matter ...... vide a positive seal against leakage from the outside or around the filter. The front filter STACK MOISTURE MEASUREMENT DATA holder shall be attached immediately at the outlet of the probe and prior to the first im- (OPTIONAL) pinger. The second filter holder shall be at- Volume of liquid water tached on the outlet of the third impinger collected and prior to the inlet of the fourth (silica gel) impinger. Impinger Silica gel volume, ml weight, g NOTE: Mention of trade names or specific product does not constitute endorsement by Final ...... the Environmental Protection Agency. Initial ...... Liquid collected ...... 2.1.5 Filter Heating System. Same as Total volume collected ...... g1 ml Method 5, Section 2.1.6. 1 Convert weight of water to volume by dividing total weight 2.1.6 Condenser. Same as Method 5, Sec- increase by density of water (1 g/ml). tion 2.1.7, used to collect condensible mate- rials and determine the stack gas moisture Increase, g content. = Volume water, ml 2.1.7 Metering System. Same as Method 5, ()1g/ ml Section 2.1.8. 2.1.8 Barometer. Mercury, aneroid, or Figure 5G–4. Analysis data sheet. other barometer capable of measuring at- mospheric pressure to within 2.5 mm Hg (0.1 METHOD 5H—DETERMINATION OF PARTICULATE in. Hg). EMISSIONS FROM WOOD HEATERS FROM A 2.2 Stack Flow Rate Measurement Sys- STACK LOCATION tem. A schematic of an example test system Applicability and Principle is shown in Figure 5H–2. The flow rate meas- urement system consists of the following 1.1 Applicability. This method is applica- components: ble for the determination of particulate mat- 2.2.1 Sample Probe. A glass or stainless ter and condensible emissions from wood steel sampling probe. heaters. 2.2.2 Gas Conditioning System. A high 1.2 Principle. Particulate matter is with- density filter to remove particulate matter drawn proportionally from the wood heater and a condenser capable of lowering the dew exhaust and is collected on two glass fiber point of the gas to less than 5 °C (40 °F). Des- filters separated by impingers immersed in iccant, such as Drierite, may be used to dry an ice bath. The first filter is maintained at the sample gas. Do not use silica gel. ° a temperature of no greater than 120 C (248 2.2.3 Pump. An inert (i.e., Teflon or stain- ° F). The second filter and the impinger sys- less steel heads) sampling pump capable of tem are cooled such that the exiting tem- delivering more than the total amount of ° perature of the gas is no greater than 20 C sample required in the manufacturer’s in- ° (68 F). The particulate mass collected in the structions for the individual instruments. A probe, on the filters, and in the impingers is means of controlling the analyzer flow rate determined gravimetrically after removal of and a device for determining proper sample uncombined water. flow rate (e.g., precision rotameter, pressure 2. Apparatus gauge downstream of all flow controls) shall be provided at the analyzer. The require- 2.1 Sampling Train. The sampling train ments for measuring and controlling the an- configuration is shown in Figure 5H–1. alyzer flow rate are not applicable if data are APTD–0576 is suggested for operating and presented that demonstrate the analyzer is maintenance procedures. The train consists insensitive to flow variations over the range of the following components: encountered during the test.
688
VerDate 11
2.2.4 CO Analyzer. Any analyzer capable 2.5 Analysis. Weighing dishes, desiccator, of providing a measure of CO in the range of analytical balance, beakers (250 ml or less), 0 to 10 percent by volume at least once every hygrometer or psychrometer, and tempera- 10 minutes. ture gauge as described in Method 5, Sec- 2.2.5 CO2 Analyzer. Any analyzer capable tions 2.3.1 through 2.3.7, respectively, are of providing a measure of CO2 in the range of needed. In addition, a separatory funnel, 0 to 25 percent by volume at least once every glass or Teflon, 500 ml or greater, is needed. 10 minutes. 3. Reagents NOTE: Analyzers with ranges less than those specified above may be used provided 3.1 Sampling. The reagents used in sam- actual concentrations do not exceed the pling are as follows: range of the analyzer. 3.1.1 Filters. Glass fiber filters, without organic binder, exhibiting at least 99.95 per- 2.2.6 Manifold. A sampling tube capable of cent efficiency (<0.05 percent penetration) on delivering the sample gas to two analyzers 0.3-micron dioctyl phthalate smoke par- and handling an excess of the total amount ticles. Gelman A/E 61631 filters have been used by the analyzers. The excess gas is ex- found acceptable for this purpose. hausted through a separate port. 3.1.2 Silica Gel. Same as Method 5, Sec- 2.2.7 Recorders (optional). To provide a tion 3.1.2. permanent record of the analyzer outputs. 3.1.3 Water. Deionized distilled to conform 2.3 Proportional Gas Flow Rate System. to ASTM Specification D1193–77, Type 3 (in- To monitor stack flow rate changes and pro- corporated by reference—see § 60.17). Run vide a measurement that can be used to ad- blanks prior to field use to eliminate a high just and maintain particulate sampling flow blank on test samples. rates proportional to the stack flow rate. A 3.1.4 Crushed Ice. schematic of the proportional flow rate sys- 3.1.5 Stopcock Grease. Same as Method 5, tem is shown in Figure 5H–2 and consists of Section 3.1.5. the following components: 3.2 Sample Recovery. Same as Method 5, 2.3.1 Tracer Gas Injection System. To in- Section 3.2. ject a known concentration of SO2 into the 3.3 Cylinder Gases. For the purposes of flue. The tracer gas injection system con- this procedure, span value is defined as the sists of a cylinder of SO2, a gas cylinder reg- upper limit of the range specified for each ulator, a stainless steel needle valve or flow analyzer as described in Section 2.2 or 2.3. If controller, a nonreactive (stainless steel and an analyzer with a range different from that glass) rotameter, and an injection loop to specified in this method is used, the span disperse the SO2 evenly in the flue. value shall be equal to the upper limit of the 2.3.2 Sample Probe. A glass or stainless range for the analyzer used (see NOTE in Sec- steel sampling probe. tion 2.2.5). 2.3.3 Gas Conditioning System. A combus- 3.3.1 Calibration Gases. The calibration tor as described in Method 16A, Sections 2.1.5 gases for the CO2, CO and SO2 analyzers shall and 2.1.6, followed by a high density filter to be CO2, CO, or SO2, as appropriate, in N2. CO2 remove particulate matter, and a condenser and CO calibration gases may be combined in capable of lowering the dew point of the gas a single cylinder. to less than 5 °C (40 °F). Desiccant, such as There are two alternatives for checking Drierite, may be used to dry the sample gas. the concentrations of the calibration gases. Do not use silica gel. (a) The first is to use calibration gases that 2.3.4 Pump. As described in Section 2.2.3. are documented traceable to National Bu- reau of Standards Reference Materials. Use 2.3.5 SO2 Analyzer. Any analyzer capable Tracebility Protocol for Establishing True Con- of providing a measure of the SO2 concentra- tion in the range of 0 to 1,000 ppm by volume centrations of Gases Used for Calibrations and Audits of Continuous Source Emission Monitors (or other range necessary to measure the SO2 concentration) at least once every 10 min- (Protocol Number 1) that is available from utes. the Environmental Monitoring and Support 2.3.6 Recorder (optional). To provide a Laboratory, Quality Assurance Branch, Mail permanent record of the analyzer outputs. Drop 77, Environmental Protection Agency, Research Triangle Park, North Carolina NOTE: Other tracer gas systems, including 27711. Obtain a certification from the gas helium gas systems, are allowed for deter- manufacturer that the protocol was fol- mining instantaneous proportional sampling lowed. These calibration gases are not to be rates. analyzed with the test methods. (b) The sec- 2.4 Sample Recovery. Probe liner and ond alternative is to use calibration gases probe nozzle brushes, wash bottles, sample not prepared according to the protocol. If storage containers, petri dishes, graduated this alternative is chosen, within 6 months cylinder or balance, plastic storage contain- prior to the certification test, analyze each ers, funnel and rubber policeman, as de- of the CO2 and CO calibration gas mixtures scribed in Method 5, Sections 2.2.1 through in triplicate using Method 3, and within 1 2.2.8, respectively, are needed. month prior to the certification test, analyze
689
VerDate 11
SO2 calibration gas mixtures using Method 6. 2.5 percent of the span value over the period For the low-level, mid-level, or high-level of the test run. gas mixtures, each of the individual SO2 ana- 4.4 Resolution. The resolution of the out- lytical results must be within 10 percent (or put for each analyzer shall be 0.5 percent of 10 ppm, whichever is greater) of the trip- span value or less. licate set average; CO2 and CO test results 4.5 Calibration Error. The linear calibra- must be within 0.5 percent CO2 and CO; oth- tion curve produced using the zero and mid- erwise, discard the entire set and repeat the level calibration gases shall predict the ac- triplicate analyses. If the average of the trip- tual response to the low-level and high-level licate test method results is within 5 percent calibration gases within 2 percent of the for SO2 gas (or 0.5 percent CO2 and CO for the span value. CO2 and CO gases) of the calibration gas manufacturer’s tag values, use the tag value; 5. Procedure otherwise, conduct at least three additional 5.1 Pretest Preparation. test method analyses until the results of six 5.1.1 Filter and Desiccant. Same as Meth- individual SO2 runs (the three original plus od 5, Section 4.1.1. three additional) agree within 10 percent (or 5.1.2 Sampling Probe and Nozzle. The 10 ppm, whichever is greater) of the average sampling location for the particulate sam- (CO2 and CO test results must be within 0.5 pling probe shall be 2.45±0.15 m (8±0.5 ft) percent). Then use this average for the cyl- above the platform upon which the wood inder value. Four calibration gas levels are heater is placed (i.e., the top of the scale). required as specified below: Select a nozzle, if used, sized for the range 3.3.1.1 High-level Gas. A gas concentra- of velocity heads, such that it is not nec- tion that is equivalent to 80 to 90 percent of essary to change the nozzle size in order to the span value. maintain proportional sampling rates. Dur- 3.3.1.2 Mid-level Gas. A gas concentration ing the run, do not change the nozzle size. that is equivalent to 45 to 55 percent of the Select a suitable probe liner and probe span value. length to effect minimum blockage. 3.3.1.3 Low-level Gas. A gas concentration 5.1.3 Preparation of Particulate Sampling that is equivalent to 20 to 30 percent of the Train. During preparation and assembly of span value. the particulate sampling train, keep all 3.3.1.4 Zero Gas. A gas concentration of openings where contamination can occur less than 0.25 percent of the span value. Puri- covered until just prior to assembly or until fied air may be used as zero gas for the CO2, sampling is about to begin. CO, and SO2 analyzers. Place 100 ml of water in each of the first 3.3.2 SO2 Injection Gas. A known con- two impingers, leave the third impinger centration of SO2 in N2. The concentration empty, and transfer approximately 200 to 300 must be at least 2 percent SO2 with a maxi- g of preweighed silica gel from its container mum of 100 percent SO2. The cylinder con- to the fourth impinger. More silica gel may centration shall be certified by the manufac- be used, but care should be taken to ensure turer to be within 2 percent of the specified that it is not entrained and carried out from concentration. the impinger during sampling. Place the con- 3.4 Analysis. Three reagents are required tainer in a clean place for later use in the for the analysis: sample recovery. Alternatively, the weight 3.4.1 Acetone. Same as 3.2. of the silica gel plus impinger may be deter- 3.4.2 Dichloromethane (Methylene Chlo- mined to the nearest 0.5 g and recorded. ride). Reagent grade, <0.001 percent residue Using a tweezer or clean surgical gloves, in glass bottles. place one labeled (identified) and weighed fil- 3.4.3 Desiccant. Anhydrous calcium sul- ter in each of the filter holders. Be sure that fate, calcium chloride, or silica gel, indicat- each of the filters is properly centered and ing type. the gasket properly placed so as to prevent the sample gas stream from circumventing 4. Gas Measurement System Performance the filter. Check the filters for tears after as- Specifications. sembly is completed. 4.1 Response Time. The amount of time When glass liners are used, install the se- required for the measurement system to dis- lected nozzle using a Viton A O-ring. Other play 95 percent of a step change in gas con- connecting systems using either 316 stainless centration. The response time for each ana- steel or Teflon ferrules may be used. Mark lyzer and gas conditioning system shall be the probe with heat resistant tape or by no more than 2 minutes. some other method to denote the proper dis- 4.2 Zero Drift. The zero drift value for tance into the stack or duct. each analyzer shall be less than 2.5 percent Set up the train as in Figure 5H 1, using (if of the span value over the period of the test necessary) a very light coat of silicone run. grease on all ground glass joints, greasing 4.3 Calibration Drift. The calibration drift only the outer portion (see APTD–0576) to value measured with the mid-level calibra- avoid possibility of contamination by the sil- tion gas for each analyzer shall be less than icone grease.
690
VerDate 11
Place crushed ice around the impingers. 5.2.2.1 Test Fuel Charge Weight. Record 5.1.4 Leak-Check Procedures. the test fuel charge weight in kilograms 5.1.4.1 Pretest Leak-Check. A pretest (wet) as specified in Method 28, Section 6.4.2. leak-check is recommended, but not re- The wood is assumed to have the following quired. If the tester opts to conduct the pre- weight percent composition: 51 percent car- test leak-check, conduct the leak-check as bon, 7.3 percent hydrogen, 41 percent oxygen. described in Method 5, Section 4.1.4.1, except Record the wood moisture for each wood that a vacuum of 130 mm Hg (5 in. Hg) may charge as described in Method 28, Section be used instead of 380 mm Hg (15 in. Hg). 6.2.5. The ash is assumed to have negligible 5.1.4.2 Leak-Checks During Sample Run. effect on associated C, H, O concentrations If, during the sampling run, a component after the test burn. (e.g., filter assembly or impinger) change be- 5.2.2.2 Measured Values. Record the CO comes necessary, conduct a leak-check as de- and CO2 concentrations in the stack on a dry scribed in Method 5, Section 4.1.4.2. basis every 10 minutes during the test run or 5.1.4.3 Post-Test Leak-Check. A leak- more often at the option of the tester. Aver- check is mandatory at the conclusion of each age these values for the test run. Use as a sampling run. The leak-check shall be done mole fraction (e.g., 10 percent CO is recorded in accordance with the procedures described 2 as 0.10) in the calculations to express total in Method 5, Section 4.1.4.3, except that a vacuum of 130 mm Hg (5 in. Hg) or the great- flow Equation 5H–7. est vacuum measured during the test run, 5.2.3 Particulate Train Operation. For whichever is greater, may be used instead of each run, record the data required on a data 380 mm Hg (15 in. Hg). sheet such as the one shown in Figure 5H–3. 5.1.5 Tracer Gas Procedure. A schematic Be sure to record the initial dry gas meter of the tracer gas injection and sampling sys- reading. Record the dry gas meter readings tems is shown in Figure 5H–2. at the beginning and end of each sampling time increment, when changes in flow rates 5.1.5.1 SO2 Injection Probe. Install the SO2 injection probe and dispersion loop in the are made, before and after each leak-check, stack at a location 2.8±0.15 m (9.5±0.5 ft) and when sampling is halted. Take other above the sampling platform. readings as indicated on Figure 5H–3 at least 5.1.5.2 SO2 Sampling Probe. Install the once each 10 minutes during the test run. SO2 sampling probe at the centroid of the Remove the nozzle cap, verify that the fil- stack at a location 4±0.15 m (13.5±0.5 ft) above ter and probe heating systems are up to tem- the sampling platform. perature, and that the probe is properly posi- 5.1.6 Flow Rate Measurement System. A tioned. Position the nozzle, if used, facing schematic of the flow rate measurement sys- into gas stream, or the probe tip in the 50 tem is shown in Figure 5H–2. Locate the flow mm (2 in.) centroidal area of the stack. rate measurement sampling probe at the Be careful not to bump the probe tip into centroid of the stack at a location 2.3±0.3 m the stack wall when removing or inserting (7.5±1 ft) above the sampling platform. the probe through the porthole; this mini- 5.2 Test Run Procedures. The start of the mizes the chance of extracting deposited ma- test run is defined as in Method 28, Section terial. 6.4.1. When the probe is in position, block off the 5.2.1 Tracer Gas Procedure. Within 1 openings around the probe and porthole to minute after closing the wood heater door at prevent unrepresentative dilution of the gas the start of the test run, meter a known con- stream. centration of SO2 tracer gas at a constant Begin sampling at the start of the test run flow rate into the wood heater stack. Mon- as defined in Method 28, Section 6.4.1, start itor the SO concentration in the stack, and 2 the sample pump, and adjust the sample flow record the SO concentrations at 10-minute 2 rate to between 0.003 and 0.015 m3/min (0.1 intervals or more often at the option of the and 0.5 cfm). Adjust the sample flow rate tester. Adjust the particulate sampling flow proportionally to the stack flow during the rate proportionally to the SO concentration 2 test run (Section 5.2.1), and maintain a pro- changes using Equation 5H–6 (e.g., the SO2 concentration at the first 10-minute reading portional sampling rate (within 10 percent of is measured to be 100 ppm; the next 10 the desired value) and a filter holder tem- perature no greater than 120 °C (248 °F). minute SO2 concentration is measured to be 75 ppm: the particulate sample flow rate is During the test run, make periodic adjust- adjusted from the initial 0.15 cfm to 0.20 ments to keep the temperature around the cfm). A check for proportional rate variation filter holder at the proper level. Add more shall be made at the completion of the test ice to the impinger box and, if necessary, run using Equation 5H–10. salt to maintain a temperature of less than 5.2.2 Volumetric Flow Rate Procedure. 20 °C (68 °F) at the condenser/silica gel out- Apply stoichiometric relationships to the let. wood combustion process in determining the If the pressure drop across the filter be- exhaust gas flow rate as follows: comes too high, making sampling difficult to
691
VerDate 11
maintain, either filter may be replaced dur- if sample recovery and analysis occur in the ing a sample run. It is recommended that an- same room. other complete filter assembly be used rath- Container No. 5. Transfer the silica gel from er than attempting to change the filter the fourth impinger to its original container itself. Before a new filter assembly is in- and seal. A funnel may make it easier to stalled, conduct a leak-check (see Section pour the silica gel without spilling. A rubber 5.1.4.2). The total particulate weight shall in- policeman may be used as an aid in removing clude the summation of all filter assembly the silica gel from the impinger. It is not catches. The total time for changing sample necessary to remove the small amount of train components shall not exceed 10 min- dust particles that may adhere to the im- utes. No more than one component change is pinger wall and are difficult to remove. allowed for any test run. Since the gain in weight is to be used for At the end of the test run, turn off the moisture calculations, do not use any water coarse adjust valve, remove the probe and or other liquids to transfer the silica gel. If nozzle from the stack, turn off the pump, a balance is available, follow the procedure record the final dry gas meter reading, and for Container No. 5 in Section 5.4. conduct a post-test leak-check, as outlined 5.4 Analysis. Record the data required on in Section 5.1.4.3. a sheet such as the one shown in Figure 5H– 5.3 Sample Recovery. Begin recovery of 4. Handle each sample container as follows: the probe and filter sample as described in Containers No. 1 and 2. Leave the contents Method 5, Section 4.2, except that an acetone in the shipping container or transfer both of blank volume of about 50 ml may be used. the filters and any loose particulate from the Treat the samples as follows: sample container to a tared glass weighing dish. Desiccate for no more than 36 hours. Container No. 1. Carefully remove the filter Weigh to a constant weight and report the from the front filter holder and place it in its results to the nearest 0.1 mg. For purposes of identified petri dish container. Use a pair of this Section, 5.6, the term ‘‘constant weight’’ tweezers and/or clean disposable surgical means a difference of no more than 0.5 mg or gloves to handle the filter. If it is necessary 1 percent of total weight less tare weight, to fold the filter, do so such that the particu- whichever is greater, between two consecu- late cake is inside the fold. Carefully trans- tive weighings, with no less than 2 hours be- fer to the petri dish any particulate matter tween weighings. and/or filter fibers which adhere to the filter Container No. 3. Note the level of liquid in holder gasket, by using a dry Nylon bristle the container and confirm on the analysis brush and/or a sharp-edged blade. Seal and sheet whether leakage occurred during label the container. transport. If a noticeable amount of leakage Container No. 2. Remove the filter from the has occurred, either void the sample or use back filter holder using the same procedures methods, subject to the approval of the Ad- as described above. ministrator, to correct the final results. De- Container No. 3. Same as Method 5, Section termination of sample leakage is not appli- 4.2 for Container No. 2. except that descrip- cable if sample recovery and analysis occur tions of capping and sample transport are in the same room. Measure the liquid in this not applicable if sample recovery and analy- container either volumetrically to within 1 sis occur in the same room. ml or gravimetrically to within 0.5 g. Trans- Container No. 4. Treat the impingers as fol- fer the contents to a tared 250-ml or smaller lows: Measure the liquid which is in the first beaker, and evaporate to dryness at ambient three impingers to within 1 ml by using a temperature and pressure. Desiccate and graduated cylinder or by weighing it to with- weigh to a constant weight. Report the re- in 0.5 g by using a balance (if one is avail- sults to the nearest 0.1 mg. able). Record the volume or weight of liquid Container No. 4. Note the level of liquid in present. This information is required to cal- the container and confirm on the analysis culate the moisture content of the effluent sheet whether leakage occurred during gas. transport. If a noticeable amount of leakage Transfer the water from the first, second has occurred, either void the sample or use and third impingers to a glass container. methods, subject to the approval of the Ad- Tighten the lid on the sample container so ministrator, to correct the final results. De- that water will not leak out. Rinse termination of sample leakage is not appli- impingers and graduated cylinder, if used, cable if sample recovery and analysis occur with acetone three times or more. Avoid di- in the same room. Measure the liquid in this rect contact between the acetone and any container either volumetrically to within 1 stopcock grease or collection of any stop- ml or gravimetrically to within 0.5 g. Trans- cock grease in the rinse solutions. Add these fer the contents to a 500 ml or larger sepa- rinse solutions to sample Container No. 3. ratory funnel. Rinse the container with Whenever possible, containers should be water, and add to the separatory funnel. Add transferred in such a way that they remain 25 ml of dichloromethane to the separatory upright at all times. Descriptions of capping funnel, stopper and vigorously shake 1 and transport of samples are not applicable minute, let separate and transfer the
692
VerDate 11
dichloromethane (lower layer) into a tared cury barometer is used, no calibration is nec- beaker or evaporating dish. Repeat twice essary. Follow the manufacturer’s instruc- more. It is necessary to rinse the Container tions for operation. No. 4 with dichloromethane. This rinse is 6.6 SO2 Injection Rotameter. Calibrate added to the impinger extract container. the SO2 injection rotameter system with a Transfer the remaining water from the sepa- soap film flowmeter or similar direct volume ratory funnel to a tared beaker or measuring device with an accuracy of ± 2 evaporating dish and evaporate to dryness at percent. Operate the rotameter at a single 220 °F (105 °C). Desiccate and weigh to a con- reading for at least three calibration runs for stant weight. Evaporate the combined im- 10 minutes each. When three consecutive pinger water extracts at ambient tempera- calibration flow rates agree within 5 percent, ture and pressure. Desiccate and weigh to a average the three flow rates, mark the ro- constant weight. Report both results to the tameter at the calibrated setting, and use nearest 0.1 mg. the calibration flow rate as the SO2 injection Container No. 5. Weigh the spent silica gel flow rate during the test run. Repeat the ro- (or silica gel plus impinger) to the nearest 0.5 tameter calibration before the first certifi- g using a balance. cation test and semiannually, thereafter. ‘‘Acetone Blank’’ Container. Measure ace- 6.7 Analyzer Calibration Error Check. tone in this container either volumetrically Conduct the analyzer calibration error check or gravimetrically. Transfer the acetone to a prior to each certification test. tared 250-ml or smaller beaker, and evapo- 6.7.1 Calibration Gas Injection. After the rate to dryness at ambient temperature and flow rate measurement system and the trac- pressure. Desiccate and weigh to a constant er gas measurement system have been pre- weight. Report the results to the nearest 0.1 pared for use (Sections 5.1.5.2 and 5.1.6), in- mg. troduce zero gases and then the mid-level ‘‘Dichloromethane’’ Container. Measure 75 calibration gases for each analyzer. Set the ml of dichloromethane in this container and analyzers’ output responses to the appro- treat it the same as the ‘‘acetone blank.’’ priate levels. Then introduce the low-level ‘‘Water Blank’’ Container. Measure 200 ml and high-level calibration gases, one at a water into this container either time, for each analyzer. Record the analyzer volumetrically or gravemetrically. Transfer responses. the water to a tared 250-ml beaker and evap- 6.7.2 Acceptability Values. If the linear orate to dryness at 105 °C (221 °F). Desiccate curve for any analyzer determined from the and weigh to a constant weight. zero and mid-level calibration gases’ re- sponses does not predict the actual responses 6. Calibration of the low-level and high-level gases within 2 Maintain a laboratory record of all calibra- percent of the span value, the calibration of tions. that analyzer shall be considered invalid. 6.1 Volume Metering System. Take corrective measures on the measure- 6.1.1 Initial and Periodic Calibration. Be- ment system before repeating the calibra- fore the first certification or audit test and tion error check and proceeding with the at least semiannually, thereafter, calibrate test runs. the volume metering system as described in 6.8 Measurement System Response Time. Method 5G, Section 5.2.1. Introduce zero gas at the calibration gas 6.1.2 Calibration After Use. Same as valve into the flow rate measurement system Method 5G, Section 5.2.2. and the tracer gas measurement system 6.1.3 Acceptable Variation in Calibration. until all readings are stable. Then, quickly Same as Method 5G, Section 5.2.3. switch to introduce the mid-level calibration 6.2 Probe Heater Calibration. (Optional) gas at the calibration value until a stable The probe heating system shall be calibrated value is obtained. A stable value is equiva- before the first certification or audit test. lent to a change of less than 1 percent of Use the procedure described in Method 5, span value for 30 seconds. Record the re- Section 5.4. sponse time. Repeat the procedure three 6.3 Temperature Gauges. Use the proce- times. Conduct the response time check for dure in Method 2, Section 4.3, to calibrate in- each analyzer separately before its initial stack temperature gauges before the first use and at least semiannually thereafter. certification or audit test and semiannually, 6.9 Measurement System Drift Checks. thereafter. Immediately prior to the start of each test 6.4 Leak-Check of Metering System run (within 1 hour of the test run start), in- Shown in Figure 5H–1. That portion of the troduce zero and mid-level calibration gases, sampling train from the pump to the orifice one at a time, to each analyzer through the meter shall be leak-checked after each cer- calibration valve. Adjust the analyzers to re- tification or audit test. Use the procedure spond appropriately. Immediately following described in Method 5, Section 5.6. each test run (within 1 hour of the end of the 6.5 Barometer. Calibrate against a mer- test run), or if adjustments to the analyzers cury barometer before the first certification or measurement systems are required during test and semiannually, thereafter. If a mer- the test run, reintroduce the zero- and mid-
693
VerDate 11
level calibration gases and record the re- PR=Percent of proportional sampling rate. sponses, as described above. Make no adjust- Pbar=Barometric pressure at the sampling ments to the analyzers or the measurement site, mm Hg (in. Hg). system until after the drift checks are made. Pstd=Standard absolute pressure, 760 mm Hg If the difference between the analyzer re- (29.92 in. Hg). 3 sponses and the known calibration gas val- Qsd=Total gas flow rate, dsm /hr (dscf/hr). ues exceed the specified limits (Sections 4.2 QT=Flow of tracer gas, liters/min. and 4.3), the test run will be considered in- Si=Concentration measured at the SO2 ana- valid and shall be repeated following correc- lyzer for the ‘‘ith’’ 10 minute interval, tions to the measurement system. Alter- ppm. natively, recalibrate the measurement sys- S1=Concentration measured at the SO2 ana- tem and recalculate the measurement data. lyzer for the first 10-minute interval, Report the test run results using both the ppm. initial and final calibration data. T1=Absolute average stack gas temperature 6.10 Analytical Balance. Perform a for the first 10-minute interval, ° K (° R). multipoint calibration (at least five points Ti=Absolute average stack gas temperature spanning the operational range) of the ana- at the ‘‘ith’’ 10-minute interval, ° K (° R). lytical balance before the first certification Tm=Absolute average dry gas meter tempera- test and semiannually, thereafter. Before ture (see Figure 5H–3), ° K (° R). each certification test, audit the balance by Tstd=Standard absolute temperature, 293 ° K weighing at least one calibration weight (528 ° R). (class F) that corresponds to 50 to 150 percent Va=volume of solvent blank, ml. of the weight of one filter. If the scale can- Vaw=Volume of solvent used in wash, ml. not reproduce the value of the calibration Vlc=Total volume of liquid collected in weight to within 0.1 mg, conduct the impingers and silica gel (see Figure 5H– multipoint calibration before use. 4), ml. Vm=Volume of gas sample as measured by 7. Calculations dry gas meter, dm 3 (dcf). Carry out calculations, retaining at least Vm(std)=Volume of gas sample measured by one extra decimal figure beyond that of the the dry gas meter, corrected to standard acquired data. Round off figures after the conditions, dsm 3 (dscf). final calculation. Other forms of the equa- Vml(std)=Volume of gas sample measured by tions may be used as long as they give equiv- the dry gas meter during the first 10- alent results. minute interval, corrected to standard 7.1 Nomenclature. conditions, dsm 3 (dscf). a=Sample flow rate adjustment factor. Vmi(std)=Volume of gas sample measured by BR=dry wood burn rate, kg/hr (lb/hr), from the dry gas meter during the ‘‘ith’’ 10- Method 28, Section 8.3. minute interval, dsm 3 (dscf). V ( )=Volume of water vapor in the gas sam- Bws=Water vapor in the gas stream, propor- w std tion by volume. ple, corrected to standard conditions, sm 3 (scf). cs=Concentration of particulate matter in stack gas, dry basis, corrected to stand- Wa=Weight of residue in solvent wash, mg. ard conditions, g/dsm 3 (g/dscf). Y=Dry gas meter calibration factor. E=Particulate emission rate, g/hr. YCO=Measured mole fraction of CO (dry), av- ∆ H=Average pressure differential across the erage from Section 5.2.2.2, g/g-mole (lb/lb- mole). orifice meter (see Figure 5H–1), mm H2O Y =Measured mole fraction of CO (dry), (in. H2O). CO2 2 average from Section 5.2.2.2, g/g-mole (lb/ La=Maximum acceptable leakage rate for ei- ther a post-test leak check or for a leak- lb-mole). check following a component change; YHC=Assumed mole fraction of HC (dry), g/g- equal to 0.00057 m 3/min (0.02 cfm) or 4 mole (lb/lb-mole); percent of the average sampling rate, =0.0088 for catalytic wood heaters; whichever is less. =0.0132 for non-catalytic wood heaters; =0.0080 for pellet-fired wood heaters. L1=Individual leakage rate observed during the leak-check conducted before a com- 10=Length of first sampling period, minutes. ponent change, m 3/min (cfm). 13.6=Specific gravity of mercury. 100=Conversion to percent. Lp=Leakage rate observed during the post- test leak-check, m 3/min (cfm). θ=Total sampling time, min. θ =Sampling time interval, from the begin- mn=Total amount of particulate matter col- 1 lected, mg. ning of a run until the first component change, min. ma=Mass of residue of solvent after evapo- ration, mg. 7.2 Average dry gas meter temperature NC=Gram atoms of carbon/gram of dry fuel and average orifice pressure drop. See data (lb/lb), equal to 0.0425. sheet (Figure 5H–3). NT=Total dry moles of exhaust gas/Kg of dry 7.3 Dry Gas Volume. Correct the sample wood burned, g-moles/kg (lb-moles/lb). volume measured by the dry gas meter to
694
VerDate 11
standard conditions (20 °C, 760 mm Hg or 68 °F, 29.92 in. Hg) by using Equation 5H–1.
where; the weights obtained from containers 1, 2, 3, 0 K1=0.3858 K/m. Hg for metric units. and 4 less the appropriate solvent blanks (see =17.64 0 R/in. Hg for English units. Figure 5H–4). NOTE: Equation 5H–1 can be used as written NOTE: Refer to Method 5, Section 4.1.5 to unless the leakage rate observed during any assist in calculation of results involving two of the mandatory leak-checks (i.e., the post- filter assemblies. test leak-check or leak-check conducted be- 7.8 Particulate Concentration. fore a component change) exceeds La. c =(0.001 g/mg) (m /V ( )) If L exceeds L , Equation 5H–1 must be s n m std p a Eq. 5H–5 modified as follows: (a) Case I. No component changes made 7.9 Sample Flow Rate Adjustment. during sampling run. In this case, replace Vm in Equation 5H–1 with the expression: = S1 [Vm¥(Lp¥La)θ] a Eq. 5 H- 6 (b) Case II. One component change made Si during the sampling run. In this case, re- 7.10 Carbon Balance for Total Moles of place Vm in Equation 5H–1 by the expression: Exhaust Gas (dry)/Kg of Wood Burned in the
Vm¥(L1¥La)θ1 Exhaust Gas. and substitute only for those leakage rates (L1 or Lp) which exceed La. 7.4 Volume of Water Vapor.
Vw(std)=K2Vlc Eq. 5H–2 where: 3 K2=0.001333 m /ml for metric units =0.04707 ft3/ml for English units. 7.5 Moisture Content.
where:
K3=1000 g/kg for metric units. K3=1.0 lb/lb for English units.
NOTE: The NOx/SOx portion of the gas is as- sumed to be negligible. 7.11 Total Stack Gas Flow Rate.
Qsd=K4 NTBR 7.6 Solvent Wash Blank. Eq. 5H–8 where: 3 K4=0.02406 for metric units, dsm /g-mole. =384.8 for English units, dscf/lb-mole. 7.12 Particulate Emission Rate.
E=cs Qsd Eq. 5H–9 7.13 Proportional Rate Variation. Cal- 7.7 Total Particulate Weight. Determine culate PR for each 10-minute interval, i, of the total particulate catch from the sum of the test run.
695
VerDate 11
sion results, but do not include the test run results in calculating the weighted average emission rate, and repeat the test.
8. Bibliography 1. Same as for Method 5, citations 1 through 11, with the addition of the follow- ing: 2. Oregon Department of Environmental Quality Standard Method for Measuring the emissions and efficiencies of Woodstoves, 7.14 Acceptable Results. If no more than July 8, 1984. Pursuant to Oregon Administra- 15 percent of the PR values for all the inter- tive Rules Chapter 340, Division 21. ≤ ≤ vals exceed 90 percent PR 110 percent, and 3. American Society for Testing Materials. ≤ if no PR value for any interval exceeds 75 Proposed Test Methods for Heating Perform- ≤ PR 125 percent, the results are acceptable. ance and Emissions of Residential Wood- If the PR values for the test runs are judged fired Closed Combustion-Chamber Heating to be unacceptable, report the test run emis- Appliances. E–6 Proposal P 180. August 1986.
696
VerDate 11
697
VerDate 11
698
VerDate 11
699
VerDate 11
METHOD 6—DETERMINATION OF SULFUR DIOX- 1.1 Principle. A gas sample is extracted IDE EMISSIONS FROM STATIONARY SOURCES from the sampling point in the stack. The sulfuric acid mist (including sulfur trioxide) 1. Principle and Applicability
700
VerDate 11
and the sulfur dioxide are separated. The sul- cations and fluorides are removed by glass fur dioxide fraction is measured by the bar- wool filters and an isopropanol bubbler, and ium-thorin titration method. hence do not affect the SO2 analysis. When 1.2 Applicability. This method is applica- samples are being taken from a gas stream ble for the determination of sulfur dioxide with high concentrations of very fine metal- emissions from stationary sources. The mini- lic fumes (such as in inlets to control de- mum detectable limit of the method has vices), a high-efficiency glass fiber filter been determined to be 3.4 milligrams (mg) of must be used in place of the glass wool plug SO /m3 (2.12×10¥7 lb/ft3). Although no upper 2 (i.e., the one in the probe) to remove the limit has been established, tests have shown that concentrations as high as 80,000 mg/m3 cation interferents. Free ammonia interferes by reacting with of SO2 can be collected efficiently in two midget impingers, each containing 15 milli- SO2 to form particulate sulfite and by react- liters of 3 percent hydrogen peroxide, at a ing with the indicator. If free ammonia is rate of 1.0 lpm for 20 minutes. Based on theo- present (this can be determined by knowl- retical calculations, the upper concentration edge of the process and the presence of white limit in a 20-liter sample is about 93,300 mg/ particulate matter in the probe and m3. isopropanol bubbler), the alternative proce- Possible interferents are free ammonia, dures in Section 7.2 shall be used. water-soluble cations, and fluorides. The
701
VerDate 11
2. Apparatus tion of substituting sampling equipment de- 2.1 Sampling. The sampling train is scribed in Method 8 in place of the midget shown in Figure 6–1, and component parts impinger equipment of Method 6. However, are discussed below. The tester has the op-
702
VerDate 11
the Method 8 train must be modified to in- used, subject to approval of the Adminis- clude a heated filter between the probe and trator. isopropanol impinger, and the operation of 2.1.7 Valve. Needle valve, to regulate sam- the sampling train and sample analysis must ple gas flow rate. be at the flow rates and solution volumes de- 2.1.8 Pump. Leak-free diaphragm pump, or fined in Method 8. equivalent, to pull gas through the train. In- The tester also has the option of determin- stall a small surge tank between the pump ing SO2 simultaneously with particulate and rate meter to eliminate the pulsation ef- matter and moisture determinations by (1) fect of the diaphragm pump on the rotam- replacing the water in a Method 5 impinger eter. system with 3 percent peroxide solution, or 2.1.9 Rate Meter. Rotameter, or equiva- (2) by replacing the Method 5 water impinger lent, capable of measuring flow rate to with- system with a Method 8 isopropanol-filter- in 2 percent of the selected flow rate of about peroxide system. The analysis for SO2 must 1000 cc/min. be consistent with the procedure in Method 2.1.10 Volume Meter. Dry gas meter, suffi- 8. ciently accurate to measure the sample vol- 2.1.1 Probe. Borosilicate glass, or stain- ume within 2 percent, calibrated at the se- less steel (other materials of construction lected flow rate and conditions actually en- may be used, subject to the approval of the countered during sampling, and equipped Administrator), approximately 6-mm inside with a temperature gauge (dial thermom- diameter, with a heating system to prevent eter, or equivalent) capable of measuring water condensation and a filter (either in- temperature to within 3°C (5.4°F). stack or heated out-stack) to remove partic- 2.1.11 Barometer. Mercury, aneroid, or ulate matter, including sulfuric acid mist. A other barometer capable of measuring at- plug of glass wool is a satisfactory filter. mospheric pressure to within 2.5 mm Hg (0.1 2.1.2 Bubbler and Impingers. One midget in. Hg). In many cases, the barometric read- bubbler, with medium-coarse glass frit and ing may be obtained from a nearby National borosilicate or quartz glass wool packed in Weather Service station, in which case the top (see Figure 6–1) to prevent sulfuric acid station value (which is the absolute baro- mist carryover, and three 30-ml midget metric pressure) shall be requested and an impingers. The bubbler and midget adjustment for elevation differences between impingers must be connected in series with the weather station and sampling point shall leak-free glass connectors. silicone grease be applied at a rate of minus 2.5 mm Hg (0.1 may be used, if necessary, to prevent leak- in. Hg) per 30 m (100 ft) elevation increase or age. vice versa for elevation decrease. At the option of the tester, a midget im- 2.1.12 Vacuum Gauge and Rotameter. At pinger may be used in place of the midget least 760 mm Hg (30 in. Hg) gauge and 0–40 cc/ bubbler. min rotameter, to be used for leak check of Other collection absorbers and flow rates the sampling train. may be used, but are subject to the approval 2.2 Sample Recovery. of the Administrator. Also, collection effi- 2.2.1 Wash Bottles. Polyethylene or glass, ciency must be shown to be at least 99 per- 500 ml, two. cent for each test run and must be docu- 2.2.2 Storage Bottles. Polyethylene, 100 mented in the report. If the efficiency is ml, to store impinger samples (one per sam- found to be acceptable after a series of three ple). tests, further documentation is not required. 2.3 Analysis. To conduct the efficiency test, an extra ab- 2.3.1 Pipettes. Volumetric type, 5-ml, 20- sorber must be added and analyzed sepa- ml (one per sample), and 25-ml sizes. rately. This extra absorber must not contain 2.3.2 Volumetric Flasks. 100-ml size (one per sample) and 1000 ml size. more than 1 percent of the total SO2. 2.1.3 Glass Wool. Borosilicate or quartz. 2.3.3 Burettes. 5- and 50-ml sizes. 2.1.4 Stopcock Grease. Acetone-insoluble, 2.3.4 Erlenmeyer Flasks. 250 ml-size (one heatstable silicone grease may be used, if for each sample, blank, and standard). necessary. 2.3.5 Dropping Bottle. 125-ml size, to add indicator. 2.1.5 Temperature Gauge. Dial thermom- 2.3.6 Graduated Cylinder. 100-ml size. eter, or equivalent, to measure temperature 2.3.7 Spectrophotometer. To measure ab- of gas leaving impinger train to within 1°C sorbance at 352 nanometers. (2°F.) 2.1.6 Drying Tube. Tube packed with 6- to 3. Reagents 16-mesh indicating type silica gel, or equiva- Unless otherwise indicated, all reagents lent, to dry the gas sample and to protect must conform to the specifications estab- the meter and pump. If the silica gel has lished by the Committee on Analytical Re- been used previously, dry at 175°C (350°F) for agents of the American Chemical Society. 2 hours. New silica gel may be used as re- Where such specifications are not available, ceived. Alternatively, other types of use the best available grade. desiccants (equivalent or better) may be 3.1 Sampling.
703
VerDate 11
3.1.1 Water. Deionized distilled to conform agement office at each EPA regional Office to ASTM Specification D1193–77, Type 3 (in- or the responsible enforcement agency. corporated by reference—see § 60.17). At the (NOTE: The tester should notify the quality option of the analyst, the KMnO4 test for ox- assurance office or the responsible enforce- idizable organic matter may be omitted ment agency at least 30 days prior to the test when high concentrations of organic matter date to allow sufficient time for sample de- are not expected to be present. Unless other- livery.) wise specified, this water shall be used 3.3.7 Hydrochloric Acid (HCl) Solution, 0.1 throughout this method. N (for use in Section 7.2). Carefully pipette 3.1.2 Isopropanol, 80 percent. Mix 80 ml of 8.6 ml of concentrated HCl into a 1-liter volu- isopropanol with 20 ml of water. Check each metric flask containing water. Dilute to vol- lot of isopropanol for peroxide impurities as ume with mixing. follows: shake 10 ml of isopropanol with 10 4. Procedure ml of freshly prepared 10 percent potassium 4.1 Sampling. iodide solution. Prepare a blank by similarly 4.1.1 Preparation of Collection Train. treating 10 ml of water. After 1 minute, read Measure 15 ml of 80 percent isopropanol into the absorbance at 352 nanometers on a spec- the midget bubbler and 15 ml of 3 percent hy- trophotometer. If absorbance exceeds 0.1, re- drogen peroxide into each of the first two ject alcohol for use. midget impingers. Leave the final midget Peroxides may be removed from impinger dry. Assemble the train as shown isopropanol by redistilling or by passage through a column of activated alumina; how- in Figure 6–1. Adjust probe heater to a tem- ever, reagent grade isopropanol with suit- perature sufficient to prevent water con- ably low peroxide levels may be obtained densation. Place crushed ice and water from commercial sources. Rejection of con- around the impingers. taminated lots may, therefore, be a more ef- 4.1.2 Leak-Check Procedure. A leak check ficient procedure. prior to the sampling run is optional; how- 3.1.3 Hydrogen Peroxide, 3 Percent. Dilute ever, a leak check after the sampling run is 30 percent hydrogen peroxide 1:9 (v/v) with mandatory. The leak-check procedure is as water (30 ml is needed per sample). Prepare follows: fresh daily. Temporarily attach a suitable (e.g., 0–40 cc/ 3.1.4 Potassium Iodide Solution, 10 Per- min) rotameter to the outlet of the dry gas cent. Dissolve 10.0 grams KI in water and di- meter and place a vacuum gauge at or near lute to 100 ml. Prepare when needed. the probe inlet. Plug the probe inlet, pull a 3.2 Sample Recovery. vaccum of at least 250 mm Hg (10 in. Hg), and 3.2.1 Water. Same as in Section 3.1.1. note the flow rate as indicated by the rotam- 3.2.2 Isopropanol, 80 Percent. Mix 80 ml of eter. A leakage rate not in excess of 2 per- isopropanol with 20 ml of water. cent of the average sampling rate is accept- 3.3 Analysis. able. 3.3.1 Water. Same as in Section 3.1.1. NOTE: Carefully release the probe inlet 3.3.2 Isopropanol, 100 Percent. plug before turning off the pump. 3.3.3 Thorin Indicator. 1-(o- It is suggested (not mandatory) that the arsonophenylazo)-2-naphthol-3,6-disulfonic pump be leak-checked separately, either acid, disodium salt, or equivalent. Dissolve prior to or after the sampling run. If done 0.20 g in 100 ml of water. prior to the sampling run, the pump leak- 3.3.4 Barium Perchlorate Solution, 0.0100 check shall precede the leak check of the N. Dissolve 1.95 g of barium perchlorate tri- sampling train described immediately above; if done after the sampling run, the pump hydrate [Ba(ClO4)2•3H2O] in 200 ml water and dilute to 1 liter with isopropanol. Alter- leak-check shall follow the train leak-check. To leak check the pump, proceed as follows: natively, 1.22 g of [BaCl2•2H2O] may be used instead of the perchlorate. Standardize as in Disconnect the drying tube from the probe- Section 5.5. impinger assembly. Place a vacuum gauge at 3.3.5 Sulfuric Acid Standard, 0.0100 N. the inlet to either the drying tube or the Purchase or standardize to ±0.0002 N against pump, pull a vacuum of 250 mm (10 in.) Hg, 0.0100 N NaOH which has previously been plug or pinch off the outlet of the flow meter standardized against potassium acid phthal- and then turn off the pump. The vacuum ate (primary standard grade). should remain stable for at least 30 seconds. 3.3.6 Quality Assurance Audit Samples. Other leak-check procedures may be used, Sulfate samples in glass vials prepared by subject to the approval of the Adminstrator, EPA’s Environmental Monitoring Systems U.S. Environmental Protection Agency. Laboratory, Quality Assurance Division, 4.1.3 Sample Collection. Record the initial Source Branch, Mail Drop 77A, Research Tri- dry gas meter reading and barometric pres- angle Park, North Carolina 27711. Each set sure. To begin sampling, position the tip of will consist of two vials having solutions of the probe at the sampling point, connect the unknown concentrations. Only when making probe to the bubbler, and start the pump. Ad- compliance determinations, obtain an audit just the sample flow to a constant rate of ap- sample set from the Quality Assurance Man- proximately 1.0 liter/min as indicated by the
704
VerDate 11
rotameter. Maintain this constant rate (±10 Section 3.3.6.) The same analysts, analytical percent) during the entire sampling run. reagents, and analytical system shall be used Take readings (dry gas meter, temperatures both for compliance samples and the EPA at dry gas meter and at impinger outlet, and audit samples; if this condition is met, audit- rate meter) at least every 5 minutes. Add ing of subsequent compliance analyses for more ice during the run to keep the tempera- the same enforcement agency within 30 days ture of the gases leaving the last impinger at is not required. An audit sample set may not 20°C (68°F) or less. At the conclusion of each be used to validate different sets of compli- run, turn off the pump, remove probe from ance samples under the jurisdiction of dif- the stack, and record the final readings. Con- ferent enforcement agencies, unless prior ar- duct a leak check as in Section 4.1.2 (This rangements are made with both enforcement leak check is mandatory.) If a leak is found, agencies. void the test run, or use procedures accept- Calculate the concentrations in mg/dscm able to the Administrator to adjust the sam- using the specified sample volume in the ple volume for the leakage. Drain the ice audit instructions. (NOTE: Indication of ac- bath, and purge the remaining part of the ceptable results may be obtained imme- train by drawing clean ambient air through diately by reporting the audit results in mg/ the system for 15 minutes at the sampling rate. dscm and compliance results in total mg SO2/ Clean ambient air can be provided by pass- sample by telephone to the responsible en- ing air through a charcoal filter or through forcement agency.) Include the results of an extra midget impinger with 15 ml of 3 per- both audit samples, their identification numbers, and the analyst’s name with the cent H2O2. The tester may opt to simply use ambient air, without purification. results of the compliance determination 4.2 Sample Recovery. Disconnect the samples in appropriate reports to the EPA impingers after purging. Discard the con- regional office or the appropriate enforce- tents of the midget bubbler. Pour the con- ment agency. Include this information with tents of the midget impingers into a leak- subsequent compliance analyses for the same free polyethylene bottle for shipment. Rinse enforcement agency during the 30-day pe- the three midget impingers and the connect- riod. ing tubes with water, and add the washings The concentrations of the audit samples to the same storage container. Mark the obtained by the analyst shall agree within 5 fluid level. Seal and identify the sample con- percent of the actual concentrations. If the tainer. 5-percent specification is not met, reanalyze 4.3 Sample Analysis. Note level of liquid the compliance samples and audit samples, in container, and confirm whether any sam- and include initial and reanalysis values in ple was lost during shipment; note this on the test report (see NOTE in first paragraph analytical data sheet. If a noticeable amount of this section). of leakage has occurred, either void the sam- Failure to meet the 5-percent specification ple or use methods, subject to the approval may require retests until the audit problems of the Administrator, to correct the final re- are resolved. However, if the audit results do sults. not affect the compliance or noncompliance Transfer the contents of the storage con- status of the affected facility, the Adminis- tainer to a 100-ml volumetric flask and di- trator may waive the reanalysis require- lute to exactly 100 ml with water. Pipette a ment, further audits, or retests and accept 20-ml aliquot of this solution into a 250-ml the results of the compliance test. While Erlenmeyer flask, add 80 ml of 100 percent steps are being taken to resolve audit analy- isopropanol and two to four drops of thorin sis problems, the Administrator may also indicator, and titrate to a pink endpoint choose to use the data to determine the com- using 0.0100 N barium perchlorate. Repeat pliance or noncompliance status of the af- and average the titration volumes. Run a fected facility. blank with each series of samples. Replicate 5. titrations must agree within 1 percent or 0.2 Calibration ml, whichever is larger. 5.1 Metering System. NOTE: Protect the 0.0100 N barium per- 5.1.1 Initial Calibration. Before its initial chlorate solution from evaporation at all use in the field, first leak check the meter- times. ing system (drying tube, needle valve, pump, 4.4 Audit Sample Analysis. Concurrently rotameter, and dry gas meter) as follows: analyze the two audit samples and a set of place a vacuum gauge at the inlet to the dry- compliance samples (Section 4.3) in the same ing tube and pull a vaccum of 250 mm (10 in.) manner to evaluate the technique of the ana- Hg; plug or pinch off the outlet of the flow lyst and the standards preparation. (NOTE: It meter, and then turn off the pump. The is recommended that known quality control vaccum shall remain stable for at least 30 samples be analyzed prior to the compliance seconds. Carefully release the vaccum gauge and audit sample analysis to optimize the before releasing the flow meter end. system accuracy and precision. One source of Next, remove the drying tube and calibrate these samples is the Source Branch listed in the metering system (at the sampling flow
705
VerDate 11
rate specified by the method) as follows: con- Pstd=Standard absolute pressure, 760 mm Hg nect an appropriately sized wet test meter (29.92 in. Hg). (e.g., 1 liter per revolution) to the inlet of Tm=Average dry gas meter absolute tempera- the drying tube. Make three independent ture, °K (°R). calibration runs, using at least five revolu- Tstd=Standard absolute temperature, 293°K tions of the dry gas meter per run. Calculate (528°R). the calibration factor, Y (wet test meter Va=Volume of sample aliquot titrated, ml. calibration volume divided by the dry gas Vm=Dry gas volume as measured by the dry meter volume, both volumes adjusted to the gas meter, dcm (dcf). same reference temperature and pressure), Vm(std)=Dry gas volume measured by the dry for each run, and average the results. If any gas meter, corrected to standard condi- Y value deviates by more than 2 percent tions, dscm (dscf). from the average, the metering system is un- Vsoln=Total volume of solution in which the acceptable for use. Otherwise, use the aver- sulfur dioxide sample is contained, 100 age as the calibration factor for subsequent ml. test runs. Vt=Volume of barium perchlorate titrant 5.1.2 Post-Test Calibration Check. After used for the sample, ml (average or rep- each field test series, conduct a calibration licate titrations). check as in Section 5.1.1 above, except for V =Volume of barium perchlorate titrant the following variations: (a) the leak check tb used for the blank, ml. is not to be conducted, (b) three, or more Y=Dry gas meter calibration factor. revolutions of the dry gas meter may be used, and (c) only two independent runs need 32.03=Equivalent weight of sulfur dioxide. be made. If the calibration factor does not 6.2 Dry Sample Gas Volume, Corrected to deviate by more than 5 percent from the ini- Standard Conditions. tial calibration factor (determined in Sec- tion 5.1.1), then the dry gas meter volumes obtained during the test series are accept- able. If the calibration factor deviates by more than 5 percent, recalibrate the meter- Where: ing system as in Section 5.1.1, and for the K1=0.3858°K/mm Hg for metric units. calculations, use the calibration factor (ini- =17.64°R/in. Hg for English units. tial or recalibration) that yields the lower 6.3 Sulfur Dioxide Concentration. gas volume for each test run. 5.2 Thermometers. Calibrate against mer- cury-in-glass thermometers. 5.3 Rotameter. The rotameter need not be calibrated but should be cleaned and main- tained according to the manufactuturer’s in- struction. Where: 5.4 Barometer. Calibrate against a mer- cury barometer. K3=32.03 mg/meq. for metric units. × ¥5 5.5 Barium Perchlorate Solution. Stand- =7.061 10 lb/meq. for English units. ardize the barium perchlorate solution 6.4 Relative Error (RE) for QA Audit Sam- against 25 ml of standard sulfuric acid to ples, Percent. which 100 ml of 100 percent isopropanol has CC− been added. RE = d a ()100Eq. 6- 3 Run duplicate analyses. Calculate the nor- C mality using the average of a pair of dupli- a cate analyses where the titrations agree Where: within 1 percent or 0.2 ml, whichever is larg- Cd=Determined audit sample concentration, er. mg/dscm.
6. Calculations Ca=Actual audit sample concentration, mg/ Carry out calculations, retaining at least dscm. one extra decimal figure beyond that of the 7. Alternative Procedures acquired data. Round off figures after final 7.1 Dry Gas Meter as a Calibration Stand- calculation. ard. A dry gas meter may be used as a cali- 6.1 Nomenclature. bration standard for volume measurements Cso2=Concentration of sulfur dioxide, dry in place of the wet test meter specified in basis corrected to standard conditions, Section 5.1, provided that it is calibrated ini- mg/dscm (lb/dscf). tially and recalibrated periodically accord- N=Normality of barium perchlorate titrant, ing to the same procedures outlined in Meth- milliequivalents/ml. od 5, Section 7.1, with the following excep- Pbar=Barometric pressure at the exit orifice tion: (1) the dry gas meter is calibrated of the dry gas meter, mm Hg (in. Hg). against a wet test meter having a capacity of
706
VerDate 11
1 liter/rev or 3 liters/rev and having the capa- 7.2.1 Preparation of Collection Train. Pre- bility of measuring volume to within ±1 per- pare the sampling train as shown in Figure cent; (2) the dry gas meter is calibrated at 1 6–2. The rotameter and surge tank are op- liter/min (2 cfh); and (3) the meter box of the tional but are recommended in order to de- Method 6 sampling train is calibrated at the tect changes in the flow rate. same flow rate. 7.2 Critical Orifices for Volume and Rate NOTE: The critical orifices can be adapted Measurements. A critical orifice may be used to a Method 6 type sampling train as follows: in place of the dry gas meter specified in Insert sleeve type, serum bottle stoppers Section 2.1.10, provided that it is selected, into two reducing unions. Insert the needle calibrated, and used as follows: into the stoppers as shown in Figure 6–3.
707
VerDate 11
708
VerDate 11
7.2.2 Selection of Critical Orifices. The that do not reach a critical value shall not procedure that follows describes the use of be used. hypodermic needles and stainless steel nee- 7.2.3 Field Procedure. dle tubings, which have been found suitable 7.2.3.1 Leak-Check Procedure. A leak- for use as critical orifices. Other materials check before the sampling run is rec- and critical orifice designs may be used pro- ommended, but is optional. The leak-check vided the orifices act as true critical orifices, procedure is as follows: i.e., a critical vacuum can be obtained, as de- Temporarily attach a suitable (e.g., 0–40 cc/ scribed in this section. Select a critical ori- min) rotameter and surge tank, or a soap fice that is sized to operate at the desired bubble meter and surge tank to the outlet of flow rate. The needle sizes and tubing the pump. Plug the probe inlet, pull an out- lengths shown below give the following ap- let vacuum of at least 254 mm Hg (10 in. Hg), proximate flow rates. and note the flow rate as indicated by the ro- tameter or bubble meter. A leakage rate not Flow rate, Flow rate, in excess of 2 percent of the average sam- Gauge/cm Gauge/cm ¯ cc/min cc/min pling rate (Qstd) is acceptable. Carefully re- lease the probe inlet plug before turning off 21/7.6 1100 23/3.8 500 the pump. 22/2.9 1000 23/5.1 450 7.2.3.2 Moisture Determination. At the 22/3.8 900 24/3.2 400 sampling location, prior to testing, deter- mine the percent moisture of the ambient air Determine the suitability and the appro- using the wet and dry bulb temperatures or, priate operating vaccum of the critical ori- if appropriate, a relative-humidity meter. fice as follows: If applicable, temporarily at- 7.2.3.3 Critical Orifice Calibration. Prior tach a rotameter and surge tank to the out- to testing, at the sampling location, cali- let of the sampling train. Turn on the pump, brate the entire sampling train using a 500- and adjust the valve to give an outlet vacu- cc soap bubble meter which is attached to um reading corresponding to about half of the inlet of the probe and an outlet vacuum the atmospheric pressure. Observe the ro- of 25 to 50 mm Hg (1 to 2 in. Hg) above the tameter reading. Slowly increase the vacu- critical vacuum. Record the information um until a stable reading is obtained on the listed in Figure 6–4. rotameter. Record the critical vacuum, Calculate the standard volume of air meas- which is the outlet vacuum when the rotam- ured by the soap bubble meter and the volu- eter first reaches a stable value. Orifices metric flow rate, using the equations below:
709
VerDate 11
where: Tstd=Standard absolute temperature, 273°K Pbar=Barometric pressure, mm Hg (in. Hg). (528°R). Pstd=Standard absolute pressure, 760 mm Hg Vsb=Volume of gas as measured by the soap (29.92 in. Hg). bubble meter, m3 (ft3). Qstd=Volumetric flow rate through critical Vsb(std)=Volume of gas as measured by the orifice, scm/min (scf/min). soap bubble meter, corrected to standard Tamb=Ambient absolute temperature of air, conditions, scm (scf). °K (°R). Θ=Time, min.
710
VerDate 11
7.2.3.4 Sampling. Operate the sampling Conduct a post-test calibration run using train for sample collection at the same vacu- the calibration procedure outlined in Section um used during the calibration run. Start 7.2.3.3. If the Qstd obtained before and after the watch and pump simultaneously. Take the test differ by more than 5 pecent, void readings (temperature, rate meter, inlet vac- the test run; if not, calculate the volume of uum, and outlet vacuum) at least every 5 the gas measured with the critical orifice, minutes. At the end of the sampling run, Vm(std), using Equation 6–6 and the average of stop the watch and pump simultaneously. Qstd of both runs, as follows:
where: Pc=Inlet vacuum reading obtained during the Vm(std)=Dry gas volume measured with the calibration run, mm Hg (in. Hg). critical orifice, corrected to standard Psr=Inlet vacuum reading obtained during conditions, dscm (dscf). the sampling run, mm Hg (in. Hg). If the percent difference between the mo- Qstd=Average flow rate of pretest and post- test calibration runs, scm/min (scf/min). lecular weight of the ambient air at satu- rated conditions and the sample gas is more B =Water vapor in ambient air, proportion wa than ±3 percent, then the molecular weight by volume. of the gas sample must be considered in the θ s=Sampling time, min. calculations using the following equation:
where: has been shown to work. Where alkaline par- Ma=Molecular weight of the ambient air ticulate matter and condensed moisture are saturated at impinger temperature, g/g- present in the gas stream, the filter shall be mole (lb/lb-mole). heated above the moisture dew point but Ms=Molecular weight of the sample gas satu- below 225 °C. rated at impinger temperature, g/g-mole 7.3.2 Sample Recovery. Recover the sample (lb/lb-mole). according to Section 4.2 except for discard- NOTE: A post-test leak-check is not nec- ing the contents of the midget bubbler. Add essary because the post-test calibration run the bubbler contents, including the rinsings results will indicate whether there is any of the bubbler with water, to the poly- leakage. ethylene bottle containing the rest of the Drain the ice bath, and purge the sampling sample. Under normal testing conditions train using the procedure described in Sec- where sulfur trioxide will not be present sig- tion 4.1.3. nificantly, the tester may opt to delete the 7.3 Elimination of Ammonia Interference. midget bubbler from the sampling train. If The following alternative procedures shall be an approximation of the sulfur trioxide con- used in addition to those specified in the centration is desired, transfer the contents method when sampling at sources having of the midget bubbler to a separate poly- ammonia emissions. ethylene bottle. 7.3.1 Sampling. The probe shall be main- 7.3.3 Sample Analysis. Follow the proce- tained at 275 °C and equipped with a high-ef- dures in Section 4.3, except add 0.5 ml of 0.1 ficiency in-stack filter (glass fiber) to re- N HC1 to the Erlenmeyer flask and mix be- move particulate matter. The filter material fore adding the indicator. The following shall be unreactive to SO2. Whatman 934AH analysis procedure may be used for an ap- (formerly Reeve Angel 934AH) filters treated proximation of the sulfur trioxide concentra- as described in Citation 10 of the Method 5 tion. The accuracy of the calculated con- bibliography is an example of a filter that centration will depend upon the ammonia to
711
VerDate 11
SO2 ratio and the level of oxygen present in Agency, Research Triangle Park, NC. EPA– the gas stream. A fraction of the SO2 will be 650/4–74–024. December 1973. counted as sulfur trioxide as the ammonia to 7. Annual Book of ASTM Standards. Part SO2 ratio and the sample oxygen content in- 31; Water, Atmospheric Analysis. American creases. Generally, when this ratio is 1 or Society for Testing and Materials. Philadel- less and the oxygen content is in the range phia, PA. 1974. pp. 40–42. of 5 percent, less than 10 percent of the SO2 8. Knoll, J. E. and M. R. Midgett. The Ap- will be counted as sulfur trioxide. Analyze plication of EPA Method 6 to High Sulfur Di- the peroxide and isopropanol sample portions oxide Concentrations. Environmental Pro- separately. Analyze the peroxide portion as tection Agency. Research Triangle Park, NC. described above. Sulfur trioxide is deter- EPA–600/4–76–038. July 1976. mined by difference using sequential titra- 9. Westlin, P. R. and R. T. Shigehara. Pro- tion of the isopropanol portion of the sam- cedure for Calibrating and Using Dry Gas ple. Transfer the contents of the isopropanol Meter Volume Meters as Calibration Stand- storage container to a 100-ml volumetric ards. Source Evaluation Society Newsletter. flask, and dilute to exactly 100 ml with 3(1):17–30. February 1978. water. Pipette a 20-ml aliquot of this solu- 10. Yu, K. K. Evaluation of Moisture Effect tion into a 250-ml Erlenmeyer flask, add 0.5 on Dry Gas Meter Calibration. Source Eval- uation Society Newsletter. 5(1):24–28. Feb- ml of 0.1 N HC1, 80 ml of 100 percent ruary 1980. isopropanol, and two to four drops of thorin 11. Lodge, J.P., Jr., J.B. Pate, B.E. indicator. Titrate to a pink endpoint using Ammons, and G.A. Swanson. The Use of 0.0100 N barium perchlorate. Repeat and av- Hypodermic Needles as Critical Orifices in erage the titration volumes that agree with- Air Sampling. J. Air Pollution Control Asso- in 1 percent or 0.2 ml, whichever is larger. ciation. 16:197–200. 1966. Use this volume in Equation 6–2 to deter- 12. Shigehara, R.T., and Candace B. mine the sulfur trioxide concentration. From Sorrell. Using Critical Orifices as Method 5 the flask containing the remainder of the Calibration Standards. Source Evaluation isopropanol sample, determine the fraction Society Newsletter. 10(3):4–15. August 1985. of SO2 collected in the bubbler by pipetting 20-ml aliquots into 250-ml Erlenmeyer flasks. METHOD 6A—DETERMINATION OF SULFUR DI- Add 5 ml of 3 percent hydrogen peroxide, 100 OXIDE, MOISTURE, AND CARBON DIOXIDE ml of 100 percent isopropanol, and two to EMISSIONS FROM FOSSIL FUEL COMBUSTION four drops of thorin indicator, and titrate as SOURCES before. From this titration volume, subtract the titrant volume determined for sulfur tri- 1. Principle and Applicability oxide, and add the titrant volume deter- 1.1 Applicability. This method applies to mined for the peroxide portion. This final the determination of sulfur dioxide (SO2) volume constitutes Vt, the volume of barium emissions from fossil fuel combustion 3 perchlorate used for the SO2 sample. sources in terms of concentration (mg/m ) and in terms of emission rate (ng/J) and to 8. Bibliography the determination of carbon dioxide (CO2) 1. Atmospheric Emissions from Sulfuric concentration (percent). Moisture, if desired, Acid Manufacturing Processes. U.S. DHEW, may also be determined by this method. PHS, Division of Air Pollution. Public The minimum detectable limit, the upper Health Service Publication No. 999–AP–13. limit, and the interferences of the method Cincinnati, OH. 1965. for the measurement of SO2 are the same as 2. Corbett, P. F. The Determination of SO2 for Method 6. For a 20-liter sample, the and SO3 in Flue Gases. Journal of the Insti- method has a precision of 0.5 percent CO2 for tute of Fuel. 24: 237–243, 1961. concentrations between 2.5 and 25 percent 3. Matty, R. E. and E. K. Diehl. Measuring CO2 and 1.0 percent moisture for moisture Flue-Gas SO2 and SO3. Power. 101: 94–97. No- concentrations greater than 5 percent. vember 1957. 1.2 Principle. The principle of sample col- 4. Patton, W. F. and J. A. Brink, Jr. New lection is the same as for Method 6 except Equipment and Techniques for Sampling that moisture and CO2 are collected in addi- Chemical Process Gases. J. Air Pollution tion to SO2 in the same sampling train. Control Association. 13: 162. 1963. Moisture and CO2 fractions are determined 5. Rom, J. J. Maintenance, Calibration, gravimetrically. and Operation of Isokinetic Source-sampling 2. Apparatus Equipment. Office of Air Programs, Environ- 2.1 Sampling. The sampling train is mental Protection Agency. Research Tri- shown in Figure 6A–1; the equipment re- angle Park, NC. APTD–0576. March 1972. quired is the same as for Method 6, Section 6. Hamil, H. F. and D. E. Camann. Collabo- 2.1, except as specified below: rative Study of Method for the Determina- 2.1.1 SO2 Absorbers. Two 30-ml midget tion of Sulfur Dioxide Emissions from Sta- impingers with a 1-mm restricted tip and tionary Sources (Fossil-Fuel Fired Steam two 30-ml midget bubblers with an unre- Generators). Environmental Protection stricted tip. Other types of impingers and
712
VerDate 11
bubblers, such as Mae West for SO2 collec- assure the desired temperature is main- tion and rigid cylinders for moisture absorb- tained. A heated Teflon connector may be ers containing Drierite, may be used with used to connect the filter holder or probe to proper attention to reagent volumes and lev- the first impinger. els, subject to the Administrator’s approval. NOTE: Mention of a brand name does not 2.1.2 CO2 Absorber. A sealable rigid cyl- constitute endorsement by the Environ- inder or bottle with an inside diameter be- mental Protection Agency. tween 30 and 90 mm and a length between 125 and 250 mm and with appropriate connec- 2.2 Sample Recovery and Analysis. The tions at both ends. equipment needed for sample recovery and analysis is the same as required for Method NOTE: For applications downstream of wet 6. In addition, a balance to measure within scrubbers, a heated out-of-stack filter (either 0.05 g is needed for analysis. borosilicate glass wool or glass fiber mat) is necessary. The filter may be a separate heat- 3. Reagents ed unit or may be within the heated portion Unless otherwise indicated, all reagents of the probe. If the filter is within the sam- must conform to the specifications estab- pling probe, the filter should not be within 15 lished by the committee on analytical re- cm of the probe inlet or any unheated sec- agents of the American Chemical Society. tion of the probe, such as the connection to Where such specifications are not available, the first SO2 absorber. The probe and filter use the best available grade. should be heated to at least 20°C above the 3.1 Sampling. The reagents required for source temperature, but not greater than sampling are the same as specified in Method 120°C. The filter temperature (i.e., the sam- 6. In addition, the following reagents are re- ple gas temperature) should be monitored to quired:
3.1.1 Drierite. Anhydrous calcium sulfate analysis are the same as for Method 6, Sec- (CaSO4) desiccant, 8 mesh, indicating type is tions 3.2 and 3.3, respectively. recommended. (Do not use silica gel or simi- 4. Procedure lar desiccant in the application.) 4.1 Sampling. 3.1.2 CO2 Absorbing Material. Ascarite II. 4.1.1 Preparation of Collection Train. Sodium hydroxide coated silica, 8 to 20 mesh. Measure 15 ml of 80 percent isopropanol into 3.2 Sample Recovery and Analysis. The the first midget bubbler and 15 ml of 3 per- reagents needed for sample recovery and cent hydrogen peroxide into each of the first two midget impingers as described in Method
713
VerDate 11
6. Insert the glass wool into the top of the more glass wool into the cylinder to make isopropanol bubbler as shown in Figure 6A–1. the CO2 absorbing material stable. Clean the Into the fourth vessel in the train, the sec- outside of the cylinder of loose dirt and ond midget bubbler, place about 25 g of moisture and weigh at room temperature to Drierite. Clean the outsides of the bubblers the nearest 0.1 g. Record this initial mass. and impingers, and weigh at room tempera- Assemble the train as shown in Figure 6A– ture (ù20°C) to the nearest 0.1 g. Weigh the 1. Adjust the probe heater to a temperature four vessels simultaneously, and record this sufficient to prevent condensation (see Note initial mass. in section 2.1.1). Place crushed ice and water With one end of the CO2 absorber sealed, around the impingers and bubblers. Mount place glass wool in the cylinder to a depth of the CO2 absorber outside the water bath in a about 1 cm. Place about 150 g of CO2 absorb- vertical flow position with the sample gas ing material in the cylinder on top of the inlet at the bottom. Flexible tubing, e.g., glass wool, and fill the remaining space in Tygon, may be used to connect the last SO2 the cylinder with glass wool. Assemble the absorbing bubbler to the Drierite absorber cylinder as shown in Figure 6A–2. With the and to connect the Drierite absorber to the cylinder in a horizontal position, rotate it CO2 absorber. A second, smaller CO2 absorber around the horizontal axis. The CO2 absorb- containing Ascarite II may be added in line ing material should remain in position dur- downstream of the primary CO2 absorber as a ing the rotation, and no open spaces or chan- breakthrough indicator. Ascarite II turns nels should be formed. If necessary, pack white when CO2 is absorbed.
4.1.2 Leak-Check Procedure and Sample leak-free polyethylene bottle for shipping. Collection. The leak-check procedure and Rinse the two midget impingers and connect- sample collection procedure are the same as ing tubes with deionized distilled water, and specified in Method 6, Sections 4.1.2 and 4.1.3, add the washings to the same storage con- respectively. tainer.
4.2 Sample Recovery. 4.2.3 CO2 Absorber. Allow the CO2 absorber 4.2.1 Moisture Measurement. Disconnect to warm to room temperature (about 10 min- the isopropanol bubbler, the SO2 impingers, utes), clean the outside of loose dirt and and the moisture absorber from the sample moisture, and weigh to the nearest 0.1 g in train. Allow about 10 minutes for them to the same manner as in Section 4.1.1. Record reach room temperature, clean the outsides this final mass. Discard used Ascarite II ma- of loose dirt and moisture, and weigh them terial. simultaneously in the same manner as in 4.3 Sample Analysis. The sample analysis Section 4.1.1. Record this final mass. procedure for SO2 is the same as specified in 4.2.2 Peroxide Solution. Discard the con- Method 6, Section 4.3. tents of the isopropanol bubbler and pour the contents of the midget impingers into a
714
VerDate 11
4.4 Quality Assurance (QA) Audit Sam- abbreviated procedure may be used. The dif- ples. Only when this method is used for com- ferences between Method 6A and the abbre- pliance determinations, obtain an audit sam- viated procedure are described below. ple set as directed in Section 3.3.6 of Method 7.1 Sample Train. The sample train is the 6. Analyze the audit samples, and report the same as shown in Figure 6A–1 and as de- results as directed in Section 4.4 of Method 6. scribed in Section 4, except that the dry gas Acceptance criteria for the audit results are meter is not needed. the same as in Method 6. 7.2 Preparation of the Collection Train. 5. Calibration Follow the same procedure as in Section The calibrations and checks are the same 4.1.1, except do not weigh the isopropanol
as required in Method 6, Section 5. bubbler, the SO2 absorbing impingers or the 6. Calculations moisture absorber. Carry out calculations, retaining at least 7.3 Sampling. Operate the train as de- one extra decimal figure beyond that of the scribed in Section 4.1.3, except that dry gas acquired data. Round off figures after final meter readings, barometric pressure, and dry calculations. The calculations, nomen- gas meter temperatures need not be re- clature, and procedures are the same as spec- corded. ified in Method 6 with the addition of the fol- 7.4 Sample Recovery. Follow the proce- lowing: dure in Section 4.2, except do not weigh the
6.1 Nomenclature. isopropanol bubbler, the SO2 absorbing
Cw= Concentration of moisture, percent. impingers, or the moisture absorber. CCO2=Concentration of CO2, dry basis, per- 7.5 Sample Analysis. Analysis of the per- cent. oxide solution is the same as described in mwi=Initial mass of impingers, bubblers, and Section 4.3. Only when making compliance
moisture absorber, g. determinations, conduct an audit of the SO2 mwf=Final mass of impingers, bubblers, and analysis procedure as described in Section moisture absorber, g. 4.4. mai=Initial mass of CO2 absorber, g. 7.6 Calculations. maf=Final mass of CO2 absorber, g. 7.6.1 SO2 Mass Collected. VCO2(std)=Equivalent volume of CO2 collected at standard conditions, dsm3. Vw(std)=Equivalent volume of moisture col- lected at standard conditions, sm3. ¥4 5.467×10 =Equivalent volume of gaseous CO2 at standard conditions per gram, sm3/g. Where: 1.336×10¥3=Equivalent volume of water vapor at standard conditions per gram, sm3/g. mSO2=Mass of SO2 collected, mg. 6.2 CO2 Volume Collected, Corrected to 7.6.2 Sulfur Dioxide Emission Rate. Standard Conditions. ¥4 VCO2(std) =5.467 x 10 (maf¥ mai) Eq. 6A–1 6.3 Moisture Volume Collected, Corrected to Standard Conditions. ¥3 Vw(std) =1.336 x 10 (mwf¥ mwi) Eq. 6A–2 Where: 6.4 SO2 Concentration. ESO2=Emission rate of SO2 (ng/J). 3 Fc=Carbon F Factor for the fuel burned, m / J, from Method 19. 8. Bibliography 1. Same as for Method 6, Citations 1
6.5 CO2 Concentration. through 8, with the addition of the following: 2. Stanley, Jon and P.R. Westlin. An Al- ternate Method for Stack Gas Moisture De- termination. Source Evaluation Society Newsletter. Vol. 3, No. 4. November 1978. 6.6 Moisture Concentration. 3. Whittle, Richard N. and P.R. Westlin. V Air Pollution Test Report: Development and = w() std Evaluation of an Intermittent Integrated Cw Eq. 6 A- 5 VVV+ + SO2/CO2 Emission Sampling Procedure. Envi- m()()() std w std CO2 std ronmental Protection Agency, Emission 7. Emission Rate Procedure Standard and Engineering Division, Emis- If the only emission measurement desired sion Measurement Branch. Research Tri- is in terms of emission rate of SO¶ (ng/J), an angle Park, NC. December 1979. 14 pages.
715
VerDate 11
METHOD 6B—DETERMINATION OF SULFUR DI- because of potential corrosion and contami- OXIDE AND CARBON DIOXIDE DAILY AVERAGE nation of sample. Glass probes or other types EMISSIONS FROM FOSSIL FUEL COMBUSTION of stainless steel, e.g., Hasteloy or Carpenter SOURCES 20, are recommended for long-term use. 1. Principle and Applicability Other sampling equipment, such as Mae West bubblers and rigid cylinders for mois- 1.1 Applicability. This method applies to ture absorption, which requires sample or re- the determination of sulfur dioxide (SO2) agent volumes other than those specified in emissions from combustion sources in terms this procedure for full effectiveness may be 3 of concentration (ng/m ) and emission rate used, subject to the approval of the Adminis- (ng/J), and for the determination of carbon trator. dioxide (CO2) concentration (percent) on a daily (24 hours) basis. 3. Reagents The minimum detectable limits, upper All reagents for sampling and analysis are limit, and the interferences for SO2 measure- the same as described in Method 6A, Section ments are the same as for Method 6. EPA- 3, except isopropanol is not used for sam- sponsored collaborative studies were under- pling. The hydrogen peroxide absorbing solu- taken to determine the magnitude of repeat- tion shall be diluted to no less than 6 percent ability and reproducibility achievable by by volume, instead of 3 percent as specified qualified testers following the procedures in in Method 6. If Method 6B is to be operated this method. The results of the studies in a low sample flow condition (less than 100 evolve from 145 field tests including compari- ml/min), molecular sieve material may be sons with Methods 3 and 6. For measure- substituted for Ascarite II as the CO2 absorb- ments of emission rates from wet, flue gas ing material. The recommended molecular desulfurization units in (ng/J), the repeat- sieve material is Union Carbide 1⁄16 inch pel- ability (within laboratory precision) is 8.0 lets, 5A°, or equivalent. Molecular sieve ma- percent and the reproducibility (between lab- terial need not be discarded following the oratory precision) is 11.1 percent. sampling run provided it is regenerated as 1.2 Principle. A gas sample is extracted per the manufacturer’s instruction. Use of from the sampling point in the stack inter- molecular sieve material at flow rates higher mittently over a 24-hour or other specified than 100 ml/min may cause erroneous CO2 re- time period. Sampling may also be con- sults. ducted continuously if the apparatus and 4. Procedure procedures are appropriately modified (see 4.1 Sampling. Note in Section 4.1.1). The SO2 and CO2 are separated and collected in the sampling 4.1.1 Preparation of Collection Train. train. The SO2 fraction is measured by the Preparation of the sample train is the same barium-thorin titration method, and CO2 is as described in Method 6A, Section 4.1.4, with determined gravimetrically. the addition of the following: 2. Apparatus The sampling train is assembled as shown in Figure 6A–1, except the isopropanol bub- The equipment required for this method is bler is not included. The probe must be heat- the same as specified for Method 6A, Section ed to a temperature sufficient to prevent 2, except the isopropanol bubbler is not used. water condensation and must include a filter An empty bubbler for the collection of liquid (either in-stack, out-of-stack, or both) to droplets and does not allow direct contact prevent particulate entrainment in the per- between the collected liquid and the gas oxide impingers. The electric supply for the sample may be included in the train. For probe heat should be continuous and sepa- intermittent operation, include an industrial rate from the timed operation of the sample timer-switch designed to operate in the ‘‘on’’ pump. position at least 2 minutes continuously and Adjust the timer-switch to operate in the ‘‘off’’ the remaining period over a repeating ‘‘on’’ position from 2 to 4 minutes on a 2- cycle. The cycle of operation in designated in the applicable regulation. At a minimum, hour repeating cycle or other cycle specified the sampling operation should include at in the applicable regulation. Other timer se- least 12, equal, evenly-spaced periods per 24 quences may be used with the restriction hours. that the total sample volume collected is be- For applications downstream of wet scrub- tween 25 and 60 liters for the amounts of bers, a heated out-of-stack filter (either sampling reagents prescribed in this method. borosilicate glass wool or glass fiber mat) is Add cold water to the tank until the necessary. The probe and filter should be impingers and bubblers are covered at least heated continuously to at least 20°C above two-thirds of their length. The impingers the source temperature, but not greater than and bubbler tank must be covered and pro- 120°C. The filter (i.e., sample gas) tempera- tected from intense heat and direct sunlight. ture should be monitored to assure the de- If freezing conditions exist, the impinger so- sired temperature is maintained. lution and the water bath must be protected. Stainless steel sampling probes, type 316, NOTE: Sampling may be conducted con- are not recommended for use with Method 6B tinuously if a low flow-rate sample pump (20
716
VerDate 11
to 40 ml/min for the reagent volumes de- except for the following variations: (1) The scribed in this method) is used. Then the leak check is not to be conducted, (2) three timer-switch is not necessary. In addition, if or more revolutions of the dry gas meter the sample pump is designed for constant must be used, and (3) only two independent rate sampling, the rate meter may be de- runs need be made. If the calibration factor leted. The total gas volume collected should does not deviate by more than 5 percent from be between 25 and 60 liters for the amounts of the initial calibration factor determined in sampling reagents prescribed in this method. Section 5.1.1, then the dry gas meter volumes 4.1.2 Leak-Check Procedure. The leak- obtained during the test series are accept- check procedure is the same as described in able and use of the train can continue. If the Method 6, Section 4.1.2. calibration factor deviates by more than 5 4.1.3 Sample Collection. Record the initial percent, recalibrate the metering system as dry gas meter reading. To begin sampling, in Section 5.1.1; and for the calculations for position the tip of the probe at the sampling the preceding 30 days of data, use the cali- point, connect the probe to the first im- bration factor (initial or recalibration) that pinger (or filter), and start the timer and the yields the lower gas volume for each test sample pump. Adjust the sample flow to a run. Use the latest calibration factor for suc- constant rate of approximately 1.0 liter/min ceeding tests. as indicated by the rotameter. Assure that 5.2 Thermometers. Calibrate against mer- the timer is operating as intended, i.e., in cury-in-glass thermometers initially and at the ‘‘on’’ position for the desired period and 30-day intervals. the cycle repeats as required. 5.3 Rotameter. The rotameter need not be During the 24-hour sampling period, record calibrated, but should be cleaned and main- the dry gas meter temperature one time be- tained according to the manufacturer’s in- tween 9:00 a.m. and 11:00 a.m., and the baro- structions. metric pressure. 5.4 Barometer. Calibrate against a mer- At the conclusion of the run, turn off the cury barometer initially and at 30-day inter- timer and the sample pump, remove the vals. 5.5 Barium Perchlorate Solution. Stand- probe from the stack, and record the final ardize the barium perchlorate solution gas meter volume reading. Conduct a leak against 25 ml of standard sulfuric acid to check as described in Section 4.1.2. If a leak which 100 ml of 100 percent isopropanol has is found, void the test run or use procedures been added. acceptable to the Administrator to adjust the sample volume for leakage. Repeat the 6. Calculations steps in this section (4.1.3) for successive The nomenclature and calculation proce- runs. dures are the same as in Method 6A with the 4.2 Sample Recovery. The procedures for following exceptions:
sample recovery (moisture measurement, Pbar= Initial barometric pressure for the test peroxide solution, and CO2 absorber) are the period, mm Hg. same as in Method 6A, Section 4.2. Tm= Absolute meter temperature for the test 4.3 Sample Analysis. Analysis of the per- period, °K. oxide impinger solutions is the same as in 7. Emission Rate Procedure Method 6, Section 4.3. 4.4 Quality Assurance (QA) Audit Sam- The emission rate procedure is the same as ples. Only when this method is used for com- described in Method 6A, Section 7, except pliance determinations, obtain an audit sam- that the timer is needed and is operated as ple set as directed in Section 3.3.6 of Method described in this method. Only when this 6. Analyze the audit samples at least once method is used for compliance determina- tions, perform the QA audit analyses as de- for every 30 days of sample collection, and scribed in Section 4.4. report the results as directed in Section 4.4 of Method 6. The analyst performing the 8. Bibliography sample analyses shall perform the audit The bibliography is the same as described analyses. If more than one analyst performed in Method 6A, with the addition of the fol- the sample analyses during the 30-day sam- lowing: pling period, each analyst shall perform the 1. Butler, Frank E; J.E. Knoll, J.C. Suggs, audit analyses and all audit results shall be M.R. Midgett, and W. Mason. The Collabo- reported. Acceptance criteria for the audit rative Test of Method 6B: Twenty-Four-Hour results are the same as in Method 6. Analysis of SO2 and CO2. JAPCA. Vol. 33, No. 5. Calibration 10. October 1983.
5.1 Metering System. METHOD 6C—DETERMINATION OF SULFUR DIOX- 5.1.1 Initial Calibration. The initial cali- IDE EMISSIONS FROM STATIONARY SOURCES bration for the volume metering system is (INSTRUMENTAL ANALYZER PROCEDURE) the same as for Method 6, Section 5.1.1. 5.1.2 Periodic Calibration Check. After 30 1. Applicability and Principle days of operation of the test train, conduct a 1.1 Applicability. This method is applicable calibration check as in Section 5.1.1 above, to the determination of sulfur dioxide (SO2)
717
VerDate 11
concentrations in controlled and uncon- 3.6 Zero Drift. The difference in the meas- trolled emissions from stationary sources urement system output reading from the ini- only when specified within the regulations. tial calibration response at the zero con- 1.2 Principle. A gas sample is continu- centration level after a stated period of oper- ously extracted from a stack, and a portion ation during which no unscheduled mainte- of the sample is conveyed to an instrumental nance, repair, or adjustment took place. analyzer for determination of SO2 gas con- 3.7 Calibration Drift. The difference in the centration using an ultraviolet (UV), non- measurement system output reading from dispersive infrared (NDIR), or fluorescence the initial calibration response at a mid- analyzer. Performance specifications and range calibration value after a stated period test procedures are provided to ensure reli- of operation during which no unscheduled able data. maintenance, repair, or adjustment took 2. Range and Sensitivity place. 2.1 Analytical Range. The analytical 3.8 Response Time. The amount of time range is determined by the instrumental de- required for the measurement system to dis- sign. For this method, a portion of the ana- play 95 percent of a step change in gas con- lytical range is selected by choosing the span centration on the data recorder. of the monitoring system. The span of the 3.9 Interference Check. A method for de- monitoring system shall be selected such tecting analytical interferences and exces- that the pollutant gas concentration equiva- sive biases through direct comparison of gas lent to the emission standard is not less than concentrations provided by the measurement 30 percent of the span. If at any time during system and by a modified Method 6 proce- a run the measured gas concentration ex- dure. For this check, the modified Method 6 ceeds the span, the run shall be considered samples are acquired at the sample by-pass invalid. discharge vent. 2.2 Sensitivity. The minimum detectable 3.10 Calibration Curve. A graph or other limit depends on the analytical range, span, systematic method of establishing the rela- and signal-to-noise ratio of the measurement tionship between the analyzer response and system. For a well designed system, the min- the actual gas concentration introduced to imum detectable limit should be less than 2 the analyzer. percent of the span. 4. Measurement System Performance Specifica- 3. Definitions tions 3.1 Measurement System. The total equip- 4.1 Analyzer Calibration Error. Less than ment required for the determination of gas ±2 percent of the span for the zero, mid- concentration. The measurement system range, and high-range calibration gases. consists of the following major subsystems: 4.2 Sampling System Bias. Less than ±5 3.1.1 Sample Interface. That portion of a percent of the span for the zero, and mid- or system used for one or more of the following: high-range calibration gases. sample acquisition, sample transport, sam- 4.3 Zero Drift. Less than ±3 percent of the ple conditioning, or protection of the analyz- span over the period of each run. ers from the effects of the stack effluent. 4.4 Calibration Drift. Less than ±3 percent 3.1.2 Gas Analyzer. That portion of the of the span over the period of each run. system that senses the gas to be measured 4.5 Interference Check. Less than ±7 per- and generates an output proportional to its cent of the modified Method 6 result for each concentration. run. 3.1.3 Data Recorder. A strip chart re- 5. Apparatus and Reagents corder, analog computer, or digital recorder for recording measurement data from the an- 5.1 Measurement System. Any measure- alyzer output. ment system for SO2 that meets the speci- 3.2 Span. The upper limit of the gas con- fications of this method. A schematic of an centration measurement range displayed on acceptable measurement system is shown in the data recorder. Figure 6C–1. The essential components of the 3.3 Calibration Gas. A known concentra- measurement system are described below: tion of a gas in an appropriate diluent gas. 5.1.1 Sample Probe. Glass, stainless steel, 3.4 Analyzer Calibration Error. The dif- or equivalent, of sufficient length to traverse ference between the gas concentration exhib- the sample points. The sampling probe shall ited by the gas analyzer and the known con- be heated to prevent condensation. centration of the calibration gas when the 5.1.2 Sample Line. Heated (sufficient to calibration gas is introduced directly to the prevent condensation) stainless steel or Tef- analyzer. lon tubing, to transport the sample gas to 3.5 Sampling System Bias. The difference the moisture removal system. between the gas concentrations exhibited by 5.1.3 Sample Transport Lines. Stainless the measurement system when a known con- steel or Teflon tubing, to transport the sam- centration gas is introduced at the outlet of ple from the moisture removal system to the the sampling probe and when the same gas is sample pump, sample flow rate control, and introduced directly to the analyzer. sample gas manifold.
718
VerDate 11
5.1.4 Calibration Valve Assembly. A three- constructed of any material that is nonreac- way valve assembly, or equivalent, for block- tive to the gas being sampled. ing the sample gas flow and introducing cali- 5.1.10 Gas Analyzer. A UV or NDIR ab- bration gases to the measurement system at sorption or fluorescence analyzer, to deter- the outlet of the sampling probe when in the mine continuously the SO2 concentration in calibration mode. the sample gas stream. The analyzer shall 5.1.5 Moisture Removal System. A refrig- meet the applicable performance specifica- erator-type condenser or similar device (e.g., tions of Section 4. A means of controlling permeation dryer), to remove condensate the analyzer flow rate and a device for deter- continuously from the sample gas while mining proper sample flow rate (e.g., preci- maintaining minimal contact between the sion rotameter, pressure gauge downstream condensate and the sample gas. The moisture of all flow controls, etc.) shall be provided at removal system is not necessary for analyz- the analyzer. ers that can measure gas concentrations on (NOTE: Housing the analyzer(s) in a clean, a wet basis; for these analyzers, (1) heat the thermally-stable, vibration-free environment sample line and all interface components up will minimize drift in the analyzer calibra- to the inlet of the analyzer sufficiently to tion.) prevent condensation, and (2) determine the 5.1.11 Data Recorder. A strip chart re- moisture content and correct the measured corder, analog computer, or digital recorder, gas concentrations to a dry basis using ap- for recording measurement data. The data propriate methods, subject to the approval of recorder resolution (i.e., readability) shall be the Administrator. The determination of 0.5 percent of span. Alternatively, a digital sample moisture content is not necessary for or analog meter having a resolution of 0.5 percent of span may be used to obtain the pollutant analyzers that measure concentra- analyzer responses and the readings may be tions on a wet basis when (1) a wet basis CO 2 recorded manually. If this alternative is analyzer operated according to Method 3A is used, the readings shall be obtained at equal- used to obtain simultaneous measurements, ly spaced intervals over the duration of the and (2) the pollutant/CO measurements are 2 sampling run. For sampling run durations of used to determine emissions in units of the less than 1 hour, measurements at 1-minute standard. intervals or a minimum of 30 measurements, 5.1.6 Particulate Filter. An in-stack or whichever is less restrictive, shall be ob- heated (sufficient to prevent water condensa- tained. For sampling run durations greater tion) out-of-stack filter. The filter shall be than 1 hour, measurements at 2-minute in- borosilicate or quartz glass wool, or glass tervals or a minimum of 96 measurements, fiber mat. Additional filters at the inlet or whichever is less restrictive, shall be ob- outlet of the moisture removal system and tained. inlet of the analyzer may be used to prevent 5.2 Method 6 Apparatus and Reagents. The accumulation of particulate material in the apparatus and reagents described in Method measurement system and extend the useful 6, and shown by the schematic of the sam- life of the components. All filters shall be pling train in Figure 6C–2, to conduct the in- fabricated of materials that are nonreactive terference check. to the gas being sampled. 5.3 SO2 Calibration Gases. The calibration 5.1.7 Sample Pump. A leak-free pump, to gases for the gas analyzer shall be SO2 in N2 pull the sample gas through the system at a or SO2 in air. Alternatively, SO2/CO2, SO2/O2, flow rate sufficient to minimize the response or SO2/CO2/O2 gas mixtures in N2 may be time of the measurement system. The pump used. For fluorescence-based analyzers, the may be constructed of any material that is O2 and CO2 concentrations of the calibration nonreactive to the gas being sampled. gases as introduced to the analyzer shall be 5.1.8 Sample Flow Rate Control. A sample within 1 percent (absolute) O2 and 1 percent flow rate control valve and rotameter, or (absolute) CO2 of the O2 and Co2 concentra- equivalent, to maintain a constant sampling tions of the effluent samples as introduced to rate within 10 percent. the analyzer. Alternatively, for fluorescence- based analyzers, use calibration blends of (NOTE: The tester may elect to install a SO in air and the nomographs provided by back-pressure regulator to maintain the 2 the vendor to determine the quenching cor- sample gas manifold at a constant pressure rection factor (the effluent O and CO con- in order to protect the analyzer(s) from over- 2 2 centrations must be known). Use three cali- pressurization, and to minimize the need for bration gases as specified below: flow rate adjustments.) 5.3.1 High-Range Gas. Concentration 5.1.9 Sample Gas Manifold. A sample gas equivalent to 80 to 100 percent of the span. manifold, to divert a portion of the sample 5.3.2 Mid-Range Gas. Concentration equiv- gas stream to the analyzer, and the remain- alent to 40 to 60 percent of the span. der to the by-pass discharge vent. The sam- 5.3.3 Zero Gas. Concentration of less than ple gas manifold should also include provi- 0.25 percent of the span. Purified ambient air sions for introducing calibration gases di- may be used for the zero gas by passing air rectly to the analyzer. The manifold may be through a charcoal filter, or through one or
719
VerDate 11
more impingers containing a solution of 3 NOTE: A calibration curve established prior percent H2O2. to the analyzer calibration error check may 6. Measurement System Performance Test Proce- be used to convert the analyzer response to dures the equivalent gas concentration introduced to the analyzer. However, the same correc- Perform the following procedures before tion procedure shall be used for all effluent measurement of emissions (Section 7). and calibration measurements obtained dur- 6.1 Calibration Gas Concentration Ver- ing the test. ification. There are two alternatives for es- tablishing the concentrations of calibration 6.3.2 The analyzer calibration error check gases. Alternative Number 1 is preferred. shall be considered invalid if the gas con- 6.1.1 Alternative Number 1—Use of cali- centration displayed by the analyzer exceeds ± bration gases that are analyzed following the 2 percent of the span for any of the calibra- Environmental Protection Agency tion gases. If an invalid calibration is exhib- Traceability Protocol Number 1 (see Citation ited, take corrective action, and repeat the 1 in the Bibliography). Obtain a certification analyzer calibration error check until ac- from the gas manufacturer that Protocol ceptable performance is achieved. Number 1 was followed. 6.4 Sampling System Bias Check. Perform 6.1.2 Alternative Number 2—Use of cali- the sampling system bias check by introduc- bration gases not prepared according to Pro- ing calibration gases at the calibration valve tocol Number 1. If this alternative is chosen, installed at the outlet of the sampling probe. obtain gas mixtures with a manufacturer’s A zero gas and either the mid-range or high- tolerance not to exceed ±2 percent of the tag range gas, whichever most closely approxi- value. Within 6 months before the emission mates the effluent concentrations, shall be test, analyze each of the calibration gases in used for this check as follows: triplicate using Method 6. Citation 2 in the 6.4.1 Introduce the upscale calibration Bibliography describes procedures and tech- gas, and record the gas concentration dis- niques that may be used for this analysis. played by the analyzer on a form similar to Record the results on a data sheet (example Figure 6C–5. Then introduce zero gas, and is shown in Figure 6C–3). Each of the individ- record the gas concentration displayed by the analyzer. During the sampling system ual SO2 analytical results for each calibra- tion gas shall be within 5 percent (or 5 ppm, bias check, operate the system at the normal whichever is greater) of the triplicate set av- sampling rate, and make no adjustments to erage; otherwise, discard the entire set, and the measurement system other than those repeat the triplicate analyses. If the average necessary to achieve proper calibration gas of the triplicate analyses is within 5 percent flow rates at the analyzer. Alternately intro- of the calibration gas manufacturer’s cyl- duce the zero and upscale gases until a stable inder tag value, use the tag value; otherwise, response is achieved. The tester shall deter- conduct at least three additional analyses mine the measurement system response time until the results of six consecutive runs by observing the times required to achieve a agree with 5 percent (or 5 ppm, whichever is stable response for both the zero and upscale greater) of their average. Then use this aver- gases. Note the longer of the two times as age for the cylinder value. the response time. 6.4.2 The sampling system bias check 6.2 Measurement System Preparation. As- shall be considered invalid if the difference semble the measurement system by follow- between the gas concentrations displayed by ing the manufacturer’s written instructions the measurement system for the analyzer for preparing and preconditioning the gas an- calibration error check and for the sampling alyzer and, as applicable, the other system system bias check exceeds ±5 percent of the components. Introduce the calibration gases span for either the zero or upscale calibra- in any sequence, and make all necessary ad- tion gas. If an invalid calibration is exhib- justments to calibrate the analyzer and the ited, take corrective action, and repeat the data recorder. Adjust system components to sampling system bias check until acceptable achieve correct sampling rates. 6.3 Analyzer Calibration Error. Conduct performance is achieved. If adjustment to the analyzer calibration error check by in- the analyzer is required, first repeat the ana- troducing calibration gases to the measure- lyzer calibration error check, then repeat ment system at any point upstream of the the sampling system bias check. gas analyzer as follows: 7. Emission Test Procedure 6.3.1 After the measurement system has 7.1 Selection of Sampling Site and Sam- been prepared for use, introduce the zero, pling Points. Select a measurement site and mid-range, and high-range gases to the ana- sampling points using the same criteria that lyzer. During this check, make no adjust- are applicable to Method 6. ments to the system except those necessary 7.2 Interference Check Preparation. For to achieve the correct calibration gas flow each individual analyzer, conduct an inter- rate at the analyzer. Record the analyzer re- ference check for at least three runs during sponses to each calibration gas on a form the initial field test on a particular source similar to Figure 6C–4. category. Retain the results, and report
720
VerDate 11
them with each test performed on that dry gas meter reading, meter temperature, source category. and barometric pressure. Recover and ana- If an interference check is being per- lyze the contents of the midget impingers, formed, assemble the modified Method 6 and determine the SO2 gas concentration train (flow control valve, two midget using the procedures of Method 6. (It is not impingers containing 3 percent H2O2, and dry necessary to analyze EPA performance audit gas meter) as shown in Figure 6C–2. Install samples for Method 6.) Determine the aver- the sampling train to obtain a sample at the age gas concentration exhibited by the ana- measurement system sample by-pass dis- lyzer for the run. If the gas concentrations charge vent. Record the initial dry gas meter provided by the analyzer and the modified reading. Method 6 differ by more than 7 percent of the 7.3 Sample Collection. Position the sam- pling probe at the first measurement point, modified Method 6 result, the run is invali- and begin sampling at the same rate as used dated. during the sampling system bias check. 8. Emission Calculation ± Maintain constant rate sampling (i.e., 10 The average gas effluent concentration is percent) during the entire run. The sampling determined from the average gas concentra- time per run shall be the same as for Method tion displayed by the gas analyzer, and is ad- 6 plus twice the system response time. For justed for the zero and upscale sampling sys- each run, use only those measurements ob- tem bias checks, as determined in accord- tained after twice response time of the meas- urement system has elapsed, to determine ance with Section 7.4. The average gas con- the average effluent concentration. If an in- centration displayed by the analyzer may be terference check is being performed, open determined by integration of the area under the flow control valve on the modified Meth- the curve for chart recorders, or by averag- od 6 train concurrent with the initiation of ing all of the effluent measurements. Alter- the sampling period, and adjust the flow to 1 natively, the average may be calculated liter per minute (±10 percent). from measurements recorded at equally (NOTE: If a pump is not used in the modi- spaced intervals over the entire duration of fied Method 6 train, caution should be exer- the run. For sampling run durations of less cised in adjusting the flow rate since over- than 1 hour, measurements at 1-minute in- pressurization of the impingers may cause tervals or a minimum of 30 measurements, leakage in the impinger train, resulting in whichever is less restrictive, shall be used. positively biased results). For sampling run durations greater than 1 7.4 Zero and Calibration Drift Tests. Im- hour, measurements at 2–minute intervals or mediately preceding and following each run, a minimum of 96 measurements, whichever is or if adjustments are necessary for the meas- less restrictive, shall be used. Calculate the urement system during the run, repeat the effluent gas concentration using Equation sampling system bias check procedure de- 6C–1. scribed in Section 6.4 (Make no adjustments C to the measurement system until after the CCC=( − ) ma Eq. 6 C- 1 drift checks are completed.) Record and ana- gas o − lyzer’s responses on a form similar to Figure CCm o 6C–5. Where: 7.4.1 If either the zero or upscale calibra- Cgas = Effluent gas concentration, dry basis, tion value exceeds the sampling system bias ppm. specification, then the run is considered in- C¯ = Average gas concentration indicated by valid. Repeat both the analyzer calibration gas analyzer, dry basis, ppm. error check procedure (Section 6.3) and the sampling system bias check procedure (Sec- Co = Average of initial and final system cali- tion 6.4) before repeating the run. bration bias check responses for the zero 7.4.2 If both the zero and upscale calibra- gas, ppm. tion values are within the sampling system Cm = Average of initial and final system cali- bias specification, then use the average of bration bias check responses for the the initial and final bias check values to cal- upscale calibration gas, ppm. culate the gas concentration for the run. If Cma = Actual concentration of the upscale the zero or upscale calibration drift value ex- calibration gas, ppm. ceeds the drift limits, based on the difference 9. Bibliography between the sampling system bias check re- sponses immediately before and after the 1. Traceability Protocol for Establishing run, repeat both the analyzer calibration True Concentrations of Gases Used for Cali- error check procedure (Section 6.3) and the brations and Audits of Continuous Source sampling system bias check procedure (Sec- Emission Monitors: Protocol Number 1. U.S. tion 6.4) before conducting additional runs. Environmental Protection Agency, Quality 7.5 Interference Check (if performed). Assurance Division. Research Triangle Park, After completing the run, record the final NC. June 1978.
721
VerDate 11
2. Westlin, Peter R. and J. W. Brown. Meth- inder Samples. Source Evaluation Society ods for Collecting and Analyzing Gas Cyl- Newsletter. 3(3):5–15. September 1978.
FIGURE 6C–3—ANALYSIS OF CALIBRATION Analytic method used ———————————— GASES Date —————————————————————
Gas concentration (indicate units) Mid- High- Zero a range b range c
Sample run: 1 ...... 2 ......
722
VerDate 11
Gas concentration (indicate units) Mid- High- Zero a range b range c
3 ...... Average ...... Maximum percent deviation ...... a Average must be less than 0.25 percent of span. b Average must be 50 to 60 percent of span. c Average must be 80 to 90 percent of span.
FIGURE 6C–4—ANALYZER CALIBRATION DATA Date: ———————————————————— Analyzer calibration data for sampling Source identification: ———————————— runs: ———————————————————— Test personnel: ——————————————— Span: ————————————————————
Analyzer Cylinder calibration Absolute Difference value (in- response difference (percent dicate (indicate (indicate of span) units) units) units)
Zero gas ...... Mid-range gas ...... High-range gas ......
FIGURE 6C–5—SYSTEM CALIBRATION BIAS AND Test personnel: ——————————————— DRIFT DATA Date: ———————————————————— Run number: ———————————————— Source identification: ———————————— Span: ————————————————————
Initial values Final values Analyzer System System Drift (per- calibration System cal. bias System cal. bias cent of response calibration (percent calibration (percent span) response of span) response of span)
Zero gas ...... Upscale gas ......
System Cal. Re sponse- Analyzer Cal . Re sponse System Calibration Bias = ×100 Span Final System Cal. Re sponse- Initial System Cal . Re sponse Drift = ×100 Span
METHOD 7—DETERMINATION OF NITROGEN oxide, are measured colorimetrically using OXIDE EMISSIONS FROM STATIONARY SOURCES the phenoldisulfonic acid (PDS) procedure. 1.2 Applicability. This method is applica- 1. Principle and Applicability ble to the measurement of nitrogen oxides 1.1 Principle. A grab sample is collected emitted from stationary sources. The range in an evacuated flask containing a dilute sul- of the method has been determined to be 2 to furic acid-hydrogen peroxide absorbing solu- 400 milligrams NOx (as NO2) per dry standard tion, and the nitrogen oxides, except nitrous cubic meter, without having to dilute the sample.
723
VerDate 11
2. Apparatus tion Agency. The following equipment is 2.1 Sampling (see Figure 7–1). Other grab used in sampling: sampling systems or equipment, capable of 2.1.1 Probe. Borosilicate glass tubing, suf- ficiently heated to prevent water condensa- measuring sample volume to within ±2.0 per- tion and equipped with an in-stack or out- cent and collecting a sufficient sample vol- stack filter to remove particulate matter (a ume to allow analytical reproducibility to plug of glass wool is satisfactory for this ± within 5 percent, will be considered accept- purpose). Stainless steel or Teflon 3 tubing able alternatives, subject to approval of the may also be used for the probe. Heating is Administrator, U.S. Environmental Protec- not necessary if the probe remains dry dur- ing the purging period.
3 Mention of trade names or specific prod- ucts does not constitute endorsement by the Environmental Protection Agency. 724
VerDate 11
2.1.2 Collection Flask. Two-liter 2.1.3 Flask Valve. T-bore stopcock con- borosilicate, round bottom flask, with short nected to a 24/40 standard taper joint. neck and 24/40 standard taper opening, pro- tected against implosion or breakage.
725
VerDate 11
2.1.4 Temperature Gauge. Dial-type ther- 2.3.4 Dropping Pipette or Dropper. Three mometer, or other temperature gauge, capa- required. ble of measuring 1°C (2°F) intervals from ¥5 2.3.5 Polyethylene Policeman. One for to 50°C (25 to 125°F). each sample and each standard. 2.1.5 Vacuum Line. Tubing capable of 2.3.6 Graduated Cylinder. 100 ml with 1-ml withstanding a vacuum of 75 mm Hg (3 in. divisions. Hg) absolute pressure, with ‘‘T’’ connection 2.3.7 Volumetric Flasks. 50 ml (one for and T-bore stopcock. each sample and each standard), 100 ml (one 2.1.6 Vacuum Gauge. U-tube manometer, 1 for each sample and each standard, and one meter (36 in.), with 1-mm (0.1-in.) divisions, for the working standard KNO3 solution), and or other gauge capable of measuring pressure 1000 ml (one). to within ±2.5 mm Hg (0.10 in. Hg). 2.3.8 Spectrophotometer. To measure ab- 2.1.7 Pump. Capable of evacuating the col- sorbance at 410 nm. lection flask to a pressure equal to or less 2.3.9 Graduated Pipette. 10 ml with 0.1-ml than 75 mm Hg (3 in. Hg) absolute. divisions. 2.1.8 Squeeze Bulb. One-way. 2.3.10 Test Paper for Indicating pH. To 2.1.9 Volumetric Pipette. 25 ml. cover the pH range of 7 to 14. 2.1.10 Stopcock and Ground Joint Grease. 2.3.11 Analytical Balance. To measure to A high-vacuum, high-temperature within 0.1 mg. chlorofluorocarbon grease is required. 3. Reagents Halocarbon 25–5S has been found to be effec- tive. Unless otherwise indicated, it is intended 2.1.11 Barometer. Mercury, aneroid, or that all reagents conform to the specifica- other barometer capable of measuring at- tions established by the Committee on Ana- mospheric pressure to within 2.5 mm Hg (0.1 lytical Reagents of the American Chemical in. Hg). In many cases, the barometric read- Society, where such specifications are avail- ing may be obtained from a nearby National able; otherwise, use the best available grade. Weather Service station, in which case the 3.1 Sampling. To prepare the absorbing station value (which is the absolute baro- solution, cautiously add 2.8 ml concentrated metric pressure) shall be requested and an H2SO4 to 1 liter of deionized, distilled water. adjustment for elevation differences between Mix well and add 6 ml of 3 percent hydrogen the weather station and sampling point shall peroxide, freshly prepared from 30 percent be applied at a rate of minus 2.5 mm Hg (0.1 hydrogen peroxide solution. The absorbing in. Hg) per 30 m (100 ft) elevation increase, or solution should be used within 1 week of its vice versa for elevation decrease. preparation. Do not expose to extreme heat 2.2 Sample Recovery. The following or direct sunlight. equipment is required for sample recovery: 3.2 Sample Recovery. Two reagents are 2.2.1 Graduated Cylinder. 50 ml with 1-ml required for sample recovery: divisions. 3.2.1 Sodium Hydroxide (1N). Dissolve 40 g 2.2.2 Storage Containers. Leak-free poly- NaOH in deionized, distilled water and dilute ethylene bottles. to 1 liter. 2.2.3 Wash Bottle. Polyethylene or glass. 3.2.2 Water. Deionized, distilled to con- 2.2.4 Glass Stirring Rod. form to ASTM Specification D1193–77, Type 3 2.2.5 Test Paper for Indicating pH. To (incorporated by reference—see § 60.17). At cover the pH range of 7 to 14. the option of the analyst, the KMnO4 test for 2.3 Analysis. For the analysis, the follow- oxidizable organic matter may be omitted ing equipment is needed: when high concentrations of organic matter 2.3.1 Volumetric Pipettes. Two 1 ml, two 2 are not expected to be present. ml, one 3 ml, one 4 ml, two 10 ml, and one 25 3.3 Analysis. For the analysis, the follow- ml for each sample and standard. ing reagents are required: 2.3.2 Porcelain Evaporating Dishes. 175- to 3.3.1 Fuming Sulfuric Acid. 15 to 18 per- 250-ml capacity with lip for pouring, one for cent by weight free sulfur trioxide. HANDLE each sample and each standard. The Coors WITH CAUTION. No. 45006 (shallow-form, 195 ml) has been 3.3.2 Phenol. White solid. found to be satisfactory. Alternatively, 3.3.3 Sulfuric Acid. Concentrated, 95 per- polymethyl pentene beakers (Nalge No. 1203, cent minimum assay. HANDLE WITH CAU- 150 ml), or glass beakers (150 ml) may be TION. used. When glass beakers are used, etching of 3.3.4 Potassium Nitrate. Dried at 105 to the beakers may cause solid matter to be 110°C (220 to 230°F) for a minimum of 2 hours present in the analytical step; the solids just prior to preparation of standard solu- should be removed by filtration (see Section tion. 4.3). 3.3.5 Standard KNO3 Solution. Dissolve 2.3.3 Steam Bath. Low-temperature ovens exactly 2.198 g of dried potassium nitrate or thermostatically controlled hot plates (KNO3) in deionized, distilled water and di- kept below 70°C (160°F) are acceptable alter- lute to 1 liter with deionized, distilled water natives. in a 1,000-ml volumetric flask.
726
VerDate 11
3.3.6 Working Standard KNO3 Solution. the flask valve area, heat the probe and Dilute 10 ml of the standard solution to 100 purge until the condensation disappears. ml with deionized distilled water. One Next, turn the pump valve to its ‘‘vent’’ posi- milliter of the working standard solution is tion. Turn the flask valve clockwise to its equivalent to 100 µg nitrogen dioxide (NO2). ‘‘evacuate’’ position and record the dif- 3.3.7 Water. Deionized, distilled as in Sec- ference in the mercury levels in the manom- tion 3.2.2. eter. The absolute internal pressure in the 3.3.8 Phenoldisulfonic Acid Solution. Dis- flask (Pi) is equal to the barometric pressure solve 25 g of pure white phenol in 150 ml con- less the manometer reading. Immediately centrated sulfuric acid on a steam bath. turn the flask valve to the ‘‘sample’’ position Cool, add 75 ml fuming sulfuric acid, and and permit the gas to enter the flask until heat at 100°C (212°F) for 2 hours. Store in a pressures in the flask and sample line (i.e., dark, stoppered bottle. duct, stack) are equal. This will usually re- 3.3.9 Quality Assurance Audit Samples. quire about 15 seconds; a longer period indi- Nitrate samples in glass vials prepared by cates a ‘‘plug’’ in the probe, which must be EPA’s Environmental Monitoring Systems corrected before sampling is continued. After Laboratory, Quality Assurance Division, collecting the sample, turn the flask valve to Source Branch, Mail Drop 77A, Research Tri- its ‘‘purge’’ position and disconnect the flask angle Park, North Carolina 27711. Each set from the sampling train. Shake the flask for will consist of two vials having solutions of at least 5 minutes. unknown concentrations. Only when making 4.1.2 If the gas being sampled contains in- compliance determinations, obtain an audit sufficient oxygen for the conversion of NO to sample set from the quality assurance man- NO2 (e.g., an applicable subpart of the stand- agement office at each EPA regional office ard may require taking a sample of a calibra- or the responsible enforcement agency. tion gas mixture of NO in N2), then oxygen (NOTE: The tester should notify the quality shall be introduced into the flask to permit assurance office or the responsible enforce- this conversion. Oxygen may be introduced ment agency at least 30 days prior to the test into the flask by one of three methods; (1) date to allow sufficient time for sample de- Before evacuating the sampling flask, flush livery.) with pure cylinder oxygen, then evacuate flask to 75 mm Hg (3 in. Hg) absolute pres- 4. Procedures sure or less; or (2) inject oxygen into the 4.1 Sampling. flask after sampling; or (3) terminate sam- 4.1.1 Pipette 25 ml of absorbing solution pling with a minimum of 50 mm Hg (2 in. Hg) into a sample flask, retaining a sufficient vacuum remaining in the flask, record this quantity for use in preparing the calibration final pressure, and then vent the flask to the standards. Insert the flask valve stopper into atmosphere until the flask pressure is al- the flask with the valve in the ‘‘purge’’ posi- most equal to atmospheric pressure. tion. Assemble the sampling train as shown 4.2 Sample Recovery. Let the flask set for in Figure 7–1 and place the probe at the sam- a minimum of 16 hours and then shake the pling point. Make sure that all fitings are contents for 2 minutes. Connect the flask to tight and leak-free, and that all ground glass a mercury filled U-tube manometer. Open joints have been properly greased with a the valve from the flask to the manometer high-vacuum, high-temperature chlorofluor- and record the flask temperature (Tf), the ocarbon-based stopcock grease. Turn the barometric pressure, and the difference be- flask valve and the pump valve to their tween the mercury levels in the manometer. ‘‘evacuate’’ positions. Evacuate the flask to The absolute internal pressure in the flask 75 mm Hg (3 in. Hg) absolute pressure, or (Pf) is the barometric pressure less the ma- less. Evacuation to a pressure approaching nometer reading. Transfer the contents of the vapor pressure of water at the existing the flask to a leak-free polyethylene bottle. temperature is desirable. Turn the pump Rinse the flask twice with 5-ml portions of valve to its ‘‘vent’’ position and turn the off deionized, distilled water and add the rinse the pump. Check for leakage by observing water to the bottle. Adjust the pH to be- the manometer for any pressure fluctuation. tween 9 and 12 by adding sodium hydroxide (1 (Any variation greater than 10 mm Hg (0.4 in. N), dropwise (about 25 to 35 drops). Check the Hg) over a period of 1 minute is not accept- pH by dipping a stirring rod into the solution able, and the flask is not to be used until the and then touching the rod to the pH test leakage problem is corrected. Pressure in the paper. Remove as little material as possible flask is not to exceed 75 mm Hg (3 in. Hg) ab- during this step. Mark the height of the liq- solute at the time sampling is commenced.) uid level so that the container can be Record the volume of the flask and valve checked for leakage after transport. Label (Vf), the flask temperature (Ti), and the baro- the container to clearly identify its con- metric pressure. Turn the flask valve coun- tents. Seal the container for shipping. terclockwise to its ‘‘purge’’ position and do 4.3 Analysis. Note the level of the liquid the same with the pump valve. Purge the in container and confirm whether or not any probe and the vacuum tube using the squeeze sample was lost during shipment; note this bulb. If condensation occurs in the probe and on the analytical data sheet. If a noticeable
727
VerDate 11
amount of leakage has occurred, either void for the same enforcement agency within 30 the sample or use methods, subject to the ap- days is not required. An audit sample set proval of the Administrator, to correct the may not be used to validate different sets of final results. Immediately prior to analysis, compliance samples under the jurisdiction of transfer the contents of the shipping con- different enforcement agencies, unless prior tainer to a 50-ml volumetric flask, and rinse arrangements are made with both enforce- the container twice with 5-ml portions of de- ment agencies. ionized, distilled water. Add the rinse water Calculate the concentrations in mg/dscm to the flask and dilute to the mark with de- using the specified sample volume in the ionized, distilled water; mix thoroughly. Pi- audit instructions. (NOTE: Indication of ac- pette a 25-ml aliquot into the procelain ceptable results may be obtained imme- evaporating dish. Return any unused portion diately by reporting the audit results in mg/ of the sample to the polyethylene storage dscm and compliance results in total µg NO / bottle. Evaporate the 25-ml aliquot to dry- 2 ness on a steam bath and allow to cool. Add sample by telephone to the responsible en- 2 ml phenoldisulfonic acid solution to the forcement agency.) Include the results of dried residue and triturate thoroughly with both audit samples, their identification a polyethylene policeman. Make sure the so- numbers, and the analyst’s name with the lution contacts all the residue. Add 1 ml de- results of the compliance determination ionized, distilled water and four drops of con- samples in appropriate reports to the EPA centrated sulfuric acid. Heat the solution on regional office or the appropriate enforce- a steam bath for 3 minutes with occasional ment agency. Include this information with stirring. Allow the solution to cool, add 20 subsequent compliance analyses for the same ml deionized, distilled water, mix well by enforcement agency during the 30-day pe- stirring, and add concentrated ammonium riod. hydroxide, dropwise, with constant stirring, The concentrations of the audit samples until the pH is 10 (as determined by pH obtained by the analyst shall agree within 10 paper). If the sample contains solids, these percent of the actual audit concentrations. If must be removed by filtration (centrifuga- the 10-percent specification is not met, rean- tion is an acceptable alternative, subject to alyze the compliance samples and audit sam- the approval of the Administrator), as fol- ples and include initial and reanalysis values lows: filter through Whatman No. 41 filter in the test report (see NOTE in the first para- paper into a 100-ml volumetric flask; rinse graph of this section). the evaporating dish with three 5-ml por- Failure to meet the 10-percent specifica- tions of deionized, distilled water; filter tion may require retests until the audit these three rinses. Wash the filter with at problems are resolved. However, it the audit least three 15-ml portions of deionized, dis- results do not affect the compliance or non- tilled water. Add the filter washings to the contents of the volumetric flask and dilute compliance status of the affected facility, to the mark with deionized, distilled water. the Administrator may waive the reanalysis If solids are absent, the solution can be requirement, further audits, or retests and transferred directly to the 100-ml volumetric accept the results of the compliance test. flask and diluted to the mark with deionized, While steps are being taken to resolve audit distilled water. Mix the contents of the flask analysis problems, the Administrator may thoroughly, and measure the absorbance at also choose to use the data to determine the the optimum wavelength used for the stand- compliance or noncompliance status of the ards (Section 5.2.1), using the blank solution affected facility. as a zero reference. Dilute the sample and 5. Calibration the blank with equal volumes of deionized, distilled water if the absorbance exceeds A , 5.1 Flask Volume. The volume of the col- 4 lection flask-flask valve combination must the absorbance of the 400 µg NO2 standard (see Section 5.2.2). be known prior to sampling. Assemble the 4.4 Audit Sample Analysis. Concurrently flask and flask valve and fill with water, to analyze the two audit samples and a set of the stopcock. Measure the volume of water ± compliance samples (Section 4.3) in the same to 10 ml. Record this volume on the flask. manner to evaluate the technique of the ana- 5.2 Spectrophotometer Calibration. lyst and the standards preparation. (NOTE: It 5.2.1 Optimum Wavelength Determina- is recommended that known quality control tion. Calibrate the wavelength scale of the samples be analyzed prior to the compliance spectrophotometer every 6 months. The cali- and audit sample analysis to optimize the bration may be accomplished by using an en- system accuracy and precision. One source of ergy source with an intense line emission these samples is the Source Branch listed in such as a mercury lamp, or by using a series Section 3.3.9.) The same analysts, analytical of glass filters spanning the measuring range reagents, and analytical system shall be used of the spectrophotometer. Calibration mate- both for the compliance samples and the rials are available commercially and from EPA audit samples; if this condition is met, the National Bureau of Standards. Specific auditing of subsequent compliance analyses details on the use of such materials should
728
VerDate 11
be supplied by the vendor; general informa- 5.2.3 Spectrophotometer Calibration Qual- tion about calibration techniques can be ob- ity Control. Multiply the absorbance value tained from general reference books on ana- obtained for each standard by the Kc factor lytical chemistry. The wavelength scale of (least squares slope) to determine the dis- the spectrophotometer must read correctly tance each calibration point lies from the within ±5 nm at all calibration points; other- theoretical calibration line. These calculated wise, the spectrophotometer shall be re- concentration values should not differ from paired and recalibrated. Once the wavelength the actual concentrations (i.e., 100, 200, 300, scale of the spectrophotometer is in proper and 400 µg NO2) by more than 7 percent for calibration, use 410 nm as the optimum three of the four standards. wavelength for the measurement of the ab- 5.3 Barometer. Calibrate against a mer- sorbance of the standards and samples. cury barometer. Alternatively, a scanning procedure may 5.4 Temperature Gauge. Calibrate dial be employed to determine the proper meas- thermometers against mercury-in-glass ther- uring wavelength. If the instrument is a dou- mometers. ble-beam spectrophotometer, scan the spec- 5.5 Vacuum Gauge. Calibrate mechanical trum between 400 and 415 nm using a 200 µg gauges, if used, against a mercury manom- NO2 standard solution in the sample cell and eter such as that specified in 2.1.6. a blank solution in the reference cell. If a 5.6 Analytical Balance. Calibrate against peak does not occur, the spectrophotometer standard weights. is probably malfunctioning and should be re- 6. Calculations paired. When a peak is obtained within the 400 to 415 nm range, the wavelength at which Carry out the calculations, retaining at this peak occurs shall be the optimum wave- least one extra decimal figure beyond that of length for the measurement of absorbance of the acquired data. Round off figures after both the standards and the samples. For a final calculations. single-beam spectrophotometer, follow the 6.1 Nomenclature. scanning procedure described above, except A=Absorbance of sample. that the blank and standard solutions shall C=Concentration of NOx as NO2, dry basis, be scanned separately. The optimum wave- corrected to standard conditions, mg/ length shall be the wavelength at which the dscm (lb/dscf). maximum difference in absorbance between F=Dilution factor (i.e., 25/5, 25/10, etc., re- the standard and the blank occurs. quired only if sample dilution was needed 5.2.2 Determination of Spectrophotometer to reduce the absorbance into the range of calibration). Calibration Factor Kc. Add 0.0 ml, 2 ml, 4 ml, K =Spectrophotometer calibration factor. 6 ml., and 8 ml of the KNO3 working standard c m=Mass of NO as NO in gas sample, µg. solution (1 ml=100 µg NO2) to a series of five x 2 50-ml volumetric flasks. To each flask, add Pf=Final absolute pressure of flask, mm Hg 25 ml of absorbing solution, 10 ml deionized, (in. Hg). distilled water, and sodium hydroxide (1 N) Pi=Initial absolute pressure of flask, mm Hg dropwise until the pH is between 9 and 12 (in. Hg). (about 25 to 35 drops each). Dilute to the Pstd=Standard absolute pressure, 760 mm Hg mark with deionized, distilled water. Mix (29.92 in. Hg). thoroughly and pipette a 25-ml aliquot of Tf=Final absolute temperature of flask, °K each solution into a separate porcelain (°R). evaporating dish. Beginning with the evapo- Ti=Initial absolute temperature of flask, °K ration step, follow the analysis procedure of (°R). Section 4.3 until the solution has been trans- Tstd=Standard absolute temperature 293°K ferred to the 100 ml volumetric flask and di- (528°R). luted to the mark. Measure the absorbance Vsc=Sample volume at standard conditions of each solution, at the optimum wave- (dry basis), ml. length, as determined in Section 5.2.1. This Vf=Volume of flask and valve, ml. calibration procedure must be repeated on Va=Volume of absorbing solution, 25 ml. each day that samples are analyzed. Cal- 2=50/25, the aliquot factor. (If other than a 25- culate the spectrophotometer calibration ml aliquot was used for analysis, the cor- factor as follows: responding factor must be substituted). 6.2 Sample Volume, Dry Basis, Corrected to AAAA+2 + 3 + 4 Standard Conditions. K = 100 1 2 3 4 Eq.7- 1 c 2+ 2 + 2 + 2 Vsc= (Tstd/Pstd)(Vf ¥Va)(Pf/Tf¥Pi/Ti) AAAA1 2 1 4 =K1 (Vf ¥25 ml)(Pf/Tf¥ Pi/Ti) Eq. 7–2 Where: Where: ° Kc=Calibration factor, µg. K1=0.3858 K/mm Hg for metric units ° A1=Absorbance of the 100-µg NO2 standard. =17.64 R/in. Hg for English units. A2=Absorbance of the 200-µg NO2 standard. 6.3 Total µg NO2 Per Sample. A3=Absorbance of the 300-µg NO2 standard. A4=Absorbance of the 400-µg NO2 standard. m= 2Kc A F Eq. 7–3 729
VerDate 11
NOTE: If other than a 25-ml aliquot is used 1.1 Applicability. This method applies to for analysis, the factor 2 must be replaced by the measurement of nitrogen oxides emitted a corresponding factor. from stationary sources; it may be used as 6.4 Sample Concentration, Dry Basis, Cor- an alternative to Method 7 (as defined in 40 rected to Standard Conditions. CFR Part 60.8(b)) to determine compliance if the stack concentration is within the analyt- ical range. The analytical range of the meth- 3 od is from 125 to 1,250 mg NOx/m as NO2 (65 to 655 ppm), and higher concentrations may be analyzed by diluting the sample. The Where: lower detection limit is approximately 19 3 3 K2=10 (mg/scm)/(µg/ml) for metric units. mg/m (10 ppm), but may vary among instru- =6.242×10¥5 (lb/scf)/(µg/ml) for English ments. units. 1.2 Principle. A grab sample is collected To convert from mg/dscm to g/dscm, divide C in an evacuated flask containing a diluted by 1,000. sulfuric acid-hydrogen peroxide absorbing 6.5 Relative Error (RE) for QA Audit Sam- solution. The nitrogen oxides, except nitrous ples, Percent. oxide, are oxidized to nitrate and measured by ion chromatography. CC− 2. Apparatus RE = d a ().100Eq 7- 5 C 2.1 Sampling. Same as in Method 7, Sec- a tion 2.1. Where: 2.2 Sampling Recovery. Same as in Meth-
Cd=Determined audit sample concentration, od 7, Section 2.2, except the stirring rod and mg/dscm. pH paper are not needed. Ca=Actual audit sample concentration, mg/ 2.3 Analysis. For the analysis, the follow- dscm. ing equipment is needed. Alternative instru- 7. Bibliography mentation and procedures will be allowed 1. Standard Methods of Chemical Analysis provided the calibration precision in Section 6th ed. New York, D. Van Nostrand Co., Inc. 5.2 and acceptable audit accuracy can be 1962. Vol. 1, p. 329–330. met. 2. Standard Method of Test for Oxides of 2.3.1 Volumetric Pipets. Class A; 1-, 2-, 4- Nitrogen in Gaseous Combustion Products , 5-ml (two for the set of standards and one (Phenoldisulfonic Acid Procedure). In: 1968 per sample), 6-, 10-, and graduated 5-ml sizes. Book of ASTM Standards, Part 26. Philadel- 2.3.2 Volumetric Flasks. 50–ml (two per phia, PA. 1968. ASTM Designation D–1608–60, sample and one per standard), 200-ml, and 1- p. 725–729. liter sizes. 3. Jacob, M. B. The Chemical Analysis of 2.3.3 Analytical Balance. To measure to Air Pollutants. New York. Interscience Pub- within 0.1 mg. lisher, Inc. 1960. Vol. 10, p. 351–356. 2.3.4 Ion Chromatograph. The ion chro- 4. Beatty, R. L., L. B. Berger, and H. H. matograph should have at least the following Schrenk. Determination of Oxides of Nitro- components: gen by the Phenoldisulfonic Acid Method. 2.3.4.1 Columns. An anion separation or Bureau of Mines, U.S. Dept. of Interior. RI. other column capable of resolving the ni- 3687. February 1943. trate ion from sulfate and other species 5. Hamil, H. F. and D. E. Camann. Collabo- present and a standard anion suppressor col- rative Study of Method for the Determina- umn (optional). Suppressor columns are pro- tion of Nitrogen Oxide Emissions from Sta- duced as proprietary items; however, one can tionary Sources (Fossil Fuel-Fired Steam be produced in the laboratory using the resin Generators). Southwest Research Institute available from BioRad Company, 32nd and report for Environmental Protection Agen- Griffin Streets, Richmond, CA. Peak resolu- cy. Research Triangle Park, NC. October 5, tion can be optimized by varying the efluent 1973. strength or column flow rate, or by experi- 6. Hamil, H. F. and R. E. Thomas. Collabo- menting with alternative columns that may rative Study of Method for the Determina- offer more efficient separation. When using tion of Nitrogen Oxide Emissions from Sta- guard columns with the stronger reagent to tionary Sources (Nitric Acid Plants). South- protect the separation column, the analyst west Research Institute report for Environ- should allow rest periods between injection mental Protection Agency. Research Tri- intervals to purge possible sulfate buildup in angle Park, NC. May 8, 1974. the guard column. 2.3.4.2 Pump. Capable of maintaining a METHOD 7A—DETERMINATION OF NITROGEN steady flow as required by the system. OXIDE EMISSIONS FROM STATIONARY 2.3.4.3 Flow Gauges. Capable of measuring SOURCES—ION CHROMATOGRAPHIC METHOD the specified system flow rate. 1. Applicability and Principle 2.3.4.4 Conductivity Detector.
730
VerDate 11
2.3.4.5 Recorder. Compatible with the out- Do not store the samples more than 4 days put voltage range of the detector. between collection and recovery. 3. Reagents 4.3 Sample. Preparation. the level of the liquid in the container and confirm whether Unless otherwise indicated, it is intended any sample was lost during shipment; note that all reagents conform to the specifica- this on the analytical data sheet. If a notice- tions established by the Committee on Ana- able amount of leakage has occurred, either lytical Reagents of the American Chemical void the sample or use methods, subject to Society, where such specifications are avail- the approval of the Administrator, to correct able; otherwise, use the best available grade. the final results. Immediately before analy- 3.1 Sampling. An absorbing solution con- sis, transfer the contents of the shipping sisting of sulfuric acid (H SO ) and hydrogen 2 4 container to a 50-ml volumetric flask, and peroxide (H O ) is required for sampling. To 2 2 rinse the container twice with 5-ml portions prepare the absorbing solution, cautiously of water. Add the rinse water to the flask, add 2.8 ml concentrated H SO to a 1-liter 2 4 and dilute to the mark with water. Mix thor- flask containing water (same as Section 3.2). oughly. Add 6 ml of 3 percent H O that has been 2 2 Pipet a 5-ml aliquot of the sample into a freshly prepared from 30 percent solution. 50-ml volumetric flask, and dilute to the Dilute to volume with water, and mix well. mark with water. Mix thoroughly. For each This absorbing solution should be used with- set of determinations, prepare a reagent in 1 week of its preparation. Do not expose to blank by diluting 5 ml of absorbing solution extreme heat or direct sunlight. to 50 ml with water. (Alternatively, eluent NOTE: Biased testing results have been ob- solution may be used in all sample, standard, served when sampling under conditions of and blank dilutions.) high sulfur dioxide concentrations (above 4.4 Analysis. Prepare a standard calibra- 2000 ppm). tion curve according to Section 5.2. Analyze 3.2 Sample Recovery. Deionized distilled the set of standards followed by the set of water that conforms to American Society for samples using the same injection volume for Testing and Materials Specification D 1193– both standards and samples. Repeat this 74, Type 3, is required for sample recovery. analysis sequence followed by a final analy- At the option of the analyst, the KMnO4 test sis of the standard set. Average the results. for oxidizable organic matter may be omit- The two sample values must agree within 5 ted when high concentrations of organic percent of their mean for the analysis to be matter are not expected to be present. valid. Perform this duplicate analysis se- 3.3 Analysis. For the analysis, the follow- quence on the same day. Dilute any sample ing reagents are required: and the blank with equal volumes of water if 3.3.1 Water. Same as in Section 3.2. the concentration exceeds that of the high- 3.3.2 Stock Standard Solution, 1 mg NO2/ est standard. ml. Dry an adequate amount of sodium ni- Document each sample chromatogram by 1 trate (NaNO3) at 105 to 110 ⁄2 C for a mini- listing the following analytical parameters: mum of 2 hours just before preparing the injection point, injection volume, nitrate standard solution. Then dissolve exactly and sulfate retention times, flow rate, detec- 1.847 g of dried NaNO3 in water, and dilute to tor sensitivity setting, and recorder chart 1 liter in a volumetric flask. Mix well. This speed. solution is stable for 1 month and should not 4.5 Audit Sample Analysis. Same as re- be used beyond this time. quired in Method 7. 3.3.3 Working Standard Solution, 25µg/ml. 5. Calibration Dilute 5 ml of the standard solution to 200 ml 5.1 Flask Volume. Same as in Method 7, with water in a volumetric flask, and mix Section 5.1. well. 5.2 Standard Calibration Curve. Prepare a 3.3.4 Eluent Solution. Weight 1.018 g of so- series of five standards by adding 1.0, 2.0, 4.0, dium carbonate (Na CO ) and 1.008 g of so- 2 3 6.0, and 10.0 ml of working standard solution dium bicarbonate (NaHCO ), and dissolve in 4 3 (25µg/ml) to a series of five 50-ml volumetric liters of water. This solution is 0.0024 M flasks. (The standard masses will equal 25, Na CO /0.003 M NaHCO Other eluents appro- 2 3 3. 50, 100, 150, and 250µg.) Dilute each flask to priate to the column type and capable of re- volume with water, and mix well. Analyze solving nitrate ion from sulfate and other with the samples as described in Section 4.4 species present may be used. and subtract the blank from each value. Pre- 3.3.5 Quality Assurance Audit Samples. pare or calculate a linear regression plot to Same as required in Method 7. the standard masses in µg (x-axis) versus 4. Procedure their peak height responses in millimeters 4.1 Sampling. Same as in Method 7, Sec- (y-axis). (Take peak height measurements tion 4.1. with symmetrical peaks; in all other cases, 4.2 Sample Recovery. Same as in Method calculate peak areas.) From this curve, or 7, Section 4.2, except delete the steps on ad- equation, determine the slope, and calculate justing and checking the pH of the sample. its reciprocal to denote as the calibration
731
VerDate 11
factor, S. If any point deviates from the line 2. Sawicki, E., J. D. Mulik, and E. by more than 7 percent of the concentration Wittgenstein. Ion Chromatographic Analysis at that point, remake and reanalyze that of Environmental Pollutants. Ann Arbor, standard. This deviation can be determined Ann Arbor Science Publishers, Inc. Vol. 1. by multiplying S times the peak height re- 1978. sponse for each standard. The resultant con- 3. Siemer, D. D. Separation of Chloride and centrations must not differ by more than 7 Bromide from Complex Matrices Prior to Ion percent from each known standard mass (i.e., Chromatographic Determination. Analytical 25, 50, 100, 150, and 250µg). Chemistry 52(12:1874–1877). October 1980. 5.3 Conductivity Detector. Calibrate ac- 4. Small, H., T. S. Stevens, and W. C. cording to manufacturer’s specifications Bauman. Novel Ion Exchange prior to initial use. Chromatographic Method Using 5.4 Barometer. Calibrate against a mer- Conductimetric Determination. Analytical cury barometer. Chemistry. 47(11:1801). 1975. 5.5 Temperature Gauge. Calibrate dial 5. Yu, King K. and Peter R. Westlin. Eval- thermometers against mercury-in-glass ther- uation of Reference Method 7 Flask Reaction mometers. Time. Source Evaluation Society News- 5.6 Vacuum Gauge. Calibrate mechanical letter. 4(4). November 1979. 10 p. gauges, if used, against a mercury manom- eter such as that specified in Section 2.1.6 of METHOD 7B—DETERMINATION OF NITROGEN Method 7. OXIDE EMISSIONS FROM STATIONARY 5.7 Analytical Balance. Calibrate against SOURCES (ULTRAVIOLET standard weights. SPECTROPHOTOMETRY) 6. Calculations 1. Applicability and Principle Carry out the calculations, retaining at 1.1 Applicability. This method is applica- least one extra decimal figure beyond that of ble to the measurement of nitrogen oxides the acquired data. Round off figures after emitted from nitric acid plants. The range of final calculations. the method as outlined has been determined 6.1 Sample Volume. Calculate the sample to be 57 to 1,500 milligrams NOx (as NO2) per volume Vsc (in ml) on a dry basis, corrected dry standard cubic meter, or 30 to 786 ppm to standard conditions, using Equation 7–2 of NOx (as NO2), assuming corresponding stand- Method 7. ards are prepared. 6.2 Sample Concentration of NOx as NO2. 1.2 Principle. A grab sample is collected Calculate the sample concentration C (in mg/ in an evacuated flask containing a dilute sul- dscm) as follows: furic acid-hydrogen peroxide absorbing solu- 4 tion; and the nitrogen oxides, except nitrous HSF ×10 oxide, are measured by ultraviolet absorp- C = Eq. 7 A- 1 tion. Vsc 2. Apparatus Where: 2.1 Sampling. Same as Method 7, Section H =Sample peak height, mm. 2.1.1 through Section 2.1.11. S =Calibration factor, µg/mm. 2.2 Sample Recovery. The following F =Dilution factor (required only if sample equipment is required for sample recovery: dilution was needed to reduce the con- 2.2.1 Wash Bottle. Polyethylene or glass. centration into the range of calibration) 2.2.2 Volumetric Flasks. 100-ml (one for 104 = 1:10 dilution times conversion factor of each sample). 2.3 Analysis. The following equipment is mg 106ml needed for analysis: × 2.3.1 Volumetric Pipettes. 5-, 10-, 15-, and 3 3 10µ g m 20-ml to make standards and sample dilu- To convert from mg/dscm to g/dscm, divide C tions. by 1000. 2.3.2 Volumetric Flasks. 1000- and 100-ml If desired, the concentration of NO may be for preparing standards and dilution of sam- 2 ples. calculated as ppm NO2 at standard condi- tions as follows: 2.3.3 Spectrophotometer. To measure ul- traviolet absorbance at 210 nm. ppm NO= 0. 5228 C Eq . 7 A -2 2.3.4 Analytical Balance. To measure to 2 within 0.1 mg. Where: 3. Reagents 0.5228 = ml/mg NO 2. Unless otherwise indicated, all reagents 7. Bibliography are to conform to the specifications estab- 1. Mulik, J. D. and E. Sawicki. Ion lished by the committee on analytical re- Chromatographic Analysis of Environmental agents of the American Chemical Society, Pollutants. Ann Arbor, Ann Arbor Science where such specifications are available. Oth- Publishers, Inc. Vol. 2, 1979. erwise, use the best available grade.
732
VerDate 11
3.1 Sampling. Same as Method 7, Section the compliance determination samples. The 3.1. It is important that the amount of hy- relative error will be determined by the re- drogen peroxide in the absorbing solution gional office or the appropriate enforcement not be increased. Higher concentrations of agency. peroxide may interfere with sample analysis. 5. Calibration 3.2 Sample Recovery. Same as for Method 7, Section 3.2.2. Same as Method 7, Section 5.1 and Sections 3.3 Analysis. Same as for Method 7, Sec- 5.3 through 5.6 with the addition of the fol- tions 3.3.4, 3.3.5, and 3.3.7 with the addition of lowing: the following: 5.1 Determination of Spectrophotometer 3.3.1 Working Standard KNO3 Solution. Standard Curve. Add 0.0 ml, 5 ml, 10 ml, 15 Dilute 10 ml of the standard solution to 1000 ml, and 20 ml of the KNO3 working standard ml with water. One milliliter of the working solution (1 ml= 10 µg NO2) to a series to five standard is equivalent to 10 µg nitrogen diox- 100-ml volumetric flasks. To each flask, add ide (NO2). 5 ml of absorbing solution. Dilute to the 3.3.2 Absorbing Solution. Same as in Sec- mark with water. The resulting solutions tion 3.1. contain 0.0, 50, 100, 150, and 200 µg NO , re- 3.3.3 Quality Assurance Audit Samples. 2 spectively. Measure the absorbance by ultra- Nitrate samples are prepared in glass vials violet spectrophotometry at 210 nm, using by the Environmental Protection Agency (EPA), Environmental Monitoring Systems the blank as a zero reference. Prepare a µ Laboratory, Research Triangle Park, North standard curve plotting absorbance vs. g Carolina. Each set will consist of two vials NO2. with two unknown concentrations. When NOTE: If other than a 20-ml aliquot of sam- making compliance determinations, obtain ple is used for analysis, then the amount of the audit samples from the quality assurance absorbing solution in the blank and stand- management office at each EPA regional of- ards must be adjusted such that the same fice. amount of absorbing solution is in the blank 4. Procedures and standards as is in the aliquot of sample 4.1 Sampling. Same as Method 7, Sections used. Calculate the spectrophotometer cali- 4.1.1 and 4.1.2. bration factor Kc as follows: 4.2 Sample Recovery. Let the flask sit for a minimum of 16 hours, and then shake the contents for 2 minutes. Connect the flask to a mercury filled U-tube manometer. Open the valve from the flask to the manometer, and record the flask temperature (Tf), the barometric pressure, and the difference be- tween the mercury levels in the manometer. The absolute internal pressure in the flask Where: (Pf) is the barometric pressure less the ma- nometer reading. mi=Mass of NO2 in standard i, µg. Transfer the contents of the flask to a 100- Ai=Absorbance of NO2 standard i. ml volumetric flask. Rinse the flask three N=Total number of calibration standards. times with 10-ml portions of water, and add For the set of calibration standards speci- to the volumetric flask. Dilute to 100 ml fied here, Equation 7–1 simplifies to the fol- with water. Mix thoroughly. The sample is lowing: now ready for analysis. 4.3 Analysis. Pipette a 20-ml aliquot of A + 2A + 3A + 4A sample into a 100-ml volumetric flask. Dilute 1 2 3 4 Kc = 50 2 2 2 2 Eq. 7B-2 to 100 ml with water. The sample is now A1 + A2 + A3 + A4 ready to be read by ultraviolet 6. Calculations spectrophotometry. Using the blank as zero reference, read the absorbance of the sample Same as Method 7, Sections 6.1, 6.2, and 6.4 at 210 nm. with the addition of the following: 4.4 Audit Analysis. With each set of com- 6.1 Total µg NO2 Per Sample: pliance samples or once per analysis day, or m=5K AF Eq. 7B–3 once per week when averaging continuous c samples, analyze each performance audit in Where: the same manner as the sample to evaluate 5=100/20, the aliquot factor. the analyst’s technique and standard prepa- NOTE: If other than a 20-ml aliquot is used ration. The same person, the same reagents, for analysis, the factor 5 must be replaced by and the same analytical system must be used a corresponding factor. both for compliance determination samples and the EPA audit samples. Report the re- 6.2 Relative Error (RE) for Quality Assur- sults of all audit samples with the results of ance Audits.
733
VerDate 11
NOx (NO+NO2) emissions are oxidized to NO2– o and NO –. The NO – is reduced to NO – with CCd a 3 3 2 RE = ×100 Eq. 7B-4 cadmium, and the NO2– is analyzed C a colorimetrically. 1.3 Interferences. Possible interferences Where: are SO2 and NH3. High concentrations of SO2 Cd=Determined audit concentration. could interfere because SO2 consumes MnO4– Ca=Actual audit concentration. (as does NOx) and, therefore, could reduce the 7. Bibliography NOx collection efficiency. However, when 1. National Institute for Occupational sampling emissions from a coal-fired electric Safety and Health Recommendations for Oc- utility plant burning 2.1-percent sulfur coal cupational Exposure to Nitric Acid. In: Occu- with no control of SO2 emissions, collection pational Safety and Health Reporter. Wash- efficiency was not reduced. In fact, calcula- ington, DC. Bureau of National Affairs, Inc. tions show that sampling 3000 ppm SO2 will 1976. p. 149. reduce the MnO4– concentration by only 5 2. Rennie, P.J., A.M. Sumner, and F.B. percent if all the SO2 is consumed in the first Basketter. ‘‘Determination of Nitrate in impinger. Raw, Potable, and Waste Waters by Ultra- NH3 is slowly oxidized to NO3– by the ab- violet Spectrophotometry.’’ ‘‘Analyst.’’ Vol. sorbing solution. At 100 ppm NH3 in the gas 104. September 1979. p. 837. stream, an interference of 6 ppm NOx (11 mg 3 NO2/m ) was observed when the sample was METHOD 7C—DETERMINATION OF NITROGEN analyzed 10 days after collection. Therefore, OXIDE EMISSIONS FROM STATIONARY the method may not be applicable to plants SOURCES—ALKALINE-PERMANGANATE/COL- using NH3 injection to control NOx emissions ORIMETRIC METHOD unless means are taken to correct the re- 1. Applicability, Principle, Interferences, Preci- sults. An equation has been developed to sion, Bias, and Stability allow quantitation of the interference and is discussed in Citation 5 of the Bibliography. 1.1 Applicability. The method is applica- 1.4 Precision and Bias. The method does ble to the determination of NO emissions x not exhibit any bias relative to Method 7. from fossil-fuel fired steam generators, elec- The within-laboratory relative standard de- tric utility plants, nitric acid plants, or viation for a single measurement is 2.8 and other sources as specified in the regulations. 2.9 percent at 201 and 268 ppm NO , respec- /m3, x The lower detectable limit is 13 mg NOx tively. as NO (7 ppm NO ) when sampling at 500 cc/ 2 x 1.5 Stability. Collected samples are stable min for 1 hour. No upper limit has been es- for at least 4 weeks. tablished; however, when using the rec- ommended sampling conditions, the method 2. Apparatus has been found to collect NOx emissions 2.1 Sampling and Sample Recovery. The 3 quantitatively up to 1,782 mg NOx/m , as NO2 sampling train is shown in Figure 7C–1, and (932 ppm NOx). component parts are discussed below. Alter- 1.2 Principle. An integrated gas sample is native apparatus and procedures are allowed extracted from the stack and collected in al- provided acceptable accuracy and precision kaline-potassium permanganate solution; can be demonstrated.
734
VerDate 11
2.1.1 Probe. Borosilicate glass tubing, suf- 2.1.2 Impingers. Three restricted-orifice ficiently heated to prevent water condensa- glass impingers, having the specifications tion and equipped with an in-stack or out- given in Figure 7C–2, are required for each stack filter to remove particulate matter (a sampling train. The impingers must be con- plug of glass wool is satisfactory for this nected in series with leak-free glass connec- purpose). Stainless steel or Teflon tubing tors. Stopcock grease may be used, if nec- may also be used for the probe. (Note: Men- essary, to prevent leakage. (The impingers tion of trade names or specific products does can be fabricated by a glass blower until not constitute endorsement by the U.S. En- they become available commercially.) vironmental Protection Agency.)
735
VerDate 11
2.1.3 Glass Wool, Stopcock Grease, Drying the sampling conditions of 400 to 500 cc/min Tube, Valve, Pump, Barometer, and Vacuum for 60 minutes within an accuracy of 2 per- Gauge and Rotameter. Same as in Method 6, cent. Sections 2.1.3, 2.1.4, 2.1.6, 2.1.7, 2.1.8, 2.1.11, 2.1.6 Filter. To remove NOx from ambient and 2.1.12, respectively. air, prepared by adding 20 g of a 5-angstrom 2.1.4 Rate Meter. Rotameter, or equiva- molecular sieve to a cylindrical tube, e.g., a lent, accurate to within 2 percent at the se- polyethylene drying tube. lected flow rate between 400 and 500 cc/min. 2.1.7 Polyethylene Bottles. 1-liter, for For rotameters, a range of 0 to 1 liter/min is sample recovery. recommended. 2.1.8 Funnel and Stirring Rods. For sam- 2.1.5 Volume Meter. Dry gas meter capa- ple recovery. ble of measuring the sample volume, under 2.2 Sample Preparation and Analysis.
736
VerDate 11
2.2.1 Hot Plate. Stirring type with 50- by w). Dissolve 40.0 g of KMnO4 and 20.0 g of 10-mm Teflon-coated stirring bars. NaOH in 940 ml of water. 2.2.2 Beakers. 400-, 600-, and 1000-ml capac- 3.2 Sample Preparation and Analysis. ities. 3.2.1 Water. Same as in Section 3.1.1. 2.2.3 Filtering Flask. 500-ml capacity with 3.2.2 Sulfuric Acid. Concentrated H2SO4. side arm. 3.2.3 Oxalic Acid Solution. Dissolve 48 g of 2.2.4 Buchner Funnel. 75-mm ID, with oxalic acid [(COOH)2•2H2O] in water, and di- spout equipped with a 13-mm ID by 90-mm lute to 500 ml. Do not heat the solution. long piece of Teflon tubing to minimize pos- 3.2.4 Sodium Hydroxide, 0.5 N. Dissolve 20 sibility of aspirating sample solution during g of NaOH in water, and dilute to 1 liter. filtration. 3.2.5 Sodium Hydroxide, 10 N. Dissolve 2.2.5 Filter Paper. Whatman GF/C, 7.0-cm diameter. 40 g of NaOH in water and dilute to 100 ml. 2.2.6 Stirring Rods. 3.2.6 Ethylenediamine Tetraacetic Acid 2.2.7 Volumetric Flasks. 100-, 200- or 250-, (EDTA) Solution, 6.5 Percent. Dissolve 6.5 g 500-, and 1000-ml capacity. of EDTA (disodium salt) in water, and dilute 2.2.8 Watch Glasses. To cover 600- and to 100 ml. Solution is best accomplished by 1,000-ml beakers. using a magnetic stirrer. 2.2.9 Graduated Cylinders. 50- and 250-ml 3.2.7 Column Rinse Solution. Add 20 ml of capacities. 6.5 percent EDTA solution to 960 ml of water, 2.2.10 Pipettes. Class A and adjust the pH to 11.7 to 12.0 with 0.5 N 2.2.11 pH Meter. To measure pH from 0.5 NaOH. to 12.0 3.2.8 Hydrochloric Acid (HCl), 2 N. Add 86 2.2.12 Burette. 50-ml with a micrometer ml of concentrated HCl to a 500-ml volu- type stopcock. (The stopcock is Catalogue metric flask containing water, dilute to vol- No. 8225–t–05, Ace Glass, Inc., Post Office Box ume, and mix well. Store in a glass-stoppered 996, Louisville, Kentucky 50201.) Place a bottle. glass wool plug in bottom of burette. Cut off 3.2.9 Sulfanilamide Solution. Add 20 g of burette at a height of 43 cm from the top of sulfanilamide (melting point 165 to 167 °C) to plug, and have a glass blower attach a glass 700 ml of water. Add, with mixing, 50 ml con- funnel to top of burette such that the diame- centrated phosphoric acid (85 percent), and ter of the burette remains essentially un- dilute to 1000 ml. This solution is stable for changed. Other means of attaching the fun- at least 1 month, if refrigerated. nel are acceptable. 3.2.10 N-(1-Naphthyl)-Ethylenediamine 2.2.13 Glass Funnel. 75-mm ID at the top. Dihydrochloride (NEDA) Solution. Dissolve 2.2.14 Spectrophotometer. Capable of 0.5 g of NEDA in 500 ml of water. An aqueous measuring absorbance at 540 nm. One-cm solution should have one absorption peak at cells are adequate. 320 nm over the range of 260 to 400 nm. 2.2.15 Metal Thermometers. Bimetallic NEDA, showing more than one absorption thermometers, range 0 to 1501⁄2 C. peak over this range, is impure and should 2.2.16 Culture Tubes. 20- by 150-mm, not be used. This solution is stable for at Kimax No. 45048. least 1 month if protected from light and re- 2.2.17 Parafilm ‘‘M.’’ Obtained from Amer- frigerated. ican Can Company, Greenwich, Connecticut 3.2.11 Cadmium. Obtained from Matheson 06830. Coleman and Bell, 2909 Highland Avenue, 2.2.18 CO2 Measurement Equipment. Same Norwood, Ohio 45212, as EM Laboratories as in Method 3. Catalogue No. 2001. Prepare by rinsing in 2 N 3. Reagents HCl for 5 minutes until the color is silver- Unless otherwise indicated, all reagents grey. Then rinse the cadmium with water should conform to the specifications estab- until the rinsings are neutral when tested lished by the Committee on Analytical Re- with pH paper. CAUTION: H2 is liberated agents of the American Chemical Society, during preparation. Prepare in an exhaust where such specifications are available; oth- hood away from any flame. erwise, use the best available grade. 3.2.12 NaNO2 Standard Solution, Nominal 3.1 Sampling. Concentration, 1000 µg NO2–/ml. Desiccate 3.1.1 Water. Deionized distilled to conform NaNO2 overnight. Accurately weigh 1.4 to 1.6 to ASTM Specification D 1193–74, Type 3 (in- g of NaNO2 (assay of 97 percent NaNO2 or corporated by reference—see § 60.17). greater), dissolve in water, and dilute to 1 3.1.2 Potassium Permanganate, 4.0 per- liter. Calculate the exact NO2– concentration cent (w/w), Sodium Hydroxide, 2.0 percent (w/ from the following relationship:
purity,%.46 01 µgNO−/ ml = g of NaNO × ×103 × 2 2 100 69. 01
737
VerDate 11
This solution is stable for at least 6 months itors, the minimum sampling time is 1 hour, under laboratory conditions. sampling 20 minutes at each traverse point. 3.2.13 KNO3 Standard Solution. Dry KNO3 [NOTE.— When the SO2 concentration is at 110 °C for 2 hours, and cool in a desiccator. greater than 1200 ppm, the sampling time Accurately weigh 9 to 10 g of KNO3 to within may have to be reduced to 30 minutes to 0.1 mg, dissolve in water, and dilute to 1 eliminate plugging of the impinger orifice liter. Calculate the exact NO3– concentration with MnO2. For RA tests with SO2 greater from the following relationship: than 1200 ppm, sample for 30 minutes (10 minutes at each point)]. Record the DGM µ − = ×3 × 62. 01 temperature, and check the flow rate at gNO− / ml gof KNO3 10 3 10110. least every 5 minutes. At the conclusion of This solution is stable for 2 months without each run, turn off the pump, remove probe preservative under laboratory conditions. from the stack, and record the final read- 3.2.14 Spiking Solution. Pipette 7 ml of ings. Divide the sample volume by the sam- pling time to determine the average flow the KNO3 standard into a 100-ml volumetric flask, and dilute to volume. rate. Conduct a leak-check as in Section 3.2.15 Blank Solution. Dissolve 2.4 g of 4.1.2. If a leak is found, void the test run, or use procedures acceptable to the Adminis- KMnO4 and 1.2 g of NaOH in 96 ml of water. trator to adjust the sample volume for the Alternatively, dilute 60 ml of KMnO4/NaOH solution to 100 ml. leakage. 3.2.16 Quality Assurance Audit Samples. 4.1.5 CO2 Measurement. During sampling, Same as in Method 7, Section 3.3.9. When re- measure the CO2 content of the stack gas questing audit samples, specify that they be near the sampling point using Method 3. The in the appropriate concentration range for single-point grab sampling procedure is ade- Method 7C. quate, provided the measurements are made at least three times—near the start, midway, 4. Procedure and before the end of a run and the average 4.1 Sampling. CO2 concentration is computed. The Orsat or 4.1.1 Preparation of Collection Train. Add Fyrite analyzer may be used for this analy- 200 ml of KMnO4/NaOH solution (3.1.2) to sis. each of three impingers, and assemble the 4.2 Sample Recovery. Disconnect the im- train as shown in Figure 7C–1. Adjust probe pingers. Pour the contents of the impingers heater to a temperature sufficient to prevent into a 1-liter polyethylene bottle using a fun- water condensation. nel and a stirring rod (or other means) to 4.1.2 Leak-Check Procedure. A leak-check prevent spillage. Complete the quantitative prior to the sampling run should be carried transfer by rinsing the impingers and con- out; a leak-check after the sampling run is necting tubes with water until the rinsings mandatory. Carry out the leak-check(s) ac- are clear to light pink, and add the rinsings cording to Method 6, Section 4.1.2. to the bottle. Mix the sample, and mark the 4.1.3 Check of Rotameter Calibration Ac- solution level. Seal and identify the sample curacy (Optional). Disconnect the probe from container. the first impinger, and connect the filter 4.3 Sample Preparation for Analysis. Pre- (2.1.6). Start the pump, and adjust the rotam- pare a cadmium reduction column as follows: eter to read between 400 and 500 cc/min. After Fill the burette (2.2.12) with water. Add the flow rate has stabilized, start measuring freshly prepared cadmium slowly with tap- the volume sampled, as recorded by the dry ping until no further settling occurs. The gas meter (DGM), and the sampling time. height of the cadmium column should be 39 Collect enough volume to measure accu- cm. When not in use, store the column under rately the flow rate, and calculate the flow rinse solution (3.2.7). (NOTE.— The column rate. This average flow rate must be less should not contain any bands of cadmium than 500 cc/min for the sample to be valid; fines. This may occur if regenerated column therefore, it is recommended that the flow is used and will greatly reduce the column rate be checked as above prior to each test. lifetime.) 4.1.4 Sample Collection. Record the initial Note the level of liquid in the sample con- DGM reading and barometric pressure. De- tainer, and determine whether any sample termine the sampling point or points accord- was lost during shipment. If a noticeable ing to the appropriate regulations, e.g., amount of leakage has occurred, the volume § 60.46(c) of 40 CFR Part 60. Position the tip lost can be determined from the difference of the probe at the sampling point, connect between initial and final solution levels, and the probe to the first impinger, and start the this value can then be used to correct the an- pump. Adjust the sample flow to a value be- alytical result. Quantitatively transfer the tween 400 and 500 cc/min. CAUTION: HIGHER contents to a 1-liter volumetric flask, and di- FLOW RATES WILL PRODUCE LOW RE- lute to volume. SULTS. Once adjusted, maintain a constant Take a 100-ml aliquot of the sample and flow rate during the entire sampling run. blank (unexposed KMnO4/NaOH) solutions, Sample for 60 minutes. For relative accuracy and transfer to 400-ml beakers containing (RA) testing of continuous emission mon- magnetic stirring bars. Using a pH meter,
738
VerDate 11
add concentrated H2SO4 with stirring until a pare a new column, and repeat the cadmium pH of 0.7 is obtained. Allow the solutions to reduction]. stand for 15 minutes. Cover the beakers with 4.4 Sample Analysis. Pipette 10 ml of sam- watch glasses, and bring the temperature of ple into a culture tube. (NOTE.— Some test ° the solutions to 50 C. Keep the temperature tubes give a high blank NO2– value but cul- below 60 °C. Dissolve 4.8 g of oxalic acid in a ture tubes do not.) Pipette in 10 ml of sulfa- minimum volume of water, approximately 50 nilamide solution and 1.4 ml of NEDA solu- ml, at room temperature. Do not heat the so- tion. Cover the culture tube with parafilm, lution. Add this solution slowly, in incre- and mix the solution. Prepare a blank in the ments, until the KMnO4 solution becomes same manner using the sample from treat- colorless. If the color is not completely re- ment of the unexposed KMnO4/NaOH solution moved, prepare some more of the above ox- (3.1.2). Also, prepare a calibration standard alic acid solution, and add until a colorless to check the slope of the calibration curve. solution is obtained. Add an excess of oxalic After a 10-minute color development inter- acid by dissolving 1.6 g of oxalic acid in 50 ml val, measure the absorbance at 540 nm of water, and add 6 ml of this solution to the against water. Read µg NO2–/ml from the colorless solution. If suspended matter is calibration curve. If the absorbance is great- present, add concentrated H2SO4 until a clear er than that of the highest calibration stand- solution is obtained. ard, pipette less than 10 ml of sample and Allow the samples to cool to near room enough water to make the total sample vol- temperature, being sure that the samples are ume 10 ml, and repeat the analysis. Deter- still clear. Adjust the pH to 11.7 to 12.0 with mine the NO2 concentration using the cali- 10 N NaOH. Quantitatively transfer the mix- bration curve obtained in Section 5.3. ture to a Buchner funnel containing GF/C fil- 4.5 Audit Analysis. This is the same as in ter paper, and filter the precipitate. Filter Method 7, Section 4.4. the mixture into a 500-ml filtering flask. Wash the solid material four times with 5. Calibration water. When filtration is complete, wash the 5.1 Dry Gas Metering System (DGM). Teflon tubing, quantitatively transfer the 5.1.1 Initial Calibration. Same as in Method filtrate to a 500-ml volumetric flask, and di- 6, Section 5.1.1. For detailed instructions on lute to volume. The samples are now ready carrying out this calibration, it is suggested for cadmium reduction. Pipette a 50-ml ali- that Section 3.5.2 of Citation 4 in the quot of the sample into a 150-ml beaker, and Bibiography be consulted. add a magnetic stirring bar. Pipette in 1.0 ml 5.1.2 Post-Test Calibration Check. Same of 6.5 percent EDTA solution, and mix. as in Method 6, Section 5.1.2. Determine the correct stopcock setting to 5.2 Thermometers for DGM and Barom- establish a flow rate of 7 to 9 ml/min of col- eter. Same as in Method 6, Sections 5.2 and umn rinse solution through the cadmium re- 5.4, respectively. duction column. Use a 50-ml graduated cyl- 5.3 Calibration Curve for Spectrophoto- inder to collect and measure the solution meter. Dilute 5.0 ml of the NaNO2 standard volume. After the last of the rinse solution solution to 200 ml with water. This solution has passed from the funnel into the burette, nominally contains 25 µg NO2–/ml. Use this but before air entrapment can occur, start solution to prepare calibration standards to adding the sample, and collect it in a 250-ml cover the range of 0.25 to 3.00 µg NO2–/ml. graduated cylinder. Complete the quan- Prepare a minimum of three standards each titative transfer of the sample to the column for the linear and slightly nonlinear (de- as the sample passes through the column. scribed below) range of the curve. Use pi- After the last of the sample has passed from pettes for all additions. the funnel into the burette, start adding 60 Run standards and a water blank as in- ml of column rinse solution, and collect the structed in Section 4.4. Plot the net absorb- rinse solution until the solution just dis- ance vs µgNO2–/ml. Draw a smooth curve appears from the funnel. Quantitatively through the points. The curve should be lin- transfer the sample to a 200-ml volumetric ear up to an absorbance of approximately 1.2 flask (250-ml may be required), and dilute to with a slope of approximately 0.53 absorb- volume. The samples are now ready for NO2– ance units/µg NO2–/ml. The curve should pass analysis. [NOTE.— Both the sample and blank through the origin. The curve is slightly should go through this procedure. Addition- nonlinear from an absorbance of 1.2 to 1.6. ally, two spiked samples should be run with every group of samples passed through the 6. Calculations column. To do this, prepare two additional Carry out calculations, retaining at least 50–ml aliquots of the sample suspected to one extra decimal figure beyond that of the have the highest NO3– concentration, and acquired data. Round off figures after final add 1 ml of the spiking solution to these calculation. aliquots. If the spike recovery or column ef- 6.1 Sample volume, dry basis, corrected to ficiency (see 6.2.1) is below 95 percent, pre- standard conditions.
739
VerDate 11
T P VP =std bar = m bar Vm() std V m XY K1 XY Eq.7 C- 1 Tm Pstd Tm
Where: Pbar=Barometric pressure, mm Hg.
Vm(std)=Dry gas volume measured by the dry Pstd=Standard absolute pressure, 760 mm Hg. gas meter, corrected to standard condi- Tm=Average dry gas meter absolute tempera- tions, dscm. ture, °K.
Vm=Dry gas volume as measured by the dry Tstd=Standard absolute temperature, 293 °K. gas meter, dcm. K1=0.3858 °K/mm Hg. Y=Dry gas meter calibration factor. 6.2 Total µg NO Per Sample. X=Correction factor for CO collection. 2 2 6.2.1 Efficiency of Cadmium Reduction Column. Calculate this value as follows: = 100 − 100 %(/)CO2 v v
Where: s=Concentration of spiking solution, µg E=Column efficiency, unitless. NO3¥/ml. x=Analysis of spiked sample, µg NO2–/ml. 1.0=Volume of spiking solution added, ml. µ µ y=Analysis of unspiked sample, µg NO2–/ml. 46.01= g NO2–/ mole. 200=Final volume of sample and blank after 62.01=µg NO3–/µ mole. passing through the column, ml. 6.2.2 Total µg NO2.
()()()SB− 500 1000 2× 104 SB − m = ×200 × × = Eq.7 C- 3 E 50 100 E
Where: 50=Aliquot of prepared sample processed
m=Mass of NOx, as NO2, in sample, µg. through cadmium column, ml. 100=Aliquot of KMnO /NaOH solution, ml. S=Analysis of sample, µg NO2–/ml. 4 1000=Total volume of KMnO4/NaOH solution B=Analysis of blank, µg NO2–/ml. 500=Total volume of prepared sample, ml. ml. 6.3 Sample Concentration.
= m CK2 Vm() std
Where: 1 ft3=2.832×10¥2 m3.
C=Concentration of NOx as NO2, dry basis, 1000 mg = 1 g. mg/dscm. 7. Quality Control K =10¥3 mg/µg. 2 Quality control procedures are specified in 6.4 Conversion Factors. Sections 4.1.3 (flow rate accuracy); 4.3 (cad- 3 1.0 ppm NO=1.247 mg NO/m at STP. mium column efficiency); 4.4 (calibration 3 1.0 ppm NO2=1.912 mg NO2/m at STP. 740
VerDate 11
curve accuracy); and 4.5 (audit analysis accu- percent if all the SO2 is consumed in the first racy). impinger. 8. Bibliography NH3 is slowly oxidized to NO3– by the ab- sorbing solution. At 100 ppm NH in the gas 1. Margeson, J.H., W.J. Mitchell, J.C. 3 stream, an interference of 6 ppm NO (11 mg Suggs, and M.R. Midgett. Integrated Sam- x NO /m3) was observed when the sample was pling and Analysis Methods for Determining 2 analyzed 10 days after collection. Therefore, NO Emissions at Electric Utility Plants. x the method may not be applicable to plants U.S. Environmental Protection Agency, Re- using NH injection to control NO emissions search Triangle Park, NC. Journal of the Air 3 x unless means are taken to correct the re- Pollution Control Association. 32:1210–1215. sults. An equation has been developed to 1982. 2. Memorandum and attachment from J.H. allow quantitation of the interference and is Margeson, Source Branch, Quality Assurance discussed in Citation 4 of the Bibliography. Division, Environmental Monitoring Sys- 1.4 Precision and Bias. The method does tems Laboratory, to The Record, EPA. not exhibit any bias relative to Method 7. The within-laboratory relative standard de- March 30, 1983. NH3 Interference in Methods 7C and 7D. viation for a single measurement was ap- 3. Margeson, J.H., J.C. Suggs, and M.R. proximately 6 percent at 200 to 270 ppm NOx. Midgett. Reduction of Nitrate to Nitrite 1.5 Stability. Collected samples are stable with Cadmium. Anal. Chem. 52:1955–57. 1980. for at least 4 weeks. 4. Quality Assurance Handbook for Air Pol- 2. Apparatus lution Measurement Systems. Volume III— 2.1 Sampling and Sample Recovery. The Stationary Source Specific Methods. August sampling train is the same as in Figure 7C– 1977. U.S. Environmental Protection Agency. 1 of Method 7C. Component parts are the Research Triangle Park, NC. Publication No. same as in Method 7C, Section 2.1. EPA–600/4–77–027b. August 1977. 2.2 Sample Preparation and Analysis. 5. Margeson, J.H., et al. An Integrated 2.2.1 Magnetic Stirrer. With 25- by 10-mm Method for Determining NOx Emissions at Teflon-coated stirring bars. Nitric Acid Plants. Manuscript submitted to 2.2.2 Filtering Flask. 500-ml capacity with Analytical Chemistry. April 1984. sidearm. 2.2.3 Buchner Funnel. 75-mm ID. The METHOD 7D—DETERMINATION OF NITROGEN spout equipped with a 13-mm ID by 90-mm OXIDE EMISSIONS FROM STATIONARY long piece of Teflon tubing to minimize pos- SOURCES—ALKALINE-PERMANGANATE/ION sibility of aspirating sample solution during CHROMATOGRAPHIC METHOD filtration. 1. Applicability, Principle, Interferences, Preci- 2.2.4 Filter Paper. Whatman GF/C, 7.0-cm sion, Bias, and Stability diameter. 1.1 Applicability. The method is applica- 2.2.5 Stirring Rods. 2.2.6 Volumetric Flask. 250-ml. ble to the determination of NOx emissions from fossil-fuel fired steam generators, elec- 2.2.7 Pipettes. Class A. tric utility plants, nitric acid plants, or 2.2.8 Erlenmeyer Flasks. 250-ml. other sources as specified in the regulations. 2.2.9 Ion Chromatograph. Equipped with The lower detectable limit is similar to that ananionseparatorcolumntoseparate NO3– ∂ for Method 7C. No upper limit has been es- , a H suppressor, and necessary auxiliary tablished; however, when using the rec- equipment. Nonsuppressed and other forms ommended sampling conditions, the method of ion chromatography may also be used pro- vided that adequate resolution of NO3– is ob- has been found to collect NOx emissions 3 tained. The system must also be able to re- quantitatively up to 1782 mg NOx/m , as NO2 solve and detect NO2–. (932 pm NOx). 1.2 Principle. An integrated gas sample is 3. Reagents extracted from the stack and collected in al- Unless otherwise indicated, all reagents kaline-potassium permanganate solution; should conform to the specifications estab- NOx (NO+NO2) emissions are oxidized to NO3– lished by the Committee on Analytical Re- . Then NO3– is analyzed by ion chroma- agents of the American Chemical Society, tography. where such specifications are available; oth- 1.3 Interferences. Possible interferences erwise, use the best available grade. are SO2 and NH3. High concentrations of SO2 3.1 Sampling. could interfere because SO2 consumes MnO4– 3.1.1 Water. Deionized distilled to conform (as does NOx) and, therefore, could reduce the to ASTM Specification D 1193–74, Type 3 (in- NOx collection efficiency. However, when corporated by reference—see § 60.17). sampling emissions from a coal-fired electric 3.1.2 Potassium Permanganate, 4.0 Per- utility plant burning 2.1-percent sulfur coal cent (w/w), Sodium Hydroxide, 2.0 Percent with no control of SO2 emissions, collection (w/w). Dissolve 40.0 g of KMnO4 and 20.0 g of efficiency was not reduced. In fact, calcula- NaOH in 940 ml of water. tions show that sampling 3000 ppm SO2 will 3.2 Sample Preparation and Analysis. reduce the MnO4– concentration by only 5 3.2.1 Water. Same as in Section 3.1.1. 741
VerDate 11
3.2.2 Hydrogen Peroxide, 5 Percent. Dilute 3.2.4 KNO3 Standard Solution. Dry 30 percent H2O2 1:5 (v/v) with water. KNO3 at 110 ° C for 2 hours, and cool in a des- 3.2.3 Blank Solution. Dissolve 2.4 g of iccator. Accurately weigh 9 to 10 g of KNO to within 0.1 mg, dissolve in water, and KMnO4 and 1.2 g of NaOH in 96 ml of water. 3 dilute to 1 liter. Calculate the exact NO – Alternatively, dilute 60 ml of KMnO4/NaOH 3 solution to 100 ml. concentration from the following relation- ship:
µ − = ×3 × 62. 01 gNO3/ ml g of KNO 3− 10 10110.
This solution is stable for 2 months without the stirring rate to as fast a rate as possible preservative under laboratory conditions. without loss of solution. Add 5 percent 3.2.5 Eluent, 0.003 M NaHCO3/0.0024 M H2O2 in increments of approximately 5 ml Na2CO3. Dissolve 1.008 g NaHCO3 and 1.018 g using a 5-ml pipette. When the KMnO4 color Na2CO3 in water, and dilute to 4 liters. Other appears to have been removed, allow the pre- eluents capable of resolving nitrate ion from cipitate to settle, and examine the super- sulfate and other species present may be natant liquid. If the liquid is clear, the used. H2O2 addition is complete. If the KMnO4 color 3.2.6 Quality Assurance Audit Samples. persists, add more H O , with stirring, until This is the same as in Method 7, Section 2 2 the supernatant liquid is clear. (NOTE: The 3.3.9. When requesting audit samples, specify faster the stirring rate, the less volume of that they be in the appropriate concentra- H O that will be required to remove the tion range for Method 7D. 2 2 KMnO4.) Quantitatively transfer the mixture 4. Procedure to a Buchner funnel containing GF/C filter 4.1 Sampling. This is the same as in Meth- paper, and filter the precipitate. The spout of od 7C, Section 4.1. the Buchner funnel should be equipped with 4.2 Sample Recovery. This is the same as a 13-mm ID by 90-mm long piece of Teflon in Method 7C, Section 4.2. tubing. This modification minimizes the pos- 4.3 Sample Preparation for Analysis. Note sibility of aspirating sample solution during the level of liquid in the sample container, filtration. Filter the mixture into a 500-ml and determine whether any sample was lost filtering flask. Wash the solid material four during shipment. If a noticeable amount of times with water. When filtration is com- leakage has occurred, the volume lost can be plete, wash the Teflon tubing, quantitatively determined from the difference between ini- transfer the filtrate to a 250-ml volumetric tial and final solution levels, and this value can then be used to correct the analytical re- flask, and dilute to volume. The sample and sult. Quantitatively transfer the contents to blank are now ready for NO3¥ analysis. a 1-liter volumetric flask, and dilute to vol- 4.4 Sample Analysis. The following ume. chromatographic conditions are rec- Sample preparation can be started 36 hours ommended: 0.003 M NaHCO3/0.0024 M Na2CO3 after collection. This time is necessary to eluent solution. (3.2.5), full scale range 3 µ ensure that all NO2– is converted to NO3– MHO; sample loop, 0.5 ml; flow rate, 2.5 ml/ Take a 50–ml aliquot of the sample and min. These conditions should give a NO3¥ blank, and transfer to 250-ml Erlenmeyer retention time of approximately 15 minutes flasks. Add a magnetic stirring bar. Adjust (Figure 7D–1).
742
VerDate 11
Establish a stable baseline. Inject a sample 5.1.1 Initial Calibration. Same as in Meth-
of water, and determine if any NO3¥ appears od 6, Section 5.1.1. For detailed instructions in the chromatogram. If NO3¥ is present, re- on carrying out this calibration, it is sug- peat the water load/injection procedure ap- gested that Section 3.5.2 of Citation 3 in the proximately five times; then re-inject a Bibliography be consulted. water sample, and observe the chromato- 5.1.2 Post-Test Calibration Check. Same as in Method 6, Section 5.1.2. gram. When no NO3¥ is present, the instru- ment is ready for use. Inject calibration 5.2 Thermometers for DGM and Barom- standards. Then inject samples and a blank. eter. Same as in Method 6, Section 5.2 and 5.4, respectively. Repeat the injection of the calibration 5.3 Calibration Curve for Ion Chromato- standards (to compensate for any drift in re- graph. Dilute a given volume (1.0 ml or sponse of the instrument). Measure the greater) of the KNO3 standard solution to a NO3¥ peak height or peak area, and deter- convenient volume with water, and use this mine the sample concentration from the solution to prepare calibration standards. calibration curve. Prepare at least four standards to cover the 4.5 Audit analysis. This is the same as in range of the samples being analyzed. Use pi- Method 7, Section 4.4. pettes for all additions. Run standards as in- 5. Calibration structed in Section 4.4. Determine peak height or area, and plot the individual values 5.1 Dry Gas Metering System (DGM). versus concentration in µgNO3–/ml. Do not 743
VerDate 11
force the curve through zero. Draw a smooth Carry out calculations, retaining at least curve through the points. The curve should one extra decimal figure beyond that of the be linear. With the linear curve, use linear acquired data. Round off figures after final regression to determine the calibration calculation. equation. 6.1 Sample Volume, Dry Basis, Corrected to Standard Conditions. Same as in Method 6. Calculations 7C, Section 6.1. 6.2 Total µg NO2 Per Sample.
1000 46. 01 m=() S − B ×250 × × =3710(). S − B Eq 7 D- 1 50 62. 01
Where: Triangle Park, NC. Publication No. EPA–600/ 4–77–027b. August 1977. m=Mass of NOx, as NO2, in sample, µg. S=Analysis of sample, µg NO3–/ml. 4. Margeson, J.H., et al. An Integrated B=Analysis of blank, µg NO3–/ml. Method for determining NOx Emissions at 250=Volume of prepared sample, ml. Nitric Acid Plants. Manuscript submitted to 46.01=Molecular weight of NO2–. Analytical Chemistry. April 1984. 62.01=Molecular weight of NO3–. 1000=Total volume of KMnO4 solution, ml. METHOD 7E—DETERMINATION OF NITROGEN 50=Aliquot KMnO4/NaOH solution, ml. OXIDES EMISSIONS FROM STATIONARY 6.3 Sample Concentration. SOURCES (INSTRUMENTAL ANALYZER PROCE- DURE) = m 1. Applicability and Principle CK2 Vm() std 1.1 Applicability. This method is applica- ble to the determination of nitrogen oxides Where: (NOx) concentrations in emissions from sta- C=Concentration of NOx as NO2, dry basis, tionary sources only when specified within mg/dscm. the regulations. K =10¥3 mg/µg. 2 1.2 Principle. A gas simple is continu- Vm(std)=Dry gas volume measured by the dry gas meter, corrected to standard condi- ously extracted from a stack, and a portion tions, dscm. of the sample is conveyed to an instrumental 6.4 Conversion Factors. chemiluminescent analyzer for determina- 1.0 ppm NO=1.247 mg NO/m3 at STP. tion of NOx concentration. Performance spec- 3 ifications and test procedures are provided to 1.0 ppm NO2=1.912 mg NO2/m at STP. 1 ft3=2.832×10¥2 m3. ensure reliable data. 1000 mg = 1 g. 2. Range and Sensitivity 7. Quality Control Same as Method 6C, Sections 2.1 and 2.2. Quality control procedures are specified in 3. Definitions Sections 4.1.3 (flow rate accuracy) and 4.5 (audit analysis accuracy) of Method 7C. 3.1 Measurement System. The total equip- ment required for the determination of NOx 8. Bibliography concentration. The measurement system 1. Margeson, J.H., W.J. Mitchell, J.C. consists of the following major subsystems: Suggs, and M.R. Midgett. Integrated Sam- 3.1.1 Sample Interface, Gas Analyzer, and pling and Analysis Methods for Determining Data Recorder. Same as Method 6C, Sections NOx Emissions at Electric Utility Plants. 3.1.1, 3.1.2, and 3.1.3. U.S. Environmental Protection Agency, Re- 3.1.2 NO to NO Converter. A device that search Triangle Park, NC. Journal of the Air 2 converts the nitrogen dioxide (NO2) in the Pollution Control Association. 32:1210–1215. sample gas to nitrogen oxide (NO). 1982. 3.2 Span, Calibration Gas, Analyzer Cali- 2. Memorandum and attachment from J.H. bration Error, Sampling System Bias, Zero Margeson, Source Branch, Quality Assurance Drift, Calibration Drift, and Response Time. Division, Environmental Monitoring Sys- Same as Method 6C, Sections 3.2 through 3.8. tems Laboratory, to The Record, EPA. 3.3 Interference Response. The output re- March 30, 1983. NH3 Interference in Methods 7C and 7D. sponse of the measurement system to a com- 3. Quality Assurance Handbook for Air Pol- ponent in the sample gas, other than the gas lution Measurement Systems. Volume III— component being measured. Stationary Source Specific Methods. U.S. 4. Measurement System Performance Specifica- Environmental Protection Agency, Research tions
744
VerDate 11
Same as Method 6C, Sections 4.1 through the NO2 concentration within the sample 4.4. stream is not greater than 5 percent of the 5. Apparatus and Reagents NOx concentration, conduct an NO2 to NO 5.1 Measurement System. Any measure- conversion efficiency test in accordance with Section 5.6 of Method 20. ment system for NOx that meets the speci- fications of this method. A schematic of an 7. Emission Test Procedure acceptable measurement system is shown in 7.1 Selection of Sampling Site and Sam- Figure 6C–1 of Method 6C. The essential com- pling Points. Select a measurement site and ponents of the measurement system are de- sampling points using the same criteria that scribed below: are applicable to tests performed using 5.1.1 Sample Probe, Sample Line, Calibra- Method 7. tion Valve Assembly, Moisture Removal Sys- 7.2 Sample Collection. Position the sam- tem, Particulate Filter, Sample Pump, Sam- pling probe at the first measurement point, ple Flow Rate Control, Sample Gas Manifold, and begin sampling at the same rate as used and Data Recorder. Same as Method 6C, Sec- during the system calibration drift test. tions 5.1.1 through 5.1.9, and 5.1.11. Maintain constant rate sampling (i.e., ±10 5.1.2 NO2 to NO Converter. That portion of percent) during the entire run. The sampling the system that converts the nitrogen diox- time per run shall be the same as the total ide (NO2) in the sample gas to nitrogen oxide time required to perform a run using Method (NO). An NO2 to NO converter is not nec- 7, plus twice the system response time. For essary if data are presented to demonstrate each run, use only those measurements ob- that the NO2 portion of the exhaust gas is tained after twice the response time of the less than 5 percent of the total NO con- x measurement system has elapsed, to deter- centration. mine the average effluent concentration. 5.1.3 NO Analyzer. An analyzer based on x 7.3 Zero and Calibration Drift Test. Follow the principles of chemiluminescence, to de- Section 7.4 of Method 6C. termine continuously the NOx concentration in the sample gas stream. The analyzer shall 8. Emission Calculation meet the applicable performance specifica- Follow Section 8 of Method 6C. tions of Section 4. A means of controlling 9. Bibliography the analyzer flow rate and a device for deter- mining proper sample flow rate (e.g., preci- Same as bibliography of Method 6C. sion rotameter, pressure gauge downstream METHOD 8—DETERMINATION OF SULFURIC ACID of all flow controls, etc.) shall be provided at MIST AND SULFUR DIOXIDE EMISSIONS FROM the analyzer. STATIONARY SOURCES 5.2 NOx Calibration Gases. The calibration gases for the NOx analyzer shall be NO in N2. 1. Principle and Applicability Three calibration gases, as specified in Sec- 1.1 Principle. A gas sample is extracted tions 5.3.1 through 5.3.3. of Method 6C, shall isokinetically from the stack. The sulfuric be used. Ambient air may be used for the acid mist (including sulfur trioxide) and the zero gas. sulfur dioxide are separated, and both frac- 6. Measurement System Performance Test Proce- tions are measured separately by the bar- dures ium-thorin titration method. Perform the following procedures before 1.2 Applicability. This method is applica- measurement of emissions (Section 7). ble for the determination of sulfuric acid 6.1 Calibration Gas Concentration Ver- mist (including sulfur trioxide, and in the ification. Follow Section 6.1 of Method 6C, absence of other particulate matter) and sul- except if calibration gas analysis is required, fur dioxide emissions from stationary use Method 7, and change all 5 percent per- sources. Collaborative tests have shown that formance values to 10 percent (or 10 ppm, the minimum detectable limits of the meth- whichever is greater). od are 0.05 milligrams/cubic meter (0.03>10¥7 6.2 Interference Response. Conduct an in- pounds/cubic foot) for sulfur trioxide and 1.2 terference response test of the analyzer prior mg/m3 (0.74 10¥7 lb/ft3) for sulfur dioxide. No to its initial use in the field. Thereafter, re- upper limits have been established. Based on check the measurement system if changes theoretical calculations for 200 milliters of 3 are made in the instrumentation that could percent hydrogen peroxide solution, the alter the interference response (e.g., changes upper concentration limit for sulfur dioxide in the gas detector). Conduct the inter- in a 1.0 m3 (35.3 ft3) gas sample is about 12,500 ference response in accordance with Section mg/m3 (7.7×10¥4 lb/ft3). The upper limit can 5.4 of Method 20. be extended by increasing the quantity of 6.3 Measurement System Preparation, An- peroxide solution in the impingers. alyzer Calibration Error, and Sample System Possible interfering agents of this method Bias Check. Follow Sections 6.2 through 6.4 are fluorides, free ammonia, and dimethyl of Method 6C. aniline. If any of these interfering agents are 6.4 NO2 to NO Conversion Efficiency. Un- present (this can be determined by knowl- less data are presented to demonstrate that edge of the process), alternative methods,
745
VerDate 11
subject to the approval of the Administrator, modifications to Figure 8–1 are discussed in U. S. E. P.A., are required. the following subsections. Filterable particulate matter may be de- The operating and maintenance procedures termined along with SO3 and SO2 (subject to for the sampling train are described in the approval of the Administrator) by insert- APTD–0576. Since correct usage is important ing a heated glass fiber filter between the in obtaining valid results, all users should probe and isopropanol impinger (see Section read the APTD–0576 document and adopt the 2.1 of Method 6.) If this option is chosen, par- operating and maintenance procedures out- ticulate analysis is gravimetric only; H2SO4 lined in it, unless otherwise specified herein. acid mist is not determined separately. Further details and guidelines on operation 2. Apparatus and maintenance are given in Method 5 and 2.1 Sampling. A schematic of the sam- should be read and followed whenever they pling train used in this method is shown in are applicable. Figure 8–1; it is similiar to the Method 5 2.1.1 Probe Nozzle. Same as Method 5, train except that the filter position is dif- Section 2.1.1. ferent and the filter holder does not have to 2.1.2 Probe Liner. Borosilicate or quartz be heated. Commercial models of this train glass, with a heating system to prevent visi- are available. For those who desire to build ble condensation during sampling. Do not their own, however, complete construction use metal probe liners. details are described in APTD–0581. Changes 2.1.3 Pitot Tube. Same as Method 5, Sec- from the APTD–0581 document and allowable tion 2.1.3.
746
VerDate 11
2.1.4 Differential Pressure Gauge. Same as Greenburg-Smith design with standard tips. Method 5, Section 2.1.4. The second and fourth shall be of the 2.1.5 Filter Holder. Borosilicate glass, Greenburg-Smith design, modified by replac- with a glass frit filter support and a silicone ing the insert with an approximately 13 mil- rubber gasket. Other gasket materials, e.g., limeter (0.5 in.) ID glass tube, having an Teflon or Viton, may be used subject to the unconstricted tip located 13 mm (0.5 in.) approval of the Administrator. The holder from the bottom of the flask. Similar collec- design shall provide a positive seal against tion systems, which have been approved by leakage from the outside or around the fil- the Administrator, may be used. ter. The filter holder shall be placed between 2.1.7 Metering System. Same as Method 5, the first and second impingers. Note: Do not Section 2.1.8. heat the filter holder. 2.1.8 Barometer. Same as Method 5, Sec- 2.1.6 Impingers. Four, as shown in Figure tion 2.1.9. 8–1. The first and third shall be of the
747
VerDate 11
2.1.9 Gas Density Determination Equip- activated alumina. However, reagent grade ment. Same as Method 5, Section 2.1.10. isopropanol with suitably low peroxide levels 2.1.10 Temperature Gauge. Thermometer, is readily available from commercial or equivalent, to measure the temperature of sources; therefore, rejection of contaminated the gas leaving the impinger train to within lots may be more efficient than following 1°C (2°F). the peroxide removal procedure. 2.2 Sample Recovery. 3.1.5 Hydrogen Peroxide, 3 Percent. Dilute 2.2.1 Wash Bottles. Polyethylene or glass, 100 ml of 30 percent hydrogen peroxide to 1 500 ml. (two). liter with deionized, distilled water. Prepare 2.2.2 Graduated Cylinders. 250 ml, 1 liter. fresh daily. (Volumetric flasks may also be used. 3.1.6 Crushed Ice. 2.2.3 Storage Bottles. Leak-free polyeth- 3.2 Sample Recovery. lene bottles, 1000 ml size (two for each sam- 3.2.1 Water. Same as 3.1.3. pling run). 3.2.2 Isopropanol, 80 Percent. Same as 2.2.4 Trip Balance. 500-gram capacity, to 3.1.4. measure to ±0.5 g (necessary only if a mois- 3.3 Analysis. ture content analysis is to be done). 3.3.1 Water. Same as 3.1.3. 2.3 Analysis. 3.3.2 Isopropanol, 100 Percent. 2.3.1 Pipettes. Volumetric 25 ml, 100 ml. 3.3.3 Thorin Indicator. 1-(o-arsonophenyl- 2.3.2 Burette, 50 ml. azo) 2-naphthol-3, 6-disulfonic acid, disodium 2.3.3 Erlenmeyer Flask. 250 ml. (one for salt, or equivalent. Dissolve 0.20 g in 100 ml each sample, blank, and standard). of deionized, distilled water. 2.3.4 Graduated Cylinder. 100 ml. 3.3.4 Barium Perchlorate (0.0100 Normal). 2.3.5 Trip Balance. 500 g capacity, to Dissolve 1.95 g of barium perchlorate tri- measure to ±0.5 g. hydrate (Ba(C104)2 · 3H2O) in 200 ml deionized, 2.3.6 Dropping Bottle. To add indicator so- distilled water, and dilute to 1 liter with lution, 125-ml size. isopropanol; 1.22 g of barium chloride dihy- 3. Reagents drate (BaC12 · 2H2O) may be used instead of Unless otherwise indicated, all reagents the barium perchlorate. Standardize with are to conform to the specifications estab- sulfuric acid as in Section 5.2. This solution lished by the Committee on Analytical Re- must be protected against evaporation at all agents of the American Chemical Society, times. where such specifications are available. Oth- 3.3.5 Sulfuric Acid Standard (0.0100 N). ± erwise, use the best available grade. Purchase or standardize to 0.0002 N against 3.1 Sampling. 0.0100 N NaOH that has previously been 3.1.1 Filters. Same as Method 5, Section standardized against primary standard po- 3.1.1. tassium acid phthalate. 3.1.2 Silica Gel. Same as Method 5, Sec- 3.3.6 Quality Assurance Audit Samples. tion 3.1.2. Same as in Method 6, Section 3.3.6. 3.1.3 Water. Deionized, distilled to con- 4. Procedure form to ASTM Specification D1193–77, Type 3 4.1 Sampling. (incorporated by reference—see § 60.17). At 4.1.1 Pretest Preparation. Follow the pro- the option of the analyst, the KMnO4 test for cedure outlined in Method 5, Section 4.1.1; oxidizable organic matter may be omitted filters should be inspected, but need not be when high concentrations of organic matter desiccated, weighed, or identified. If the ef- are not expected to be present. fluent gas can be considered dry, i.e., mois- 3.1.4 Isopropanol. 80 Percent. Mix 800 ml ture free, the silica gel need not be weighed. of isopropanol with 200 ml of deionized, dis- 4.1.2 Preliminary Determinations. Follow tilled water. the procedure outlined in Method 5, Section NOTE: Experience has shown that only 4.1.2. A.C.S. grade isopropanol is satisfactory. 4.1.3 Preparation of Collection Train. Fol- Tests have shown that isopropanol obtained low the procedure outlined in Method 5, Sec- from commercial sources occasionally has tion 4.1.3 (except for the second paragraph peroxide impurities that will cause erro- and other obviously inapplicable parts) and neously high sulfuric acid mist measure- use Figure 8–1 instead of Figure 5–1. Replace ment. Use the following test for detecting the second paragraph with: Place 100 ml of 80 peroxides in each lot of isopropanol: Shake percent isopropanol in the first impinger, 100 10 ml of the isopropanol with 10 ml of freshly ml of 3 percent hydrogen peroxide in both prepared 10 percent potassium iodide solu- the second and third impingers; retain a por- tion. Prepare a blank by similarly treating tion of each reagent for use as a blank solu- 10 ml of distilled water. After 1 minute, read tion. Place about 200 g of silica gel in the the absorbance on a spectrophotometer at fourth impinger. 352 nanometers. If the absorbance exceeds NOTE: If moisture content is to be deter- 0.1, the isopropanol shall not be used. Perox- mined by impinger analysis, weigh each of ides may be removed from isopropanol by re- the first three impingers (plus absorbing so- distilling, or by passage through a column of lution) to the nearest 0.5 g and record these
748
VerDate 11
weights. The weight of the silica gel (or sili- the test, observe the connecting line between ca gel plus container) must also be deter- the probe and first impinger for signs of con- mined to the nearest 0.5 g and recorded. densation. If it does occur, adjust the probe 4.1.4 Pretest Leak-Check Procedure. Fol- heater setting upward to the minimum tem- low the basic procedure outlined in Method perature required to prevent condensation. If 5, Section 4.1.4.1, noting that the probe heat- component changes become necessary during er shall be adjusted to the minimum tem- a run, a leak-check shall be done imme- perature required to prevent condensation, diately before each change, according to the and also that verbage such as, ‘‘. . . plugging procedure outlined in Section 4.1.4.2 of Meth- the inlet to the filter holder . . .,’’ shall be od 5 (with appropriate modifications, as replaced by, ‘‘. . . plugging the inlet to the mentioned in Section 4.1.4 of this method); first impinger . . .’’ The pretest leak-check record all leak rates. If the leakage rate(s) is optional. exceed the specified rate, the tester shall ei- 4.1.5 Train Operation. Follow the basic ther void the run or shall plan to correct the procedures outlined in Method 5, Section sample volume as outlined in Section 6.3 of 4.1.5, in conjunction with the following spe- Method 5. Immediately after component cial instructions. Data shall be recorded on a changes, leak-checks are optional. If these sheet similar to the one in Figure 8–2. The leak-checks are done, the procedure outlined sampling rate shall not exceed 0.030 m3/min in Section 4.1.4.1 of Method 5 (with appro- (1.0 cfm) during the run. Periodically during priate modifications) shall be used.
749
VerDate 11
750
VerDate 11
After turning off the pump and recording 4.3.1 Container No. 1. Shake the container the final readings at the conclusion of each holding the isopropanol solution and the fil- run, remove the probe from the stack. Con- ter. If the filter breaks up, allow the frag- duct a post-test (mandatory) leak-check as ments to settle for a few minutes before re- in Section 4.1.4.3 of Method 5 (with appro- moving a sample. Pipette a 100–ml aliquot of priate modification) and record the leak this solution into a 250–ml Erlenmeyer flask, rate. If the post-test leakage rate exceeds add 2 to 4 drops of thorin indicator, and ti- the specified acceptable rate, the tester shall trate to a pink endpoint using 0.0100 N bar- either correct the sample volume, as out- ium perchlorate. Repeat the titration with a lined in Section 6.3 of Method 5, or shall void second aliquot of sample and average the ti- the run. tration values. Replicate titrations must Drain the ice bath and, with the probe dis- agree within 1 percent or 0.2 ml, whichever is connected, purge the remaining part of the greater. train, by drawing clean ambient air through 4.3.2 Container No. 2. Thoroughly mix the the system for 15 minutes at the average solution in the container holding the con- flow rate used for sampling. tents of the second and third impingers. Pi- NOTE: Clean ambient air can be provided pette a 10–ml aliquot of sample into a 250–ml by passing air through a charcoal filter. At Erlenmeyer flask. Add 40 ml of isopropanol, the option of the tester, ambient air (with- 2 to 4 drops of thorin indicator, and titrate out cleaning) may be used. to a pink endpoint using 0.0100 N barium per- chlorate. Repeat the titration with a second 4.1.6 Calculation of Percent Isokinetic. aliquot of sample and average the titration Follow the procedure outlined in Method 5, values. Replicate titrations must agree with- Section 4.1.6. 4.2 Sample Recovery. in 1 percent or 0.2 ml, whichever is greater. 4.2.1 Container No. 1. If a moisture con- 4.3.3 Blanks. Prepare blanks by adding 2 tent analysis is to be done, weigh the first to 4 drops of thorin indicator to 100 ml of 80 impinger plus contents to the nearest 0.5 g percent isopropanol. Titrate the blanks in and record this weight. the same manner as the samples. Transfer the contents of the first impinger 4.4 Quality Control Procedures. Same as to a 250–ml graduated cylinder. Rinse the in Method 5, Section 4.4. probe, first impinger, all connecting glass- 4.5 Audit Sample Analysis. Same as in ware before the filter, and the front half of Method 6, Section 4.4. the filter holder with 80 percent isopropanol. 5. Calibration Add the rinse solution to the cylinder. Dilute 5.1 Calibrate equipment using the proce- to 250 ml with 80 percent isopropanol. Add dures specified in the following sections of the filter to the solution, mix, and transfer Method 5: Section 5.3 (metering system); to the storage container. Protect the solu- Section 5.5 (temperature gauges); Section 5.7 tion against evaporation. Mark the level of (barometer). Note that the recommended liquid on the container and identify the sam- leak-check of the metering system, described ple container. in Section 5.6 of Method 5, also applies to 4.2.2 Container No. 2. If a moisture con- this method. tent analysis is to be done, weigh the second 5.2 Standardize the barium perchlorate and third impingers (plus contents) to the solution with 25 ml of standard sulfuric acid, nearest 0.5 g and record these weights. Also, to which 100 ml of 100 percent isopropanol weigh the spent silica gel (or silica gel plus has been added. impinger) to the nearest 0.5 g. 6. Calculations Transfer the solutions from the second and third impingers to a 1000–ml graduated cyl- NOTE: Carry out calculations retaining at inder. Rinse all connecting glassware (in- least one extra decimal figure beyond that of cluding back half of filter holder) between the acquired data. Round off figures after the filter and silica gel impinger with deion- final calculation. ized, distilled water, and add this rinse water 6.1 Nomenclature. 2 2 to the cylinder. Dilute to a volume of 1000 ml An=Cross-sectional area of nozzle, m (ft ). with deionized, distilled water. Transfer the Bws=Water vapor in the gas stream, propor- solution to a storage container. Mark the tion by volume.
level of liquid on the container. Seal and CH2SO4=Sulfuric acid (including SO3) con- identify the sample container. centration, g/dscm (lb/dscf).
4.3 Analysis. CSO2=Sulfur dioxide concentration, g/dscm Note the level of liquid in Containers 1 and (lb/dscf). 2, and confirm whether or not any sample I=Percent of isokinetic sampling. was lost during shipment; note this on the N=Normality of barium perchlorate titrant, analytical data sheet. If a noticeable amount meq/ml. of leakage has occured, either void the sam- Pbar=Barometric pressure at the sampling ple or use methods, subject to the approval site, mm Hg (in. Hg). of the Administrator, to correct the final re- Ps=Absolute stack gas pressure, mm Hg (in. sults. Hg).
751
VerDate 11
Pstd=Standard absolute pressure, 760 mm Hg 6.3 Dry Gas Volume. Correct the sample (29.92 in. Hg). volume measured by the dry gas meter to Tm=Average absolute dry gas meter tempera- standard conditions (20°C and 760 mm Hg or ture (see Figure 8–2), ° K (° R). 68°F and 29.92 in. Hg) by using Equation 8–1. Ts=Average absolute stack gas temperature (see Figure 8–2), ° K (° R). Tstd=Standard absolute temperature, 293°K (528°R). Va=Volume of sample aliquot titrated, 100 ml for H2SO4 and 10 ml for SO2. Vlc=Total volume of liquid collected in impingers and silica gel, ml. Where:
Vm=Volume of gas sample as measured by K1=0.3858 °K/mm Hg for metric units. dry gas meter, dcm (dcf). =17.64 °R/in., Hg for English units. Vm(std)=Volume of gas sample measured by the dry gas meter corrected to standard NOTE: If the leak rate observed during any conditions, dscm (dscf). mandatory leak-checks exceeds the specified vs=Average stack gas velocity, calculated by acceptable rate, the tester shall either cor- Method 2, Equation 2–9, using data ob- rect the value of Vm in Equation 8–1 (as de- tained from Method 8, m/sec (ft/sec). scribed in Section 6.3 of Method 5), or shall Vsoln=Total volume of solution in which the invalidate the test run. sulfuric acid or sulfur dioxide sample is 6.4 Volume of Water Vapor and Moisture contained, 250 ml or 1,000 ml, respec- Content. Calculate the volume of water tively. vapor using Equation 5–2 of Method 5; the Vt=Volume of barium perchlorate titrant used for the sample, ml. weight of water collected in the impingers and silica gel can be directly converted to Vtb=Volume of barium perchlorate titrant used for the blank, ml. milliliters (the specific gravity of water is 1 Y=Dry gas meter calibration factor. g/ml). Calculate the moisture content of the ∆H=Average pressure drop across orifice stack gas, using Equation 5–3 of Method 5. meter, mm (in.) H2O. The ‘‘Note’’ in Section 6.5 of Method 5 also θ=Total sampling time, min. applies to this method. Note that if the efflu- 13.6=Specific gravity of mercury. ent gas stream can be considered dry, the 60=sec/min. volume of water vapor and moisture content 100=Conversion to percent. need not be calculated.
6.2 Average Dry Gas Meter Temperature 6.5 Sulfuric Acid Mist (including SO3) and Average Orifice Pressure Drop. See data Concentration. sheet (Figure 8–2).
Where: Where:
K2=0.04904 g/milliequivalent for metric units. K3=0.03203 g/meq for metric units. =1.081×10¥4lb/meq for English units. =7.061×10¥5 lb/meq for English units. 6.6 Sulfur Dioxide Concentration. 6.7 Isokinetic Variation. 6.7.1 Calculation from Raw Data.
752
VerDate 11
+ +∆ 100Ts[] K4 V 1c () V m Y/T m/.() P bar H/ 13 6 I = Eq. 8- 4 θ 60 vs P s A n
Where: search Triangle Park, NC. EPA–650/4–74–024. 3 K4=0.003464 mm Hg-m /ml-°K for metric December, 1973. units. 7. Annual Book of ASTM Standards. Part =0.002676 in. Hg-ft3/ml-°R for English units. 31; Water, Atmospheric Analysis. pp. 40–42. 6.7.2 Calculation from Intermediate Values. American Society for Testing and Materials. Philadelphia, Pa. 1974.
METHOD 9—VISUAL DETERMINATION OF THE OPACITY OF EMISSIONS FROM STATIONARY SOURCES Many stationary sources discharge visible emissions into the atmosphere; these emis- where: sions are usually in the shape of a plume.
K5=4.320 for metric units. This method involves the determination of =0.09450 for English units plume opacity by qualified observers. The 6.8 Acceptable Results. If 90 percent < I method includes procedures for the training <110 percent, the results are acceptable. If and certification of observers, and proce- the results are low in comparison to the dures to be used in the field for determina- standards and I is beyond the acceptable tion of plume opacity. The appearance of a range, the Administrator may opt to accept plume as viewed by an observer depends upon the results. Use Citation 4 in the Bibliog- a number of variables, some of which may be raphy of Method 5 to make judgments. Oth- controllable and some of which may not be erwise, reject the results and repeat the test. controllable in the field. Variables which can 6.9 Stack Gas Velocity and Volumetric be controlled to an extent to which they no Flow Rate. Calculate the average stack gas longer exert a significant influence upon velocity and volumetric flow rate, if needed, plume appearance include: Angle of the ob- using data obtained in this method and equa- server with respect to the plume; angle of tions in Sections 5.2 and 5.3 of Method 2. the observer with respect to the sun; point of 6.10 Relative Error (RE) for QA Audit observation of attached and detached steam Samples. Same as in Method 6, Section 6.4. plume; and angle of the observer with re- 7. Bibliography spect to a plume emitted from a rectangular 1. Atmospheric Emissions from Sulfuric stack with a large length to width ratio. The Acid Manufacturing Processes. U.S. DHEW, method includes specific criteria applicable PHS, Division of Air Pollution. Public to these variables. Health Service Publication No. 999–AP–13. Other variables which may not be control- Cincinnati, OH. 1965. lable in the field are luminescence and color 2. Corbett, P. F. The Determination of SO2 contrast between the plume and the back- and SO3 in Flue Gases. Journal of the Insti- ground against which the plume is viewed. tute of Fuel. 24:237–243. 1961. These variables exert an influence upon the 3. Martin, Robert M. Construction Details appearance of a plume as viewed by an ob- of Isokinetic Source Sampling Equipment. server, and can affect the ability of the ob- Environmental Protection Agency. Research Triangle Park, NC. Air Pollution Control Of- server to accurately assign opacity values to fice Publication No. APTD–0581. April, 1971. the observed plume. Studies of the theory of 4. Patton, W. F. and J. A. Brink, Jr. New plume opacity and field studies have dem- Equipment and Techniques for Sampling onstrated that a plume is most visible and Chemical Process Gases. Journal of Air Pol- presents the greatest apparent opacity when lution Control Association. 13:162. 1963. viewed against a contrasting background. It 5. Rom, J. J. Maintenance, Calibration, follows from this, and is confirmed by field and Operation of Isokinetic Source-Sampling trials, that the opacity of a plume, viewed Equipment. Office of Air Programs, Environ- under conditions where a contrasting back- mental Protection Agency. Research Tri- ground is present can be assigned with the angle Park, NC. APTD–0576. March, 1972. greatest degree of accuracy. However, the 6. Hamil, H. F. and D. E. Camann. Collabo- potential for a positive error is also the rative Study of Method for Determination of greatest when a plume is viewed under such Sulfur Dioxide Emissions from Stationary contrasting conditions. Under conditions Sources (Fossil Fuel-Fired Steam Genera- presenting a less contrasting background, tors). Environmental Protection Agency. Re- the apparent opacity of a plume is less and
753
VerDate 11
approaches zero as the color and lumines- baghouses, noncircular stacks), approxi- cence contrast decrease toward zero. As a re- mately perpendicular to the longer axis of sult, significant negative bias and negative the outlet. The observer’s line of sight errors can be made when a plume is viewed should not include more than one plume at a under less contrasting conditions. A negative time when multiple stacks are involved, and bias decreases rather than increases the pos- in any case the observer should make his ob- sibility that a plant operator will be cited servations with his line of sight perpendicu- for a violation of opacity standards due to lar to the longer axis of such a set of mul- observer error. tiple stacks (e.g., stub stacks on baghouses). Studies have been undertaken to deter- 2.2 Field Records. The observer shall record mine the magnitude of positive errors which the name of the plant, emission location, can be made by qualified observers while type facility, observer’s name and affili- reading plumes under contrasting conditions ation, a sketch of the observer’s position rel- and using the procedures set forth in this ative to the source, and the date on a field method. The results of these studies (field data sheet (Figure 9–1). The time, estimated trials) which involve a total of 769 sets of 25 distance to the emission location, approxi- readings each are as follows: mate wind direction, estimated wind speed, (1) For black plumes (133 sets at a smoke description of the sky condition (presence generator), 100 percent of the sets were read and color of clouds), and plume background with a positive error 1 of less than 7.5 percent are recorded on a field data sheet at the time opacity; 99 percent were read with a positive opacity readings are initiated and com- error of less than 5 percent opacity. pleted. (2) For white plumes (170 sets at a smoke 2.3 Observations. Opacity observations generator, 168 sets at a coal-fired power shall be made at the point of greatest opac- plant, 298 sets at a sulfuric acid plant), 99 ity in that portion of the plume where con- percent of the sets were read with a positive densed water vapor is not present. The ob- error of less than 7.5 percent opacity; 95 per- server shall not look continuously at the cent were read with a positive error of less plume, but instead shall observe the plume than 5 percent opacity. momentarily at 15-second intervals. The positive observational error associated 2.3.1 Attached Steam Plumes. When con- with an average of twenty-five readings is densed water vapor is present within the therefore established. The accuracy of the plume as it emerges from the emission out- method must be taken into account when de- let, opacity observations shall be made be- yond the point in the plume at which con- termining possible violations of applicable densed water vapor is no longer visible. The opacity standards. observer shall record the approximate dis- 1. Principle and Applicability tance from the emission outlet to the point 1.1 Principle. The opacity of emissions in the plume at which the observations are from stationary sources is determined vis- made. ually by a qualified observer. 2.3.2 Detached Steam Plume. When water 1.2 Applicability. This method is applicable vapor in the plume condenses and becomes for the determination of the opacity of emis- visible at a distinct distance from the emis- sions from stationary sources pursuant to sion outlet, the opacity of emissions should § 60.11(b) and for qualifying observers for vis- be evaluated at the emission outlet prior to ually determining opacity of emissions. the condensation of water vapor and the for- 2. Procedures mation of the steam plume. 2.4 Recording Observations. Opacity obser- The observer qualified in accordance with vations shall be recorded to the nearest 5 section 3 of this method shall use the follow- percent at 15-second intervals on an observa- ing procedures for visually determining the tional record sheet. (See Figure 9–2 for an ex- opacity of emissions: ample.) A minimum of 24 observations shall 2.1 Position. The qualified observer shall be recorded. Each momentary observation stand at a distance sufficient to provide a recorded shall be deemed to represent the av- clear view of the emissions with the sun ori- erage opacity of emissions for a 15-second pe- ° ented in the 140 sector to his back. Consist- riod. ent with maintaining the above requirement, 2.5 Data Reduction. Opacity shall be deter- the observer shall, as much as possible, make mined as an average of 24 consecutive obser- his observations from a position such that vations recorded at 15-second intervals. Di- his line of vision is approximately per- vide the observations recorded on the record pendicular to the plume direction, and when sheet into sets of 24 consecutive observa- observing opacity of emissions from rectan- tions. A set is composed of any 24 consecu- gular outlets (e.g., roof monitors, open tive observations. Sets need not be consecu- tive in time and in no case shall two sets 1 For a set, positive error = average opac- overlap. For each set of 24 observations, cal- ity determined by observers’ 25 observa- culate the average by summing the opacity tions—average opacity determined from of the 24 observations and dividing this sum transmissometer’s 25 recordings. by 24. If an applicable standard specifies an
754
VerDate 11
averaging time requiring more than 24 obser- shall be checked and if the drift exceeds ±1 vations, calculate the average for all obser- percent opacity, the condition shall be cor- vations made during the specified time pe- rected prior to conducting any subsequent riod. Record the average opacity on a record test runs. The smoke meter shall be dem- sheet. (See Figure 9–1 for an example.) onstrated, at the time of installation, to 3. Qualifications and Testing meet the specifications listed in Table 9–1. 3.1 Certification Requirements. To receive This demonstration shall be repeated follow- certification as a qualified observer, a can- ing any subsequent repair or replacement of didate must be tested and demonstrate the the photocell or associated electronic cir- ability to assign opacity readings in 5 per- cuitry including the chart recorder or output cent increments to 25 different black plumes meter, or every 6 months, whichever occurs and 25 different white plumes, with an error first. not to exceed 15 percent opacity on any one reading and an average error not to exceed TABLE 9±1ÐSMOKE METER DESIGN AND 7.5 percent opacity in each category. Can- PERFORMANCE SPECIFICATIONS didates shall be tested according to the pro- cedures described in section 3.2. Smoke gen- Parameter Specification erators used pursuant to section 3.2 shall be a. Light source ...... Incandescent lamp operated at equipped with a smoke meter which meets nominal rated voltage. the requirements of section 3.3. b. Spectral response of Photopic (daylight spectral re- The certification shall be valid for a period photocell. sponse of the human eyeÐCi- of 6 months, at which time the qualification tation 3). procedure must be repeated by any observer c. Angle of view ...... 15° maximum total angle. in order to retain certification. d. Angle of projection ...... 15° maximum total angle. 3.2 Certification Procedure. The certifi- e. Calibration error ...... ±3% opacity, maximum. ± cation test consists of showing the candidate f. Zero and span drift ...... 1% opacity, 30 minutes g. Response time ...... 5 seconds. a complete run of 50 plumes—25 black plumes and 25 white plumes—generated by a smoke 3.3.1 Calibration. The smoke meter is cali- generator. Plumes within each set of 25 brated after allowing a minimum of 30 min- black and 25 white runs shall be presented in random order. The candidate assigns an utes warmup by alternately producing simu- opacity value to each plume and records his lated opacity of 0 percent and 100 percent. observation on a suitable form. At the com- When stable response at 0 percent or 100 per- pletion of each run of 50 readings, the score cent is noted, the smoke meter is adjusted to of the candidate is determined. If a can- produce an output of 0 percent or 100 per- didate fails to qualify, the complete run of 50 cent, as appropriate. This calibration shall readings must be repeated in any retest. The be repeated until stable 0 percent and 100 smoke test may be administered as part of a percent readings are produced without ad- smoke school or training program, and may justment. Simulated 0 percent and 100 per- be preceded by training or familiarization cent opacity values may be produced by al- runs of the smoke generator during which ternately switching the power to the light candidates are shown black and white source on and off while the smoke generator plumes of known opacity. is not producing smoke. 3.3 Smoke Generator Specifications. Any 3.3.2 Smoke Meter Evaluation. The smoke smoke generator used for the purposes of meter design and performance are to be eval- section 3.2 shall be equipped with a smoke uated as follows: meter installed to measure opacity across 3.3.2.1 Light Source. Verify from manufac- the diameter of the smoke generator stack. turer’s data and from voltage measurements The smoke meter output shall display made at the lamp, as installed, that the lamp instack opacity based upon a pathlength is operated within ±5 percent of the nominal equal to the stack exit diameter, on a full 0 rated voltage. to 100 percent chart recorder scale. The 3.3.2.2 Spectral Response of Photocell. Ver- smoke meter optical design and performance ify from manufacturer’s data that the photo- shall meet the specifications shown in Table cell has a photopic response; i.e., the spectral 9–1. The smoke meter shall be calibrated as sensitivity of the cell shall closely approxi- prescribed in section 3.3.1 prior to the con- mate the standard spectral-luminosity curve duct of each smoke reading test. At the com- for photopic vision which is referenced in (b) pletion of each test, the zero and span drift of Table 9–1.
755
VerDate 11
756
VerDate 11
FIGURE 9±2ÐOBSERVATION RECORD Page ll of ll Company ...... Observer ...... Location ...... Type facility ...... Test Number ...... Point of emissions ...... Date.
Seconds Steam plume (check if applicable) Hr. Min. Comments 0 15 30 45 Attached Detached
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
FIGURE 9±2ÐOBSERVATION RECORDÐ(CONTINUED) Page ll of ll Company ...... Observer ...... Location ...... Type facility ......
757
VerDate 11
FIGURE 9±2ÐOBSERVATION RECORDÐ(CONTINUED)ÐContinued Page ll of ll Test Number ...... Point of emissions ...... Date.
Seconds Steam plume (check if applicable) Hr. Min. Comments 0 15 30 45 Attached Detached
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
3.3.2.3 Angle of View. Check construction where θ=total angle of view; d=the sum of geometry to ensure that the total angle of the photocell diameter+the diameter of the view of the smoke plume, as seen by the pho- limiting aperture; and L=the distance from tocell, does not exceed 15°. The total angle of the photocell to the limiting aperture. The view may be calculated from: θ= 2 tan¥ 1d/2L,
758
VerDate 11
limiting aperture is the point in the path be- system (laser radar; Light Detection and tween the photocell and the smoke plume Ranging). The method includes procedures where the angle of view is most restricted. In for the calibration of the lidar and proce- smoke generator smoke meters this is nor- dures to be used in the field for the lidar de- mally an orifice plate. termination of plume opacity. The lidar is 3.3.2.4 Angle of Projection. Check construc- used to measure plume opacity during either tion geometry to ensure that the total angle day or nighttime hours because it contains of projection of the lamp on the smoke its own pulsed light source or transmitter. plume does not exceed 15°. The total angle of The operation of the lidar is not dependent projection may be calculated from: θ=2 upon ambient lighting conditions (light, tan¥ 1d/2L, where θ= total angle of projec- dark, sunny or cloudy). tion; d= the sum of the length of the lamp The lidar mechanism or technique is appli- filament + the diameter of the limiting aper- cable to measuring plume opacity at numer- ture; and L= the distance from the lamp to ous wavelengths of laser radiation. However, the limiting aperture. the performance evaluation and calibration 3.3.2.5 Calibration Error. Using neutral- test results given in support of this method density filters of known opacity, check the apply only to a lidar that employs a ruby error between the actual response and the (red light) laser [Reference 5.1]. theoretical linear response of the smoke 1. Principle and Applicability meter. This check is accomplished by first calibrating the smoke meter according to 1.1 Principle. The opacity of visible emis- 3.3.1 and then inserting a series of three neu- sions from stationary sources (stacks, roof tral-density filters of nominal opacity of 20, vents, etc.) is measured remotely by a mo- 50, and 75 percent in the smoke meter bile lidar (laser radar). pathlength. Filters calibrated within ±2 per- 1.2 Applicability. This method is applica- cent shall be used. Care should be taken ble for the remote measurement of the opac- when inserting the filters to prevent stray ity of visible emissions from stationary light from affecting the meter. Make a total sources during both nighttime and daylight of five nonconsecutive readings for each fil- conditions, pursuant to 40 CFR § 60.11(b). It is ter. The maximum error on any one reading also applicable for the calibration and per- shall be 3 percent opacity. formance verification of the mobile lidar for 3.3.2.6 Zero and Span Drift. Determine the the measurement of the opacity of emis- zero and span drift by calibrating and oper- sions. A performance/design specification for ating the smoke generator in a normal man- a basic lidar system is also incorporated into ner over a 1-hour period. The drift is meas- this method. ured by checking the zero and span at the 1.3 Definitions. end of this period. Azimuth angle: The angle in the horizontal 3.3.2.7 Response Time. Determine the re- plane that designates where the laser beam sponse time by producing the series of five is pointed. It is measured from an arbitrary simulated 0 percent and 100 percent opacity fixed reference line in that plane. values and observing the time required to Backscatter: The scattering of laser light reach stable response. Opacity values of 0 in a direction opposite to that of the inci- percent and 100 percent may be simulated by dent laser beam due to reflection from par- alternately switching the power to the light ticulates along the beam’s atmospheric path source off and on while the smoke generator which may include a smoke plume. is not operating. Backscatter signal: The general term for the lidar return signal which results from 4. Bibliography. laser light being backscattered by atmos- 1. Air Pollution Control District Rules and pheric and smoke plume particulates. Regulations, Los Angeles County Air Pollu- Convergence distance: The distance from tion Control District, Regulation IV, Prohi- the lidar to the point of overlap of the lidar bitions, Rule 50. receiver’s field-of-view and the laser beam. 2. Weisburd, Melvin I., Field Operations Elevation angle: The angle of inclination and Enforcement Manual for Air, U.S. Envi- of the laser beam referenced to the hori- ronmental Protection Agency, Research Tri- zontal plane. angle Park, NC. APTD–1100, August 1972, pp. Far region: The region of the atmosphere’s 4.1–4.36. path along the lidar line-of-sight beyond or 3. Condon, E.U., and Odishaw, H., Hand- behind the plume being measured. book of Physics, McGraw-Hill Co., New York, Lidar: Acronym for Light Detection and NY, 1958, Table 3.1, p. 6–52. Ranging. Lidar range: The range or distance from ALTERNATE METHOD 1—DETERMINATION OF the lidar to a point of interest along the THE OPACITY OF EMISSIONS FROM STATION- lidar line-of-sight. ARY SOURCES REMOTELY BY LIDAR Near region: The region of the atmospheric This alternate method provides the quan- path along the lidar line-of-sight between titative determination of the opacity of an the lidar’s convergence distance and the emissions plume remotely by a mobile lidar plume being measured.
759
VerDate 11
Opacity: One minus the optical transmit- atmospheric backscatter signal [Reference tance of a smoke plume, screen target, etc. 5.1]. This backscatter signal should be re- Pick interval: The time or range intervals corded. in the lidar backscatter signal whose mini- When there is more than one source of mum average amplitude is used to calculate emissions in the immediate vicinity of the opacity. Two pick intervals are required, one plume, the lidar shall be positioned so that in the near region and one in the far region. the laser beam passes through only a single Plume: The plume being measured by lidar. plume, free from any interference of the Plume signal: The backscatter signal re- other plumes for a minimum of 50 meters or sulting from the laser light pulse passing three consecutive pick intervals (whichever through a plume. is greater) in each region before and beyond 1/R2correction: The correction made for the the plume along the line-of-sight (deter- systematic decrease in lidar backscatter sig- mined from the backscatter signals). The nal amplitude with range. lidar shall initially be positioned so that its Reference signal: The backscatter signal resulting from the laser light pulse passing line-of-sight is approximately perpendicular through ambient air. to the plume. Sample interval: The time period between When measuring the opacity of emissions successive samples for a digital signal or be- from rectangular outlets (e.g., roof monitors, tween successive measurements for an ana- open baghouses, noncircular stacks, etc.), log signal. the lidar shall be placed in a position so that Signal spike: An abrupt, momentary in- its line-of-sight is approximately perpendicu- crease and decrease in signal amplitude. lar to the longer (major) axis of the outlet. Source: The source being tested by lidar. 2.2 Lidar Operational Restrictions. The Time reference: The time (to) when the lidar receiver shall not be aimed within an laser pulse emerges from the laser, used as angle of ± 15° (cone angle) of the sun. the reference in all lidar time or range meas- This method shall not be used to make urements. opacity measurements if thunderstorms, 2. Procedures snowstorms, hail storms, high wind, high- The mobile lidar calibrated in accordance ambient dust levels, fog or other atmos- with Paragraph 3 of this method shall use pheric conditions cause the reference signals the following procedures for remotely meas- to consistently exceed the limits specified in uring the opacity of stationary source emis- Section 2.3. sions: 2.3 Reference Signal Requirements. Once 2.1 Lidar Position. The lidar shall be posi- placed in its proper position for opacity tioned at a distance from the plume suffi- measurement, the laser is aimed and fired cient to provide an unobstructed view of the with the line-of-sight near the outlet height source emissions. The plume must be at a and rotated horizontally to a position clear range of at least 50 meters or three consecu- of the source structure and the associated tive pick intervals (whichever is greater) plume. The backscatter signal obtained from from the lidar’s transmitter/receiver conver- this position is called the ambient-air or ref- gence distance along the line-of-sight. The erence signal. The lidar operator shall in- maximum effective opacity measurement spect this signal [Section V of Reference 5.1] distance of the lidar is a function of local at- to: (1) determine if the lidar line-of-sight is mospheric conditions, laser beam diameter, free from interference from other plumes and and plume diameter. The test position of the from physical obstructions such as cables, lidar shall be selected so that the diameter power lines, etc., for a minimum of 50 meters of the laser beam at the measurement point or three consecutive pick intervals (which- within the plume shall be no larger than ever is greater) in each region before and be- three-fourths the plume diameter. The beam yond the plume, and (2) obtain a qualitative diameter is calculated by Equation (AM1–1): measure of the homogeneity of the ambient D(lidar)=A+Rφ≤0.75 D(Plume) (AM1–1) air by noting any signal spikes. Where: Should there be any signal spikes on the D(Plume)=diameter of the plume (cm), reference signal within a minimum of 50 me- φ=laser beam divergence measured in radians ters or three consecutive pick intervals R=range from the lidar to the source (cm) (whichever is greater) in each region before D(Lidar)=diameter of the laser beam at and beyond the plume, the laser shall be range R (cm), fired three more times and the operator shall A=diameter of the laser beam or pulse where inspect the reference signals on the display. it leaves the laser. If the spike(s) remains, the azimuth angle The lidar range, R, is obtained by aiming and shall be changed and the above procedures firing the laser at the emissions source conducted again. If the spike(s) disappears in structure immediately below the outlet. The all three reference signals, the lidar line-of- range value is then determined from the sight is acceptable if there is shot-to-shot backscatter signal which consists of a signal consistency and there is no interference from spike (return from source structure) and the other plumes.
760
VerDate 11
Shot-to-shot consistency of a series of ref- been recorded, and the total stop or inter- erence signals over a period of twenty sec- rupt time does not exceed 3 minutes. onds is verified in either of two ways. (1) The 2.4.1 Non-hydrated Plumes. The laser lidar operator shall observe the reference shall be aimed at the region of the plume signal amplitudes. For shot-to-shot consist- which displays the greatest opacity. The ency the ratio of Rf to Rn [amplitudes of the lidar operator must visually verify that the near and far region pick intervals (Section laser is aimed clearly above the source exit 2.6.1)] shall vary by not more than ± 6% be- structure. tween shots; or (2) the lidar operator shall 2.4.2 Hydrated Plumes. The lidar will be accept any one of the reference signals and used to measure the opacity of hydrated or treat the other two as plume signals; then so-called steam plumes. As listed in the ref- the opacity for each of the subsequent ref- erence method, there are two types, i.e., at- erence signals is calculated (Equation AM1– tached and detached steam plumes. 2). For shot-to-shot consistency, the opacity 2.4.2.1 Attached Steam Plumes. When con- values shall be within ± 3% of 0% opacity and densed water vapor is present within a the associated So values less than or equal to plume, lidar opacity measurements shall be 8% (full scale) [Section 2.6]. made at a point within the residual plume If a set of reference signals fails to meet where the condensed water vapor is no the requirements of this section, then all longer visible. The laser shall be aimed into plume signals [Section 2.4] from the last set the most dense region (region of highest of acceptable reference signals to the failed opacity) of the residual plume. set shall be discarded. During daylight hours the lidar operator locates the most dense portion of the resid- 2.3.1 Initial and Final Reference Signals. ual plume visually. During nighttime hours Three reference signals shall be obtained a high-intensity spotlight, night vision within a 90-second time period prior to any scope, or low light level TV, etc., can be used data run. A final set of three reference sig- as an aid to locate the residual plume. If vis- nals shall be obtained within three (3) min- ual determination is ineffective, the lidar utes after the completion of the same data may be used to locate the most dense region run. of the residual plume by repeatedly measur- 2.3.2 Temporal Criterion for Additional ing opacity, along the longitudinal axis or Reference Signals. An additional set of ref- center of the plume from the emissions out- erence signals shall be obtained during a let to a point just beyond the steam plume. data run if there is a change in wind direc- The lidar operator should also observe color tion or plume drift of 30° or more from the differences and plume reflectivity to ensure direction that was prevalent when the last that the lidar is aimed completely within the set of reference signals was obtained. An ad- residual plume. If the operator does not ob- ditional set of reference signals shall also be tain a clear indication of the location of the obtained if there is an increase in value of residual plume, this method shall not be SIn (near region standard deviation, Equa- used. tion AM1–5) or SIf (far region standard devi- Once the region of highest opacity of the ation, Equation AM1–6) that is greater than residual plume has been located, aiming ad- 6% (full scale) over the respective values cal- justments shall be made to the laser line-of- culated from the immediately previous sight to correct for the following: movement plume signal, and this increase in value re- to the region of highest opacity out of the mains for 30 seconds or longer. An additional lidar line-of-sight (away from the laser set of reference signals shall also be obtained beam) for more than 15 seconds, expansion of if there is a change in amplitude in either the steam plume (air temperature lowers the near or the far region of the plume sig- and/or relative humidity increases) so that it nal, that is greater than 6% of the near sig- just begins to encroach on the field-of-view nal amplitude and this change in amplitude of the lidar’s optical telescope receiver, or a remains for 30 seconds or more. decrease in the size of the steam plume (air 2.4 Plume Signal Requirements. Once temperature higher and/or relative humidity properly aimed, the lidar is placed in oper- decreases) so that regions within the resid- ation with the nominal pulse or firing rate of ual plume whose opacity is higher than the six pulses/minute (1 pulse/10 seconds). The one being monitored, are present. lidar operator shall observe the plume 2.4.2.2 Detached Steam Plumes. When the backscatter signals to determine the need water vapor in a hydrated plume condenses for additional reference signals as required and becomes visible at a finite distance from by Section 2.3.2. The plume signals are re- the stack or source emissions outlet, the corded from lidar start to stop and are called opacity of the emissions shall be measured in a data run. The length of a data run is deter- the region of the plume clearly above the mined by operator discretion. Short-term emissions outlet and below condensation of stops of the lidar to record additional ref- the water vapor. erence signals do not constitute the end of a During daylight hours the lidar operators data run if plume signals are resumed within can visually determine if the steam plume is 90 seconds after the reference signals have detached from the stack outlet. During
761
VerDate 11
nighttime hours a high-intensity spotlight, lidar to make interference-free opacity night vision scope, low light level TV, etc., measurements, this method shall not be can be used as an aid in determining if the used. steam plume is detached. If visual deter- 2.5 Field Records. In addition to the re- mination is ineffective, the lidar may be cording recommendations listed in other sec- used to determine if the steam plume is de- tions of this method the following records tached by repeatedly measuring plume opac- should be maintained. Each plume measured ity from the outlet to the steam plume along should be uniquely identified. The name of the plume’s longitudinal axis or center line. the facility, type of facility, emission source The lidar operator should also observe color type, geographic location of the lidar with differences and plume reflectivity to detect a respect to the plume, and plume characteris- detached plume. If the operator does not ob- tics should be recorded. The date of the test, tain a clear indication of the location of the the time period that a source was monitored, detached plume, this method shall not be the time (to the nearest second) of each used to make opacity measurements between opacity measurement, and the sample inter- the outlet and the detached plume. val should also be recorded. The wind speed, Once the determination of a detached wind direction, air temperature, relative hu- steam plume has been confirmed, the laser midity, visibility (measured at the lidar’s shall be aimed into the region of highest position), and cloud cover should be recorded opacity in the plume between the outlet and at the beginning and end of each time period the formation of the steam plume. Aiming for a given source. A small sketch depicting adjustments shall be made to the lidar’s line- the location of the laser beam within the of-sight within the plume to correct for plume should be recorded. changes in the location of the most dense re- If a detached or attached steam plume is gion of the plume due to changes in wind di- present at the emissions source, this fact rection and speed or if the detached steam should be recorded. Figures AM1–I and AM1– plume moves closer to the source outlet en- II are examples of logbook forms that may croaching on the most dense region of the be used to record this type of data. Magnetic plume. If the detached steam plume should tape or paper tape may also be used to record move too close to the source outlet for the data.
762
VerDate 11
763
VerDate 11
764
VerDate 11
2.6 Opacity Calculation and Data Analy- measurement is calculated using Equation sis. Referring to the reference signal and AM1–2. (Op=1¥Tp; Tp is the plume transmit- plume signal in Figure AM1–III, the meas- tance.) ured opacity (Op) in percent for each lidar
765
VerDate 11
Where: val of minimal average amplitude. If So is
In=near-region pick interval signal ampli- still greater than 8%, then this procedure is tude, plume signal, 1/R2corrected, repeated for the far pick interval. This pro- If=far-region pick interval signal amplitude, cedure may be repeated once again for the plume signal, 1/R2corrected, near pick interval, but if So remains greater Rn=near-region pick interval signal ampli- than 8%, the plume signal shall be discarded. tude, reference signal, 1/R2corrected, and The reference signal pick intervals, Rn and Rf=far-region pick interval signal amplitude, Rf, must be chosen over the same time inter- 2 reference signal, 1/R corrected. val as the plume signal pick intervals, In and 2 The 1/R correction to the plume and ref- If, respectively [Figure AM1–III]. Other erence signal amplitudes is made by mul- methods of selecting pick intervals may be tiplying the amplitude for each successive used if they give equivalent results. Field- sample interval from the time reference, by oriented examples of pick interval selection the square of the lidar time (or range) associ- are available in Reference 5.1. ated with that sample interval [Reference The average amplitudes for each of the 5.1]. pick intervals, In, If, Rn, Rf, shall be cal- The first step in selecting the pick inter- culated by averaging the respective individ- vals for Equation AM1–2 is to divide the ual amplitudes of the sample intervals from plume signal amplitude by the reference sig- the plume signal and the associated ref- nal amplitude at the same respective ranges erence signal each corrected for 1/R2. The to obtain a ‘‘normalized’’ signal. The pick in- amplitude of In shall be calculated according tervals selected using this normalized signal, to Equation (AM–3). are a minimum of 15 m (100 nanoseconds) in length and consist of at least 5 contiguous sample intervals. In addition, the following criteria, listed in order of importance, gov- ern pick interval selection. (1) The intervals shall be in a region of the normalized signal Where: where the reference signal meets the require- ments of Section 2.3 and is everywhere great- Ini=the amplitude of the ith sample interval er than zero. (2) The intervals (near and far) (near-region), with the minimum average amplitude are Σ=sum of the individual amplitudes for the chosen. (3) If more than one interval with sample intervals, the same minimum average amplitude is m=number of sample intervals in the pick found, the interval closest to the plume is interval, and chosen. (4) The standard deviation, So, for In=average amplitude of the near-region pick the calculated opacity shall be 8% or less. interval. (So is calculated by Equation AM1–7). Similarly, the amplitudes for If, Rn, and Rf If So is greater than 8%, then the far pick are calculated with the three expressions in interval shall be changed to the next inter- Equation (AM1–4).
The standard deviation, SIn, of the set of Similarly, the standard deviations SIf, SRn, amplitudes for the near-region pick interval, and SRf are calculated with the three expres- In, shall be calculated using Equation (AM1– sions in Equation (AM1–6). 5).
766
VerDate 11
The standard deviation, So, for each associ- ated opacity value, Op, shall be calculated using Equation (AM1–7).
The calculated values of In, If, Rn, Rf, SIn, equivalent results, shall be used to deter- SIf, SRn, SRf, Op, and So should be recorded. mine the need for a correction, to calculate Any plume signal with an So greater than 8% the correction, and to document the point shall be discarded. within the plume at which the opacity was 2.6.1 Azimuth Angle Correction. If the azi- measured. muth angle correction to opacity specified in Figure AM1–IV(b) shows the geometry of this section is performed, then the elevation the opacity correction. L′ is the path angle correction specified in Section 2.6.2 through the plume along which the opacity shall not be performed. When opacity is ′ measured in the residual region of an at- measurement is made. P is the path per- tached steam plume, and the lidar line-of- pendicular to the plume at the same point. ε ′ sight is not perpendicular to the plume, it The angle is the angle between L and the may be necessary to correct the opacity plume center line. The angle (π/2-ε), is the measured by the lidar to obtain the opacity angle between the L′ and P′. The measured that would be measured on a path per- opacity, Op, measured along the path L′ shall pendicular to the plume. The following be corrected to obtain the corrected opacity, method, or any other method which produces Opc, for the path P′, using Equation (AM1–8).
767
VerDate 11
The correction in Equation (AM1–8) shall be performed if the inequality in Equation (AM1–9) is true.
Figure AM1–IV(a) shows the geometry used the plume, the elevation angle of the lidar to calculate ε and the position in the plume from the horizontal plane, and the azimuth at which the lidar measurement is made. angle of the lidar from an arbitrary fixed ref- This analysis assumes that for a given lidar erence in the horizontal plane can all be ob- measurement, the range from the lidar to tained directly.
768
VerDate 11
769
VerDate 11
Rs=range from lidar to source* ...... βs=elevation angle of Rs* ...... Rp=range from lidar to plume at the opacity measurement point* ...... βp=elevation angle of Rp* ...... Ra=range from lidar to plume at some arbitrary point, Pa, so the drift angle of the plume can be determined* βa=elevation angle of Ra* ...... α=angle between Rp and Ra ...... R′s=projection of Rs in the horizontal plane ...... R′p=projection of Rp in the horizontal plane ...... R′a=projection of Ra in the horizontal plane ψ′=angle between R′s and R′p* ...... α′=angle between R′p and R′a* ...... R≤=distance from the source to the opacity measurement point projected in the horizontal plane ...... Rθ=distance from opacity measurement point Pp to the point in the plume Pa......
1/2 The correction angle ε shall be determined Rθ=(Rp2+Ra2¥2 Rp Ra Cosα) using Equation AM1–10. R≤, the distance from the source to the Where: opacity measurement point projected in the ¥1 α=Cos (Cosβp Cosβa Cosα′+Sinβp Sinβa), horizontal plane, shall be determined using and Equation AM1–11.
Where: In the special case where the plume center-
R′s=Rs Cos βs, and line at the opacity measurement point is R′p=Rp Cos βp. horizontal, parallel to the ground, Equation AM1–12 may be used to determine ε instead of Equation AM1–10.
Where: 2.6.2 Elevation Angle Correction. An indi- 2 2 2 1/2 R′′s=(R′ s+Rp Sin βp) . vidual lidar-measured opacity, Op, shall be If the angle ε is such that ε≤ 30° or ε ≥ 150°, corrected for elevation angle if the laser ele- β the azimuth angle correction shall not be vation or inclination angle, p [Figure AM1– performed and the associated opacity value V], is greater than or equal to the value cal- shall be discarded. culated in Equation AM1–13.
The measured opacity, Op, along the lidar plume (horizontal) path, P, by using path L, is adjusted to obtain the cor- Equation (AM1–14). rected opacity, Opc, for the actual
*Obtained directly from lidar. These values should be recorded. 770
VerDate 11
Cosβ O= ()100% 1−() 1 − 0., 01 O p() AM 1- 14 pc p
Where: Opc=corrected opacity for the actual plume thickness P. βp=lidar elevation or inclination angle, The values for β , O and O should be re- O =measured opacity along path L, and p p pc p corded.
771
VerDate 11
2.6.3 Determination of Actual Plume 2.6.4 Calculation of Average Actual Plume Opacity. Actual opacity of the plume shall Opacity. The average of the actual plume be determined by Equation AM1–15. opacity, Opa, shall be calculated as the aver- age of the consecutive individual actual opacity values, Opa, by Equation AM1–16.
772
VerDate 11
The lidar shall be aimed through the cen- ter of the plume within 1 stack diameter of the exit, or through the geometric center of the screen target selected. The lidar shall be set in operation for a 6-minute data run at a Where: nominal pulse rate of 1 pulse every 10 sec- onds. Each backscatter return signal and (O ) =the kth actual opacity value in an pa k each respective opacity value obtained from averaging interval containing n opacity the smoke generator transmissometer, shall values; k is a summing index. Σ=the sum of the individual actual opacity be obtained in temporal coincidence. The values. data shall be analyzed and reduced in accord- n=the number of individual actual opacity ance with Section 2.6 of this method. This values contained in the averaging inter- calibration shall be performed for 0% (clean val. air), and at least five other opacities (nomi- Opa=average actual opacity calculated over nally 10, 20, 40, 60, and 80%). the averaging interval. The average of the lidar opacity values ob- 3. Lidar Performance Verification tained during a 6-minute calibration run shall be calculated and should be recorded. The lidar shall be subjected to two types of Also the average of the opacity values ob- performance verifications that shall be per- formed in the field. The annual calibration, tained from the smoke generator transmis- conducted at least once a year, shall be used someter for the same 6-minute run shall be to directly verify operation and performance calculated and should be recorded. of the entire lidar system. The routine ver- Alternate calibration procedures that do ification, conducted for each emission source not meet the above requirements but measured, shall be used to insure proper per- produce equivalent results may be used. formance of the optical receiver and associ- 3.2 Routine Verification Procedures. Ei- ated electronics. ther one of two techniques shall be used to 3.1 Annual Calibration Procedures. Either conduct this verification. It shall be per- a plume from a smoke generator or screen formed at least once every 4 hours for each targets shall be used to conduct this calibra- emission source measured. The following pa- tion. rameters shall be directly verified. If the screen target method is selected, five 1) The opacity value of 0% plus a minimum screens shall be fabricated by placing an of 5 (nominally 10, 20, 40, 60, and 80%) opacity opaque mesh material over a narrow frame values shall be verified through the PMT de- (wood, metal extrusion, etc.). The screen tector and data processing electronics. shall have a surface area of at least one 2) The zero-signal level (receiver signal square meter. The screen material should be with no optical signal from the source chosen for precise optical opacities of about 10, 20, 40, 60, and 80%. Opacity of each target present) shall be inspected to insure that no shall be optically determined and should be spurious noise is present in the signal. With recorded. If a smoke generator plume is se- the entire lidar receiver and analog/digital lected, it shall meet the requirements of Sec- electronics turned on and adjusted for nor- tion 3.3 of Reference Method 9. This calibra- mal operating performance, the following tion shall be performed in the field during procedures shall be used for Techniques 1 and calm (as practical) atmospheric conditions. 2, respectively. The lidar shall be positioned in accordance 3.2.1 Procedure for Technique 1. This test with Section 2.1. shall be performed with no ambient or stray The screen targets must be placed per- light reaching the PMT detector. The narrow pendicular to and coincident with the lidar band filter (694.3 nanometers peak) shall be line-of-sight at sufficient height above the removed from its position in front of the ground (suggest about 30 ft) to avoid ground- PMT detector. Neutral density filters of level dust contamination. Reference signals nominal opacities of 10, 20, 40, 60, and 80% shall be obtained just prior to conducting shall be used. The recommended test con- the calibration test. figuration is depicted in Figure AM1–VI.
773
VerDate 11
The zero-signal level shall be measured and The light source either shall be a continu- should be recorded, as indicated in Figure ous wave (CW) laser with the beam mechani- AM1–VI(a). This simulated clear-air or 0% cally chopped or a light emitting diode con- opacity value shall be tested in using the se- trolled with a pulse generator (rectangular lected light source depicted in Figure AM1– pulse). (A laser beam may have to be attenu- VI(b). ated so as not to saturate the PMT detector). This signal level shall be measured and
774
VerDate 11
should be recorded. The opacity value is cal- shall be performed at least three times for culated by taking two pick intervals [Sec- each selected opacity value. While the order tion 2.6] about 1 microsecond apart in time is not important, each of the opacity values and using Equation (AM1–2) setting the ratio from the optical generator shall be verified. Rn/Rf=1. This calculated value should be re- The calibrated optical generator opacity corded. value for each selection should be recorded. The simulated clear-air signal level is also The optical generator used for Technique 2 employed in the optical test using the neu- shall be calibrated for actual opacity with an tral density filters. Using the test configura- accuracy of ±1% or better. This calibration tion in Figure AM1–VI(c), each neutral den- shall be done monthly while the generator is sity filter shall be separately placed into the in use and calibrated value should be re- light path from the light source to the PMT corded. detector. The signal level shall be measured Alternate verification procedures that do and should be recorded. The opacity value not meet the above requirements but for each filter is calculated by taking the produce equivalent results may be used. signal level for that respective filter (I ), di- f 3.3 Deviation. The permissible error for viding it by the 0% opacity signal level (I ) n the annual calibration and routine verifica- and performing the remainder of the calcula- tion are: tion by Equation (AM1–2) with Rn/Rf=1. The calculated opacity value for each filter 3.3.1 Annual Calibration Deviation. should be recorded. 3.3.1.1 Smoke Generator. If the lidar- The neutral density filters used for Tech- measured average opacity for each data run nique 1 shall be calibrated for actual opacity is not within ±5% (full scale) of the respec- with accuracy of ±2% or better. This calibra- tive smoke generator’s average opacity over tion shall be done monthly while the filters the range of 0% through 80%, then the lidar are in use and the calibrated values should shall be considered out of calibration. be recorded. 3.3.1.2 Screens. If the lidar-measured aver- 3.2.2 Procedure for Technique 2. An opti- age opacity for each data run is not within cal generator (built-in calibration mecha- ±3% (full scale) of the laboratory-determined nism) that contains a light-emitting diode opacity for each respective simulation (red light for a lidar containing a ruby laser) screen target over the range of 0% through is used. By injecting an optical signal into 80%, then the lidar shall be considered out of the lidar receiver immediately ahead of the calibration. PMT detector, a backscatter signal is simu- 3.3.2 Routine Verification Error. If the lated. With the entire lidar receiver elec- lidar-measured average opacity for each neu- tronics turned on and adjusted for normal tral density filter (Technique 1) or optical operating performance, the optical generator generator selection (Technique 2) is not is turned on and the simulation signal (cor- within ±3% (full scale) of the respective lab- rected for 1/R2) is selected with no plume oratory calibration value then the lidar shall spike signal and with the opacity value equal be considered non-operational. to 0%. This simulated clear-air atmospheric return signal is displayed on the system’s 4. Performance/Design Specification for Basic video display. The lidar operator then makes Lidar System any fine adjustments that may be necessary 4.1 Lidar Design Specification. The essen- to maintain the system’s normal operating tial components of the basic lidar system are range. a pulsed laser (transmitter), optical receiver, The opacity values of 0% and the other five detector, signal processor, recorder, and an values are selected one at a time in any aiming device that is used in aiming the order. The simulated return signal data lidar transmitter and receiver. Figure AM1– should be recorded. The opacity value shall VII shows a functional block diagram of a be calculated. This measurement/calculation basic lidar system.
775
VerDate 11
4.2 Performance Evaluation Tests. The of this method. The annual calibration shall owner of a lidar system shall subject such a be performed for three separate, complete lidar system to the performance verification tests described in Section 3, prior to first use
776
VerDate 11
runs and the results of each should be re- 2. Range and Sensitivity corded. The requirements of Section 3.3.1 2.1 Range. 0 to 1,000 ppm. must be fulfilled for each of the three runs. 2.2 Sensitivity. Minimum detectable con- Once the conditions of the annual calibra- centration is 20 ppm for a 0 to 1,000 ppm tion are fulfilled the lidar shall be subjected span. to the routine verification for three separate 3. Interferences complete runs. The requirements of Section 3.3.2 must be fulfilled for each of the three Any substance having a strong absorption runs and the results should be recorded. The of infrared energy will interfere to some ex- Administrator may request that the results tent. For example, discrimination ratios for of the performance evaluation be submitted water (H2O) and carbon dioxide (CO2) are 3.5 for review. percent H2O per 7 ppm CO and 10 percent CO2 per 10 ppm CO, respectively, for devices 5. References measuring in the 1,500 to 3,000 ppm range. 5.1 The Use of Lidar for Emissions Source For devices measuring in the 0 to 100 ppm Opacity Determination, U.S. Environmental range, interference ratios can be as high as Protection Agency, National Enforcement 3.5 percent H2O per 25 ppm CO and 10 percent Investigations Center, Denver, CO. EPA–330/ CO2 per 50 ppm CO. The use of silica gel and 1–79–003–R, Arthur W. Dybdahl, current edi- ascarite traps will alleviate the major inter- tion [NTIS No. PB81–246662]. ference problems. The measured gas volume 5.2 Field Evaluation of Mobile Lidar for must be corrected if these traps are used. the Measurement of Smoke Plume Opacity, 4. Precision and Accuracy U.S. Environmental Protection Agency, Na- tional Enforcement Investigations Center, 4.1 Precision. The precision of most NDIR Denver, CO. EPA/NEIC–TS–128, February analyzers is approximately ±2 percent of 1976. span. 5.3 Remote Measurement of Smoke Plume 4.2 Accuracy. The accuracy of most NDIR Transmittance Using Lidar, C. S. Cook, G. analyzers is approximately ±5 percent of W. Bethke, W. D. Conner (EPA/RTP). Applied span after calibration. Optics 11, pg 1742. August 1972. 5. Apparatus 5.4 Lidar Studies of Stack Plumes in 5.1 Continuous Sample (Figure 10–1). Rural and Urban Environments, EPA–650/4– 5.1.1 Probe. Stainless steel or sheathed 73–002, October 1973. Pyrex1 glass, equipped with a filter to re- 5.5 American National Standard for the move particulate matter. Safe Use of Lasers ANSI Z 136.1–176, March 8, 5.1.2 Air-Cooled Condenser or Equivalent. 1976. To remove any excess moisture. 5.6 U.S. Army Technical Manual TB MED 5.2 Integrated Sample (Figure 10–2). 279, Control of Hazards to Health from Laser 5.2.1 Probe. Stainless steel or sheathed Radiation, February 1969. Pyrex glass, equipped with a filter to remove 5.7 Laser Institute of America Laser Safe- particulate matter. ty Manual, 4th Edition. 5.2.2 Air-Cooled Condenser or Equivalent. 5.8 U.S. Department of Health, Education To remove any excess moisture. and Welfare, Regulations for the Administra- 5.2.3 Valve. Needle valve, or equivalent, to tion and Enforcement of the Radiation Con- adjust flow rate. trol for Health and Safety Act of 1968, Janu- 5.2.4 Pump. Leak-free diaphragm type, or ary 1976. equivalent, to transport gas. 5.9 Laser Safety Handbook, Alex Mallow, 5.2.5 Rate Meter. Rotameter, or equivalent, Leon Chabot, Van Nostrand Reinhold Co., to measure a flow range from 0 to 1.0 liter 1978. per min (0.035 cfm). 5.2.6 Flexible Bag. Tedlar, or equivalent, METHOD 10—DETERMINATION OF CARBON MON- with a capacity of 60 to 90 liters (2 to 3 ft 3). OXIDE EMISSIONS FROM STATIONARY Leak-test the bag in the laboratory before SOURCES using by evacuating bag with a pump fol- 1. Principle and Applicability lowed by a dry gas meter. When evacuation 1.1 Principle. An integrated or continuous is complete, there should be no flow through gas sample is extracted from a sampling the meter. point and analyzed for carbon monoxide (CO) 5.2.7 Pitot Tube. Type S, or equivalent, at- content using a Luft-type nondispersive in- tached to the probe so that the sampling frared analyzer (NDIR) or equivalent. rate can be regulated proportional to the 1.2 Applicability. This method is applicable stack gas velocity when velocity is varying for the determination of carbon monoxide with the time or a sample traverse is con- emissions from stationary sources only when ducted. specified by the test procedures for deter- 5.3 Analysis (Figure 10–3). mining compliance with new source perform- ance standards. The test procedure will indi- 1 Mention of trade names or specific prod- cate whether a continuous or an integrated ucts does not constitute endorsement by the sample is to be used. Environmental Protection Agency.
777
VerDate 11
5.3.1 Carbon Monoxide Analyzer. Nondis- persive infrared spectrometer, or equivalent. This instrument should be demonstrated, preferably by the manufacturer, to meet or exceed manufacturer’s specifications and those described in this method. 5.3.2 Drying Tube. To contain approxi- mately 200 g of silica gel. 5.3.3 Calibration Gas. Refer to section 6.1. 5.3.4 Filter. As recommended by NDIR manufacturer.
6.2 Silica Gel. Indicating type, 6 to 16 mesh, dried at 175°C (347°F) for 2 hours. 6.3 Ascarite. Commercially available. 7. Procedure 7.1 Sampling. 7.1.1 Continuous Sampling. Set up the equipment as shown in Figure 10–1 making sure all connections are leak free. Place the probe in the stack at a sampling point and purge the sampling line. Connect the ana- lyzer and begin drawing sample into the ana- lyzer. Allow 5 minutes for the system to sta- bilize, then record the analyzer reading as required by the test procedure. (See sec- tion 7.2 and 8). CO2 content of the gas may be determined by using the Method 3 integrated sample procedure, or by weighing the ascarite CO2 removal tube and computing CO2 concentration from the gas volume sam- pled and the weight gain of the tube. 7.1.2 Integrated Sampling. Evacuate the flexible bag. Set up the equipment as shown in Figure 10–2 with the bag disconnected. Place the probe in the stack and purge the sampling line. Connect the bag, making sure 5.3.5 CO2 Removal Tube. To contain ap- proximately 500 g of ascarite. that all connections are leak free. Sample at a rate proportional to the stack velocity. 5.3.6 Ice Water Bath. For ascarite and sili- CO content of the gas may be determined by ca gel tubes. 2 using the Method 3 integrated sample proce- 5.3.7 Valve. Needle valve, or equivalent, to dures, or by weighing the ascarite CO2 re- adjust flow rate moval tube and computing CO2 concentra- 5.3.8 Rate Meter. Rotameter or equivalent tion from the gas volume sampled and the to measure gas flow rate of 0 to 1.0 liter per weight gain of the tube. min (0.035 cfm) through NDIR. 7.2 CO Analysis. Assemble the apparatus as 5.3.9 Recorder (optional). To provide perma- shown in Figure 10–3, calibrate the instru- nent record of NDIR readings. ment, and perform other required operations 6. Reagents as described in section 8. Purge analyzer with N2 prior to introduction of each sample. 6.1 Calibration Gases. Known concentra- Direct the sample stream through the in- tion of CO in nitrogen (N2) for instrument strument for the test period, recording the span, prepurified grade of N2 for zero, and readings. Check the zero and span again two additional concentrations corresponding after the test to assure that any drift or mal- approximately to 60 percent and 30 percent function is detected. Record the sample data span. The span concentration shall not ex- on Table 10–1. ceed 1.5 times the applicable source perform- 8. Calibration ance standard. The calibration gases shall be certified by the manufacturer to be within ±2 Assemble the apparatus according to Fig- ure 10–3. Generally an instrument requires a percent of the specified concentration. warm-up period before stability is obtained. Follow the manufacturer’s instructions for specific procedure. Allow a minimum time of 1 hour for warm-up. During this time check
778
VerDate 11
the sample conditioning apparatus, i.e., fil- ADDENDA ter, condenser, drying tube, and CO2 removal tube, to ensure that each component is in A. PERFORMANCE SPECIFICATIONS FOR NDIR good operating condition. Zero and calibrate CARBON MONOXIDE ANALYZERS the instrument according to the manufactur- er’s procedures using, respectively, nitrogen Range (minimum) ...... 0±1000 ppm. and the calibration gases. Output (minimum) ...... 0±10mV. Minimum detectable sensitiv- 20 ppm. ity. TABLE 10±1ÐFIELD DATA Rise time, 90 percent (maxi- 30 seconds. mum). Comments Fall time, 90 percent (maxi- 30 seconds. mum). Location ...... Zero drift (maximum) ...... 10% in 8 hours. Test ...... Span drift (maximum) ...... 10% in 8 hours. Date ...... Precision (minimum) ...... ±2% of full scale. Operator ...... Noise (maximum) ...... ±1% of full scale. Linearity (maximum deviation) 2% of full scale. Interference rejection ratio ..... CO Ð1000 to 1, H OÐ500 Rotameter setting, liters per 2 2 Clock time minute (cubic feet per minute) to 1. B. Definitions of Performance Specifications. Range— The minimum and maximum measurement limits. 9. Calculation Output— Electrical signal which is propor- Calculate the concentration of carbon tional to the measurement; intended for con- monoxide in the stack using Equation 10–1. nection to readout or data processing de- CCO stack=CCO NDIR(1¥Fco2) vices. Usually expressed as millivolts or milliamps full scale at a given impedance. Eq. 10–1 Full scale— The maximum measuring limit Where: for a given range. CCO stack=Concentration of CO in stack, ppm Minimum detectable sensitivity— The small- by volume (dry basis). est amount of input concentration that can CCO NDIR=Concentration of CO measured by be detected as the concentration approaches NDIR analyzer, ppm by volume (dry zero. basis). Accuracy— The degree of agreement be- FCO 2=Volume fraction of CO2 in sample, i.e., tween a measured value and the true value; percent CO2 from Orsat analysis divided usually expressed as ± percent of full scale. by 100. Time to 90 percent response— The time inter- val from a step change in the input con- 10. Alternative Procedures centration at the instrument inlet to a read- 10.1 Interference Trap. The sample condi- ing of 90 percent of the ultimate recorded tioning system described in Method 10A, sec- concentration. tions 2.1.2 and 4.2, may be used as an alter- Rise Time (90 percent)—The interval be- native to the silica gel and ascarite traps. tween initial response time and time to 90 percent response after a step increase in the 11. Bibliography inlet concentration. 1. McElroy, Frank, The Intertech NDIR–CO Fall Time (90 percent)—The interval be- Analyzer, Presented at 11th Methods tween initial response time and time to 90 Conference on Air Pollution, University percent response after a step decrease in the of California, Berkeley, CA. April 1, 1970. inlet concentration. 2. Jacobs, M. B., et al., Continuous Deter- Zero Drift— The change in instrument out- mination of Carbon Monoxide and Hydro- put over a stated time period, usually 24 carbons in Air by a Modified Infrared An- hours, of unadjusted continuous operation alyzer, J. Air Pollution Control Associa- when the input concentration is zero; usu- tion, 9(2): 110–114. August 1959. ally expressed as percent full scale. 3. MSA LIRA Infrared Gas and Liquid Ana- Span Drift— The change in instrument out- lyzer Instruction Book, Mine Safety Ap- put over a stated time period, usually 24 pliances Co., Technical Products Divi- hours, of unadjusted continuous operation sion, Pittsburgh, PA. when the input concentration is a stated 4. Models 215A, 315A, and 415A Infrared Ana- upscale value; usually expressed as percent lyzers, Beckman Instruments, Inc., Beck- full scale. man Instructions 1635–B, Fullerton, CA. Precision— The degree of agreement be- October 1967. tween repeated measurements of the same 5. Continuous CO Monitoring System, concentration, expressed as the average devi- Model A5611, Intertech Corp., Princeton, ation of the single results from the mean. NJ. Noise—Spontaneous deviations from a 6. UNOR Infrared Gas Analyzers, Bendix mean output not caused by input concentra- Corp., Ronceverte, WV tion changes.
779
VerDate 11
Linearity—The maximum deviation be- 1.3.2 Sensitivity. The detection limit is 3 tween an actual instrument reading and the ppm based on three times the standard devi- reading predicted by a straight line drawn ation of the mean reagent blank values. between upper and lower calibration points. 1.4 Interferences. Sulfur oxides, nitric oxide, and other acid gases interfere with the col- METHOD 10A—DETERMINATION OF CARBON orimetric reaction. They are removed by MONOXIDE EMISSIONS IN CERTIFYING CONTIN- passing the sampled gas through an alkaline UOUS EMISSION MONITORING SYSTEMS AT PE- potassium permanganate scrubbing solution. TROLEUM REFINERIES Carbon dioxide (CO2) does not interfere, but, because it is removed by the scrubbing solu- 1. Applicability and Principle tion, its concentration must be measured independently and an appropriate volume 1.1 Applicability. This method applies to the correction made to the sampled gas. measurement of carbon monoxide (CO) at pe- 1.5 Precision, Accuracy, and Stability. troleum refineries. This method serves as the 1.5.1 Precision. The estimated intralabora- reference method in the relative accuracy tory standard deviation of the method is 3 test for nondispersive infrared (NDIR) CO percent of the mean for gas samples analyzed continuous emission monitoring systems in duplicate in the concentration range of 39 (CEMS’s) that are required to be installed in to 412 ppm. The interlaboratory precision petroleum refineries on fluid catalytic has not been established. cracking unit catalyst regenerators [40 CFR 1.5.2 Accuracy. The method contains no sig- Part 60.105(a)(2)]. nificant biases when compared to an NDIR 1.2 Principle. An integrated gas sample is analyzer calibrated with National Bureau of extracted from the stack, passed through an Standards (NBS) standards. alkaline permanganate solution to remove 1.5.3 Stability. The individual components sulfur and nitrogen oxides, and collected in a of the colorimetric reagent are stable for at Tedlar bag. The CO concentration in the least 1 month. The colorimetric reagent sample is measured spectrophotometrically must be used within 2 days after preparation using the reaction of CO with p-sulfa- to avoid excessive blank correction. The minobenzoic acid. samples in the Tedlar 1 bag should be stable 1.3 Range and Sensitivity. for at least 1 week if the bags are leak-free. 1.3.1 Range. Approximately 3 to 1800 ppm CO. Samples having concentrations below 400 2. Apparatus ppm are analyzed at 425 nm, and samples 2.1 Sampling. The sampling train is shown having concentrations above 400 ppm are in Figure 10A–1, and component parts are analyzed at 600 nm. discussed below:
1 Mention of trade names or commercial for use by the Environmental Protection products in this publication does not con- Agency. stitute the endorsement or recommendation
780
VerDate 11
2.1.1 Probe. Stainless steel, sheathed Pyrex 2.1.8 CO2 Analyzer. Method 3 or its ap- glass, or equivalent, equipped with a glass proved alternative to measure CO2 con- wool plug to remove particulate matter. centration to within 0.5 percent. 2.1.2 Sample Conditioning System. Three 2.1.9 Volume Meter. Dry gas meter, cali- Greenburg-Smith impingers connected in se- brated and capable of measuring the sample ries with leak-free connections. volume under rotameter calibration condi- 2.1.3 Pump. Leak-free pump with stainless tions of 300 ml/min for 10 minutes. steel and Teflon parts to transport sample at 2.1.10 Pressure Gauge. A water filled U-tube a flow rate of 300 ml/min to the flexible bag. manometer, or equivalent, of about 28 cm (12 2.1.4 Surge Tank. Installed between the in.) to leak-check the flexible bag. pump and the rate meter to eliminate the 2.2 Analysis. pulsation effect of the pump on the rate Spectrophotometer. Single- or double- meter. 2.2.1 2.1.5 Rate Meter. Rotameter, or equivalent, beam to measure absorbance at 425 and 600 to measure flow rate at 300 ml/min. Calibrate nm. Slit width should not exceed 20 nm. according to Section 5.2. 2.2.2 Spectrophotometer Cells. 1-cm 2.1.6 Flexible Bag. Tedlar, or equivalent, pathlength. with a capacity of 10 liters and equipped with 2.2.3 Vacuum Gauge. U-tube mercury ma- a sealing quick-connect plug. The bag must nometer, 1 meter (39 in.), with 1-mm divi- be leak-free according to Section 4.1. For sions, or other gauge capable of measuring protection, it is recommened that the bag be pressure to within 1 mm Hg. enclosed with a rigid container. 2.2.4 Pump. Capable of evacuating the gas 2.1.7 Valves. Stainless-steel needle valve to reaction bulb to a pressure equal to or less adjust flow rate, and stainless-steel three- than 40 mm Hg absolute, equipped with way valve, or equivalent. coarse and fine flow control valves.
781
VerDate 11
2.2.5 Barometer. Mercury, aneroid, or other free at 40 mm Hg, designed so that 10 ml of barometer capable of measuring atmospheric the colorimetric reagent can be added and pressure to within 1 mm Hg. removed easily and accurately. Commer- 2.2.6 Reaction Bulbs. Pyrex glass, 100.ml cially available gas sample bulbs such as with Teflon stopcock (Figure 10A-2), leak- Supelco Catalog No. 2-2161 may also be used.
782
VerDate 11
2.2.7 Manifold. Stainless steel, with con- 2.2.8 Pipets. Class A, 10-ml size. nections for three reaction bulbs and the ap- 2.2.9 Shaker Table. Reciprocating-stroke propriate connections for the manometer type such as Eberbach Corporation, Model and sampling bag as shown in Figure 10A-3. 6015. A rocking arm or rotary-motion type
783
VerDate 11
shaker may also be used. The shaker must be for most commercial shakers to provide suf- large enough to accommodate at least six ficient space for the needed bulbs (Figure gas sample bulbs simultaneously. It may be 10A–4). necessary to construct a table top extension 2.2.10 Valve. Stainless steel shut-off valve.
2.2.11 Analytical Balance. Capable of accu- 3.1.2 Alkaline Permanganate Solution, 0.25 rately weighing to 0.1 mg. M KMn04/1.5 M NaOH. Dissolve 40 g KMn04 and 60 g NaOH in water, and dilute to 1 liter. 3. Reagents 3.2 Analysis. 3.2.1 Water. Same as in Section 3.1.1. Unless otherwise indicated, all reagents 3.2.2 1 M Sodium Hydroxide (NaOH) Solu- shall conform to the specifications estab- tion. Dissolve 40 g NaOH in approximately 900 lished by the Committee on Analytical Re- ml of water, cool, and dilute to 1 liter. agents of the American Chemical Society, 3.2.3 0.1 M Silver Nitrate (AgNO3) Solution. where such specifications are available, oth- Dissolve 8.5 g AgNO3 in water, and dilute to erwise, the best available grade shall be 500 ml. used. 3.2.4 0.1 M Para-Sulfaminobenzoic Acid (p- 3.1 Sampling. SABA) Solution. Dissolve 10.0 g p-SABA in 0.1 3.1.1 Water. Deionized distilled, as de- M NaOH (prepared by diluting 50 ml of 1 M scribed in Method 6, Section 3.1.1. NaOH to 500 ml), and dilute to 500 ml with 0.1 M NaOH.
784
VerDate 11
3.2.5 Colorimetric Solution. To a flask, add sample collection. The bag should be leak- 100 ml of p-SABA solution and 100 ml of checked in the inflated and deflated condi- AgNO3 solution. Mix, and add 50 ml of 1 M tion according to the following procedures. NaOH with shaking. The resultant solution Connect the bag to a water manometer, should be clear and colorless. This solution and pressurize the bag to 5 to 10 cm H20 (2 to is acceptable for use for a period of 2 days. 4 in. H20). Allow the bag to stand for 60 min- 3.2.6 Standard Gas Mixtures. Traceable to utes. Any displacement in the water manom- NBS standards and containing between 50 eter indicates a leak. Now, evacuate the bag and 1000 ppm CO in nitrogen. At least two with a leakless pump that is connected on concentrations are needed to span each cali- the downstream side of a flow-indicating de- bration range used (Section 5.3). vice such as a 0-to 100-ml/min rotameter or The calibration gases shall be certified by an impinger containing water. When the bag the manufacturer to be within 2 percent of is completely evacuated, no flow should be the specified concentrations. evident if the bag is leak-free. 4.2 Sampling. Evacuate the Tedlar bag 4. Procedure completely using a vacuum pump. Assemble 4.1 Sample Bag Leak-checks. While a bag the apparatus as shown in Figure 10A-1. leak-check is required after bag use, it Loosely pack glass wool in the tip of the should also be done before the bag is used for probe. Place 400 ml of alkaline permanganate
785
VerDate 11
solution in the first two impingers and 250 each is less than 1,000 ppm and the CO2 con- ml in the third. Connect the pump to the centration is less than 15 percent. Replace third impinger, and follow this with the the scrubber solution after every fifth sam- surge tank, rate meter, and three-way valve. ple. Do not connect the Tedlar bag to the system 4.3 Carbon Dioxide Measurement. Measure at this time. the CO2 content in the stack to the nearest Leak-check the sampling system by plac- 0.5 percent each time a CO sample is col- ing a vacuum gauge at or near the probe lected. A simultaneous grab sample analyzed inlet, plugging the probe inlet, opening the by the Fyrite analyzer is acceptable. three-way valve, and pulling a vacuum of ap- 4.4 Analysis. Assemble the system shown proximately 250 mm Hg on the system while in Figure 10A–3, and record the information observing the rate meter for flow. If flow is indicated on the rate meter, do not proceed required in Table 10A-1 as it is obtained. further until the leak is found and corrected. Pipet 10.0 ml of the colorimetric reagent into Purge the system with sample gas by in- each gas reaction bulb, and attach the bulbs serting the probe into the stack and drawing to the system. Open the stopcocks to the re- sample through the system at 300 ml/min ±10 action bulbs, but leave the valve to the percent for 5 minutes. Connect the evacuated Tedlar bag closed. Turn on the pump, fully Tedlar bag to the system, record the starting open the coarse-adjust flow valve, and slowly time, and sample at a rate of 300 ml/min for open the fine-adjust valve until the pressure 30 minutes, or until the Tedlar bag is nearly is reduced to at least 40 mm Hg. Now close full. Record the sampling time, the baro- the coarse adjust valve, and observe the ma- metric pressure, and the ambient tempera- nometer to be certain that the system is ture. Purge the system as described above leak-free. Wait a minimum of 2 minutes. If immediately before each sample. the pressure has increased less than 1 mm, The scrubbing solution is adequate for re- proceed as described below. If a leak is moving sulfur and nitrogen oxides from 50 li- present, find and correct it before proceeding ters of stack gas when the concentration of further.
Record the vacuum pressure (Pv) to the Close the bulb stopcocks, remove the bulbs, nearest 1 mm Hg, and close the reaction bulb record the room temperature and barametric stopcocks. Open the Tedlar bag valve, and pressure (Pbar, to nearest mm Hg), and place allow the system to come to atmospheric the bulbs on the shaker table with their pressure. Close the bag valve, open the pump main axis either parallel to or perpendicular coarse adjust valve, and evacuate the system to the plane of the table top. Purge the bulb- again. Repeat this fill/evacuation procedure filling system with ambient air for several at least twice to flush the manifold com- minutes between samples. Shake the sam- pletely. Close the pump coarse adjust valve, ples for exactly 2 hours. open the Tedlar bag valve, and let the sys- Immediately after shaking, measure the tem fill to atmospheric pressure. Open the absorbance (A) of each bulb sample at 425 nm stopcocks to the reaction bulbs, and let the if the concentration is less than or equal to entire system come to atmospheric pressure. 400 ppm CO or at 600 nm if the concentration
786
VerDate 11
is above 400 ppm. This may be accomplished ppm. Draw a smooth curve through the with multiple bulb sets by sequentially col- points. The curve should be linear over the lecting sets and adding to the shaker at stag- two concentration ranges discussed in Sec- gered intervals, followed by sequentially re- tion 1.3.1. moving sets from the shaker for absorbance measurement after the two-hour designated 6. Calculations intervals have elapsed. Carry out calculations retaining at least Use a small portion of the sample to rinse one extra decimal figure beyond that of the a spectrophotometer cell several times be- acquired data. Round off figures after final fore taking an aliquot for analysis. If one calculation. cell is used to analyze multiple samples, rinse the cell several times between samples 6.1 Nomenclature. with water. Prepare and analyze standards and a rea- A=Sample absorbance, uncorrected for the gent blank as described in Section 5.3. Use reagent blank.
water as the reference. Reject the analysis if Ar=Absorbance of the reagent blank. the blank absorbance is greater than 0.1. All As=Average sample absorbance per liter, conditions should be the same for analysis of units/liter. samples and standards. Measure the Bw=Moisture content in the bag sample. absorbances as soon as possible after shaking C=CO concentration in the stack gas, dry is completed. Determine the CO concentra- basis, ppm. tion of each bag sample using the calibration C =CO concentration of the bag sample, dry curve for the appropriate concentration b basis, ppm. range as discussed in Section 5.3. Cg=CO concentration from the calibration 5. Calibration curve, ppm. F=Volume fraction of CO in the stack. 5.1 Bulb Calibration. Weigh the empty bulb 2 n=Number of reaction bulbs used per bag to the nearest 0.1 g. Fill the bulb to the stop- sample. cock with water, and again weigh to the nearest 0.1 g. Subtract the tare weight, and Pbar=Barometric pressure, mm Hg. calculate the volume in liters to three sig- Pv=Residual pressure in the sample bulb nificant figures using the density of water after evacuation, mm Hg. (at the measurement temperature). Record Pw=Vapor pressure of H20 in the bag (from the volume on the bulb; alternatively, mark Table 10A–2), mm Hg. an identification number on the bulb, and Vb=Volume of the sample bulb, liters. record the volume in a notebook. Vr=Volume of reagent added to the sample 5.2 Rate Meter Calibration. Assemble the bulb, 0.0100 liter. system as shown in Figure 10A–1 (the impingers may be removed), and attach a 6.2 Average Sample Absorbance per Liter. volume meter to the probe inlet. Set the ro- Average the three absorbance values for tameter at 300 ml/min, record the volume each bulb set. Then calculate A for each set meter reading, start the pump, and pull gas s of gas bulbs using Equation 10A–1. Use As to through the system for 10 minutes. Record determine the CO concentration from the the final volume meter reading. Repeat the calibration curve (C ). procedure and average the results to deter- g mine the volume of gas that passed through ()AAP− () the system. = r bar 5.3 Spectrophotometer Calibration Curve. A s Eq.10 A- 1 The calibration curve is established by tak- − − ()VVPPb r() bar v ing at least two sets of three bulbs of known CO collected from Tedlar bags through the NOTE: A and Ar must be at the same wave- analysis procedure. Reject the standard set length. where any of the individual bulb absorbances differ from the set mean by more than 10 per- 6.3 CO Concentration in the Bag. cent. Collect the standards as described in Calculate C using Equations 10A–2 and Section 4.2. Prepare standards to span the 0- b 10A–3. If condensate is visible in the Tedlar to 400- or 400- to 1000-ppm range. If any sam- bag, calculate B using Table 10A–2 and the ples span both concentration ranges, prepare w temperature and barometric pressure in the a calibration curve for each range. A set of analysis room. If condensate is not visible, three bulbs containing colorimetric reagent calculate B using the temperature and baro- but no CO should serve as a reagent blank w and be taken through the analysis procedure. metric pressure recorded at the sampling Calculate the average absorbance for each site. set (3 bulbs) of standards using Equation 10A–1 and Table 10A–1. Construct a graph of = Pw average absorbance for each standard Bw Eq.10 A- 2 against its corresponding concentration in Pbar
787
VerDate 11
6.4 CO Concentration in the Stack. C C = g Eq.10 A- 3 = − b − C Cb ().1 F Eq 10 A- 4 ()1 Bw
7. BIBLIOGRAPHY 5. Repp, M. Evaluation of Continuous Mon- itors for Carbon Monoxide in Stationary 1. Butler, F.E., J.E. Knoll, and M.R. Sources. U.S. Environmental Protection Midgett. Development and Evaluation of Agency. Research Triangle Park, NC. Publi- Methods for Determining Carbon Monoxide cation No. EPA–600/2–77–063. March 1977. 155 Emissions. Quality Assurance Division, En- p. vironmental Monitoring Systems Labora- 6. Smith, F., D.E. Wagoner, and R.P. Dono- tory, U.S. Environmental Protection Agen- cy, Research Triangle Park, NC 27711. June van. Guidelines for Development of a Quality 1985. 33 p. Assurance Program: Volume VIII—Deter- mination of CO Emissions from Stationary 2. Ferguson, B.B., R.E. Lester, and W.J. Sources by NDIR Spectrometry. U.S. Envi- Mitchell. Field Evaluation of Carbon Mon- ronmental Protection Agency. Research Tri- oxide and Hydrogen Sulfide Continuous Emission Monitors at an Oil Refinery. U.S. angle Park, NC. Publication No. EPA–650/4– Environmental Protection Agency. Research 74–005–h. February 1975. 96 p. Triangle Park, NC. Publication No. EPA–600/ METHOD 10B—DETERMINATION OF CARBON 4–82–054. August 1982. 100 p. MONOXIDE EMISSIONS FROM STATIONARY 3. Lambert, J.L., and R.E. Weins. Induced SOURCES Colorimetric Method for Carbon Monoxide. Analytical Chemistry. 46(7):929–930. June 1. Applicability and Principle 1974. 4. Levaggi, D.A., and M. Feldstein. The 1.1 Applicability. This method applies to Colorimetric Determination of Low Con- the measurement of carbon monoxide (CO) centrations of Carbon Monoxide. Industrial emissions at petroleum refineries and from Hygiene Journal. 25:64–66. January-February other sources when specified in an applicable 1964. subpart of the regulations.
788
VerDate 11
1.2 Principle. An integrated gas sample is 3.2.1 Carrier, Fuel, and Combustion Gases. extracted from the sampling point and ana- Same as in Method 25, sections 3.2.1, 3.2.2, lyzed for CO. The sample is passed through a and 3.2.3. conditioning system to remove interferences 3.2.2 Linearity and Calibration Gases. and collected in a Tedlar bag. The CO is sep- Three standard gases with nominal CO con- arated from the sample by gas chroma- centrations of 20–, 200–, and 1,000–ppm CO in tography (GC) and catalytically reduced to nitrogen. methane (CH4) prior to analysis by flame 3.2.3 Reduction Catalyst Efficiency Check ionization detection FID. The analytical por- Calibration Gas. Standard CH4 gas with a tion of this method is identical to applicable concentration of 1,000 ppm in air. sections in Method 25 detailing CO measure- ment. The oxidation catalyst required in 4. Procedure Method 25 is not needed for sample analysis. 4.1 Sample Bag Leak-checks, Sampling, Complete Method 25 analytical systems are and CO2 Measurement. Same as in Method acceptable alternatives when calibrated for 10A, sections 4.1, 4.2, and 4.3. CO and operated by the Method 25 analytical 4.2 Preparation for Analysis. Before put- procedures. ting the GC analyzer into routine operation, NOTE: Mention of trade names or commer- conduct the calibration procedures listed in cial products in this method does not con- section 5. Establish an appropriate carrier stitute the endorsement or recommendation flow rate and detector temperature for the for use by the Environmental Protection specific instrument used. Agency. 4.3 Sample Analysis. Purge the sample 1.3 Interferences. Carbon dioxide (CO2) loop with sample, and then inject the sam- and organics potentially can interfere with ple. Analyze each sample in triplicate, and the analysis. Carbon dioxide is primarily re- calculate the average sample area (A). Deter- moved from the sample by the alkaline per- mine the bag CO concentration according to manganate conditioning system; any resid- section 6.2. ual CO2 and organics are separated from the CO by GC. 5. Calibration
2. Apparatus 5.1 Carrier Gas Blank Check. Analyze each new tank of carrier gas with the GC an- 2.1 Sampling. Same as in Method 10A, sec- alyzer according to section 4.3 to check for tion 2.1. contamination. The corresponding con- 2.2 Analysis. centration must be less than 5 ppm for the 2.2.1 Gas Chromatographic (GC) Analyzer. tank to be acceptable for use. A semicontinuous GC/FID analyzer capable 5.2 Reduction Catalyst Efficiency Check. of quantifying CO in the sample and contain- Prior to initial use, the reduction catalyst ing at least the following major components. shall be tested for reduction efficiency. With 2.2.1.1 Separation Column. A column that the heated reduction catalyst bypassed, make triplicate injections of the 1,000-ppm separates CO from CO2 and organic com- CH gas (section 3.2.3) to calibrate the ana- pounds that may be present. A 1⁄8-in. OD 4 stainless-steel column packed with 5.5 ft of lyzer. Repeat the procedure using 1,000-ppm 60/80 mesh Carbosieve S–II (available from CO (section 3.2.2) with the catalyst in oper- Supelco) has been used successfully for this ation. The reduction catalyst operation is purpose. The column listed in Addendum 1 of acceptable if the CO response is within 5 per- Method 25 is also acceptable. cent of the certified gas value. 2.2.1.2 Reduction Catalyst. Same as in 5.3 Analyzer Linearity Check and Calibra- Method 25, section 2.3.2. tion. Perform this test before the system is first placed into operation. With the reduc- 2.2.1.3 Sample Injection System. Same as tion catalyst in operation, conduct a linear- in Method 25, section 2.3.4, equipped to ac- ity check of the analyzer using the standards cept a sample line from the Tedlar bag. specified in section 3.2.2. Make triplicate in- 2.2.1.4 Flame Ionization Detector. Linear- jections of each calibration gas, and then ity meeting the specifications in section calculate the average response factor (area/ 2.3.5.1 of Method 25 where the linearity check ppm) for each gas, as well as the overall is carried out using standard gases contain- mean of the response factor values. The in- ing 20–, 200–, and 1,000–ppm CO. The minimal strument linearity is acceptable if the aver- instrument range shall span 10 to 1,000 ppm age response factor of each calibration gas is CO. within 2.5 percent of the overall mean value 2.2.1.5 Data Recording System. Same as in and if the relative standard deviation (cal- Method 25, section 2.3.6. culated in section 6.9 of Method 25) for each 3. Reagents set of triplicate injections is less than 2 per- 3.1 Sampling. Same as in Method 10A, sec- cent. Record the overall mean of the re- tion 3.1. sponse factor values as the calibration re- 3.2 Analysis. sponse factor (R).
789
VerDate 11
6. Calculations Control Association. Denver, CO. June 9, 1974.) 25 p. Carry out calculations retaining at least one extra decimal figure beyond that of the METHOD 11—DETERMINATION OF HYDROGEN acquired data. Round off results only after SULFIDE CONTENT OF FUEL GAS STREAMS IN the final calculation. PETROLEUM REFINERIES 6.1 Nomenclature. 1. Principle and Applicability A=Average sample area. 1.1 Principle. Hydrogen sulfide (H2S) is col- Bw=Moisture content in the bag sample, fraction. lected from a source in a series of midget C=CO concentration in the stack gas, dry impingers and absorbed in pH 3.0 cadmium basis, ppm. sulfate (CdSO4) solution to form cadmium sulfide (CdS). The latter compound is then C =CO concentration in the bag sample, b measured iodometrically. An impinger con- dry basis, ppm. taining hydrogen peroxide is included to re- F=Volume fraction of CO2 in the stack, move SO2 as an interfering species. This fraction. method is a revision of the H S method origi- P =Barometric pressure, mm Hg. 2 bar nally published in the FEDERAL REGISTER, Pw=Vapor pressure H2O in the bag (from Volume 39, No. 47, dated Friday, March 8, Table 10–2, Method 10A), mm Hg. 1974. R=Mean calibration response factor, area/ 1.2 Applicability. This method is applica- ppm. ble for the determination of the hydrogen 6.2 CO Concentration in the Bag. Cal- sulfide content of fuel gas streams at petro- culate Cb using Equations 10B–1 and 10B–2. If leum refineries. condensate is visible in the Tedlar bag, cal- 2. Range and Sensitivity culate Bw using Table 10A–1 of Method 10A and the temperature and barometric pres- The lower limit of detection is approxi- 3 sure in the analysis room. If condensate is mately 8 mg/m (6 ppm). The maximum of the range is 740 mg/m3 (520 ppm). not visible, calculate Bw using the tempera- ture and barometric pressure at the sam- 3. Interferences pling site. Any compound that reduces iodine or oxi- dizes iodide ion will interfere in this proce- = Pw dure, provided it is collected in the cadmium Bw Eq.10 B- 1 sulfate impingers. Sulfur dioxide in con- Pbar centrations of up to 2,600 mg/m3 is elimi- nated by the hydrogen peroxide solution. A Thiols precipitate with hydrogen sulfide. In C = Eq.10 B- 2 the absence of H2S, only co-traces of thiols b ()− are collected. When methane- and ethane- RB1 w thiols at a total level of 300 mg/m3 are 6.3 CO Concentration in the Stack. present in addition to H2S, the results vary from 2 percent low at an H2S concentration C=C (1¥F) 3 b of 400 mg/m to 14 percent high at an H2S Eq. 10B–3 concentration of 100 mg/m3. Carbon oxysul- fide at a concentration of 20 percent does not 7. Bibliography interfere. Certain carbonyl-containing com- 1. Butler, F.E, J.E. Knoll, and M.R. pounds react with iodine and produce recur- Midgett. Development and Evaluation of ring end points. However, acetaldehyde and Methods for Determining Carbon Monoxide acetone at concentrations of 1 and 3 percent, Emissions. Quality Assurance Division, En- respectively, do not interfere. vironmental Monitoring Systems Labora- Entrained hydrogen peroxide produces a tory, U.S. Environmental Protection Agen- negative interference equivalent to 100 per- cy, Research Triangle Park, NC 27711. June cent of that of an equimolar quantity of hy- 1985. 33p. drogen sulfide. Avoid the ejection of hydro- 2. Salo, A.E., S. Witz, and R.D. MacPhee. gen peroxide into the cadmium sulfate Determination of Solvent Vapor Concentra- impingers. tions by Total Combustion Analysis: A Com- 4. Precision and Accuracy parison of Infrared with Flame Ionization Collaborative testing has shown the with- Detectors. Paper No. 75–33.2. (Presented at in-laboratory coefficient of variation to be the 68th Annual Meeting of the Air Pollution 2.2 percent and the overall coefficient of var- Control Association. Boston, MA. June 15, iation to be 5 percent. The method bias was 1975.) 14 p. shown to be ¥4.8 percent when only H2S was 3. Salo, A.E., W.L. Oaks, and R.D. present. In the presence of the interferences MacPhee. Measuring the Organic Carbon cited in section 3, the bias was positive at Content of Source Emissions for Air Pollu- low H2S concentration and negative at high- 3 tion Control. Paper No. 74–190. (Presented at er concentrations. At 230 mg H2S/m , the the 67th Annual Meeting of the Air Pollution level of the compliance standard, the bias
790
VerDate 11
was +2.7 percent. Thiols had no effect on the (0.1 in. Hg) per 30 m (100 ft) elevation in- precision. crease or vice-versa for elevation decrease. 5. Apparatus 5.1.11 U-tube Manometer. 0–30 cm water column. For leak check procedure. 5.1 Sampling Apparatus. 5.1.12 Rubber Squeeze Bulb. To pressurize 5.1.1 Sampling Line. Six to 7 mm (1⁄4 in.) train for leak check. Teflon 1 tubing to connect the sampling train 5.1.13 Tee, Pinchclamp, and Connecting to the sampling valve. Tubing. For leak check. 5.1.2 Impingers. Five midget impingers, 5.1.14 Pump. Diaphragm pump, or equiva- each with 30 ml capacity. The internal diam- lent. Insert a small surge tank between the eter of the impinger tip must be 1 mm ±0.05 pump and rate meter to eliminate the pulsa- mm. The impinger tip must be positioned 4 tion effect of the diaphragm pump on the ro- to 6 mm from the bottom of the impinger. tameter. The pump is used for the air purge 5.1.3 Tubing. Glass or Teflon connecting at the end of the sample run; the pump is not tubing for the impingers. ordinarily used during sampling, because 5.1.4 Ice Bath Container. To maintain ab- fuel gas streams are usually sufficiently sorbing solution at a low temperature. pressurized to force sample gas through the 5.1.5 Drying Tube. Tube packed with 6- to train at the required flow rate. The pump 16-mesh indicating-type silica gel, or equiva- need not be leak-free unless it is used for lent, to dry the gas sample and protect the sampling. meter and pump. If the silica gel has been 5.1.15 Needle Valve or Critical Orifice. To used previously, dry at 175°C (350°F) for 2 set air purge flow to 1 liter/min. hours. New silica gel may be used as re- 5.1.16 Tube Packed With Active Carbon. ceived. Alternatively, other types of To filter air during purge. desiccants (equivalent or better) may be 5.1.17 Volumetric Flask. One 1,000 ml. used, subject to approval of the Adminis- 5.1.18 Volumetric Pipette. One 15 ml. trator. 5.1.19 Pressure-Reduction Regulator. De- NOTE: Do not use more than 30 g of silica pending on the sampling stream pressure, a gel. Silica gel absorbs gases such as propane pressure-reduction regulator may be needed from the fuel gas stream, and use of exces- to reduce the pressure of the gas stream en- sive amounts of silica gel could result in er- tering the Teflon sample line to a safe level. rors in the determination of sample volume. 5.1.20 Cold Trap. If condensed water or 5.1.6 Sampling Valve. Needle valve or amine is present in the sample stream, a cor- equivalent to adjust gas flow rate. Stainless rosion-resistant cold trap shall be used im- steel or other corrosion-resistant material. mediately after the sample tap. The trap 5.1.7 Volume Meter. Dry gas meter, suffi- shall not be operated below 0°C (32°F) to ciently accurate to measure the sample vol- avoid condensation of C3 or C4 hydrocarbons. ume within 2 percent, calibrated at the se- 5.2 Sample Recovery. lected flow rate (∼1.0 liter/min) and condi- 5.2.1 Sample Container. Iodine flask, tions actually encountered during sampling. glass-stoppered: 500 ml size. The meter shall be equipped with a tempera- 5.2.2 Pipette. 50 ml volumetric type. ture gauge (dial thermometer or equivalent) 5.2.3 Graduated Cylinders. One each 25 capable of measuring temperature to within and 250 ml. 3°C (5.4°F). The gas meter should have a pet- 5.2.4 Flasks. 125 ml, Erlenmeyer. cock, or equivalent, on the outlet connector 5.2.5 Wash Bottle. which can be closed during the leak check. 5.2.6 Volumetric Flasks. Three 1,000 ml. Gas volume for one revolution of the meter 5.3 Analysis. must not be more than 10 liters. 5.3.1 Flask. 500 ml glass-stoppered iodine 5.1.8 Flow Meter. Rotameter or equiva- flask. lent, to measure flow rates in the range from 5.3.2 Burette. 50 ml. 0.5 to 2 liters/min (1 to 4 cfh). 5.3.3 Flask. 125 ml, Erlenmeyer. 5.1.9 Graduated Cylinder, 25 ml size. 5.3.4 Pipettes, Volumetric. One 25 ml; two 5.1.10 Barometer. Mercury, aneroid, or each 50 and 100 ml. other barometer capable of measuring at- 5.3.5 Volumetric Flasks. One 1,000 ml; two mospheric pressure to within 2.5 mm Hg (0.1 500 ml. in. Hg). In many cases, the barometric read- 5.3.6 Graduated Cylinders. One each 10 ing may be obtained from a nearby National and 100 ml. Weather Service station, in which case, the 6. Reagents station value (which is the absolute baro- Unless otherwise indicated, it is intended metric pressure) shall be requested and an that all reagents conform to the specifica- adjustment for elevation differences between tions established by the Committee on Ana- the weather station and the sampling point lytical Reagents of the American Chemical shall be applied at a rate of minus 2.5 mm Hg Society, where such specifications are avail- able. Otherwise, use best available grade. 1 Mention of trade names or specific prod- 6.1 Sampling. ucts does not constitute endorsement by the 6.1.1 Cadmium Sulfate Absorbing Solu- Environmental Protection Agency. tion. Dissolve 41 g of 3CdSO4•8H2 O and 15 ml
791
VerDate 11
of 0.1 M sulfuric acid in a 1-liter volumetric droxide. After settling, decant 140 ml of this flask that contains approximately 3⁄4 liter of solution into 800 ml of distilled water. Bring deionized distilled water. Dilute to volume the solution to pH 6–7 with 6N hydrochloric with deionized water. Mix thoroughly. pH acid and dilute to 1 liter. Standardize as in should be 3±0.1. Add 10 drops of Dow-Corning Section 8.1.3. Antifoam B. Shake well before use. If 6.3.4 Starch Indicator Solution. Suspend Antifoam B is not used, the alternative 10 g of soluble starch in 100 ml of deionized, acidified iodine extraction procedure (Sec- distilled water and add 15 g of potassium hy- tion 7.2.2) must be used. droxide (KOH) pellets. Stir until dissolved, 6.1.2 Hydrogen Peroxide, 3 Percent. Dilute dilute with 900 ml of deionized distilled 30 percent hydrogen peroxide to 3 percent as water and let stand for 1 hour. Neutralize the needed. Prepare fresh daily. alkali with concentrated hydrochloric acid, 6.1.3 Water. Deionized, distilled to con- using an indicator paper similar to Alkacid form to ASTM specifications D1193–72, Type test ribbon, then add 2 ml of glacial acetic 3. At the option of the analyst, the KMnO4 acid as a preservative. test for oxidizable organic matter may be omitted when high concentrations of organic NOTE: Test starch indicator solution for de- matter are not expected to be present. composition by titrating, with 0.01 N iodine 6.2 Sample Recovery. solution, 4 ml of starch solution in 200 ml of 6.2.1 Hydrochloric Acid Solution (HCl), distilled water that contains 1 g potassium 3M. Add 240 ml of concentrated HCl (specific iodide. If more than 4 drops of the 0.01 N io- gravity 1.19) to 500 ml of deionized, distilled dine solution are required to obtain the blue water in a 1-liter volumetric flask. Dilute to color, a fresh solution must be prepared. 1 liter with deionized water. Mix thoroughly. 7. Procedure 6.2.2 Iodine Solution 0.1 N. Dissolve 24 g of 7.1 Sampling. potassium iodide (KI) in 30 ml of deionized, 7.1.1 Assemble the sampling train as distilled water. Add 12.7 g of resublimed io- shown in Figure 11–1, connecting the five dine (I2) to the potassium iodide solution. midget impingers in series. Place 15 ml of 3 Shake the mixture until the iodine is com- percent hydrogen peroxide solution in the pletely dissolved. If possible, let the solution first impinger. Leave the second impinger stand overnight in the dark. Slowly dilute empty. Place 15 ml of the cadmium sulfate the solution to 1 liter with deionized, dis- absorbing solution in the third, fourth, and tilled water, with swirling. Filter the solu- fifth impingers. Place the impinger assembly tion if it is cloudy. Store solution in a in an ice bath container and place crushed brown-glass reagent bottle. ice around the impingers. Add more ice dur- 6.2.3 Standard Iodine Solution, 0.01 N. Pi- ing the run, if needed. pette 100.0 ml of the 0.1 N iodine solution 7.1.2 Connect the rubber bulb and manom- into a 1-liter volumetric flask and dilute to eter to first impinger, as shown in Figure 11– volume with deionized, distilled water. 1. Close the petcock on the dry gas meter Standardize daily as in Section 8.1.1. This so- outlet. Pressurize the train to 25-cm water lution must be protected from light. Reagent pressure with the bulb and close off tubing bottles and flasks must be kept tightly connected to rubber bulb. The train must stoppered. hold a 25-cm water pressure with not more 6.3 Analysis. than a 1-cm drop in pressure in a 1-minute 6.3.1 Sodium Thiosulfate Solution, Stand- interval. Stopcock grease is acceptable for ard 0.1 N. Dissolve 24.8 g of sodium sealing ground glass joints. thiosulfate pentahydrate (Na2S2O3•5H2O) or 15.8 g of anhydrous sodium thiosulfate NOTE: This leak check procedure is op- (Na2S2O3) in 1 liter of deionized, distilled tional at the beginning of the sample run, water and add 0.01 g of anhydrous sodium but is mandatory at the conclusion. Note carbonate (Na2CO3) and 0.4 ml of chloroform also that if the pump is used for sampling, it (CHCl3) to stabilize. Mix thoroughly by shak- is recommended (but not required) that the ing or by aerating with nitrogen for approxi- pump be leak-checked separately, using a mately 15 minutes and store in a glass- method consistent with the leak-check pro- stoppered, reagent bottle. Standardize as in cedure for diaphragm pumps outlined in Sec- Section 8.1.2. tion 4.1.2 of Method 6, 40 CFR part 60, appen- 6.3.2 Sodium Thiosulfate Solution, Stand- dix A. ard 0.01 N. Pipette 50.0 ml of the standard 0.1 7.1.3 Purge the connecting line between N thiosulfate solution into a volumetric the sampling valve and first impinger, by flask and dilute to 500 ml with distilled disconnecting the line from the first im- water. pinger, opening the sampling valve, and al- NOTE: A 0.01 N phenylarsine oxide solution lowing process gas to flow through the line may be prepared instead of 0.01 N thiosulfate for a minute or two. Then, close the sam- (see Section 6.3.3). pling valve and reconnect the line to the im- 6.3.3 Phenylarsine Oxide Solution, Stand- pinger train. Open the petcock on the dry ard 0.01 N. Dissolve 1.80 g of phenylarsine gas meter outlet. Record the initial dry gas oxide (C6H5AsO) in 150 ml of 0.3 N sodium hy- meter reading. 792
VerDate 11
7.1.4 Open the sampling valve and then For sample recovery, cap the open ends and adjust the valve to obtain a rate of approxi- remove the impinger train to a clean area mately 1 liter/min. Maintain a constant (±10 that is away from sources of heat. The area percent) flow rate during the test. Record should be well lighted, but not exposed to di- the meter temperature. rect sunlight. 7.1.5 Sample for at least 10 min. At the 7.2 Sample Recovery. end of the sampling time, close the sampling 7.2.1 Discard the contents of the hydrogen valve and record the final volume and tem- peroxide impinger. Carefully rinse the con- perature readings. Conduct a leak check as tents of the third, fourth, and fifth impingers described in Section 7.1.2 above. into a 500 ml iodine flask. 7.1.6 Disconnect the impinger train from the sampling line. Connect the charcoal tube NOTE: The impingers normally have only a and the pump, as shown in Figure 11–1. Purge thin film of cadmium sulfide remaining after the train (at a rate of 1 liter/min) with clean a water rinse. If Antifoam B was not used or ambient air for 15 minutes to ensure that all if significant quantities of yellow cadmium H2S is removed from the hydrogen peroxide.
793
VerDate 11
sulfide remain in the impingers, the alter- 7.3.1 Using 0.01 N sodium thiosulfate solu- native recovery procedure described below tion (or 0.01 N phenylarsine oxide, if applica- must be used. ble), rapidly titrate each sample in an iodine 7.2.2 Pipette exactly 50 ml of 0.01 N iodine flask using gentle mixing, until solution is solution into a 125 ml Erlenmeyer flask. Add light yellow. Add 4 ml of starch indicator so- 10 ml of 3 M HCl to the solution. Quan- lution and continue titrating slowly until titatively rinse the acidified iodine into the the blue color just disappears. Record VTT, iodine flask. Stopper the flask immediately the volume of sodium thiosulfate solution and shake briefly. used, or VAT, the volume of phenylarsine 7.2.2 (Alternative). Extract the remaining oxide solution used (ml). cadmium sulfide from the third, fourth, and 7.3.2 Titrate the blanks in the same man- fifth impingers using the acidified iodine so- ner as the samples. Run blanks each day lution. Immediately after pouring the acidi- until replicate values agree within 0.05 ml. fied iodine into an impinger, stopper it and Average the replicate titration values which shake for a few moments, then transfer the agree within 0.05 ml. liquid to the iodine flask. Do not transfer 8. Calibration and Standards any rinse portion from one impinger to an- other; transfer it directly to the iodine flask. 8.1 Standardizations. Once the acidified iodine solution has been 8.1.1 Standardize the 0.01 N iodine solu- poured into any glassware containing cad- tion daily as follows: Pipette 25 ml of the io- mium sulfide, the container must be tightly dine solution into a 125 ml Erlenmeyer flask. stoppered at all times except when adding Add 2 ml of 3 M HCl. Titrate rapidly with more solution, and this must be done as standard 0.01 N thiosulfate solution or with quickly and carefully as possible. After add- 0.01 N phenylarsine oxide until the solution ing any acidified iodine solution to the io- is light yellow, using gentle mixing. Add four dine flask, allow a few minutes for absorp- drops of starch indicator solution and con- tion of the H2S before adding any further tinue titrating slowly until the blue color rinses. Repeat the iodine extraction until all just disappears. Record VT, the volume of cadmium sulfide is removed from the thiosulfate solution used, or VAS, the volume impingers. Extract that part of the connect- of phenylarsine oxide solution used (ml). Re- ing glassware that contains visible cadmium peat until replicate values agree within 0.05 sulfide. ml. Average the replicate titration values Quantitatively rinse all of the iodine from which agree within 0.05 ml and calculate the the impingers, connectors, and the beaker exact normality of the iodine solution using into the iodine flask using deionized, dis- Equation 11–3. Repeat the standardization tilled water. Stopper the flask and shake daily. briefly. 8.1.2 Standardize the 0.1 N thiosulfate so- 7.2.3 Allow the iodine flask to stand about lution as follows: Oven-dry potassium di- 30 minutes in the dark for absorption of the chromate (K2Cr2O7) at 180 to 200°C (360 to H2S into the iodine, then complete the titra- 390°F). Weigh to the nearest milligram, 2 g of tion analysis as in Section 7.3. potassium dichromate. Transfer the dichro- NOTE: Caution! Iodine evaporates from mate to a 500 ml volumetric flask, dissolve acidified iodine solutions. Samples to which in deionized, distilled water and dilute to ex- acidified iodine have been added may not be actly 500 ml. In a 500 ml iodine flask, dissolve stored, but must be analyzed in the time approximately 3 g of potassium iodide (KI) in schedule stated in Section 7.2.3. 45 ml of deionized, distilled water, then add 10 ml of 3 M hydrochloric acid solution. Pi- 7.2.4 Prepare a blank by adding 45 ml of pette 50 ml of the dichromate solution into cadmium sulfate absorbing solution to an io- this mixture. Gently swirl the solution once dine flask. Pipette exactly 50 ml of 0.01 N io- and allow it to stand in the dark for 5 min- dine solution into a 125-ml Erlenmeyer flask. utes. Dilute the solution with 100 to 200 ml of Add 10 ml of 3 M HCl. Follow the same im- deionized distilled water, washing down the pinger extracting and quantitative rinsing sides of the flask with part of the water. Ti- procedure carried out in sample analysis. trate with 0.1 N thiosulfate until the solu- Stopper the flask, shake briefly, let stand 30 tion is light yellow. Add 4 ml of starch indi- minutes in the dark, and titrate with the cator and continue titrating slowly to a samples. green end point. Record VS, the volume of NOTE: The blank must be handled by ex- thiosulfate solution used (ml). Repeat until actly the same procedure as that used for the replicate analyses agree within 0.05 ml. Cal- samples. culate the normality using Equation 11–1. 7.3 Analysis. Repeat the standardization each week, or NOTE: Titration analyses should be con- after each test series, whichever time is ducted at the sample-cleanup area in order shorter. to prevent loss of iodine from the sample. Ti- 8.1.3 Standardize the 0.01 N Phenylarsine tration should never be made in direct sun- oxide (if applicable) as follows: oven dry po- light. tassium dichromate (K2Cr2O7) at 180 to 200°C
794
VerDate 11
(360 to 390°F). Weigh to the nearest milli- calibration factor (determined in Section gram, 2 g of the K2Cr2O7; transfer the dichro- 8.2.1.1.), then the dry gas meter volumes ob- mate to a 500 ml volumetric flask, dissolve tained during the test series are acceptable. in deionized, distilled water, and dilute to If the calibration factor deviates by more exactly 500 ml. In a 500 ml iodine flask, dis- than 5 percent, recalibrate the dry gas meter solve approximately 0.3 g of potassium iodide as in Section 8.2.1.1, and for the calculations, (KI) in 45 ml of deionized, distilled water; add use the calibration factor (initial or re- 10 ml of 3M hydrochloric acid. Pipette 5 ml of calibration) that yields the lower gas volume the K2Cr2O7 solution into the iodine flask. for each test run. Gently swirl the contents of the flask once 8.2.2 Thermometers. Calibrate against and allow to stand in the dark for 5 minutes. mercury-in-glass thermometers. Dilute the solution with 100 to 200 ml of de- 8.2.3 Rotameter. The rotameter need not ionized, distilled water, washing down the be calibrated, but should be cleaned and sides of the flask with part of the water. Ti- maintained according to the manufacturer’s trate with 0.01 N phenylarsine oxide until instruction. the solution is light yellow. Add 4 ml of 8.2.4 Barometer. Calibrate against a mer- starch indicator and continue titrating slow- cury barometer. ly to a green end point. Record V , the vol- A 9. Calculations ume of phenylarsine oxide used (ml). Repeat until replicate analyses agree within 0.05 ml. Carry out calculations retaining at least Calculate the normality using Equation 11–2. one extra decimal figure beyond that of the Repeat the standardization each week or acquired data. Round off results only after after each test series, whichever time is the final calculation. shorter. 9.1 Normality of the Standard (∼0.1 N) 8.2 Sampling Train Calibration. Calibrate Thiosulfate Solution. the sampling train components as follows: NS=2.039W/VS 8.2.1 Dry Gas Meter. Eq. 11–1 8.2.1.1 Initial Calibration. The dry gas meter shall be calibrated before its initial where: use in the field. Proceed as follows: First, as- W=Weight of K2Cr2O7 used, g. semble the following components in series: VS=Volume of Na2S2O3 solution used, ml. Drying tube, needle valve, pump, rotameter, NS=Normality of standard thiosulfate solu- and dry gas meter. Then, leak-check the sys- tion, g-eq/liter. tem as follows: Place a vacuum gauge (at 2.039=Conversion factor least 760 mm Hg) at the inlet to the drying
tube and pull a vacuum of 250 mm (10 in.) Hg; (6 eq. I2/MOLE K2CR2O7) (1,000 ML/LITER)/ (294.2 G plug or pinch off the outlet of the flow K2CR2O7/MOLE) (10 ALIQUOT FACTOR) meter, and then turn off the pump. The vacu- um shall remain stable for at least 30 sec- 9.2 Normality of Standard Phenylarsine onds. Carefully release the vacuum gauge be- Oxide Solution (if applicable). fore releasing the flow meter end. NA=0.2039 W/VA Next, calibrate the dry gas meter (at the Eq. 11–2 sampling flow rate specified by the method) where: as follows: Connect an appropriately sized wet test meter (e.g., 1 liter per revolution) to W=Weight of K2Cr2O7 used, g. the inlet of the drying tube. Make three VA=Volume of C6H5AsO used, ml. independent calibration runs, using at least NA=Normality of standard phenylarsine five revolutions of the dry gas meter per run. oxide solution, g-eq/liter. Calculate the calibration factor, Y (wet test 0.2039=Conversion factor meter calibration volume divided by the dry (6 eq. I /MOLE K CR O ) (1,000 ML/LITER)/(249.2 G gas meter volume, both volumes adjusted to 2 2 2 7 K CR O /MOLE) (100 ALIQUOT FACTOR) the same reference temperature and pres- 2 2 7 sure), for each run, and average the results. 9.3 Normality of Standard Iodine Solu- If any Y value deviates by more than 2 per- tion. cent from the average, the dry gas meter is unacceptable for use. Otherwise, use the av- NI=NTVT/VI erage as the calibration factor for subse- 11–3 quent test runs. 8.2.1.2 Post-test Calibration Check. After where: each field test series, conduct a calibration NI=Normality of standard iodine solution, g- check as in Section 8.2.1.1. above, except for eq/liter. the following variations: (a) The leak check VI=Volume of standard iodine solution used, is not to be conducted, (b) three or more rev- ml. olutions of the dry gas meter may be used, NT=Normality of standard (∼0.01 N) and (c) only two independent runs need be thiosulfate solution; assumed to be 0.1 made. If the calibration factor does not devi- NS, g-eq/liter. ate by more than 5 percent from the initial VT=Volume of thiosulfate solution used, ml.
795
VerDate 11
NOTE: If phenylarsine oxide is used instead 1. Determination of Hydrogen Sulfide, of thiosulfate, replace NT and VT in Equation Ammoniacal Cadmium Chloride Method. API 11–3 with NA and VAS, respectively (see Sec- Method 772–54. In: Manual on Disposal of Re- tions 8.1.1 and 8.1.3). finery Wastes, Vol. V: Sampling and Analy- 9.4 Dry Gas Volume. Correct the sample sis of Waste Gases and Particulate Matter, volume measured by the dry gas meter to American Petroleum Institute, Washington, standard conditions (20°C and 760 mm Hg.) DC. 1954. 2. Tentative Method of Determination of
Vm(std)=Vm Y [(Tstd/Tm) (Pbar/Pstd)] Hydrogen Sulfide and Mercaptan Sulfur in Natural Gas, Natural Gas Processors Asso- Eq. 11–4 ciation, Tulsa, OK. NGPA Publication No. Where: 2265–65. 1965. 3. Knoll, J. E., and M. R. Midgett. Deter- Vm(std)=Volume at standard conditions of gas mination of Hydrogen Sulfide in Refinery sample through the dry gas meter, stand- Fuel Gases, Environmental Monitoring Se- ard liters. ries, Office of Research and Development, Vm=Volume of gas sample through the dry USEPA, Research Triangle Park, NC 27711, gas meter (meter conditions), liters. EPA 600/4–77–007. Tstd=Absolute temperature at standard con- 4. Scheil, G. W., and M. C. Sharp. Stand- ditions, 293°K. ardization of Method 11 at a Petroleum Re- Tm=Average dry gas meter temperature, °K. finery, Midwest Research Institute Draft Re- Pbar=Barometric pressure at the sampling port for USEPA, Office of Research and De- site, mm Hg. velopment, Research Triangle Park, NC Pstd=Absolute pressure at standard condi- 27711, EPA Contract No. 68–02–1098. August tions, 760 mm Hg. 1976, EPA 600/4–77–088a (Volume 1) and EPA Y=Dry gas meter calibration factor. 600/4–77–088b (Volume 2). 9.5 Concentration of H2S. Calculate the METHOD 12—DETERMINATION OF INORGANIC concentration of H2S in the gas stream at standard conditions using the following LEAD EMISSIONS FROM STATIONARY SOURCES equation: 1. Principle and Applicability CH2S=K[(VITNI¥VTTNT) sample— 1.1 Applicability. This method applies to (VIT NI¥VTT NT)]/Vm(std) the determination of inorganic lead (Pb) Eq. 11–5 emissions from specified stationary sources only. Where (metric units): 1.2 Principle. Particulate and gaseous Pb
CH2S=Concentration of H2S at standard con- emissions are withdrawn isokinetically from ditions, mg/dscm. the source and collected on a filter and in di- K=Conversion factor 17.04×10 3 lute nitric acid. The collected samples are digested in acid solution and analyzed by 3 (34.07 g/mole H2S) (1,000 LITERS/M ) (1,000 MG/G)/ atomic absorption spectrometry using an air (1,000 ML/LITER) (2H2S EQ/MOLE) acetylene flame.
VIT=Volume of standard iodine solution=50.0 2. Range, Sensitivity, Precision, and Inter- ml. ferences NI=Normality of standard iodine solution, g- 2.1 Range. For a minimum analytical ac- eq/liter. curacy of ±10 percent, the lower limit of the VTT=Volume of standard (∼0.01 N) sodium range is 100 µg. The upper limit can be con- thiosulfate solution, ml. siderably extended by dilution. NT=Normality of standard sodium 2.2 Analytical Sensitivity. Typical sen- thiosulfate solution, g-eq/liter. sitivities for a 1-percent change in absorp- Vm(std)=Dry gas volume at standard condi- tion (0.0044 absorbance units) are 0.2 and 0.5 tions, liters. µg Pb/ml for the 217.0 and 283.3 nm lines, re- NOTE: If phenylarsine oxide is used instead spectively. of thiosulfate, replace NT and VTT in Equa- 2.3 Precision. The within-laboratory pre- tion 11–5 with NA and VAT, respectively (see cision, as measured by the coefficient of var- Sections 7.3.1 and 8.1.3). iation ranges from 0.2 to 9.5 percent relative to a run-mean concentration. These values 10. Stability were based on tests conducted at a gray iron The absorbing solution is stable for at foundry, a lead storage battery manufactur- least 1 month. Sample recovery and analysis ing plant, a secondary lead smelter, and a should begin within 1 hour of sampling to lead recovery furnace of an alkyl lead manu- minimize oxidation of the acidified cadmium facturing plant. The concentrations encoun- sulfide. Once iodine has been added to the tered during these tests ranged from 0.61 to sample, the remainder of the analysis proce- 123.3 mg Pb/m3. dure must be completed according to Sec- 2.4 Interferences. Sample matrix effects tions 7.2.2 through 7.3.2. may interfere with the analysis for Pb by 11. Bibliography flame atomic absorption. If this interference
796
VerDate 11
is suspected, the analyst may confirm the System, Barometer, and Gas Density Deter- presence of these matrix effects and fre- mination Equipment. Same as Method 5, quently eliminate the interference by using Sections 2.1.1 to 2.1.6 and 2.1.8 to 2.1.10, re- the Method of Standard Additions. spectively. High concentrations of copper may inter- 3.1.2 Impingers. Four impingers connected fere with the analysis of Pb at 217.0 nm. This in series with leak-free ground glass fittings interference can be avoided by analyzing the or any similar leak-free noncontaminating samples at 283.3 nm. fittings. For the first, third, and fourth impingers, use the Greenburg-Smith design, 3. Apparatus modified by replacing the tip with a 1.3 cm 3.1 Sampling Train. A schematic of the (1⁄2 in.) ID glass tube extending to about 1.3 sampling train is shown in Figure 12–1; it is cm (1⁄2 in.) from the bottom of the flask. For similar to the Method 5 train. The sampling the second impinger, use the Greenburg- train consists of the following components: Smith design with the standard tip. Place a 3.1.1 Probe Nozzle, Probe Liner, Pitot thermometer, capable of measuring tempera- Tube, Differential Pressure Gauge, Filter ture to within 1°C (2°F) at the outlet of the Holder, Filter Heating System, Metering fourth impinger for monitoring purposes.
3.2 Sample Recovery. The following items liners that are either rubber-backed Teflon* are needed: or leak-free and resistant to chemical attack 3.2.1 Probe-Liner and Probe-Nozzle Brush- by 0.1 N HNO3. (Narrow mouth glass bottles es, Petri Dishes, Plastic Storage Containers, have been found to be less prone to leakage.) and Funnel and Rubber Policeman. Same as 3.2.4 Graduated Cylinder and/or Balance. Method 5, Sections 2.2.1, 2.2.4, 2.2.6, and 2.2.7, To measure condensed water to within 2 ml respectively. or 1 g. Use a graduated cylinder that has a 3.2.2 Wash Bottles. Glass (2). 3.2.3 Sample Storage Containers. Chemi- cally resistant, borosilicate glass bottles, for *Mention of trade names or specific prod- 0.1 N nitric acid (HNO3) impinger and probe ucts does not constitute endorsement by the solutions and washes, 1000-ml. Use screw-cap U.S. Environmental Protection Agency.
797
VerDate 11
minimum capacity of 500 ml, and subdivi- tilled water, add 2 ml concentrated HNO3, sions no greater than 5 ml. (Most laboratory and dilute to 100 ml with deionized distilled balances are capable of weighing to the near- water. est 0.5 g or less.) 4.4.5 Working Lead Standards. Pipet 0.0, 3.2.5 Funnel. Glass, to aid in sample re- 1.0, 2.0, 3.0, 4.0, and 5.0 ml of the stock lead covery. standard solution (4.4.4) into 250-ml volu- 3.3 Analysis. The following equipment is metric flasks. Add 5 ml of concentrated needed: HNO3 to each flask and dilute to volume 3.3.1 Atomic Absorption Spectrophotom- with deionized distilled water. These work- eter. With lead hollow cathode lamp and ing standards contain 0.0, 4.0, 8.0, 12.0, 16.0, burner for air/acetylene flame. and 20.0 µg Pb/ml, respectively. Prepare, as 3.3.2 Hot Plate. needed, additional standards at other con- 3.3.3 Erlenmeyer Flasks. 125-ml, 24/40 $¯. centrations in a similar manner. 3.3.4 Membrane Filters. Millipore SCWPO 4.4.6 Air. Suitable quality for atomic ab- 4700 or equivalent. sorption analysis. 3.3.5 Filtration Apparatus. Millipore vac- 4.4.7 Acetylene. Suitable quality for uum filtration unit, or equivalent, for use atomic absorption analysis. with the above membrane filter. 4.4.8 Hydrogen Peroxide, 3 percent (V/V). 3.3.6 Volumetric Flasks. 100-ml, 250-ml, Dilute 10 ml of 30 percent H2O2 to 100 ml with and 1000-ml. deionized distilled water. 4. Reagents 5. Procedure 4.1 Sampling. The reagents used in sam- 5.1 Sampling. The complexity of this pling are as follows: method is such that, in order to obtain reli- 4.1.1 Filter. Gelman Spectro Grade, Reeve able results, testers should be trained and Angel 934 AH, MSA 1106 BH, all with lot experienced with the test procedures. assay for Pb, or other high-purity glass fiber 5.1.1 Pretest Preparation. Follow the filters, without organic binder, exhibiting at same general procedure given in Method 5, least 99.95 percent efficiency (<0.05 percent Section 4.1.1, except the filter need not be penetration) on 0.3 micron dioctyl phthalate weighed. smoke particles. Conduct the filter effi- 5.1.2 Preliminary Determinations. Follow ciency test using ASTM Standard Method the same general procedure given in Method D2986–71 (incorporated by reference—see 5, Section 4.1.2. § 60.17) or use test data from the supplier’s 5.1.3 Preparation of Collection Train. Fol- quality control program. low the same general procedure given in 4.1.2 Silica Gel, Crushed Ice, and Stopcock Method 5, Section 4.1.3, except place 100 ml Grease. Same as Method 5, Sections 3.1.2, of 0.1 N HNO3 in each of the first two 3.1.4, and 3.1.5, respectively. impingers, leave the third impinger empty, 4.1.3 Water. Deionized distilled, to con- and transfer approximately 200 to 300 g of form to ASTM Specification D1192–77 (incor- preweighed silica gel from its container to porated by reference—see § 60.17), Type 3. If the fourth impinger. Set up the train as high concentrations of organic matter are shown in Figure 12–1. not expected to be present, the analyst may 5.1.4 Leak-Check Procedures. Follow the delete the potassium permanganate test for general leak-check procedures given in oxidizable organic matter. Method 5, Sections 4.1.4.1. (Pretest Leak- 4.1.4 Nitric Acid, 0.1 N. Dilute 6.5 ml of Check), 4.1.4.2 (Leak-Checks During the concentrated HNO3 to 1 liter with deionized Sample Run), and 4.1.4.3 (Post-Test Leak- distilled water. (It may be desirable to run Check). blanks before field use to eliminate a high 5.1.5 Sampling Train Operation. Follow blank on test samples.) the same general procedure given in Method 4.2 Pretest Preparation. 6 N HNO3 is need- 5, Section 4.1.5. For each run, record the data ed. Dilute 390 ml of concentrated HNO3 to 1 required on a data sheet such as the one liter with deionized distilled water. shown in EPA Method 5, Figure 5–2. 4.3 Sample Recovery. 0.1 N HNO3 (same as 5.1.6 Calculation of Percent Isokinetic. 4.1.4 above) is needed for sample recovery. Same as Method 5, Section 4.1.6. 4.4 Analysis. The following reagents are 5.2 Sample Recovery. Begin proper clean- needed for analysis (use ACS reagent grade up procedure as soon as the probe is removed chemicals or equivalent, unless otherwise from the stack at the end of the sampling pe- specified): riod. 4.4.1 Water. Same as 4.1.3 above. Allow the probe to cool. When it can be 4.4.2 Nitric Acid. Concentrated. safely handled, wipe off all external particu- 4.4.3 Nitric Acid, 50 percent (V/V). Dilute late matter near the tip of the probe nozzle 500 ml of concentrated HNO3 to 1 liter with and place a cap over it. Do not cap off the deionized distilled water. probe tip tightly while the sampling train is 4.4.4 Stock Lead Standard Solution, 1000 cooling down as this would create a vacuum µg Pb/ml. Dissolve 0.1598 g of lead nitrate in the filter holder, thus drawing liquid from [Pb(NO3)2] in about 60 ml of deionized dis- the impingers into the filter.
798
VerDate 11
Before moving the sampling train to the matter that is brushed from the probe. Run cleanup site, remove the probe from the sam- the brush through the probe three times or pling train, wipe off the silicone grease, and more until no visible sample matter is car- cap the open outlet of the probe. Be careful ried out with the 0.1 N HNO3 and none re- not to lose any condensate that might be mains on the probe liner on visual inspec- present. Wipe off the silicone grease from the tion. With stainless steel or other metal glassware inlet where the probe was fastened probes, run the brush through in the above and cap the inlet. Remove the umbilical cord prescribed manner at least six times, since from the last impinger and cap the impinger. metal probes have small crevices in which The tester may use ground-glass stoppers, sample matter can be entrapped. Rinse the plastic caps, or serum caps to close these brush with 0.1 N HNO3 and quantitatively openings. collect these washings in the sample con- Transfer the probe and filter-impinger as- tainer. After the brushing make a final rinse sembly to a cleanup area, which is clean and of the probe as described above. protected from the wind so that the chances It is recommended that two people clean of contaminating or losing the sample are the probe to minimize loss of sample. Be- minimized. tween sampling runs, keep brushes clean and Inspect the train prior to and during dis- protected from contamination. assembly and note any abnormal conditions. After insuring that all joints are wiped Treat the samples as follows: clean of silicone grease, brush and rinse with 5.2.1 Container No. 1 (Filter), Carefully re- 0.1 N HNO3 the inside of the front half of the move the filter from the filter holder and filter holder. Brush and rinse each surface place it in its identified petri dish container. three times or more, if needed, to remove If it is necessary to fold the filter, do so such visible sample matter. Make a final rinse of that the sample-exposed side is inside the the brush and filter holder. After all 0.1 N fold. Carefully transfer to the petri dish any HNO3 washings and sample matter are col- visible sample matter and/or filter fibers lected in the sample container, tighten the that adhere to the filter holder gasket by lid on the sample container so that the fluid using a dry Nylon bristle brush and/or a will not leak out when it is shipped to the sharp-edged blade. Seal the container. laboratory. Mark the height of the fluid level 5.2.2 Container No. 2 (Probe). Taking care to determine whether leakage occurs during that dust on the outside of the probe or other transport. Label the container to clearly exterior surfaces does not get into the sam- identify its contents. ple, quantitatively recover sample matter or 5.2.3 Container No. 3 (Silica Gel). Check any condensate from the probe nozzle, probe the color of the indicating silica gel to deter- fitting, probe liner, and front half of the fil- mine if it has been completely spent and ter holder by washing these components with make a notation of its condition. Transfer 0.1 N HNO3 and placing the wash into a glass the silica gel from the fourth impinger to the sample storage container. Measure and original container and seal. The tester may record (to the nearest 2-ml) the total amount use a funnel to pour the silica gel and a rub- of 0.1 N HNO3 used for each rinse. Perform ber policeman to remove the silica gel from the 0.1 N HNO3 rinses as follows: the impinger. It is not necessary to remove Carefully remove the probe nozzle and the small amount of particles that may ad- rinse the inside surfaces with 0.1 N HNO3 here to the walls and are difficult to remove. from a wash bottle while brushing with a Since the gain in weight is to be used for stainless steel, Nylon-bristle brush. Brush moisture calculations, do not use any water until the 0.1 N HNO3 rinse shows no visible or other liquids to transfer the silica gel. If particles, then make a final rinse of the in- a balance is available in the field, the tester side surface. may follow procedure for Container No. 3 Brush and rinse with 0.1 N HNO3 the inside under Section 5.4 (Analysis). parts of the Swagelok fitting in a similar 5.2.4 Container No. 4 (Impingers). Due to way until no visible particles remain. the large quantity of liquid involved, the Rinse the probe liner with 0.1 N HNO3. tester may place the impinger solutions in While rotating the probe so that all inside several containers. Clean each of the first surfaces will be rinsed with 0.1 N HNO3, tilt three impingers and connecting glassware in the probe and squirt 0.1 N HNO3 into its the following manner: upper end. Let the 0.1 N HNO3 drain from the 1. Wipe the impinger ball joints free of sili- lower end into the sample container. The cone grease and cap the joints. tester may use a glass funnel to aid in trans- 2. Rotate and agitate each impinger, so ferring liquid washes to the container. Fol- that the impinger contents might serve as a low the rinse with a probe brush. Hold the rinse solution. probe in an inclined position, squirt 0.1 N 3. Transfer the contents of the impingers HNO3 into the upper end of the probe as the to a 500-ml graduated cylinder. Remove the probe brush is being pushed with a twisting outlet ball joint cap and drain the contents action through the probe; hold the sample through this opening. Do not separate the container underneath the lower end of the impinger parts (inner and outer tubes) while probe and catch any 0.1 N HNO3 and sample transferring their contents to the cylinder.
799
VerDate 11
Measure the liquid volume to within ±2 ml. a temperature just below boiling. If the sam- Alternatively, determine the weight of the ple volume falls below 15 ml, add more 50 liquid to within ±0.5 g. Record in the log the percent HNO3. Add 10 ml of 3 percent H2O2 volume or weight of the liquid present, along and continue heating for 10 min. Add 50 ml of with a notation of any color or film observed hot (80°C) deionized distilled water and heat in the impinger catch. The liquid volume or for 20 min. Remove the flask from the hot weight is needed, along with the silica gel plate and allow to cool. Filter the sample data, to calculate the stack gas moisture through a Millipore membrane filter or content (see Method 5, Figure 5–3). equivalent and transfer the filtrate to a 250- 4. Transfer the contents to Container No. 4. ml volumetric flask. Dilute to volume with 5. Note: In steps 5 and 6 below, measure and deionized distilled water. record the total amount of 0.1 N HNO3 used 5.3.4 Filter Blank. Determine a filter for rinsing. Pour approximately 30 ml of 0.1 blank using two filters from each lot of fil- N HNO3 into each of the first three impingers ters used in the sampling train. Cut each fil- and agitate the impingers. Drain the 0.1 N ter into strips and place each filter in a sepa- HNO3 through the outlet arm of each im- rate 125-ml Erlenmeyer flask. Add 15 ml of 50 pinger into Container No. 4. Repeat this op- percent HNO3 and treat as described in Sec- eration a second time; inspect the impingers tion 5.3.3 using 10 ml of 3 percent H2O2 and 50 for any abnormal conditions. ml of hot, deionized distilled water. Filter 6. Wipe the ball joints of the glassware con- and dilute to a total volume of 100 ml using necting the impingers free of silicone grease deionized distilled water. and rinse each piece of glassware twice with 5.3.5 0.1 N HNO3 Blank. Take the entire 0.1 N HNO3 ; transfer this rinse into Con- 200 ml of 0.1 N HNO3 to dryness on a steam tainer No. 4. (Do not rinse or brush the glass- bath, add 15 ml of 50 percent HNO3, and treat fritted filter support.) Mark the height of the as described in Section 5.3.3 using 10 ml of 3 fluid level to determine whether leakage oc- percent H2O2 and 50 ml of hot, deionized dis- curs during transport. Label the container to tilled water. Dilute to a total volume of 100 clearly identify its contents. ml using deionized distilled water. 5.2.5 Blanks. Save 200 ml of the 0.1 N HNO3 5.4 Analysis. used for sampling and cleanup as a blank. 5.4.1 Lead Determination. Calibrate the Take the solution directly from the bottle spectrophotometer as described in Section being used and place into a glass sample con- 6.2 and determine the absorbance for each tainer labeled ‘‘0.1 N HNO blank.’’ 3 source sample, the filter blank, and 0.1 N 5.3 Sample Preparation. HNO blank. Analyze each sample three 5.3.1 Container No. 1 (Filter). Cut the filter 3 times in this manner. Make appropriate dilu- into strips and transfer the strips and all tions, as required, to bring all sample Pb loose particulate matter into a 125-ml Erlen- concentrations into the linear absorbance meyer flask. Rinse the petri dish with 10 ml range of the spectrophotometer. of 50 percent HNO to insure a quantitative 3 If the Pb concentration of a sample is at transfer and add to the flask. (Note: If the the low end of the calibration curve and high total volume required in Section 5.3.3 is ex- accuracy is required, the sample can be pected to exceed 80 ml, use a 250-ml Erlen- taken to dryness on a hot plate and the resi- meyer flask in place of the 125-ml flask.) 5.3.2 Containers No. 2 and No. 4 (Probe and due dissolved in the appropriate volume of Impingers). (Check the liquid level in Con- water to bring it into the optimum range of tainers No. 2 and/or No. 4 and confirm as to the calibration curve. whether or not leakage occurred during 5.4.2 Check for Matrix Effects on the Lead transport; note observation on the analysis Results. Since the analysis for Pb by atomic sheet. If a noticeable amount of leakage had absorption is sensitive to the chemical com- occurred, either void the sample or take position and to the physical properties (vis- steps, subject to the approval of the Admin- cosity, pH) of the sample (matrix effects), istrator, to adjust the final results.) Combine the analyst shall check at least one sample the contents of Containers No. 2 and No. 4 from each source using the method of addi- and take to dryness on a hot plate. tions as follows: 5.3.3 Sample Extraction for Lead. Based Add or spike an equal volume of standard on the approximate stack gas particulate solution to an aliquot of the sample solution, concentration and the total volume of stack then measure the absorbance of the resulting gas sampled, estimate the total weight of solution and the absorbance of an aliquot of particulate sample collected. Then transfer unspiked sample. the residue from Containers No. 2 and No. 4 Next, calculate the Pb concentration Cs in µ to the 125-ml Erlenmeyer flask that contains g/ml of the sample solution by using the fol- the filter using rubber policeman and 10 ml lowing equation: of 50 percent HNO3 for every 100 mg of sam- ple collected in the train or a minimum of 30 = As CCs a Eq.12- 1 ml of 50 percent HNO3 whichever is larger. − Place the Erlenmeyer flask on a hot plate AAt s and heat with periodic stirring for 30 min at Where:
800
VerDate 11
Ca=Pb concentration of the standard solu- 7.2 Volume of Water Vapor and Moisture tion, µg/ml. Content. Using data obtained in this test and As=Absorbance of the sample solution. Equations 5–2 and 5–3 of Method 5, calculate At=Absorbance of the spiked sample solution. the volume of water vapor Vw(std) and the Volume corrections will not be required if moisture content Bws of the stack gas. the solutions as analyzed have been made to 7.3 Total Lead in Source Sample. For each source sample correct the average ab- the same final volume. Therefore, Cs and Ca represent Pb concentration before dilutions. sorbance for the contribution of the filter Method of additions procedures described blank and the 0.1 N HNO3 blank. Use the cali- bration curve and this corrected absorbance on pages 9–4 and 9–5 of the section entitled µ ‘‘General Information’’ of the Perkin Elmer to determine the g Pb concentration in the sample aspirated into the spectrophoto- Corporation Atomic Absorption Spectropho- meter. Calculate the total Pb content C° tometry Manual, Number 303–0152 (see Cita- Pb (in µg) in the original source sample; correct tion 1 of Bibliography) may also be used. In for all the dilutions that were made to bring any event, if the results of the method of ad- the Pb concentration of the sample into the ditions procedure used on the single source linear range of the spectrophotometer. sample do not agree to within 5 percent of 7.4 Lead Concentration. Calculate the the value obtained by the routine atomic ab- stack gas Pb concentration CPb in mg/dscm sorption analysis, then reanalyze all samples as follows: from the source using a method of additions procedure. o 5.4.3 Container No. 3 (Silica Gel). The test- = C Pb er may conduct this step in the field. Weigh CKPb Eq.12- 2 the spent silica gel (or silica gel plus im- Vm() std pinger) to the nearest 0.5 g; record this Where: weight. K=0.001 mg/µg for metric units. 6. Calibration =2.205 lb/µg×10¥9 for English units. Maintain a laboratory log of all calibra- 7.5 Isokinetic Variation and Acceptable tions. Results. Same as Method 5, Sections 6.11 and 6.1 Sampling Train Calibration. Calibrate 6.12, respectively. To calculate vs, the aver- the sampling train components according to age stack gas velocity, use Equation 2–9 of the indicated sections of Method 5: Probe Method 2 and the data from this field test. Nozzle (Section 5.1); Pitot Tube (Section 5.2); 8. Alternative Test Methods for Inorganic Lead Metering System (Section 5.3); Probe Heater 8.1 Simultaneous Determination of Par- (Section 5.4); Temperature Gauges (Section ticulate and Lead Emissions. The tester may 5.5); Leak-Check of the Metering System use Method 5 to simultaneously determine (Section 5.6); and Barometer (Section 5.7). Pb provided that (1) he uses acetone to re- 6.2 Spectrophotometer. Measure the ab- move particulate from the probe and inside sorbance of the standard solutions using the of the filter holder as specified by Method 5, instrument settings recommended by the (2) he uses 0.1 N HNO3 in the impingers, (3) he spectrophotometer manufacturer. Repeat uses a glass fiber filter with a low Pb back- until good agreement (±3 percent) is obtained ground, and (4) he treats and analyzes the between two consecutive readings. Plot the entire train contents, including the absorbance (y-axis) versus concentration in impingers, for Pb as described in Section 5 of µg Pb/ml (x-axis). Draw or compute a this method. straight line through the linear portion of 8.2 Filter Location. The tester may use a the curve. Do not force the calibration curve filter between the third and fourth impinger through zero, but if the curve does not pass provided that he includes the filter in the through the origin or at least lie closer to analysis for Pb. the origin than ±0.003 absorbance units, 8.3 In-stack Filter. The tester may use an check for incorrectly prepared standards and in-stack filter provided that (1) he uses a for curvature in the calibration curve. glass-lined probe and at least two impingers, To determine stability of the calibration each containing 100 ml of 0.1 N HNO3, after curve, run a blank and a standard after every the in-stack filter and (2) he recovers and five samples and recalibrate, as necessary. analyzes the probe and impinger contents for 7. Calculations Pb. Recover sample from the nozzle with ace- 7.1 Dry Gas Volume. Using the data from tone if a particulate analysis is to be made. this test, calculate Vm(std), the total volume 9. Bibliography of dry gas metered corrected to standard 1. Perkin Elmer Corporation. Analytical conditions (20°C and 760 mm Hg), by using Methods for Atomic Absorption Spectropho- Equation 5–1 of Method 5. If necessary, ad- tometry. Norwalk, CT. September 1976. just Vw(std) for leakages as outlined in Sec- 2. American Society for Testing and Ma- tion 6.3 of Method 5. See the field data sheet terials. Annual Book of ASTM Standards. for the average dry gas meter temperature Part 31; Water, Atmospheric Analysis. Phila- and average orifice pressure drop. delphia, PA. 1974. p. 40–42.
801
VerDate 11
3. Klein, R. and C. Hach. Standard Addi- standard solutions must be measured at the tions—Uses and Limitations in same temperature. Spectrophotometric Analysis. Amer. Lab. 5. Apparatus 9:21–27. 1977. 4. Mitchell, W.J. and M.R. Midgett. Deter- 5.1 Sampling Train. A schematic of the mining Inorganic and Alkyl Lead Emissions sampling train is shown in Figure 13A–1; it is from Stationary Sources. U.S. Environ- similar to the Method 5 train except the fil- mental Protection Agency, Emission Mon- ter position is interchangeable. The sam- itoring and Support Laboratory. Research pling train consists of the following compo- Triangle Park, NC. (Presented at National nents: APCA Meeting. Houston. June 26, 1978). 5.1.1 Probe Nozzle, Pitot Tube, Differen- 5. Same as Method 5, Citations 2 to 5 and tial Pressure Gauge, Filter Heating System, 7 of bibliography. Metering System, Barometer, and Gas Den- sity Determination Equipment. Same as METHOD 13A—DETERMINATION OF TOTAL FLU- Method 5, Sections 2.1.1, 2.1.3, 2.1.4, 2.1.6, ORIDE EMISSIONS FROM STATIONARY 2.1.8, 2.1.9, and 2.1.10. When moisture con- SOURCES; SPADNS ZIRCONIUM LAKE METH- densation is a problem, the filter heating OD system is used. 1. Principle and Applicability 5.1.2 Probe Liner. Borosilicate glass or 316 stainless steel. When the filter is located 1.1 Applicability. This method applies to immediately after the probe, the tester may the determination of fluoride (F) emissions use a probe heating system to prevent filter from sources as specified in the regulations. plugging resulting from moisture condensa- It does not measure fluorocarbons, such as tion, but the tester shall not allow the tem- freons. perature in the probe to exceed 120±14°C 1.2 Principle. Gaseous and particulate F (248±25°F). are withdrawn isokinetically from the source 5.1.3 Filter Holder. With positive seal and collected in water and on a filter. The against leakage from the outside or around total F is then determined by the SPADNS the filter. If the filter is located between the Zirconium Lake Colorimetric Method. probe and first impinger, use borosilicate 2. Range and Sensitivity glass or stainless steel with a 20-mesh stain- The range of this method is 0 to 1.4 µg F/ less steel screen filter support and a silicone ml. Sensitivity has not been determined. rubber gasket; do not use a glass frit or a 3. Interferences sintered metal filter support. If the filter is located between the third and fourth Large quantities of chloride will interfere impingers, the tester may use borosilicate with the analysis, but this interference can glass with a glass frit filter support and a sil- be prevented by adding silver sulfate into the icone rubber gasket. The tester may also use distillation flask (see Section 7.3.4). If chlo- other materials of construction with ap- ride ion is present, it may be easier to use proval from the Administrator. the Specific Ion Electrode Method (Method 5.1.4 Impingers. Four impingers con- 13B). Grease on sample-exposed surfaces may nected as shown in Figure 13A–1 with cause low F results due to adsorption. ground-glass (or equivalent), vacuum-tight 4. Precision, Accuracy, and Stability fittings. For the first, third, and fourth 4.1 Precision. The following estimates impingers, use the Greenburg-Smith design, are based on a collaborative test done at a modified by replacing the tip with a 1.3-cm- primary aluminum smelter. In the test, six inside-diameter (1⁄2 in.) glass tube extending laboratories each sampled the stack simulta- to 1.3 cm (1⁄2 in.) from the bottom of the neously using two sampling trains for a total flask. For the second impinger, use a of 12 samples per sampling run. Fluoride con- Greenburg-Smith impinger with the stand- centrations encountered during the test ard tip. The tester may use modifications ranged from 0.1 to 1.4 mg F/m3. The within- (e.g., flexible connections between the laboratory and between-laboratory standard impingers or materials other than glass), deviations, which include sampling and anal- subject to the approval of the Administrator. ysis errors, were 0.044 mg F/m3with 60 de- Place a thermometer, capable of measuring grees of freedom and 0.064 mg F/m3with five temperature to within 1°C (2°F), at the outlet degrees of freedom, respectively. of the fourth impinger for monitoring pur- 4.2 Accuracy. The collaborative test did poses. not find any bias in the analytical method. 5.2 Sample Recovery. The following 4.3 Stability. After the sample and col- items are needed: orimetric reagent are mixed, the color 5.2.1 Probe-Liner and Probe-Nozzle Brush- formed is stable for approximately 2 hours. A es, Wash Bottles, Graduated Cylinder and/or 3 °C temperature difference between the sam- Balance, Plastic Storage Containers, Rubber ple and standard solutions produces an error Policeman, Funnel. Same as Method 5, Sec- of approximately 0.005 mg F/liter. To avoid tions 2.2.1 to 2.2.2 and 2.2.5 to 2.2.8, respec- this error, the absorbances of the sample and tively.
802
VerDate 11
5.2.2 Sample Storage Container. Wide- 5.3.9 Balance. 300-g capacity to measure mouth, high-density-polyethylene bottles for to ±0.5 g. impinger water samples, 1-liter. 5.3.10 Spectrophotometer. Instrument 5.3 Analysis. The following equipment is that measures absorbance at 570 nm and pro- needed: vides at least a 1-cm light path. 5.3.1 Distillation Apparatus. Glass dis- 5.3.11 Spectrophotometer Cells. 1-cm tillation apparatus assembled as shown in pathlength. Figure 13A–2. 5.3.2 Bunsen Burner. 6. Reagents 5.3.3 Electric Muffle Furnace. Capable of 6.1 Sampling. Use ACS reagent-grade ° heating to 600 C. chemicals or equivalent, unless otherwise 5.3.4 Crucibles. Nickel, 75- to 100-ml. specified. The reagents used in sampling are 5.3.5 Beakers. 500-ml and 1500-ml. as follows: 5.3.6 Volumetric Flasks. 50-ml. 5.3.7 Erlenmeyer Flasks or Plastic Bot- 6.1.1 Filters. tles. 500-ml. 6.1.1.1 If the filter is located between the 5.3.8 Constant Temperature Bath. Capa- third and fourth impingers, use a Whatman 1 ble of maintaining a constant temperature of No. 1 filter, or equivalent, sized to fit the fil- ±1.0°C at room temperature conditions. ter holder.
1 Mention of company or product names does not constitute endorsement by the U.S. Environmental Protection Agency. 803
VerDate 11
804
VerDate 11
6.3.1 Calcium Oxide (CaO). Certified grade containing 0.005 percent F or less. 6.3.2 Phenolphthalein Indicator. Dissolve 0.1 g of phenolphthalein in a mixture of 50 ml of 90 percent ethanol and 50 ml of deionized distilled water. 6.3.3 Silver Sulfate (Ag2 SO4). 6.3.4 Sodium Hydroxide (NaOH). Pellets. 6.3.5 Sulfuric Acid (H2SO4), Concentrated. 6.3.6 Sulfuric Acid, 25 percent (V/V). Mix 1 part of concentrated H2SO4 with 3 parts of deionized distilled water. 6.3.7 Filters. Whatman No. 541, or equiv- alent. 6.3.8 Hydrochloric Acid (HCl), Con- centrated. 6.3.9 Water. From same container as de- scribed in Section 6.1.2. 6.3.10 Fluoride Standard Solution, 0.01 mg F/ml. Dry in an oven at 110°C for at least 2 hours. Dissolve 0.2210 g of NaF in 1 liter of deionized distilled water. Dilute 100 ml of this solution to 1 liter with deionized dis- tilled water. 6.3.11 SPADNS Solution [4, 5 dihydroxy-3- (p-sulfophenylazo)-2,7-naphthalene-disulfonic acid trisodium salt]. Dissolve 0.960 ± 0.010 g of SPADNS reagent in 500 ml deionized dis- tilled water. If stored in a well-sealed bottle 6.1.1.2 If the filter is located between the protected from the sunlight, this solution is probe and first impinger, use any suitable stable for at least 1 month. medium (e.g., paper, organic membrane) that 6.3.12 Spectrophotometer Zero Reference conforms to the following specifications: (1) Solution. Prepare daily. Add 10 ml of The filter can withstand prolonged exposure SPADNS solution (6.3.11) to 100 ml deionized to temperatures up to 135°C (275°F). (2) The distilled water, and acidify with a solution filter has at least 95 percent collection effi- prepared by diluting 7 ml of concentrated ciency (≤5 percent penetration) for 0.3 µm HCl to 10 ml with deionized distilled water. dioctyl phthalate smoke particles. Conduct 6.3.13 SPADNS Mixed Reagent. Dissolve the filter efficiency test before the test se- 0.135 ± 0.005 g of zirconyl chloride octahydrate ries, using ASTM Standard Method D 2986–71, (ZrOCl2. 8H2O) in 25 ml of deionized distilled or use test data from the supplier’s quality water. Add 350 ml of concentrated HCl, and control program. (3) The filter has a low F dilute to 500 ml with deionized distilled blank value (≤0.015 mg F/cm2of filter area). water. Mix equal volumes of this solution Before the test series, determine the average and SPADNS solution to form a single rea- F blank value of at least three filters (from gent. This reagent is stable for at least 2 the lot to be used for sampling) using the ap- months. plicable procedures described in Sections 7.3 7. Procedure and 7.4 of this method. In general, glass fiber 7.1 Sampling. Because of the complexity filters have high and/or variable F blank val- of this method, testers should be trained and ues, and will not be acceptable for use. experienced with the test procedures to as- 6.1.2 Water. Deionized distilled, to con- sure reliable results. form to ASTM Specification D 1193–74, Type 7.1.1 Pretest Preparation. Follow the 3. If high concentrations of organic matter general procedure given in Method 5, Section are not expected to be present, the analyst 4.1.1, except the filter need not be weighed. may delete the potassium permanganate test 7.1.2 Preliminary Determinations. Fol- for oxidizable organic matter. low the general procedure given in Method 5, 6.1.3 Silica Gel, Crushed Ice, and Stopcock Section 4.1.2., except the nozzle size selected Grease. Same as Method 5, Sections 3.1.2, must maintain isokinetic sampling rates 3.1.4, and 3.1.5, respectively. below 28 liters/min (1.0 cfm). 6.2 Sample Recovery. Water, from same 7.1.3 Preparation of Collection Train. container as described in Section 6.1.2, is Follow the general procedure given in Meth- needed for sample recovery. od 5, Section 4.1.3, except for the following 6.3 Sample Preparation and Analysis. variations: The reagents needed for sample preparation Place 100 ml of deionized distilled water in and analysis are as follows: each of the first two impingers, and leave the
805
VerDate 11
third impinger empty. Transfer approxi- probe in this determination. Transfer the im- mately 200 to 300 g of preweighed silica gel pinger water from the graduated cylinder from its container to the fourth impinger. into this polyethylene container. Add the fil- Assemble the train as shown in Figure 13A– ter to this container. (The filter may be han- 1 with the filter between the third and fourth dled separately using procedures subject to impingers. Alternatively, if a 20-mesh stain- the Administrator’s approval.) Taking care less steel screen is used for the filter sup- that dust on the outside of the probe or other port, the tester may place the filter between exterior surfaces does not get into the sam- the probe and first impinger. The tester may ple, clean all sample-exposed surfaces (in- also use a filter heating system to prevent cluding the probe nozzle, probe fitting, probe moisture condensation, but shall not allow liner, first three impingers, impinger connec- the temperature around the filter holder to tors, and filter holder) with deionized dis- exceed 120 ± 14°C (248 ± 25°F). Record the filter tilled water. Use less than 500 ml for the en- location on the data sheet. tire wash. Add the washings to the sampler 7.1.4 Leak-Check Procedures. Follow the container. Perform the deionized distilled leak-check procedures given in Method 5, water rinses as follows: Sections 4.1.4.1 (Pretest Leak-Check), 4.1.4.2 Carefully remove the probe nozzle and (Leak-Checks During the Sample Run), and rinse the inside surface with deionized dis- 4.1.4.3 (Post-Test Leak-Check). tilled water from a wash bottle. Brush with 7.1.5 Fluoride Train Operation. Follow a Nylon bristle brush, and rinse until the the general procedure given in Method 5, rinse shows no visible particles, after which Section 4.1.5, keeping the filter and probe make a final rinse of the inside surface. ± ° temperatures (if applicable) at 120 14 C Brush and rinse the inside parts of the ± ° (248 25 F) and isokinetic sampling rates Swagelok fitting with deionized distilled below 28 liters/min (1.0 cfm). For each run, water in a similar way. record the data required on a data sheet such Rinse the probe liner with deionized dis- as the one shown in Method 5, Figure 5–2. tilled water. While squirting the water into 7.2 Sample Recovery. Begin proper the upper end of the probe, tilt and rotate cleanup procedure as soon as the probe is re- moved from the stack at the end of the sam- the probe so that all inside surfaces will be pling period. wetted with water. Let the water drain from Allow the probe to cool. When it can be the lower end into the sample container. The safely handled, wipe off all external particu- tester may use a funnel (glass or poly- late matter near the tip of the probe nozzle ethylene) to aid in transferring the liquid and place a cap over it to keep from losing washes to the container. Follow the rinse part of the sample. Do not cap off the probe with a probe brush. Hold the probe in an in- tip tightly while the sampling train is cool- clined position, and squirt deionized distilled ing down, because a vacuum would form in water into the upper end as the probe brush the filter holder, thus drawing impinger is being pushed with a twisting action water backwards. through the probe. Hold the sample con- Before moving the sample train to the tainer underneath the lower end of the cleanup site, remove the probe from the sam- probe, and catch any water and particulate ple train, wipe off the silicone grease, and matter that is brushed from the probe. Run cap the open outlet of the probe. Be careful the brush through the probe three times or not to lose any condensate, if present. Re- more. With stainless steel or other metal move the filter assembly, wipe off the sili- probes, run the brush through in the above cone grease from the filter holder inlet, and prescribed manner at least six times since cap this inlet. Remove the umbilical cord metal probes have small crevices in which from the last impinger, and cap the im- particulate matter can be entrapped. Rinse pinger. After wiping off the silicone grease, the brush with deionized distilled water, and cap off the filter holder outlet and any open quantitatively collect these washings in the impinger inlets and outlets. The tester may sample container. After the brushing, make use ground-glass stoppers, plastic caps, or a final rinse of the probe as described above. serum caps to close these openings. It is recommended that two people clean Transfer the probe and filter-impinger as- the probe to minimize sample losses. Be- sembly to an area that is clean and protected tween sampling runs, keep brushes clean and from the wind so that the chances of con- protected from contamination. taminating or losing the sample is mini- Rinse the inside surface of each of the first mized. three impingers (and connecting glassware) Inspect the train before and during dis- three separate times. Use a small portion of assembly, and note any abnormal conditions. deionized distilled water for each rinse, and Treat the samples as follows: brush each sample-exposed surface with a 7.2.1 Container No. 1 (Probe, Filter, and Nylon bristle brush, to ensure recovery of Impinger Catches). Using a graduated cyl- fine particulate matter. Make a final rinse of inder, measure to the nearest ml, and record each surface and of the brush. the volume of the water in the first three After ensuring that all joints have been impingers; include any condensate in the wiped clean of the silicone grease, brush and
806
VerDate 11
rinse with deionized distilled water the in- ration of the water, keep the slurry basic side of the filter holder (front-half only, if (red to phenolphthalein) to avoid loss of F. If filter is positioned between the third and the indicator turns colorless (acidic) during fourth impingers). Brush and rinse each sur- the evaporation, add CaO until the color face three times or more if needed. Make a turns red again. final rinse of the brush and filter holder. After evaporation of the water, place the After all water washings and particulate crucible on a hot plate under a hood and matter have been collected in the sample slowly increase the temperature until the container, tighten the lid so that water will Whatman No. 541 and sampling filters char. not leak out when it is shipped to the labora- It may take several hours to completely char tory. Mark the height of the fluid level to de- the filters. termine whether leakage occurs during transport. Label the container clearly to Place the crucible in a cold muffle furnace. identify its contents. Gradually (to prevent smoking) increase the ° 7.2.2 Container No. 2 (Sample Blank). temperature to 600 C, and maintain until the Prepare a blank by placing an unused filter contents are reduced to an ash. Remove the in a polyethylene container and adding a vol- crucible from the furnace and allow to cool. ume of water equal to the total volume in Add approximately 4 g of crushed NaOH to Container No. 1. Process the blank in the the crucible and mix. Return the crucible to same manner as for Container No. 1. the muffle furnace, and fuse the sample for 7.2.3 Container No. 3 (Silica Gel). Note 10 minutes at 600°C. the color of the indicating silica gel to deter- Remove the sample from the furnace, and mine whether it has been completely spent cool to ambient temperature. Using several and make a notation of its condition. Trans- rinsings of warm deionized distilled water, fer the silica gel from the fourth impinger to transfer the contents of the crucible to the its original container and seal. The tester beaker containing the filtrate. To assure may use a funnel to pour the silica gel and complete sample removal, rinse finally with a rubber policeman to remove the silica gel two 20-ml portions of 25 percent H2SO4, and from the impinger. It is not necessary to re- carefully add to the beaker. Mix well, and move the small amount of dust particles transfer to a 1-liter volumetric flask. Dilute that may adhere to the impinger wall and to volume with deionized distilled water, and are difficult to remove. Since the gain in mix thoroughly. Allow any undissolved sol- weight is to be used for moisture calcula- ids to settle. tions, do not use any water or other liquids to transfer the silica gel. If a balance is 7.3.2 Container No. 2 (Sample Blank). available in the field, the tester may follow Treat in the same manner as described in the analytical procedure for Container No. 3 Section 7.3.1 above. in Section 7.4.2. 7.3.3 Adjustment of Acid/Water Ratio in 7.3 Sample Preparation and Distillation. Distillation Flask. (Use a protective shield (Note the liquid levels in Containers No. 1 when carrying out this procedure.) Place 400 and No. 2 and confirm on the analysis sheet ml of deionized distilled water in the dis- whether or not leakage occurred during tillation flask, and add 200 ml of con- transport. If noticeable leakage had oc- centrated H2SO4. (Caution: Observe standard curred, either void the sample or use meth- precautions when mixing H2SO4 with water. ods, subject to the approval of the Adminis- Slowly add the acid to the flask with con- trator, to correct the final results.) Treat stant swirling.) Add some soft glass beads the contents of each sample container as de- and several small pieces of broken glass tub- scribed below: ing, and assemble the apparatus as shown in 7.3.1 Container No. 1 (Probe, Filter, and Figure 13A–2. Heat the flask until it reaches Impinger Catches). Filter this container’s a temperature of 175°C to adjust the acid/ contents, including the sampling filter, water ratio for subsequent distillations. Dis- through Whatman No. 541 filter paper, or card the distillate. equivalent, into a 1500-ml beaker. 7.3.4 Distillation. Cool the contents of 7.3.1.1 If the filtrate volume exceeds 900 the distillation flask to below 80°C. Pipet an ml, make the filtrate basic (red to phenol- aliquot of sample containing less than 10.0 phthalein) with NaOH, and evaporate to less mg F directly into the distillation flask, and than 900 ml. add deionized distilled water to make a total 7.3.1.2 Place the filtered material (includ- volume of 220 ml added to the distillation ing sampling filter) in a nickel crucible, add flask. (To estimate the appropriate aliquot a few ml of deionized distilled water, and size, select an aliquot of the solution and macerate the filters with a glass rod. Add 100 mg CaO to the crucible, and mix treat as described in Section 7.4.1. This will the contents thoroughly to form a slurry. be an approximation of the F content be- Add two drops of phenolphthalein indicator. cause of possible interfering ions.) Place the crucible in a hood under infrared NOTE: If the sample contains chloride, add lamps or on a hot plate at low heat. Evapo- 5 mg of Ag2 SO4 to the flask for every mg of rate the water completely. During the evapo- chloride.
807
VerDate 11
Place a 250-ml volumetric flask at the con- Leak Check of Metering System (Section denser exit. Heat the flask as rapidly as pos- 5.6); and Barometer (Section 5.7). sible with a Bunsen burner, and collect all 8.2 Spectrophotometer. Prepare the the distillate up to 175°C. During heatup, blank standard by adding 10 ml of SPADNS play the burner flame up and down the side mixed reagent to 50 ml of deionized distilled of the flask to prevent bumping. Conduct the water. Accurately prepare a series of stand- distillation as rapidly as possible (15 minutes ards from the 0.01 mg F/ml standard fluoride or less). Slow distillations have been found solution (6.3.10) by diluting 0, 2, 4, 6, 8, 10, 12, to produce low F recoveries. Caution: Be and 14 ml to 100 ml with deionized distilled careful not to exceed 175°C to avoid causing water. Pipet 50 ml from each solution and H2SO4 to distill over. transfer each to a separate 100-ml beaker. If F distillation in the mg range is to be Then add 10 ml of SPADNS mixed reagent to followed by a distillation in the fractional each. These standards will contain 0, 10, 20, mg range, add 220 ml of deionized distilled 30, 40 50, 60, and 70 µg F (0 to 1.4 µg/ml), re- water and distill it over as in the acid ad- spectively. justment step to remove residual F from the After mixing, place the reference standards distillation system. and reference solution in a constant tem- The tester may use the acid in the distilla- perature bath for 30 minutes before reading tion flask until there is carry-over of inter- the absorbance with the spectrophotometer. ferences or poor F recovery. Check for these Adjust all samples to this same temperature every tenth distillation using a deionized before analyzing. distilled water blank and a standard solu- With the spectrophotometer at 570 nm, use tion. Change the acid whenever the F recov- the reference solution (6.3.12) to set the ab- ery is less than 90 percent or the blank value sorbance to zero. exceeds 0.1 µg/ml. Determine the absorbance of the stand- 7.4 Analysis. ards. Prepare a calibration curve by plotting 7.4.1 Containers No. 1 and No. 2. After µg F/50 ml versus absorbance on linear graph distilling suitable aliquots from Containers paper. Prepare the standard curve initially No. 1 and No. 2 according to Section 7.3.4, di- and thereafter whenever the SPADNS mixed lute the distillate in the volumetric flasks to reagent is newly made. Also, run a calibra- exactly 250 ml with deionized distilled water, tion standard with each set of samples and if and mix thoroughly. Pipet a suitable aliquot it differs from the calibration curve by ±2 of each sample distillate (containing 10 to 40 percent, prepare a new standard curve. µg F/ml) into a beaker, and dilute to 50 ml with deionized distilled water. Use the same 9. Calculations aliquot size for the blank. Add 10 ml of Carry out calculations, retaining at least SPADNS mixed reagent (6.3.13), and mix one extra decimal figure beyond that of the thoroughly. acquired data. Round off figures after final After mixing, place the sample in a con- calculation. Other forms of the equations stant-temperature bath containing the may be used, provided that they yield equiv- standard solutions (see Section 8.2) for 30 alent results. minutes before reading the absorbance on 9.1 Nomenclature the spectrophotometer. A = Aliquot of distillate taken for color de- Set the spectrophotometer to zero absorb- d velopment, ml. ance at 570 nm with the reference solution A = Aliquot of total sample added to still, (6.3.12), and check the spectrophotometer t ml. calibration with the standard solution. De- B = Water vapor in the gas stream, propor- termine the absorbance of the samples, and ws tion by volume. determine the concentration from the cali- C = Concentration of F in stack gas, mg/m3 bration curve. If the concentration does not s (mg/ft3), dry basis, corrected to standard fall within the range of the calibration conditions of 760 mm Hg (29.92 in. Hg) and curve, repeat the procedure using a different 293°K (528°R). size aliquot. 7.4.2 Container No. 3 (Silica Gel). Weigh Ft = Total F in sample, mg. µ the spent silica gel (or silica gel plus im- g F = Concentration from the calibration µ pinger) to the nearest 0.5 g using a balance. curve, g. The tester may conduct this step in the field. Tm = Absolute average dry gas meter tem- perature (see Figure 5–2 of Method 5), °K 8. Calibration (°R). Maintain a laboratory log of all calibra- Ts = Absolute average stack gas temperature tions. (see Figure 5–2 of Method 5), °K (°R). 8.1 Sampling Train. Calibrate the sam- Vd = Volume of distillate as diluted, ml. pling train components according to the in- Vm(std) = Volume of gas sample as measured dicated sections in Method 5: Probe Nozzle by dry gas meter, corrected to standard (Section 5.1); Pitot Tube (Section 5.2); Meter- conditions, dscm (dscf). ing System (Section 5.3); Probe Heater (Sec- Vt = Total volume of F sample, after final di- tion 5.4); Temperature Gauges (Section 5.5); lution, ml.
808
VerDate 11
Vw(std) = Volume of water vapor in the gas µg F/ml require extra care. Sensitivity has sample, corrected to standard conditions, not been determined. scm (scf). 3. Interferences 9.2 Average Dry Gas Meter Temperature Grease on sample-exposed surfaces may and Average Orifice Pressure Drop. See data cause low F results because of adsorption. sheet (Figure 5–2 of Method 5). 4. Precision and Accuracy 9.3 Dry Gas Volume. Calculate Vm(std) and adjust for leakage, if necessary, using the 4.1 Precision. The following estimates equation in Section 6.3 of Method 5. are based on a collaborative test done at a 9.4 Volume of Water Vapor and Moisture primary aluminum smelter. In the test, six Content. Calculate the volume of water laboratories each sampled the stack simulta- vapor Vw(std) and moisture content Bws from neously using two sampling trains for a total the data obtained in this method (Figure of 12 samples per sampling run. Fluoride con- 13A–1); use Equations 5–2 and 5–3 of Method 5. centrations encountered during the test 9.5 Concentration. ranged from 0.1 to 1.4 mg F/m3. The within- 9.5.1 Total Fluoride in Sample. Calculate laboratory and between-laboratory standard the amount of F in the sample using the fol- deviations, which include sampling and anal- lowing equation: ysis errors, are 0.037 mg F/m3with 60 degrees of freedom and 0.056 mg F/m3with five de- = −3 Vt Vd µ grees of freedom, respectively. Ft 10().g F Eq 13 A- 1 4.2 Accuracy. The collaborative test did At Ad not find any bias in the analytical method. 9.5.2 Fluoride Concentration in Stack 5. Apparatus Gas. Determine the F concentration in the 5.1 Sampling Train and Sample Recovery. stack gas using the following equation: Same as Method 13A, Sections 5.1 and 5.2, re- spectively. = Ft 5.2 Analysis. The following items are Cs Eq.13 A- 2 needed: Vm() std 5.2.1 Distillation Apparatus, Bunsen 9.6 Isokinetic Variation and Acceptable Burner, Electric Muffle Furnace, Crucibles, Results. Use Method 5, Sections 6.11 and Beakers, Volumetric Flasks, Erlenmeyer 6.12. Flasks or Plastic Bottles, Constant Tem- perature Bath, and Balance. Same as Meth- 10. Bibliography od 13A, Sections 5.3.1 to 5.3.9, respectively, 1. Bellack, Ervin, Simplified Fluoride Dis- except include also 100-ml polyethylene tillation Method. Journal of the American beakers. Water Works Association. 50:5306. 1958. 5.2.2 Fluoride Ion Activity Sensing Elec- 2. Mitchell, W. J., J. C. Suggs, and F. J. trode. Bergman. Collaborative Study of EPA Meth- 5.2.3 Reference Electrode. Single junc- od 13A and Method 13B. Publication No. tion, sleeve type. EPA–600/4–77–050. Environmental Protection 5.2.4 Electrometer. A pH meter with mil- Agency. Research Triangle Park, NC. Decem- livolt-scale capable of ±0.1-mv resolution, or ber 1977. a specific ion meter made specifically for 3. Mitchell, W. J. and M. R. Midgett. Ade- specific ion use. quacy of Sampling Trains and Analytical 5.2.5 Magnetic Stirrer and TFE 2 Fluoro- Procedures Used for Fluoride. Atm. Environ. carbon-Coated Stirring Bars. 10:865–872. 1976. 6. Reagents METHOD 13B—DETERMINATION OF TOTAL FLU- 6.1 Sampling and Sample Recovery. ORIDE EMISSIONS FROM STATIONARY Same as Method 13A, Sections 6.1 and 6.2, re- SOURCES—SPECIFIC ION ELECTRODE METHOD spectively. 6.2 Analysis. Use ACS reagent grade 1. Principle and Applicability chemicals (or equivalent), unless otherwise 1.1 Applicability. This method applies to specified. The reagents needed for analysis the determination of fluoride (F) emissions are as follows: from stationary sources as specified in the 6.2.1 Calcium Oxide (CaO). Certified regulations. It does not measure fluoro- grade containing 0.005 percent F or less. carbons, such as freons. 6.2.2 Phenolphthalein Indicator. Dissolve 1.2 Principle. Gaseous and particulate F 0.1 g of phenolphthalein in a mixture of 50 ml are withdrawn isokinetically from the source of 90 percent ethanol and 50 ml deionized dis- and collected in water and on a filter. The tilled water. total F is then determined by the specific ion 6.2.3 Sodium Hydroxide (NaOH). Pellets. electrode method. 2. Range and Sensitivity 2 Mention of any trade name or specific The range of this method is 0.02 to 2,000 µg product does not constitute endorsement by F/ml; however, measurements of less than 0.1 the U.S. Environmental Protection Agency.
809
VerDate 11
6.2.4 Sulfuric Acid (H2SO4), Concentrated. reading is obtained, record it. This may take 6.2.5 Filters. Whatman No. 541, or equiv- several minutes. Determine concentration alent. from the calibration curve. Between elec- 6.2.6 Water. From same container as 6.1.2 trode measurements, rinse the electrode of Method 13A. with deionized distilled water. 6.2.7 Sodium Hydroxide, 5 M. Dissolve 20 7.2.2 Container No. 3 (Silica Gel). Same g of NaOH in 100 ml of deionized distilled as Method 13A, Section 7.4.2. water. 8. Calibration 6.2.8 Sulfuric Acid, 25 percent (V/V). Mix Maintain a laboratory log of all calibra- 1 part of concentrated H2SO4 with 3 parts of tions. deionized distilled water. 8.1 Sampling Train. Same as Method 6.2.9 Total Ionic Strength Adjustment 13A. Buffer (TISAB). Place approximately 500 ml 8.2 Fluoride Electrode. Prepare fluoride of deionized distilled water in a 1-liter beak- standardizing solutions by serial dilution of er. Add 57 ml of glacial acetic acid, 58 g of so- the 0.1 M fluoride standard solution. Pipet 10 dium chloride, and 4 g of cyclohexylene ml of 0.1 M fluoride standard solution into a dinitrilo tetraacetic acid. Stir to dissolve. 100-ml volumetric flask, and make up to the Place the beaker in a water bath to cool it. mark with deionized distilled water for a Slowly add 5 M NaOH to the solution, meas- 10¥2M standard solution. Use 10 ml of 10¥2M uring the pH continuously with a calibrated solution to make a 10¥3M solution in the pH/reference electrode pair, until the pH is same manner. Repeat the dilution procedure 5.3. Cool to room temperature. Pour into a 1- and make 10¥4and 10¥5solutions. liter volumetric flask, and dilute to volume Pipet 50 ml of each standard into a sepa- with deionized distilled water. Commercially rate beaker. Add 50 ml of TISAB to each prepared TISAB may be substituted for the beaker. Place the electrode in the most di- above. lute standard solution. When a steady milli- 6.2.10 Fluoride Standard Solution, 0.1 M. volt reading is obtained, plot the value on Oven dry some sodium fluoride (NaF) for a the linear axis of semilog graph paper versus minimum of 2 hours at 110°C, and store in a concentration on the log axis. Plot the nomi- desiccator. Then add 4.2 g of NaF to a 1-liter nal value for concentration of the standard volumetric flask, and add enough deionized on the log axis, e.g., when 50 ml of 10¥2M distilled water to dissolve. Dilute to volume standard is diluted with 50 ml of TISAB, the with deionized distilled water. concentration is still designated ‘‘10¥2M.’’ 7. Procedure Between measurements soak the fluoride 7.1 Sampling, Sample Recovery, and Sam- sensing electrode in deionized distilled water ple Preparation and Distillation. Same as for 30 seconds, and then remove and blot dry. Analyze the standards going from dilute to Method 13A, Sections 7.1, 7.2, and 7.3, respec- concentrated standards. A straight-line cali- tively, except the notes concerning chloride bration curve will be obtained, with nominal and sulfate interferences are not applicable. concentrations of 10¥4, 10¥3, 10¥2, and 7.2 Analysis. 10¥1fluoride molarity on the log axis plotted 7.2.1 Containers No. 1 and No. 2. Distill versus electrode potential (in mv) on the lin- suitable aliquots from Containers No. 1 and ear scale. Some electrodes may be slightly No. 2. Dilute the distillate in the volumetric nonlinear between 10¥5and 10¥4M. If this oc- flasks to exactly 250 ml with deionized dis- curs, use additional standards between these tilled water and mix thoroughly. Pipet a 25- two concentrations. ml aliquot from each of the distillate and Calibrate the fluoride electrode daily, and separate beakers. Add an equal volume of check it hourly. Prepare fresh fluoride stand- TISAB, and mix. The sample should be at the ardizing solutions daily (10¥2M or less). same temperature as the calibration stand- Store fluoride standardizing solutions in pol- ards when measurements are made. If ambi- yethylene or polypropylene containers. ent laboratory temperature fluctuates more than ±2°C from the temperature at which the NOTE: Certain specific ion meters have calibration standards were measured, condi- been designed specifically for fluoride elec- tion samples and standards in a constant- trode use and give a direct readout of fluo- temperature bath before measurement. Stir ride ion concentration. These meters may be the sample with a magnetic stirrer during used in lieu of calibration curves for fluoride measurement to minimize electrode response measurements over narrow concentration time. If the stirrer generates enough heat to ranges. Calibrate the meter according to the change solution temperature, place a piece manufacturer’s instructions.) of temperature insulating material such as 9. Calculations cork, between the stirrer and the beaker. Carry out calculations, retaining at least Hold dilute samples (below 10¥4M fluoride one extra decimal figure beyond that of the ion content) in polyethylene beakers during acquired data. Round off figures after final measurement. calculation. Insert the fluoride and reference electrodes 9.1 Nomenclature. Same as Method 13A, into the solution. When a steady millivolt Section 9.1. In addition:
810
VerDate 11
M=F concentration from calibration curve, mometer shall be able to withstand pro- molarity. longed exposure to dusty and corrosive envi- 9.2 Average Dry Gas Meter Temperature ronments; one way of achieving this is to and Average Orifice Pressure Drop, Dry Gas continuously purge the bearings of the ane- Volume, Volume of Water Vapor and Mois- mometer with filtered air during operation; ture Content, Fluoride Concentration in (6) All anemometer components shall be Stack Gas, and Isokinetic Variation and Ac- properly shielded or encased, such that the ceptable Results. Same as Method 13A, Sec- performance of the anemometer is uninflu- tions 9.2 to 9.4, 9.5.2, and 9.6, respectively. enced by potroom magnetic field effects; (7) 9.3 Fluoride in Sample. Calculate the A known relationship shall exist between the amount of F in the sample using the follow- electrical output signal from the anemom- ing: eter generator and the propeller shaft rpm, V at a minimum of three evenly spaced rpm = t settings between 60 and 1800 rpm; for the 3 FKt Vd M Eq.13 B- 1 settings, use 60±15, 900±100, and 1800±100 rpm. At Anemometers having other types of output Where: signals (e.g., optical) may be used, subject to K=19 mg/millimole. the approval of the Administrator. If other 10. Bibliography types of anemometers are used, there must 1. Same as Method 13A, Citations 1 and 2 of be a known relationship (as described above) Bibliography. between output signal and shaft rpm; also, 2. MacLeod, Kathryn E. and Howard L. each anemometer must be equipped with a Crist. Comparison of the SPADNS—Zir- suitable readout system (See Section 2.1.3). conium Lake and Specific Ion Electrode 2.1.2 Installation of Anemometers. Methods of Fluoride Determination in Stack 2.1.2.1 If the affected facility consists of a Emission Samples. Analytical Chemistry. single, isolated potroom (or potroom seg- 45:1272–1273. 1973. ment), install at least one anemometer for METHOD 14—DETERMINATION OF FLUORIDE every 85 m of roof monitor length. If the EMISSIONS FROM POTROOM ROOF MONITORS length of the roof monitor divided by 85 m is FOR PRIMARY ALUMINUM PLANTS not a whole number, round the fraction to the nearest whole number to determine the 1. Applicability and Principle number of anemometers needed. For mon- 1.1 Applicability. This method is applica- itors that are less than 130 m in length, use ble for the determination of fluoride emis- at least two anemometers. Divide the mon- sions from stationary sources only when itor cross-section into as many equal areas specified by the test procedures for deter- as anemometers and locate an anemometer mining compliance with new source perform- at the centroid of each equal area. See excep- ance standards. tion in Section 2.1.2.3. 1.2 Principle. Gaseous and particulate flu- 2.1.2.2 If the affected facility consists of oride roof monitor emissions are drawn into a permanent sampling manifold through sev- two or more potrooms (or potroom segments) eral large nozzles. The sample is transported ducted to a common control device, install from the sampling manifold to ground level anemometers in each potroom (or segment) through a duct. The gas in the duct is sam- that contains a sampling manifold. Install at pled using Method 13A or 13B—Determina- least one anemometer for every 85 m of roof tion of Total Fluoride Emissions from Sta- monitor length of the potroom (or segment). tionary Sources. Effluent velocity and volu- If the potroom (or segment) length divided metric flow rate are determined with by 85 is not a whole number, round the frac- anemometers located in the roof monitor. tion to the nearest whole number to deter- 2. Apparatus mine the number of anemometers needed. If the potroom (or segment) length is less than 2.1 Velocity Measurement Apparatus. 2.1.1 Anemometers. Propeller anemome- 130 m, use at least two anemometers. Divide ters, or equivalent. Each anemometer shall the potroom (or segment) monitor cross-sec- meet the following specifications: (1) Its pro- tion into as many equal areas as peller shall be made of polystyrene, or simi- anemometers and locate an anemometer at lar material of uniform density. To insure the centroid of each equal area. See excep- uniformity of performance among propellers, tion in Section 2.1.2.3. it is desirable that all propellers be made 2.1.2.3 At least one anemometer shall be from the same mold; (2) The propeller shall installed in the immediate vicinity (i.e., be properly balanced, to optimize perform- within 10 m) of the center of the manifold ance; (3) When the anemometer is mounted (See Section 2.2.1). For its placement in rela- horizontally, its threshold velocity shall not tion to the width of the monitor, there are exceed 15 m/min (50 fpm); (4) The measure- two alternatives. The first is to make a ve- ment range of the anemometer shall extend locity traverse of the width of the roof mon- to at least 600 m/min (2,000 fpm); (5) The ane- itor where an anemometer is to be placed
811
VerDate 11
and install the anemometer at a point of av- cient, determined as outlined in Section 4.1 erage velocity along this traverse. The tra- of Method 2. verse may be made with any suitable low ve- 2.1.6 Differential Pressure Gauge. Inclined locity measuring device, and shall be made manometer or equivalent, as described in during normal process operating conditions. Section 2.1.2 of Method 2. The second alternative, at the option of 2.2 Roof Monitor Air Sampling System. the tester, is to install the anemometer half- 2.2.1 Sampling Ductwork. A minimum of way across the width of the roof monitor. In one manifold system shall be installed for this latter case, the velocity traverse need each potroom group (as defined in Subpart S, not be conducted. Section 60.191). The manifold system and 2.1.3 Recorders. Recorders, equipped with connecting duct shall be permanently in- suitable auxiliary equipment (e.g. trans- stalled to draw an air sample from the roof monitor to ground level. A typical installa- ducers) for converting the output signal from tion of a duct for drawing a sample from a each anemometer to a continuous recording roof monitor to ground level is shown in Fig- of air flow velocity, or to an integrated ure 14–1. A plan of a manifold system that is measure of volumetric flowrate. A suitable located in a roof monitor is shown in Figure recorder is one that allows the output signal 14.2. These drawings represent a typical in- from the propeller anemometer to be read to stallation for a generalized roof monitor. within 1 percent when the velocity is be- The dimensions on these figures may be al- tween 100 and 120 m/min (350 and 400 fpm). tered slightly to make the manifold system For the purpose of recording velocity, ‘‘con- fit into a particular roof monitor, but the tinuous’’ shall mean one readout per 15- general configuration shall be followed. minute or shorter time interval. A constant There shall be eight nozzles, each having a amount of time shall elapse between read- diameter of 0.40 to 0.50 m. Unless otherwise ings. Volumetric flow rate may be deter- specified by the Administrator, the length of mined by an electrical count of anemometer the manifold system from the first nozzle to revolutions. The recorders or counters shall the eighth shall be 35 m or eight percent of permit identification of the velocities or the length of the potroom (or potroom seg- flowrate measured by each individual ane- ment) roof monitor, whichever is greater. mometer. The duct leading from the roof monitor 2.1.4 Pitot Tube. Standard-type pitot manifold shall be round with a diameter of tube, as described in Section 2.7 of Method 2, 0.30 to 0.40 m. As shown in Figure 14–2, each and having a coefficient of 0.99±0.01. of the sample legs of the manifold shall have 2.1.5 Pitot Tube (Optional). Isolated, Type a device, such as a blast gate or valve, to en- S pitot, as described in Section 2.1 of Method able adjustment of the flow into each sample 2. The pitot tube shall have a known coeffi- nozzle.
812
VerDate 11
813
VerDate 11
The manifold shall be located in the imme- Locate two sample ports in a vertical sec- diate vicinity of one of the propeller tion of the duct between the roof monitor anemometers (see Section 2.1.2.3) and as and exhaust fan. The sample ports shall be at close as possible to the midsection of the least 10 duct diameters downstream and potroom (or potroom segment). Avoid locat- three diameters upstream from any flow dis- ing the manifold near the end of a potroom turbance such as a bend or contraction. The or in a section where the aluminum reduc- two sample ports shall be situated 90° apart. tion pot arrangement is not typical of the One of the sample ports shall be situated so rest of the potroom (or potroom segment). that the duct can be traversed in the plane of Center the sample nozzles in the throat of the nearest upstream duct bend. the roof monitor (see Figure 14–1). Construct 2.2.2 Exhaust Fan. An industrial fan or all sample-exposed surfaces within the noz- blower shall be attached to the sample duct zles, manifold and sample duct of 316 stain- at ground level (see Figure 14–1). This ex- less steel. Aluminum may be used if a new haust fan shall have a capacity such that a ductwork system is conditioned with fluo- large enough volume of air can be pulled ride-laden roof monitor air for a period of six through the ductwork to maintain an weeks prior to initial testing. Other mate- isokinetic sampling rate in all the sample rials of construction may be used if it is nozzles for all flow rates normally encoun- demonstrated through comparative testing tered in the roof monitor. that there is no loss of fluorides in the sys- The exhaust fan volumetric flow rate shall tem. All connections in the ductwork shall be adjustable so that the roof monitor air be leak free. can be drawn isokinetically into the sample
814
VerDate 11
nozzles. This control of flow may be achieved 4.1.1.1 through 4.1.1.3, below. Alternatively, by a damper on the inlet to the exhauster or the tester may use any other suitable meth- by any other workable method. od, subject to the approval of the Adminis- 2.3 Temperature Measurement Apparatus. trator, that takes into account the signal 2.3.1 Thermocouple. Install a thermo- output, propeller condition and threshold ve- couple in the roof monitor near the sample locity of the anemometer. duct. The thermocouple shall conform to the 4.1.1.1 Check the signal output of the ane- specifications outlined in Section 2.3 of mometer by using an accurate rpm generator Method 2. (see Figure 14–3) or synchronous motors to 2.3.2 Signal Transducer. Transducer, to spin the propeller shaft at each of the three change the thermocouple voltage output to a rpm settings described in Section 2.1.1 above temperature readout. (specification No. 7), and measuring the out- 2.3.3 Thermocouple Wire. To reach from put signal at each setting. If, at each setting, roof monitor to signal transducer and re- the output signal is within ±5 percent of the corder. manufacturer’s value, the anemometer can 2.3.4 Recorder. Suitable recorder to mon- be used. If the anemometer performance is itor the output from the thermocouple signal unsatisfactory, the anemometer shall either transducer. be replaced or repaired. 2.4 Fluoride Sampling Train. Use the 4.1.1.2 Check the propeller condition, by train described in Method 13A or 13B. visually inspecting the propeller, making 3. Reagents note of any significant damage or warpage; damaged or deformed propellers shall be re- 3.1 Sampling and Analysis. Use reagents placed. described in Method 13A or 13B. 4.1.1.3 Check the anemometer threshold 4. Calibration velocity as follows: With the anemometer 4.1 Initial Performance Checks. Conduct mounted as shown in Figure 14–4(A), fasten a these checks within 60 days prior to the first known weight (a straight-pin will suffice) to performance test. the anemometer propeller at a fixed distance 4.1.1 Propeller Anemometers. Anemome- from the center of the propeller shaft. This ters which meet the specifications outlined will generate a known torque; for example, a in Section 2.1.1 need not be calibrated, pro- 0.1 g weight, placed 10 cm from the center of vided that a reference performance curve re- the shaft, will generate a torque of 1.0 g-cm. lating anemometer signal output to air ve- If the known torque causes the propeller to locity (covering the velocity range of inter- rotate downward, approximately 90° [see Fig- est) is available from the manufacturer. For ure 14–4(B)], then the known torque is great- the purpose of this method, a ‘‘reference’’ er than or equal to the starting torque; if the performance curve is defined as one that has propeller fails to rotate approximately 90°, been derived from primary standard calibra- the known torque is less than the starting tion data, with the anemometer mounted torque. By trying different combinations of vertically. ‘‘Primary standard’’ data are ob- weight and distance, the starting torque of a tainable by: (1) Direct calibration of one or particular anemometer can be satisfactorily more of the anemometers by the National estimated. Once an estimate of the starting Bureau of Standards (NBS); (2) NBS-trace- torque has been obtained, the threshold ve- able calibration; or (3) Calibration by direct locity of the anemometer (for horizontal measurement of fundamental parameters mounting) can be estimated from a graph such as length and time (e.g., by moving the such as Figure 14–5 (obtained from the manu- anemometers through still air at measured facturer). If the horizontal threshold veloc- rates of speed, and recording the output sig- ity is acceptable [<15 m/min (50 fpm), when nals). If a reference performance curve is not this technique is used], the anemometer can available from the manufacturer, such a be used. If the threshold velocity of an ane- curve shall be generated, using one of the mometer is found to be unacceptably high, three methods described as above. Conduct a the anemometer shall either be replaced or performance-check as outlined in Sections repaired.
815
VerDate 11
816
VerDate 11
817
VerDate 11
4.1.2 Thermocouple. Check the calibration a recorder or counter is found to be out of of the thermocouple-potentiometer system, calibration, by an average amount greater using the procedures outlined in Section 4.3 than 5 percent for the three calibration of Method 2, at temperatures of 0, 100, and points, replace or repair the system; other- 150°C. If the calibration is off by more than wise, the system can be used. 5°C at any of the temperatures, repair or re- 4.1.4 Manifold Intake Nozzles. In order to place the system; otherwise, the system can balance the flow rates in the eight individual be used. nozzles, proceed as follows: Adjust the ex- 4.1.3 Recorders and/or Counters. Check haust fan to draw a volumetric flow rate the calibration of each recorder and/or (refer to Equation 14–1) such that the en- counter (see Section 2.1.3) at a minimum of trance velocity into each manifold nozzle ap- three points, approximately spanning the ex- proximates the average effluent velocity in pected range of velocities. Use the calibra- the roof monitor. Measure the velocity of the tion procedures recommended by the manu- air entering each nozzle by inserting a stand- facturer, or other suitable procedures (sub- ard pitot tube into a 2.5 cm or less diameter ject to the approval of the Administrator). If hole (see Figure 14–2) located in the manifold
818
VerDate 11
between each blast gate (or valve) and noz- 5.1.2 Velocity Determination During a zle. Note that a standard pitot tube is used, Test Run. During the actual test run, record rather than a type S, to eliminate possible the velocity or volumetric flowrate readings velocity measurement errors due to cross- of each propeller anemometer in the roof section blockage in the small (0.13 m diame- monitor. Readings shall be taken for each ter) manifold leg ducts. The pitot tube tip anemometer every 15 minutes or at shorter shall be positioned at the center of each equal time intervals (or continuously). manifold leg duct. Take care to insure that 5.2 Temperature Recording. Record the there is no leakage around the pitot tube, temperature of the roof monitor every 2 which could affect the indicated velocity in hours during the test run. the manifold leg. If the velocity of air being 5.3 Sampling. drawn into each nozzle is not the same, open 5.3.1 Preliminary Air Flow in Duct. Dur- or close each blast gate (or valve) until the ing 24 hours preceding the test, turn on the velocity in each nozzle is the same. Fasten exhaust fan and draw roof monitor air each blast gate (or valve) so that it will re- through the manifold duct to condition the main in this position and close the pitot port ductwork. Adjust the fan to draw a volu- holes. This calibration shall be performed metric flow through the duct such that the when the manifold system is installed. Alter- velocity of gas entering the manifold nozzles natively, the manifold may be preassembled approximates the average velocity of the air and the flow rates balanced on the ground, exiting the roof monitor in the vicinity of before being installed. the sampling manifold. 4.2 Periodical Performance Checks. 5.3.2 Manifold Isokinetic Sample Rate Ad- Twelve months after their initial installa- justment(s). tion, check the calibration of the propeller 5.3.2.1 Initial Adjustment. Prior to the anemometers, thermocouple-potentiometer test run (or first sub-run, if applicable; see system, and the recorders and/or counters as Sections 5.1.1 and 5.3.2.2), adjust the fan to in Section 4.1. If the above systems pass the provide the necessary volumetric flowrate in performance checks, (i.e., if no repair or re- the sampling duct, so that air enters the placement of any component is necessary), manifold sample nozzles at a velocity equal continue with the performance checks on a to the appropriate estimated average veloc- 12-month interval basis. However, if any of ity determined under Section 5.1.1. Equation the above systems fail the performance 14–1 gives the correct stream velocity needed checks, repair or replace the system(s) that in the duct at the sampling location, in order failed and conduct the periodical perform- for sample gas to be drawn isokinetically ance checks on a 3-month interval basis, into the manifold nozzles. Next, verify that until sufficient information (consult with the correct stream velocity has been the Administrator) is obtained to establish a achieved, by performing a pitot tube traverse modified performance check schedule and of the sample duct (using either a standard calculation procedure. or type S pitot tube); use the procedure out- NOTE: If any of the above systems fail the lined in Method 2. initial performance checks, the data for the past year need not be recalculated. 2 8()D 1min 5. Procedure v = n ()v Eq.14- 1 5.1 Roof Monitor Velocity Determination. d ()2 m 60sec 5.1.1 Velocity Estimate(s) for Setting Dd Isokinetic Flow. To assist in setting Where: isokinetic flow in the manifold sample noz- v =Desired velocity in duct at sampling loca- zles, the anticipated average velocity in the d tion, m/sec. section of the roof monitor containing the sampling manifold shall be estimated prior Dn=Diameter of a roof monitor manifold noz- to each test run. The tester may use any zle, m. convenient means to make this estimate Dd=Diameter of duct at sampling location, (e.g., the velocity indicated by the anemom- m. eter in the section of the roof monitor con- vm=Average velocity of the air stream in the taining the sampling manifold may be con- roof monitor, m/min, as determined tinuously monitored during the 24-hour pe- under Section 5.1.1. riod prior to the test run). 5.3.2.2 Adjustments During Run. If the If there is question as to whether a single test run is divided into two or more ‘‘sub- estimate of average velocity is adequate for runs’’ (see Section 5.1.1), additional an entire test run (e.g., if velocities are an- isokinetic rate adjustment(s) may become ticipated to be significantly different during necessary during the run. Any such adjust- different potroom operations), the tester ment shall be made just before the start of a may opt to divide the test run into two or sub-run, using the procedure outlined in Sec- more ‘‘sub-runs,’’ and to use a different esti- tion 5.3.2.1 above. mated average velocity for each sub-run (see NOTE: Isokinetic rate adjustments are not Section 5.3.2.2.) permissible during a sub-run.
819
VerDate 11
5.3.3 Sample Train Operation. Sample the 6.1.4.1 If vs is less than or equal to 120 per- duct using the standard fluoride train and cent of vd, the results are acceptable (note methods described in Methods 13A and 13B. that in cases where the above calculations Determine the number and location of the have been performed for each sub-run, the re- sampling points in accordance with Method sults are acceptable if the average percent- 1. A single train shall be used for the entire age for all sub-runs is less than or equal to sampling run. Alternatively, if two or more 120 percent). sub-runs are performed, a separate train may 6.1.4.2 If vs is more than 120 percent of vd, be used for each sub-run; note, however, that multiply the reported emission rate by the if this option is chosen, the area of the sam- following factor. pling nozzle shall be the same (± 2 percent) for each train. If the test run is divided into ()100v/ v − 120 sub-runs, a complete traverse of the duct 1+ s d shall be performed during each sub-run. 200 5.3.4 Time Per Run. Each test run shall 6.2 Average Velocity of Roof Monitor last 8 hours or more; if more than one run is Gases. Calculate the average roof monitor to be performed, all runs shall be of approxi- velocity using all the velocity or volumetric mately the same (± 10 percent) length. If flow readings from Section 5.1.2. question exists as to the representativeness 6.3 Roof Monitor Temperature. Calculate of an 8-hour test, a longer period should be the mean value of the temperatures recorded selected. Conduct each run during a period in Section 5.2. when all normal operations are performed 6.4 Concentration of Fluorides in Roof underneath the sampling manifold. For most Monitor Air. recently-constructed plants, 24 hours are re- 6.4.1 If a single sampling train was used quired for all potroom operations and events throughout the run, calculate the average to occur in the area beneath the sampling fluoride concentration for the roof monitor manifold. During the test period, all pots in using Equation 13A–2 of Method 13A. the potroom group shall be operated such 6.4.2 If two or more sampling trains were that emissions are representative of normal used (i.e., one per sub-run), calculate the av- operating conditions in the potroom group. erage fluoride concentration for the run, as 5.3.5 Sample Recovery. Use the sample re- follows: covery procedure described in Method 13A or 13B. n 5.4 Analysis. Use the analysis procedures ∑ described in Method 13A or 13B. − ()F = i=1 t i 6. Calculations CS n Eq.14- 2 6.1 Isokinetic Sampling Check. ()Vm() std ∑ i 6.1.1 Calculate the mean velocity (vm) for the sampling run, as measured by the ane- i=l mometer in the section of the roof monitor Where: containing the sampling manifold. If two or ¯ Cs=Average fluoride concentration in roof more sub-runs have been performed, the test- monitor air, mg F/dscm (mg F/dscf). er may opt to calculate the mean velocity Ft=Total fluoride mass collected during a for each sub-run. particular sub-run, mg F (from Equation 6.1.2 Using Equation 14–1, calculate the 13A–1 of Method 13A or Equation 13B–1 of expected average velocity (vd) in the sam- Method 13B). pling duct, corresponding to each value of vm Vm(std)=Total volume of sample gas passing obtained under Section 6.1.1. through the dry gas meter during a par- 6.1.3 Calculate the actual average veloc- ticular sub-run, dscm (dscf) (see Equa- ity (vs) in the sampling duct for each run or tion 5–1 of Method 5). sub-run, according to Equation 2–9 of Method n=Total number of sub-runs. 2, and using data obtained from Method 13. 6.5 Average volumetric flow from the roof 6.1.4 Express each value vs from Section monitor of the potroom(s) (or potroom seg- 6.1.3 as a percentage of the corresponding vd ment(s)) containing the anemometers is value from Section 6.1.2. given in Equation 14–3.
o VMPKAmt d m ()293 Q = Eq.14- 3 sd + o ()tm 273() 760 mm Hg
820
VerDate 11
Where: cific cassette arrangement. The cassettes are connected by tubing to flowmeters and a Qsd=Average volumetric flow from roof mon- itor at standard conditions on a dry manifold system that allows for the equal basis, m3/min. distribution of volume pulled through each A=Roof monitor open area, m2. cassette, and finally to a dry gas meter. The vmt=Average velocity of air in the roof mon- cassettes have a specific internal arrange- itor, m/min, from Section 6.2. ment of one unaltered cellulose filter and Pm=Pressure in the roof monitor; equal to support pad in the first section of the cas- barometric pressure for this application, sette for solid fluoride retention and two cel- mm Hg. lulose filters with support pads that are im- tm=Roof monitor temperature, °C, from Sec- pregnated with sodium formate for the tion 6.3. chemical absorption of gaseous fluorides in Md=Mole fraction of dry gas, which is given the following two sections of the cassette. A by: minimum of eight cassettes shall be used for Md=(1Ø Bws) a potline and shall be strategically located at equal intervals across the potroom roof so NOTE: B is the proportion by volume of ws as to encompass a minimum of 8 percent of water vapor in the gas stream, from Equa- the total length of the potroom. A greater tion 5–3, Method 5. number of cassettes may be used should the 6.6 Conversion Factors. regulated facility choose to do so. The mass 1 ft3=0.02832 m3 flow rate of pollutants is determined with 1 hr=60 min anemometers and temperature sensing de- 7. Bibliography vices located immediately below the opening 1. Shigehara, R. T., A Guideline for Evalu- of the roof monitor and spaced evenly within ating Compliance Test Results (Isokinetic the cassette group. 3.0 Definitions Sampling Rate Criterion). U.S. Environ- 3.1 Cassette. A segmented, styrene acrylo- mental Protection Agency, Emission Meas- nitrile cassette configuration with three sep- urement Branch. Research Triangle Park, arate segments and a base, for the purpose of NC. August 1977. this method, to capture and retain fluoride METHOD 14A—DETERMINATION OF TOTAL FLU- from potroom gases. ORIDE EMISSIONS FROM SELECTED SOURCES 3.2 Cassette arrangement. The cassettes, AT PRIMARY ALUMINUM PRODUCTION FACILI- tubing, manifold system, flowmeters, dry gas TIES meter, and any other related equipment as- sociated with the actual extraction of the NOTE: This method does not include all the sample gas stream. specifications (e.g., equipment and supplies) 3.3 Cassette group. That section of the and procedures (e.g., sampling) essential to potroom roof monitor where a distinct group its performance. Some material is incor- of cassettes is located. porated by reference from other methods in 3.4 Potline. A single, discrete group of this part. Therefore, to obtain reliable re- electrolytic reduction cells electrically con- sults, persons using this method should have nected in series, in which alumina is reduced a thorough knowledge of at least the follow- to form aluminum. ing additional test methods: Method 5, Meth- 3.5 Potroom. A building unit that houses ods 13A and 13B, and Method 14 of this appen- a group of electrolytic reduction cells in dix. which aluminum is produced. 1.0 Scope and Application 3.6 Potroom group. An uncontrolled 1.1 Analytes. potroom, a potroom that is controlled indi- vidually, or a group of potrooms or potroom Analyte CAS No. Sensitivity segments ducted to a common primary con- Total fluorides ...... None assigned ..... Not determined. trol system. Includes hydrogen 007664±39±3 ...... Not determined. 3.7 Primary control system. The equip- fluoride. ment used to capture the gases and particu- late matter generated during the reduction 1.2 Applicability. This method is applica- process and the emission control device(s) ble for the determination of total fluorides used to remove pollutants prior to discharge (TF) emissions from sources specified in the of the cleaned gas to the atmosphere. applicable regulation. This method was de- 3.8 Roof monitor. That portion of the roof veloped by consensus with the Aluminum As- of a potroom building where gases, not cap- sociation and the U.S. Environmental Pro- tured at the cell, exit from the potroom. tection Agency (EPA). 3.9 Total fluorides (TF). Elemental fluo- 2.0 Summary of Method rine and all fluoride compounds as measured 2.1 Total fluorides, in the form of solid by Methods 13A or 13B of this appendix or by and gaseous fluorides, are withdrawn from an approved alternative method. the ascending air stream inside of an alu- 4.0 Interferences and Known Limitations minum reduction potroom and, prior to 4.1 There are two principal categories of exiting the potroom roof monitor, into a spe- limitations that must be addressed when
821
VerDate 11
using this method. The first category is sam- nate. Treat residual chemical burn as ther- pling bias and the second is analytical bias. mal burn. Biases in sampling can occur when there is 5.3 Sodium Hydroxide (NaOH). Causes se- an insufficient number of cassettes located vere damage to eyes and skin. Inhalation along the roof monitor of a potroom or if the causes irritation to nose, throat, and lungs. distribution of those cassettes is spatially Reacts exothermically with limited amounts unequal. Known sampling biases also can of water. occur when there are leaks within the cas- 5.4 Perchloric Acid (HClO ). Corrosive to sette arrangement and if anemometers and 4 eyes, skin, nose, and throat. Provide ventila- temperature devices are not providing accu- rate data. Applicable instruments must be tion to limit exposure. Very strong oxidizer. properly calibrated to avoid sampling bias. Keep separate from water and oxidizable ma- Analytical biases can occur when instrumen- terials to prevent vigorous evolution of heat, tation is not calibrated or fails calibration spontaneous combustion, or explosion. Heat and the instrument is used out of proper solutions containing HClO4 only in hoods calibration. Additionally, biases can occur in specifically designed for HClO4. the laboratory if fusion crucibles retain re- 6.0 Equipment and Supplies sidual fluorides over lengthy periods of use. 6.1 Sampling. This condition could result in falsely ele- 6.1.1 Cassette arrangement. The cassette vated fluoride values. Maintaining a clean itself is a three-piece, styrene acrylonitrile work environment in the laboratory is cru- cassette unit (a Gelman Sciences product), 37 cial to producing accurate values. millimeter (mm), with plastic connectors. In 4.2 Biases during sampling can be avoided the first section (the intake section), an un- by properly spacing the appropriate number treated Gelman Sciences 37 mm, 0.8 microm- of cassettes along the roof monitor, conduct- eter (µm) DM–800 metricel membrane filter ing leak checks of the cassette arrangement, and cellulose support pad, or equivalent, is calibrating the dry gas meter every 30 days, verifying the accuracy of individual flow- situated. In the second and third segments of meters (so that there is no more than 5 per- the cassette there is placed one each of cent difference in the volume pulled between Gelman Sciences 37 mm, 5 µm GLA–5000 low- any two flowmeters), and calibrating or re- ash PVC filter with a cellulose support pad placing anemometers and temperature sens- or equivalent product. Each of these two fil- ing devices as necessary to maintain true ters and support pads shall have been im- data generation. mersed in a solution of 10 percent sodium 4.3 Analytical biases can be avoided by formate (volume/volume in an ethyl alcohol calibrating instruments according to the solution). The impregnated pads shall be manufacturer’s specifications prior to con- placed in the cassette segments while still ducting any analyses, by performing internal wet and heated at 50°C (122°F) until the pad and external audits of up to 10 percent of all is completely dry. It is important to check samples analyzed, and by rotating individual for a proper fit of the filter and support pad crucibles as the ‘‘blank’’ crucible to detect to the cassette segment to ensure that there any potential residual fluoride carry-over to are no areas where gases could bypass the fil- samples. Should any contamination be dis- ter. Once all of the cassette segments have covered in the blank crucible, the crucible been prepared, the cassette shall be assem- shall be thoroughly cleaned to remove any bled and a plastic plug shall be inserted into detected residual fluorides and a ‘‘blank’’ the exhaust hole of the cassette. Prior to analysis conducted again to evaluate the ef- placing the cassette into service, the space fectiveness of the cleaning. The crucible between each segment shall be taped with an shall remain in service as long as no detect- appropriately durable tape to prevent the in- able residual fluorides are present. filtration of gases through the points of con- 5.0 Safety nection, and an aluminum nozzle shall be in- 5.1 This method may involve the handling serted into the intake hole of the cassette. of hazardous materials in the analytical phase. This method does not purport to ad- The aluminum nozzle shall have a short sec- dress all of the potential safety hazards asso- tion of tubing placed over the opening of the ciated with its use. It is the responsibility of nozzle, with the tubing plugged to prevent the user to establish appropriate safety and dust from entering the nozzle and to prepare health practices and determine the applica- the nozzle for the cassette arrangement leak bility of regulatory limitations prior to per- check. An alternate nozzle type can be used forming this test method. if historical results or scientific demonstra- 5.2 Corrosive reagents. The following re- tion of applicability can be shown. agents are hazardous. Personal protective 6.1.2 Anemometers and temperature sens- equipment and safe procedures are useful in ing devices. To calculate the mass flow rate preventing chemical splashes. If contact oc- of TF from the roof monitor under standard curs, immediately flush with copious conditions, anemometers that meet the spec- amounts of water for at least 15 minutes. Re- ifications in section 2.1.1 in Method 14 of this move clothing under shower and decontami- appendix or an equivalent device yielding
822
VerDate 11
equivalent information shall be used. A re- The manifold system is connected to a dry cording mechanism capable of accurately re- gas meter (Research Appliance Company cording the exit gas temperature at least model 201009 or equivalent). The length of every 2 hours shall be used. tubing is managed by pneumatically or elec- 6.1.3 Barometer. To correct the volu- trically operated hoists located in the roof metric flow from the potline roof monitor to monitor, and the travel of the tubing is con- standard conditions, a mercury (Hg), aner- trolled by encasing the tubing in aluminum oid, or other barometer capable of measuring conduit. The tubing is lowered for cassette atmospheric pressure to within 2.5 mm [0.1 insertion by operating a control box at floor inch (in)] Hg shall be used. level. Once the cassette has been securely in- serted into the tubing and the leak check NOTE: The barometric reading may be ob- performed, the tubing and cassette are raised tained from a nearby National Weather Serv- to the roof monitor level using the floor ice Station. In this case, the station value level control box. Arrangements similar to (which is absolute barometric pressure) shall the one described are acceptable if the sci- be requested and an adjustment for elevation entific sample collection principles are fol- differences between the weather station and lowed. the sampling point shall be made at a rate of minus 2.5 mm (0.1 in) Hg per 30 meters (m) 8.2 Test run sampling period. A test run [100 feet (ft)] elevation increase or plus 2.5 shall comprise a minimum of a 24-hour sam- mm (0.1 in) Hg per 30 m (100 ft) elevation de- pling event encompassing at least eight cas- crease. settes per potline (or four cassettes per 6.2 Sample recovery. potroom group). Monthly compliance shall 6.2.1 Hot plate. be based on three test runs during the 6.2.2 Muffle furnace. month. Test runs of greater than 24 hours 6.2.3 Nickel crucible. are allowed; however, three such runs shall 6.2.4 Stirring rod. Teflon’. be conducted during the month. 6.2.5 Volumetric flask. 50-milliliter (ml). 8.3 Leak-check procedures. 6.2.6 Plastic vial. 50-ml. 8.3.1 Pretest leak check. A pretest leak- 6.3 Analysis. check is recommended; however, it is not re- 6.3.1 Primary analytical method. An auto- quired. To perform a pretest leak-check after mated analyzer having the following compo- the cassettes have been inserted into the nents or equivalent: a multichannel propor- tubing, isolate the cassette to be leak- tioning pump, multiposition sampler, volt- checked by turning the valves on the mani- age stabilizer, colorimeter, instrument re- fold to stop all flows to the other sampling cording device, microdistillation apparatus, points connected to the manifold and meter. flexible Teflon heating bath, vacuum pump, The cassette, with the plugged tubing sec- pulse suppressers and an air flow system. tion securing the intake of the nozzle, is sub- 6.3.2 Secondary analytical method. Spe- jected to the highest vacuum expected dur- cific Ion Electrode (SIE). ing the run. If no leaks are detected, the tub- 7.0 Reagents and Standards ing plug can be briefly removed as the dry 7.1 Water. Deionized distilled to conform gas meter is rapidly turned off. to ASTM Specification D 1193–77, Type 3 (in- 8.3.2 Post-test leak check. A leak check is corporated by reference in § 60.17(a)(22) of required at the conclusion of each test run this part). The KMnO4 test for oxidizable or- for each cassette. The leak check shall be ganic matter may be omitted when high con- performed in accordance with the procedure centrations of organic matter are not ex- outlined in section 8.3.1 of this method ex- pected to be present. cept that it shall be performed at a vacuum 7.2 Calcium oxide. greater than the maximum vacuum reached 7.3 Sodium hydroxide (NaOH). Pellets. during the test run. If the leakage rate is 7.4 Perchloric acid (HClO4). Mix 1:1 with found to be no greater than 4 percent of the water. Sulfuric acid (H2SO4) may be used in average sampling rate, the results are ac- place of HClO4. ceptable. If the leakage rate is greater than 7.5 Audit samples. The audit samples dis- 4 percent of the average sampling rate, ei- cussed in section 9.1 shall be prepared from ther record the leakage rate and correct the reagent grade, water soluble stock reagents, sampling volume as discussed in section 12.4 or purchased as an aqueous solution from a of this method or void the test run if the commercial supplier. If the audit stock solu- minimum number of cassettes were used. If tion is purchased from a commercial sup- the number of cassettes used was greater plier, the standard solution must be accom- than the minimum required, discard the panied by a certificate of analysis or an leaking cassette and use the remaining cas- equivalent proof of fluoride concentration. settes for the emission determination. 8.0 Sample Collection and Analysis 8.3.3 Anemometers and temperature sens- 8.1 Preparing cassette arrangement for ing device placement. Install the recording sampling. The cassettes are initially con- mechanism to record the exit gas tempera- nected to flexible tubing. The tubing is con- ture. Anemometers shall be installed as re- nected to flowmeters and a manifold system. quired in section 6.1.2 of Method 14 of this
823
VerDate 11
appendix, except replace the word ‘‘mani- evaporated and allow the mixture to gradu- fold’’ with ‘‘cassette group’’ in section 6.1.2.3. ally char for 1 hour. These two different instruments shall be lo- 8.5.3 Transfer the crucible to a cold muffle cated near each other along the roof mon- furnace and ash at 600°C (1,112°F). Remove itor. See conceptual configurations in Fig- the crucible after the ashing phase and, after ures 14A–1, 14A–2, and 14A–3 of this method. the crucible cools, add 3.0 g (±0.1 g) of NaOH Fewer temperature devices than pellets. Place this mixture in a muffle fur- anemometers may be used if at least one nace at 600°C (1,112°F) for 3 minutes. Remove temperature device is located within the the crucible and roll the melt so as to reach span of the cassette group. Other anemom- all of the ash with the molten NaOH. Let the eter location siting scenarios may be accept- melt cool to room temperature. Add 10 to 15 able as long as the exit velocity of the roof ml of water to the crucible and place it on a monitor gases is representative of the entire hot plate at a low temperature setting until section of the potline being sampled. the melt is soft or suspended. Transfer the 8.4 Sampling. The actual sample run shall contents of the crucible to a 50-ml volu- begin with the removal of the tubing and metric flask. Rinse the crucible with 20 ml of plug from the cassette nozzle. Each cassette 1:1 perchloric acid or 20 ml of 1:1 sulfuric acid is then raised to the roof monitor area, the in two (2) 10 ml portions. Pour the acid rinse dry gas meter is turned on, and the flow- slowly into the volumetric flask and swirl meters are set to the calibration point, the flask after each addition. Cool to room which allows an equal volume of sampled gas temperature. The product of this procedure to enter each cassette. The dry gas meter is particulate fluorides. shall be set to a range suitable for the spe- 8.5.4 Gaseous fluorides can be isolated for cific potroom type being sampled that will analysis by folding the gaseous fluoride fil- yield valid data known from previous experi- ters and support pads to approximately 1⁄4 of ence or a range determined by the use of the their original size and placing them in a 50- calculation in section 12 of this method. Pa- ml plastic vial. To the vial add exactly 10 ml rameters related to the test run that shall be of water and leach the sample for a mini- recorded, either during the test run or after mum of 1 hour. The leachate from this proc- the test run if recording devices are used, in- ess yields the gaseous fluorides for analysis. clude: anemometer data, roof monitor exit 9.0 Quality Control gas temperature, dry gas meter temperature, 9.1 Laboratory auditing. Laboratory au- dry gas meter volume, and barometric pres- dits of specific and known concentrations of sure. At the conclusion of the test run, the fluoride shall be submitted to the laboratory cassettes shall be lowered, the dry gas meter with each group of samples submitted for turned off, and the volume registered on the analysis. An auditor shall prepare and dry gas meter recorded. The post-test leak present the audit samples as a ‘‘blind’’ eval- check procedures described in section 8.3.2 of uation of laboratory performance with each this method shall be performed. All data rel- group of samples submitted to the labora- evant to the test shall be recorded on a field tory. The audits shall be prepared to rep- data sheet and maintained on file. resent concentrations of fluoride that could 8.5 Sample recovery. be expected to be in the low, medium and 8.5.1 The cassettes shall be brought to the high range of actual results. Average recov- laboratory with the intake nozzle contents eries of all three audits must equal 90 to 110 protected with the section of plugged tubing percent for acceptable results; otherwise, the previously described. The exterior of cas- laboratory must investigate procedures and settes shall carefully be wiped free of any instruments for potential problems. dust or debris, making sure that any falling dust or debris does not present a potential NOTE: The analytical procedure allows for laboratory contamination problem. the analysis of individual or combined filters 8.5.2 Carefully remove all tape from the and pads from the cassettes provided that cassettes and remove the initial filter, sup- equal volumes (±10 percent) are sampled port pad, and all loose solids from the first through each cassette. (intake) section of the cassette. Fold the fil- 10.0 Calibrations ter and support pad several times and, along 10.1 Equipment evaluations. To ensure with all loose solids removed from the inte- the integrity of this method, periodic cali- rior of the first section of the cassette, place brations and equipment replacements are them into a nickel crucible. Using water, necessary. wash the interior of the nozzle into the same 10.1.1 Metering system. At 30-day inter- nickel crucible. Add 0.1 gram (g) [±0.1 milli- vals the metering system shall be calibrated. gram (mg)] of calcium oxide and a sufficient Connect the metering system inlet to the amount of water to make a loose slurry. Mix outlet of a wet test meter that is accurate to the contents of the crucible thoroughly with 1 percent. Refer to Figure 5–4 of Method 5 of a Teflon’’ stirring rod. After rinsing any ad- this appendix. The wet-test meter shall have hering residue from the stirring rod back a capacity of 30 liters/revolution [1 cubic foot into the crucible, place the crucible on a hot (ft3)/revolution]. A spirometer of 400 liters (14 plate or in a muffle furnace until all liquid is ft3) or more capacity, or equivalent, may be
824
VerDate 11
used for calibration; however, a wet-test meter. The connection is then made to the meter is usually more practical. The wet- wet-test meter and finally to a dry gas test meter shall be periodically tested with a meter. All connections are made with tub- spirometer or a liquid displacement meter to ing. ensure the accuracy. Spirometers or wet-test 10.1.4.2 Turn the dry gas meter on for 15 meters of other sizes may be used, provided min. in preparation for the calibration. Turn that the specified accuracies of the proce- the dry gas meter off and plug the intake dure are maintained. Run the metering sys- hole of the cassette. Turn the dry gas meter tem pump for about 15 min. with the orifice back on to evaluate the entire system for manometer indicating a median reading as leaks. If the dry gas meter shows a leakage expected in field use to allow the pump to rate of less than 0.02 ft3/min at 10 in. of Hg warm up and to thoroughly wet the interior vacuum as noted on the dry gas meter, the of the wet-test meter. Then, at each of a system is acceptable to further calibration. minimum of three orifice manometer set- 10.1.4.3 With the dry gas meter turned on tings, pass an exact quantity of gas through and the flow indicator ball at a selected flow the wet-test meter and record the volume in- rate, record the exact amount of gas pulled dicated by the dry gas meter. Also record the through the flowmeter by taking measure- barometric pressure, the temperatures of the ments from the wet test meter after exactly wet test meter, the inlet temperatures of the 10 min. Record the room temperature and dry gas meter, and the temperatures of the barometric pressure. Conduct this test for all outlet of the dry gas meter. Record all cali- flowmeters in the system with all flow- bration data on a form similar to the one meters set at the same indicator ball read- shown in Figure 5–5 of Method 5 of this ap- ing. When all flowmeters have gone through pendix and calculate Y, the dry gas meter the procedure above, correct the volume calibration factor, and ∆H@, the orifice cali- pulled through each flowmeter to standard bration factor at each orifice setting. Allow- conditions. The acceptable difference be- able tolerances for Y and ∆H@ are given in tween the highest and lowest flowmeter rate Figure 5–6 of Method 5 of this appendix. is 5 percent. Should one or more flowmeters 10.1.2 Estimating volumes for initial test be outside of the acceptable limit of 5 per- runs. For a facility’s initial test runs, the cent, repeat the calibration procedure at a regulated facility must have a target or de- lower or higher indicator ball reading until sired volume of gases to be sampled and a all flowmeters show no more than 5 percent target range of volumes to use during the difference among them. calibration of the dry gas meter. Use Equa- 10.1.4.4 This flowmeter calibration shall tions 14A–1 and 14A–2 in section 12 of this be conducted at least once per year. method to derive the target dry gas meter 10.1.5 Miscellaneous equipment calibra- volume (Fv) for these purposes. tions. Miscellaneous equipment used such as 10.1.3 Calibration of anemometers and an automatic recorder/ printer used to meas- temperature sensing devices. If the standard ure dry gas meter temperatures shall be cali- anemometers in Method 14 of this appendix brated according to the manufacturer’s spec- are used, the calibration and integrity eval- ifications in order to maintain the accuracy uations in sections 10.3.1.1 through 10.3.1.3 of of the equipment. Method 14 of this appendix shall be used as 11.0 Analytical Procedure well as the recording device described in sec- 11.1 The preferred primary analytical de- tion 2.1.3 of Method 14. The calibrations or termination of the individual isolated sam- complete change-outs of anemometers shall ples or the combined particulate and gaseous take place at a minimum of once per year. samples shall be performed by an automated The temperature sensing and recording de- methodology. The analytical method for this vices shall be calibrated according to the technology shall be based on the manufac- manufacturer’s specifications. turer’s instructions for equipment operation 10.1.4 Calibration of flowmeters. The cali- and shall also include the analysis of five bration of flowmeters is necessary to ensure standards with concentrations in the ex- that an equal volume of sampled gas is en- pected range of the actual samples. The re- tering each of the individual cassettes and sults of the analysis of the five standards that no large differences, which could pos- shall have a coefficient of correlation of at sibly bias the sample, exist between the cas- least 0.99. A check standard shall be analyzed settes. as the last sample of the group to determine 10.1.4.1 Variable area, 65 mm flowmeters if instrument drift has occurred. The accept- or equivalent shall be used. These flow- able result for the check standard is 95 to 105 meters can be mounted on a common base percent of the standard’s true value. for convenience. These flowmeters shall be 11.2 The secondary analytical method calibrated by attaching a prepared cassette, shall be by specific ion electrode if the sam- complete with filters and pads, to the flow- ples are distilled or if a TISAB IV buffer is meter and then to the system manifold. This used to eliminate aluminum interferences. manifold is an aluminum cylinder with Five standards with concentrations in the valved inlets for connections to the flow- expected range of the actual samples shall be meters/cassettes and one outlet to a dry gas analyzed, and a coefficient of correlation of
825
VerDate 11
at least 0.99 is the minimum acceptable limit 12.1 Carry out calculations, retaining at for linearity. An exception for this limit for least one extra decimal point beyond that of linearity is a condition when low-level the acquired data. Round off values after the standards in the range of 0.01 to 0.48 µg fluo- final calculation. Other forms of calculations ride/ml are analyzed. In this situation, a may be used as long as they give equivalent minimum coefficient of correlation of 0.97 is results. required. TISAB II shall be used for low-level 12.2 Estimating volumes for initial test analyses. runs. 12.0 Data Analysis and Calculations
()FX() = d Fv Eq. 14A-1 Fe
Where X = Number of cassettes used.
Fv = Desired volume of dry gas to be sam- Fe = Typical concentration of TF in emis- pled, ft3. sions to be sampled, µg/ft 3, calculated Fd = Desired or analytically optimum mass from Equation 14A–2. of TF per cassette, micrograms of TF per cassette (µg/cassette).
()R() R()4./ 536× 108 µ g lb = e p Fe Eq. 14A-2 (AVr )( r )
Where 12.2.1 Example calculation. Assume that the typical emission rate (Re) is 1.0 lb TF/ton Re = Typical emission rate from the facility, pounds of TF per ton (lb/ton) of alu- of aluminum, the typical roof vent gas exit velocity (V ) is 250 ft/min, the typical produc- minum. r tion rate (Rp) is 0.10 ton/min, the known open Rp = Typical production rate of the facility, 2 area for the roof monitor (Ar) is 8,700 ft , and tons of aluminum per minute (ton/min). the desired (analytically optimum) mass of Vr = Typical exit velocity of the roof monitor TF per cassette is 1,500 µg. First calculate gases, feet per minute (ft/min). the concentration of TF per cassette (Fe) in Ar=Open area of the roof monitor, square feet µg/ft3 using Equation 14A–2. Then calculate 2 (ft ). the desired volume of gas to be sampled (Fv) using Equation 14A–1.
(1. 0lb/ton )( 0 . 1tons /min )() 4 . 536× 108 µ g / lb = = Fe 20. 855 Eq.14 A- 3 ()8, 700ft2 () 250 ft /min
()1, 500µg() 8 cassettes =3 = Fv 575.40 ft Eq.14 A-4 ()20./ 855 µg ft3
This is a total of 575.40 ft3 for eight cas- 12.3.1 Obtain a standard cubic feet (scf) settes or 71.925 ft3/cassette. value for the volume pulled through the dry 12.3 Calculations of TF emissions from gas meter for all cassettes by using the field field and laboratory data that would yield a and calibration data and Equation 5–1 of production related emission rate can be cal- Method 5 of this appendix. culated as follows:
826
VerDate 11
12.3.2 Derive the average quantity of TF corrected dry gas meter volume for each cas- 3 per cassette (in µg TF/cassette) by adding all sette; this value then becomes TFstd (µg/ft ). laboratory data for all cassettes and dividing 12.3.3 Calculate the production-based this value by the total number of cassettes emission rate (Re) in lb/ton using Equation used. Divide this average TF value by the 14A–5.
− ()TF( V )( A )()2./ 2× 10 9 lbµ g = std r r Re Eq.14 A-5 ()R p
2 12.3.4 As an example calculation, assume monitor for the potline (Ar) is 17,400 ft . The eight cassettes located in a potline were used exit velocity of the roof monitor gases (Vr) is to sample for 72 hours during the run. The 250 ft/min. The production rate of aluminum analysis of all eight cassettes yielded a total over the previous 720 hours was 5,000 tons, of 3,000 µg of TF. The dry gas meter volume which is 6.94 tons/hr or 0.116 ton/min (Rp). was corrected to yield a total of 75 scf per Substituting these values into Equation 14A– cassette, which yields a value for TFstd of 5 yields: 3,000/75=5 µg/ft3. The open area of the roof
− ()5µg/ ft3() 250 ft /min() 17 , 400 ft 2() 2 . 2× 10 9 lb / µ g R = Eq.14 A-6 e ()0. 116 ton /min = Re 0.41lb / ton of aluminum produced. Eq.14A-7
12.4 Corrections to volumes due to leak- 8.3.2 of this method, correct the volume as age. Should the post-test leak check leakage detailed in Case I in section 6.3 of Method 5 rate exceed 4 percent as described in section of this appendix.
827
VerDate 11
828
VerDate 11
829
VerDate 11
830
VerDate 11
METHOD 15—DETERMINATION OF HYDROGEN eluting CO and CO2 before any of the sulfur SULFIDE, CARBONYL SULFIDE, AND CARBON compounds to be measured.) Compliance DISULFIDE EMISSIONS FROM STATIONARY with this requirement can be demonstrated SOURCES by submitting chromatograms of calibration gases with and without CO in the diluent Introduction 2 gas. The CO2 level should be approximately The method described below uses the prin- 10 percent for the case with CO2 present. The ciple of gas chromatographic separation and two chromatograms should show agreement flame photometric detection (FPD). Since within the precision limits of Section 4.1. there are many systems or sets of operating 3.3 Elemental Sulfur. The condensation of conditions that represent useable methods of sulfur vapor in the sampling system can lead determining sulfur emissions, all systems to blockage of the particulate filter. This which employ this principle, but differ only problem can be minimized by observing the in details of equipment and operation, may filter for buildup and changing as needed. be used as alternative methods, provided 3.4 Sulfur Dioxide (SO2). Sulfur dioxide is that the calibration precision and sample- not a specific interferent but may be present line loss criteria are met. in such large amounts that it cannot be ef- 1. Principle and Applicability fectively separated from the other com- 1.1 Principle. A gas sample is extracted pounds of interest. The SO2 scrubber de- from the emission source and diluted with scribed in Section 5.1.3 will effectively re- clean dry air. An aliquot of the diluted sam- move SO2 from the sample. ple is then analyzed for hydrogen sulfide 3.5 Alkali Mist. Alkali mist in the emis- (H2S), carbonyl sulfide (COS), and carbon di- sions of some control devices may cause a sulfide (CS2) by gas chromatographic (GC) rapid increase in the SO2 scrubber pH to give separation and flame photometric detection low sample recoveries. Replacing the SO2 (FPD). scrubber contents after each run will mini- 1.2 Applicability. This method is applicable mize the chances of interference in these for determination of the above sulfur com- cases. pounds from tail gas control units of sulfur 4. Precision recovery plants. 4.1 Calibration Precision. A series of three 2. Range and Sensitivity consecutive injections of the same calibra- 2.1 Range. Coupled with a gas chromto- tion gas, at any dilution, shall produce re- graphic system utilizing a 1-milliliter sam- sults which do not vary by more than ±13 ple size, the maximum limit of the FPD for percent from the mean of the three injec- each sulfur compound is approximately 10 tions. ppm. It may be necessary to dilute gas sam- 4.2 Calibration Drift. The calibration drift ples from sulfur recovery plants hundredfold determined from the mean of three injec- (99:1) resulting in an upper limit of about tions made at the beginning and end of any 1000 ppm for each compound. run or series of runs within a 24-hour period 2.2 Sensitivity. The minimum detectable shall not exceed ±5 percent. concentration of the FPD is also dependent 5. Apparatus on sample size and would be about 0.5 ppm for a 1 ml sample. 5.1 Sampling (Figure 15–1). 5.1.1 Probe. The probe shall be made of 3. Interferences Teflon or Teflon-lined stainless steel and 3.1 Moisture Condensation. Moisture con- heated to prevent moisture condensation. It densation in the sample delivery system, the shall be designed to allow calibration gas to analytical column, or the FPD burner block enter the probe at or near the sample point can cause losses or interferences. This poten- entry. Any portion of the probe that con- tial is eliminated by heating the probe, filter tacts the stack gas must be heated to pre- box, and conncections, and by maintaining vent moisture condensation. The probe de- the SO2 scrubber in an ice water bath. Mois- scribed in Section 2.1.1 of Method 16A having ture is removed in the SO2 scrubber and a nozzle directed away from the gas stream heating the sample beyond this point is not is recommended for sources having particu- necessary provided the ambient temperature late or mist emissions. Where very high ° is above 0 C. Alternatively, moisture may be stack temperatures prohibit the use of Tef- eliminated by heating the sample line, and lon probe components, glass or quartz-lined by conditioning the sample with dry dilution probes may serve as substitutes. air to lower its dew point below the operat- ing temperature of the GC/FPD analytical NOTE.— Mention of trade names or specific system prior to analysis. products does not constitute an endorsement 3.2 Carbon Monoxide and Carbon Dioxide. by the Environmental Protection Agency. CO and CO2 have substantial desensitizing ef- 5.1.2 Particulate Filter. 50-mm Teflon fil- fects on the flame photometric detector even ter holder and a 1- to 2-micron porosity Tef- after 9:1 dilution. (Acceptable systems must lon filter (available through Savillex Cor- demonstrate that they have eliminated this poration, 5325 Highway 101, Minnetonka, interference by some procedure such as Minnesota 55343). The filter holder must be
831
VerDate 11
maintained in a hot box at a temperature of 5.1.3.2 Connections between the probe, at least 120°C (248°F). particulate filter, and S02 scrubber shall be
5.1.3 SO2 Scrubber. made of Teflon and as short in length as pos- 5.1.3.1 Three 300-ml Teflon segment sible. All portions of the probe, particulate impingers connected in series with flexible, filter, and connections prior to the S02 scrub- thick-walled, Teflon tubing. (Impinger parts ber (or alternative point of moisture re- and tubing available through Savillex.) The moval) shall be maintained at a temperature of at least 120 °Ψ (248 °F). first two impingers contain 100 ml of citrate 5.1.4 Sample Line. Teflon, no greater than buffer, and the third impinger is initially 1.3-cm (1⁄2-in.) ID. Alternative materials, dry. The tip of the tube inserted into the so- such as virgin Nylon, may be used provided lution should be constricted to less than 3- the line loss test is acceptable. 1 mm ( ⁄8-in.) ID and should be immersed to a 5.1.5 Sample Pump. The sample pump depth of at least 5 cm (2 in.). Immerse the shall be a leakless Teflon-coated diaphragm impingers in an ice water bath and maintain type or equivalent. near 0°C. The scrubber solution will nor- 5.2 Dilution System. The dilution system mally last for a 3-hour run before needing re- must be constructed such that all sample placement. This will depend upon the effects contacts are made of Teflon, glass, or stain- of moisture and particulate matter on the less-steel. It must be capable of approxi- solution strength and pH. mately a 9:1 dilution of the sample.
832
VerDate 11
5.3 Gas Chromatograph (Figure 15–2). The 5.3.1 Oven. Capable of maintaining the sep- gas chromatograph must have at least the aration column at the proper operating tem- following components: perature ±1°C.
833
VerDate 11
5.3.2 Temperature Gauge. To monitor col- 5.3.4.1 Electrometer. Capable of full scale umn oven, detector, and exhaust tempera- amplification of linear ranges of 10¥9 to ture ±1°C. 10¥4amperes full scale. 5.3.3 Flow System. Gas metering system to 5.3.4.2 Power Supply. Capable of delivering measure sample, fuel, combustion gas, and up to 750 volts. carrier gas flows. 5.3.4.3 Recorder. Compatible with the out- 5.3.4 Flame Photometric Detector. put voltage range of the electrometer.
5.3.4.4 Rotary Gas Valves. Multiport Tef- teria may be considered alternate methods lon-lined valves equipped with sample loop. subject to the approval of the Administrator. Sample loop volumes shall be chosen to pro- 5.5 Calibration System (Figure 15–3). The vide the needed analytical range. Teflon tub- calibration system must contain the follow- ing and fittings shall be used throughout to ing components. present an inert surface for sample gas. The 5.5.1 Flow System. To measure air flow gas chromatograph shall be calibrated with over permeation tubes within ±2 percent. the sample loop used for sample analysis. Each flowmeter shall be calibrated after a 5.4 Gas Chromatograph Columns. The col- complete test series with a wet-test meter. If umn system must be demonstrated to be ca- the flow measuring device differs from the pable of resolving three major reduced sulfur wet-test meter by more than 5 percent, the compounds: H2S, COS, and CS2. completed test shall be discarded. Alter- To demonstrate that adequate resolution natively, the tester may elect to use the flow has been achieved the tester must submit a data that will yield the lowest flow measure- chromatogram of a calibration gas contain- ment. Calibration with a wet-test meter be- ing all three reduced sulfur compounds in fore a test is optional. Flow over the perme- the concentration range of the applicable ation device may also be determined using a standard. Adequate resolution will be defined soap bubble flowmeter. as base line separation of adjacent peaks 5.5.2 Constant Temperature Bath. Device when the amplifier attenuation is set so that capable of maintaining the permeation tubes the smaller peak is at least 50 percent of full at the calibration temperature within 0.1 °C. scale. Base line separation is defined as a re- 5.5.3 Temperature Gauge. Thermometer turn to zero ±5 percent in the interval be- or equivalent to monitor bath temperature tween peaks. Systems not meeting this cri- within 0.1 °C.
834
VerDate 11
6. Reagents in 1 liter of water. Alternatively, 284 g of so- dium citrate may be substituted for the po- 6.1 Fuel. Hydrogen (H2) prepurified grade or better. tassium citrate. Adjust the pH to between 5.4 and 5.6 with potassium citrate or citric acid, 6.2 Combustion Gas. Oxygen (O2) or air, re- search purity or better. as required. 6.3 Carrier Gas. Prepurified grade or better. 6.7 Sample Line Loss Gas (Optional). As 6.4 Diluent. Air containing less than 0.5 an alternative, H2S cylinder gas may be used ppm total sulfur compounds and less than 10 for the sample line loss test. The gas shall be ppm each of moisture and total hydro- calibrated against permeation devices hav- carbons. ing known permeation rates or by the proce- 6.5 Calibration Gases. Permeation tubes, dure in Section 7 of Method 16A. one each of H2S, COS, and CS2, gravimetri- 7. Pretest Procedures cally calibrated and certified at some con- The following procedures are optional but venient operating temperature. These tubes would be helpful in preventing any problem consist of hermetically sealed FEP Teflon which might occur later and invalidate the tubing in which a liquified gaseous substance entire test. is enclosed. The enclosed gas permeates 7.1 After the complete measurement sys- through the tubing wall at a constant rate. tem has been set up at the site and deemed When the temperature is constant, calibra- to be operational, the following procedures tion gases covering a wide range of known should be completed before sampling is initi- concentrations can be generated by varying ated. and accurately measuring the flow rate of 7.1.1 Leak Test. Appropriate leak test pro- diluent gas passing over the tubes. These cedures should be employed to verify the in- calibration gases are used to calibrate the tegrity of all components, sample lines, and GC/FPD system and the dilution system. connections. The following leak test proce- 6.6 Citrate Buffer. Dissolve 300 g of potas- dure is suggested: For components upstream sium citrate and 41 g of anhydrous citric acid of the sample pump, attach the probe end of
835
VerDate 11
the sample line to a manometer or vacuum Vary the amount of air flowing over the gauge, start the pump and pull greater than tubes to produce the desired concentrations 50 mm (2 in.) Hg vacuum, close off the pump for calibrating the analytical and dilution outlet, and then stop the pump and ascertain systems. The air flow across the tubes must that there is no leak for 1 minute. For com- at all times exceed the flow requirement of ponents after the pump, apply a slight posi- the analytical systems. The concentration in tive pressure and check for leaks by applying parts per million generated by a tube con- a liquid (detergent in water, for example) at taining a specific permeant can be calculated each joint. Bubbling indicates the presence as follows: of a leak. As an alternative to the initial C=K×Pr/ML leak-test, the sample line loss test described Eq. 15–1 in Section 10.1 may be performed to verify Where: the integrity of components. C=Concentration of permeant produced in 7.1.2 System Performance. Since the com- ppm. plete system is calibrated following each P =Permeation rate of the tube in µg/min. test, the precise calibration of each compo- r M=Molecular weight of the permeant: g/g- nent is not critical. However, these compo- mole. nents should be verified to be operating prop- L=Flow rate, l/min, of air over permeant @ erly. This verification can be performed by 20°C, 760 mm Hg. observing the response of flowmeters or of K=Gas constant at 20°C and 760 mm Hg=24.04 the GC output to changes in flow rates or l/g mole. calibration gas concentrations and 8.3 Calibration of Analysis System. Gen- ascertaining the response to be within pre- erate a series of three or more known con- dicted limits. If any component or the com- centrations spanning the linear range of the plete system fails to respond in a normal and FPD (approximately 0.5 to 10 ppm for a 1—ml predictable manner, the source of the dis- sample) for each of the three major sulfur crepancy should be identified and corrected compounds. Bypassing the dilution system, before proceeding. inject these standards into the GC/FPD ana- 8. Calibration lyzers and monitor the responses. Three in- jects for each concentration must yield the Prior to any sampling run, calibrate the precision described in Section 4.1. Failure to system using the following procedures. (If attain this precision is an indication of a more than one run is performed during any problem in the calibration or analytical sys- 24-hour period, a calibration need not be per- tem. Any such problem must be identified formed prior to the second and any subse- and corrected before proceeding. quent runs. The calibration must, however, 8.4 Calibration Curves. Plot the GC/FPD re- be verified as prescribed in Section 10, after sponse in current (amperes) versus their the last run made within the 24-hour period.) causative concentrations in ppm on log-log 8.1 General Considerations. This section coordinate graph paper for each sulfur com- outlines steps to be followed for use of the pound. Alternatively, a least squares equa- GC/FPD and the dilution system. The proce- tion may be generated from the calibration dure does not include detailed instructions data. Alternatively, a least squares equation because the operation of these systems is may be generated from the calibration data complex, and it requires an understanding of using concentrations versus the appropriate the individual system being used. Each sys- instrument response units. tem should include a written operating man- 8.5 Calibration of Dilution System. Gen- ual describing in detail the operating proce- erate a known concentration of hydrogen dures associated with each component in the sulfied using the permeation tube system. measurement system. In addition, the opera- Adjust the flow rate of diluent air for the tor should be familiar with the operating first dilution stage so that the desired level principles of the components; particularly of dilution is approximated. Inject the di- the GC/FPD. The citations in the Bibliog- luted calibration gas into the GC/FPD sys- raphy at the end of this method are rec- tem and monitor its response. Three injec- ommended for review for this purpose. tions for each dilution must yield the preci- 8.2 Calibration Procedure. Insert the per- sion described in Section 4.1. Failure to at- meation tubes into the tube chamber. Check tain this precision in this step is an indica- the bath temperature to assure agreement tion of a problem in the dilution system. with the calibration temperature of the Any such problem must be identified and tubes within ±0.1°C. Allow 24 hours for the corrected before proceeding. Using the cali- tubes to equilibrate. Alternatively equilibra- bration data for H2S (developed under 8.3) de- tion may be verified by injecting samples of termine the diluted calibration gas con- calibration gas at 1-hour intervals. The per- centration in ppm. Then calculate the dilu- meation tubes can be assumed to have tion factor as the ratio of the calibration gas reached equilibrium when consecutive hour- concentration before dilution to the diluted ly samples agree within the precision limits calibration gas concentration determined of Section 4.1. under this section. Repeat this procedure for
836
VerDate 11
each stage of dilution required. Alter- guideline for determining if there are leaks natively, the GC/FPD system may be cali- in the sampling system. brated by generating a series of three or 10.2 Recalibration. After each run, or after more concentrations of each sulfur com- a series of runs made within a 24-hour period, pound and diluting these samples before in- perform a partial recalibration using the jecting them into the GC/FPD system. This procedures in Section 8. Only H2S (or other data will then serve as the calibration data permeant) need be used to recalibrate the for the unknown samples and a separate de- GC/FPD analysis system (8.3) and the dilu- termination of the dilution factor will not be tion system (8.5). necessary. However, the precision require- 10.3 Determination of Calibration Drift. ments of Section 4.1 are still applicable. Compare the calibration curves obtained 9. Sampling and Analysis Procedure prior to the runs, to the calibration curves 9.1 Sampling. Insert the sampling probe obtained under Section 10.2. The calibration into the test port making certain that no di- drift should not exceed the limits set forth in lution air enters the stack through the port. Section 4.2. If the drift exceeds this limit, Begin sampling and dilute the sample ap- the intervening run or runs should be consid- proximately 9:1 using the dilution system. ered not valid. The tester, however, may in- Note that the precise dilution factor is that stead have the option of choosing the cali- which is determined in section 8.5. Condition bration data set which would give the high- the entire system with sample for a mini- est sample values. mum of 15 minutes prior to commencing 11. Calculations analysis. 11.1 Determine the concentrations of each 9.2 Analysis. Aliquots of diluted sample are reduced sulfur compound detected directly injected into the GC/FPD analyzer for analy- from the calibration curves. Alternatively, sis. the concentrations may be calculated using 9.2.1 Sample Run. A sample run is com- the equation for the least squares line. posed of 16 individual analyses (injects) per- 11.2 Calculation of SO Equivalent. SO formed over a period of not less than 3 hours 2 2 equivalent will be determined for each anal- or more than 6 hours. ysis made by summing the concentrations of 9.2.2 Observation for Clogging of Probe or each reduced sulfur compound resolved dur- Filter. If reductions in sample concentra- ing the given analysis. tions are observed during a sample run that cannot be explained by process conditions, SO2 equivalent=Σ(H2S, COS, 2 CS2)d the sampling must be interrupted to deter- Eq. 15–2 mine if the probe or filter is clogged with particulate matter. If either is found to be Where: clogged, the test must be stopped and the re- SO2 equivalent=The sum of the concentra- sults up to that point discarded. Testing may tion of each of the measured compounds resume after cleaning or replacing the probe (COS, H2S, CS2) expressed as sulfur diox- and filter. After each run, the probe and fil- ide in ppm. ter shall be inspected and, if necessary, re- H2S=Hydrogen sulfide, ppm. placed. COS=Carbonyl sulfide, ppm.
10. Post-Test Procedures CS2=Carbon disulfide, ppm. 10.1 Sample Line Loss. A known concentra- d=Dilution factor, dimensionless. tion of hydrogen sulfide at the level of the 11.3 Average SO2 Equivalent. This is de- applicable standard, ±20 percent, must be in- termined using the following equation. Sys- troduced into the sampling system at the tems that do not remove moisture from the opening of the probe in sufficient quantities sample but conditions the gas to prevent to ensure that there is an excess of sample condensation must correct the average SO2 which must be vented to the atmosphere. equivalent for the fraction of water vapor The sample must be transported through the present. entire sampling system to the measurement system in the normal manner. The resulting measured concentration should be compared to the known value to determine the sam- pling system loss. A sampling system loss of more than 20 percent is unacceptable. Sam- pling losses of 0–20 percent must be corrected where: by dividing the resulting sample concentra- Average SO equivalent = Average SO equiv- tion by the fraction of recovery. The known 2 2 alent in ppm, dry basis. gas sample may be generated using perme- Average SO equivalent = SO in ppm as de- ation tubes. Alternatively, cylinders of hy- 2 i 2 termined by Equation 15–2. drogen sulfide mixed in nitrogen and verified according to Section 6.7 may be used. The N = Number of analyses performed. optional pretest procedures provide a good 12. Bibliography
837
VerDate 11
12.1 O’Keeffe, A. E. and G. C. Ortman. ‘‘Pri- air is added to the oxygen (O2)-deficient gas mary Standards for Trace Gas Analysis.’’ at a known rate. The TRS compounds (hy- Anal. Chem. 38,760 (1966). drogen sulfide, carbonyl sulfide, and carbon 12.2 Stevens, R. K., A. E. O’Keeffe, and disulfide) are thermally oxidized to sulfur di- G. C. Ortman. ‘‘Absolute Calibration of a oxide, collected in hydrogen peroxide as sul- Flame Photometric Detector to Volatile Sul- fate ion, and then analyzed according to the fur Compounds at Sub-Part-Per-Million Lev- Method 6 barium-thorin titration procedure. els.’’ Environmental Science and Technology 1.3 Interferences. Reduced sulfur com- 3:7 (July 1969). pounds, other than TRS, that are present in 12.3 Mulik, J. D., R. K. Stevens, and R. the emissions will also be oxidized to SO2. Baumgardner. ‘‘An Analytical System De- For example, thiophene has been identified signed to Measure Multiple Malodorous Com- in emissions from a Stretford unit and pro- pounds Related to Kraft Mill Activities.’’ duced a positive bias of 30 percent in the Presented at the 12th Conference on Methods Method 15A result. However, these biases in Air Pollution and Industrial Hygiene may not affect the outcome of the test at Studies, University of Southern California, units where emissions are low relative to the Los Angeles, CA, April 6–8, 1971. standard. 12.4 Devonald, R. H., R. S. Serenius, and A. Calcium and aluminum have been shown to D. McIntyre. ‘‘Evaluation of the Flame Pho- interfere in the Method 6 titration proce- tometric Detector for Analysis of Sulfur dure. Since these metals have been identified Compounds.’’ Pulp and Paper Magazine of in particulate matter emissions from Canada, 73,3 (March, 1972). Stretford units, a Teflon filter is required to 12.5 Grimley, K. W., W. S. Smith, and R. remove this interference. M. Martin. ‘‘The Use of a Dynamic Dilution NOTE: Mention of trade name or commer- System in the Conditioning of Stack Gases cial products in this publication does not for Automated Analysis by a Mobile Sam- constitute the endorsement or recommenda- pling Van.’’ Presented at the 63rd Annual tion for use by the Environmental Protec- APCA Meeting in St. Louis, MO. June 14–19, tion Agency. 1970. 12.6 General Reference. Standard Methods When used to sample emissions containing of Chemical Analysis Volume III A and B In- 7 percent moisture or less, the midget strumental Methods. Sixth Edition. Van impingers have sufficient volume to contain Nostrand Reinhold Co. the condensate collected during sampling. Dilution of the H2O2 does not affect the col- METHOD 15A—DETERMINATION OF TOTAL RE- lection of SO2. At higher moisture contents, DUCED SULFUR EMISSIONS FROM SULFUR RE- the potassium citrate-citric acid buffer sys- COVERY PLANTS IN PETROLEUM REFINERIES tem used with Method 16A should be used to collect the condensate. 1. Applicability, Principle, Interferences, Preci- 1.4 Precision and bias. Relative standard sion, and Bias deviations of 2.8 and 6.9 percent at 41 ppm 1.1 Applicability. This method is applica- TRS have been obtained when sampling for 1 ble to the determination of total reduced sul- and 3 hours, respectively. Results obtained fur (TRS) emissions from sulfur recovery with this method are likely to contain a plants where the emissions are in a reducing positive bias due to the presence of nonregu- atmosphere, such as in Stretford units. The lated sulfur compounds (that are present in lower detectable limit is 0.1 ppm of sulfur di- petroleum) in the emissions. oxide (SO2) when sampling at 2 liters/min for 3 hours or 0.3 ppm when sampling at 2 liters/ 2. Apparatus min for 1 hour. The upper concentration 2.1 Sampling. The sampling train is limit of the method exceeds TRS levels gen- shown in Figure 15A–1, and component parts erally encountered in sulfur recovery plants. are discussed below. Modifications to this 1.2 Principle. An integrated gas sample is sampling train are acceptable provided that extracted from the stack, and combustion the system performance check is met.
838
VerDate 11
2.1.1 Probe. 0.6-cm (1⁄4-in.) OD Teflon tub- sive tape. A flexible thermocouple or some ing sequentially wrapped with heat-resistant other suitable temperature-measuring device fiber strips, a rubberized heating tape (with shall be placed between the Teflon tubing a plug at one end), and heat-resistant adhe- and the fiber strips so that the temperature
839
VerDate 11
can be monitored. The probe should be sheath will prevent flue gas from entering sheathed in stainless steel to provide in- between the probe and sheath. The sampling stack rigidity. A series of bored-out stainless probe is depicted in Figure 15A–2. steel fittings placed at the front of the
Figure 15A±2. Method 15A sampling probe.
2.1.2 Particulate filter. A 50-mm Teflon maintained in a hot box at a high enough filter holder and a 1- to 2-µm porosity Teflon temperature to prevent condensation. filter (available through Savillex Corpora- 2.1.3 Combustion air delivery system. As tion, 5325 Highway 101, Minnetonka, Min- shown in the schematic diagram in Figure nesota 55345). The filter holder must be 15A–3. The rotameter should be selected to measure an air flow rate of 0.5 liter/min.
840
VerDate 11
2.1.4 Combustion tube. Quartz glass tub- 2.1.8 Volume meter. Dry gas meter capa- ing with an expanded combustion chamber ble of measuring the sample volume under 2.54 cm (1 in.) in diameter and at least 30.5 the particular sampling conditions with an cm (12 in.) long. The tube ends should have accuracy of ± 2 percent. an outside diameter of 0.6 cm (1⁄4 in.) and be 2.1.9 U-tube manometer. To measure the at least 15.3 cm (6 in.) long. This length is pressure at the exit of the combustion gas necessary to maintain the quartz-glass con- dry gas meter. nector at ambient temperature and thereby 2.2 Sample recovery and analysis. Same avoid leaks. Alternatively, the outlet may be as in Method 6, Sections 2.2 and 2.3, except a constructed with a 90-degree glass elbow and 10-ml buret with 0.05-ml graduations is re- socket that would fit directly onto the inlet quired for titrant volumes of less than 10.0 of the first peroxide impinger. ml, and the spectrophotometer is not needed. 2.1.5 Furnace. Of sufficient size to enclose 3. Reagents the combustion tube. The furnace shall have Unless otherwise indicated, all reagents a temperature regulator capable of main- must conform to the specifications estab- taining the temperature at 1100 ± 50°C. The lished by the Committee on Analytical Re- furnace operating temperature shall be agents of the American Chemical Society. checked with a thermocouple to ensure accu- When such specifications are not available, racy. Lindberg furnaces have been found to the best available grade shall be used. be satisfactory. 3.1 Sampling. The following reagents are 2.1.6 Peroxide impingers, stopcock grease, needed: thermometer, drying tube, valve, pump, ba- 3.1.1 Water. Same as in Method 6, Section rometer, and vacuum gauge. Same as in 3.1.1. Method 6, Sections 2.1.2, 2.1.4, 2.1.5, 2.1.6, 3.1.2 Hydrogen peroxide, 3 percent. Same 2.1.7, 2.1.8, 2.1.11, and 2.1.12, respectively. as in Method 6, Section 3.1.5 (40 ml is needed 2.1.7 Rate meters. Rotameters (or equiva- per sample). lent) capable of measuring flow rate to with- 3.1.3 Recovery check gas. Carbonyl sulfide in 5 percent of the selected flow rate and (COS) in nitrogen (100 ppm or greater, if nec- calibrated as in Section 5.2. essary) in an aluminum cylinder. Verify the
841
VerDate 11
concentration by gas chromatography where nometer reading at regular intervals during the instrument is calibrated with a COS per- the sampling period. Sample for 1 or 3 hours. meation tube. At the end of sampling, turn off the sample 3.1.4 Combustion gas. Air, contained in a pump and combustion air simultaneously gas cylinder equipped with a two-stage regu- (within 15 to 30 seconds of each other). All lator. The gas should contain less than 50 other procedures are the same as in Method ppb of reduced sulfur compounds and less 6, Section 4.1.3, except that the sampling than 10 ppm total hydrocarbons. train should not be purged. After collecting 3.2 Sample recovery and analysis. Same the sample, remove the probe from the stack as in Method 6, Sections 3.2 and 3.3. and conduct a leak-check (mandatory). 4. Procedure After each 3-hour test run (or after three 1- 4.1 Sampling. Before any source sampling hour samples), conduct one system perform- is done, conduct two 30-minute system per- ance check (see Section 4.3). After this sys- formance checks in the field, as detailed in tem performance check and before the next Section 4.3, to validate the sampling train test run, it is recommended that the probe components and procedures (optional). be rinsed and brushed and the filter replaced. 4.1.1 Preparation of sampling train. For In Method 15, a test run is composed of 16 the Method 6 part of the train, measure 20 ml individual analyses (injects) performed over of 3 percent hydrogen peroxide into the first a period of not less than 3 hours or more and second midget impingers. Leave the than 6 hours. For Method 15A to be consist- third midget impinger empty and add silica ent with Method 15, the following may be gel to the fourth impinger. Alternatively, a used to obtain a test run: (1) Collect three 60- silica gel drying tube may be used in place of minute samples or (2) collect one 3-hour the fourth impinger. Place crushed ice and sample. (Three test runs constitute a test.) water around all impingers. Maintain the ox- 4.2 Sample recovery. Recover the hydro- idation furnace at 1100 ± 50°C to ensure 100 gen peroxide-containing impingers as de- percent oxidation of COS. Maintain the tailed in Method 6, Section 4.2. probe and filter temperatures at a high 4.3 System performance check. A system enough level (no visible condensation) to performance check is done (1) to validate the prevent moisture condensation and monitor sampling train components and procedure the temperatures with a thermocouple. (before testing, optional) and (2) to validate 4.1.2 Leak-check procedure. Assemble the a test run (after a run). Perform a check in sampling train and leak-check as described the field before testing consisting of at least in Method 6, Section 4.1.2. Include the com- two samples (optional), and perform an addi- bustion air delivery system from the needle tional check after each 3-hour run or after valve forward in the leak-check. three 1-hour samples (mandatory). 4.1.3 Sample collection. Adjust the pres- The checks involve sampling a known con- sure on the second stage of the regulator on centration of COS and comparing the ana- the combustion air cylinder to 10 psig. Ad- lyzed concentration with the known con- just the combustion air flow rate to 0.50 centration. Mix the recovery gas with N2 as liter/min (±10 percent) before injecting com- shown in Figure 15A–4 if dilution is required. bustion air into the sampling train. Then in- Adjust the flow rates to generate a COS con- ject combustion air into the sampling train, centration in the range of the stack gas or start the sample pump, and open the stack within 20 percent of the applicable standard sample gas valve. Carry out these three oper- at a total flow rate of at least 2.5 liters/min. ations within 15 to 30 seconds to avoid pres- Use Equation 15A–4 to calculate the con- surizing the sampling train. Adjust the total centration of recovery gas generated. Cali- sample flow rate to 2.0 liters/min (±10 per- brate the flow rate from both sources with a cent). The combustion air flow rate of 0.50 soap bubble flow tube so that the diluted liter/min and the total sample flow rate of concentration of COS can be accurately cal- 2.0 liters/min produce an 02 concentration of culated. Collect 30-minute samples, and ana- 5.0 percent in the stack gas. This 02 con- lyze in the normal manner. Collect the sam- centration must be maintained constantly to ples through the probe of the sampling train allow oxidation of TRS to SO2. Adjust these using a manifold or some other suitable de- flow rates during sampling as necessary. vice that will ensure extraction of a rep- Monitor and record the combustion air ma- resentative sample.
842
VerDate 11
The recovery check must be performed in In the calculations, retain at least one the field before replacing the particulate fil- extra decimal figure beyond that of the ac- ter and before cleaning the probe. A sample quired data. Round off figures after final cal- recovery of 100 ± 20 percent must be obtained culations. for the data to be valid and should be re- 6.1 Nomenclature. ported with the emission data, but should not be used to correct the data. However, if CTRS=Concentration of TRS as SO2, dry basis, the performance check results do not affect corrected to standard conditions, ppm. the compliance or noncompliance status of N=Normality of barium perchlorate titrant, the affected facility, the Administrator may milliequivalents/ml. decide to accept the results of the compli- Pbar=Barometric pressure at exit orifice of ance test. Use Equation 15A–5 to calculate the dry gas meter, mm Hg. the recovery efficiency. P =Standard absolute pressure, 760 mm Hg. 4.4 Sample analysis. Same as in Method 6, std Section 4.3. For compliance tests only, an Tm=Average dry gas meter absolute tempera- °. EPA SO field audit sample shall be analyzed ture, 2 °. with each set of samples. Such audit samples Tstd=Standard absolute temperature, 293 are available from the Quality Assurance Di- Va=Volume of sample aliquot titrated, ml. vision, Environmental Monitoring Systems Vms=Dry gas volume as measured by the sam- Laboratory, U.S. Environmental Protection ple train dry gas meter, liters.
Agency, Research Triangle Park, NC 27711. Vmc=Dry gas volume as measured by the 5. Calibration. combustion air dry gas meter, liters. 5.1 Metering system, thermometers, ba- Vms(std)=Dry gas volume measured by the sam- rometer, and barium perchlorate solution. ple train dry gas meter, corrected to Calibration procedures are presented in standard conditions, liters. Method 6, Sections 5.1, 5.2, 5.4, and 5.5. 5.2 Rotameters. Calibrate with a bubble Vmc(std)=Dry gas volume measured by the flow tube. combustion air dry gas meter, corrected 6. Calculations. to standard conditions, liters.
843
VerDate 11
Vsoln=Total volume of solution in which the CRG=Concentration of generated recovery sulfur dioxide sample is contained, 100 gas, ppm.
ml. CCOS=Concentration of COS recovery gas, Vt=Volume of barium perchlorate titrant ppm. used for the sample (average of replicate QCOS=Flow rate of COS recovery gas, liters/ titrations), ml. min. V =Volume of barium perchlorate titrant tb Q =Flow rate of diluent N liters/min. used for the blank, ml. N2 2, Y=Calibration factor for sampling train dry R=Recovery efficiency for the system per- gas meter. formance check, percent. Yc=Calibration factor for combustion air dry 32.03=Equivalent weight of sulfur dioxide, gas meter. mg/meq.
3 3 µl ()32.. 03mg () 24 05liters ()mole()1 g( 10 ml )( 10 µ l ) 12025 = meq ()meq ()mole ()64. 06g() 103 g( liter )( ml ) 6.2 Dry Sample Gas Volume, Corrected to Standard Conditions. VYTP( )( ) KYVP()() = ms std bar = l m bar Vms() std Eq.15 A- 1 ()TPm() std Tm
where: K1=0.3858°Κ/mm Hg for metric units. 6.3 Combustion Air Gas Volume, Gorrected to Standard Conditions.
k Y( V )( P ) = l c mc bar Vmc() std Eq.15 A- 2 Tm
NOTE: Correct Pbar for the average pressure 6.4 Concentration of TRS as ppm SO2. of the manometer during the sampling pe- riod.
KVVNVV()− ()/ = 2 t tb soln a C TRS Eq.15 A- 3 VV− ms() std mc() sytd
where: K2=12025 µl/meq for metric units. Annual Book of ASTM Standards. Part 31: 6.5 Concentration of Generated Recovery Water, Atmospheric Analysis. Philadelphia, Gas. Pennsylvania. 1974. p. 40–42. 2. Blosser, R.O., H.S. Oglesby, and A.K. Jain ()()CQCOS COS A Study of Alternate SO2 Scrubber Designs C = Eq.15 A- 4 Used for TRS Monitoring. National Council RG + QQCOS N2 of the Paper Industry for Air and Stream Im- provement, Inc., New York, New York. Spe- 6.6 Recovery Efficiency. cial Report 77–05. July 1977. C 3. Curtis, F., and G.D. McAlister R =TRS × 100Eq. 15 A- 5 Development and Evaluation of an Oxida- C tion/Method 6 TRS Emission Sampling Pro- RG cedure. Emission Measurement Branch, 7. Bibliography Emission Standards and Engineering Divi- 1. American Society for Testing and Materials sion, U.S. Environmental Protection Agency,
844
VerDate 11
Research Triangle Park, North Carolina can cause losses or interferences. This is pre- 27711. February 1980. vented by maintaining the probe, filter box, 4. Gellman, I. and connections at a temperature of at least A Laboratory and Field Study of Reduced 120°C (248°F). Moisture is removed in the SO2 Sulfur Sampling and Monitoring Systems. scrubber and heating the sample beyond this National Council of the Paper Industry for point is not necessary provided the ambient Air and Stream Improvement, Inc., New temperature is above 0°C. Alternatively, York, New York. Atmospheric Quality Im- moisture may be eliminated by heating the provement Technical Bulletin No. 81. Octo- sample line, and by conditioning the sample ber 1975. with dry dilution air to lower its dew point 5. Margeson, J.H., J.E. Knoll, M.R. Midgett, below the operating temperature of the GC/ B.B. Ferguson, and P.J. Schworer FPD analytical system prior to analysis. A Manual Method for TRS Determination. 3.2 Carbon Monoxide and Carbon Dioxide. Journal of Air Pollution Control Associa- CO and CO2 have a substantial desensitizing tion. 35:1280–1286. December 1985. effect on the flame photometric detector even after dilution. Acceptable systems must METHOD 16—SEMICONTINUOUS DETERMINATION demonstrate that they have eliminated this OF SULFUR EMISSIONS FROM STATIONARY interference by some procedure such as SOURCES eluting these compounds before any of the Introduction compounds to be measured. Compliance with this requirement can be demonstrated by The method described below uses the prin- submitting chromatograms of calibration ciple of gas chromatographic separation and gases with and without CO in the diluent flame photometric detection (FPD). Since 2 gas. The CO level should be approximately there are many systems or sets of operating 2 10 percent for the case with CO present. The conditions that represent useable methods of 2 two chromatograms should show agreement determining sulfur emissions, all systems within the precision limits of Section 4.1. which employ this principle, but differ only in details of equipment and operation, may 3.3 Particulate Matter. Particulate mat- be used as alternative methods, provided ter in gas samples can cause interference by that the calibration precision and sample eventual clogging of the analytical system. line loss criteria are met. This interference is eliminated by using the Teflon filter after the probe. 1. Principle and Applicability 3.4 Sulfur Dioxide (SO2). Sulfur dioxide is 1.1 Principle. A gas sample is extracted not a specific interferent but may be present from the emission source and an aliquot is in such large amounts that it cannot be ef- analyzed for hydrogen sulfide (H2S), methyl fectively separated from the other com- mercaptan (MeSH), dimethly sulfide (DMS), pounds of interest. The SO2 scrubber de- and dimethyl disulfide (DMDS) by gas scribed in Section 5.1.3 will effectively re- chromatographic (GC) separation and flame move SO2 from the sample. photometric detection (FPD). These four compounds are know collectively as total re- 4. Precision and Accuracy duced sulfur (TRS). 4.1 GC/FPD Calibration Precision. A se- 1.2 Applicability. This method is applica- ries of three consecutive injections of the ble for determination of TRS compounds same calibration gas, at any dilution, shall from recovery furnaces, lime kilns, and produce results which do not vary by more smelt dissolving tanks at kraft pulp mills. than ±5 percent from the mean of the three 2. Range and Sensitivity injections. 2.1 Range. The analytical range will vary 4.2 Calibration Drift. The calibration drift with the sample loop size. Typically, the an- determined from the mean of three injec- alytical range may extend from 0.1 to 100 tions made at the beginning and end of any ppm using 10 to 0.1-ml sample loop sizes. run or series of runs within a 24-hour period ± This eliminates the need for sample dilution shall not exceed 5 percent. in most cases. 4.3 System Calibration Accuracy. Losses 2.2 Sensitivity. Using the 10-ml sample through the sample transport system must size, the minimum detectable concentration be measured and a correction factor devel- is approximately 50 ppb. oped to adjust the calibration accuracy to 100 percent. 3. Interferences 3.1 Moisture Condensation. Moisture con- 5. Apparatus densation in the sample delivery system, the 5.1. Sampling. analytical column, or the FPD burner block 5.1.1 Probe.
845
VerDate 11
5.1.1.1 Teflon or Teflon-lined stainless or near the sample point entry. Any portion steel. The probe must be heated to prevent of the probe that contacts the stack gas moisture condensation. It shall be designed must be heated to prevent moisture con- to allow calibration gas to enter the probe at densation.
846
VerDate 11
5.1.1.2 Figure 16–1 illustrates the probe 5.3.2 Temperature Gauge. To monitor col- used in lime kilns and other sources where umn oven, detector, and exhaust tempera- significant amounts of particulate matter ture ±1°C. are present. The probe is designed with the 5.3.3 Flow System. Gas metering system deflector shield placed between the sample to measure sample, fuel, combustion gas, and and the gas inlet holes to reduce clogging of carrier gas flows. the filter and possible adsorption of sample 5.3.4 Flame Photometric Detector. gas. As an alternative, the probe described in 5.3.4.1 Electrometer. Capable of full scale Section 2.1.1 of Methods 16A having a nozzle amplification of linear ranges of 10¥9 to 10¥4 directed away from the gas stream may be amperes full scale. used at sources having significant amounts 5.3.4.2 Power Supply. Capable of deliver- of particulate matter. ing up to 750 volts. 5.1.1.3 NOTE: Mention of trade names or 5.3.4.3 Recorder. Compatible with the out- specific products does not constitute an en- put voltage range of the electrometer. dorsement by the Environmental Protection 5.3.4.4 Rotary Gas Valves. Multiport Tef- Agency. lon-lined valves equipped with sample loop. 5.1.2 Particulate Filter. 50-mm Teflon fil- Sample loop volumes shall be chosen to pro- ter holder and a 1- to 2-micron porosity Tef- vide the needed analytical range. Teflon tub- lon filter (available through Savillex Cor- ing and fittings shall be used throughout to poration, 5325 Highway 101, Minnetonka, present an inert surface for sample gas. The Minnesota 55343). The filter holder must be gas chromatograph shall be calibrated with maintained in a hot box at a temperature of the sample loop used for sample analysis. 5.4 Gas Chromatogram Columns. The col- at least 120 °C (248 °F). umn system must be demonstrated to be ca- 5.1.3 SO Scrubber. 2 pable of resolving the four major reduced 5.1.3.1 Three 300-ml Teflon segmented sulfur compounds: H2S, MeSH, DMS, and impingers connected in series with flexible, DMDS. It must also demonstrate freedom thick-walled, Teflon tubing. (Impinger parts from known interferences. and tubing available through Savillex.) The To demonstrate that adequate resolution first two impingers contain 100 ml of citrate has been achieved, the tester must submit a buffer and the third impinger is initially dry. chromatogram of a calibration gas contain- The tip of the tube inserted into the solution ing all four of the TRS compounds in the 1 should be constricted to less than 3-mm ( ⁄8- concentration range of the applicable stand- in.) ID and should be immersed to a depth of ard. Adequate resolution will be defined as at least 5 cm (2 in.). Immerse the impingers base line separation of adjacent peaks when ° in an ice water bath and maintain near 0 C. the amplifier attenuation is set so that the The scrubber solution will normally last for smaller peak is at least 50 percent of full a 3-hour run before needing replacement. scale. Baseline separation is defined as a re- This will depend upon the effects of moisture turn to zero ±5 percent in the interval be- and particulate matter on the solution tween peaks. Systems not meeting this cri- strength and pH. teria may be considered alternate methods 5.1.3.2 Connections between the probe, subject to the approval of the Administrator. particulate filter, and SO2 scrubber shall be 5.5 Calibration System. The calibration made of Teflon and as short in length as pos- system must contain the following compo- sible. All portions of the probe, particulate nents. (Figure 16–2) filter, and connections prior to the SO2 5.5.1 Tube Chamber. Chamber of glass or scrubber (or alternative point of moisture re- Teflon of sufficient dimensions to house per- moval) shall be maintained at a temperature meation tubes. of at least 120 °C (248 °F). 5.5.2 Flow System. To measure air flow 5.1.4 Sample Line. Teflon, no greater than over permeation tubes at ±2 percent. Each 1.3-cm (1⁄2-in.) ID. Alternative materials, flowmeter shall be calibrated after a com- such as virgin Nylon, may be used provided plete test series with a wet test meter. If the the line loss test is acceptable. flow measuring device differs from the wet 5.1.5 Sample Pump. The sample pump test meter by 5 percent, the completed test shall be leakless Teflon-coated diaphragm shall be discarded. Alternatively, the tester type or equivalent. may elect to use the flow data that would 5.2 Dilution System. Needed only for high yield the lower flow measurement. Calibra- sample concentrations. The dilution system tion with a wet test meter before a test is must be constructed such that all sample optional. Flow over the permeation device contacts are made of Teflon, glass, or stain- may also be determined using a soap bubble less steel. flowmeter. 5.3 Gas Chromatograph. The gas chro- 5.5.3 Constant Temperature Bath. Device matograph must have at least the following capable of maintaining the permeation tubes components: at the calibration temperature within ±0.1°C. 5.3.1 Oven. Capable of maintaining the 5.5.4 Temperature Gauge. Thermometer separation column at the proper operating or equivalent to monitor bath temperature temperature ±1°C. within ±1°C.
847
VerDate 11
6. Reagents indicates the presence of a leak. As an alter- native to the initial leak-test, the sample 6.1 Fuel. Hydrogen (H2), prepurified grade or better. line loss test described in Section 10.1 may 6.2 Combustion Gas. Oxygen (O2) or air, be performed to verify the integrity of com- research purity or better. ponents. 6.3 Carrier Gas. Prepurified grade or bet- 7.1.2 System Performance. Since the com- ter. plete system is calibrated following each 6.4 Diluent (If required). Air containing test, the precise calibration of each compo- less than 50 ppb total sulfur compounds and nent is not critical. However, these compo- less than 10 ppm each of moisture and total nents should be verified to be operating prop- hydrocarbons. erly. This verification can be performed by 6.5 Calibration Gases. Permeation tubes, observing the response of flowmeters or of one each of H2S, MeSH, DMS, and DMDS, the GC output to changes in flow rates or gravimetrically calibrated and certified at calibration gas concentrations and some convenient operating temperature. ascertaining the response to be within pre- These tubes consist of hermetically sealed dicted limits. In any component, or if the FEP Teflon tubing in which a liquified gase- complete system fails to respond in a normal ous substance is enclosed. The enclosed gas and predictable manner, the source of the permeates through the tubing wall at a con- discrepancy should be identified and cor- stant rate. When the temperature is con- rected before proceeding. stant, calibration gases covering a wide 8. Calibration range of known concentrations can be gen- Prior to any sampling run, calibrate the erated by varying and accurately measuring system using the following procedures. (If the flow rate of diluent gas passing over the more than one run is performed during any tubes. These calibration gases are used to 24-hour period, a calibration need not be per- calibrate the GC/FPD system and the dilu- formed prior to the second and any subse- tion system. quent runs. The calibration must, however, 6.6 Citrate Buffer. Dissolve 300 grams of be verified as prescribed in Section 10, after potassium citrate and 41 grams of anhydrous the last run made within the 24-hour period.) citric acid in 1 liter of deionized water. 284 grams of sodium citrate may be substituted 8.1 General Considerations. This section for the potassium citrate. Adjust the pH to outlines steps to be followed for use of the between 5.4 and 5.6 with potassium citrate or GC/FPD and the dilution system (if applica- citric acid, as required. ble). The procedure does not include detailed 6.7 Sample Line Loss Gas (Optional). As instructions because the operation of these systems is complex, and it requires an under- an alternative to permeation gas, H2S cyl- inder gas may be used for the sample line standing of the individual system being used. loss test. The gas shall be calibrated against Each system should include a written oper- permeation devices having known perme- ating manual describing in detail the operat- ation rates or by the procedure in Section 7 ing procedures associated with each compo- of Method 16A. nent in the measurement system. In addi- tion, the operator should be familiar with 7. Pretest Procedures the operating principles of the components, The following procedures are optional but particularly the GC/FPD. The citations in would be helpful in preventing any problem the Bibliography at the end of this method which might occur later and invalidate the are recommended for review for this purpose. entire test. 8.2 Calibration Procedure. Insert the per- 7.1 After the complete measurement sys- meation tubes into the tube chamber. Check tem has been set up at the site and deemed the bath temperature to assure agreement to be operational, the following procedures with the calibration temperature of the should be completed before sampling is initi- tubes within ±0.1°C. Allow 24 hours for the ated. tubes to equilibrate. Alternatively equilibra- 7.1.1 Leak Test. Appropriate leak test tion may be verified by injecting samples of procedures should be employed to verify the calibration gas at 1-hour intervals. The per- integrity of all components, sample lines, meation tubes can be assumed to have and connections. The following leak test pro- reached equilibrium when consecutive hour- cedure is suggested: For components up- ly samples agree within the precision limits stream of the sample pump, attach the probe of Section 4.1. end of the sample line to a manometer or Vary the amount of air flowing over the vacuum gauge, start the pump and pull tubes to produce the desired concentrations greater than 50 mm (2 in.) Hg vacuum, close for calibrating the analytical and dilution off the pump outlet, and then stop the pump systems. The air flow across the tubes must and ascertain that there is no leak for 1 at all times exceed the flow requirement of minute. For components after the pump, the analytical systems. The concentration in apply a slight positive pressure and check for parts per million generated by a tube con- leaks by applying a liquid (detergent in taining a specific permeant can be calculated water, for example) at each joint. Bubbling as follows:
848
VerDate 11
9.2.1 Sample Run. A sample run is com- Pr posed of 16 individual analyses (injects) per- CK= Eq.16- 1 formed over a period of not less than 3 hours ML or more than 6 hours. Where: 9.2.2 Observation for Clogging of Probe or C=Concentration of permeant produced in Filter. If reductions in sample concentra- ppm. tions are observed during a sample run that P =Permeation rate of the tube in µg/min. cannot be explained by process conditions, r the sampling must be interrupted to deter- M=Molecular weight of the permeant (g/g- mine if the probe or filter is clogged with mole). particulate matter. If either is found to be L=Flow rate, 1/min, of air over permeant @ clogged, the test must be stopped and the re- 20°C, 760 mm Hg. sults up to that point discarded. Testing may ° K=Gas constant at 20 C and 760 mm Hg=24.04 resume after cleaning or replacing the probe 1/g mole. and filter. After each run, the probe and fil- 8.3 Calibration of Analysis System. Gen- ter shall be inspected and, if necessary, re- erate a series of three or more known con- placed. centrations spanning the linear range of the 10. Post-Test Procedures FPD (approximately 0.5 to 10 ppm for a 1-ml sample) for each of the four major sulfur 10.1 Sample line loss. A known concentra- compounds. Inject these standards into the tion of hydrogen sulfide at the level of the GC/FPD analyzer and monitor the responses. applicable standard, ±20 percent, must be in- Three injects for each concentration must troduced into the sampling system at the not vary by more than 5 percent from the opening of the probe in sufficient quantities mean of the three injections. Failure to at- to ensure that there is an excess of sample tain this precision is an indication of a prob- which must be vented to the atmosphere. lem in the calibration or analytical system. The sample must be transported through the Any such problem must be identified and entire sampling system to the measurement corrected before proceeding. system in the normal manner. (See figure 16– 8.4 Calibration Curves. Plot the GC/FPD 1). The resulting measured concentration response in current (amperes) versus their should be compared to the known value to causative concentrations in ppm on log-log determine the sampling system loss. coordinate graph paper for each sulfur com- For sampling losses greater than 20 per- pound. Alternatively, a least squares equa- cent in a sample run, the sample run is not tion may be generated from the calibration to be used when determining the arithmetic data. Alternatively, a least squares equation mean of the performance test. For sampling may be generated from the calibration data losses of 0–20 percent, the sample concentra- using concentrations versus the appropriate tion must be corrected by dividing the sam- instrument response units. ple concentration by the fraction of recov- ery. The fraction of recovery is equal to one 9. Sampling and Analysis Procedure minus the ratio of the measured concentra- 9.1 Sampling. Insert the sampling probe tion to the known concentration of hydrogen into the test port making certain that no di- sulfide in the sample line loss procedure. The lution air enters the stack through the port. known gas sample may be generated using Begin sampling. Condition the entire system permeation tubes. Alternatively, cylinders of with sample for a minimum of 15 minutes hydrogen sulfide mixed in nitrogen and cer- prior to commencing analysis. tified according to section 6.7 may be used. 9.2 Analysis. Aliquots of sample are in- The optional pretest procedures provide a jected into the GC/FPD analyzer for analy- good guideline for determining if there are sis. leaks in the sampling system.
849
VerDate 11
10.2 Recalibration. After each run, or TRS=Total reduced sulfur in ppm, dry basis. after a series of runs made within a 24-hour H2S=Hydrogen sulfide, ppm. period, perform a partial recalibration using MeSH=Methyl mercaptan, ppm. the procedures in Section 8. Only H2S (or DMS=Dimethyl sulfide, ppm. other calibration gas) need be used to recali- DMDS=Dimethyl disulfide, ppm. brate the GC/FPD analysis system (Section d=Dilution factor, dimensionless. 8.3). 11.3 Average TRS. The average TRS will 10.3 Determination of Calibration Drift. be determined as follows: Compare the calibration curves obtained prior to the runs, to the calibration curves N obtained under Section 10.2. The calibration ∑ TRS drift should not exceed the limits set forth in i Section 4.2. If the drift exceeds this limit, Average TRS = i=l the intervening run or runs should be consid- ()− ered not valid. The tester, however, may in- NB1 WO stead have the option of choosing the cali- Where: bration data set which would give the high- Average TRS=Average total reduced sulfur est sample values. in ppm, dry basis. 11. Calculations TRSi=Total reduced sulfur in ppm as deter- 11.1 Determine the concentrations of each mined by Equation 16–2. reduced sulfur compound detected directly N=Number of samples. from the calibration curves. Alternatively, Bwo=Fraction of volume of water vapor in the concentrations may be calculated using the gas stream as determined by ref- the equation for the least squares line. erence Method 4—Determination of Mois- 11.2 Calculation of TRS. Total reduced sul- ture in Stack Gases. fur will be determined for each analysis 11.4 Average Concentration of Individual made by summing the concentrations of each Reduced Sulfur Compounds. reduced sulfur compound resolved during a N given analysis. ∑ S TRS=Σ (H2S, MeSH, DMS, 2DMDS)d i Eq. 16–2 C = i=l Eq.16- 4 Where: N
850
VerDate 11
Where: The flue gas must contain at least 1 per-
Si=Concentration of any reduced sulfur com- cent oxygen for complete oxidation of all pound from the ith sample injection, TRS to sulfur dioxide (SO2). The lower de- ppm. tectable limit is 0.1 ppm SO2 when sampling C=Average concentration of any one of the at 2 liters/min for 3 hours or 0.3 ppm when reduced sulfur compounds for the entire sampling at 2 liters/min for 1 hour. The run, ppm. upper concentration limit of the method ex- N=Number of injections in any run period. ceeds TRS levels generally encountered at 12. Bibliography kraft pulp mills. 1.2 Principle. An integrated gas sample is 12.1 O’Keeffe, A. E. and G. C. Ortman. extracted from the stack. SO is removed se- ‘‘Primary Standards for Trace Gas Analy- 2 lectively from the sample using a citrate sis.’’ Analytical Chemical Journal, 38,760 buffer solution. TRS compounds are then (1966). thermally oxidized to SO , collected in hy- 12.2 Stevens, R. K., A. E. O’Keeffe, and G. 2 drogen peroxide as sulfate, and analyzed by C. Ortman. ‘‘Absolute Calibration of a Flame the Method 6 barium-thorin titration proce- Photometric Detector to Volatile Sulfur dure. Compounds at Sub-Part-Per-Million Levels.’’ 1.3 Interferences. TRS compounds other Environmental Science and Technology, 3:7 than those regulated by the emission stand- (July, 1969). 12.3 Mulik, J. D., R. K. Stevens, and R. ards, if present, may be measured by this Baumgardner. ‘‘An Analytical System De- method. Therefore, carbonyl sulfide, which is signed to Measure Multiple Malodorous Com- partially oxidized to SO2 and may be present pounds Related to Kraft Mill Activities.’’ in a lime kiln exit stack, would be a positive Presented at the 12th Conference on Methods interferent. in Air Pollution and Industrial Hygiene Particulate matter from the lime kiln Studies, University of Southern California, stack gas (primarily calcium carbonate) can Los Angeles, CA. April 6–8, 1971. cause a negative bias if it is allowed to enter 12.4 Devonald, R. H., R. S. Serenius, and the citrate scrubber; the particulate matter A. D. McIntyre. ‘‘Evaluation of the Flame will cause the pH to rise and H2S to be ab- Photometric Detector for Analysis of Sulfur sorbed prior to oxidation. Furthermore, if Compounds.’’ Pulp and Paper Magazine of the calcium carbonate enters the hydrogen Canada, 73,3 (March, 1972). peroxide impingers, the calcium will precipi- 12.5 Grimley, K. W., W. S. Smith, and R. tate sulfate ion. Proper use of the particu- M. Martin. ‘‘The Use of a Dynamic Dilution late filter described in Section 2.1.3 will System in the Conditioning of Stack Gases eliminate this interference. for Automated Analysis by a Mobile Sam- 1.4 Precision and Bias. Relative standard pling Van.’’ Presented at the 63rd Annual deviations of 2.0 and 2.6 percent were ob- APCA Meeting in St. Louis, MO. June 14–19, tained when sampling a recovery boiler for 1 1970. and 3 hours, respectively. 12.6 General Reference. Standard Methods In a separate study at a recovery boiler, of Chemical Analysis Volume III A and B In- Method 16A was found to be unbiased rel- strumental Methods. Sixth Edition. Van ative to Method 16. Comparison of Method Nostrand Reinhold Co. 16A with Method 16 at a lime kiln indicated that there was no bias in Method 16A. How- METHOD 16A—DETERMINATION OF TOTAL RE- ever, instability of the source emissions ad- DUCED SULFUR EMISSIONS FROM STATION- versely affected the comparison. The preci- ARY SOURCES (IMPINGER TECHNIQUE) sion of Method 16A at the lime kiln was simi- 1. Applicability, Principle, Interferences, Preci- lar to that obtained at the recovery boiler. sion, and Bias Relative standard deviations of 2.7 and 7.7 percent have been obtained for system per- 1.1 Applicability. This method is applica- formance checks. ble to the determination of total reduced sul- fur (TRS) emissions from recovery boilers, 2. Apparatus lime kilns, and smelt dissolving tanks at 2.1 Sampling. The sampling train is kraft pulp mills, and from other sources shown in Figure 16A–1 and component parts when specified in an applicable subpart of are discussed below. Modifications to this the regulations. The TRS compounds include sampling train are acceptable provided the hydrogen sulfide, methyl mercaptan, di- system performance check (Section 4.3) is methyl sulfide, and dimethyl disulfide. met.
851
VerDate 11
2.1.1 Probe. Teflon (mention of trade ries of bored-out stainless steel fittings names or specific products does not con- placed at the front of the sheath will prevent stitute endorsement by the U.S. Environ- moisture and particulate from entering be- mental Protection Agency) tubing, 0.6-cm tween the probe and sheath. A 0.6-cm (1⁄4-in.) (1⁄4-in.) diameter, sequentially wrapped with Teflon elbow (bored out) should be attached heat-resistant fiber strips, a rubberized heat to the inlet of the probe, and a 2.54-cm (1-in.) tape (plug at one end), and heat-resistant ad- piece of Teflon tubing should be attached at hesive tape. A flexible thermocouple or other the open end of the elbow to permit the suitable temperature measuring device opening of the probe to be turned away from should be placed between the Teflon tubing the particulate stream; this will reduce the and the fiber strips so that the temperature amount of particulate drawn into the sam- can be monitored to prevent softening of the probe. The probe should be sheathed in stain- pling train. The sampling probe is depicted less steel to provide in-stack rigidity. A se- in Figure 16A–2.
852
VerDate 11
2.1.2 Probe Brush. Nylon bristle brush lished by the Committee on Analytical Re- with handle inserted into a 3.2-mm (1⁄8-in.) agents of the American Chemical Society. Teflon tubing. The Teflon tubing should be When such specifications are not available, long enough to pass the brush through the the best available grade shall be used. length of the probe. 3.1 Sampling. The following reagents are 2.1.3 Particulate Filter. 50-mm Teflon fil- needed: ter holder and a 1- to 2-µ porosity, Teflon fil- 3.1.1 Water. Same as in Method 6, Section ter (available through Savillex Corporation, 3.1.1. 5325 Highway 101, Minnetonka, Minnesota 3.1.2 Citrate Buffer. 300 g of potassium cit- 55343). The filter holder must be maintained rate (or 284 g of sodium citrate) and 41 g of in a hot box at a temperature sufficient to anhydrous citric acid dissolved in 1 liter of prevent moisture condensation. A tempera- water (200 ml is needed per test). Adjust the ture of 121 °C (250 °F) was found to be suffi- pH to between 5.4 and 5.6 with potassium cit- cient when testing a lime kiln under sub- rate or citric acid, as required. freezing ambient conditions. 3.1.3 Hydrogen Peroxide, 3 percent. Same 2.1.4 SO2 Scrubber. Three 300-ml Teflon as in Method 6, Section 3.1.3 (40 ml is needed segmented impingers connected in series per sample). with flexible, thick-walled, Teflon tubing. 3.1.4 Recovery Check Gas. Hydrogen sul- (Impinger parts and tubing available through fide (100 ppm or less) in nitrogen, stored in Savillex.) The first two impingers contain aluminum cylinders. Verify the concentra- 100 ml of citrate buffer and the third im- tion by Method 11 or by gas chromatography pinger is initially dry. The tip of the tube in- where the instrument is calibrated with an serted into the solution should be con- H2S permeation tube as described below. For stricted to less than 3 mm (1⁄8 in.) ID and Method 11, the standard deviation should not should be immersed to a depth of at least 5 exceed 5 percent on at least three 20-minute cm (2 in.). runs. 2.1.5 Combustion Tube. Quartz glass tub- Alternatively, hydrogen sulfide recovery ing with an expanded combustion chamber gas generated from a permeation device 2.54 cm (1 in.) in diameter and at least 30.5 gravimetrically calibrated and certified at cm (12 in.) long. The tube ends should have some convenient operating temperature may an outside diameter of 0.6 cm (1⁄4 in.) and be be used. The permeation rate of the device at least 15.3 cm (6 in.) long. This length is must be such that at a dilution gas flow rate necessary to maintain the quartz-glass con- of 3 liters/min, an H2S concentration in the nector at ambient temperature and thereby range of the stack gas or within 20 percent of avoid leaks. Alternatively, the outlet may be the standard can be generated. constructed with a 90-degree glass elbow and 3.1.5 Combustion Gas. Gas containing less socket that would fit directly onto the inlet than 50 ppb reduced sulfur compounds and of the first peroxide impinger. less than 10 ppm total hydrocarbons. The gas 2.1.6 Furnace. A furnace of sufficient size may be generated from a clean-air system to enclose the combustion chamber of the that purifies ambient air and consists of the combustion tube with a temperature regu- following components: Diaphragm pump, lator capable of maintaining the tempera- silica gel drying tube, activated charcoal ture at 800±100 °C. The furnace operating tube, and flow rate measuring device. Flow temperature should be checked with a ther- from a compressed air cylinder is also ac- mocouple to ensure accuracy. ceptable. 2.1.7 Peroxide Impingers, Stopcock 3.2 Sample Recovery and Analysis. Same Grease, Thermometer, Drying Tube, Valve, as in Method 6, Sections 3.2.1 and 3.3. Pump, Barometer, and Vacuum Gauge. Same 4. Procedure as in Method 6, Sections 2.1.2, 2.1.4, 2.1.6, 4.1 Sampling. Before any source sampling 2.1.7, 2.1.8, 2.1.11, and 2.1.12, respectively. is done, conduct two 30-minute system per- 2.1.8 Rate Meter. Rotameter, or equiva- formance checks in the field as detailed in lent, accurate to within 5 percent at the se- Section 4.3 to validate the sampling train lected flow rate of 2 liters/min. components and procedure (optional). 2.1.9 Volume Meter. Dry gas meter capa- 4.1.1 Preparation of Collection Train. For ble of measuring the sample volume under the SO scrubber, measure 100 ml of citrate the sampling conditions of 2 liters/min with 2 buffer into the first and second impingers; an accuracy of ±2 percent. leave the third impinger empty. Immerse the 2.1.10 Polyethylene Bottles. 250-ml bottles impingers in an ice bath, and locate them as for hydrogen peroxide solution recovery. close as possible to the filter heat box. The 2.2 Sample Preparation and Analysis. connecting tubing should be free of loops. Same as in Method 6, Section 2.3, except a 10- Maintain the probe and filter temperatures ml buret with 0.05-ml graduations is required sufficiently high to prevent moisture con- and the spectrophotometer is not needed. densation, and monitor with a suitable tem- 3. Reagents perature indicator. Unless otherwise indicated, all reagents For the Method 6 part of the train, meas- must conform to the specifications estab- ure 20 ml of 3 percent hydrogen peroxide into
853
VerDate 11
the first and second midget impingers. Leave minute samples or (2) collect one 3-hour the third midget impinger empty, and place sample. (Three test runs constitute a test.) silica gel in the fourth midget impinger. Al- 4.2 Sample Recovery. Disconnect the ternatively, a silica gel drying tube may be impingers. Quantitatively transfer the con- used in place of the fourth impinger. Main- tents of the midget impingers of the Method tain the oxidation furnace at 800±100 °C. 6 part of the train into a leak-free poly- Place crushed ice and water around all ethylene bottle for shipment. Rinse the three impingers. midget impingers and the connecting tubes 4.1.2 Citrate Scrubber Conditioning Pro- with water and add the washings to the same cedure. Condition the citrate buffer scrub- storage container. Mark the fluid level. Seal bing solution by pulling stack gas through and identify the sample container. the Teflon impingers and bypassing all other 4.3 System Performance Check. A system sampling train components. A purge rate of performance check is done (1) to validate the 2 liters/min for 10 minutes has been found to sampling train components and procedure be sufficient to obtain equilibrium. After the (prior to testing; optional) and (2) to validate citrate scrubber has been conditioned, as- a test run (after a run). Perform a check in semble the sampling train, and conduct (op- the field prior to testing consisting of a least tional) a leak-check as described in Method two samples (optional), and perform an addi- 6, Section 4.1.2. tional check after each 3-hour run or after 4.1.3 Sample Collection. Same as in Meth- three 1-hour samples (mandatory). od 6, Section 4.1.3, except the sampling rate The checks involve sampling a known con- is 2 liters/min (± 10 percent) for 1 or 3 hours. centration of H2S and comparing the ana- After the sample is collected, remove the lyzed concentration with the known con- probe from the stack, and conduct (manda- centration. Mix the H2S recovery gas (Sec- tory) a post-test leak check as described in tion 3.1.4) and combustion gas in a dilution Method 6, Section 4.1.2. The 15-minute purge system such as is shown in Figure 16A–3. Ad- of the train following collection should not just the flow rates to generate an H2S con- be performed. After each 3-hour test run (or centration in the range of the stack gas or after three 1-hour samples), conduct one sys- within 20 percent of the applicable standard tem performance check (see Section 4.3) to and an oxygen concentration greater than 1 determine the reduced sulfur recovery effi- percent at a total flow rate of at least 2.5 li- ciency through the sampling train. After ters/min. Use Equation 16A–3 to calculate this system performance check and before the concentration of recovery gas generated. the next test run, rinse and brush the probe Calibrate the flow rate from both sources with water, replace the filter, and change the with a soap bubble flow tube so that the di- citrate scrubber (recommended but op- luted concentration of H2S can be accurately tional). calculated. Collect 30-minute samples, and In Method 16, a test run is composed of 16 analyze in the normal manner (as discussed individual analyses (injects) performed over in Section 4.1.3). Collect the sample through a period of not less than 3 hours or more the probe of the sampling train using a than 6 hours. For Method 16A to be consist- manifold or some other suitable device that ent with Method 16, the following may be will ensure extraction of a representative used to obtain a test run: (1) collect three 60- sample.
854
VerDate 11
The recovery check must be performed in cent isopropanol, and four drops of thorin. the field prior to replacing the SO2 scrubber Analyze an EPA SO2 field audit sample with and particulate filter and before the probe is each set of samples. Such audit samples are cleaned. A sample recovery of 100 ±20 percent available from the Source Branch, Quality must be obtained for the data to be valid and Assurance Division, Environmental Monitor- should be reported with the emission data, ing Systems Laboratory, U. S. Environ- but should not be used to correct the data. mental Protection Agency, Research Tri- However, if the performance check results do angle Park, North Carolina 27711. not affect the compliance or noncompliance 5. Calibration status of the affected facility, the Adminis- 5.1 Metering System, Thermometers, trator may decide to accept the results of the compliance test. Use Equation 16A–4 to Rotameters, Barometers, and Barium Per- calculate the recovery efficiency. chlorate Solution. Calibration procedures are presented in Method 6, Sections 5.1 4.4 Sample Analysis. Same as in Method through 5.5. 6, Section 4.3, except for 1-hour sampling, take a 40-ml aliquot, add 160 ml of 100 per- 6. Calculations
855
VerDate 11
In the calculations, at least one extra deci- Tm=Average dry gas meter absolute tempera- mal figure should be retained beyond that of ture, °K (°R). the acquired data. Figures should be rounded Tstd=Standard absolute temperature, 293 °K, off after final calculations. (528 °R). 6.1 Nomenclature. Va=Volume of sample aliquot titrated, ml.
CTRS=Concentration of TRS as SO2, dry basis Vm=Dry gas volume as measured by the dry corrected to standard conditions, ppm. gas meter, liters (dcf). CRG=Concentration of recovery gas gen- Vm(std)=Dry gas volume measured by the dry erated, ppm. gas meter, corrected to standard condi- CH2S=Verified concentration of H2S recovery tions, liters (dscf). gas. Vsoln=Total volume of solution in which the N=Normality of barium perchlorate titrant, sulfur dioxide sample is contained, 100 milliequivalents/ml. ml. Pbar=Barometric pressure at exit orifice of Vt=Volume of barium perchlorate titrant the dry gas meter, mm Hg (in. Hg). used for the sample, ml (average of rep- Pstd=Standard absolute pressure, 760 mm Hg licate titrations). (29.92 in. Hg). Vtb=Volume of barium perchlorate titrant QH2S=Calibrated flow rate of H2S recovery used for the blank, ml. gas, liters/min. Y=Dry gas meter calibration factor. QCG=Calibrated flow rate of combustion gas, 32.03=Equivalent weight of sulfur dioxide, liters/min. mg/meq. R=Recovery efficiency for the system per- 6.2 Dry Sample Gas Volume, Corrected to formance check, percent. Standard Conditions.
− =Tstd Pbar = VPm bar VVYm() std m KYl Eq.16 A- 1 Tm Pstd Tm
Where: K1=0.3858 °K/mm Hg for metric units.
6.3 Concentration of TRS as ppm SO2.
KVVNVV()()− / = 2 t tb soln a CTRS() ppm Eq.16 A- 2 Vm() std Where: 32.. 03mg 24 05liters 1 mole 1g 1000ml 1000µl µl K = = 12025 2 meq mole 64. 06g 1000mg liter 1ml meq
6.4 Concentration of Recovery Gas Gen- procedure may be used to verify the con- erated in the System Performance Check. centration of the recovery check gas. The H2S is collected from the calibration gas cyl- inder and is absorbed in zinc acetate solution (QCHSHS2 )( 2 ) C = Eq.16 A- 3 to form zinc sulfide. The latter compound is RG + QQH2 S CG then measured iodometrically. The method has been examined in the range of 5 to 1500 6.5 Recovery Efficiency for the System ppm. There are no known interferences to Performance Check. this method when used to analyze cylinder C gases containing H2S in nitrogen. Labora- R =TRS ×100Eq. 16 A- 4 tory tests have shown a relative standard de- C viation of less than 3 percent. The method RG showed no bias when compared to a gas 7. Alternative Procedures chromatographic method that used gravi- 7.1 Determination of H2S Content in Cyl- metrically certified permeation tubes for inder Gases. As an alternative to the proce- calibration. dures specified in section 3.1.4, the following
856
VerDate 11
7.1.1 Sampling Apparatus. The sampling train is shown in Figure 16A–4 and consists of the following components:
857
VerDate 11
1 7.1.1.1 Sampling Line. Teflon tubing ( ⁄4-in.) 7.1.3.4 Sodium Thiosulfate (Na2S203) Solu- to connect the cylinder regulator to the sam- tion, Standard 0.1 N. Same as in Method 11, pling valve. Section 6.3.1. Standardize according to Sec- 7.1.1.2 Needle Valve. Stainless steel or Tef- tion 7.1.8.2. lon needle valve to control the flow rate of 7.1.3.5 Na2S203 Solution, Standard 0.01 N. Pi- gases to the impingers. pette 100.0 ml of 0.1 N Na2S203 solution into a 7.1.1.3 Impingers. Three impingers of ap- 1-liter volumetric flask, and dilute to the proximately 100-ml capacity, constructed to mark with water. permit the addition of reagents through the 7.1.3.6 Iodine Solution, 0.1 N. Same as in gas inlet stem. The impingers shall be con- Method 11, Section 6.2.2. nected in series with leak-free glass or Tef- lon connectors. The impinger bottoms have a 7.1.3.7 Standard Iodine Solution, 0.01 N. Same as in Method 11, Section 6.2.3. Stand- standard 24⁄25 ground-glass fitting. The stems are from standard 1⁄4-in. (0.64-cm) ball joint ardize according to Section 7.1.8.3. midget impingers, custom lengthened by 7.1.3.8 Hydrochloric Acid (HCl) Solution, 10 about 1 in. When fitted together, the stem Percent by Weight. Add 230 ml concentrated end should be approximately 1⁄2 in. (1.27-cm) HCl (specific gravity 1.19) to 770 ml water. from the bottom (Southern Scientific, Inc., 7.1.3.9 Starch Indicator Solution. To 5 g Micanopy, Florida: Set Number S6962–048). starch (potato, arrowroot, or soluble), add a The third in-line impinger acts as a drop-out little cold water, and grind in a mortar to a bottle. thin paste. Pour into 1 liter of boiling water, 7.1.1.4 Drying Tube, Flowmeter, and Ba- stir, and let settle overnight. Use the clear rometer. Same as in Method 11, Sections supernatant. Preserve with 1.25 g salicylic 5.1.5, 5.1.8, and 5.1.10. acid, 4 g zinc chloride, or a combination of 4 7.1.1.5 Cylinder Gas Regulator. Stainless g sodium propionate and 2 g sodium azide per steel, to reduce the pressure of the gas liter of starch solution. Some commercial stream entering the Teflon sampling line to starch substitutes are satisfactory. a safe level. 7.1.4 Sampling Procedure. 7.1.1.6 Soap Bubble Meter. Calibrated for 100 and 500 ml, or two separate bubble me- 7.1.4.1 Selection of Gas Sample Volumes. ters. This procedure has been validated for esti- 7.1.1.7 Critical Orifice. For volume and rate mating the volume of cylinder gas sample measurements. The critical orifice may be needed when the H2S concentration is in the fabricated according to Section 7.1.4.3 and range of 5 to 1500 ppm. The sample volume must be calibrated as specified in Section ranges were selected in order to ensure a 35 7.1.8.4. to 60 percent consumption of the 20 ml of 0.01 7.1.1.8 Graduated Cylinder. 50-ml size. N iodine (thus ensuring a 0.01 N Na2S2O3 titer 7.1.1.9 Volumetric Flask. 1-liter size. of approximately 7 to 12 ml). The sample vol- 7.1.1.10 Volumetric Pipette. 15-ml size. umes for various H2S concentrations can be 7.1.1.11 Vacuum Gauge. Minimum 20-in. Hg estimated by dividing the approximate ppm- capacity. liters desired for a given concentration range 7.1.1.12 Stopwatch. by the H2S concentration stated by the man- 7.1.2 Sample Recovery and Analysis Appara- ufacturer. tus. 7.1.2.1 Erlenmeyer Flasks. 125- and 250-ml Approxi- sizes. Approximate cylinder gas H2S concentration mate ppm- 7.1.2.2 Pipettes. 2-, 10-, 20-, and 100-ml volu- (ppm) liters de- sired metric. 7.1.2.3 Burette. 50-ml size. 5 to <30 ...... 650 7.1.2.4 Volumetric Flask. 1-liter size. 30 to <500 ...... 800 7.1.2.5 Graduated Cylinder. 50-ml size. 500 to <1500 ...... 1000 7.1.2.6 Wash Bottle. 7.1.2.7 Stirring Plate and Bars. For example, for analyzing a cylinder gas 7.1.3 Reagents. Unless otherwise indicated, containing approximately 10 ppm H S, the all reagents shall conform to the specifica- 2 optimum sample volume is 65 liters (650 ppm- tions established by the Committee on Ana- liters/10 ppm). For analyzing a cylinder gas lytical Reagents of the American Chemical Society, where such specifications are avail- containing approximately 1000 ppm H2S, the able. Otherwise, use the best available grade. optimum sample volume is 1 liter (1000 ppm- 7.1.3.1 Water. Same as in Method 11, Sec- liters/1000 ppm). tion 6.1.3. 7.1.4.2 Critical Orifice Flow Rate Selection. 7.1.3.2 Zinc Acetate Absorbing Solution. Dis- The following table shows the ranges of sam- solve 20 g zinc acetate in water and dilute to ple flow rates that are desirable in order to 1 liter. ensure capture of H2S in the impinger solu- 7.1.3.3 Potassium Bi-iodate [KH(IO3)2 Solu- tion. Slight deviations from these ranges tion, Standard 0.100 N. Dissolve 3.249 g anhy- will not have an impact on measured con- drous KH(IO3)2 in water, and dilute to 1 liter. centrations. 858
VerDate 11
Critical orifice 7.1.4.4 Determination of Critical Orifice Ap- Cylinder gas H2S concentration (ppm) flow rate (ml/ proximate Flow Rate. Connect the critical ori- min) fice to the sampling system as shown in Fig- 5 to <50 ppm ...... 1500 ± 500 ure 16A–4 but without the H2S cylinder. Con- 50 to <250 ppm ...... 500 ± 250 nect a rotameter in the line to the first im- 250 to <1000 ppm ...... 200 ± 50 pinger. Turn on the pump, and adjust the ± <1000 ppm ...... 75 25 valve to give a reading of about half atmos- pheric pressure. Observe the rotameter read- 7.1.4.3 Critical Orifice Fabrication. Critical ing. Slowly increase the vacuum until a sta- orifice of desired flow rates may be fab- ricated by selecting an orifice tube of desired ble flow rate is reached, and record this as the critical vacuum. The measured flow rate length and connecting 1⁄16-in. x 1⁄4-in. (0.16-cm x 0.64-cm) reducing fittings to both ends. The indicates the expected critical flow rate of inside diameters and lengths of orifice tubes the orifice. If this flow rate is in the range needed to obtain specific flow rates are shown in Section 7.1.4.2, proceed with the shown below. critical orifice calibration according to Sec- tion 7.1.8.4. Flow 7.1.4.5 Determination of Approximate Sam- Tube Length rate Altech Tube (in. OD) catalog pling Time. Determine the approximate sam- (in. ID) (in.) (ml/ 1 min) No. pling time for a cylinder of known con- centration. Use the optimum sample volume 1/16 ...... 0.007 1.2 85 301430 1/16 ...... 0.01 3.2 215 300530 obtained in Section 7.1.4.1. 1/16 ...... 0.01 1.2 350 300530 1/16 ...... 0.02 1.2 1400 300230 1 Altech Associates, 2051 Waukegon Road., Deerfield, Illi- nois 60015.
Optimum volume ×1000 Approximate sampling time ()min = Critical orifice flow rate() ml / min
7.1.4.6 Sample Collection. Connect the Tef- the decrease in flow rate through the excess lon tubing, Teflon tee, and rotameter to the flow rotameter. This decrease should equal flow control needle valve as shown in Figure the known flow rate of the critical orifice 16A–4. Vent the rotameter to an exhaust being used. Continue sampling for the period hood. Plug the open end of the tee. Five to 10 determined in Section 7.1.4.5. minutes prior to sampling, open the cylinder When sampling is complete, turn off the valve while keeping the flow control needle pump and stopwatch. Disconnect the sam- valve closed. Adjust the delivery pressure to pling line from the tee and plug it. Close the 20 psi. Open the needle valve slowly until the needle valve followed by the cylinder valve. rotameter shows a flow rate approximately Record the sampling time. 50 to 100 ml above the flow rate of the criti- 7.1.5 Blank Analysis. While the sample is cal orifice being used in the system. being collected, run a blank as follows: To a Place 50 ml of zinc acetate solution in two 250-ml Erlenmeyer flask, add 100 ml of zinc of the impingers, connect them and the acetate solution, 20.0 ml. 0.01 N iodine solu- empty third impinger (dropout bottle) and tion, and 2 ml HCl solution. Titrate, while the rest of the equipment as shown in Figure stirring, with 0.01 N Na2S203 until the solu- 16A–4. Make sure the ground-glass fittings tion is light yellow. Add starch, and con- are tight. The impingers can be easily sta- tinue titrating until the blue color dis- bilized by using a small cardboard box in appears. Analyze a blank with each sample, which three holes have been cut, to act as a as the blank titer has been observed to holder. Connect the Teflon sample line to the change over the course of a day. first impinger. Cover the impingers with a NOTE: Iodine titration of zinc acetate solu- dark cloth or piece of plastic to protect the tions is difficult to perform because the solu- absorbing solution from light during sam- tion turns slightly white in color near the pling. end point, and the disappearance of the blue Record the temperature and barometric color is hard to recognize. In addition, a blue pressure. Note the gas flow rate through the color may reappear in the solution about 30 rotameter. Open the closed end of the tee. to 45 seconds after the titration endpoint is Connect the sampling tube to the tee, ensur- reached. This should not be taken to mean ing a tight connection. Start the sampling the original endpoint was in error. It is rec- pump and stopwatch simultaneously. Note ommended that persons conducting this test
859
VerDate 11
perform several titrations to be able to cor- agree within 0.05 ml. Take the average vol- rectly identify the endpoint. The importance ume of Na2S2O3 consumed to calculate the of this should be recognized because the re- normality to three decimal figures using sults of this analytical procedure are ex- Equation 16A–5. tremely sensitive to errors in titration. 7.1.8.3 Iodine Solution, 0.01 N. Standardize 7.1.6 Sample Analysis. Sample treatment is the 0.01 N iodine solution as follows: Pipet similar to the blank treatment. Before de- 20.0 ml of 0.01 N iodine solution into a 125-ml taching the stems from the bottoms of the Erlenmeyer flask. Titrate with standard 0.01 impingers, add 20.0 ml of 0.01 N iodine solu- N Na2S2O3 solution until the solution is light tion through the stems of the impingers hold- yellow. Add 3 ml starch solution, and con- ing the zinc acetate solution, dividing it be- tinue titrating until the blue color just dis- tween the two (add about 15 ml to the first appears. impinger and the rest to the second). Add 2 If the normality of the iodine tested is not ml HCl solution through the stems, dividing 0.010, add a few ml of 0.1 N iodine solution if it as with the iodine. Disconnect the sam- it is low, or a few ml of water if it is high, pling line, and store the impingers for 30 and standardize again. Repeat the titration minutes. At the end of 30 minutes, rinse the until replicate values agree within 0.05 ml. impinger stems into the impinger bottoms. Take the average volume to calculate the Titrate the impinger contents with 0.01 N normality to three decimal figures using Equation 16A–6. Na2S203. Do not transfer the contents of the impinger to a flask because this may result 7.1.8.4 Critical Orifice. Calibrate the criti- in a loss of iodine and cause a positive bias. cal orifice using the sampling train shown in 7.1.7 Post-test Orifice Calibration. Conduct a Figure 16A–4 but without the H2S cylinder post-test critical orifice calibration run and vent rotameter. Connect the soap bubble using the calibration procedures outlined in meter to the Teflon line that is connected to the first impinger. Turn on the pump, and Section 7.1.8.4. If the Qstd obtained before and after the test differs by more than 5 percent, adjust the needle valve until the vaccum is void the sample; if not, proceed to perform higher than the critical vacuum determined the calculations. in Section 7.1.4.4. Record the time required 7.1.8 Calibrations and Standardizations. for gas flow to equal the soap bubble meter volume (use the 100-ml soap bubble meter for 7.1.8.1 Rotameter and Barometer. Same as gas flow rates below 100 ml/min, otherwise in Method 11, Sections 8.2.3 and 8.2.4. use the 500-ml soap bubble meter). Make 7.1.8.2 Na S O Solution, 0.1 N. Standard- 2 2 3 three runs, and record the data listed in ize the 0.1 N Na S O solution as follows: To 2 2 3 Table 1. Use these data to calculate the volu- 80 ml water, stirring constantly, add 1 ml metric flow rate of the orifice. concentrated H SO , 10.0 ml 0.100 N KH(IO ) 2 4 3 2 7.1.9 Calculations. and 1 g potassium iodide. Titrate imme- 7.1.9.1 Nomenclature. diately with 0.1 N Na2S2O3 until the solution is light yellow. Add 3 ml starch solution, and Bwa=Fraction of water vapor in ambient air titrate until the blue color just disappears. during orifice calibration. Repeat the titration until replicate analyses CH2S=H2S concentration in cylinder gas, ppm.
17. 03g 24.05liters H S 1 mole H S 103 ml K= Conversion factor =12025 ml/ eq = 2 2 − g eq mole H2 S 34.06 g H2 S liter
Ma=Molecular weight of ambient air satu- Qstd=Volumetric flow rate through critical rated at impinger temperature, g/g-mole. orifice, liters/min. Ms=Molecular weight of sample gas (nitro- Datellllllllll gen) saturated at impinger temperature, llllllllll g/g-mole. (For tests carried out in a lab- Critical orifice ID oratory where the impinger temperature Soap bubble meter volume, Vsb lll liters Θ is 25 °C, Ma=28.5 g/g-mole and Ms=27.7 g/g- Time, sb mole.) Run no. 1 lll min lll sec lll lll NI=Normality of standard iodine solution Run no. 2 min sec (0.01 N), g-eq/liter. Run no. 3 lll min lll sec Average lll min lll sec NT=Normality of standard Na2S2O3 solution (0.01 N), g-eq/liter. Convert the seconds to fraction of minute: Time Pbar=Barometric pressure, mm Hg. = lll min + lll Sec/60 Pstd=Standard absolute pressure, 760 mm Hg. = lll min
860
VerDate 11
Barometric pressure, Pbar = lll mm Hg Pump vacuum, = lll mm Hg. (This should Ambient temperature, tamb = 273 + lll °C be approximately 0.4 times barometric pres- = lll °K sure.)
− (VTP )( )( )()10 3 = sb std bar = − − − − − Vsb() std liters (TPamb )( std )
Vsb() std KNVV()− Q = = − − − − − − liters / min C = T TB T Eq. 16 A- 8 std θ HS2 sb Vm() std
TABLE 1—CRITICAL ORIFICE CALIBRATION DATA. 8. Bibliography Qstd, average= Average standard flow rate through critical orifice, liters/min. 1. American Public Health Association, Qstd, before= Average standard flow rate through American Water Works Association, and critical orifice determined before H2S Water Pollution Control Federation. Stand- sampling (Section 7.1.4.4), liters/min. ard Methods for the Examination of Water Qstd, after= Average standard flow rate through and Wastewater. Washington, DC. American critical orifice determined after H2S Public Health Association. 1975. p. 316–317. sampling (Section 7.1.7), liters/min. 2. American Society for Testing and Mate- ° Tamb = Absolute ambient temperature, K. rials. Annual Book of ASTM Standards. Part T = Standard absolute temperature, 293 °K. std 31: Water, Atmospheric Analysis. Philadel- Θ = Sampling time, min. s phia, PA. 1974. p. 40–42. Θsb = Time for soap bubble meter flow rate measurement, min. 3. Blosser, R.O. A Study of TRS Measure- ment Methods. National Council of the Paper Vm(std) = Sample gas volume measured by the critical orifice, corrected to standard Industry for Air and Stream Improvement, conditions, liters. Inc., New York, NY. Technical Bulletin No. 434. May 1984. 14 p. Vsb = Volume of gas as measured by the soap bubble meter, ml. 4. Blosser, R.O., H.S. Oglesby, and A.K. Vsb(std) = Volume of gas as measured by the Jain. A Study of Alternate SO2 Scrubber De- soap bubble meter, corrected to standard signs Used for TRS Monitoring. A Special conditions, liters. Report by the National Council of the Paper VI = Volume of standard iodine solution (0.01 Industry for Air and Stream Improvement, N) used, ml. Inc., New York, NY. July 1977. VT = Volume of standard Na2S2O3 solution 5. Curtis, F., and G.D. McAlister. Develop- (0.01 N) used, ml. ment and Evaluation of an Oxidation/Method VTB = Volume of standard Na2S2O3 solution 6 TRS Emission Sampling Procedure. Emis- (0.01 N) used for the blank, ml. sion Measurement Branch, Emission Stand- 7.1.9.2 Normality of Standard Na2S2O3 So- ards and Engineering Division, U.S. Environ- lution (0.1. N). mental Protection Agency, Research Tri- angle Park, NC 27711. February 1980. 1 N = Eq. 16 A- 5 6. Gellman, I. A Laboratory and Field T ml Na S O Consumed Study of Reduced Sulfur Sampling and Mon- 2 2 3 itoring Systems. National Council of the 7.1.9.3 Normality of Standard Iodine Solu- Paper Industry for Air and Stream Improve- tion (0.01 N). ment, Inc., New York, NY. Atmospheric Quality Improvement Technical Bulletin No. = NVTT 81. October 1975. NI Eq. 16 A- 6 7. Margeson, J.H., J.E. Knoll, and M.R. VI Midgett. A Manual Method for TRS Deter- 7.1.9.4 Sample Gas Volume. mination. Draft Available from the authors. Source Branch, Quality Assurance Division, = θ − Ma U.S. Environmental Protection Agency, Re- VQBm() std() std() s()1 wa Eq. 16 A- 7 search Triangle Park, NC 27711. Mb 8. National Council of the Paper Industry 7.1.9.5 Concentration of H2S in the Gas for Air and Stream Improvement. An Inves- Cylinder. tigation of H2S and SO2 Calibration Cylinder 861
VerDate 11
Gas Stability and Their Standardization sources with emission levels between 10 and Using Wet Chemical Techniques. Special Re- 100 ppm, the measuring range can be best ex- port 76–06. New York, NY. August 1976. tended by reducing the sample size. 9. National Council of the Paper Industry 1.4 Interferences. The TRS compounds for Air and Stream Improvement. Wet Chem- other than those regulated by the emission ical Method for Determining the H2S Con- standards, if present, may be measured by centration of Calibration Cylinder Gases. this method. Therefore, carbonyl sulfide, Technical Bulletin Number 450. New York, which is partially oxidized to SO2 and may NY. January 1985. 23 p. be present in a lime kiln exit stack, would be 10. National Council of the Paper Industry a positive interferent. for Air and Stream Improvement. Modified Particulate matter from the lime kiln Wet Chemcial Method for Determining the stack gas (primarily calcium carbonate) can H2S Concentration of Calibration Cylinder cause a negative bias if it is allowed to enter Gases. Draft Report. New York, NY. March the citrate scrubber; the particulate matter 1987. 29 p. will cause the pH to rise and H2S to be ab- sorbed before oxidation. Proper use of the METHOD 16B—DETERMINATION OF TOTAL RE- particulate filter, described in Section 2.1.3 DUCED SULFUR EMISSIONS FROM STATION- of Method 16A, will eliminate this inter- ARY SOURCES ference. 1. Applicability, Principle, Range and Sensitiv- Carbon monoxide (CO) and carbon dioxide ity, Interferences, and Precision and Accuracy (CO2) have substantial desensitizing effects 1.1 Applicability. This method is applicable on the FPD even after dilution. Acceptable to the determination of total reduced sulfur systems must demonstrate that they have (TRS) emissions from recovery furnaces, eliminated this interference by some proce- lime kilns, and smelt dissolving tanks at dure such as eluting these compounds before kraft pulp mills, and from other sources the SO2. Compliance with this requirement when specified in an applicable subpart of can be demonstrated by submitting the regulations. The TRS compounds include chromatograms of calibration gases with and hydrogen sulfide (H2S), methyl mercaptan, without CO2 in diluent gas. The CO2 level dimethyl sulfide, and dimethyl disulfide. should be approximately 10 percent for the The flue gas must contain at least 1 per- case with CO2 present. The two cent oxygen for complete oxidation of all chromatograms should show agreement TRS to sulfur dioxide (SO2). within the precision limits of Section 1.5. 1.2 Principle. An integrated gas sample is 1.5 Precision and Accuracy. The GC/FPD extracted from the stack. The SO2 is re- and dilution calibration precision and drift, moved selectively from the sample using a and the system calibration accuracy are the citrate buffer solution. The TRS compounds same as in Method 16, Sections 4.1 to 4.3. are then thermally oxidized to SO2 and ana- Field tests between this method and Meth- lyzed as SO2 by gas chromatography (GC) od 16A showed an average difference of less using flame photometric detection (FPD). than 4.0 percent. This difference was not de- 1.3 Range and Sensitivity. Coupled with a termined to be significant. GC utilizing a 1-ml sample size, the maxi- 2. Apparatus mum limit of the FPD for SO2 is approxi- 2.1 Sampling. A sampling train is shown in mately 10 ppm. This limit is expanded by di- Figure 16B–1. Modifications to the apparatus lution of the sample gas before analysis or are accepted provided the system perform- by reducing the sample aliquot size. For ance check is met.
862
VerDate 11
2.1.1 Probe, Probe Brush, Particulate Filter, be used for calibration. The calibration gas SO2 Scrubber, Combustion Tube, and Furnace. is used to calibrate the GC/FPD system and Same as in Method 16A, Sections 2.1.1 to the dilution system. 2.1.6. 3.2 Recovery Check Gas. Hydrogen sulfide 2.1.2 Sampling Pump. Leakless Teflon- (100 ppm or less) in nitrogen, stored in alu- coated diaphragm type or equivalent. minum cylinders. Verify the concentration 2.2 Analysis. by Method 11, the procedure discussed in 2.2.1 Dilution System (optional), Gas Chro- Section 7.1 of Method 16A, or gas chroma- matograph, Oven, Temperature Gauges, Flow tography where the instrument is calibrated System, Flame Photometric Detector, Elec- with an H2S permeation tube as described trometer, Power Supply, Recorder, Calibration below. For the wet-chemical methods, the System, Tube Chamber, Flow System, and Con- standard deviation should not exceed 5 per- stant Temperature Bath. Same as in Method cent on at least three 20-minute runs. 16, Sections 5.2, 5.4, and 5.5. Hydrogen sulfide recovery gas generated 2.2.2 Gas Chromatograph Columns. Same as from a permeation device gravimetically in Method 16, Section 12.1.4.1.1. Other col- calibrated and certified at some convenient umns with demonstrated ability to resolve operation temperature may be used. The per- SO2 and be free from known interferences are meation rate of the device must be such that acceptable alternatives. at a dilution gas flow rate of 3 liters/min, an 3. Reagents H2S concentration in the range of the stack Same as in Method 16, Section 6, except gas or within 20 percent of the standard can the following: be generated. 3.1 Calibration Gas. SO2 permeation tube 3.3 Combustion Gas. Gas containing less gravimetrically calibrated and certified at than 50 ppb reduced sulfur compounds and some convenient operating temperature. less than 10 ppm total hydrocarbons. The gas These tubes consist of hermetically sealed may be generated from a clean-air system FEP Teflon tubing in which a liquefied gase- that purifies ambient air and consists of the ous substance is enclosed. The enclosed gas following components: diaphragm pump, sili- permeates through the tubing wall at a con- ca gel drying tube, activated charcoal tube, stant rate. When the temperature is con- and flow rate measuring device. Gas from a stant, calibration gases covering a wide compressed air cylinder is also acceptable. range of known concentrations can be gen- 4. Pretest Procedures erated by varying and accurately measuring Same as in Method 16, Section 7. the flow rate of diluent gas passing over the 5. Calibration tubes. In place of SO2 permeation tubes, Na- Same as in Method 16, Section 8, except tional Bureau of Standards traceable cyl- SO2 is used instead of H2S. inder gases containing SO2 in nitrogen may 6. Sampling and Analysis Procedure 863
VerDate 11
6.1 Sampling. Before any source sampling nical Bulletin No. 434. New York, NY. May is done, conduct a system performance check 1984. 12 p. as detailed in Section 7.1 to validate the 3. Margeson, J.H., J.E. Knoll, and M.R. sampling train components and procedures. Midgett. A Manual Method for TRS Deter- Although this test is optional, it would sig- mination. Draft available from the authors. nificantly reduce the possibility of rejecting Source Branch, Quality Assurance Division, tests as a result of failing the post-test per- U.S. Environmental Protection Agency, Re- formance check. At the completion of the search Triangle Park, NC 27711. pretest system performance check, insert the sampling probe into the test port making METHOD 17—DETERMINATION OF PARTICULATE certain that no dilution air enters the stack EMISSIONS FROM STATIONARY SOURCES (IN- through the port. Condition the entire sys- STACK FILTRATION METHOD) tem with sample for a minimum of 15 min- Introduction utes before beginning analysis. If the sample Particulate matter is not an absolute is diluted, determine the precise dilution fac- quantity; rather, it is a function of tempera- tor as in Section 8.5 of Method 16. ture and pressure. Therefore, to prevent vari- 6.2 Analysis. Pass aliquots of diluted sam- ability in particulate matter emission regu- ple through the SO scrubber and oxidation 2 lations and/or associated test methods, the furnace, and then inject into the GC/FPD an- temperature and pressure at which particu- alyzer for analysis. The rest of the analysis late matter is to be measured must be care- is the same as in Method 16, Sections 9.2.1 fully defined. Of the two variables (i.e., tem- and 9.2.2. perature and pressure), temperature has the 7. Post-Test Procedures greater effect upon the amount of particu- 7.1 System Performance Check. Same as in late matter in an effluent gas stream; in Method 16A, Section 4.3. Sufficient numbers most stationary source categories, the effect of sample injections should be made so that of pressure appears to be negligible. the precision requirements of Section 4.1 of In Method 5, 250°F is established as a nomi- Method 16 are satisfied. nal reference temperature. Thus, where 7.2 Recalibration. Same as in Method 16, Method 5 is specified in an applicable sub- Section 10.2. part of the standards, particulate matter is 7.3 Determination of Calibration Drift. Same defined with respect to temperature. In order as in Method 16, Section 10.3. to maintain a collection temperature of 8. Calculations 250°F, Method 5 employs a heated glass sam- 8.1 Nomenclature. ple probe and a heated filter holder. This CSO2 = Sulfur dioxide concentration, ppm. equipment is somewhat cumbersome and re- CTRS = Total reduced sulfur concentration as quires care in its operation. Therefore, where determined by Equation 16B–1, ppm. particulate matter concentrations (over the d = Dilution factor, dimensionless. normal range of temperature associated with N = Number of samples. a specified source category) are known to be 8.2 SO2 Concentration. Determine the con- independent of temperature, it is desirable centration of SO2 (CSO2) directly from the to eliminate the glass probe and heating sys- calibration curves. Alternatively, the con- tems, and sample at stack temperature. centration may be calculated using the equa- This method describes an in-stack sam- tion for the least-squares line. pling system and sampling procedures for 8.3 TRS Concentration. use in such cases. It is intended to be used
CTRS = (CSO2) (d) only when specified by an applicable subpart Eq. 16B–1 of the standards, and only within the appli- cable temperature limits (if specified), or 8.4 Average TRS Concentration. when otherwise approved by the Adminis- trator. 1. Principle and Applicability 1.1 Principle. Particulate matter is with- drawn isokinetically from the source and collected on a glass fiber filter maintained at 9. Example System stack temperature. The particulate mass is Same as in Method 16, Section 12. Single determined gravimetrically after removal of column systems using the column in Section uncombined water. 12.1.4.1.1 of Method 16 or a 7-ft Carbosorb B 1.2 Applicability. This method applies to HT 100 column have been found satisfactory the determination of particulate emissions in resolving SO2 from CO2. from stationary sources for determining 10. Bibliography compliance with new source performance 1. Same as in Method 16, Sections 13.1 to standards, only when specifically provided 13.6. for in an applicable subpart of the standards. 2. National Council of the Paper Industry This method is not applicable to stacks that for Air and Stream Improvement, Inc. A contain liquid droplets or are saturated with Study of TRS Measurement Methods. Tech- water vapor. In addition, this method shall
864
VerDate 11
not be used as written if the projected cross- 2.1 Sampling Train. A schematic of the sectional area of the probe extension-filter sampling train used in this method is shown holder assembly covers more than 5 percent in Figure 17–1. Construction details for of the stack cross-sectional area (see Section many, but not all, of the train components 4.1.2). are given in APTD–0581 (Citation 2 in Bibli- ography); for changes from the APTD–0581 2. Apparatus document and for allowable modifications to Figure 17–1, consult with the Administrator.
865
VerDate 11
The operating and maintenance procedures Source-sampling assemblies that do not for many of the sampling train components meet the minimum spacing requirements of are described in APTD–0576 (Citation 3 in Figure 17–1 (or the equivalent of these re- Bibliography). Since correct usage is impor- quirements, e.g., Figure 2–7 of Method 2) may tant in obtaining valid results, all users be used; however, the pitot tube coefficients should read the APTD–0576 document and of such assemblies shall be determined by adopt the operating and maintenance proce- calibration, using methods subject to the ap- dures outlined in it, unless otherwise speci- proval of the Administrator. fied herein. The sampling train consists of 2.1.5 Differential Pressure Gauge. Inclined the following components: manometer or equivalent device (two), as de- 2.1.1 Probe Nozzle. Stainless steel (316) or scribed in Section 2.2 of Method 2. One ma- glass, with sharp, tapered leading edge. The nometer shall be used for velocity head (∆ p) angle of taper shall be 30° and the taper shall readings, and the other, for orifice differen- be on the outside to preserve a constant in- tial pressure readings. ternal diameter. The probe nozzle shall be of 2.1.6 Condenser. It is recommended that the button-hook or elbow design, unless oth- the impinger system described in Method 5 erwise specified by the Administrator. If be used to determine the moisture content of made of stainless steel, the nozzle shall be the stack gas. Alternatively, any system constructed from seamless tubing. Other ma- that allows measurement of both the water terials of construction may be used subject condensed and the moisture leaving the con- to the approval of the Administrator. denser, each to within 1 ml or 1 g, may be A range of sizes suitable for isokinetic used. The moisture leaving the condenser can be measured either by: (1) monitoring sampling should be available, e.g., 0.32 to 1.27 the temperature and pressure at the exit of cm (1⁄8 to 1⁄2 in.)—or larger if higher volume the condenser and using Dalton’s law of par- sampling trains are used—inside diameter tial pressures; or (2) passing the sample gas (ID) nozzles in increments of 0.16 cm (1⁄16 in.). stream through a silica gel trap with exit Each nozzle shall be calibrated according to gases kept below 20°C (68°F) and determining the procedures outlined in Section 5.1. the weight gain. 2.1.2 Filter Holder. The in-stack filter Flexible tubing may be used between the holder shall be constructed of borosilicate or probe extension and condenser. If means quartz glass, or stainless steel; if a gasket is other than silica gel are used to determine used, it shall be made of silicone rubber, Tef- the amount of moisture leaving the con- lon, or stainless steel. Other holder and gas- denser, it is recommended that silica gel ket materials may be used subject to the ap- still be used between the condenser system proval of the Administrator. The filter hold- and pump to prevent moisture condensation er shall be designed to provide a positive seal in the pump and metering devices and to against leakage from the outside or around avoid the need to make corrections for mois- the filter. ture in the metered volume. 2.1.3 Probe Extension. Any suitable rigid 2.1.7 Metering System. Vacuum gauge, probe extension may be used after the filter leak-free pump, thermometers capable of holder. measuring temperature to within 3°C (5.4°F), 2.1.4 Pitot Tube. Type S, as described in dry gas meter capable of measuring volume Section 2.1 of Method 2, or other device ap- to within 2 percent, and related equipment, proved by the Administrator; the pitot tube as shown in Figure 17–1. Other metering sys- shall be attached to the probe extension to tems capable of maintaining sampling rates allow constant monitoring of the stack gas within 10 percent of isokinetic and of deter- velocity (see Figure 17–1). The impact (high mining sample volumes to within 2 percent pressure) opening plane of the pitot tube may be used, subject to the approval of the shall be even with or above the nozzle entry Administrator. When the metering system is plane during sampling (see Method 2, Figure used in conjunction with a pitot tube, the 2–6b). It is recommended: (1) that the pitot system shall enable checks of isokinetic tube have a known baseline coefficient, de- rates. termined as outlined in Section 4 of Method Sampling trains utilizing metering sys- 2; and (2) that this known coefficient be pre- tems designed for higher flow rates than that served by placing the pitot tube in an inter- described in APTD–0581 or APTD–0576 may be ference-free arrangement with respect to the used provided that the specifications of this sampling nozzle, filter holder, and tempera- method are met. ture sensor (see Figure 17–1). Note that the 2.1.8 Barometer. Mercury, aneroid, or 1.9 cm (0.75 in.) free-space between the nozzle other barometer capable of measuring at- and pitot tube shown in Figure 17–1, is based mospheric pressure to within 2.5 mm Hg (0.1 on a 1.3 cm (0.5 in.) ID nozzle. If the sampling in. Hg). In many cases, the barometric read- train is designed for sampling at higher flow ing may be obtained from a nearby National rates than that described in APTD–0581, thus Weather Service station, in which case the necessitating the use of larger sized nozzles, station value (which is the absolute baro- the free-space shall be 1.9 cm (0.75 in.) with metric pressure) shall be requested and an the largest sized nozzle in place. adjustment for elevation differences between
866
VerDate 11
the weather station and sampling point shall 2.3.4 Balance. To measure to within 0.5 be applied at a rate of minus 2.5 mm Hg (0.1 mg. in. Hg) per 30 m (100 ft) elevation increase or 2.3.5 Beakers. 250 ml. vice versa for elevation decrease. 2.3.6 Hygrometer. To measure the relative 2.1.9 Gas Density Determination Equip- humidity of the laboratory environment. ment. Temperature sensor and pressure 2.3.7 Temperature Gauge. To measure the gauge, as described in Sections 2.3 and 2.4 of temperature of the laboratory environment. Method 2, and gas analyzer, if necessary, as 3. Reagents described in Method 3. 3.1 Sampling. The temperature sensor shall be attached 3.1.1 Filters. The in-stack filters shall be to either the pitot tube or to the probe ex- glass mats or thimble fiber filters, without tension, in a fixed configuration. If the tem- organic binders, and shall exhibit at least perature sensor is attached in the field, the 99.95 percent efficiency (0.05 percent penetra- sensor shall be placed in an interference-free tion) on 0.3 micron dioctyl phthalate smoke arrangement with respect to the Type S particles. The filter efficiency tests shall be pitot tube openings (as shown in Figure 17–1 conducted in accordance with ASTM Stand- or in Figure 2–7 of Method 2). Alternatively, ard Method D2986–71 (Reapproved 1978) (in- the temperature sensor need not be attached corporated by reference—see § 60.17). Test to either the probe extension or pitot tube data from the supplier’s quality control pro- during sampling, provided that a difference gram are sufficient for this purpose. of not more than 1 percent in the average ve- 3.1.2 Silica Gel. Indicating type, 6- to 16- locity measurement is introduced. This al- mesh. If previously used, dry at 175°C (350°F) ternative is subject to the approval of the for 2 hours. New silica gel may be used as re- Administrator. ceived. Alternatively, other types of 2.2 Sample Recovery. desiccants (equivalent or better) may be 2.2.1 Probe Nozzle Brush. Nylon bristle used, subject to the approval of the Adminis- brush with stainless steel wire handle. The trator. brush shall be properly sized and shaped to 3.1.3 Crushed Ice. brush out the probe nozzle. 3.1.4 Stopcock Grease. Acetone-insoluble, 2.2.2 Wash Bottles—Two. Glass wash bot- heat-stable silicone grease. This is not nec- tles are recommended; polyethylene wash essary if screw-on connectors with Teflon bottles may be used at the option of the test- sleeves, or similar, are used. Alternatively, er. It is recommended that acetone not be other types of stopcock grease may be used, stored in polyethylene bottles for longer subject to the approval of the Administrator. than a month. 3.1.5 Water. Same as in Method 5, section 2.2.3 Glass Sample Storage Containers. 3.1.3. Chemically resistant, borosilicate glass bot- 3.2 Sample Recovery. Acetone, reagent tles, for acetone washes, 500 ml or 1000 ml. grade, 0.001 percent residue, in glass bottles. Screw cap liners shall either be rubber- Acetone from metal containers generally has backed Teflon or shall be constructed so as a high residue blank and should not be used. to be leak-free and resistant to chemical at- Sometimes, suppliers transfer acetone to tack by acetone. (Narrow mouth glass bot- glass bottles from metal containers. Thus, tles have been found to be less prone to leak- acetone blanks shall be run prior to field use age.) Alternatively, polyethylene bottles and only acetone with low blank values ( may be used. 0.001 percent) shall be used. In no case shall 2.2.4 Petri Dishes. For filter samples; a blank value of greater than 0.001 percent of glass or polyethylene, unless otherwise spec- the weight of acetone used be subtracted ified by the Administrator. from the sample weight. 2.2.5 Graduated Cylinder and/or Balance. 3.3 Analysis. To measure condensed water to within 1 ml 3.3.1 Acetone. Same as 3.2. or 1 g. Graduated cylinders shall have sub- 3.3.2 Desiccant. Anhydrous calcium sul- divisions no greater than 2 ml. Most labora- fate, indicating type. Alternatively, other tory balances are capable of weighing to the types of desiccants may be used, subject to nearest 0.5 g or less. Any of these balances is the approval of the Administrator. suitable for use here and in Section 2.3.4. 2.2.6 Plastic Storage Containers. Air tight 4. Procedure containers to store silica gel. 4.1 Sampling. The complexity of this 2.2.7 Funnel and Rubber Policeman. To method is such that, in order to obtain reli- aid in transfer of silica gel to container; not able results, testers should be trained and necessary if silica gel is weighed in the field. experienced with the test procedures. 2.2.8 Funnel. Glass or polyethylene, to aid 4.1.1 Pretest Preparation. All components in sample recovery. shall be maintained and calibrated according 2.3 Analysis. to the procedure described in APTD–0576, un- 2.3.1 Glass Weighing Dishes. less otherwise specified herein. 2.3.2 Desiccator. Weigh several 200 to 300 g portions of silica 2.3.3 Analytical Balance. To measure to gel in air-tight containers to the nearest 0.5 within 0.1 mg. g. Record the total weight of the silica gel
867
VerDate 11
plus container, on each container. As an al- as specified by the Administrator. Make a ternative, the silica gel need not be projected-area model of the probe extension- preweighed, but may be weighed directly in filter holder assembly, with the pitot tube its impinger or sampling holder just prior to face openings positioned along the centerline train assembly. of the stack, as shown in Figure 17–2. Cal- Check filters visually against light for culate the estimated cross-section blockage, irregularities and flaws or pinhole leaks. as shown in Figure 17–2. If the blockage ex- Label filters of the proper size on the back ceeds 5 percent of the duct cross sectional side near the edge using numbering machine area, the tester has the following options: (1) ink. As an alternative, label the shipping a suitable out-of-stack filtration method containers (glass or plastic petri dishes) and may be used instead of in-stack filtration; or keep the filters in these containers at all (2) a special in-stack arrangement, in which times except during sampling and weighing. the sampling and velocity measurement Desiccate the filters at 20±5.6°C (68±10°F) sites are separate, may be used; for details and ambient pressure for at least 24 hours and weigh at intervals of at least 6 hours to concerning this approach, consult with the a constant weight, i.e., 0.5 mg change from Administrator (see also Citation 10 in Bibli- previous weighing; record results to the ography). Determine the stack pressure, nearest 0.1 mg. During each weighing the fil- temperature, and the range of velocity heads ter must not be exposed to the laboratory at- using Method 2; it is recommended that a mosphere for a period greater than 2 minutes leak-check of the pitot lines (see Method 2, and a relative humidity above 50 percent. Al- Section 3.1) be performed. Determine the ternatively (unless otherwise specified by moisture content using Approximation the Administrator), the filters may be oven Method 4 or its alternatives for the purpose dried at 105°C (220°F) for 2 to 3 hours, des- of making isokinetic sampling rate settings. iccated for 2 hours, and weighed. Procedures Determine the stack gas dry molecular other than those described, which account weight, as described in Method 2, Section 3.6; for relative humidity effects, may be used, if integrated Method 3 sampling is used for subject to the approval of the Administrator. molecular weight determination, the inte- 4.1.2 Preliminary Determinations. Select grated bag sample shall be taken simulta- the sampling site and the minimum number neously with, and for the same total length of sampling points according to Method 1 or of time as, the particulate sample run.
868
VerDate 11
Select a nozzle size based on the range of chosen for the range of velocity heads en- velocity heads, such that it is not necessary countered (see Section 2.2 of Method 2). to change the nozzle size in order to main- Select a probe extension length such that tain isokinetic sampling rates. During the all traverse points can be sampled. For large run, do not change the nozzle size. Ensure stacks, consider sampling from opposite that the proper differential pressure gauge is
869
VerDate 11
sides of the stack to reduce the length of 4.1.4.1 Pretest Leak-Check. A pretest probes. leak-check is recommended, but not re- Select a total sampling time greater than quired. If the tester opts to conduct the pre- or equal to the minimum total sampling test leak-check, the following procedure time specified in the test procedures for the shall be used. specific industry such that (1) the sampling After the sampling train has been assem- time per point is not less than 2 minutes (or bled, plug the inlet to the probe nozzle with some greater time interval if specified by the a material that will be able to withstand the Administrator), and (2) the sample volume stack temperature. Insert the filter holder taken (corrected to standard conditions) will into the stack and wait approximately 5 exceed the required minimum total gas sam- minutes (or longer, if necessary) to allow the ple volume. The latter is based on an approx- system to come to equilibrium with the tem- imate average sampling rate. perature of the stack gas stream. Turn on It is recommended that the number of min- the pump and draw a vacuum of at least 380 utes sampled at each point be an integer or mm Hg (15 in. Hg); note that a lower vacuum an integer plus one-half minute, in order to may be used, provided that it is not exceeded avoid timekeeping errors. during the test. Determine the leakage rate. In some circumstances, e.g., batch cycles, A leakage rate in excess of 4 percent of the 3 it may be necessary to sample for shorter average sampling rate or 0.00057 m /min. (0.02 times at the traverse points and to obtain cfm), whichever is less, is unacceptable. smaller gas sample volumes. In these cases, The following leak-check instructions for the sampling train described in APTD–0576 the Administrator’s approval must first be and APTD–0581 may be helpful. Start the obtained. pump with by-pass valve fully open and 4.1.3 Preparation of Collection Train. Dur- coarse adjust valve completely closed. Par- ing preparation and assembly of the sam- tially open the coarse adjust valve and slow- pling train, keep all openings where con- ly close the by-pass valve until the desired tamination can occur covered until just vacuum is reached. Do not reverse direction prior to assembly or until sampling is about of by-pass valve. If the desired vacuum is ex- to begin. ceeded, either leak-check at this higher vac- If impingers are used to condense stack gas uum or end the leak-check as shown below moisture, prepare them as follows: place 100 and start over. ml of water in each of the first two When the leak-check is completed, first impingers, leave the third impinger empty, slowly remove the plug from the inlet to the and transfer approximately 200 to 300 g of probe nozzle and immediately turn off the preweighed silica gel from its container to vacuum pump. This prevents water from the fourth impinger. More silica gel may be being forced backward and keeps silica gel used, but care should be taken to ensure that from being entrained backward. it is not entrained and carried out from the 4.1.4.2 Leak-Checks During Sample Run. impinger during sampling. Place the con- If, during the sampling run, a component tainer in a clean place for later use in the (e.g., filter assembly or impinger) change be- sample recovery. Alternatively, the weight comes necessary, a leak-check shall be con- of the silica gel plus impinger may be deter- ducted immediately before the change is mined to the nearest 0.5 g and recorded. made. The leak-check shall be done accord- If some means other than impingers is used ing to the procedure outlined in Section to condense moisture, prepare the condenser 4.1.4.1 above, except that it shall be done at (and, if appropriate, silica gel for condenser a vacuum equal to or greater than the maxi- outlet) for use. mum value recorded up to that point in the Using a tweezer or clean disposable sur- test. If the leakage rate is found to be no gical gloves, place a labeled (identified) and greater than 0.00057 m3/min (0.02 cfm) or 4 weighed filter in the filter holder. Be sure percent of the average sampling rate (which- that the filter is properly centered and the ever is less), the results are acceptable, and gasket properly placed so as not to allow the no correction will need to be applied to the sample gas stream to circumvent the filter. total volume of dry gas metered; if, however, Check filter for tears after assembly is com- a higher leakage rate is obtained, the tester pleted. Mark the probe extension with heat shall either record the leakage rate and plan resistant tape or by some other method to to correct the sample volume as shown in denote the proper distance into the stack or Section 6.3 of this method, or shall void the duct for each sampling point. sampling run. Assemble the train as in Figure 17–1, using Immediately after component changes, a very light coat of silicone grease on all leak-checks are optional; if such leak-checks ground glass joints and greasing only the are done, the procedure outlined in Section outer portion (see APTD–0576) to avoid possi- 4.1.4.1 above shall be used. bility of contamination by the silicone 4.1.4.3 Post-Test Leak-Check. A leak- grease. Place crushed ice around the check is mandatory at the conclusion of each impingers. sampling run. The leak-check shall be done 4.1.4 Leak Check Procedures. in accordance with the procedures outlined
870
VerDate 11
in Section 4.1.4.1, except that it shall be con- For each run, record the data required on ducted at a vacuum equal to or greater than the example data sheet shown in Figure 17– the maximum value reached during the sam- 3. Be sure to record the initial dry gas meter pling run. If the leakage rate is found to be reading. Record the dry gas meter readings no greater than 0.00057 m3/min (0.02 cfm) or 4 at the beginning and end of each sampling percent of the average sampling rate (which- time increment, when changes in flow rates ever is less), the results are acceptable, and are made, before and after each leak check, no correction need be applied to the total and when sampling is halted. Take other volume of dry gas metered. If, however, a readings required by Figure 17–3 at least higher leakage rate is obtained, the tester shall either record the leakage rate and cor- once at each sample point during each time rect the sample volume as shown in Section increment and additional readings when sig- 6.3 of this method, or shall void the sampling nificant changes (20 percent variation in ve- run. locity head readings) necessitate additional 4.1.5 Particulate Train Operation. During adjustments in flow rate. Level and zero the the sampling run, maintain a sampling rate manometer. Because the manometer level such that sampling is within 10 percent of and zero may drift due to vibrations and true isokinetic, unless otherwise specified by temperature changes, make periodic checks the Administrator. during the traverse.
871
VerDate 11
872
VerDate 11
Clean the portholes prior to the test run to use of two or more trains will be subject to minimize the chance of sampling the depos- the approval of the Administrator. Note that ited material. To begin sampling, remove the when two or more trains are used, a separate nozzle cap and verify that the pitot tube and analysis of the collected particulate from probe extension are properly positioned. Po- each train shall be performed, unless iden- sition the nozzle at the first traverse point tical nozzle sizes were used on all trains, in with the tip pointing directly into the gas which case the particulate catches from the stream. Immediately start the pump and ad- individual trains may be combined and a sin- just the flow to isokinetic conditions. Nomo- gle analysis performed. graphs are available, which aid in the rapid At the end of the sample run, turn off the adjustment to the isokinetic sampling rate pump, remove the probe extension assembly without excessive computations. These from the stack, and record the final dry gas nomographs are designed for use when the meter reading. Perform a leak-check, as out- Type S pitot tube coefficient is 0.85±0.02, and lined in Section 4.1.4.3. Also, leak-check the the stack gas equivalent density (dry molec- pitot lines as described in Section 3.1 of ular weight) is equal to 29±4. APTD–0576 de- Method 2; the lines must pass this leak- tails the procedure for using the check, in order to validate the velocity head nomographs. If Cp and Md are outside the data. above stated ranges, do not use the 4.1.6 Calculation of Percent Isokinetic. nomographs unless appropriate steps (see Ci- Calculate percent isokinetic (see Section tation 7 in Bibliography) are taken to com- 6.11) to determine whether another test run pensate for the deviations. should be made. If there is difficulty in When the stack is under significant nega- maintaining isokinetic rates due to source tive pressure (height of impinger stem), take conditions, consult with the Administrator care to close the coarse adjust valve before for possible variance on the isokinetic rates. inserting the probe extension assembly into 4.2 Sample Recovery. Proper cleanup pro- the stack to prevent water from being forced cedure begins as soon as the probe extension backward. If necessary, the pump may be assembly is removed from the stack at the turned on with the coarse adjust valve end of the sampling period. Allow the assem- closed. When the probe is in position, block off the bly to cool. openings around the probe and porthole to When the assembly can be safely handled, prevent unrepresentative dilution of the gas wipe off all external particulate matter near stream. the tip of the probe nozzle and place a cap Traverse the stack cross section, as re- over it to prevent losing or gaining particu- quired by Method 1 or as specified by the Ad- late matter. Do not cap off the probe tip ministrator, being careful not to bump the tightly while the sampling train is cooling probe nozzle into the stack walls when sam- down as this would create a vacuum in the pling near the walls or when removing or in- filter holder, forcing condenser water back- serting the probe extension through the ward. portholes, to minimize chance of extracting Before moving the sample train to the deposited material. cleanup site, disconnect the filter holder- During the test run, take appropriate steps probe nozzle assembly from the probe exten- (e.g., adding crushed ice to the impinger ice sion; cap the open inlet of the probe exten- bath) to maintain a temperature of less than sion. Be careful not to lose any condensate, 20°C (68°F) at the condenser outlet; this will if present. Remove the umbilical cord from prevent excessive moisture losses. Also, peri- the condenser outlet and cap the outlet. If a odically check the level and zero of the ma- flexible line is used between the first im- nometer. pinger (or condenser) and the probe exten- If the pressure drop across the filter be- sion, disconnect the line at the probe exten- comes too high, making isokinetic sampling sion and let any condensed water or liquid difficult to maintain, the filter may be re- drain into the impingers or condenser. Dis- placed in the midst of a sample run. It is rec- connect the probe extension from the con- ommended that another complete filter hold- denser; cap the probe extension outlet. After er assembly be used rather than attempting wiping off the silicone grease, cap off the to change the filter itself. Before a new filter condenser inlet. Ground glass stoppers, plas- holder is installed, conduct a leak check, as tic caps, or serum caps (whichever are appro- outlined in Section 4.1.4.2. The total particu- priate) may be used to close these openings. late weight shall include the summation of Transfer both the filter holder-probe nozzle all filter assembly catches. assembly and the condenser to the cleanup A single train shall be used for the entire area. This area should be clean and protected sample run, except in cases where simulta- from the wind so that the chances of con- neous sampling is required in two or more taminating or losing the sample will be separate ducts or at two or more different lo- minimized. cations within the same duct, or, in cases Save a portion of the acetone used for where equipment failure necessitates a cleanup as a blank. Take 200 ml of this ace- change of trains. In all other situations, the tone directly from the wash bottle being
873
VerDate 11
used and place it in a glass sample container termine if it has been completely spent; labeled ‘‘acetone blank.’’ make a notation of its condition. Transfer Inspect the train prior to and during dis- the silica gel back to its original container assembly and note any abnormal conditions. and seal. A funnel may make it easier to Treat the samples as follows: pour the silica gel without spilling, and a Container No. 1. Carefully remove the filter rubber policeman may be used as an aid in from the filter holder and place it in its iden- removing the silica gel. It is not necessary to tified petri dish container. Use a pair of remove the small amount of dust particles tweezers and/or clean disposable surgical that may adhere to the walls and are dif- gloves to handle the filter. If it is necessary ficult to remove. Since the gain in weight is to fold the filter, do so such that the particu- to be used for moisture calculations, do not late cake is inside the fold. Carefully trans- use any water or other liquids to transfer the fer to the petri dish any particulate matter silica gel. If a balance is available in the and/or filter fibers which adhere to the filter field, follow the procedure for Container No. holder gasket, by using a dry Nylon bristle 3 under ‘‘Analysis.’’ brush and/or a sharp-edged blade. Seal the Condenser Water. Treat the condenser or container. impinger water as follows: make a notation Container No. 2. Taking care to see that of any color or film in the liquid catch. dust on the outside of the probe nozzle or Measure the liquid volume to within ±1 ml other exterior surfaces does not get into the by using a graduated cylinder or, if a balance sample, quantitatively recover particulate is available, determine the liquid weight to matter or any condensate from the probe within ±0.5 g. Record the total volume or nozzle, fitting, and front half of the filter weight of liquid present. This information is holder by washing these components with ac- required to calculate the moisture content of etone and placing the wash in a glass con- the effluent gas. Discard the liquid after tainer. Distilled water may be used instead measuring and recording the volume or of acetone when approved by the Adminis- weight. trator and shall be used when specified by 4.3 Analysis. Record the data required on the Administrator; in these cases, save a the example sheet shown in Figure 17–4. Han- water blank and follow Administrator’s di- dle each sample container as follows: rections on analysis. Perform the acetone Container No. 1. Leave the contents in the rinses as follows: shipping container or transfer the filter and Carefully remove the probe nozzle and any loose particulate from the sample con- clean the inside surface by rinsing with ace- tainer to a tared glass weighing dish. Des- tone from a wash bottle and brushing with a iccate for 24 hours in a desiccator containing Nylon bristle brush. Brush until acetone anhydrous calcium sulfate. Weigh to a con- rinse shows no visible particles, after which stant weight and report the results to the make a final rinse of the inside surface with nearest 0.1 mg. For purposes of this Section, acetone. 4.3, the term ‘‘constant weight’’ means a dif- Brush and rinse with acetone the inside ference of no more than 0.5 mg or 1 percent parts of the fitting in a similar way until no of total weight less tare weight, whichever is visible particles remain. A funnel (glass or greater, between two consecutive weighings, polyethylene) may be used to aid in transfer- with no less than 6 hours of desiccation time ring liquid washes to the container. Rinse between weighings. the brush with acetone and quantitatively Alternatively, the sample may be oven collect these washings in the sample con- dried at the average stack temperature or tainer. Between sampling runs, keep brushes 105°C (220°F), whichever is less, for 2 to 3 clean and protected from contamination. hours, cooled in the desiccator, and weighed After ensuring that all joints are wiped to a constant weight, unless otherwise speci- clean of silicone grease (if applicable), clean fied by the Administrator. The tester may the inside of the front half of the filter hold- also opt to oven dry the sample at the aver- er by rubbing the surfaces with a Nylon bris- age stack temperature or 105°C (220°F), tle brush and rinsing with acetone. Rinse whichever is less, for 2 to 3 hours, weigh the each surface three times or more if needed to sample, and use this weight as a final remove visible particulate. Make final rinse weight. of the brush and filter holder. After all ace- tone washings and particulate matter are FIGURE 17–4—ANALYTICAL DATA collected in the sample container, tighten Plant ———————————————————— the lid on the sample container so that ace- Date ————————————————————— tone will not leak out when it is shipped to Run No. ——————————————————— the laboratory. Mark the height of the fluid Filter No. —————————————————— level to determine whether or not leakage Amount liquid lost during transport ———— occurred during transport. Label the con- Acetone blank volume, ml ————————— tainer to clearly identify its contents. Acetone wash volume, ml —————————— Container No. 3. If silica gel is used in the Acetone blank concentration, mg/mg (Equa- condenser system for mositure content de- tion 17–4) ————————————————— termination, note the color of the gel to de- Acetone wash blank, mg (Equation 17–5) ——
874
VerDate 11
Weight of particulate collected, mg tone is highly flammable and has a low flash Container point. number Final weight Tare weight Weight gain 5. Calibration 1. Maintain a laboratory log of all calibra- tions. 2. 5.1 Probe Nozzle. Probe nozzles shall be Total. calibrated before their initial use in the field. Using a micrometer, measure the in- Less acetone blank. side diameter of the nozzle to the nearest 0.025 mm (0.001 in.). Make three separate Weight of particulate matter. measurements using different diameters each time, and obtain the average of the Volume of liquid water col- measurements. The difference between the lected high and low numbers shall not exceed 0.1 Impinger vol- Silica gel mm (0.004 in.). When nozzles become nicked, ume, ml weight, g dented, or corroded, they shall be reshaped, sharpened, and recalibrated before use. Each Final. nozzle shall be permanently and uniquely Initial. identified. Liquid collected. 5.2 Pitot Tube. If the pitot tube is placed Total volume collected ...... g* ml in an interference-free arrangement with re- *Convert weight of water to volume by dividing total weight spect to the other probe assembly compo- increase by density of water (1 g/ml). nents, its baseline (isolated tube) coefficient Increase, g shall be determined as outlined in Section 4 ÐÐÐÐÐ = Volume water, ml of Method 2. If the probe assembly is not in- (1 g/ml) terference-free, the pitot tube assembly coef- Container No. 2. Note the level of liquid in ficient shall be determined by calibration, the container and confirm on the analysis using methods subject to the approval of the sheet whether or not leakage occurred dur- Administrator. ing transport. If a noticeable amount of 5.3 Metering System. Before its initial use leakage has occurred, either void the sample in the field, the metering system shall be or use methods, subject to the approval of calibrated according to the procedure out- the Administrator, to correct the final re- lined in APTD–0576. Instead of physically ad- sults. Measure the liquid in this container ei- justing the dry gas meter dial readings to ther volumetrically to ±1 ml or gravimetri- correspond to the wet test meter readings, cally to ±0.5 g. Transfer the contents to a calibration factors may be used to mathe- tared 250-ml beaker and evaporate to dryness matically correct the gas meter dial read- at ambient temperature and pressure. Des- ings to the proper values. iccate for 24 hours and weigh to a constant Before calibrating the metering system, it weight. Report the results to the nearest 0.1 is suggested that a leak-check be conducted. mg. For metering systems having diaphragm Container No. 3. This step may be con- pumps, the normal leak-check procedure will ducted in the field. Weigh the spent silica gel not detect leakages within the pump. For (or silica gel plus impinger) to the nearest 0.5 these cases the following leak-check proce- g using a balance. dure is suggested: make a 10-minute calibra- ‘‘Acetone Blank’’ Container. Measure ace- tion run at 0.00057 m 3/min (0.02 cfm); at the tone in this container either volumetrically end of the run, take the difference of the or gravimetrically. Transfer the acetone to a measured wet test meter and dry gas meter tared 250-ml beaker and evaporate to dryness volumes; divide the difference by 10, to get at ambient temperature and pressure. Des- the leak rate. The leak rate should not ex- iccate for 24 hours and weigh to a constant ceed 0.00057 m 3/min (0.02 cfm). weight. Report the results to the nearest 0.1 After each field use, the calibration of the mg. metering system shall be checked by per- NOTE: At the option of the tester, the con- forming three calibration runs at a single, tents of Container No. 2 as well as the ace- intermediate orifice setting (based on the tone blank container may be evaporated at previous field test), with the vacuum set at temperatures higher than ambient. If evapo- the maximum value reached during the test ration is done at an elevated temperature, series. To adjust the vacuum, insert a valve the temperature must be below the boiling between the wet test meter and the inlet of point of the solvent; also, to prevent ‘‘bump- the metering system. Calculate the average ing,’’ the evaporation process must be close- value of the calibration factor. If the calibra- ly supervised, and the contents of the beaker tion has changed by more than 5 percent, re- must be swirled occasionally to maintain an calibrate the meter over the full range of even temperature. Use extreme care, as ace- orifice settings, as outlined in APTD–0576.
875
VerDate 11
Alternative procedures, e.g., using the ori- sampling train from the pump to the orifice fice meter coefficients, may be used, subject meter should be leak checked prior to initial to the approval of the Administrator. use and after each shipment. Leakage after NOTE: If the dry gas meter coefficient val- the pump will result in less volume being re- ues obtained before and after a test series corded than is actually sampled. The follow- differ by more than 5 percent, the test series ing procedure is suggested (see Figure 17–5). shall either be voided, or calculations for the Close the main valve on the meter box. In- test series shall be performed using which- sert a one-hole rubber stopper with rubber ever meter coefficient value (i.e., before or tubing attached into the orifice exhaust after) gives the lower value of total sample pipe. Disconnect and vent the low side of the volume. orifice manometer. Close off the low side ori- 5.4 Temperature Gauges. Use the proce- fice tap. Pressurize the system to 13 to 18 cm dure in Section 4.3 of Method 2 to calibrate (5 to 7 in.) water column by blowing into the in-stack temperature gauges. Dial thermom- eters, such as are used for the dry gas meter rubber tubing. Pinch off the tubing and ob- and condenser outlet, shall be calibrated serve the manometer for one minute. A loss against mercury-in-glass thermometers. of pressure on the manometer indicates a 5.5 Leak Check of Metering System leak in the meter box; leaks, if present, must Shown in Figure 17–1. That portion of the be corrected.
876
VerDate 11
5.6 Barometer. Calibrate against a mer- Carry out calculations, retaining at least cury barometer. one extra decimal figure beyond that of the 6. Calculations acquired data. Round off figures after the
877
VerDate 11
final calculation. Other forms of the equa- ρa=Density of acetone, mg/ml (see label on tions may be used as long as they give equiv- bottle). alent results. ρw=Density of water, 0.9982 g/ml (0.002201 lb/ 6.1 Nomenclature. ml). θ=Total sampling time, min. A =Cross-sectional area of nozzle, m 2(ft2). n θ =Sampling time interval, from the begin- B =Water vapor in the gas stream, propor- 1 ws ning of a run until the first component tion by volume. change, min. C =Acetone blank residue concentration, mg/ a θ =Sampling time interval, between two suc- mg. i cessive component changes, beginning c =Concentration of particulate matter in s with the interval between the first and stack gas, dry basis, corrected to stand- second changes, min. ard conditions, g/dscm (g/dscf). θ =Sampling time interval, from the final I=Percent of isokinetic sampling. p (nth) component change until the end of L =Maximum acceptable leakage rate for ei- a the sampling run, min. ther a pretest leak check or for a leak 13.6=Specific gravity of mercury. check following a component change; 60=Sec/min. 3 equal to 0.00057 m /min (0.02 cfm) or 4 100=Conversion to percent. percent of the average sampling rate, 6.2 Average Dry Gas Meter Temperature whichever is less. and Average Orifice Pressure Drop. See data Li=Individual leakage rate observed during sheet (Figure 17–3). the leak check conducted prior to the 6.3 Dry Gas Volume. Correct the sample th ‘‘i ’’ component change (i=1, 2, 3 . . . n), volume measured by the dry gas meter to 3 m /min (cfm). standard conditions (20°C, 760 mm Hg or 68°F, Lp=Leakage rate observed during the post- 29.92 in. Hg) by using Equation 17–1. test leak check, m 3/min (cfm). ma=Mass of residue of acetone after evapo- ration, mg. mn=Total amount of particulate matter col- lected, mg. Mw=Molecular weight of water, 18.0 g/g-mole (18.0 lb/lb-mole). Pbar=Barometric pressure at the sampling site, mm Hg (in. Hg). Ps=Absolute stack gas pressure, mm Hg (in. Hg). Pstd=Standard absolute pressure, 760 mm Hg Eq. 17–1 (29.92 in. Hg). Where: R=Ideal gas constant, 0.06236 mm Hg-m3/°K-g- K =0.3858°K/mm Hg for metric units; 17.64°R/ mole (21.85 in. Hg-ft 3/°R-lb-mole). 1 in. Hg for English units. Tm=Absolute average dry gas meter tem- perature (see Figure 17–3), °K (°R). NOTE: Equation 17–1 can be used as written unless the leakage rate observed during any Ts=Absolute average stack gas temperature (see Figure 17–3), °K (°R). of the mandatory leak checks (i.e., the post- test leak check or leak checks conducted Tstd=Standard absolute temperature, 293°K (528°R). prior to component changes) exceeds La. If Lp or Li exceeds La, Equation 17–1 must be Va=Volume of acetone blank, ml. modified as follows: Vaw=Volume of acetone used in wash, ml. Vlc=Total volume of liquid collected in (a) Case I. No component changes made impingers and silica gel (see Figure 17–4), during sampling run. In this case, replace Vm ml. in Equation 17–1 with the expression:
Vm=Volume of gas sample as measured by [Vm¥(Lp¥La)θ] dry gas meter, dcm (dcf). (b) Case II. One or more component Vm(std)=Volume of gas sample measured by changes made during the sampling run. In the dry gas meter, corrected to standard this case, replace Vm in Equation 17–1 by the conditions, dscm (dscf). expression: Vw(std)=Volume of water vapor in the gas sample, corrected to standard conditions, scm (scf). vs=Stack gas velocity, calculated by Method 2, Equation 2–9, using data obtained from Method 17, m/sec (ft/sec). Wa=Weight of residue in acetone wash, mg. Y=Dry gas meter calibration coefficient. ∆ H=Average pressure differential across the and substitute only for those leakage rates orifice meter (see Figure 17–3), mm H2O (Li or Lp) which exceed La. (in. H2O). 6.4 Volume of Water Vapor.
878
VerDate 11
Where:
K4=4.320 for metric units; 0.09450 for English units. 6.12 Acceptable Results. If 90 percent < I <110 percent, the results are acceptable. If Eq. 17–2 the results are low in comparison to the Where: standard and I is beyond the acceptable range, or, if I is less than 90 percent, the Ad- 3 K2=0.001333 m /ml for metric units; 0.04707 ministrator may opt to accept the results. ft 3/ml for English units. Use Citation 4 in Bibliography to make judg- 6.5 Moisture Content. ments. Otherwise, reject the results and re- peat the test. V B = w() std Eq. 17- 3 7. Bibliography ws VV+ 1. Addendum to Specifications for Inciner- m()() std w std ator Testing at Federal Facilities. PHS, 6.6 Acetone Blank Concentration. NCAPC. December 6, 1967. 2. Martin, Robert M., Construction Details M of Isokinetic Source-Sampling Equipment. C = a Eq. 17- 4 a ρ Environmental Protection Agency. Research Va a Triangle Park, NC, APTD–0581. April, 1971. 6.7 Acetone Wash Blank. 3. Rom, Jerome J., Maintenance, Calibra- tion, and Operation of Isokinetic Source- Wa=Ca Vawρa Eq. 17–5 Sampling Equipment. Environmental Pro- 6.8 Total Particulate Weight. Determine tection Agency. Research Triangle Park, NC the total particulate catch from the sum of APTD–0576. March, 1972. the weights obtained from Containers 1 and 4. Smith, W. S., R. T. Shigehara, and W. F. 2 less the acetone blank (see Figure 17–4). Todd. A Method of Interpreting Stack Sam- pling Data. Paper Presented at the 63rd An- NOTE: Refer to Section 4.1.5 to assist in cal- nual Meeting of the Air Pollution Control culation of results involving two or more fil- Association, St. Louis, MO June 14–19, 1970. ter assemblies or two or more sampling 5. Smith, W. S., et al., Stack Gas Sampling trains. Improved and Simplified with New Equip- 6.9 Particulate Concentration. ment. APCA Paper No. 67–119. 1967. cs=(0.001 g/mg) (mn/Vm(std)) 6. Specifications for Incinerator Testing at Eq. 17–6 Federal Facilities. PHS, NCAPC. 1967. 6.10 Conversion Factors: 7. Shigehara, R. T., Adjustments in the From To Multiply by EPA Nomograph for Different Pitot Tube Co- efficients and Dry Molecular Weights. Stack scf ...... m 3 ...... 0.02832 Sampling News 2:4–11. October, 1974. g/ft 3 ...... gr/ft 3 ...... 15.43 8. Vollaro, R. F., A Survey of Commer- g/ft 3 ...... lb/ft 3 ...... 2.205×10¥3 cially Available Instrumentation for the 3 3 g/ft ...... g/m ...... 35.31 Measurement of Low-Range Gas Velocities. U.S. Environmental Protection Agency, 6.11 Isokinetic Variation. Emission Measurement Branch. Research 6.11.1 Calculation from Raw Data. Triangle Park, NC, November, 1976 (unpub- lished paper). 9. Annual Book of ASTM Standards. Part 26. Gaseous Fuels; Coal and Coke; Atmos- pheric Analysis. American Society for Test- Eq. 17–7 ing and Materials. Philadelphia, PA 1974. pp. Where: 617–622. 3 K3=0.003454 mm Hg–m /ml–°K for metric 10. Vollaro, R. F., Recommended Procedure units; 0.002669 in. Hg-ft 3/ml–°R for for Sample Traverses in Ducts Smaller than English units. 12 Inches in Diameter. U.S. Environmental 6.11.2 Calculation from Intermediate Val- Protection Agency, Emission Measurement ues. Branch. Research Triangle Park, NC, Novem- ber, 1976.
METHOD 18—MEASUREMENT OF GASEOUS OR- GANIC COMPOUND EMISSIONS BY GAS CHRO- MATOGRAPHY Introduction This method should not be attempted by persons unfamiliar with the performance characteristics of gas chromatography, nor Eq. 17–8 by those persons who are unfamiliar with
879
VerDate 11
source sampling. Particular care should be standard deviation (RSD), but an experi- exercised in the area of safety concerning enced GC operator with a reliable instru- choice of equipment and operation in poten- ment can readily achieve 5 percent RSD. For tially explosive atmospheres. this method, the following combined GC/op- 1. Applicability and Principle erator values are required. (a) Precision. Duplicate analyses are with- 1.1 Applicability. This method applies to in 5 percent of their mean value. the analysis of approximately 90 percent of (b) Accuracy. Analysis results of prepared the total gaseous organics emitted from an audit samples are within 10 percent of prepa- industrial source. It does not include tech- ration values. niques to identify and measure trace (c) Recovery. After developing an appro- amounts of organic compounds, such as priate sampling and analytical system for those found in building air and fugitive emis- the pollutants of interest, conduct the proce- sion sources. dure in Section 7.6. Conduct the appropriate This method will not determine compounds recovery study in Section 7.6 at each sam- that (1) are polymeric (high molecular pling point where the method is being ap- weight), (2) can polymerize before analysis, plied. Submit the data and results of the re- or (3) have very low vapor pressures at stack covery procedure with the reporting of re- or instrument conditions. sults under Section 7.5. 1.2 Principle. 4. Interferences The major organic components of a gas mixture are separated by gas chroma- Resolution interferences that may occur tography (GC) and individually quantified by can be eliminated by appropriate GC column flame ionization, photoionization, electron and detector choice or by shifting the reten- capture, or other appropriate detection prin- tion times through changes in the column ciples. flow rate and the use of temperature pro- The retention times of each separated com- gramming. ponent are compared with those of known The analytical system is demonstrated to compounds under identical conditions. be essentially free from contaminants by pe- Therefore, the analyst confirms the identity riodically analyzing blanks that consist of and approximate concentrations of the or- hydrocarbon-free air or nitrogen. Sample cross-contamination that occurs ganic emission components beforehand. With when high-level and low-level samples or this information, the analyst then prepares standards are analyzed alternately, is best or purchases commercially available stand- dealt with by thorough purging of the GC ard mixtures to calibrate the GC under con- sample loop between samples. ditions identical to those of the samples. The To assure consistent detector response, analyst also determines the need for sample calibration gases are contained in dry air. To dilution to avoid detector saturation, gas adjust gaseous organic concentrations when stream filtration to eliminate particulate water vapor is present in the sample, water matter, and prevention of moisture con- vapor concentrations are determined for densation. those samples, and a correction factor is ap- 2. Range and Sensitivity plied. 2.1 Range. The lower range of this method 5. Presurvey and Presurvey Sampling. is determined by the sampling system; ad- Perform a presurvey for each source to be sorbents may be used to concentrate the tested. Refer to Figure 18–1. Some of the in- sample, thus lowering the limit of detection formation can be collected from literature below the 1 part per million (ppm) typically surveys and source personnel. Collect gas achievable with direct interface or bag sam- samples that can be analyzed to confirm the pling. The upper limit is governed by GC de- identities and approximate concentrations of tector saturation or column overloading; the the organic emissions. upper range can be extended by dilution of 5.1 Apparatus. This apparatus list also ap- sample with an inert gas or by using smaller plies to Sections 6 and 7. volume gas sampling loops. The upper limit 5.1.1 Teflon Tubing. (Mention of trade can also be governed by condensation of names or specific products does not con- higher boiling compounds. stitute endorsement by the U.S. Environ- 2.2 Sensitivity. The sensitivity limit for a mental Protection Agency.) Diameter and compound is defined as the minimum detect- length determined by connection require- able concentration of that compound, or the ments of cylinder regulators and the GC. Ad- concentration that produces a signal-to- ditional tubing is necessary to connect the noise ratio of three to one. The minimum de- GC sample loop to the sample. tectable concentration is determined during 5.1.2 Gas Chromatograph. GC with suit- the presurvey calibration for each able detector, columns, temperature-con- compound. trolled sample loop and valve assembly, and 3. Precision and Accuracy temperature programable oven, if necessary. Gas chromatographic techniques typically The GC shall achieve sensitivity require- provide a precision of 5 to 10 percent relative ments for the compounds under study.
880
VerDate 11
5.1.3 Pump. Capable of pumping 100 ml/ 5.2.8 Zero Gas. Hydrocarbon free air or ni- min. For flushing sample loop. trogen, to be used for dilutions, blank prepa- 5.1.4 Flowmeters. To measure flow rates. ration, and standard preparation. 5.1.5 Regulators. Used on gas cylinders for 5.3 Sampling. GC and for cylinder standards. 5.3.1 Collection of Samples with Glass 5.1.6 Recorder. Recorder with linear strip Sampling Flasks. Presurvey samples can be chart is minimum acceptable. Integrator collected in precleaned 250-ml double-ended (optional) is recommended. glass sampling flasks. Teflon stopcocks, 5.1.7 Syringes. 0.5-ml, 1.0- and 10- without grease, are preferred. Flasks should microliter sizes, calibrated, maximum accu- be cleaned as follows: Remove the stopcocks racy (gas tight), for preparing calibration from both ends of the flasks, and wipe the standards. Other appropriate sizes can be parts to remove any grease. Clean the stop- used. cocks, barrels, and receivers with methylene 5.1.8 Tubing Fittings. To plumb GC and dichloride. Clean all glass ports with a soap gas cylinders. solution, then rinse with tap and deionized 5.1.9 Septums. For syringe injections. distilled water. Place the flask in a cool 5.1.10 Glass Jars. If necessary, clean-col- glass annealing furnace and apply heat up to ored glass jars with Teflon-lined lids for con- 500°C. Maintain at this temperature for 1 densate sample collection. Size depends on hour. After this time period, shut off and volume of condensate. open the furnace to allow the flask to cool. 5.1.11 Soap Film Flow Meter. To deter- Grease the stopcocks with stopcock grease mine flow rates. and return them to the flask receivers. 5.1.12 Tedlar Bags. 10- and 50-liter capac- Purge the assembly with high-purity nitro- ity, for preparation of standards. gen for 2 to 5 minutes. Close off the stop- 5.1.13 Dry Gas Meter with Temperature cocks after purging to maintain a slight and Pressure Gauges. Accurate to ±2 percent, positive nitrogen pressure. Secure the stop- for perparation of gas standards. cocks with tape. 5.1.14 Midget Impinger/Hot Plate Assem- Presurvey samples can be obtained either bly. For preparation of gas standards. by drawing the gases into the previously 5.1.15 Sample Flasks. For presurvey sam- evacuated flask or by drawing the gases into ples, must have gas-tight seals. and purging the flask with a rubber suction 5.1.16 Adsorption Tubes. If necessary, bulb. blank tubes filled with necessary adsorbent 5.3.1.1 Evacuated Flask Procedure. Use a (charcoal, Tenax, XAD–2, etc.) for presurvey high-vacuum pump to evacuate the flask to samples. the capacity of the pump; then close off the 5.1.17 Personnel Sampling Pump. Cali- stopcock leading to the pump. Attach a 6- brated, for collecting adsorbent tube mm outside diameter (OD) glass tee to the presurvey samples. flask inlet with a short piece of Teflon tub- 5.1.18 Dilution System. Calibrated, the di- ing. Select a 6-mm OD borosilicate sampling lution system is to be constructed following probe, enlarged at one end to a 12-mm OD the specifications of an acceptable method. and of sufficient length to reach the centroid 5.1.19 Sample Probes. Pyrex or stainless of the duct to be sampled. Insert a glass wool steel, of sufficient length to reach centroid plug in the enlarged end of the probe to re- of stack, or a point no closer to the walls move particulate matter. Attach the other than 1 m. end of the probe to the tee with a short piece 5.1.20 Barometer. To measure barometric of Teflon tubing. Connect a rubber suction pressure. bulb to the third leg of the tee. Place the fil- 5.2 Reagents. ter end of the probe at the centroid of the 5.2.1 Deionized Distilled Water. duct, or at a point no closer to the walls 5.2.2 Methylene Dichloride. than 1 m, and purge the probe with the rub- 5.2.3 Calibration Gases. A series of stand- ber suction bulb. After the probe is com- ards prepared for every compound of inter- pletely purged and filled with duct gases, est. open the stopcock to the grab flask until the 5.2.4 Organic Compound Solutions. Pure pressure in the flask reaches duct pressure. (99.9 percent), or as pure as can reasonably Close off the stopcock, and remove the probe be obtained, liquid samples of all the organic from the duct. Remove the tee from the flask compounds needed to prepare calibration and tape the stopcocks to prevent leaks dur- standards. ing shipment. Measure and record the duct 5.2.5 Extraction Solvents. For extraction temperature and pressure. of adsorbent tube samples in preparation for 5.3.1.2 Purged Flask Procedure. Attach analysis. one end of the sampling flask to a rubber 5.2.6 Fuel. As recommended by the manu- suction bulb. Attach the other end to a 6-mm facturer for operation of the GC. OD glass probe as described in Section 5.3.1.1. 5.2.7 Carrier Gas. Hydrocarbon free, as Place the filter end of the probe at the cen- recommended by the manufacturer for oper- troid of the duct, or at a point no closer to ation of the detector and compatability with the walls than 1 m, and apply suction with the column. the bulb to completely purge the probe and
881
VerDate 11
flask. After the flask has been purged, close 6.1.2 Preliminary GC Adjustment. Using off the stopcock near the suction bulb, and the standards and column obtained in Sec- then close the stopcock near the probe. Re- tion 6.1.1, perform initial tests to determine move the probe from the duct, and dis- appropriate GC conditions that provide good connect both the probe and suction bulb. resolution and minimum analysis time for Tape the stopcocks to prevent leakage dur- the compounds of interest. ing shipment. Measure and record the duct 6.1.3 Preparation of Presurvey Samples. If temperature and pressure. the samples were collected on an adsorbent, 5.3.2 Flexible Bag Procedure. Tedlar or extract the sample as recommended by the aluminized Mylar bags can also be used to manufacturer for removal of the compounds obtain the presurvey sample. Use new bags, with a solvent suitable to the type of GC and leak check them before field use. In ad- analysis. Prepare other samples in an appro- dition, check the bag before use for contami- priate manner. nation by filling it with nitrogen or air, and 6.1.4 Presurvey Sample Analysis. Before analyzing the gas by GC at high sensitivity. analysis, heat the presurvey sample to the Experience indicates that it is desirable to duct temperature to vaporize any condensed allow the inert gas to remain in the bag material. Analyze the samples by the GC about 24 hours or longer to check for procedure, and compare the retention times desorption of organics from the bag. Follow against those of the calibration samples that the leak check and sample collection proce- contain the components expected to be in dures given in Section 7.1. the stream. If any compounds cannot be 5.3.3 Determination of Moisture Content. identified with certainty by this procedure, For combustion or water-controlled proc- identify them by other means such as GC/ esses, obtain the moisture content from mass spectroscopy (GC/MS) or GC/infrared plant personnel or by measurement during techniques. A GC/MS system is rec- the presurvey. If the source is below 59°C, ommended. measure the wet bulb and dry bulb tempera- Use the GC conditions determined by the tures, and calculate the moisture content procedures of Section 6.1.2 for the first injec- using a psychrometric chart. At higher tem- tion. Vary the GC parameters during subse- peratures, use Method 4 to determine the quent injections to determine the optimum moisture content. settings. Once the optimum settings have 5.4 Determination of Static Pressure. Ob- been determined, perform repeat injections tain the static pressure from the plant per- of the sample to determine the retention sonnel or measurement. If a type S pitot time of each compound. To inject a sample, tube and an inclined manometer are used, draw sample through the loop at a constant take care to align the pitot tube 90° from the rate (100 ml/min for 30 seconds). Be careful direction of the flow. Disconnect one of the not to pressurize the gas in the loop. Turn off tubes to the manometer, and read the static the pump and allow the gas in the sample pressure; note whether the reading is posi- loop to come to ambient pressure. Activate tive or negative. the sample valve, and record injection time, 5.5 Collection of Presurvey Samples with loop temperature, column temperature, car- Adsorption Tube. Follow Section 7.4 for rier flow rate, chart speed, and attenuator presurvey sampling. setting. Calculate the retention time of each peak using the distance from injection to the 6. Analysis Development peak maximum divided by the chart speed. 6.1 Selection of GC Parameters. Retention times should be repeatable within 6.1.1 Column Choice. Based on the initial 0.5 seconds. contact with plant personnel concerning the If the concentrations are too high for ap- plant process and the anticipated emissions, propriate detector response, a smaller sam- choose a column that provides good resolu- ple loop or dilutions may be used for gas tion and rapid analysis time. The choice of samples, and, for liquid samples, dilution an appropriate column can be aided by a lit- with solvent is appropriate. Use the standard erature search, contact with manufacturers curves (Section 6.3) to obtain an estimate of of GC columns, and discussion with person- the concentrations. nel at the emission source. Identify all peaks by comparing the known Most column manufacturers keep excellent retention times of compounds expected to be records of their products. Their technical in the retention times of peaks in the sam- service departments may be able to rec- ple. Identify any remaining unidentified ommend appropriate columns and detector peaks which have areas larger than 5 percent type for separating the anticipated com- of the total using a GC/MS, or estimation of pounds, and they may be able to provide in- possible compounds by their retention times formation on interferences, optimum operat- compared to known compounds, with con- ing conditions, and column limitations. firmation by further GC analysis. Plants with analytical laboratories may 6.2 Calibration Standards. Prepare or ob- also be able to provide information on appro- tain enough calibration standards so that priate analytical procedures. there are three different concentrations of
882
VerDate 11
each organic compound expected to be meas- Once the flowmeters are calibrated, con- ured in the source sample. For each organic nect the flowmeters to the calibration and compound, select those concentrations that diluent gas supplies using 6-mm Teflon tub- bracket the concentrations expected in the ing. Connect the outlet side of the flow- source samples. A calibration standard may meters through a connector to a leak-free contain more than one organic compound. If Tedlar bag as shown in Figure 18–5. (See Sec- available, commercial cylinder gases may be tion 7.1 for bag leak-check procedures.) Ad- used if their concentrations have been cer- just the gas flow to provide the desired dilu- tified by direct analysis. tion, and fill the bag with sufficient gas for If samples are collected in adsorbent tubes GC calibration. Be careful not to overfill and (charcoal, XAD–2, Tenax, etc.), prepare or cause the bag to apply additional pressure on obtain standards in the same solvent used the dilution system. Record the flow rates of for the sample extraction procedure. Refer to both flowmeters, and the laboratory tem- Section 7.4.3. perature and atmospheric pressure. Cal- Verify the stability of all standards for the culate the concentration Cs in ppm of each time periods they are used. If gas standards organic in the diluted gas as follows: are prepared in the laboratory, use one or more of the following procedures. 6.2.1 Preparation of Standards from High Concentration Cylinder Standards. Obtain enough high concentration cylinder stand- ards to represent all the organic compounds expected in the source samples. Use these high concentration standards to prepare lower concentration standards by di- where: lution, as shown by Figures 18–5 and 18–6. 106=Conversion to ppm. To prepare the diluted calibration samples, X=Mole or volume fraction of the organic in calibrated rotameters are normally used to the calibration gas to be diluted. meter both the high concentration calibra- qc=Flow rate of the calibration gas to be di- tion gas and the diluent gas. Other types of luted. flowmeters and commercially available dilu- qd=Diluent gas flow rate. tion systems can also be used. Single-stage dilutions should be used to pre- Calibrate each flowmeter before use by pare calibration mixtures up to about 1:20 di- placing it between the diluent gas supply and lution factor. suitably sized bubble meter, spirometer, or For greater dilutions, a double dilution wet test meter. Record all data shown on system is recommended, as shown in Figure Figure 18–4. While it is desirable to calibrate 18–6. Fill the Tedlar bag with the dilute gas the cylinder gas flowmeter with cylinder gas, from the second stage. Record the laboratory the available quantity and cost may preclude temperature, barometric pressure, and static it. The error introduced by using the diluent pressure readings. Correct the flow reading gas for calibration is insignificant for gas for temperature and pressure. Calculate the mixtures of up to 1,000 to 2,000 ppm of each concentration Cs in ppm of the organic in the organic component. final gas mixture as follows:
Where: Further details of the calibration methods 106=Conversion to ppm. for flowmeters and the dilution system can X=Mole or volume fraction of the organic in be found in Citation 21 in the Bibliography. the calibration gas to be diluted. 6.2.2 Preparation of Standards from Vola- qc1=Flow rate of the calibration gas to be di- tile Materials. Record all data shown on Fig- luted in stage 1. ure 18–3. qc2=Flow rate of the calibration gas to be di- 6.2.2.1 Gas Injection Technique. This pro- luted in stage 2. cedure is applicable to organic compounds qd1=Flow rate of diluent gas in stage 1. that exist entirely as a gas at ambient condi- qd2=Flow rate of diluent gas in stage 2. tions. Evacuate a 10-liter Tedlar bag that has 883
VerDate 11
passed a leak-check (see Section 7.1), and pare dilutions having other concentrations. meter in 5.0 liters of air or nitrogen through Prepare a minimum of three concentrations. a dry gas meter that has been calibrated in Place each bag on a smooth surface, and al- a manner consistent with the procedure de- ternately depress opposite sides of the bag 50 scribed in Section 5.1.1 of Method 5. While times to mix the gases. Record the average the bag is filling use a 0.5-ml syringe to in- meter temperature and pressure, the gas vol- ject a known quantity of ‘‘pure’’ gas of the ume and the barometric pressure. Record the organic compound through the wall of the syringe temperature and pressure before in- bag, or through a septum-capped tee at the jection. bag inlet. Withdraw the syringe needle, and Calculate each organic standard con- immediately cover the resulting hole with a piece of masking tape. In a like manner, pre- centration Cs in ppm as follows:
where: ing, open the bag inlet valve, and open the cylinder. Adjust the rate so that the bag will Gv=Gas volume or organic compound in- jected, ml. be completely filled in approximately 15 minutes. Record meter pressure and tem- 106=Conversion to ppm. perature, and local barometric pressure. Ps=Absolute pressure of syringe before injec- tion, mm Hg. Allow the liquid organic to equilibrate to Ts=Absolute temperature of syringe before room temperature. Fill the 1.0- or 10- injection, °K. microliter syringe to the desired liquid vol- Vm=Gas volume indicated by dry gas meter, ume with the organic. Place the syringe nee- liters. dle into the impinger inlet using the septum Y=Dry gas meter calibration factor, provided, and inject the liquid into the flow- dimensionless. ing air stream. Use a needle of sufficient Pm=Absolute pressure of dry gas meter, mm length to permit injection of the liquid Hg. below the air inlet branch of the tee. Remove Tm=Absolute temperature of dry gas meter, the syringe. °K. When the bag is filled, stop the pump, and 1000=Conversion factor, ml/liter. close the bag inlet valve. Record the final 6.2.2.2 Liquid Injection Technique. Use meter reading, temperature, and pressure. the equipment shown in Figure 18–8. Cali- Disconnect the bag from the impinger out- brate the dry gas meter as described in Sec- let, and either set it aside for at least 1 hour, tion 6.2.2.1 with a wet test meter or a spirom- or massage the bag to insure complete mix- eter. Use a water manometer for the pressure ing. gauge and glass, Teflon, brass, or stainless Measure the solvent liquid density at room steel for all connections. Connect a valve to temperature by accurately weighing a the inlet of the 50-liter Tedlar bag. known volume of the material on an analyt- To prepare the standards, assemble the ical balance to the nearest 1.0 milligram. A equipment as shown in Figure 18–8, and leak- ground-glass stoppered 25-mil volumetric check the system. Completely evacuate the flask or a glass-stoppered specific gravity bag. Fill the bag with hydrocarbon-free air, bottle is suitable for weighing. Calculate the and evacuate the bag again. Close the inlet result in terms of g/ml. As an alternative, valve. literature values of the density of the liquid Turn on the hot plate, and allow the water at 20 °C may be used. to reach boiling, Connect the bag to the im- Calculate each organic standard con- pinger outlet. Record the initial meter read- centration Cs in ppm as follows:
884
VerDate 11
where: mental Protection Agency, Environmental Monitoring Systems Laboratory, Quality As- Lv=Liquid volume of organic injected, µl. µl=Liquid organic density as determined, g/ surance Division (MD–77), Research Triangle ml. Park, North Carolina 27711. Audit cylinders M=Molecular weight of organic, g/g-mole. obtained from a commercial gas manufac- 24.055=Ideal gas molar volume at 293 °K and turer may be used provided that (a) the gas 760 mm Hg, liters/g-mole. manufacturer certifies the audit cylinder in 106=Conversion to ppm. a manner similar to the procedure described 1000=Conversion factor, µl/ml. in 40 CFR Part 61, Appendix B, Method 106, 6.3 Preparation of Calibration Curves. Es- Section 5.2.3.1, and (b) the gas manufacturer tablish proper GC conditions, then flush the obtains an independent analysis of the audit sampling loop for 30 seconds at a rate of 100 cylinders to verify this analysis. Independent ml/min. Allow the sample loop pressure to analysis is defined as an analysis performed equilibrate to atmospheric pressure, and ac- by an individual other than the individual tivate the injection valve. Record the stand- who performs the gas manufacturer’s analy- ard concentration, attenuator factor, injec- sis, while using calibration standards and tion time, chart speed, retention time, peak analysis equipment different from those used area, sample loop temperature, column tem- for the gas manufacturer’s analysis. Ver- perature, and carrier gas flow rate. Repeat ification is complete and acceptable when the standard injection until two consecutive the independent analysis concentration is injections give area counts within 5 percent within 5 percent of the gas manufacturer’s of their average. The average value concentration. multipled by the attenuator factor is then 7. Final Sampling and Analysis Procedure the calibration area value for the concentra- Considering safety (flame hazards) and the tion. source conditions, select an appropriate sam- Repeat this procedure for each standard. pling and analysis procedure (Sections 7.1, Prepare a graphical plot of concentration 7.2, 7.3, or 7.4). In situations where a hydro- (Cs) versus the calibration area values. Per- gen flame is a hazard and no intrinsically form a regression analysis, and draw the safe GC is suitable, use the flexible bag col- least squares line. lection technique or an adsorption tech- 6.4 Relative Response Factors. The cali- nique. If the source temperature is below bration curve generated from the standards 100°C, and the organic concentrations are for a single organic can usually be related to suitable for the detector to be used, use the each of the individual GC response curves direct interface method. If the source gases that are developed in the laboratory for all require dilution, use a dilution interface and the compounds in the source. In the field, either the bag sample or adsorption tubes. standards for that single organic can then be The choice between these two techniques used to ‘‘calibrate’’ the GC for all the will depend on the physical layout of the organics present. This procedure should first site, the source temperature, and the storage be confirmed in the laboratory by preparing stability of the compounds if collected in the and analyzing calibration standards contain- bag. Sample polar compounds by direct ing multiple organic compounds. interfacing or dilution interfacing to prevent 6.5 Quality Assurance for Laboratory Pro- sample loss by adsorption on the bag. cedures. Immediately after the preparation 7.1 Integrated Bag Sampling and Analy- of the calibration curves and prior to the sis. presurvey sample analysis, the analysis 7.1.1 Evacuated Container Sampling Pro- audit described in 40 CFR Part 61, Appendix cedure. In this procedure, the bags are filled C, Procedure 2: ‘‘Procedure for Field Audit- by evacuating the rigid air-tight containers ing GC Analysis,’’ should be performed. The that hold the bags. Use a field sample data information required to document the analy- sheet as shown in Figure 18–10. Collect trip- sis of the audit samples has been included on licate sample from each sample location. the example data sheets shown in Figures 18– 7.1.1.1 Apparatus. 3 and 18–7. The audit analyses should agree 7.1.1.1.1 Probe. Stainless steel, Pyrex with the audit concentrations within 10 per- glass, or Teflon tubing probe, according to cent. When available, the tester may obtain the duct temperature, with 6.4-mm OD Tef- audit cylinders by contacting: U.S. Environ- lon tubing of sufficient length to connect to
885
VerDate 11
the sample bag. Use stainless steel or Teflon to leak check the system prior to the dilu- unions to connect probe and sample line. tions so as not to create a potentially explo- 7.1.1.1.2 Quick Connects. Male (2) and fe- sive atmosphere.) As an alternative, collect male (2) of stainless steel construction. the sample gas, and simultaneously dilute it 7.1.1.1.3 Needle Valve. To control gas flow. in the Tedlar bag. 7.1.1.1.4 Pump. Leakless Teflon-coated di- In the first procedure, heat the box con- aphragm-type pump or equivalent. To de- taining the sample bag to the source tem- liver at least 1 liter/min. perature, provided the components of the bag 7.1.1.1.5 Charcoal Adsorption Tube. Tube and the surrounding box can withstand this filled with activated charcoal, with glass temperature. Then transport the bag as rap- wool plugs at each end, to adsorb organic va- idly as possible to the analytical area while pors. maintaining the heating, or cover the box 7.1.1.1.6 Flowmeter. 0 to 500-ml flow with an insulating blanket. In the analytical range; with manufacturer’s calibration area, keep the box heated to source tempera- curve. ture until analysis. Be sure that the method 7.1.1.2 Sampling Procedure. To obtain a of heating the box and the control for the sample, assemble the sample train as shown heating circuit are compatible with the safe- in Figure 18–9. Leak check both the bag and ty restrictions required in each area. the container. Connect the vacuum line from To use the second procedure, prefill the the needle valve to the Teflon sample line Tedlar bag with a known quantity of inert from the probe. Place the end of the probe at gas. Meter the inert gas into the bag accord- the centroid of the stack, or at a point no ing to the procedure for the preparation of closer to the walls than 1 m, and start the gas concentration standards of volatile liq- pump with the needle valve adjusted to yield uid materials (Section 6.2.2.2), but eliminate a flow of 0.5 liter/minute. After allowing suf- the midget impinger section. Take the partly ficient time to purge the line several times, filled bag to the source, and meter the source connect the vacuum line to the bag, and gas into the bag through heated sampling evacuate until the rotameter indicates no lines and a heated flowmeter, or Teflon posi- flow. Then position the sample and vacuum tive displacement pump. Verify the dilution lines for sampling, and begin the actual sam- factors periodically through dilution and pling, keeping the rate proportional to the analysis of gases of known concentration. stack velocity. As a precaution, direct the 7.1.5 Analysis of Bag Samples. gas exiting the rotameter away from sam- 7.1.5.1 Apparatus. Same as Section 5. A pling personnel. At the end of the sample pe- minimum of three gas standards are re- riod, shut off the pump, disconnect the sam- quired. ple line from the bag, and disconnect the 7.1.5.2 Procedure. Establish proper GC op- vacuum line from the bag container, Record erating conditions as described in Section the source temperature, barometric pressure, 6.3, and record all data listed in Figure 18–7. ambient temperature, sampling flow rate, Prepare the GC so that gas can be drawn and initial and final sampling time on the through the sample valve. Flush the sample data sheet shown in Figure 18–10. Protect the loop with gas from one of the three calibra- Tedlar bag and its container from sunlight. tion mixtures, and activate the valve. Obtain When possible, perform the analysis within 2 at least two chromatograms for the mixture. hours of sample collection. The results are acceptable when the peak 7.1.2 Direct Pump Sampling Procedure. areas from two consecutive injections agree Follow 7.1.1, except place the pump and nee- to within 5 percent of their average. If they dle valve between the probe and the bag. Use do not, run additional analyses or correct a pump and needle valve constructed of the analytical techniques until this require- stainless steel or some other material not af- ment is met. Then analyze the other two fected by the stack gas. Leak check the sys- calibration mixtures in the same manner. tem, and then purge with stack gas before Prepare a calibration curve as described in the connecting to the previously evacuated the same manner. Prepare a calibration bag. curve as described in Section 6.3. 7.1.3 Explosion Risk Area Bag Sampling Analyze the source gas samples by con- Procedure. Follow 7.1.1 except replace the necting each bag to the sampling valve with pump with another evacuated can (see Fig- a piece of Teflon tubing identified for that ure 18–9a). Use this method whenever there is bag. Follow the specifications on replicate a possibility of an explosion due to pumps, analyses specified for the calibration gases. heated probes, or other flame producing Record the data listed in Figure 18–11. If cer- equipment. tain items do not apply, use the notation 7.1.4 Other Modified Bag Sampling Proce- ‘‘N.A.’’ After all samples have been analyzed, dures. In the event that condensation is ob- repeat the analyses of the calibration gas served in the bag while collecting the sample mixtures, and generate a second calibration and a direct interface system cannot be used, curve. Use an average of the two curves to heat the bag during collection, and maintain determine the sample gas concentrations. If it at a suitably elevated temperature during the two calibration curves differ by more all subsequent operations. (Note: Take care than 5 percent from their mean value, then
886
VerDate 11
report the final results by comparison to plug. If necessary, heat the probe with heat- both calibration curves. ing tape or a special heating unit capable of 7.1.6 Determination of Bag Water Vapor maintaining duct temperature. Content. Measure and record the ambient 7.2.1.2 Sample Lines. 6.4-mm OD Teflon temperature and barometric pressure near lines, heat-traced to prevent condensation of the bag. From a water saturation vapor pres- material. sure table, determine and record the water 7.2.1.3 Quick Connects. To connect sample vapor content as a decimal figure. (Assume line to gas sampling valve on GC instrument the relative humidity to be 100 percent un- and to pump unit used to withdraw source less a lesser value is known.) If the bag has gas. Use a quick connect or equivalent on been maintained at an elevated temperature the cylinder or bag containing calibration as described in Section 7.1.4, determine the gas to allow connection of the calibration stack gas water content by Method 4. gas to the gas sampling valve. 7.1.7 Quality Assurance. Immediately 7.2.1.4 Thermocouple Readout Device. Po- prior to the analysis of the stack gas sam- tentiometer or digital thermometer, to ples, perform audit analyses as described in measure source temperature and probe tem- Section 6.5. The audit analyses must agree perature. with the audit concentrations within 10 per- 7.2.1.5 Heated Gas Sampling Valve. Of cent. If the results are acceptable, proceed two-position, six-port design, to allow sam- with the analyses of the source samples. If ple loop to be purged with source gas or to they do not agree within 10 percent, then de- direct source gas into the GC instrument. termine the reason for the discrepancy, and 7.2.1.6 Needle Valve. To control gas sam- take corrective action before proceeding. pling rate from the source. 7.1.8 Emission Calculations. From the av- 7.2.1.7 Pump. Leakless Teflon-coated dia- erage calibration curve described in Section phragm-type pump or equivalent, capable of 7.1.5., select the value of Cs that corresponds at least 1 liter/minute sampling rate. to the peak area. Calculate the concentra- 7.2.1.8 Flowmeter. Of suitable range to tion Cc in ppm, dry basis, of each organic in measure sampling rate. the sample as follows: 7.2.1.9 Charcoal Adsorber. To adsorb or- ganic vapor collected from the source to pre- vent exposure of personnel to source gas. 7.2.1.10 Gas Cylinders. Carrier gas (helium or nitrogen), and oxygen and hydrogen for a flame ionization detector (FID) if one is used. where: 7.2.1.11 Gas Chromatograph. Capable of being moved into the field, with detector, Cs=Concentration of the organic from the calibration curve, ppm. heated gas sampling valve, column required to complete separation of desired compo- Pr=Reference pressure, the barometric pres- sure or absolute sample loop pressure re- nents, and option for temperature program- corded during calibration, mm Hg. ming. 7.2.1.12 Recorder/Integrator. To record re- Ti=Sample loop temperature at the time of sample analysis, °K. sults. 7.2.2 Procedure. To obtain a sample, as- Fr=Relative response factor (if applicable, see Section 6.4). semble the sampling system as shown in Fig- ure 18–12. Make sure all connections are Pi=Barometric or absolute sample loop pres- sure at time of sample analysis, mm Hg. tight. Turn on the probe and sample line heaters. As the temperature of the probe and Tr=Reference temperature, the termperature of the sample loop recorded during cali- heated line approaches the source tempera- bration, °K. ture as indicated on the thermocouple read- Bws=Water vapor content of the bag sample out device, control the heating to maintain a or stack gas, proportion by volume. temperature of 0 to 3°C above the source 7.2 Direct Interface Sampling and Analy- temperature. While the probe and heated sis Procedure. The direct interface procedure line are being heated, disconnect the sample can be used provided that the moisture con- line from the gas sampling valve, and attach tent of the gas does not interfere with the the line from the calibration gas mixture. analysis procedure, the physical require- Flush the sample loop with calibration gas ments of the equipment can be met at the and analyze a portion of that gas. Record the site, and the source gas concentration is low results. After the calibration gas sample has enough that detector saturation is not a been flushed into the GC instrument, turn problem. Adhere to all safety requirements the gas sampling valve to flush position, with this method. then reconnect the probe sample line to the 7.2.1 Apparatus. valve. Place the inlet of the probe at the cen- 7.2.1.1 Probe. Constructed of stainless troid of the duct, or at a point no closer to steel, Pyrex glass, or Teflon tubing as re- the walls than 1 m, and draw source gas into quired by duct temperature, 6.4-mm OD, en- the probe, heated line, and sample loop. larged at duct end to contain glass wool After thorough flushing, analyze the sample
887
VerDate 11
using the same conditions as for the calibra- tions. The box should be equipped with quick tion gas mixture. Repeat the analysis on an connect fittings to facilitate connection of: additional sample. Measure the peak areas (1) The heated sample line from the probe, (2) for the two samples, and if they do not agree the gas sampling valve, (3) the calibration to within 5 percent of their mean value, ana- gas mixtures, and (4) diluent gas lines. A lyze additional samples until two consecu- schematic diagram of the components and tive analyses meet this criteria. Record the connections is shown in Figure 18–13. data. After consistent results are obtained, (NOTE: Care must be taken to leak check remove the probe from the source and ana- the system prior to the dilutions so as not to lyze a second calibration gas mixture. create a potentially explosive atmosphere.) Record this calibration data and the other The heated box shown in Figure 18–13 is de- required data on the data sheet shown in Figure 18–11, deleting the dilution gas infor- signed to receive a heated line from the mation. probe. An optional design is to build a probe unit that attaches directly to the heated (NOTE: Take care to draw all samples, cali- bration mixtures, and audits through the box. In this way, the heated box contains the sample loop at the same pressure.) controls for the probe heaters, or, if the box is placed against the duct being sampled, it 7.2.3 Determination of Stack Gas Mois- may be possible to eliminate the probe heat- ture Content. Use Method 4 to measure the ers. In either case, a heated Teflon line is stack gas moisture content. used to connect the heated box to the gas 7.2.4 Quality Assurance. Same as Section sampling valve on the chromatograph. 7.1.7. Introduce the audit gases in the sample 7.3.2 Procedure. Assemble the apparatus line immediately following the probe. by connecting the heated box, shown in Fig- 7.2.5 Emission Calculations. Same as Sec- ure 18–13, between the heated sample line tion 7.1.8. from the probe and the gas sampling valve 7.3 Dilution Interface Sampling and Anal- on the chromatograph. Vent the source gas ysis Procedure. Source samples that contain from the gas sampling valve directly to the a high concentration of organic materials may require dilution prior to analysis to pre- charcoal filter, eliminating the pump and ro- vent saturating the GC detector. The appara- tameter. Heat the sample probe, sample line, tus required for this direct interface proce- and heated box. Insert the probe and source dure is basically the same as that described thermocouple to the centroid of the duct, or in the Section 7.2, except a dilution system to a point no closer to the walls than 1 m. is added between the heated sample line and Measure the source temperature, and adjust ° the gas sampling valve. The apparatus is ar- all heating units to a temperature 0 to 3 C ranged so that either a 10:1 or 100:1 dilution above this temperature. If this temperature of the source gas can be directed to the chro- is above the safe operating temperature of matograph. A pump of larger capacity is also the Teflon components, adjust the heating to required, and this pump must be heated and maintain a temperature high enough to pre- placed in the system between the sample line vent condensation of water and organic com- and the dilution apparatus. pounds. Verify the operation of the dilution 7.3.1 Apparatus. The equipment required system by analyzing a high concentration in addition to that specified for the direct gas of known composition through either the interface system is as follows: 10:1 or 100:1 dilution stages, as appropriate. 7.3.1.1 Sample Pump. Leakless Teflon- (If necessary, vary the flow of the diluent gas coated diaphragm-type that can withstand to obtain other dilution ratios.) Determine being heated to 120°C and deliver 1.5 liters/ the concentration of the diluted calibration minute. gas using the dilution factor and the calibra- 7.3.1.2 Dilution Pumps. Two Model A–150 tion curves prepared in the laboratory. Komhyr Teflon positive displacement type Record the pertinent data on the data sheet delivering 150 cc/minute, or equivalent. As an shown in Figure 18–11. If the data on the di- option, calibrated flowmeters can be used in luted calibration gas are not within 10 per- conjunction with Teflon-coated diaphragm cent of the expected values, determine pumps. whether the chromatograph or the dilution 7.3.1.3 Valves. Two Teflon three-way system is in error, and correct it. Verify the valves, suitable for connecting to 6.4-mm OD GC operation using a low concentration Teflon tubing. standard by diverting the gas into the sam- 7.3.1.4 Flowmeters. Two, for measurement ple loop, bypassing the dilution system. If of diluent gas, expected delivery flow rate to these analyses are not within acceptable be 1,350 cc/min. limits, correct the dilution system to pro- 7.3.1.5 Diluent Gas with Cylinders and vide the desired dilution factors. Make this Regulators. Gas can be nitrogen or clean dry correction by diluting a high-concentration air, depending on the nature of the source standard gas mixture to adjust the dilution gases. ratio as required. 7.3.1.6 Heated Box. Suitable for being Once the dilution system and GC oper- heated to 120°C, to contain the three pumps, ations are satisfactory, proceed with the three-way valves, and associated connec- analysis of source gas, maintaining the same
888
VerDate 11
dilution settings as used for the standards. 7.4.2 Sampling and Analysis. It is sug- Repeat the analyses until two consecutive gested that the tester follow the sampling values do not vary by more than 5 percent and analysis portion of the respective NIOSH from their mean value are obtained. method section entitled ‘‘Procedure.’’ Cali- Repeat the analysis of the calibration gas brate the pump and limiting orifice flow rate mixtures to verify equipment operation. through adsorption tubes with the bubble Analyze the two field audit samples using ei- tube flowmeter before sampling. The sample ther the dilution system, or directly connect system can be operated as a ‘‘recirculating to the gas sampling valve as required. loop’’ for this operation. Record the ambient Record all data and report the results to the temperature and barometric pressure. Then, audit supervisor. during sampling, use the rotameter to verify 7.3.3 Determination of Stack Gas Mois- that the pump and orifice sampling rate re- ture Content. Same as Section 7.2.3. mains constant. 7.3.4 Quality Assurance. Same as Section Use a sample probe, if required, to obtain 7.2.4. the sample at the centroid of the duct, or at 7.3.5 Emission Calculations. Same as Sec- a point no closer to the walls than 1 m. Mini- tion 7.2.5, with the dilution factor applied. mize the length of flexible tubing between 7.4 Adsorption Tube Procedure (Alter- the probe and adsorption tubes. Several ad- native Procedure). It is suggested that the sorption tubes can be connected in series, if tester refer to the National Institute of Oc- the extra adsorptive capacity is needed. Pro- cupational Safety and Health (NIOSH) meth- vide the gas sample to the sample system at od for the particular organics to be sampled. a pressure sufficient for the limiting orifice The principal interferent will be water to function as a sonic orifice. Record the total time and sample flow rate (or the num- vapor. If water vapor is present at concentra- ber of pump strokes), the barometric pres- tions above 3 percent, silica gel should be sure, and ambient temperature. Obtain a used in front of the charcoal. Where more total sample volume commensurate with the than one compound is present in the emis- expected concentration(s) of the volatile or- sions, then develop relative adsorptive ca- ganic(s) present, and recommended sample pacity information. loading factors (weight sample per weight 7.4.1 Additional Apparatus. In addition to adsorption media). Laboratory tests prior to the equipment listed in the NIOSH method actual sampling may be necessary to pre- for the particular organic(s) to be sampled, determine this volume. When more than one the following items (or equivalent) are sug- organic is present in the emissions, then de- gested. velop relative adsorptive capacity informa- 7.4.1.1 Probe (Optional). Borosilicate glass tion. If water vapor is present in the sample or stainless steel, approximately 6-mm ID, at concentrations above 2 to 3 percent, the with a heating system if water condensation adsorptive capacity may be severely reduced. is a problem, and a filter (either in-stack or Operate the gas chromatograph according to out-stack heated to stack temperature) to the manufacture’s instructions. After estab- remove particulate matter. In most in- lishing optimum conditions, verify and docu- stances, a plug of glass wool is a satisfactory ment these conditions during all operations. filter. Analyze the audit samples (see Section 7.4.1.2 Flexible Tubing. To connect probe 7.4.4.3), then the emission samples. Repeat to adsorption tubes. Use a material that ex- the analysis of each sample until the relative hibits minimal sample adsorption. deviation of two consecutive injections does 7.4.1.3 Leakless Sample Pump. Flow con- not exceed 5 percent. trolled, constant rate pump, with a set of 7.4.3 Standards and Calibration. The limiting (sonic) orifices to provide pumping standards can be prepared according to the rates from approximately 10 to 100 cc/min. respective NIOSH method. Use a minimum of 7.4.1.4 Bubble-Tube Flowmeter. Volume three different standards; select the con- accuracy within ± 1 percent, to calibrate centrations to bracket the expected average pump. sample concentration. Perform the calibra- 7.4.1.5 Stopwatch. To time sampling and tion before and after each day’s sample anal- pump rate calibration. yses. Prepare the calibration curve by using 7.4.1.6 Adsorption Tubes. Similar to ones the least squares method. specified by NIOSH, except the amounts of 7.4.4 Quality Assurance. adsorbent per primary/backup sections are 7.4.4.1 Determine the recovery efficiency 800/200 mg for charcoal tubes and 1040/260 mg of the pollutants of interest according to for silica gel tubes. As an alternative, the Section 7.6. tubes may contain a porous polymer adsorb- 7.4.4.2 Determination of Sample Collec- ent such as Tenax GC or XAD–2. tion Efficiency. For the source samples, ana- 7.4.1.7 Barometer. Accurate to 5 mm Hg, lyze the primary and backup portions of the to measure atmospheric pressure during adsorption tubes separately. If the backup sampling and pump calibration. portion exceeds 10 percent of the total 7.4.1.8 Rotameter. 0 to 100 cc/min, to de- amount (primary and backup), repeat the tect changes in flow rate during sampling. sampling with a larger sampling portion.
889
VerDate 11
7.4.4.3 Analysis Audit. Immediately before lyzer and the mean of the calibration gas re- the sample analyses, analyze the two audits sponse sampled through the probe shall be in accordance with Section 7.4.2. The analy- within 10 percent of each other. If the dif- sis audit shall agree with the audit con- ference in the two means is greater than 10 centration within 10 percent. percent, check for leaks throughout the sam- 7.4.4.4 Pump Leak Checks and Volume pling system and repeat the analysis of the Flow Rate Checks. Perform both of these standard through the sampling system until checks immediately after sampling with all this criterion is met. sampling train components in place. Perform 7.6.2 Recovery Study for Bag Sampling. all leak checks according to the manufactur- Follow the procedures for bag sampling and er’s instructions, and record the results. Use analysis in Section 7.1. After analyzing all the bubble-tube flowmeter to measure the three bag samples, choose one of the bag pump volume flow rate with the orifice used samples and analyze twice more (this bag in the test sampling, and the result. If it has will become the spiked bag). Spike the cho- changed by more than 5 but less than 20 per- sen bag sample with a known mixture (gase- cent, calculate an average flow rate for the ous or liquid) of all of the target pollutants. test. If the flow rate has changed by more Follow a procedure similar to the calibration than 20 percent, recalibrate the pump and re- standard preparation procedure listed in Sec- peat the sampling. tion 6.2, as appropriate. The theoretical con- 7.4.4.5 Calculations. All calculations can centration, in ppm, of each spiked compound be performed according to the respective in the bag shall be 40 to 60 percent of the av- NIOSH method. Correct all sample volumes erage concentration measured in the three to standard conditions. If a sample dilution bag samples. If a target compound was not system has been used, multiply the results detected in the bag samples, the concentra- by the appropriate dilution ratio. Correct all tion of that compound to be spiked shall be results according to the applicable procedure 5 times the limit of detection for that com- in Section 7.6. Report results as ppm by vol- pound. Analyze the bag three times after ume, dry basis. spiking. Calculate the average fraction re- 7.5 Reporting of Results. At the comple- covered (R) of each spiked target compound tion of the field analysis portion of the with the following equation: study, ensure that the data sheets shown in Figure 18–11 have been completed. Summa- t− u rize this data on the data sheets shown in R = Figure 18–15. s 7.6 Recovery Study. After conducting the presurvey and identifying all of the pollut- where ants of interest, conduct the appropriate re- t = measured average concentration (ppm) covery study during the test based on the of target compound and source sample sampling system chosen for the compounds (analysis results subsequent to bag spik- of interest. ing) 7.6.1 Recovery Study for Direct Interface u = source sample average concentration or Dilution Interface Sampling. If the proce- (ppm) of target compound in the bag dures in Section 7.2 or 7.3 are to be used to (analysis results before bag spiking) analyze the stack gas, conduct the calibra- s = theoretical concentration (ppm) of tion procedure as stated in Section 7.2.2 or spiked target compound in the bag 7.3.2, as appropriate. Upon successful comple- For the bag sampling technique to be con- tion of the appropriate calibration proce- sidered valid for a compound, 0.70≤R≤1.30. If dure, attach the mid-level calibration gas for the R value does not meet this criterion for at least one target compound to the inlet of a target compound, the sampling technique the probe or as close as possible to the inlet is not acceptable for that compound, and of the probe, but before the filter. Repeat the therefore another sampling technique shall calibration procedure by sampling and ana- be evaluated for acceptance (by repeating lyzing the mid-level calibration gas through the recovery study with another sampling the entire sampling and analytical system technique). Report the R value in the test re- until two consecutive samples are within 5 port and correct all field measurements with percent of their mean value. The mean of the the calculated R value for that compound by calibration gas response directly to the ana- using the following equation:
Measured Concentration (ppm) Reported Re sult = R
890
VerDate 11
7.6.3 Recovery Study for Adsorption Tube ms = total mass of compound measured on Sampling. If following the adsorption tube adsorbent with spiked train (µg).
procedure in Section 7.4, conduct a recovery vs = volume of stack gas sampled with study of the compounds of interest during spiked train (L). the actual field test. Set up two identical mu = total mass of compound measured on sampling trains. Collocate the two sampling adsorbent with unspiked train (µg). probes in the stack. The probes shall be v = volume of stack gas sampled with placed in the same horizontal plane, where u unspiked train (L). the first probe tip is 2.5 cm from the outside edge of the other and with a pitot tube on m× v the outside of each probe. One of the sam- = v s pling trains shall be designated the spiked R train and the other the unspiked train. Spike S all of the compounds of interest (in gaseous where S = theoretical mass of compound or liquid form) onto the adsorbent tube(s) in spiked onto adsorbent in spiked train the spiked train before sampling. The mass (µg). of each spiked compound shall be 40 to 60 7.6.3.1 Repeat the procedure in Section percent of the mass expected to be collected 7.6.3 twice more, for a total of three runs. In with the unspiked train. Sample the stack order for the adsorbent tube sampling and gas into the two trains simultaneously. Ana- analytical procedure to be acceptable for a lyze the adsorbents from the two trains uti- compound, 0.70≤R≤1.30 (R in this case is the lizing the same analytical procedure and in- strumentation. Determine the fraction of average of three runs). If the average R value spiked compound recovered (R) using the fol- does not meet this criterion for a target lowing equations. compound, the sampling technique is not ac- ceptable for that compound, and therefore m m another sampling technique shall be evalu- =s − u ated for acceptance (by repeating the recov- mv v v ery study with another sampling technique). s u Report the R value in the test report and where correct all field measurements with the cal-
mv = mass per volume of spiked compound culated R value for that compound by using measured (µg/L). the following equation:
Measured Concentration (ppm) Re ported Result = R
8. Bibliography 7. FR, 39 FR 32857–32860. 1974. 1. American Society for Testing and Mate- 8. FR, 41 FR 23069–23072 and 23076–23090. rials. C1 Through C5 Hydrocarbons in the At- 1976. mosphere by Gas Chromatography. ASTM D 9. FR, 41 FR 46569–46571. 1976. 2820–72, Part 23. Philadelphia, Pa. 23:950–958. 10. FR, 42 FR 41771–41776. 1977. 1973. 11. Fishbein, L. Chromatography of Envi- 2. Corazon, V. V. Methodology for Collect- ronmental Hazards, Volume II. Elsevier Sci- ing and Analyzing Organic Air Pollutants. entific Publishing Company. NY, NY. 1973. U.S. Environmental Protection Agency. Pub- 12. Hamersma, J. W., S. L. Reynolds, and lication No. EPA–600/2–79–042. February 1979. R. F. Maddalone. EPA/IERL–RTP Procedures 3. Dravnieks, A., B. K. Krotoszynski, J. Manual: Level 1 Environmental Assessment. Whitfield, A. O’Donnell, and T. Burgwald. U.S. Environmental Protection Agency. Re- Environmental Science and Technology. search Triangle Park, NC. Publication No. 5(12):1200–1222. 1971. EPA 600/276–160a. June 1976. 130 p. 4. Eggertsen, F. T., and F. M. Nelsen. Gas Chromatographic Analysis of Engine Ex- 13. Harris, J. C., M. J. Hayes, P. L. Levins, haust and Atmosphere. Analytical Chem- and D. B. Lindsay. EPA/IERL–RTP Proce- istry. 30(6): 1040–1043. 1958. dures for Level 2 Sampling and Analysis of 5. Feairheller, W. R., P. J. Marn, D. H. Har- Organic Materials. U.S. Environmental Pro- ris, and D. L. Harris. Technical Manual for tection Agency. Research Triangle Park, NC. Process Sampling Strategies for Organic Ma- Publication No. EPA 600/7–79–033. February terials. U.S. Environmental Protection 1979. 154 p. Agency. Research Triangle Park, NC. Publi- 14. Harris, W. E., H. W. Habgood. Pro- cation No. EPA 600/2–76–122. April 1976. 172 p. grammed Temperature Gas Chromatography. 6. FR, 39 FR 9319–9323. 1974. John Wiley & Sons, Inc. New York. 1966.
891
VerDate 11
15. Intersociety Committee. Methods of Air $7.25, Volume 5—017–033–00349–1/$10, Volume Sampling and Analysis. American Health As- 6—017–033–00369–6/$9, and Volume 7—017–033– sociation. Washington, DC. 1972. 00396–5/$7. Prices subject to change. Foreign 16. Jones, P. W., R. D. Grammar, P. E. orders add 25 percent. Strup, and T. B. Stanford. Environmental 20. Schuetzle, D., T. J. Prater, and S. R. Science and Technology.------10:806–810. 1976. Ruddell. Sampling and Analysis of Emissions 17. McNair Han Bunelli, E. J. Basic Gas from Stationary Sources; I. Odor and Total Chromatography. Consolidated Printers. Hydrocarbons. Journal of the Air Pollution Berkeley. 1969. Control Association. 25(9):925–932. 1975. 18. Nelson, G. O. Controlled Test 21. Snyder, A. D., F. N. Hodgson, M. A. Atmospheres, Principles and Techniques. Kemmer and J. R. McKendree. Utility of Ann Arbor. Ann Arbor Science Publishers. Solid Sorbents for Sampling Organic Emis- 1971. 247 p. 19. NIOSH Manual of Analytical Methods, sions from Stationary Sources. U.S. Environ- Volumes 1, 2, 3, 4, 5, 6, 7. U.S. Department of mental Protection Agency. Research Tri- Health and Human Services National Insti- angle Park, NC Publication No. EPA 600/2–76– tute for Occupational Safety and Health. 201. July 1976. 71 p. Center for Disease Control. 4676 Columbia 22. Tentative Method for Continuous Anal- Parkway, Cincinnati, Ohio 45226. April 1977– ysis of Total Hydrocarbons in the Atmos- August 1981. May be available from the Su- phere. Intersociety Committee, American perintendent of Documents, Government Public Health Association. Washington, DC Printing Office, Washington, DC 20402. Stock 1972. p. 184–186. Number/Price: Volume 1—017–033–00267–3/$13, 23. Zwerg, G., CRC Handbook of Chroma- Volume 2—017–033–00260–6/$11, Volume 3—017– tography, Volumes I and II. Sherma, Joseph 033–00261–4/$14, Volume 4—017–033–00317–3/ (ed.). CRC Press. Cleveland. 1972.
892
VerDate 11
893
VerDate 11
894
VerDate 11
895
VerDate 11
896
VerDate 11
897
VerDate 11
898
VerDate 11
899
VerDate 11
900
VerDate 11
901
VerDate 11
902
VerDate 11
903
VerDate 11
904
VerDate 11
905
VerDate 11
906
VerDate 11
907
VerDate 11
908
VerDate 11
909
VerDate 11
GASEOUS ORGANIC SAMPLING AND ANALYSIS Source Source Source CHECK LIST sample sample sample 1 2 3 (Respond with initials or number as appropriate) Sample collection time (24-hr basis) ...... Date Column tempera- 1. Presurvey data: ture: ° A. Grab sample collected ...... b Initial ( C) ...... B. Grab sample analyzed for Program rate ° composition ...... b ( C/min) ...... ° Method GC ...... b b Final ( C) ...... GC/MS ...... b b Carrier gas flow Other ...... b b rate (ml/min) ...... C. GC–FID analysis performed b ...... Detector tempera- 2. Laboratory calibration data: ture (°C) ...... A. Calibration curves pre- Chart speed (cm/ pared ...... b ...... min) ...... Number of compo- nents ...... b b Dilution gas flow Number of concentra- rate (ml/min) ...... tions/component (3 Diluent gas used required) ...... b b (symbol) ...... B. Audit samples (optional): Dilution ratio ...... Analysis completed ... b ...... Verified for con- Performed by (signa- centration ...... b ...... ture): ...... OK obtained for field Date: ...... work ...... b ...... 3. Sampling procedures: Figure 18–14. Sampling and analysis sheet. A. Method: Bag sample ...... b b METHOD 19—DETERMINATION OF SULFUR DIOX- Direct interface ...... b b Dilution interface ...... b b IDE REMOVAL EFFICIENCY AND PARTICULATE B. Number of samples col- MATTER, SULFUR DIOXIDE, AND NITROGEN lected ...... 1b ...... OXIDES EMISSION RATES 4. Field analysis: A. Total hydrocarbon analy- 1. Applicability and Principle sis performed ...... b ...... B. Calibration curve prepared b ...... 1.1 Applicability. This method is applica- Number of compo- ble for (a) determining particulate matter nents ...... b b (PM), sulfur dioxide (SO2), and nitrogen ox- Number of concentra- ides (NOx) emission rates; (b) determining tions per component sulfur removal efficiencies of fuel (3 required) ...... b b pretreatment and SO2 control devices; (c) de- Figure 18–14. Sampling and analysis check. termining overall reduction of potential SO2 emissions from steam generating units or GASEOUS ORGANIC SAMPLING AND ANALYSIS other sources as specified in applicable regu- ATA D lations; and (d) determining SO2 rates based Plant ———————————————————— on fuel sampling and analysis procedures. Date ————————————————————— 1.2 Principle. Location —————————————————— 1.2.1 Pollutant emission rates are deter- mined from concentrations of PM, SO2, or Source Source Source NOx, and oxygen (O2) or carbon dioxide (CO2) sample sample sample along with F factors (ratios of combustion 1 2 3 gas volumes to heat inputs). 1. General information: 1.2.2 An overall SO2 emission reduction ef- Source tempera- ficiency is computed from the efficiency of ture (°C) ...... fuel pretreatment systems (optional) and the Probe temperature efficiency of SO control devices. ° 2 ( C) ...... 1.2.3 The sulfur removal efficiency of a Ambient tempera- ture (°C) ...... fuel pretreatment system is determined by Atmospheric pres- fuel sampling and analysis of the sulfur and sure (mm Hg) ...... heat contents of the fuel before and after the Source pressure pretreatment system. (mm Hg) ...... 1.2.4 The SO2 removal efficiency of a con- Sampling rate (ml/ trol device is determined by measuring the min) ...... SO rates before and after the control device. Sample loop vol- 2 ume (ml) ...... 1.2.5 The inlet rates to SO2 control sys- Sample loop tem- tems and when SO2 control systems are not perature (°C) ...... used, SO2 emission rates to the atmosphere 910
VerDate 11
may be determined by fuel sampling and 2.2.1.1 Bwa=0.027. This value may be used analysis (optional). at any location at all times. 2.2.1.2 B =Highest monthly average of 2. Emission Rates of Particulate Matter, Sulfur wa Bwa that occurred within the previous cal- Dioxide, and Nitrogen Oxides endar year at the nearest Weather Service Select from the following sections the ap- Station. This value shall be determined an- plicable procedure to compute the PM, SO2, nually and may be used as an estimate for or NOx emission rate (E) in ng/J (lb/million the entire current calendar year. Btu). The pollutant concentration must be in 2.2.1.3 Bwa=Highest daily average of Bwa ng/scm (lb/scf) and the F factor must be in that occurred within a calendar month at scm/J (scf/million Btu). If the pollutant con- the nearest Weather Service Station, cal- centration (C) is not in the appropriate culated from the data from the past 3 years. units, use the following table to make the This value shall be computed for each month proper conversion: and may be used as an estimate for the cur- rent respective calendar month.
CONVERSION FACTORS FOR CONCENTRATION 2.2.2 If the moisture fraction (Bws) of the effluent gas is measured: From To Multiply by E=Cw Fd {20.9/[20.9(1¥Bws)¥%O2w]} g/scm ...... ng/scm ...... 109 mg/scm ...... ng/scm ...... 106 Eq. 19–3 lb/scf ...... ng/scm ...... 1.602×1013 2.3 Oxygen-Based F Factor, Dry/Wet × 6 ppm SO2 ...... ng/scm ...... 2.66 10 Basis. 6 ppm NOx ...... ng/scm ...... 1.912×10 ¥7 2.3.1 When the pollutant concentration is ppm SO2 ...... lb/scf ...... 1.660×10 ¥7 measured on a wet basis (C ) and O con- ppm NOx ...... lb/scf ...... 1.194×10 w 2 centration is measured on a dry basis (%O2d), An F factor is the ratio of the gas volume use the following equation: of the products of combustion to the heat E=[(Cw Fd)/(1¥Bws)]/[20.9/(20.9¥%O2d)] content of the fuel. The dry F factor (Fd) in- cludes all components of combustion less Eq. 19–4 water, the wet F factor (Fw) includes all 2.3.2 When the pollutant concentration is components of combustion, and the carbon F measured on a dry basis (Cd) and the O2 con- factor (Fc) includes only carbon dioxide. centration is measured on a wet basis (%O ), use the following equation: NOTE: Since Fw factors include water re- 2w sulting only from the combustion of hydro- E=[Cd Fd20.9]/[20.9¥O2w/(1¥Bws)] gen in the fuel, the procedures using Fw fac- tors are not applicable for computing E from Eq. 19–5 steam generating units with wet scrubbers 2.4 Carbon Dioxide-Based F Factor, Dry or with other processes that add water (e.g., Basis. When measurements are on a dry basis steam injection) for both CO2 (%CO2d) and pollutant (Cd) con- 2.1 Oxygen-Based F Factor, Dry Basis. centrations, use the following equation: When measurements are on a dry basis for E=Cd Fc(100/%CO2d) both O2 (%O2d) and pollutant (Cd) concentra- tions, use the following equation: Eq. 19–6 2.5 Carbon Dioxide-Based F Factor, Wet E=Cd Fd [20.9/(20.9¥%O2d)] Basis. When measurements are on a wet Eq. 19–1 basis for both CO2 (%CO2w) and pollutant (Cw) 2.2 Oxygen-Based F Factor, Wet Basis. concentrations, use the following equation: When measurements are on a wet basis for E=C F (100/%CO ) both O2 (%O2w) and pollutant (Cw) concentra- w c 2w tions, use either of the following: Eq. 19–7 2.2.1 If the moisture fraction of ambient 2.6 Carbon Dioxide-Based F Factor, Dry/ air (Bwa) is measured: Wet Basis.
E=[Cw Fw20.9]/[20.9(1¥Bwa)¥%O2w] 2.6.1 When the pollutant concentration is Eq. 19–2 measured on a wet basis (Cw) and CO2 con- centration is measured on a dry basis Instead of actual measurement, B may be wa (%CO ), use the following equation: estimated according to the procedure below. 2d (NOTE: The estimates are selected to en- E=[Cw Fc/(1¥Bws)] (100/%CO2d) sure that negative errors will not be larger Eq. 19–8 than 1.5 percent. However, positive errors, ¥ 2.6.2 When the pollutant concentration is or over-estimation of emissions, of as much measured on a dry basis (C ) and CO con- as 5 percent may be introduced depending d 2 centration is measured on a wet basis upon the geographic location of the facility (%CO ), use the following equation: and the associated range of ambient mois- 2w ture): E=Cd(1¥Bws)Fc(100/%CO2w) 911
VerDate 11
Eq. 19–9 and Eg. Do not use Fw factors for determin- 2.7 Direct-Fired Reheat Fuel Burning. ing Eg or Eco. If an SO2 control device is used, The effect of direct-fired reheat fuel burning measure Eco after the control device. (for the purpose of raising the temperature 2.8.1.2 Suitable methods shall be used to of the exhaust effluent from wet scrubbers to determine the heat input rates to the steam above the moisture dew-point) on emission generating units (Hb) and the gas turbine rates will be less than ±1.0 percent and, (Hg). therefore, may be ignored. 2.8.2 If a control device is used, compute 2.8 Combined Cycle-Gas Turbine Systems. the percent of potential SO2 emissions (% Ps) For gas turbine-steam generator combined using the following equations: cycle systems, determine the emissions from the steam generating unit or the percent re- Ebi=Eci¥(Hg/Hb)(Eci¥Eg) duction in potential SO2 emissions as fol- Eq. 19–11 lows: 2.8.1 Compute the emission rate from the % Ps=100 (1¥Ebo/Ebi) steam generating unit using the following Eq. 19–12 equation: where: Ebo=Eco+(Hg/Hb)(Eco¥Eg) Eq. 19–10 Ebi=pollutant rate from the steam generating unit, ng/J (lb/million Btu) where: Eci=pollutant rate in combined effluent, ng/J Ebo=pollutant emission rate from the steam (lb/million Btu). generating unit, ng/J (lb/million Btu). Use the test methods and procedures sec- E =pollutant emission rate in combined ef- co tion of Subpart GG to obtain Eci and Eg. Do fluent, ng/J (lb/million Btu). not use Fw factors for determining Eg or Eci. Eg=pollutant rate from gas turbine, ng/J (lb/ million Btu). 3. F Factors Hb=heat input rate to the steam generating unit from fuels fired in the steam gener- Use an average F factor according to Sec- ating unit, J/hr (million Btu/hr). tion 3.1 or determine an applicable F factor Hg=heat input rate to gas turbine from all according to Section 3.2. If combined fuels fuels fired in the gas turbine, J/hr (mil- are fired, prorate the applicable F factors lion Btu/hr). using the procedure in Section 3.3. 2.8.1.1 Use the test methods and proce- 3.1 Average F Factors. Average F factors dures section of Subpart GG to obtain Eco (Fd, Fw, or Fc) from Table 19–1 may be used.
TABLE 19±1ÐF FACTORS FOR VARIOUS FUELS 1
Fd Fw Fc
Fuel type 6 6 dscf/10 wscf/10 6 dscm/J Btu wscm/J Btu scm/J scf/10 Btu
Coal: 2 7 7 7 Anthracite ...... 2.71×10¥ 10,100 2.83×10¥ 10,540 0.530×10¥ 1,970 2 7 7 7 Bituminous ...... 2.63×10¥ 9,780 2.86×10¥ 10,640 0.484×10¥ 1,800 7 7 7 Lignite ...... 2.65×10¥ 9,860 3.21×10¥ 11,950 0.513×10¥ 1,910 3 7 7 7 Oil ...... 2.47×10¥ 9,190 2.77×10¥ 10,320 0.383×10¥ 1,420 Gas: 7 7 7 Natural ...... 2.43×10¥ 8,710 2.85×10¥ 10,610 0.287×10¥ 1,040 7 7 7 Propane ...... 2.34×10¥ 8,710 2.74×10¥ 10,200 0.321×10¥ 1,190 7 7 7 Butane ...... 2.34×10¥ 8,710 2.79×10¥ 10,390 0.337×10¥ 1,250 7 7 Wood ...... 2.48×10¥ 9,240 ...... 0.492×10¥ 1,830 7 7 Wood Bark ...... 2.58×10¥ 9,600 ...... 0.516×10¥ 1,920 7 7 Municipal ...... 2.57×10¥ 9,570 ...... 0.488×10¥ 1,820 Solid Waste ...... 1 Determined at standard conditions: 20 °C (68 °F) and 760 mm Hg (29.92 in. Hg). 2 As classified according to ASTM D388±77. 3 Crude, residual, or distillate.
3.2 Determined F Factors. If the fuel Fd = K[(Khd%H) + (Kc%C) + (Ks%S) + (Kn%N) burned is not listed in Table 19–1 or if the ¥ (Ko%0)]/GCVw owner or operator chooses to determine an F Eq. 19–13 factor rather than use the values in Table 19– F = K[(K %H) + (K %C) + (K %S) + (K %N) 1, use the procedure below: w hw c s n ¥ (K %0) + (K %H O)]/GCV 3.2.1 Equations. Use the equations below, o w 2 w as appropriate, to compute the F factors: Eq. 19–14 Fc=K(Kcc%C)/GCV 912
VerDate 11
Eq. 19–15
(NOTE.— Omit the %H2O term in the equa- tions for Fw if %H and %0 include the un- available hydrogen and oxygen in the form of
H2O.) where:
Fd,Fw,Fc=volumes of combustion components per unit of heat content, scm/J (scf/mil- lion Btu). %H, %C, %S, %N, %0, and
%H2O=concentrations of hydrogen, car- bon, sulfur, nitrogen, oxygen, and water from an ultimate analysis of fuel, weight percent. GCV=gross calorific value of the fuel consist- ent with the ultimate analysis, kJ/kg (Btu/lb). K=conversion factor, 10¥5 (kJ/J)/(%) [10 6 Btu/ million Btu].
Khd=22.7 (scm/kg))[(3.64 (scf/lb)/(%)]. Kc=9.57 (scm/kg)[(1.53 (scf/lb)/(%)]. Ks=3.54 (scm/kg) [(0.57 (scf/lb)/(%)]. Kn=0.86 (scm/kg [0.14 (scf/lb)/(%)]. Ko=2.85 (scm/kg) [0.46 (scf/lb)/(%)]. Khw=34.74 (scm/kg) [(5.57 (scf/lb)/(%)]. Kw=1.30 (scm/kg) [(0.21 (scf/lb)/(%)]. Kcc=2.0 (scm/kg) [(0.321 (scf/lb)/(%)]. 3.2.2 Use applicable sampling procedures in Section 5.2.1 or 5.2.2 to obtain samples for analyses. 3.2.3 Use ASTM D3176–74 (incorporated by where: reference—see § 60.17) for ultimate analysis Xk=fraction of total heat input from each of the fuel. type of fuel k. 3.2.4 Use applicable methods in Section n=number of fuels being burned in combina- 5.2.1 or 5.2.2 to determine the heat content of tion. solid or liquid fuels. For gaseous fuels, use 4. Determination of Average Pollutant Rates ASTM D1826–77 (IBR—see § 60.17) to deter- mine the heat content. 4.1 Average Pollutant Rates from Hourly 3.3 F Factors for Combination of Fuels. If Values. When hourly average pollutant rates (E ), inlet or outlet, are obtained (e.g., CEMS combinations of fuels are burned, use the fol- h values), compute the average pollutant rate lowing equations, as applicable unless other- (Ea) for the performance test period (e.g., 30 wise specified in applicable subpart: days) specified in the applicable regulation using the following equation:
where: Ea=average pollutant rate for the specified performance test period, ng/J (lb/million Btu). Eh=hourly average pollutant, ng/J (lb/million Btu). H=total number of operating hours for which pollutant rates are determined in the performance test period. 4.2 Average Pollutant Rates from Other than Hourly Averages. When pollutant rates
913
VerDate 11
are determined from measured values rep- where: resenting longer than 1-hour periods (e.g., Ed=average pollutant rate for each sampling daily fuel sampling and analyses or Method period (e.g., 24-hr Method 6B sample or 6B values), or when pollutant rates are deter- 24-hr fuel sample) or for each fuel lot mined from combinations of 1-hour and (e.g., amount of fuel bunkered), ng/J (lb/ longer than 1-hour periods (e.g., CEMS and million Btu). Method 6B values), compute the average pol- nd=number of operating hours of the affected lutant rate (Ea) for the performance test pe- riod (e.g., 30 days) specified in the applicable facility within the performance test pe- regulation using the following equation: riod for each Ed determined. D=number of sampling periods during the performance test period. 4.3 Daily Geometric Average Pollutant Rates from Hourly Values. The geometric
average pollutant rate (Ega) is computed using the following equation:
n = 1 Ega EXP ∑[]1n(). E hj Eq 19- 20 a n j=1
where: 5. Determination of Overall Reduction in Potential Sulfur Dioxide Emission Ega = daily geometric average pollutant rate, ng/J (lbs/million Btu) or ppm corrected 5.1 Overall Percent Reduction. Compute to 7 percent O2. the overall percent SO2 reduction (%Ro) Ehj = hourly arithmetic average pollutant using the following equation: rate for hour ‘‘j,’’ ng/J (lb/million Btu) or %R =100 [1.0¥(1.0¥%R /100)(1.0¥%R /100)] ppm corrected to 7 percent O . o f g 2 Eq. 19–21 n = total number of hourly averages for which pollutant rates are available with- where: in the 24 hr midnight to midnight daily %Rf=SO2 removal efficiency from fuel period. pretreatment, percent. ln = natural log of indicated value. %Rg=SO2 removal efficiency of the control EXP = the natural logarithmic base (2.718) device, percent. raised to the value enclosed by brackets. 5.2 Pretreatment Removal Efficiency (Op- tional). Compute the SO2 removal efficiency from fuel pretreatment (%Rf) for the averag- ing period (e.g., 90 days) as specified in the applicable regulation using the following equation:
where: n=number of fuel lots during the averaging period. %Sp, %Sr=sulfur content of the product and raw fuel lots, respectively, dry basis NOTE: In calculating %Rf, include %S and weight percent. GCV values for all fuel lots that are not pretreated and are used during the averaging GCVp, GCVr=gross calorific value for the product and raw fuel lots, respectively, period. dry basis, kg/kg (Btu/lb). 5.2.1 Solid Fossil (Including Waste) Fuel—
Lp, Lr=weight of the product and raw fuel Sampling and Analysis. lots, respectively, metric ton (ton).
914
VerDate 11
NOTE: For the purposes of this method, raw ASTM D4057–81 to determine the sulfur con- fuel (coal or oil) is the fuel delivered to the tent (%S) and ASTM D240–76 (all methods desulfurization (pretreatment) facility. For cited IBR—see § 60.17) to determine the GCV oil, the input oil to the oil desulfurization of each gross sample. These values may be process (e.g., hydrotreatment) is considered assumed to be on a dry basis. The owner or to be the raw fuel. operator of an affected facility may elect to 5.2.1.1 Sample Increment Collection. Use determine the GCV by sampling the oil com- ASTM D2234–76 (IBR—see § 60.17), Type I, busted on the first steam generating unit op- Conditions A, B, or C, and systematic spac- erating day of each calendar month and then ing. As used in this method, systematic spac- using the lowest GCV value of the three GCV ing is intended to include evenly spaced in- values per quarter for the GCV of all oil com- crements in time or increments based on busted in that calendar quarter. equal weights of coal passing the collection 5.2.3 Use appropriate procedures, subject area. to the approval of the Administrator, to de- As a minimum, determine the number and termine the fraction of total mass input de- weight of increments required per gross sam- rived from each type of fuel. ple representing each coal lot according to 5.3 Control Device Removal Efficiency. Table 2 or Paragraph 7.1.5.2 of ASTM D2234– Compute the percent removal efficiency 76. Collect one gross sample for each lot of (%Rg of the control device using the follow- raw coal and one gross sample for each lot of ing equation: product coal. %R =100[1.0 E /E ] 5.2.1.2 ASTM Lot Size. For the purpose of g ¥ ao ai Section 5.2 (fuel pretreatment), the lot size Eq. 19–23 of product coal is the weight of product coal where: from one type of raw coal. The lot size of raw E , E =average pollutant rate of the control coal is the weight of raw coal used to ao ai device, outlet and inlet, respectively, for produce one lot of product coal. Typically, the performance test period, ng/J (lb/mil- the lot size is the weight of coal processed in lion Btu). a 1-day (24-hour) period. If more than one 5.3.1 Use continuous emission monitoring type of coal is treated and produced in 1 day, then gross samples must be collected and systems or test methods, as appropriate, to analyzed for each type of coal. A coal lot size determine the outlet SO2 rates and, if appro- equaling the 90-day quarterly fuel quantity priate, the inlet SO2 rates. The rates may be for a steam generating unit may be used if determined as hourly (Eh) or other sampling representative sampling can be conducted period averages (Ed). Then, compute the av- for each raw coal and product coal. erage pollutant rates for the performance test period (Eao and Eai) using the procedures NOTE: Alternative definitions of lot sizes in Section 4. may be used, subject to prior approval of the 5.3.2 As an alternative, as-fired fuel sam- Administrator. pling and analysis may be used to determine 5.2.1.3 Gross Sample Analysis. Use ASTM inlet SO2 rates as follows: D2013–72 to prepare the sample, ASTM D3177– 5.3.2.1 Compute the average inlet SO2 75 or ASTM D4239–85 to determine sulfur con- rate (Edi) for each sampling period using the tent (%S), ASTM D3173–73 to determine following equation: moisture content, and ASTM D2015–77 or ASTM D3286–85 to determine gross calorific Edi=K (%S/GCV) value (GCV) (all methods cited IBR—see Eq. 19–24 § 60.17) on a dry basis for each gross sample. where: 5.2.2 Liquid Fossil Fuel—Sampling and Analysis. See NOTE under Section 5.2.1. Edi=average inlet SO2 rate for each sampling 5.2.2.1 Sample Collection. Follow the pro- period d, ng/J (lb/million Btu) cedures for continuous sampling in ASTM % S=sulfur content of as-fired fuel lot, dry D270–65 (Reapproved 1975) (IBR—see § 60.17) basis, weight percent. for each gross sample from each fuel lot. GCV=gross calorific value of the fuel lot con- 5.2.2.2 Lot Size. For the purpose of Sec- sistent with the sulfur analysis, kJ/kg tion 5.2 (fuel pretreatment), the lot size of a (Btu/lb). product oil is the weight of product oil from K=2×107[(kg)(ng)/(%)(J)]{2×104(lb)(Btu/ one pretreatment facility and intended as (%))(million Btu)} one shipment (ship load, barge load, etc.). After calculating E use the procedures in The lot size of raw oil is the weight of each di Section 4.2 to determine the average inlet crude liquid fuel type used to produce a lot SO rate for the performance test period of product oil. 2 (Eai). NOTE: Alternative definitions of lot sizes 5.3.2.2 Collect the fuel samples from a lo- may be used, subject to prior approval of the cation in the fuel handling system that pro- Administrator. vides a sample representative of the fuel 5.2.2.3 Sample Analysis. Use ASTM D129– bunkered or consumed during a steam gener- 64 (Reapproved 1978), ASTM D1552–83, or ating unit operating day.
915
VerDate 11
For the purpose of as-fired fuel sampling unit operating days when an oil-fired steam under Section 5.3.2 or Section 6, the lot size generating unit is operated without oil being for coal is the weight of coal bunkered or added to the oil ‘‘day’’ tank, the oil analysis consumed during each steam generating unit from the previous day shall be used until the operating day. The lot size for oil is the ‘‘day’’ tank is filled again. weight of oil supplied to the ‘‘day’’ tank or Alternative definitions of fuel lot size may consumed during each steam generating unit be used, subject to prior approval of the Ad- operating day. ministrator. For reporting and calculation purposes, the gross sample shall be identified with the 5.3.2.3 Use ASTM procedures specified in calendar day on which sampling began. For Section 5.2.1 or 5.2.2 to determine the sulfur steam generating unit operating days when a contents (%S) and gross calorific values coal-fired steam generating unit is operated (GCV). without coal being added to the bunkers, the 5.4 Daily Geometric Average Percent Re- coal analysis from the previous ‘‘as duction from Hourly Values. The geometric bunkered’’ coal sample shall be used until average percent reduction (%Rga) is com- coal is bunkered again. For steam generating puted using the following equation:
1 n %/.R=100 1 − EXP ∑1n() E E Eq 19- 24 a ga jo ji n j=1
where: GCV=gross calorific value of the fuel lot con- sistent with the sulfur analysis, kJ/kg %Rga = daily geometric average percent re- duction. (Btu/lb). K=2×107[(kg)(ng)/(%)(J)] {2×104(lb)(Btu/ Ejo, Eji = matched pair hourly arithmetic av- erage pollutant rate, outlet and inlet, re- (%))(million Btu)}
spectively, ng/J (lb/million Btu) or ppm After calculating Edi use the procedures in corrected to 7 percent O2. Section 4–2 to determine the average SO2 n = total number of hourly averages for emission rate to the atmosphere for the per- which paired inlet and outlet pollutant formance test period (Eao). rates are available within the 24-hr mid- night to midnight daily period. 7. Determination of Compliance When Minimum ln=natural log of indicated value. Data Requirement Is Not Met EXP = the natural logarithmic base (2.718) 7.1 Adjusted Emission Rates and Control raised to the value enclosed by brackets. Device Removal Efficiency. When the mini- NOTE: The calculation includes only paired mum data requirement is not met, the Ad- data sets (hourly average) for the inlet and ministrator may use the following adjusted outlet pollutant measurements. emission rates or control device removal ef- ficiencies to determine compliance with the 6. Sulfur Retention Credit for Compliance Fuel applicable standards. 7.1.1 Emission Rate. Compliance with the If fuel sampling and analysis procedures in emission rate standard may be determined Section 5.2.1 are being used to determine av- by using the lower confidence limit of the erage SO emission rates (E ) to the atmos- 2 as emission rate (Eao*) as follows: phere from a coal-fired steam generating E *=E t S unit when there is no SO2 control device, the ao ao¥ 0.95 o following equation may be used to adjust the Eq. 19–26 emission rate for sulfur retention credits (no where: credits are allowed for oil-fired systems) (Edi) for each sampling period using the following So=standard deviation of the hourly average equation: emission rates for each performance test period, ng/J (lb/million Btu). Edi=0.97 K (%S/GCV) t0.95=values shown in Table 19–2 for the indi- Eq. 19–25 cated number of data points n. 7.1.2 Control Device Removal Efficiency. where: Compliance with the overall emission reduc- Edi=average inlet SO2 rate for each sampling tion (%Ro) may be determined by using the period d, ng/J (lb/million Btu) lower confidence limit of the emission rate %S=sulfur content of as-fired fuel lot, dry (Eao*) and the upper confidence limit of the basis, weight percent. inlet pollutant rate (Eai*) in calculating the 916
VerDate 11
control device removal efficiency (%Rg) as Eq. 19–28 follows: where: %Rg=100 [1.0¥Eao*/Eai*] Si=standard deviation of the hourly average Eq. 19–27 inlet pollutant rates for each perform-
Eai*=Eai+t0.95 Si ance test period, ng/J (lb/million Btu).
TABLE 19±2ÐVALUES FOR T0.95
1 1 1 n t0.95 n t0.95 n t0.95
2 6.31 8 1.89 22±26 1.71 3 2.42 9 1.86 27±31 1.70 4 2.35 10 1.83 32±51 1.68 5 2.13 11 1.81 59±91 1.67 6 2.02 12±16 1.77 92±151 1.66 7 1.94 17±21 1.73 152 or more 1.65 1 The values of this table are corrected for n-1 degrees of freedom. Use n equal to the number (H) of hourly average data points.
7.2 Standard Deviation of Hourly Average ation (Se) of the hourly average pollutant Pollutant Rates. Compute the standard devi- rates using the following equation:
where: equivalent. The diluent determination is S=standard deviation of the hourly average used to adjust the NOx and SO2 concentra- pollutant rates for each performance test tions to a reference condition. period, ng/J (lb/million Btu). 2. Definitions Hr=total numbers of hours in the perform- 2.1 Measurement System. The total equip- ance test period (e.g., 720 hours for 30-day ment required for the determination of a gas performance test period). concentration or a gas emission rate. The Equation 19–29 may be used to compute the system consists of the following major sub- standard deviation for both the outlet (So) systems: and, if applicable, inlet (S ) pollutant rates. i 2.1.1 Sample Interface. That portion of a METHOD 20—DETERMINATION OF NITROGEN OX- system that is used for one or more of the IDES, SULFUR DIOXIDE, AND DILUENT EMIS- following: sample acquisition, sample trans- SIONS FROM STATIONARY GAS TURBINES portation, sample conditioning, or protec- tion of the analyzers from the effects of the 1. Principle and Applicability stack effluent. 1.1 Applicability. This method is applica- 2.1.2 NOx Analyzer. That portion of the ble for the determination of nitrogen oxides system that senses NOx and generates an (NOx), sulfur dioxide (SO2), and a diluent gas, output proportional to the gas concentra- either oxygen (O2) or carbon dioxide (CO2), tion. emissions from stationary gas turbines. For 2.1.3 O2 Analyzer. That portion of the sys- the NOx and diluent concentration deter- tem that senses O2 and generates an output minations, this method includes: (1) Meas- proportional to the gas concentration. urement system design criteria; (2) Analyzer 2.1.4 CO2 Analyzer. That portion of the performance specifications and performance system that senses CO2 and generates an out- test procedures; and (3) Procedures for emis- put proportional to the gas concentration. sion testing. 2.1.5 Data Recorder. That portion of the 1.2 Principle. A gas sample is continu- measurement system that provides a perma- ously extracted from the exhaust stream of a nent record of the analyzer(s) output. The stationary gas turbine; a portion of the sam- data recorder may include automatic data ple stream is conveyed to instrumental ana- reduction capabilities. lyzers for determination of NOx and diluent 2.2 Span Value. The upper limit of a gas content. During each NOx and diluent deter- concentration measurement range that is mination, a separate measurement of SO2 specified for affected source categories in the emissions is made, using Method 6, or its applicable part of the regulations.
917
VerDate 11
2.3 Calibration Gas. A known concentra- 2.8 Interference Response. The output re- tion of a gas in an appropriate diluent gas. sponse of the measurement system to a com- 2.4 Calibration Error. The difference be- ponent in the sample gas, other than the gas tween the gas concentration indicated by the component being measured. measurement system and the known con- 3. Measurement System Performance Specifica- centration of the calibration gas. tions
2.5 Zero Drift. The difference in the meas- 3.1 NO2 to NO Converter. Greater than 90 urement system output readings from zero percent conversion efficiency of NO2 to NO. after a stated period of operation during 3.2 Interference Response. Less than ±2 which no unscheduled maintenance, repair, percent of the span value. or adjustment took place and the input con- 3.3 Response Time. No greater than 30 centration at the time of the measurements seconds. was zero. 3.4 Zero Drift. Less than ±2 percent of the 2.6 Calibration Drift. The difference in the span value over the period of each test run. measurement system output readings from 3.5 Calibration Drift. Less than ±2 percent the known concentration of the calibration of the span value over the period of each test gas after a stated period of operation during run. which no unscheduled maintenance, repair, 4. Apparatus and Reagents or adjustment took place and the input at 4.1 Measurement System. Use any meas- the time of the measurements was a high- urement system for NOx and diluent that is level value. expected to meet the specifications in this 2.7 Response Time. The amount of time method. A schematic of an acceptable meas- required for the measurement system to dis- urement system is shown in Figure 20–1. The play on the data output 95 percent of a step essential components of the measurement change in pollutant concentration. system are described below:
918
VerDate 11
4.1.1 Sample Probe. Heated stainless steel, NO converter is not necessary if the NO2 por- or equivalent, open-ended, straight tube of tion of the exhaust gas is less than 5 percent sufficient length to traverse the sample of the total NOx concentration. As a guide- points. line, an NO2 to NO converter is not necessary 4.1.2 Sample Line. Heated (>95°C) stain- if the gas turbine is operated at 90 percent or less steel or Teflon tubing to transport the more of peak load capacity. A converter is sample gas to the sample conditioners and necessary under lower load conditions. analyzers. 4.1.5 Moisture Removal Trap. A refrig- 4.1.3 Calibration Valve Assembly. A three- erator-type condenser or other type device way valve assembly to direct the zero and designed to continuously remove condensate calibration gases to the sample conditioners from the sample gas while maintaining mini- and to the analyzers. The calibration valve mal contact between any condensate and the assembly shall be capable of blocking the sample gas. The moisture removal trap is sample gas flow and of introducing calibra- not necessary for analyzers that can measure tion gases to the measurement system when NO concentrations on a wet basis; for these in the calibration mode. x analyzers, (a) heat the sample line up to the 4.1.4 NO2 to NO Converter. That portion of inlet of the analyzers, (b) determine the the system that converts the nitrogen diox- moisture content using methods subject to ide (NO ) in the sample gas to nitrogen oxide 2 the approval of the Administrator, and (c) (NO). Some analyzers are designed to meas- correct the NO and diluent concentrations ure NO as NO on a wet basis and can be x x 2 to a dry basis. used without an NO2 to NO converter or a moisture removal trap provided the sample 4.1.6 Particulate Filter. An in-stack or an line to the analyzer is heated (>95°C) to the out-of-stack glass fiber filter, of the type specified in EPA Method 5; however, an out- inlet of the analyzer. In addition, an NO2 to
919
VerDate 11
of-stack filter is recommended when the for Establishing True Concentrations of Gases stack gas temperature exceeds 250 to 300°C. Used for Calibrations and Audits of Continuous 4.1.7 Sample Pump. A nonreactive leak- Source Emission Monitors (Protocol Number 1) free sample pump to pull the sample gas that is available from the Environmental through the system at a flow rate sufficient Monitoring Systems Laboratory, Quality As- to minimize transport delay. The pump shall surance Branch, Mail Drop 77, Environ- be made from stainless steel or coated with mental Protection Agency, Research Tri- Teflon or equivalent. angle Park, NC 27711. Obtain a certification 4.1.8 Sample Gas Manifold. A sample gas from the gas manufacturer that the protocol manifold to divert portions of the sample gas was followed. These calibration gases are not stream to the analyzers. The manifold may to be analyzed with the Reference Methods. be constructed of glass, Teflon, stainless (b) The second alternative is to use calibra- steel, or equivalent. tion gases not prepared according to the pro- 4.1.9 Diluent Gas Analyzer. An analyzer to tocol. If this alternative is chosen, within 1 determine the percent O2 or CO2 concentra- month prior to the emission test, analyze tion of the sample gas. each of the calibration gas mixtures in trip- 4.1.10 Nitrogen Oxides Analyzer. An ana- licate using Method 7 or the procedure out- lyzer to determine the ppm NOx concentra- lined in Citation 1 for NOx and use Method 3 tion in the sample gas stream. for O2 or CO2. Record the results on a data 4.1.11 Data Recorder. A strip-chart re- sheet (example is shown in Figure 20–2). For corder, analog computer, or digital recorder the low-level, mid-level, or high-level gas for recording measurement data. mixtures, each of the individual NOx analyt- 4.2 Sulfur Dioxide Analysis. EPA Method ical results must be within 10 percent (or 10 6 apparatus and reagents. ppm, whichever is greater) of the triplicate 4.3 NOx Calibration Gases. The calibra- set average (O2 or CO2 test results must be tion gases for the NOx analyzer shall be NO within 0.5 percent O2 or CO2); otherwise, dis- in N2. Use four calibration gas mixtures as card the entire set and repeat the triplicate specified below: analyses. If the average of the triplicate ref- 4.3.1 High-level Gas. A gas concentration erence method test results is within 5 per- that is equivalent to 80 to 90 percent of the cent for NOx gas or 0.5 percent O2 or CO2 for span value. the O2 or CO2 gas of the calibration gas man- 4.3.2 Mid-level Gas. A gas concentration ufacturer’s tag value, use the tag value; oth- that is equivalent to 45 to 55 percent of the erwise, conduct at least three additional ref- span value. erence method test analyses until the results 4.3.3 Low-level Gas. A gas concentration of six individual NOx runs (the three original that is equivalent to 20 to 30 percent of the plus three additional) agree within 10 per- span value. cent (or 10 ppm, whichever is greater) of the 4.3.4 Zero Gas. A gas concentration of less average (O2 or CO2 test results must be with- than 0.25 percent of the span value. Ambient in 0.5 percent O2 or CO2). Then use this aver- air may be used for the NOx zero gas. age for the cylinder value. 4.4 Diluent Calibration Gases. 5.2 Measurement System Preparation. 4.4.1 For O2 calibration gases, use purified Prior to the emission test, assemble the air at 20.9 percent O2 as the high-level O2 gas. measurement system following the manufac- Use a gas concentration between 11 and 15 turer’s written instructions in preparing and percent O2 in nitrogen for the mid-level gas, operating the NO2 to NO converter, the NOx and use purified nitrogen for the zero gas. analyzer, the diluent analyzer, and other 4.4.2 For CO2 calibration gases, use a gas components. concentration between 8 and 12 percent CO2 in air for the high-level calibration gas. Use FIGURE 20–2—ANALYSIS OF CALIBRATION a gas concentration between 2 and 5 percent GASES CO2 in air for the mid-level calibration gas, Date ————— (Must be within 1 month and use purified air (<100 ppm CO2) as the prior to the test period) zero level calibration gas. Reference method used ——————————— 5. Measurement System Performance Test Proce- Gas concentration, ppm dures Sample run Perform the following procedures prior to Low level a Mid level b High level c measurement of emissions (Section 6) and 1 only once for each test program, i.e., the se- ries of all test runs for a given gas turbine 2 engine. 3 5.1 Calibration Gas Checks. There are two alternatives for checking the concentrations Average of the calibration gases. (a) The first is to use calibration gases that are documented Maximum % devi- d traceable to National Bureau of Standards ation . Reference Materials. Use Traceability Protocol a Average must be 20 to 30% of span value.
920
VerDate 11
b Average must be 45 to 55% of span value. of the low-level (not applicable for the dilu- c Average must be 80 to 90% of span value. d Must be ≤±10% of applicable average or 10 ppm, which- ent analyzer) and high-level gases within 2 ever is greater. percent of the span value, the calibration shall be considered invalid. Take corrective 5.3 Calibration Check. Conduct the cali- measures on the measurement system before bration checks for both the NOx and the dilu- ent analyzers as follows: proceeding with the test. 5.3.1 After the measurement system has 5.4 Interference Response. Introduce the been prepared for use (Section 5.2), introduce gaseous components listed in Table 20–1 into zero gases and the mid-level calibration the measurement system separately, or as gases; set the analyzer output responses to gas mixtures. Determine the total inter- the appropriate levels. Then introduce each ference output response of the system to of the remainder of the calibration gases de- these components in concentration units; scribed in Sections 4.3 or 4.4, one at a time, record the values on a form similar to Figure to the measurement system. Record the re- 20–4. If the sum of the interference responses sponses on a form similar to Figure 20–3. of the test gases for either the NOx or diluent 5.3.2 If the linear curve determined from analyzers is greater than 2 percent of the ap- the zero and mid-level calibration gas re- plicable span value, take corrective measure sponses does not predict the actual response on the measurement system.
FIGURE 20±3ÐZERO AND CALIBRATION DATA Turbine type ...... Identification number. Date ...... Test number. Analyzer type ...... Identification number.
Cylinder value, ppm or Initial analyzer re- Final analyzer re- Difference: initial-final, % sponse, ppm or % sponses, ppm or % ppm or %
Zero gas.
Low-level gas.
Mid-level gas.
High-level gas.
Test Absolute difference gas Concentration, ppm Analyzer output re- % of Percent drift = ×100 type sponse span Span value
TABLE 20±1ÐINTERFERENCE TEST GAS CONCENTRATION
CO 500±50 ppm ...... CO2 ...... 10±1 percent. SO2 200±20 ppm ...... O2 ...... 20.9±1 percent.
FIGURE 20–4—INTERFERENCE RESPONSE Date of test ————— Analyzer type ———————————————— Serial No. —————————————————— Analyzeroutput response %of span = ×100 Instrumentspan Conduct an interference response test of each analyzer prior to its initial use in the field. Thereafter, recheck the measurement system if changes are made in the instru- mentation that could alter the interference response, e.g., changes in the type of gas de- tector. In lieu of conducting the interference re- sponse test, instrument vendor data, which demonstrate that for the test gases of Table
921
VerDate 11
20–1 the interference performance specifica- 5.6.2 Alternatively, the NO2 to NO con- tion is not exceeded, are acceptable. verter check described in Title 40, Part 86: 5.5 Response Time. To determine response Certification and Test Procedures for Heavy- time, first introduce zero gas into the sys- duty Engines for 1979 and Later Model Years tem at the calibration valve until all read- may be used. Other alternative procedures ings are stable; then, switch to monitor the may be used with approval of the Adminis- stack effluent until a stable reading can be trator. obtained. Record the upscale response time. 6. Emission Measurement Test Procedure Next, introduce high-level calibration gas 6.1 Preliminaries. into the system. Once the system has sta- 6.1.1 Selection of a Sampling Site. Select bilized at the high-level concentration, a sampling site as close as practical to the switch to monitor the stack effluent and exhaust of the turbine. Turbine geometry, wait until a stable value is reached. Record stack configuration, internal baffling, and the downscale response time. Repeat the pro- point of introduction of dilution air will vary cedure three times. A stable value is equiva- for different turbine designs. Thus, each of lent to a change of less than 1 percent of these factors must be given special consider- span value for 30 seconds or less than 5 per- ation in order to obtain a representative cent of the measured average concentration sample. Whenever possible, the sampling site for 2 minutes. Record the response time data shall be located upstream of the point of in- on a form similar to Figure 20–5, the read- troduction of dilution air into the duct. ings of the upscale or downscale reponse Sample ports may be located before or after time, and report the greater time as the ‘‘re- the upturn elbow, in order to accommodate sponse time’’ for the analyzer. Conduct a re- the configuration of the turning vanes and sponse time test prior to the initial field use baffles and to permit a complete, unob- of the measurement system, and repeat if structed traverse of the stack. The sample changes are made in the measurement sys- ports shall not be located within 5 feet or 2 tem. diameters (whichever is less) of the gas dis- charge to atmosphere. For supplementary- FIGURE 20–5—RESPONSE TIME fired, combined-cycle plants, the sampling Date of test ————— site shall be located between the gas turbine Analyzer type ———————————————— and the boiler. The diameter of the sample ports shall be sufficient to allow entry of the S/N ————————————————————— sample probe. Span gas concentration: ———— ppm. 6.1.2 A preliminary O2 or CO2 traverse is Analyzer span setting: ———— ppm. made for the purpose of selecting sampling Upscale: points of low O2 or high CO2 concentrations, 1 ———— seconds. as appropriate for the measurement system. 2 ———— seconds. Conduct this test at the turbine operating 3 ———— seconds. condition that is the lowest percentage of Average upscale response —— seconds. peak load operation included in the test pro- Downscale: gram. Follow the procedure below, or use an 1 ———— seconds. alternative procedure subject to the ap- 2 ———— seconds. proval of the Administrator. 3 ———— seconds. 6.1.2.1 Minimum Number of Points. Select Average downscale response —— seconds. a minimum number of points as follows: (1) Eight, for stacks having cross-sectional System response time= areas less than 1.5 m2 (16.1 ft2); (2) eight plus slower average time= one additional sample point for each 0.2 m2 ———— seconds. (2.2 ft2 of areas, for stacks of 1.5 m2 to 10.0 m2 2 5.6 NO2 to NO Conversion Efficiency. (16.1–107.6 ft ) in cross-sectional area; and (3) 5.6.1 Add gas from the mid-level NO in N2 49 sample points (48 for circular stacks) for calibration gas cylinder to a clean, evacu- stacks greater than 10.0 m 2 (107.6 ft 2) in ated, leak-tight Tedlar bag. Dilute this gas cross-sectional area. Note that for circular approximately 1:1 with 20.9 percent O2, puri- ducts, the number of sample points must be fied air. Immediately attach the bag outlet a multiple of 4, and for rectangular ducts, to the calibration valve assembly and begin the number of points must be one of those operation of the sampling system. Operate listed in Table 20–2; therefore, round off the the sampling system, recording the NOx re- number of points (upward), when appro- sponse, for at least 30 minutes. If the NO2 to priate. NO conversion is 100 percent, the instrument 6.1.2.2 Cross-sectional Layout and Loca- response will be stable at the highest peak tion of Traverse Points. After the number of value observed. If the response at the end of traverse points for the preliminary diluent 30 minutes decreases more than 2.0 percent sampling has been determined, use Method 1 of the highest peak value, the system is not to located the traverse points. acceptable and corrections must be made be- 6.1.2.3 Preliminary Diluent Measurement. fore repeating the check. While the gas turbine is operating at the
922
VerDate 11
lowest percent of peak load, conduct a pre- tion of diluent and NOx at each point and liminary diluent measurement as follows: record the data on Figure 20–8. Position the probe at the first traverse point and begin sampling. The minimum sampling FIGURE 20–7—STATIONARY GAS TURBINE DATA time at each point shall be 1 minute plus the average system response time. Determine TURBINE OPERATION RECORD the average steady-state concentration of Test operator —————————— Date —— diluent at each point and record the data on Figure 20–6. Turbine identification: 6.1.2.4 Selection of Emission Test Sam- Type ———————————————————— pling Points. Select the eight sampling Serial No. ————————————————— points at which the lowest O concentrations 2 Location: or highest CO2 concentrations were obtained. Sample at each of these selected points dur- Plant ——————————————————— ing each run at the different turbine load City ———————————————————— conditions. More than eight points may be Ambient temperature ———————————— used, if desired, providing that the points se- Ambient humidity ————————————— lected as described above are included. Test time start ——————————————— Test time finish ——————————————— TABLE 20±2ÐCROSS-SECTIONAL LAYOUT FOR Fuel flow rate a ——————————————— RECTANGULAR STACKS Water or steam flow ratea ————————— Matrix lay- Ambient pressure —————————————— out Ultimate fuel analysis: C ————————————————————— No. of traverse points: 9 ...... 3 x 3 H ————————————————————— 12 ...... 4 x 3 O ————————————————————— 16 ...... 4 x 4 N ————————————————————— 20 ...... 5 x 4 S ————————————————————— 25 ...... 5 x 5 30 ...... 6 x 5 Ash ———————————————————— 36 ...... 6 x 6 H20 ———————————————————— 42 ...... 7 x 6 49 ...... 7 x 7 Trace metals: Na ————————————————————— FIGURE 20–6—PRELIMINARY DILUENT Va ————————————————————— TRAVERSE K ————————————————————— etc b ———————————————————— Date ————— Operating load ——————————————— Location: a Plant ——————————————————— Describe measurement method, i.e., con- City, State ———————————————— tinuous flow meter, start finish volumes, etc. bi.e., additional elements added for smoke Turbine identification: suppression. Manufacturer ——————————————— Model, serial number ——————————— FIGURE 20–8—STATIONARY GAS TURBINE SAMPLE POINT RECORD Sample point Diluent concentration, ppm Turbine identification: Manufacturer ——————————————— Model, serial No. ————————————— 6.2 NOx and Diluent Measurement. This Location: test is to be conducted at each of the speci- Plant ——————————————————— fied load conditions. Three test runs at each City, State ———————————————— load condition constitute a complete test. Ambient temperature ———————————— 6.2.1 At the beginning of each NOx test run and, as applicable, during the run, record Ambient pressure —————————————— turbine data as indicated in Figure 20–7. Date ————— Also, record the location and number of the Test time: start ——————————————— traverse points on a diagram. Test time: finish —————————————— 6.2.2 Position the probe at the first point Test operator name ————————————— determined in the preceding section and begin sampling. The minimum sampling Diluent instrument type —————————— time at each point shall be at least 1 minute Serial No ————————————————— plus the average system response time. De- NOx instrument type ———————————— termine the average steady-state concentra- Serial No. —————————————————
923
VerDate 11
Sample point Time, min Diluenta, % NO a, ppm x C C = w Eq. 20- 1 d − 1 Bws where: Cd=Pollutant or diluent concentration ad- justed to dry conditions, ppm or percent. Cw=Pollutant or diluent concentration meas- ured under moist sample conditions, ppm or percent. Bws=Moisture content of sample gas as meas- ured with Method 4, reference method, or other approved method, percent/100. 7.2 CO Correction Factor. If pollutant aAverage steady-state value from recorder or instrument 2 readout. concentrations are to be corrected to 15 per- cent O2 and CO2 concentration is measured in 6.2.3 After sampling the last point, con- lieu of O2 concentration measurement, a CO2 clude the test run by recording the final tur- correction factor is needed. Calculate the bine operating parameters and by determin- CO2 correction factor as follows: ing the zero and calibration drift, as follows: 7.2.1 Calculate the fuel-specific F0 value Immediately following the test run at each for the fuel burned during the test using val- load condition, or if adjustments are nec- ues obtained from Method 19, Section 5.2, essary for the measurement system during and the following equation. the tests, reintroduce the zero and mid-level calibration gases as described in Sections 4.3 FF= 0. 209 and 4.4, one at a time, to the measurement o d Eq. 20- 2 system at the calibration valve assembly. Fc (Make no adjustments to the measurement system until after the drift checks are where: made). Record the analyzers’ responses on a FO=Fuel factor based on the ratio of oxygen form similar to Figure 20–3. If the drift val- volume to the ultimate CO2 volume pro- ues exceed the specified limits, the test run duced by the fuel at zero percent excess preceding the check is considered invalid and air, dimensionless. will be repeated following corrections to the 0.209=Fraction of air that is oxygen, percent/ measurement system. Alternatively, recali- 100. brate the measurement system and recal- Fd=Ratio of the volume of dry effluent gas to culate the measurement data. Report the the gross calorific value of the fuel from test results based on both the initial calibra- Method 19, dsm3/J (dscf/106 Btu). tion and the recalibration data. Fc=Ratio of the volume of carbon dioxide produced to the gross calorific value of 6.3 SO2 Measurement. This test is con- ducted only at the 100 percent peak load con- the fuel from Method 19, dsm3/J (dscf6 Btu). dition. Determine SO2 using Method 6, or equivalent, during the test. Select a mini- 7.2.2. Calculate the CO2 correction factor mum of six total points from those required for correcting measurement data to 15 per- cent oxygen, as follows: for the NOx measurements; use two points for each sample run. The sample time at 5. 9 each point shall be at least 10 minutes. Aver- X = Eq. 20- 3 age the diluent readings taken during the CO2 F NOx test runs at sample points corresponding o to the SO2 traverse points (see Section 6.2.2) where: and use this average diluent concentration XCO2=CO2 Correction factor, percent. to correct the integrated SO2 concentration 5.9=20.9 percent O2¥15 percent O2, the defined obtained by Method 6 to 15 percent diluent O2 correction value, percent. (see Equation 20–1). 7.3 Correction of Pollutant Concentra- If the applicable regulation allows fuel tions to 15 percent O2. Calculate the NOx and sampling and analysis for fuel sulfur content SO2 gas concentrations adjusted to 15 per- to demonstrate compliance with sulfur emis- cent O2 using Equation 20–4 or 20–5, as appro- sion unit, emission sampling with Method 6 priate. The correction to 15 percent O2 is is not required, provided the fuel sulfur con- very sensitive to the accuracy of the O2 or tent meets the limits of the regulation. CO2 concentration measurement. At the 7. Emission Calculations level of the analyzer drift specified in Sec- 7.1 Moisture Correction. Measurement tion 3, the O2 or CO2 correction can exceed 5 data used in most of these calculations must percent at the concentration levels expected be on a dry basis. If measurements must be in gas turbine exhaust gases. Therefore, O2 corrected to dry conditions, use the follow- or CO2 analyzer stability and careful calibra- ing equation: tion are necessary.
924
VerDate 11
7.3.1 Correction of Pollutant Concentra- E=Mass emission rate of pollutant, ng/J (lb/ 6 tion Using O2 Concentration. Calculate the 10 Btu). O2 corrected pollutant concentration, as fol- 7.5.2 Calculation of Emission Rate Using lows: Carbon Dioxide Correction. The CO2 con- centration and the pollutant concentration 5. 9 CC= Eq. 20- 4 may be on either a dry basis or a wet basis, adj d 20.% 9 − O but both concentrations must be on the same 2 basis for the calculations. Calculate the pol- where: lutant emission rate using Equation 20–7 or Cadj=Pollutant concentration corrected to 15 20–8: percent O2 ppm. Cd=Pollutant concentration measured, dry = 100 basis, ppm. ECFd c Eq. 20- 7 %O2=Measured O2 concentration dry basis, %CO2 percent. 7.3.2 Correction of Pollutant Concentra- = 100 tion Using CO2 Concentration. Calculate the ECFw c Eq. 20- 8 CO2 corrected pollutant concentration, as %CO2w follows: where: X Cw=Pollutant concentration measured on a = CO2 3 CCadj d Eq. 20- 5 moist sample basis, ng/sm (lb/scf). %CO2 %CO2w=Measured CO2 concentration meas- ured on a moist sample basis, percent. where: 8. Bibliography %CO2=Measured CO2 concentration meas- ured, dry basis, percent. 1. Curtis, F. A Method for Analyzing NOx Cylinder Gases-Specific Ion Electrode Proce- 7.4 Average Adjusted NOx Concentration. dure, Monograph available from Emission Calculate the average adjusted NOx con- centration by summing the adjusted values Measurement Laboratory, ESED, Research for each sample point and dividing by the Triangle Park, NC 27711, October 1978. number of points for each run. 2. Sigsby, John E., F. M. Black, T. A. Bellar, and D. L. Klosterman. Chem- 7.5 NOx and SO2 Emission Rate Calcula- iluminescent Method for Analysis of Nitro- tions. The emission rates for NOx and SO2 in units of pollutant mass per quantity of heat gen Compounds in Mobile Source Emissions input can be calculated using the pollutant (NO, NO2, and NH3 ). ‘‘Environmental Science and diluent concentrations and fuel-specific and Technology,’’ 7:51–54. January 1973. F-factors based on the fuel combustion char- 3. Shigehara, R.T., R.M. Neulicht, and W.S. acteristics. The measured concentrations of Smith. Validating Orsat Analysis Data from pollutant in units of parts per million by vol- Fossil Fuel-Fired Units. Emission Measure- ume (ppm) must be converted to mass per ment Branch, Emission Standards and Engi- unit volume concentration units for these neering Division, Office of Air Quality Plan- calculations. Use the following table for such ning and Standards, U.S. Environmental conversions: Protection Agency, Research Triangle Park, NC 27711. June 1975. CONVERSION FACTORS FOR CONCENTRATION METHOD 21—DETERMINATION OF VOLATILE From To Multiply by ORGANIC COMPOUNDS LEAKS 1. Applicability and Principle g/sm3 ...... ng/sm3 ...... 109 mg/sm3 ...... ng/sm3 ...... 106 1.1 Applicability. This method applies to lb/scf ...... ng/sm3 ...... 1.602 x 1013 the determination of volatile organic com- 3 6 ppm (SO2) ...... ng/sm ...... 2.660 x 10 pound (VOC) leaks from process equipment. 3 6 ppm (NOx) ...... ng/sm ...... 1.912 x 10 These sources include, but are not limited 7 ppm (SO2) ...... lb/scf ...... 1.660 x 10¥ to, valves, flanges and other connections, 7 ppm (NOx) ...... lb/scf ...... 1.194 x 10¥ pumps and compressors, pressure relief de- vices, process drains, open-ended valves, 7.5.1 Calculation of Emission Rate Using pump and compressor seal system degassing Oxygen Correction. Both the O2 concentra- vents, accumulator vessel vents, agitator tion and the pollutant concentration must seals, and access door seals. be on a dry basis. Calculate the pollutant 1.2 Principle. A portable instrument is emission rate, as follows: used to detect VOC leaks from individual 20. 9 sources. The instrument detector type is not ECF= Eq. 20- 6 specified, but it must meet the specifications d d 20.% 9 − O and performance criteria contained in Sec- 2 tion 3. A leak definition concentration based where: on a reference compound is specified in each
925
VerDate 11
applicable regulation. This procedure is in- calibration gas that is used for calibration, tended to locate and classify leaks only, and shall encompass the leak definition con- is not to be used as a direct measure of mass centration specified in the regulation. A di- emission rates from individual sources. lution probe assembly may be used to bring 2. Definitions the VOC concentration within both ranges; 2.1 Leak Definition Concentration. The however, the specifications for instrument local VOC concentration at the surface of a response time and sample probe diameter leak source that indicates that a VOC emis- shall still be met. sion (leak) is present. The leak definition is c. The scale of the instrument meter shall ± an instrument meter reading based on a ref- be readable to 2.5 percent of the specified erence compound. leak definition concentration when perform- 2.2 Reference Compound. The VOC species ing a no detectable emission survey. selected as an instrument calibration basis d. The instrument shall be equipped with for specification of the leak definition con- an electrically driven pump to insure that a centration. (For example: If a leak definition sample is provided to the detector at a con- concentration is 10,000 ppmv as methane, stant flow rate. The nominal sample flow then any source emission that results in a rate, as measured at the sample probe tip, local concentration that yields a meter read- shall be 0.10 to 3.0 liters per minute when the ing of 10,000 on an instrument calibrated probe is fitted with a glass wool plug or filter with methane would be classified as a leak. that may be used to prevent plugging of the In this example, the leak definition is 10,000 instrument. ppmv, and the reference compound is meth- e. The instrument shall be intrinsically ane.) safe as defined by the applicable U.S.A. 2.3 Calibration Gas. The VOC compound standards (e.g., National Electric Code by used to adjust the instrument meter reading the National Fire Prevention Association) to a known value. The calibration gas is usu- for operation in any explosive atmospheres ally the reference compound at a concentra- that may be encountered in its use. The in- tion approximately equal to the leak defini- strument shall, at a minimum, be intrinsi- tion concentration. cally safe for Class 1, Division 1 conditions, 2.4 No Detectable Emission. Any VOC and Class 2, Division 1 conditions, as defined concentration at a potential leak source (ad- by the example Code. The instrument shall justed for local VOC ambient concentration) not be operated with any safety device, such that is less than a value corresponding to the as an exhaust flame arrestor, removed. instrument readability specification of sec- f. The instrument shall be equipped with a tion 3.1.1(c) indicates that a leak is not probe or probe extension for sampling not to present. exceed 1⁄4 in. in outside diameter, with a sin- 2.5 Response Factor. The ratio of the gle end opening for admission of sample. known concentration of a VOC compound to 3.1.2 Performance Criteria. the observed meter reading when measured (a) The instrument response factors for using an instrument calibrated with the ref- each of the VOC to be measured shall be less erence compound specified in the application than 10. When no instrument is available regulation. that meets this specification when cali- 2.6 Calibration Precision. The degree of brated with the reference VOC specified in agreement between measurements of the the applicable regulation, the available in- same known value, expressed as the relative strument may be calibrated with one of the percentage of the average difference between VOC to be measured, or any other VOC, so the meter readings and the known con- long as the instrument then has a response centration to the known concentration. factor of less than 10 for each of the VOC to 2.7 Response Time. The time interval be measured. from a step change in VOC concentration at (b) The instrument response time shall be the input of the sampling system to the time equal to or less than 30 seconds. The instru- at which 90 percent of the corresponding ment pump, dilution probe (if any), sample final value is reached as displayed on the in- probe, and probe filter, that will be used dur- strument readout meter. ing testing, shall all be in place during the 3. Apparatus response time determination. 3.1 Monitoring Instrument. c. The calibration precision must be equal 3.1.1 Specifications. to or less than 10 percent of the calibration a. The VOC instrument detector shall re- gas value. spond to the compounds being processed. De- d. The evaluation procedure for each pa- tector types which may meet this require- rameter is given in Section 4.4. ment include, but are not limited to, cata- 3.1.3 Performance Evaluation Require- lytic oxidation, flame ionization, infrared ments. absorption, and photoionization. a. A response factor must be determined b. Both the linear response range and the for each compound that is to be measured, measurable range of the instrument for each either by testing or from reference sources. of the VOC to be measured, and for the VOC The response factor tests are required before
926
VerDate 11
placing the analyzer into service, but do not instrument readout. If an increased meter have to be repeated at subsequent intervals. reading is observed, slowly sample the inter- b. The calibration precision test must be face where leakage is indicated until the completed prior to placing the analyzer into maximum meter reading is obtained. Leave service, and at subsequent 3-month intervals the probe inlet at this maximum reading lo- or at the next use whichever is later. cation for approximately two times the in- c. The response time test is required prior strument response time. If the maximum ob- to placing the instrument into service. If a served meter reading is greater than the leak modification to the sample pumping system definition in the applicable regulation, or flow configuration is made that would record and report the results as specified in change the response time, a new test is re- the regulation reporting requirements. Ex- quired prior to further use. amples of the application of this general 3.2 Calibration Gases. The monitoring in- technique to specific equipment types are: strument is calibrated in terms of parts per a. Valves—The most common source of million by volume (ppmv) of the reference leaks from valves is at the seal between the compound specified in the applicable regula- stem and housing. Place the probe at the tion. The calibration gases required for mon- interface where the stem exits the packing itoring and instrument performance evalua- gland and sample the stem circumference. tion are a zero gas (air, less than 10 ppmv Also, place the probe at the interface of the VOC) and a calibration gas in air mixture ap- packing gland take-up flange seat and sam- proximately equal to the leak definition ple the periphery. In addition, survey valve specified in the regulation. If cylinder cali- housings of multipart assembly at the sur- bration gas mixtures are used, they must be face of all interfaces where a leak could analyzed and certified by the manufacturer occur. to be within ±2 percent accuracy, and a shelf b. Flanges and Other Connections—For life must be specified. Cylinder standards welded flanges, place the probe at the outer must be either reanalyzed or replaced at the edge of the flange-gasket interface and sam- end of the specified shelf life. Alternately, ple the circumference of the flange. Sample calibration gases may be prepared by the other types of nonpermanent joints (such as user according to any accepted gaseous threaded connections) with a similar tra- standards preparation procedure that will verse. yield a mixture accurate to within ±2 per- cent. Prepared standards must be replaced c. Pumps and Compressors—Conduct a cir- each day of use unless it can be dem- cumferential traverse at the outer surface of onstrated that degradation does not occur the pump or compressor shaft and seal inter- during storage. face. If the source is a rotating shaft, posi- Calibrations may be performed using a tion the probe inlet within 1 cm of the shaft- compound other than the reference com- seal interface for the survey. If the housing pound if a conversion factor is determined configuration prevents a complete traverse for that alternative compound so that the re- of the shaft periphery, sample all accessible sulting meter readings during source surveys portions. Sample all other joints on the can be converted to reference compound re- pump or compressor housing where leakage sults. could occur. d. Pressure Relief Devices—The configura- 4. Procedures tion of most pressure relief devices prevents 4.1 Pretest Preparations. Perform the in- sampling at the sealing seat interface. For strument evaluation procedures given in those devices equipped with an enclosed ex- Section 4.4 if the evaluation requirements of tension, or horn, place the probe inlet at ap- Section 3.1.3 have not been met. proximately the center of the exhaust area 4.2 Calibration Procedures. Assemble and to the atmosphere. start up the VOC analyzer according to the e. Process Drains—For open drains, place manufacturer’s instructions. After the ap- the probe inlet at approximately the center propriate warmup period and zero internal of the area open to the atmosphere. For cov- calibration procedure, introduce the calibra- ered drains, place the probe at the surface of tion gas into the instrument sample probe. the cover interface and conduct a peripheral Adjust the instrument meter readout to cor- traverse. respond to the calibration gas value. f. Open-Ended Lines or Valves—Place the NOTE: If the meter readout cannot be ad- probe inlet at approximately the center of justed to the proper value, a malfunction of the opening to the atmosphere. the analyzer is indicated and corrective ac- g. Seal System Degassing Vents and Accu- tions are necessary before use. mulator Vents—Place the probe inlet at ap- 4.3 Individual Source Surveys. proximately the center of the opening to the 4.3.1 Type I—Leak Definition Based on atmosphere. Concentration. Place the probe inlet at the h. Access Door Seals—Place the probe inlet surface of the component interface where at the surface of the door seal interface and leakage could occur. Move the probe along conduct a peripheral traverse. the interface periphery while observing the 4.3.2 Type II—‘‘No Detectable Emission’’.
927
VerDate 11
Determine the local ambient concentra- solution. Observe the potential leak sites to tion around the source by moving the probe determine if any bubbles are formed. If no inlet randomly upwind and downwind at a bubbles are observed, the source is presumed distance of one to two meters from the to have no detectable emissions or leaks as source. If an interference exists with this de- applicable. If any bubbles are observed, the termination due to a nearby emission or instrument techniques of 4.3.1 or 4.3.2 shall leak, the local ambient concentration may be used to determine if a leak exists, or if be determined at distances closer to the the source has detectable emissions, as ap- source, but in no case shall the distance be plicable. less than 25 centimeters. Then move the 4.4 Instrument Evaluation Procedures. At probe inlet to the surface of the source and the beginning of the instrument performance determine the concentration described in evaluation test, assemble and start up the 4.3.1. The difference between these con- instrument according to the manufacturer’s centrations determines whether there are no instructions for recommended warmup pe- detectable emissions. Record and report the riod and preliminary adjustments. results as specified by the regulation. 4.4.1 Response Factor. Calibrate the in- For those cases where the regulation re- strument with the reference compound as quires a specific device installation, or that specified in the applicable regulation. For specified vents be ducted or piped to a con- each organic species that is to be measured trol device, the existence of these conditions during individual source surveys, obtain or shall be visually confirmed. When the regu- prepare a known standard in air at a con- lation also requires that no detectable emis- centration of approximately 80 percent of the sions exist, visual observations and sampling applicable leak definition unless limited by surveys are required. Examples of this tech- volatility or explosivity. In these cases, pre- nique are: pare a standard at 90 percent of the satura- (a) Pump or Compressor Seals—If applica- tion concentration, or 70 percent of the lower ble, determine the type of shaft seal. explosive limit, respectively. Introduce this Preform a survey of the local area ambient mixture to the analyzer and record the ob- VOC concentration and determine if detect- served meter reading. Introduce zero air able emissions exist as described above. until a stable reading is obtained. Make a (b) Seal System Degassing Vents, Accumu- total of three measurements by alternating lator Vessel Vents, Pressure Relief Devices— between the known mixture and zero air. If applicable, observe whether or not the ap- Calculate the response factor for each repeti- plicable ducting or piping exists. Also, deter- tion and the average response factor. mine if any sources exist in the ducting or Alternatively, if response factors have piping where emissions could occur prior to been published for the compounds of interest the control device. If the required ducting or for the instrument or detector type, the re- piping exists and there are no sources where sponse factor determination is not required, the emissions could be vented to the atmos- and existing results may be referenced. Ex- phere prior to the control device, then it is amples of published response factors for presumed that no detectable emissions are flame ionization and catalytic oxidation de- present. If there are sources in the ducting tectors are included in Bibliography. or piping where emissions could be vented or 4.4.2 Calibration Precision. Make a total of sources where leaks could occur, the sam- three measurements by alternately using pling surveys described in this paragraph zero gas and the specified calibration gas. shall be used to determine if detectable Record the meter readings. Calculate the av- emissions exist. erage algebraic difference between the meter 4.3.3 Alternative Screening Procedure. A readings and the known value. Divide this screening procedure based on the formation average difference by the known calibration of bubbles in a soap solution that is sprayed value and mutiply by 100 to express the re- on a potential leak source may be used for sulting calibration precision as a percentage. those sources that do not have continuously 4.4.3 Response Time. Introduce zero gas moving parts, that do not have surface tem- into the instrument sample probe. When the peratures greater than the boiling point or meter reading has stabilized, switch quickly less than the freezing point of the soap solu- to the specified calibration gas. Measure the tion, that do not have open areas to the at- time from switching to when 90 percent of mosphere that the soap solution cannot the final stable reading is attained. Perform bridge, or that do not exhibit evidence of liq- this test sequence three times and record the uid leakage. Sources that have these condi- results. Calculate the average response time. tions present must be surveyed using the in- strument techniques of 4.3.1 or 4.3.2. 5. Bibliography Spray a soap solution over all potential 1. DuBose, D.A., and G.E. Harris. Re- leak sources. The soap solution may be a sponse Factors of VOC Analyzers at a Meter commercially available leak detection solu- Reading of 10,000 ppmv for Selected Organic tion or may be prepared using concentrated Compounds. U.S. Environmental Protection detergent and water. A pressure sprayer or a Agency, Research Triangle Park, NC. Publi- squeeze bottle may be used to dispense the cation No. EPA 600/2–81–051. September 1981.
928
VerDate 11
2. Brown, G.E., et al. Response Factors of flares are visually determined by an observer VOC Analyzers Calibrated with Methane for without the aid of instruments. Selected Organic Compounds. U.S. Environ- 3. Definitions mental Protection Agency, Research Tri- angle Park, NC. Publication No. EPA 600/2– 3.1 Emission Frequency. Percentage of 81–022. May 1981. time that emissions are visible during the 3. DuBose, D.A., et al. Response of Port- observation period. able VOC Analyzers to Chemical Mixtures. 3.2 Emission Time. Accumulated amount U.S. Environmental Protection Agency, Re- of time that emissions are visible during the search Triangle Park, NC. Publication No. observation period. EPA 600/2–81–110. September 1981. 3.3 Fugitive Emissions. Pollutant gen- erated by an affected facility which is not METHOD 22—VISUAL DETERMINATION OF FUGI- collected by a capture system and is released TIVE EMISSIONS FROM MATERIAL SOURCES to the atmosphere. AND SMOKE EMISSIONS FROM FLARES 3.4 Smoke Emissions. Pollutant generated 1. Introduction by combustion in a flare and occurring im- This method involves the visual deter- mediately downstream of the flame. Smoke mination of fugitive emissions, i.e., emis- occurring within the flame, but not down- sions not emitted directly from a process stream of the flame, is not considered a stack or duct. Fugitive emissions include smoke emission. emissions that (1) escape capture by process 3.5 Observation Period. Accumulated time equipment exhaust hoods; (2) are emitted period during which observations are con- during material transfer; (3) are emitted ducted, not to be less than the period speci- from buildings housing material processing fied in the applicable regulation. or handling equipment; and (4) are emitted 4. Equipment directly from process equipment. This meth- od is used also to determine visible smoke 4.1 Stopwatches. Accumulative type with emissions from flares used for combustion of unit divisions of at least 0.5 seconds; two re- waste process materials. quired. This method determines the amount of 4.2 Light Meter. Light meter capable of time that any visible emissions occur during measuring illuminance in the 50- to 200-lux the observation period, i.e., the accumulated range; required for indoor observations only. emission time. This method does not require 5. Procedure that the opacity of emissions be determined. Since this procedure requires only the deter- 5.1 Position. Survey the affected facility mination of whether a visible emission oc- or building or structure housing the process curs and does not require the determination to be observed and determine the locations of opacity levels, observer certification ac- of potential emissions. If the affected facil- cording to the procedures of Method 9 are ity is located inside a building, determine an not required. However, it is necessary that observation location that is consistent with the observer is educated on the general pro- the requirements of the applicable regula- cedures for determining the presence of visi- tion (i.e., outside observation of emissions ble emissions. As a minimum, the observer escaping the building/structure or inside ob- must be trained and knowledgeable regard- servation of emissions directly emitted from ing the effects on the visibility of emissions the affected facility process unit). Then se- caused by background contrast, ambient lect a position that enables a clear view of lighting, observer position relative to light- the potential emission point(s) of the af- ing, wind, and the presence of uncombined fected facility or of the building or structure water (condensing water vapor). This train- housing the affected facility, as appropriate ing is to be obtained from written materials for the applicable subpart. A position at found in Citations 1 and 2 of Bibliography or least 15 feet, but not more than 0.25 miles, from the lecture portion of the Method 9 cer- from the emission source is recommended. tification course. For outdoor locations, select a position 2. Applicability and Principle where the sun is not directly in the observ- 2.1 Applicability. This method applies to er’s eyes. the determination of the frequency of fugi- 5.2 Field Records. tive emissions from stationary sources (lo- 5.2.1 Outdoor Location. Record the follow- cated indoors or outdoors) when specified as ing information on the field data sheet (Fig- the test method for determining compliance ure 22–1): company name, industry, process with new source performance standards. unit, observer’s name, observer’s affiliation, This method also is applicable for the de- and date. Record also the estimated wind termination of the frequency of visible speed, wind direction, and sky condition. smoke emissions from flares. Sketch the process unit being observed and 2.2 Principle. Fugitive emissions produced note the observer location relative to the during material processing, handling, and source and the sun. Indicate the potential transfer operations or smoke emissions from and actual emission points on the sketch.
929
VerDate 11
5.2.2 Indoor Location. Record the follow- observed since 6 minutes is 10 percent of an ing information on the field data sheet (Fig- hour. In any case, the observation period ure 22–2): company name, industry, process shall not be less than 6 minutes in duration. unit, observer’s name, observer’s affiliation, In some cases, the process operation may be and date. Record as appropriate the type, lo- intermittent or cyclic. In such cases, it may cation, and intensity of lighting on the data be convenient for the observation period to sheet. Sketch the process unit being ob- coincide with the length of the process cycle. served and note observer location relative to 5.4.2 Observer Rest Breaks. Do not ob- the source. Indicate the potential and actual serve emissions continuously for a period of fugitive emission points on the sketch. more than 15 to 20 minutes without taking a 5.3 Indoor Lighting Requirements. For in- rest break. For sources requiring observation door locations, use a light meter to measure periods of greater than 20 minutes, the ob- the level of illumination at a location as server shall take a break of not less than 5 close to the emission source(s) as is feasible. minutes and not more than 10 minutes after An illumination of greater than 100 lux (10 every 15 to 20 minutes of observation. If con- foot candles) is considered necessary for tinuous observations are desired for extended proper application of this method. time periods, two observers can alternate be- 5.4 Observations. Record the clock time tween making observations and taking when observations begin. Use one stopwatch breaks. to monitor the duration of the observation 5.4.3 Visual Interference. Occasionally, fu- period; start this stopwatch when the obser- gitive emissions from sources other than the vation period begins. If the observation pe- affected facility (e.g., road dust) may pre- riod is divided into two or more segments by vent a clear view of the affected facility. process shutdowns or observer rest breaks, This may particularly be a problem during stop the stopwatch when a break begins and periods of high wind. If the view of the po- restart it without resetting when the break tential emission points is obscured to such a ends. Stop the stopwatch at the end of the degree that the observer questions the valid- observation period. The accumulated time ity of continuing observations, then the ob- indicated by this stopwatch is the duration servations are terminated, and the observer of the observation period. When the observa- clearly notes this fact on the data form. tion period is completed, record the clock time. 5.5 Recording Observations. Record the During the observation period, continously accumulated time of the observation period watch the emission source. Upon observing on the data sheet as the observation period an emission (condensed water vapor is not duration. Record the accumulated time considered an emission), start the second ac- emissions were observed on the data sheet as cumulative stopwatch; stop the watch when the emission time. Record the clock time the the emission stops. Continue this procedure observation period began and ended, as well for the entire observation period. The accu- as the clock time any observer breaks began mulated elapsed time on this stopwatch is and ended. the total time emissions were visible during 6. Calculations the observation period, i.e., the emission If the applicable subpart requires that the time. emission rate be expressed as an emission 5.4.1 Observation Period. Choose an obser- frequency (in percent), determine this value vation period of sufficient length to meet the as follows: Divide the accumulated emission requirements for determining compliance time (in seconds) by the duration of the ob- with the emission regulation in the applica- servation period (in seconds) or by any mini- ble subpart. When the length of the observa- mum observation period required in the ap- tion period is specifically stated in the appli- plicable subpart, if the acutal observation cable subpart, it may not be necessary to ob- period is less than the required period and serve the source for this entire period if the multiply this quotient by 100. emission time required to indicate non- compliance (based on the specified observa- 7. Bibliography tion period) is observed in a shorter time pe- 1. Missan, Robert and Arnold Stein. riod. In other words, if the regulation pro- Guidelines for Evaluation of Visible Emis- hibits emissions for more than 6 minutes in sions Certification, Field Procedures, Legal any hour, then observations may (optional) Aspects, and Background Material. EPA be stopped after an emission time of 6 min- Publication No. EPA–340/1–75–007. April 1975 utes is exceeded. Similarly, when the regula- 2. Wohlschlegel, P. and D. E. Wagoner. tion is expressed as an emission frequency Guideline for Development of a Quality As- and the regulation prohibits emissions for surance Program: Volume IX—Visual Deter- greater than 10 percent of the time in any mination of Opacity Emissions From Sta- hour, then observations may (optional) be tionary Sources. EPA Publication No. EPA– terminated after 6 minutes of emissions are 650/4–74–005–i. November 1975.
930
VerDate 11
931
VerDate 11
METHOD 23—DETERMINATION OF POLY- in the sample probe, on a glass fiber filter, CHLORINATED DIBENZO-P-DIOXINS AND POLY- and on a packed column of adsorbent mate- CHLORINATED DIBENZOFURANS FROM STA- rial. The sample cannot be separated into a TIONARY SOURCES particle vapor fraction. The PCDD’s and PCDF’s are extracted from the sample, sepa- 1. Applicability and Principle rated by high resolution gas chroma- 1.1 Applicability. This method is applica- tography, and measured by high resolution ble to the determination of polychlorinated mass spectrometry. dibenzo-p-dioxins (PCDD’s) and polychlor- 2. Apparatus inated dibenzofurans (PCDF’s) from station- ary sources. 2.1 Sampling. A schematic of the sam- 1.2 Principle. A sample is withdrawn from pling train used in this method is shown in the gas stream isokinetically and collected Figure 23–1. Sealing greases may not be used
932
VerDate 11
in assembling the train. The train is iden- od 5 of this appendix with the following addi- tical to that described in section 2.1 of Meth- tions:
2.1.1 Nozzle. The nozzle shall be made of 2.1.2 Sample Transfer Lines. The sample nickel, nickel-plated stainless steel, quartz, transfer lines, if needed, shall be heat traced, or borosilicate glass. heavy walled TFE (1⁄2 in. OD with 1⁄8 in. wall) with connecting fittings that are capable of
933
VerDate 11
forming leak-free, vacuum-tight connections tings shall form leak-free, vacuum tight without using sealing greases. The line shall seals. No sealant greases shall be used in the be as short as possible and must be main- sampling train. A coarse glass frit is in- tained at 120 °C. cluded to retain the adsorbent. 2.1.1 Filter Support. Teflon or Teflon- 2.2 Sample Recovery. coated wire. 2.2.1 Fitting Caps. Ground glass, Teflon 2.1.2 Condenser. Glass, coil type with tape, or aluminum foil (Section 2.2.6) to cap compatible fittings. A schematic diagram is off the sample exposed sections of the train. shown in Figure 23–2. 2.2.2 Wash Bottles. Teflon, 500-ml. 2.1.3 Water Bath. Thermostatically con- 2.2.3 Probe-Liner Probe-Nozzle, and Fil- trolled to maintain the gas temperature ter-Holder Brushes. Inert bristle brushes exiting the condenser at <20 °C (68 °F). with precleaned stainless steel or Teflon 2.1.4 Adsorbent Module. Glass container handles. The probe brush shall have exten- to hold the solid adsorbent. A shematic dia- sions of stainless steel or Teflon, at least as gram is shown in Figure 23–2. Other physical long as the probe. The brushes shall be prop- configurations of the resin trap/condenser as- erly sized and shaped to brush out the nozzle, sembly are acceptable. The connecting fit- probe liner, and transfer line, if used.
934
VerDate 11
935
VerDate 11
2.2.4 Filter Storage Container. Sealed fil- 3.1.1 Filters. Glass fiber filters, without ter holder, wide-mouth amber glass jar with organic binder, exhibiting at least 99.95 per- Teflon-lined cap, or glass petri dish. cent efficiency (<0.05 percent penetration) on 2.2.5 Balance. Triple beam. 0.3-micron dioctyl phthalate smoke par- 2.2.6 Aluminum Foil. Heavy duty, hexane- ticles. The filter efficiency test shall be con- rinsed. ducted in accordance with ASTM Standard 2.2.7 Metal Storage Container. Air tight Method D 2986–71 (Reapproved 1978) (incor- container to store silica gel. porated by reference—see § 60.17). 2.2.8 Graduated Cylinder. Glass, 250-ml 3.1.1.1 Precleaning. All filters shall be with 2-ml graduation. cleaned before their initial use. Place a glass 2.2.9 Glass Sample Storage Container. extraction thimble and 1 g of silica gel and a Amber glass bottle for sample glassware plug of glass wool into a Soxhlet apparatus, washes, 500- or 1000-ml, with leak free Teflon- charge the apparatus with toluene, and lined caps. reflux for a minimum of 3 hours. Remove the 2.3 Analysis. toluene and discard it, but retain the silica 2.3.1 Sample Container. 125- and 250-ml gel. Place no more than 50 filters in the flint glass bottles with Teflon-lined caps. thimble onto the silica gel bed and top with 2.3.2 Test Tube. Glass. the cleaned glass wool. Charge the Soxhlet 2.3.3 Soxhlet Extraction Apparatus. Capa- with toluene and reflux for 16 hours. After ble of holding 43 x 123 mm extraction thim- extraction, allow the Soxhlet to cool, re- bles. move the filters, and dry them under a clean 2.3.4 Extraction Thimble. Glass, pre- N2 stream. Store the filters in a glass petri cleaned cellulosic, or glass fiber. dish sealed with Teflon tape. 2.3.5 Pasteur Pipettes. For preparing liq- 3.1.2 Adsorbent Resin. Amberlite XAD–2 uid chromatographic columns. resin. Thoroughly cleaned before initial use. 2.3.6 Reacti-vials. Amber glass, 2-ml, 3.1.2.1 Cleaning Procedure. This procedure silanized prior to use. may be carried out in a giant Soxhlet extrac- 2.3.7 Rotary Evaporator. Buchi/Brinkman tor. An all-glass filter thimble containing an RF–121 or equivalent. extra-course frit is used for extraction of 2.3.8 Nitrogen Evaporative Concentrator. XAD–2. The frit is recessed 10–15 mm above a N-Evap Analytical Evaporator Model III or crenelated ring at the bottom of the thimble equivalent. to facilitate drainage. The resin must be 2.3.9 Separatory Funnels. Glass, 2-liter. carefully retained in the extractor cup with 2.3.10 Gas Chromatograph. Consisting of a glass wool plug and a stainless steel ring the following components: because it floats on methylene chloride. This 2.3.10.1 Oven. Capable of maintaining the process involves sequential extraction in the separation column at the proper operating following order. temperature ± °C and performing pro- grammed increases in temperature at rates Solvent Procedure of at least 40 °C/min. Water ...... Initial rinse: Place resin in a 2.3.10.2 Temperature Gauge. To monitor beaker, rinse once with column oven, detector, and exhaust tempera- water, and discard. Fill with tures ±1 °C. water a second time, let 2.3.10.3 Flow System. Gas metering sys- stand overnight, and dis- tem to measure sample, fuel, combustion card. gas, and carrier gas flows. Water ...... Extract with water for 8 2.3.10.4 Capillary Columns. A fused silica hours. × Methanol ...... Extract for 22 hours. column, 60 0.25 mm inside diameter (ID), Methylene Chloride ...... Extract for 22 hours. coated with DB–5 and a fused silica column, Toluene ...... Extract for 22 hours. 30 m × 0.25 mm ID coated with DB–225. Other column systems may be used provided that 3.1.2.2 Drying. the user is able to demonstrate using cali- 3.1.2.2.1 Drying Column. Pyrex pipe, 10.2 bration and performance checks that the col- cm ID by 0.6 m long, with suitable retainers. umn system is able to meet the specifica- 3.1.2.2.2 Procedure. The adsorbent must be tions of section 6.1.2.2. dried with clean inert gas. Liquid nitrogen 2.3.11 Mass Spectrometer. Capable of rou- from a standard commercial liquid nitrogen tine operation at a resolution of 1:10000 with cylinder has proven to be a reliable source of a stability of ±5 ppm. large volumes of gas free from organic con- 2.3.12 Data System. Compatible with the taminants. Connect the liquid nitrogen cyl- mass spectrometer and capable of monitor- inder to the column by a length of cleaned ing at least five groups of 25 ions. copper tubing, 0.95 cm ID, coiled to pass 2.3.13 Analytical Balance. To measure through a heat source. A convenient heat within 0.1 mg. source is a water-bath heated from a steam line. The final nitrogen temperature should 3. Reagents only be warm to the touch and not over 40 3.1 Sampling. °C. Continue flowing nitrogen through the
936
VerDate 11
adsorbent until all the residual solvent is re- 3.3 Analysis. moved. The flow rate should be sufficient to 3.3.1 Potassium Hydroxide. ACS grade, 2- gently agitate the particles but not so exces- percent (weight/volume) in water. sive as the cause the particles to fracture. 3.3.2 Sodium Sulfate. Granulated, reagent 3.1.2.3 Quality Control Check. The adsorb- grade. Purify prior to use by rinsing with ent must be checked for residual toluene. methylene chloride and oven drying. Store 3.1.2.3.1 Extraction. Weigh 1.0 g sample of the cleaned material in a glass container dried resin into a small vial, add 3 ml of tolu- with a Teflon-lined screw cap. ene, cap the vial, and shake it well. 3.3.3 Sulfuric Acid. Reagent grade. 3.1.2.3.2 Analysis. Inject a 2 µl sample of 3.3.4 Sodium Hydroxide. 1.0 N. Weigh 40 g the extract into a gas chromatograph oper- of sodium hydroxide into a 1-liter volumetric ated under the following conditions: flask. Dilute to 1 liter with water.
Column: 6 ft × 1⁄8 in stainless steel contain- 3.3.5 Hexane. Pesticide grade. ing 10 percent OV–101 on 100/120 3.3.6 Methylene Chloride. Pesticide grade. Supelcoport. 3.3.7 Benzene. Pesticide Grade. Carrier Gas: Helium at a rate of 30 ml/min. 3.3.8 Ethyl Acetate. Detector: Flame ionization detector oper- 3.3.9 Methanol. Pesticide Grade. ated at a sensitivity of 4 × 10¥11 A/mV. 3.3.10 Toluene. Pesticide Grade. Injection Port Temperature: 250 °C. 3.3.11 Nonane. Pesticide Grade. Detector Temperature: 305 °C. 3.3.12 Cyclohexane. Pesticide Grade. Oven Temperature: 30 °C for 4 min; pro- 3.3.13 Basic Alumina. Activity grade 1, grammed to rise at 40 °C/min until it 100–200 mesh. Prior to use, activate the alu- reaches 250 °C; return to 30 °C after 17 mina by heating for 16 hours at 130 °C before minutes. use. Store in a desiccator. Pre-activated alu- Compare the results of the analysis to the mina may be purchased from a supplier and results from the reference solution. Prepare may be used as received. the reference solution by injection 2.5 µl of 3.3.14 Silica Gel. Bio-Sil A, 100–200 mesh. methylene chloride into 100 ml of toluene. Prior to use, activate the silica gel by heat- This corresponds to 100 µg of methylene chlo- ing for at least 30 minutes at 180 °C. After ride per g of adsorbent. The maximum ac- cooling, rinse the silica gel sequentially with ceptable concentration is 1000 µg/g of adsorb- methanol and methylene chloride. Heat the ent. If the adsorbent exceeds this level, dry- rinsed silica gel at 50 °C for 10 minutes, then ing must be continued until the excess meth- increase the temperature gradually to 180 °C ylene chloride is removed. over 25 minutes and maintain it at this tem- 3.1.2.4 Storage. The adsorbent must be perature for 90 minutes. Cool at room tem- used within 4 weeks of cleaning. After clean- perature and store in a glass container with ing, it may be stored in a wide mouth amber a Teflon-lined screw cap. glass container with a Teflon-lined cap or 3.3.15 Silica Gel Impregnated with Sul- placed in one of the glass adsorbent modules furic Acid. Combine 100 g of silica gel with 44 tightly sealed with glass stoppers. If g of concentrated sulfuric acid in a screw precleaned adsorbent is purchased in sealed capped glass bottle and agitate thoroughly. containers, it must be used within 4 weeks Disperse the solids with a stirring rod until after the seal is broken. a uniform mixture is obtained. Store the 3.1.3 Glass Wool. Cleaned by sequential mixture in a glass container with a Teflon- immersion in three aliquots of methylene lined screw cap. chloride, dried in a 110 °C oven, and stored in 3.3.16 Silica Gel Impregnated with Sodium a methylene chloride-washed glass jar with a Hydroxide. Combine 39 g of 1 N sodium hy- Teflon-lined screw cap. droxide with 100 g of silica gel in a screw 3.1.4 Water. Deionized distilled and stored capped glass bottle and agitate thoroughly. in a methylene chloride-rinsed glass con- Disperse solids with a stirring rod until a tainer with a Teflon-lined screw cap. uniform mixture is obtained. Store the mix- 3.1.5 Silica Gel. Indicating type, 6 to 16 ture in glass container with a Teflon-lined mesh. If previously used, dry at 175 °C (350 screw cap. °F) for two hours. New silica gel may be used 3.3.17 Carbon/Celite. Combine 10.7 g of as received. Alternately other types of AX–21 carbon with 124 g of Celite 545 in a 250- desiccants (equivalent or better) may be ml glass bottle with a Teflon-lined screw used, subject to the approval of the Adminis- cap. Agitate the mixture thoroughly until a trator. uniform mixture is obtained. Store in the 3.1.6 Chromic Acid Cleaning Solution. glass container. Dissolve 20 g of sodium dichromate in 15 ml 3.3.18 Nitrogen. Ultra high purity. of water, and then carefully add 400 ml of 3.3.19 Hydrogen. Ultra high purity. concentrated sulfuric acid. 3.3.20 Internal Standard Solution. Prepare 3.2 Sample Recovery. a stock standard solution containing the 3.2.2 Acetone. Pesticide quality. isotopically labelled PCDD’s and PCDF’s at 3.2.2 Methylene Chloride. Pesticide the concentrations shown in Table 1 under qualtity. the heading ‘‘Internal Standards’’ in 10 ml of 3.2.3 Toluene. Pesticide quality. nonane.
937
VerDate 11
3.3.21 Surrogate Standard Solution. Pre- leave the first and fourth impingers empty, pare a stock standard solution containing and transfer approximately 200 to 300 g of the isotopically labelled PCDD’s and PCDF’s preweighed silica gel from its container to at the concentrations shown in Table 1 under the fifth impinger. the heading ‘‘Surrogate Standards’’ in 10 ml 4.1.3.3 Place the silica gel container in a of nonane. clean place for later use in the sample recov- 3.3.22 Recovery Standard Solution. Pre- ery. Alternatively, the weight of the silica pare a stock standard solution containing gel plus impinger may be determined to the the isotopically labelled PCDD’s and PCDF’s nearest 0.5 g and recorded. at the concentrations shown in Table 1 under 4.1.3.4 Assemble the train as shown in Fig- the heading ‘‘Recovery Standards’’ in 10 ml ure 23–1. of nonane. 4.1.3.5 Turn on the adsorbent module and condenser coil recirculating pump and begin 4. Procedure monitoring the adsorbent module gas entry 4.1 Sampling. The complexity of this temperature. Ensure proper sorbent tem- method is such that, in order to obtain reli- perature gas entry temperature before pro- able results, testers should be trained and ceeding and before sampling is initiated. It is experienced with the test procedures. extremely important that the XAD–2 adsorb- ° 4.1.1 Pretest Preparation. ent resin temperature never exceed 50 C be- 4.1.1.1 Cleaning Glassware. All glass com- cause thermal decomposition will occur. ponents of the train upstream of and includ- During testing, the XAD–2 temperature must ° ing the adsorbent module, shall be cleaned as not exceed 20 C for efficient capture of the described in section 3A of the ‘‘Manual of PCDD’s and PCDF’s. Analytical Methods for the Analysis of Pes- 4.1.4 Leak-Check Procedure. Same as Method 5, section 4.1.4. ticides in Human and Environmental Sam- 4.1.5 Sample Train Operation. Same as ples.’’ Special care shall be devoted to the re- Method 5, section 4.1.5. moval of residual silicone grease sealants on 4.2 Sample Recovery. Proper cleanup pro- ground glass connections of used glassware. cedure begins as soon as the probe is re- Any residue shall be removed by soaking the moved from the stack at the end of the sam- glassware for several hours in a chromic acid pling period. Seal the nozzle end of the sam- cleaning solution prior to cleaning as de- pling probe with Teflon tape or aluminum scribed above. foil. 4.1.1.2 Adsorbent Trap. The traps must be When the probe can be safely handled, wipe loaded in a clean area to avoid contamina- off all external particulate matter near the tion. They may not be loaded in the field. tip of the probe. Remove the probe from the Fill a trap with 20 to 40 g of XAD–2. Follow train and close off both ends with aluminum the XAD–2 with glass wool and tightly cap foil. Seal off the inlet to the train with Tef- µ both ends of the trap. Add 100 l of the surro- lon tape, a ground glass cap, or aluminum gate standard solution (section 3.3.21) to foil. each trap. Transfer the probe and impinger assembly 4.1.1.3 Sample Train. It is suggested that to the cleanup area. This area shall be clean all components be maintained according to and enclosed so that the chances of losing or the procedure described in APTD–0576. contaminating the sample are minimized. 4.1.1.4 Silica Gel. Weigh several 200 to 300 Smoking, which could contaminate the sam- g portions of silica gel in an air tight con- ple, shall not be allowed in the cleanup area. tainer to the nearest 0.5 g. Record the total Inspect the train prior to and during dis- weight of the silica gel plus container, on assembly and note any abnormal conditions, each container. As an alternative, the silica e.g., broken filters, colored impinger liquid, gel may be weighed directly in its impinger etc. Treat the samples as follows: or sampling holder just prior to sampling. 4.2.1 Container No. 1. Either seal the filter 4.1.1.5 Filter. Check each filter against holder or carefully remove the filter from light for irregularities and flaws or pinhole the filter holder and place it in its identified leaks. Pack the filters flat in a clean glass container. Use a pair of cleaned tweezers to container. handle the filter. If it is necessary to fold the 4.1.2 Preliminary Determinations. Same filter, do so such that the particulate cake is as section 4.1.2 of Method 5. inside the fold. Carefully transfer to the con- 4.1.3 Preparation of Collection Train. tainer any particulate matter and filter fi- 4.1.3.1 During preparation and assembly of bers which adhere to the filter holder gasket, the sampling train, keep all train openings by using a dry inert bristle brush and a where contamination can enter, sealed until sharp-edged blade. Seal the container. just prior to assembly or until sampling is 4.2.2 Adsorbent Module. Remove the mod- about to begin. ule from the train, tightly cap both ends, NOTE: Do not use sealant grease in assem- label it, cover with aluminum foil, and store bling the train. it on ice for transport to the laboratory. 4.1.3.2 Place approximately 100 ml of 4.2.3 Container No. 2. Quantitatively re- water in the second and third impingers, cover material deposited in the nozzle, probe
938
VerDate 11
transfer lines, the front half of the filter Thoroughly rinse the glass module catching holder, and the cyclone, if used, first, by the rinsings in the beaker containing the brushing while rinsing three times each with thimble. If the resin is wet, effective extrac- acetone and then, by rinsing the probe three tion can be accomplished by loosely packing times with methylene chloride. Collect all the resin in the thimble. Add the XAD–2 the rinses in Container No. 2. glass wool plug into the thimble. Rinse the back half of the filter holder 5.1.4 Container No. 2 (Acetone and Meth- three times with acetone. Rinse the connect- ylene Chloride). Concentrate the sample to a ing line between the filter and the condenser volume of about 1–5 ml using the rotary three times with acetone. Soak the connect- evaporator apparatus, at a temperature of ing line with three separate portions of less than 37 °C. Rinse the sample container methylene chloride for 5 minutes each. If three times with small portions of methyl- using a separate condenser and adsorbent ene chloride and add these to the con- trap, rinse the condenser in the same manner centrated solution and concentrate further as the connecting line. Collect all the rinses to near dryness. This residue contains par- in Container No. 2 and mark the level of the ticulate matter removed in the rinse of the liquid on the container. train probe and nozzle. Add the concentrate 4.2.4 Container No. 3. Repeat the methyl- to the filter and the XAD–2 resin in the ene chloride-rinsing described in Section Soxhlet apparatus described in section 5.1.1. 4.2.3 using toluene as the rinse solvent. Col- 5.1.5 Extraction. Add 100 µl of the internal lect the rinses in Container No. 3 and mark standard solution (Section 3.3.20) to the ex- the level of the liquid on the container. traction thimble containing the contents of 4.2.5 Impinger Water. Measure the liquid the adsorbent cartridge, the contents of Con- in the first three impingers to within ±1 ml tainer No. 1, and the concentrate from sec- by using a graduated cylinder or by weighing tion 5.1.4. Cover the contents of the extrac- it to within ±0.5 g by using a balance. Record tion thimble with the cleaned glass wool the volume or weight of liquid present. This plug to prevent the XAD–2 resin from float- information is required to calculate the ing into the solvent reservoir of the extrac- moisture content of the effluent gas. tor. Place the thimble in the extractor, and Discard the liquid after measuring and re- add the toluene contained in the beaker to cording the volume or weight. the solvent reservoir. Pour additional tolu- 4.2.7 Silica Gel. Note the color of the indi- ene to fill the reservoir approximately 2/3 cating silica gel to determine if it has been full. Add Teflon boiling chips and assemble completely spent and make a mention of its the apparatus. Adjust the heat source to condition. Transfer the silica gel from the cause the extractor to cycle three times per fifth impinger to its original container and hour. Extract the sample for 16 hours. After seal. extraction, allow the Soxhlet to cool. Trans- fer the toluene extract and three 10-ml rinses 5. Analysis to the rotary evaporator. Concentrate the All glassware shall be cleaned as described extract to approximately 10 ml. At this point in section 3A of the ‘‘Manual of Analytical the analyst may choose to split the sample Methods for the Analysis of Pesticides in in half. If so, split the sample, store one half Human and Environmental Samples.’’ All for future use, and analyze the other accord- samples must be extracted within 30 days of ing to the procedures in sections 5.2 and 5.3. collection and analyzed within 45 days of ex- In either case, use a nitrogen evaporative traction. concentrator to reduce the volume of the 5.1 Sample Extraction. sample being analyzed to near dryness. Dis- 5.1.1 Extraction System. Place an extrac- solve the residue in 5 ml of hexane. tion thimble (section 2.3.4), 1 g of silica gel, 5.1.6 Container No. 3 (Toluene Rinse). Add and a plug of glass wool into the Soxhlet ap- 100 µl of the Internal Standard solution (sec- paratus, charge the apparatus with toluene, tion 3.3.2) to the contents of the container. and reflux for a minimum of 3 hours. Remove Concentrate the sample to a volume of about the toluene and discard it, but retain the 1–5 ml using the rotary evaporator apparatus silica gel. Remove the extraction thimble at a temperature of less than 37 °C. Rinse the from the extraction system and place it in a sample container apparatus at a temperature glass beaker to catch the solvent rinses. of less than 37 °C. Rinse the sample container 5.1.2 Container No. 1 (Filter). Transfer the three times with small portions of toluene contents directly to the glass thimble of the and add these to the concentrated solution extraction system and extract them simulta- and concentrate further to near dryness. neously with the XAD–2 resin. Analyze the extract separately according to 5.1.3 Adsorbent Cartridge. Suspend the ad- the procedures in sections 5.2 and 5.3, but sorbent module directly over the extraction concentrate the solution in a rotary evapo- thimble in the beaker (See section 5.1.1). The rator apparatus rather than a nitrogen evap- glass frit of the module should be in the up orative concentrator. position. Using a Teflon squeeze bottle con- 5.2 Sample Cleanup and Fractionation. taining toluene, flush the XAD–2 into the 5.2.1 Silica Gel Column. Pack one end of a thimble onto the bed of cleaned silica gel. glass column, 20 mm x 230 mm, with glass
939
VerDate 11
wool. Add in sequence, 1 g silica gel, 2 g of dibenzofurans are detected in this analysis, sodium hydroxide impregnated silica gel, 1 g then analyze another aliquot of the sample silica gel, 4 g of acid-modified silica gel, and in a separate run, using the DB–225 column 1 g of silica gel. Wash the column with 30 ml to measure the 2,3,7,8 tetra-chloro of hexane and discard it. Add the sample ex- dibenzofuran isomer. Other column systems tract, dissolved in 5 ml of hexane to the col- may be used, provided that the user is able umn with two additional 5-ml rinses. Elute to demonstrate using calibration and per- the column with an additional 90 ml of formance checks that the column system is hexane and retain the entire eluate. Con- able to meet the specifications of section centrate this solution to a volume of about 1 6.1.2.2. ml using the nitrogen evaporative concentra- 5.3.1 Gas Chromatograph Operating Con- tor (section 2.3.7). ditions. 5.2.2 Basic Alumina Column. Shorten a 25- 5.3.1.1 Injector. Configured for capillary ml disposable Pasteur pipette to about 16 ml. column, splitless, 250°C. Pack the lower section with glass wool and 5.3.1.2 Carrier Gas. Helium, 1–2 ml/min. 12 g of basic alumina. Transfer the con- 5.3.1.3 Oven. Initially at 150°C. Raise by at centrated extract from the silica gel column least 40°C/min to 190°C and then at 3°C/min to the top of the basic alumina column and up to 300°C. elute the column sequentially with 120 ml of 5.3.2 High Resolution Mass Spectrometer. 0.5 percent methylene chloride in hexane fol- 5.3.2.1 Resolution. 10000 m/e. lowed by 120 ml of 35 percent methylene 5.3.2.2 Ionization Mode. Electron impact. chloride in hexane. Discard the first 120 ml of 5.3.2.3 Source Temperature 250°C. eluate. Collect the second 120 ml of eluate 5.3.2.4 Monitoring Mode. Selected ion and concentrate it to about 0.5 ml using the monitoring. A list of the various ions to be nitrogen evaporative concentrator. monitored is summarized in Table 3. 5.2.3 AX–21 Carbon/Celite 545 Column. Re- 5.3.2.5 Identification Criteria. The follow- move the botton 0.5 in. from the tip of a 9-ml ing identification criteria shall be used for disposable Pasteur pipette. Insert a glass the characterization of polychlorinated fiber filter disk in the top of the pipette 2.5 dibenzodioxins and dibenzofurans. cm from the constriction. Add sufficient car- 1. The integrated ion-abundance ratio (M/ bon/celite mixture to form a 2 cm column. M+2 or M+2/M+4) shall be within 15 percent Top with a glass wool plug. In some cases of the theoretical value. The acceptable ion- AX–21 carbon fines may wash through the abundance ratio ranges for the identification glass wool plug and enter the sample. This of chlorine-containing compounds are given may be prevented by adding a celite plug to in Table 4. the exit end of the column. Rinse the column 2. The retention time for the analytes in sequence with 2 ml of 50 percent benzene must be within 3 seconds of the correspond- in ethyl acetate, 1 ml of 50 percent methyl- ing 13 C-labeled internal standard, surrogate ene chloride in cyclohexane, and 2 ml of or alternate standard. hexane. Discard these rinses. Transfer the 3. The monitored ions, shown in Table 3 for concentrate in 1 ml of hexane from the basic a given analyte, shall reach their maximum alumina column to the carbon/celite column within 2 seconds of each other. along with 1 ml of hexane rinse. Elute the 4. The identification of specific isomers column sequentially with 2 ml of 50 percent that do not have corresponding 13 C-labeled methylene chloride in hexane and 2 ml of 50 standards is done by comparison of the rel- percent benzene in ethyl acetate and discard ative retention time (RRT) of the analyte to these eluates. Invert the column and elute in the nearest internal standard retention time the reverse direction with 13 ml of toluene. with reference (i.e., within 0.005 RRT units) Collect this eluate. Concentrate the eluate to the comparable RRT’s found in the con- in a rotary evaporator at 50 °C to about 1 ml. tinuing calibration. Transfer the concentrate to a Reacti-vial 5. The signal to noise ratio for all mon- using a toluene rinse and concentrate to a itored ions must be greater than 2.5. volume of 200 µl using a stream of N2. Store 6. The confirmation of 2, 3, 7, 8–TCDD and extracts at room temperature, shielded from 2, 3, 7, 8–TCDF shall satisfy all of the above light, until the analysis is performed. identification criteria. 5.3 Analysis. Analyze the sample with a 7. For the identification of PCDF’s, no sig- gas chromatograph coupled to a mass spec- nal may be found in the corresponding trometer (GC/MS) using the instrumental pa- PCDPE channels. rameters in sections 5.3.1 and 5.3.2. Imme- 5.3.2.6 Quantification. The peak areas for diately prior to analysis, add a 20 µl aliquot the two ions monitored for each analyte are of the Recovery Standard solution from summed to yield the total response for each Table 1 to each sample. A 2 µl aliquot of the analyte. Each internal standard is used to extract is injected into the GC. Sample ex- quantify the indigenous PCDD’s or PCDF’s tracts are first analyzed using the DB–5 cap- in its homologous series. For example, the 13 illary column to determine the concentra- C 12–2,3,7,8-tetra chlorinated dibenzodioxin tion of each isomer of PCDD’s and PCDF’s is used to calculate the concentrations of all (tetra-through octa-). If tetra-chlorinated other tetra chlorinated isomers. Recoveries
940
VerDate 11
of the tetra- and penta- internal standards the internal standards is to quantify the na- 13 are calculated using the C 12–1,2,3,4–TCDD. tive PCDD’s and PCDF’s present in the sam- Recoveries of the hexa- through octa- inter- ple as well as to determine the overall meth- 13 nal standards are calculated using C 12– od efficiency. Recoveries of the internal 1,2,3,7,8,9–HxCDD. Recoveries of the surro- standards must be between 40 to 130 percent gate standards are calculated using the cor- for the tetra-through hexachlorinated com- responding homolog from the internal stand- pounds while the range is 25 to 130 percent ard. for the higher hepta- and octachlorinated homologues. 6. Calibration 7.3 Surrogate Recoveries. The five surro- Same as Method 5 with the following addi- gate compounds in Table 2 are added to the tions. resin in the adsorbent sampling cartridge be- 6.1 GC/MS System. fore the sample is collected. The surrogate 6.1.1 Initial Calibration. Calibrate the GC/ recoveries are measured relative to the in- MS system using the set of five standards ternal standards and are a measure of collec- shown in Table 2. The relative standard devi- tion efficiency. They are not used to measure ation for the mean response factor from each native PCDD’s and PCDF’s. All recoveries of the unlabeled analytes (Table 2) and of the shall be between 70 and 130 percent. Poor re- internal, surrogate, and alternate standards coveries for all the surrogates may be an in- shall be less than or equal to the values in dication of breakthrough in the sampling Table 5. The signal to noise ratio for the GC train. If the recovery of all standards is signal present in every selected ion current below 70 percent, the sampling runs must be profile shall be greater than or equal to 2.5. repeated. As an alternative, the sampling The ion abundance ratios shall be within the control limits in Table 4. runs do not have to be repeated if the final 6.1.2 Daily Performance Check. results are divided by the fraction of surro- 6.1.2.1 Calibration Check. Inject on µl of gate recovery. Poor recoveries of isolated solution Number 3 from Table 2. Calculate surrogate compounds should not be grounds the relative response factor (RRF) for each for rejecting an entire set of the samples. compound and compare each RRF to the cor- 7.4 Toluene QA Rinse. Report the results responding mean RRF obtained during the of the toluene QA rinse separately from the initial calibration. The analyzer perform- total sample catch. Do not add it to the total ance is acceptable if the measured RRF’s for sample. the labeled and unlabeled compounds for the daily run are within the limits of the mean 8. Quality Assurance values shown in Table 5. In addition, the ion- 8.1 Applicability. When the method is abundance ratios shall be within the allow- used to analyze samples to demonstrate com- able control limits shown in Table 4. pliance with a source emission regulation, an 6.1.2.2 Column Separation Check. Inject a audit sample must be analyzed, subject to solution of a mixture of PCDD’s and PCDF’s availability. that documents resolution between 2,3,7,8– 8.2 Audit Procedure. Analyze an audit TCDD and other TCDD isomers. Resolution sample with each set of compliance samples. is defined as a valley between peaks that is less than 25 percent of the lower of the two The audit sample contains tetra through peaks. Identify and record the retention time octa isomers of PCDD and PCDF. Concur- windows for each homologous series. rently, analyze the audit sample and a set of Perform a similar resolution check on the compliance samples in the same manner to confirmation column to document the reso- evaluate the technique of the analyst and lution between 2,3,7,8 TCDF and other TCDF the standards preparation. The same ana- isomers. lyst, analytical reagents, and analytical sys- 6.2 Lock Channels. Set mass spectrometer tem shall be used both for the compliance lock channels as specified in Table 3. Mon- samples and the EPA audit sample. itor the quality control check channels spec- 8.3 Audit Sample Availability. Audit sam- ified in Table 3 to verify instrument stability ples will be supplied only to enforcement during the analysis. agencies for compliance tests. The availabil- ity of audit samples may be obtained by 7. Quality Control writing: Source Test Audit Coordinator (MD– 7.1 Sampling Train Collection Efficiency 77B), Quality Assurance Division, Atmos- Check. Add 100 µl of the surrogate standards pheric Research and Exposure Assessment in Table 1 to the absorbent cartridge of each Laboratory, U.S. Environmental Protection train before collecting the field samples. Agency, Research Triangle Park, NC 27711, 7.2 Internal Standard Percent Recoveries. or by calling the Source Test Audit Coordi- A group of nine carbon labeled PCDD’s and nator (STAC) at (919) 541–7834. The request PCDF’s representing, the tetra-through for the audit sample must be made at least octachlorinated homologues, is added to 30 days prior to the scheduled compliance every sample prior to extraction. The role of sample analysis.
941
VerDate 11
8.4 Audit Results. Calculate the audit sample concentration according to the cal- n A m * = 1 cij ci culation procedure described in the audit in- RRFi ∑ Eq. 23- 1 structions included with the audit sample. n j=1 A*cij m ci Fill in the audit sample concentration and the analyst’s name on the audit response 9.3 Concentration of the PCDD’s and form included with the audit instructions. PCDF’s. Send one copy to the EPA Regional Office or m* A the appropriate enforcement agency and a C = i i Eq. 23 - 2 second copy to the STAC. The EPA Regional i Ai* RRF i V mstd office or the appropriate enforcement agency will report the results of the audit to the 9.4 Recovery Standard Response Factor. laboratory being audited. Include this re- sponse with the results of the compliance = Aci* m rs samples in relevant reports to the EPA Re- RRFrs Eq. 23 - 3 gional Office or the appropriate enforcement Ars m ci * agency. 9.5 Recovery of Internal Standards (R*).
9. Calculations Ai* m rs × Same as Method 5, section 6 with the fol- R*=100%Eq . 23 - 4 lowing additions. Ars RF rs m i * 9.1 Nomenclature. 9.6 Surrogate Compound Response Factor.
Aai=Integrated ion current of the noise at the retention time of the analyte. = Aci* m s RRFs Eq. 23 - 5 A*ci=Integrated ion current of the two ions A m * characteristic of the internal standard i cis ci in the calibration standard. 9.7 Recovery of Surrogate Compounds (R ). Acij=Integrated ion current of the two ions s characteristic of compound i in the jth A m * calibration standard. R =s i ×100%Eq.- 23 6 A*cij=Integrated ion current of the two ions s characteristic of the internal standard i Ai* RRF s m s in the jth calibration standard. 9.8 Minimum Detectable Limit (MDL).
Acsi=Integrated ion current of the two ions characteristic of surrogate compound i in 2. 5 Aai m i * the calibration standard. MDL = Eq. 23 - 7 A* RRF Ai=Integrated ion current of the two ions ci i characteristic of compound i in the sam- 9.9 Total Concentration of PCDD’s and ple. PCDF’s in the Sample. A*i=Integrated ion current of the two ions characteristic of internal standard i in n the sample. = CTr∑ C i Eq. 23 - 8 A =Integrated ion current of the two ions rs i=1 characteristic of the recovery standard. Any PCDD’s or PCDF’s that are reported Asi=Integrated ion current of the two ions characteristic of surrogate compound i in as nondetected (below the MDL) shall be counted as zero for the purpose of calculat- the sample. ing the total concentration of PCDD’s and C =Concentration of PCDD or PCDF i in i PCDF’s in the sample. the sample, pg/M 3. CT=Total concentration of PCDD’s or 10. Bibliography PCDF’s in the sample, pg/M 3. 1. American Society of Mechanical Engi- mci=Mass of compound i in the calibration standard injected into the analyzer, pg. neers. Sampling for the Determination of Chlorinated Organic Compounds in Stack m =Mass of recovery standard in the cali- rs Emissions. Prepared for U.S. Department of bration standard injected into the ana- Energy and U.S. Environmental Protection lyzer, pg. Agency. Washington DC. December 1984. 25 p. msi=Mass of surrogate compound in the 2. American Society of Mechanical Engi- calibration standard, pg. neers. Analytical Procedures to Assay Stack RRFi=Relative response factor. Effluent Samples and Residual Combustion RRFrs=Recovery standard response factor. Products for Polychlorinated Dibenzo-p- RRFs=Surrogate compound response fac- Dioxins (PCDD) and Polychlorinated Diben- tor. zofurans (PCDF). Prepared for the U.S. De- 9.2 Average Relative Response Factor. partment of Energy and U.S. Environmental
942
VerDate 11
Protection Agency. Washington, DC. Decem- TABLE 1ÐCOMPOSITION OF THE SAMPLE FOR- ber 1984. 23 p. TIFICATION AND RECOVERY STANDARDS SOLU- 3. Thompson, J. R. (ed.). Analysis of Pes- TIONSÐContinued ticide Residues in Human and Environ- mental Samples. U.S. Environmental Protec- Concentra- tion Agency. Research Triangle Park, NC. Analyte tion (pg/µl) 1974. 13 4. Triangle Laboratories. Case Study: Anal- C12-1,2,3,7,8-PeCDF ...... 100 13 ysis of Samples for the Presence of Tetra C12-1,2,3,6,7,8-HxCDF ...... 100 13 C -1,2,3,4,6,7,8-HpCDF ...... 100 Through Octachloro-p-Dibenzodioxins and 12 Surrogate Standards: Dibenzofurans. Research Triangle Park, NC. 37 Cl4-2,3,7,8-TCDD ...... 100 13 1988. 26 p. C12-1,2,3,4,7,8-HxCDD ...... 100 13 5. U.S. Environmental Protection Agency. C12-2,3,4,7,8-PeCDF ...... 100 13 Method 8290—The Analysis of Poly- C12-1,2,3,4,7,8-HxCDF ...... 100 13 chlorinated Dibenzo-p-dioxin and Poly- C12-1,2,3,4,7,8,9-HpCDF ...... 100 chlorinated Dibenzofurans by High-Resolu- Recovery Standards: 13 tion Gas Chromotography/High-Resolution C12-1,2,3,4-TCDD ...... 500 13 Mass Spectrometry. In: Test Methods for C12-1,2,3,7,8,9-HxCDD ...... 500 Evaluating Solid Waste. Washington, DC. SW–846. TABLE 2ÐCOMPOSITION OF THE INITIAL CALIBRATION SOLUTIONS TABLE 1ÐCOMPOSITION OF THE SAMPLE FOR- TIFICATION AND RECOVERY STANDARDS SOLU- Concentrations (pg/µL) TIONS Compound Solution No. Concentra- 1 2 3 4 5 Analyte tion (pg/µl) Alternate Standard: 13 Internal Standards: C12-1,2,3,7,8,9- 13 C12-2,3,7,8-TCDD ...... 100 HxCDF ...... 2.5 5 25 250 500 13 C12-1,2,3,7,8-PeCDD ...... 100 Recovery Standards: 13 C -1,2,3,6,7,8-HxCDD ...... 100 13 12 C12-1,2,3,4-TCDD ... 100 100 100 100 100 13 C12-1,2,3,4,6,7,8-HpCDD ...... 100 13 C12-1,2,3,7,8,9- 13 C12-OCDD ...... 100 HxCDD ...... 100 100 100 100 100 13 C12-2,3,7,8-TCDF ...... 100
TABLE 3ÐELEMENTAL COMPOSITIONS AND EXACT MASSES OF THE IONS MONITORED BY HIGH RESOLUTION MASS SPECTROMETRY FOR PCDD'S AND PCDF'S
Descriptor No. Accurate mass Ion type Elemental composition Analyte
2 292.9825 LOCK C7F11 PFK 35 303.9016 MC12H4 Cl4O TCDF 35 37 305.8987 M+2 C12H4 Cl O TCDF 13 35 315.9419 M C12H4 Cl4O TCDF (S) 13 35 37 317.9389 M+2 C12H4 Cl3 ClO TCDF (S) 35 319.8965 MC12H4 ClO2 TCDD 35 37 321.8936 M+2 C12H4 Cl3 ClO2 TCDD 37 327.8847 MC12H4 Cl4O2 TCDD (S) 330.9792 QC C7F13 PFK 13 35 331.9368 M C12H4 Cl4O2 TCDD (S) 13 35 37 333.9339 M+2 C12H4 Cl ClO2 TCDD (S) 35 37 339.8597 M+2 C12H3 Cl4 ClO PECDF 35 37 341.8567 M+4 C12H3 Cl3 Cl2O PeCDF 13 35 37 351.9000 M+2 C12H3 Cl4 ClO PeCDF (S) 13 35 3537 353.8970 M+4 C12H3 Cl Cl2O PeCDF (S) 35 355.8546 M+2 C12H3 Cl337ClO2 PeCDD 35 37 357.8516 M+4 C12H3 Cl3 Cl2O2 PeCDD 13 35 37 367.8949 M+2 C12H3 Cl4 ClO2 PeCDD (S) 13 35 37 369.8919 M+4 C12H3 Cl3 Cl2O2 PeCDD (S) 35 37 375.8364 M+2 C12H4 Cl5 ClO HxCDPE 35 37 409.7974 M+2 C12H3 Cl6 ClO HpCPDE 37 3 373.8208 M+2 C12H235Cl5 ClO HxCDF 35 37 375.8178 M+4 C12H2 Cl4 Cl2O HxCDF 13 35 383.8639 M C12H2 Cl6O HxCDF (S) 13 35 37 385.8610 M+2 C12H2 Cl5 ClO HxCDF (S) 35 37 389.8157 M+2 C12H2 Cl5 ClO2 HxCDD 35 37 391.8127 M+4 C12H2 Cl4 Cl2O2 HxCDD 392.9760 LOCK C9F15 PFK 13 35 37 401.8559 M+2 C12H2 Cl5 ClO2 HxCDD (S) 13 35 37 403.8529 M+4 C12H2 Cl4 Cl2O HxCDD (S) 35 37 445.7555 M+4 C12H2 Cl6 Cl2O OCDPE
943
VerDate 11
TABLE 3ÐELEMENTAL COMPOSITIONS AND EXACT MASSES OF THE IONS MONITORED BY HIGH RESOLUTION MASS SPECTROMETRY FOR PCDD'S AND PCDF'SÐContinued
Descriptor No. Accurate mass Ion type Elemental composition Analyte
430.9729 QC C9F17 PFK 35 37 4 407.7818 M+2 C12H Cl6 ClO HpCDF 35 37 409.7789 M+4 C12H Cl5 Cl2O HpCDF 13 35 417.8253 M C12H Cl7O HpCDF (S) 13 35 37 419.8220 M+2 C12H Cl6 ClO HpCDF (S) 35 37 423.7766 M+2 C12H Cl6 ClO2 HpCDD 35 37 425.7737 M+4 C12H Cl5 Cl2O2 HpCDD 13 35 37 435.8169 M+2 C12H Cl6 ClO2 HpCDD (S) 13 35 37 437.8140 M+4 C12H Cl5 Cl2O2 HpCDD (S) 35 37 479.7165 M+4 C12H Cl7 Cl2O NCPDE 430.9729 LOCK C9F17 PFK 35 37 441.7428 M+2 C12 Cl7 ClO OCDF 35 37 443.7399 M+4 C12 Cl6 Cl2O OCDF 35 37 457.7377 M+2 C12 Cl7 ClO2 OCDD 35 37 459.7348 M+4 C12 Cl6 Cl2O2 OCDD 13 35 37 469.7779 M+2 C12 Cl7 ClO2 OCDD (S) 13 35 37 471.7750 M+4 C12 Cl6 Cl2O2 OCDD (S) 35 37 513.6775 M+4 C12 Cl8 Cl2O2 DCDPE 442.9728 QC C10F17 PFK (a) The following nuclidic masses were used: H = 1.007825 C = 12.000000 13C = 13.003355 F = 18.9984 O = 15.994915 35Cl = 34.968853 37Cl = 36.965903 S = Labeled Standard QC = Ion selected for monitoring instrument stability during the GC/MS analysis.
TABLE 4ÐACCEPTABLE RANGES FOR ION- TABLE 5ÐMINIMUM REQUIREMENTS FOR INITIAL ABUNDANCE RATIOS OF PCDD'S AND PCDF'S AND DAILY CALIBRATION RESPONSE FAC- TORSÐContinued No. of Theo- Control limits chlorine Ion type retical atoms ratio Lower Upper Relative response factors Compound Daily calibra- 4 M/M+2 0.77 0.65 0.89 Initial calibra- tion % dif- 5 M+2/M+4 1.55 1.32 1.78 tion RSD ference 6 M+2/M+4 1.24 1.05 1.43 6 a M/M+2 0.51 0.43 0.59 1,2,3,4,6,7,8-HpCDD ...... 25 25 7 b M/M+2 0.44 0.37 0.51 1,2,3,4,6,7,8-HpCDF ...... 25 25 7 M+2/M+4 1.04 0.88 1.20 OCDD ...... 25 25 8 M+2/M+4 0.89 0.76 1.02 OCDF ...... 30 30 a Used only for 13C-HxCDF. Internal b Used only for 13C-HpCDF. Standards: 13 C12-2,3,7,8-TCDD ...... 25 25 13 TABLE 5ÐMINIMUM REQUIREMENTS FOR INITIAL C12-1,2,3,7,8-PeCDD .. 30 30 13 AND DAILY CALIBRATION RESPONSE FACTORS C12-1,2,3,6,7,8-HxCDD 25 25 13 C12-1,2,3,4,6,7,8- Relative response factors HpCDD ...... 30 30 13 C12-OCDD ...... 30 30 Compound Daily calibra- 13 Initial calibra- C12-2,3,7,8-TCDF ...... 30 30 tion % dif- 13 tion RSD ference C12-1,2,3,7,8-PeCDF ... 30 30 13 C12-1,2,3,6,7,8-HxCDF 30 30 13 Unlabeled C12-1,2,3,4,6,7,8- Analytes: HpCDF ...... 30 30 2,3,7,8-TCDD ...... 25 25 Surrogate 2,3,7,8-TCDF ...... 25 25 Standards: 37 1,2,3,7,8-PeCDD ...... 25 25 Cl4-2,3,7,8-TCDD ...... 25 25 13 1,2,3,7,8-PeCDF ...... 25 25 C12-2,3,4,7,8-PeCDF ... 25 25 13 2,3,4,7,8-PeCDF ...... 25 25 C12-1,2,3,4,7,8-HxCDD 25 25 13 1,2,4,5,7,8-HxCDD ...... 25 25 C12-1,2,3,4,7,8-HxCDF 25 25 1,2,3,6,7,8-HxCDD ...... 25 25 13 C12-1,2,3,4,7,8,9- 1,2,3,7,8,9-HxCDD ...... 25 25 HpCDF ...... 25 25 1,2,3,4,7,8-HxCDF ...... 25 25 Alternate 1,2,3,6,7,8-HxCDF ...... 25 25 Standard: 1,2,3,7,8,9-HxCDF ...... 25 25 13C -1,2,3,7,8,9-HxCDF 25 25 2,3,4,6,7,8-HxCDF ...... 25 25 12
944
VerDate 11
METHOD 24—DETERMINATION OF VOLATILE ings are coatings which contain unreacted MATTER CONTENT, WATER CONTENT, DEN- monomers that are polymerized by exposure SITY, VOLUME SOLIDS, AND WEIGHT SOLIDS to ultraviolet light. To determine if a coat- OF SURFACE COATINGS ing or ink can be classified as a thin-film UV cured coating or ink, use the following equa- 1. Applicability and Principle tion: 1.1 Applicability. This method applies to the determination of volatile matter con- C=F A D Eq. 24–1 tent, water content, density, volume solids, Where: and weight solids of paint, varnish, lacquer, A=Area of substrate, in 2, cm 2. or related surface coatings. C=Amount of coating or ink added to the 1.2 Principle. Standard methods are used substrate, g. to determine the volatile matter content, 3 3 water content, density, volume solids, and D=Density of coating or ink, g/in (g/cm ) weight solids of the paint, varnish, lacquer, F=Manufacturer’s recommended film thick- or related surface coatings. ness, in (cm). 2. Applicable Standard Methods If C is less than 0.2 g and A is greater than or equal to 35 in 2 (225 cm 2) then the coating Use the apparatus, reagents, and proce- or ink is considered a thin-film UV radi- dures specified in the standard methods ation-cured coating for determining applica- below: bility of ASTM D 5403–93. 2.1 ASTM D1475–60 (Reapproved 1980), Standard Test Method for Density of Paint, NOTE: As noted in Section 1.4 of ASTM D Varnish, Lacquer, and Related Products (in- 5403–93, this method may not be applicable to corporated by reference—see § 60.17). radiation curable materials wherein the 2.2 ASTM D2369–81, Standard Test Method volatile material is water. For all other for Volatile Content of Coatings (incor- coatings not covered by Sections 3.1 or 3.2 porated by reference—see § 60.17). analyze as follows: 2.3 ASTM D3792–79, Standard Test Method 3.3 Volatile Matter Content. Use the pro- for Water Content of Water-Reducible Paints cedure in ASTM D2369–81 (incorporated by by Direct Injection into a Gas Chro- reference—see § 60.17) to determine the vola- matograph (incorporated by reference—see tile matter content (may include water) of § 60.17). the coating. Record the following informa- 2.4 ASTM D4017–81, Standard Test Method tion: for Water in Paints and Paint Materials by W =Weight of dish and sample before heat- the Karl Fischer Titration Method (incor- 1 ing, g. porated by reference—see § 60.17). W =Weight of dish and sample after heating, 2.5 ASTM D4457–85 Standard Test Method 2 g. for Determination of Dichloromethane and 1,1,1–Trichloroethane in Paints and Coatings W3=Sample weight, g. by Direct Injection into a Gas Chro- Run analyses in pairs (duplicate sets) for matograph (incorporated by reference—see each coating until the criterion in Section § 60.17). 4.3 is met. Calculate the weight fraction of 2.6 ASTM D 5403–93 Standard Test Meth- the volatile matter (Wv) for each analysis as ods for Volatile Content of Radiation Cur- follows: able Materials (incorporated by reference— see § 60.17). WW− W = 1 2 Eq. 24 - 2 3. Procedure v W 3.1 Multicomponent Coatings. Multicom- 3 ¯ ponent coatings are coatings that are pack- Record the arithmetic average (Wv). aged in two or more parts, which are com- 3.4 Water Content. For waterborne (water bined before application. Upon combination reducible) coatings only, determine the a coreactant from one part of the coating weight fraction of water (WW) using either chemically reacts, at ambient conditions, ‘‘Standard Content Method Test for Water of with a coreactant from another part of the Water-Reducible Paints by Direct Injection coating. To determine the total volatile con- into a Gas Chromatograph’’ or ‘‘Standard tent, water content, and density of multi- Test Method for Water in Paint and Paint component coatings, follow the procedures in Materials by Karl Fischer Method.’’ (These section 3.7. two methods are incorporated by reference— 3.2 Non Thin-film Ultraviolet Radiation- see § 60.17.) A waterborne coating is any coat- cured Coating. To determine volatile content ing which contains more than 5 percent of non thin-film ultraviolet radiation-cured water by weight in its volatile fraction. Run (UV radiation-cured) coatings, follow the duplicate sets of determinations until the procedures in Section 3.9. Determine water criterion in Section 4.3 is met. Record the ¯ content, density and solids content of the arithmetic average (Ww). UV-cured coatings according to Sections 3.4, 3.5 Coating Density. Determine the den- 3.5, and 3.6, respectively. The UV-cured coat- sity (Dc, kg/liter) of the surface coating
945
VerDate 11
using the procedure in ASTM D1475–60 (Re- greater than 40 weight percent, a suitable approved 1980) (incorporated by reference— size is 0.5±0.10 g. see § 60.17). NOTE: If the volatile content determined Run duplicate sets of determinations for pursuant to section 5 is not in the range cor- each coating until the criterion in Section responding to the sample size chosen repeat 4.3 is met. Record the arithmetic average the test with the appropriate sample size. ¯ (Dc). Add the specimen dropwise, shaking (swirl- 3.6 Solids Content. Determine the volume ing) the dish to disperse the specimen com- fraction (Vs) solids of the coating by calcula- pletely in the solvent. If the material forms tion using the manufacturer’s formulation. a lump that cannot be dispersed, discard the 3.7 Exempt Solvent Content. Determine specimen and prepare a new one. Similarly, the weight fraction of exempt solvents (WE) prepare a duplicate. The sample shall stand by using ASTM Method D4457–85 (incor- for a minimum of 1 hour, but no more than porated by reference—see § 60.17). Run a du- 24 hours prior to being oven dried at plicate set of determinations and record the 110°C±5°C. for 1 hour. arithmetic average (WE) 3.8.2.2 Heat the aluminum foil dishes con- NOTE: Exempt solvents are defined as those taining the dispersed specimens in the forced solvents listed in 57 FR 3941, February 3, draft oven for 60 min at 110±5°C. Caution— 1992. Dichloromethane and 1,1,1-trichloro- provide adequate ventilation, consistent ethane are listed exempt solvents and may with accepted laboratory practice, to pre- be used in coatings. vent solvent vapors from accumulating to a 3.8 To determine the total volatile con- dangerous level. tent, water content, and density of multi- 3.8.2.3 Remove the dishes from the oven, component coatings, use the following proce- place immediately in a desiccator, cool to dures: ambient temperature, and weigh to within 1 mg. 3.8.1 Prepare about 100 ml of sample by mixing the components in a storage con- 3.8.2.4 Run analyses in pairs (duplicate tainer, such as a glass jar with a screw top or sets) for each coating mixture until the cri- a metal can with a cap. The storage con- terion in section 4.3 is met. Calculate Wv fol- tainer should be just large enough to hold lowing Equation 24–2 and record the arith- the mixture. Combine the components (by metic average. weight or volume) in the ratio recommended 3.9 UV-cured Coating’s Volatile Matter by the manufacturer. Tightly close the con- Content. Use the procedure in ASTM D 5403– tainer between additions and during mixing 93 (incorporated by reference—see § 60.17) to to prevent loss of volatile materials. How- determine the volatile matter content of the ever, most manufacturers mixing instruc- coating except the curing test described in tions are by volume. Because of possible NOTE 2 of ASTM D 5403–93 is required. error caused by expansion of the liquid when 4. Data Validation Procedure measuring the volume, it is recommended 4.1 Summary. The variety of coatings that the components be combined by weight. that may be subject to analysis makes it When weight is used to combine the compo- necessary to verify the ability of the analyst nents and the manufacturer’s recommended and the analytical procedures to obtain re- ratio is by volume, the density must be de- producible results for the coatings tested. termined by section 3.5. This is done by running duplicate analyses 3.8.2 Immediately after mixing, take on each sample tested and comparing results aliquots from this 100 ml sample for deter- with the within-laboratory precision state- mination of the total volatile content, water ments for each parameter. Because of the in- content, and density. To determine water herent increased imprecision in the deter- content, follow section 3.4. To determine mination of the VOC content of waterborne density, follow section 3.5. To determine coatings as the weight percent water in- total volatile content, use the apparatus and creases, measured parameters for waterborne reagents described in ASTM D2369–81, sec- coatings are modified by the appropriate tions 3 and 4, respectively (incorporated by confidence limits based on between-labora- reference, and see § 60.17) the following proce- tory precision statements. dures: 4.2 Analytical Precision Statements. The 3.8.2.1 Weigh and record the weight of an within-laboratory and between-laboratory aluminum foil weighing dish. Add 31±l of precision statements are given below: suitable solvent as specified in ASTM D2369– 81 to the weighing dish. Using a syringe as Within- Between- specified in ASTM D2369–81, weigh to 1 mg, labora- laboratory by difference, a sample of coating into the tory weighing dish. For coatings believed to have Volatile matter content, Wv ...... 1.5 pct 4.7 pct a volatile content less than 40 weight per- WÅ v. WÅ v. ± cent, a suitable size is 0.3 0.10 g, but for coat- Water content, Ww ...... 2.9 pct 7.5 pct ings believed to have a volatile content WÅ w. WÅ w.
946
VerDate 11
METHOD 24A—DETERMINATION OF VOLATILE Within- Between- labora- laboratory MATTER CONTENT AND DENSITY OF PRINTING tory INKS AND RELATED COATINGS
Density, Dc ...... 0.001 kg/ 0.002 kg/ 1. Applicability and Principle liter. liter. 1.1 Applicability. This method applies to the determination of the volatile organic 4.3 Sample Analysis Criteria. For Wv and compound (VOC) content and density of sol- Ww, run duplicate analyses until the dif- ference between the two values in a set is vent-borne (solvent reducible) printing inks less than or equal to the within-laboratory or related coatings. precision statement for that parameter. For 1.2 Principle. Separate procedures are used to determine the VOC weight fraction Dc run duplicate analyses until each value in a set deviates from the mean of the set by no and density of the coating and the density of more than the within-laboratory precision the solvent in the coating. The VOC weight statement. If after several attempts it is fraction is determined by measuring the concluded that the ASTM procedures cannot weight loss of a known sample quantity be used for the specific coating with the es- which has been heated for a specified length tablished within-laboratory precision, the of time at a specified temperature. The den- Administrator will assume responsibility for sity of both the coating and solvent are providing the necessary procedures for revis- measured by a standard procedure. From ing the method or precision statements upon this information, the VOC volume fraction is written request to: Director, Emission calculated. Standards and Engineering Division, (MD–13) 2. Procedure Office of Air Quality Planning and Stand- 2.1 Weight Fraction VOC. ards, U.S. Environmental Protection Agen- 2.1.1 Apparatus. cy, Research Triangle Park, NC 27711. 4.4 Confidence Limit Calculations for Wa- 2.1.1.1 Weighing Dishes. Aluminum foil, 58 terborne Coatings. Based on the between-lab- mm in diameter by 18 mm high, with a flat oratory precision statements, calculate the bottom. There must be at least three weigh- confidence limits for waterborne coatings as ing dishes per sample. follows: 2.1.1.2 Disposable Syringe. 5 ml. To calculate the lower confidence limit, 2.1.1.3 Analytical Balance. To measure to subtract the appropriate between-laboratory within 0.1 mg. precision value from the measured mean 2.1.1.4 Oven. Vacuum oven capable of value for that parameter. To calculate the maintaining a temperature of 120±2°C and an upper confidence limit, add the appropriate absolute pressure of 510 ±51 mm Hg for 4 between-laboratory precision value to the hours. Alternatively, a forced draft oven ca- measured mean value for that parameter. pable of maintaining a temperature of ± ° For Wv and Dc, use the lower confidence lim- 120 2 C for 24 hours. its, and for Ww, use the upper confidence 2.1.2 Analysis. Shake or mix the sample limit. Because Vs is calculated, there is no thoroughly to assure that all the solids are adjustment for the parameter. completely suspended. Label and weigh to 5. Calculations the nearest 0.1 mg a weighing dish and record this weight (Mxl). 5.1 Nonaqueous Volatile Matter. Using a 5-ml syringe without a needle re- 5.1.1 Solvent-borne Coatings. move a sample of the coating. Weigh the sy- Wo=Wv Eq. 24–3 ringe and sample to the nearest 0.1 mg and Where: record this weight (McYl). Transfer 1 to 3 g of the sample to the tared weighing dish. Re- W =Weight fraction nonaqueous volatile o weigh the syringe and sample to the nearest matter, g/g. 0.1 mg and record this weight (M ). Heat 5.1.2 Waterborne Coatings. cY2 the weighing dish and sample in a vacuum Wo=Wv¥Ww Eq. 24–4 oven at an absolute pressure of 510±51 mm Hg 5.1.3 Coatings Containing Exempt Sol- and a temperature of 120≤2°C for 4 hours. Al- vents. ternatively, heat the weighing dish and sam- = − − ple in a forced draft oven at a temperature of Wo W v W E W w Eq. 24 - 5 120≤2°C for 24 hours. After the weighing dish where: has cooled, reweigh it to the nearest 0.1 mg and record the weight (Mx2). Repeat this pro- WE=weight fraction of exempt solvents, g/ g. cedure for a total of three determinations for 5.2 Weight Fraction Solids. each sample. 2.2 Coating Density. Determine the den- W=1 − W Eq. 25 - 5 sity of the ink or related coating according s v to the procedure outlined in ASTM D 1475–60 where: (Reapproved 1980), (incorporated by ref- Ws=weight fraction of solids, g/g. erence—see § 60.17).
947
VerDate 11
2.3 Solvent Density. Determine the den- having 10 percent CO2 and 20 percent water sity of the solvent according to the proce- vapor. dure outlined in ASTM D 1475–60 (reapproved This method is not the only method that 1980). Make a total of three determinations applies to the measurement of TGNMO. for each coating. Report the density Do as Costs, logistics, and other practicalities of the arithmetic average of the three deter- source testing may make other test methods minations. more desirable for measuring VOC contents 3. Calculations of certain effluent streams. Proper judgment is required in determining the most applica- 3.1 Weight Fraction VOC. Calculate the ble VOC test method. For example, depend- weight fraction volatile organic content W o ing upon the molecular weight of the using the following equation: organics in the effluent stream, a totally MMMM+ − − automated semicontinuous nonmethane W = x1 cY 1 cY 2 x 2 Eq. 24 A - 1 organics (NMO) analyzer interfaced directly o − MMcY1 cY 2 to the source may yield accurate results. This approach has the advantage of provid- Report the weight fraction VOC Wo as the ing emission data semicontinuously over an arithmetic average of the three determina- extended time period. tions. Direct measurement of an effluent with a 3.2 Volume Fraction VOC. Calculate the flame ionization detector (FID) analyzer volume fraction volatile organic content Vo may be appropriate with prior characteriza- using the following equation: tion of the gas stream and knowledge that ¯ ¯ ¯ the detector responds predictably to the or- Vo = (WoDc/Do ganic compounds in the stream. If present, Eq. 24A–2 methane (CH4) will, of course, also be meas- ured. The FID can be applied to the deter- 4. Bibliography mination of the mass concentration of the 1. Standard Test Method for Density of total molecular structure of the organic Paint, Varnish, Lacquer, and Related Prod- emissions under any of the following limited ucts. ASTM Designation D 1475–60 (Re- conditions: (1) Where only one compound is approved 1980). known to exist; (2) when the organic com- 2. Teleconversation. Wright, Chuck, pounds consist of only hydrogen and carbon; Inmont Corporation with Reich, R. A., Ra- (3) where the relative percentages of the dian Corporation. September 25, 1979. Gra- compounds are known or can be determined, vure Ink Analysis. and the FID responses to the compounds are 3. Teleconversation. Oppenheimer, Rob- known; (4) where a consistent mixture of the ert, Gravure Research Institute with Burt, compounds exists before and after emission Rick, Radian Corporation, November 5, 1979. control and only the relative concentrations Gravure Ink Analysis. are to be assessed; or (5) where the FID can be calibrated against mass standards of the METHOD 25—DETERMINATION OF TOTAL GASE- compounds emitted (solvent emissions, for OUS NONMETHANE ORGANIC EMISSIONS AS example). CARBON Another example of the use of a direct FID is as a screening method. If there is enough 1. Applicability and Principle information available to provide a rough es- 1.1 Applicability. This method applies to timate of the analyzer accuracy, the FID an- the measurement of volatile organic com- alyzer can be used to determine the VOC pounds (VOC) as total gaseous nonmethane content of an uncharacterized gas stream. organics (TGNMO) as carbon in source emis- With a sufficient buffer to account for pos- sions. Organic particulate matter will inter- sible inaccuracies, the direct FID can be a fere with the analysis and, therefore, a par- useful tool to obtain the desired results ticulate filter is required. The minimum de- without costly exact determination. tectable for the method is 50 ppm as carbon. In situations where a qualitative/quan- When carbon dioxide (CO2) and water vapor titative analysis of an effluent stream is de- are present together in the stack, they can sired or required, a gas chromatographic FID produce a positive bias in the sample. The system may apply. However, for sources magnitude of the bias depends on the con- emitting numerous organics, the time and centrations of CO2 and water vapor. As a expense of this approach will be formidable. guideline, multiply the CO2 concentration, 1.2 Principle. An emission sample is with- expressed as volume percent, times the water drawn from the stack at a constant rate vapor concentration. If this product does not through a heated filter and a chilled conden- exceed 100, the bias can be considered insig- sate trap by means of an evacuated sample nificant. For example, the bias is not signifi- tank. After sampling is completed, the cant for a source having 10 percent CO2 and TGNMO are determined by independently 10 percent water vapor, but it would be sig- analyzing the condensate trap and sample nificant for a source near the detection limit tank fractions and combining the analytical
948
VerDate 11
results. The organic content of the conden- fiberfrax insulation which is sealed in place sate trap fraction is determined by oxidizing by means of a silicon rubber bead around the the NMO to CO2 and quantitatively collect- upper sides of the box. A removable lid made ing the effluent in an evacuated vessel; then in a similar manner, with a 25-mm (1-in.) gap a portion of the CO2 is reduced to CH4 and between the parts, is used to cover the heat- measured by an FID. The organic content of ing chamber. the sample tank fraction is measured by in- The inner box is heated witn a 250-watt jecting a portion of the sample into a gas cartridge heater, shielded by a stainless steel chromatographic column to separate the shroud. The heater is regulated by a NMO from carbon monoxide (CO), CO2, and thermostatic temperature controller which CH4; the NMO are oxidized to CO2, reduced to is set to maintain a temperature of 121 °C as CH4, and measured by an FID. In this man- measured by a thermocouple in the gas line ner, the variable response of the FID associ- just before the filter. An additional thermo- ated with different types of organics is elimi- couple is used to monitor the temperature of nated. the gas behind the filter. 2.1.4 Condensate Trap. 9.5-mm (3⁄8-in.) OD 2. Apparatus 316 stainless steel tubing bent into a U- 2.1 Sampling. The sampling system con- shape. Exact dimensions are shown in Figure sists of a heated probe, heated filter, conden- 25–3. The tubing shall be packed with coarse sate trap, flow control system, and sample quartz wool, to a density of approximately tank (Figure 25–1). The TGNMO sampling 0.11 g/cc before bending. While the conden- equipment can be constructed from commer- sate trap is packed with dry ice in the cially available components and components Dewar, an ice bridge may form between the fabricated in a machine shop. The following arms of the condensate trap making it dif- equipment is required: ficult to remove the condensate trap. This 2.1.1 Heated Probe. 6.4-mm (1⁄4-in.) OD problem can be prevented by attaching a stainless steel tubing with a heating system steel plate between the arms of the conden- capable of maintaining a gas temperature at sate trap in the same plane as the arms to the exit end of at least 129°C (265°F). The completely fill the intervening space. probe shall be equipped with a thermocouple 2.1.5 Valve. Stainless steel shut-off valve at the exit end to monitor the gas tempera- for starting and stopping sample flow. ture. 2.1.6 Metering Valve. Stainless steel con- A suitable probe is shown in Figure 25–1. trol valve for regulating the sample flow rate The nozzle is an elbow fitting attached to through the sample train. the front end of the probe while the thermo- 2.1.7 Rotameter. Glass tube with stainless couple is inserted in the side arm of a tee fit- steel fittings, capable of measuring sample ting attached to the rear of the probe. The flow in the range of 60 to 100 cc/min. probe is wrapped with a suitable length of 2.1.8 Sample Tank. Stainless steel or alu- high temperature heating tape, and then minum tank with a minimum volume of 4 li- covered with two layers of glass cloth insula- ters. tion and one layer of aluminum foil. 2.1.9 Mercury Manometer or Absolute Pressure Gauge. Capable of measuring pres- NOTE. If it is not possible to use a heating sure to within 1 mm Hg in the range of 0 to system for safety reasons, an unheated sys- 900 mm. tem with an in-stack filter is a suitable al- 2.1.10 Vacuum Pump. Capable of evacuat- ternative. ing to an absolute pressure of 10 mm Hg. 2.1.2 Filter Holder. 25-mm (15⁄16-in.) ID 2.2 Condensate Recovery Apparatus. The Gelman filter holder with stainless steel system for the recovery of the organics cap- body and stainless steel support screen with tured in the condensate trap consists of a the Viton O-ring replaced by a Teflon O-ring. heat source, oxidation catalyst, nondisper- NOTE. Mention of trade names or specific sive infrared (NDIR) analyzer and an inter- products does not constitute endorsement by mediate collection vessel (ICV). Figure 25–4 the Environmental Protection Agency. is a schematic of a typical system. The sys- 2.1.3 Filter Heating System. A metal box tem shall be capable of proper oxidation and consisting of an inner and an outer shell sep- recovery, as specified in Section 5.1. The fol- arated by insulating material with a heating lowing major components are required: element in the inner shell capable of main- 2.2.1 Heat Source. Sufficient to heat the taining a gas temperature at the filter of condensate trap (including connecting tub- 121±3 °C (250±5 °F). ing) to a temperature of 200 °C. A system A suitable heating box is shown in Figure using both a heat gun and an electric tube 25–2. The outer shell is a metal box that furnace is recommended. measures 102 mm×280 mm×292 mm (4 in.×11 2.2.2 Heat Tape. Sufficient to heat the con- in.×111⁄2 in.), while the inner shell is a metal necting tubing between the water trap and box measuring 76 mm×229 mm×241 mm (3 the oxidation catalyst to 100 °C. in.×9 in.×91⁄2 in.). The inner box is supported 2.2.3 Oxidation Catalyst. A suitable length by 13-mm (1⁄2-in.) phenolic rods. The void of 9.5-mm (3⁄8-in.) OD Inconel 600 tubing space between the boxes is filled with packed with 15 cm (6 in.) of 3.2-mm (1⁄8-1n.)
949
VerDate 11
diameter 19 percent chromia on alumina pel- 2.3.4 Sample Injection System. A 10-port lets. The catalyst material is packed in the GC sample injection valve fitted with a sam- center of the catalyst tube with quartz wool ple loop properly sized to interface with the packed on either end to hold it in place. The NMO analyzer (1-cc loop recommended). catalyst tube shall be mounted vertically in 2.3.5 FID. An FID meeting the following a 650 °C tube furnace. specifications is required: 2.2.4 Water Trap. Leak proof, capable of 2.3.5.1 Linearity. A linear response (±5 removing moisture from the gas stream percent) over the operating range as dem- 2.2.5 Syringe Port. A 6.4-mm (1⁄4-in.) OD onstrated by the procedures established in stainless steel tee fitting with a rubber sep- Section 5.2.3. tum placed in the side arm. 2.3.5.2 Range. A full scale range of 10 to 2.2.6 NDIR Detector. Capable of indicating 50,000 ppm CH4. Signal attenuators shall be available to produce a minimum signal re- CO2 concentration in the range of zero to 5 percent, to monitor the progress of combus- sponse of 10 percent of full scale. tion of the organic compounds from the con- 2.3.6 Data Recording System. Analog strip densate trap. chart recorder or digital integration system 2.2.7 Flow-Control Valve. Stainless steel, compatible with the FID for permanently re- to maintain the trap conditioning system cording the analytical results. 2.4 Other Analysis Apparatus. near atmospheric pressure. 2.4.1 Barometer. Mercury, aneroid, or 2.2.8 Intermediate Collection Vessel. other barometer capable of measuring at- Stainless steel or aluminum, equipped with a mospheric pressure to within 1 mm Hg. female quick connect. Tanks with nominal 2.4.2 Thermometer. Capable of measuring volumes of at least 6 liters are rec- the laboratory temperature to within 1°C. ommended. 2.4.3 Vacuum Pump. Capable of evacuat- 2.2.9 Mercury Manometer or Absolute ing to an absolute pressure of 10 mm Hg. Pressure Gauge. Capable of measuring pres- 2.4.4 Syringes. 10-µl and 50-µl liquid injec- sure to within 1 mm Hg in the range of 0 to tion syringes. 900 mm. 2.4.5 Liquid Sample Injection Unit. 316 SS 2.2.10 Syringe. 10-ml gas-tight, glass sy- U-tube fitted with an injection septum, see ringe equipped with an appropriate needle. Figure 25–7. 2.3 NMO Analyzer. The NMO analyzer is a gas chromatograph (GC) with backflush ca- 3. Reagents pability for NMO analysis and is equipped 3.1 Sampling. The following are required with an oxidation catalyst, reduction cata- for sampling: lyst, and FID. Figures 25–5 and 25–6 are sche- 3.1.1 Crushed Dry Ice. matics of a typical NMO analyzer. This 3.1.2 Coarse Quartz Wool. 8 to 15 µm. semicontinuous GC/FID analyzer shall be ca- 3.1.3 Filters. Glass fiber filters, without pable of: (1) Separating CO, CO2, and CH4 organic binder. from NMO, (2) reducing the CO2 to CH4 and 3.2 NMO Analysis. The following gases are quantifying as CH4, and (3) oxidizing the needed: NMO to CO2, reducing the CO2 to CH4 and 3.2.1 Carrier Gases. Zero grade helium quantifying as CH , according to Section 5.2. 4 (He) and oxygen (O2 containing less than 1 The analyzer consists of the following major ppm CO2 and less than 0.1 ppm C as hydro- components: carbon. 2.3.1 Oxidation Catalyst. A suitable 3.2.2 Fuel Gas. Zero grade hydrogen (H2), length of 9.5-mm (3⁄8-in.) OD Inconel 600 tub- 99.999 percent pure. ing packed with 5.1 cm (2 in.) of 19 percent 3.2.3 Combustion Gas. Zero grade air or O2 chromia on 3.2-mm (1⁄8-in.) alumina pellets. as required by the detector. The catalyst material is packed in the center 3.3 Condensate Analysis. The following of the tube supported on either side by gases are needed: quartz wool. The catalyst tube must be 3.3.1 Carrier Gas. Zero grade air, contain- mounted vertically in a 650 °C furnace. ing less than 1 ppm C. 2.3.2 Reduction Catalyst. A 7.6-cm (3-in.) 3.3.2 Auxiliary O2. Zero grade O2, contain- length of 6.4-mm (1⁄4-in.) OD Inconel tubing ing less than 1 ppm C. fully packed with 100-mesh pure nickel pow- 3.3.3 Hexane. ACS grade, for liquid injec- der. The catalyst tube must be mounted ver- tion. tically in a 400 °C furnace. 3.3.4 Decane. ACS grade, for liquid injec- 2.3.3 Separation Column(s). A 30-cm (1-ft) tion. length of 3.2-mm (1⁄8-in.) OD stainless steel 3.4 Calibration. For all calibration gases, tubing packed with 60/80 mesh Unibeads 1S the manufacturer must recommend a maxi- followed by a 61-cm (2-ft) length of 3.2-mm mum shelf life for each cylinder (i.e., the (1⁄8-in.) OD stainless steel tubing packed with length of time the gas concentration is not 60/80 mesh Carbosieve G. The Carbosieve and expected to change more than ±5 percent Unibeads columns must be baked separately from its certified value). The date of gas cyl- at 200 °C with carrier gas flowing through inder preparation, certified organic con- them for 24 hours before initial use. centration, and recommended maximum
950
VerDate 11
shelf life must be affixed to each cylinder be- assemble the sampling system as shown in fore shipment from the gas manufacturer to Figure 25–1. Immerse the condensate trap the buyer. The following calibration gases body in dry ice. The point where the inlet are required: tube joins the trap body should be 2.5 to 5 cm 3.4.1 Oxidation Catalyst Efficiency Check above the top of the dry ice. Calibration Gas. Gas mixture standard with 4.1.4 Pretest Leak Check. A pretest leak nominal concentration of 1 percent methane check is required. Calculate or measure the in air. approximate volume of the sampling train 3.4.2 FID Linearity and NMO Calibration from the probe trip to the sample tank valve. Gases. Three gas mixture standards with After assembling the sampling train, plug nominal propane concentrations of 20 ppm, the probe tip, and make certain that the 200 ppm, and 3000 ppm, in air. sample tank valve is closed. Turn on the vac- 3.4.3 CO Calibration Gases. Three gas 2 uum pump, and evacuate the sampling sys- mixture standards with nominal CO con- 2 tem from the probe tip to the sample tank centrations of 50 ppm, 500 ppm, and 1 per- valve to an absolute pressure of 10 ppm Hg or cent, in air. less. Close the purge valve, turn off the NOTE.— Total NMO of less than 1 ppm re- pump, wait a minimum period of 5 minutes, quired for 1 percent mixture. and recheck the indicated vacuum. Calculate 3.4.4 NMO Analyzer System Check Cali- the maximum allowable pressure change bration Gases. Four calibration gases are based on a leak rate of 1 percent of the sam- needed as follows: pling rate using Equation 25–1, Section 6.2. If 3.4.4.1 Propane Mixture. Gas mixture the measured pressure change exceeds the standard containing (nominal) 50 ppm CO, 50 calculated limit, correct the problem before ppm CH4, 2 percent CO2, and 20 ppm C3H8, beginning sampling. The results of the leak prepared in air. check should be included in the test report. 3.4.4.2 Hexane. Gas mixture standard con- 4.1.5 Sample Train Operation. Unplug the taining (nominal) 50 ppm hexane in air. probe tip, and place the probe into the stack 3.4.4.3 Toluene. Gas mixture standard con- such that the probe is perpendicular to the taining (nominal) 20 ppm toluene in air. duct or stack axis; locate the probe tip at a 3.4.4.4 Methanol. Gas mixture standard single preselected point of average velocity containing (nominal) 100 ppm methanol in facing away from the direction of gas flow. air. For stacks having a negative static pressure, 4. Procedure seal the sample port sufficiently to prevent air in-leakage around the probe. Set the 4.1 Sampling. probe temperature controller to 129 °C (265 4.1.1 Cleaning Sampling Equipment. Be- °F) and the filter temperature controller to fore its initial use and after each subsequent 121 °C (250 °F). Allow the probe and filter to use, a condensate trap should be thoroughly heat for about 30 minutes before purging the cleaned and checked to ensure that it is not sample train. contaminated. Both cleaning and checking Close the sample valve, open the purge can be accomplished by installing the trap in valve, and start the vacuum pump. Set the the condensate recovery system and treating it as if it were a sample. The trap should be flow rate between 60 and 100 cc/min, and heated as described in the final paragraph of purge the train with stack gas for at least 10 Section 4.3.3. A trap may be considered clean minutes. When the temperatures at the exit ends of the probe and filter are within their when the CO2 concentration in its effluent gas drops below 10 ppm. This check is op- specified range, sampling may begin. tional for traps that have been used to col- Check the dry ice level around the conden- lect samples which were then recovered ac- sate trap, and add dry ice if necessary. cording to the procedure in Section 4.3.3. Record the clock time. To begin sampling, 4.1.2 Sample Tank Evacuation and Leak close the purge valve and stop the pump. Check. Evacuate the sample tank to 10 mm Open the sample valve and the sample tank Hg absolute pressure or less. Then close the valve. Using the flow control valve, set the sample tank valve, and allow the tank to sit flow through the sample train to the proper for 60 minutes. The tank is acceptable if no rate. Adjust the flow rate as necessary to change in tank vacuum is noted. The evacu- maintain a constant rate (±10 percent) ation and leak check may be conducted ei- throughout the duration of the sampling pe- ther in the laboratory or the field. The re- riod. Record the sample tank vacuum and sults of the leak check should be included in flowmeter setting at 5-minute intervals. (See the test report. Figure 25–8.) Select a total sample time 4.1.3 Sample Train Assembly. Just before greater than or equal to the minimum sam- assembly, measure the tank vacuum using a pling time specified in the applicable subpart mercury U-tube manometer or absolute pres- of the regulation; end the sampling when sure gauge. Record this vacuum, the ambient this time period is reached or when a con- temperature, and the barometric pressure at stant flow rate can no longer be maintained this time. Close the sample tank valve and because of reduced sample tank vacuum.
951
VerDate 11
NOTE: If sampling had to be stopped before tain condensed water and organics and a obtaining the minimum sampling time (spec- small volume of sampled gas. This gas from ified in the applicable subpart) because a the stack may contain a significant amount constant flow rate could not be maintained, of CO2 which must be removed from the con- proceed as follows: After closing the sample densate trap before the sample is recovered. tank valve, remove the used sample tank This is accomplished by purging the conden- from the sampling train (without disconnect- sate trap with zero air and collecting the ing other portions of the sampling train). purged gas in the original sample tank. Take another evacuated and leak-checked Begin with the sample tank and conden- sample tank, measure and record the tank sate trap from the test run to be analyzed. vacuum, and attach the new tank to the Set the four-port valve of the condensate re- sampling train. After the new tank is at- covery system in the CO2 purge position as tached to the sample train, proceed with the shown in Figure 25–9. With the sample tank sampling until the required minimum sam- valve closed, attach the sample tank to the pling time has been exceeded. sample recovery system. With the sample re- 4.2 Sample Recovery. After sampling is covery valve in the vent position and the completed, close the flow control valve, and flow control valve fully open, evacuate the record the final tank vacuum; then record manometer or pressure gauge to the vacuum the tank temperature and barometric pres- of the sample tank. Next, close the vacuum sure. Close the sample tank valve, and dis- pump valve, open the sample tank valve, and connect the sample tank from the sample record the tank pressure. system. Disconnect the condensate trap at Attach the dry-ice-cooled condensate trap the flowmetering system, and tightly seal to the recovery system, and initiate the both ends of the condensate trap. Do not in- clude the probe from the stack to the filter purge by switching the sample recovery as part of the condensate sample. Keep the valve from vent to collect position. Adjust trap packed in dry ice until the samples are the flow control valve to maintain atmos- returned to the laboratory for analysis. En- pheric pressure in the recovery system. Con- sure that the test run number is properly tinue the purge until the CO2 concentration identified on the condensate trap and the of the trap effluent is less than 5 ppm. CO2 sample tank(s). concentration in the trap effluent should be 4.3 Condensate Recovery. See Figure 25–9. measured by extracting syringe samples Set the carrier gas flow rate, and heat the from the recovery system and analyzing the catalyst to its operating temperature to con- samples with the NMO analyzer. This proce- dition the apparatus. dure should be used only after the NDIR re- 4.3.1 Daily Performance Checks. Each day sponse has reached a minimum level. Using a before analyzing any samples, perform the 10-ml syringe, extract a sample from the sy- following tests: ringe port prior to the NDIR, and inject this 4.3.1.1 Leak Check. With the carrier gas sample into the NMO analyzer.
inlets and the flow control valve closed, in- After the completion of the CO2 purge, use stall a clean condensate trap in the system, the carrier gas bypass valve to pressurize the and evacuate the system to 10 mm Hg abso- sample tank to approximately 1060 mm Hg lute pressure or less. Close the vacuum pump absolute pressure with zero air. valve and turn off the vacuum pump. Mon- 4.3.3 Recovery of the Condensate Trap itor the system pressure for 10 minutes. The Sample. See Figure 25–10. Attach the ICV to system is acceptable if the pressure change the sample recovery system. With the sam- is less than 2 mm Hg. ple recovery valve in a closed position, be- 4.3.1.2 System Background Test. Adjust tween vent and collect, and the flow control the carrier gas and auxiliary oxygen flow and ICV valves fully open, evacuate the ma- rate to their normal values of 100 cc/min and nometer or gauge, the connecting tubing, 150 cc/min, respectively, with the sample re- and the ICV to 10 mm Hg absolute pressure. covery valve in vent position. Using a 10-ml Close the flow-control and vacuum pump syringe withdraw a sample from the system valves. effluent through the syringe port. Inject this sample into the NMO analyzer, and measure Begin auxiliary oxygen flow to the oxida- the CO content. The system background is tion catalyst at a rate of 150 cc/min, then 2 switch the four-way valve to the trap recov- acceptable if the CO2 concentration is less than 10 ppm. ery position and the sample recovery valve 4.3.1.3 Oxidation Catalyst Efficiency to collect position. The system should now Check. Conduct a catalyst efficiency test as be set up to operate as indicated in Figure specified in Section 5.1.2 of this method. If 25–10. After the manometer or pressure gauge the criterion of this test cannot be met, begins to register a slight positive pressure, make the necessary repairs to the system be- open the flow control valve. Adjust the flow- fore proceeding. control valve to maintain atmospheric pres- 4.3.2 Condensate Trap CO2 Purge and sure in the system within 10 percent. Sample Tank Pressurization. After sampling Now, remove the condensate trap from the is completed, the condensate trap will con- dry ice, and allow it to warm to ambient
952
VerDate 11
temperature while monitoring the NDIR re- as rapidly as possible. A rate of 30 °C/min has sponse. If after 5 minutes, the CO2 concentra- been shown to be adequate. Record the value tion of the catalyst effluent is below 10,000 obtained for the condensible organic mate- ppm, discontinue the auxiliary oxygen flow rial (Ccm) measured as CO2 and any measured to the oxidation catalyst. Begin heating the NMO. Return the column oven temperature trap by placing it in a furnace preheated to to 85 °C in preparation for the next analysis. 200 °C. Once heating has begun, carefully Analyze each sample in triplicate, and report monitor the NDIR response to ensure that the average Ccm. the catalyst effluent concentration does not 4.4.4 Analysis of Sample Tank. Perform exceed 50,000 ppm. Whenever the CO2 con- the analysis as described in Section 4.4.3, but centration exceeds 50,000 ppm, supply auxil- record only the value measured for NMO iary oxygen to the catalyst at the rate of 150 (Ctm). cc/min. Begin heating the tubing that con- 4.5 Audit Samples. Analyze a set of two nected the heated sample box to the conden- audit samples concurrently with any compli- sate trap only after the CO2 concentration ance samples and in exactly the same man- falls below 10,000 ppm. This tubing may be ner to evaluate the analyst’s technique and heated in the same oven as the condensate the instrument calibration. The same ana- trap or with an auxiliary heat source such as lysts, analytical reagents, and analytical a heat gun. Heating temperature must not system shall be used for the compliance sam- exceed 200 °C. If a heat gun is used, heat the ples and the EPA audit samples; if this con- tubing slowly along its entire length from dition is met, auditing of subsequent compli- the upstream end to the downstream end, ance analyses for the same enforcement and repeat the pattern for a total of three agency within 30 days is not required. An times. Continue the recovery until the CO2 audit sample set may not be used to validate concentration drops to less than 10 ppm as different sets of compliance samples under determined by syringe injection as described the jurisdiction of different enforcement under the condensate trap CO2 purge Proce- agencies, unless prior arrangements are dure, Section 4.3.2. made with both enforcement agencies. After the sample recovery is completed, Calculate the concentrations of the audit use the carrier gas bypass valve to pressurize samples in ppm using the specified sample the ICV to approximately 1060 mm Hg abso- volume in the audit instructions. (NOTE.— lute pressure with zero air. Indication of acceptable results may be ob- 4.4 Analysis. Before putting the NMO ana- tained immediately by reporting the audit lyzer into routine operation, conduct an ini- results in ppm and compliance results in tial performance test. Start the analyzer, ppm by telephone to the responsible enforce- and perform all the necessary functions in ment agency.) Include the results of both order to put the analyzer into proper work- audit samples, their identification numbers, ing order; then conduct the performance test according to the procedures established in and the analyst’s name with the results of Section 5.2. Once the performance test has the compliance determination samples in ap- propriate reports to the EPA regional office been successfully completed and the CO2 and NMO calibration response factors have been or the appropriate enforcement agency dur- determined, proceed with sample analysis as ing the 30-day period. follows: The concentration of the audit samples ob- 4.4.1 Daily Operations and Calibration tained by the analyst shall agree within 20 Checks. Before and immediately after the percent of the actual concentrations. Failure analysis of each set of samples or on a daily to meet the 20-percent specification may re- basis (whichever occurs first), conduct a cali- quire retests until the audit problems are re- bration test according to the procedures es- solved. However, if the audit results do not tablished in Section 5.3. If the criteria of the affect the compliance or noncompliance sta- daily calibration test cannot be met, repeat tus of the affected facility, the Adminis- the NMO analyzer performance test (Section trator may waive the reanalysis require- 5.2) before proceeding. ment, further audits, or retests and accept 4.4.2 Operating Conditions. The carrier the results of the compliance test. While gas flow rate is 29.5 cc/min He and 2.2 cc/min steps are being taken to resolve audit analy- sis problems, the Administrator may also O2. The column oven is heated to 85 °C. The order of elution for the sample from the col- choose to use the data to determine the com- pliance or noncompliance of the affected fa- umn is CO, CH4, CO2, and NMO. 4.4.3 Analysis of Recovered Condensate cility. Sample. Purge the sample loop with sample, 5. Calibration and Operational Checks and then inject the sample. Under the speci- fied operating conditions, the CO2 in the Maintain a record of performance of each sample will elute in approximately 100 sec- item. onds. As soon as the detector response re- 5.1 Initial Performance Check of Conden- turns to baseline following the CO2 peak, sate Recovery Apparatus. Perform these switch the carrier gas flow to backflush, and tests before the system is first placed in op- raise the column oven temperature to 195 °C eration, after any shutdown of 6 months or
953
VerDate 11
more, and after any major modification of 5.1.3.1 50 µl Hexane. the system, or at the specified frequency. 5.1.3.2 10 µl Hexane. 5.1.1 Carrier Gas and Auxiliary O2 Blank 5.1.3.3 50 µl Decane. Check. Analyze each new tank of carrier gas 5.1.3.4 10 µl Decane. or auxiliary O2 with the NMO analyzer to 5.2 Initial NMO Analyzer Performance check for contamination. Treat the gas cyl- Test. Perform these tests before the system inders as noncondensible gas samples, and is first placed in operation, after any shut- analyze according to the procedure in Sec- down longer than 6 months, and after any tion 4.4.3. Add together any measured CH4, major modification of the system. CO, CO2, or NMO. The total concentration 5.2.1 Oxidation Catalyst Efficiency Check. must be less than 5 ppm. Turn off or bypass the NMO analyzer reduc- 5.1.2 Catalyst Efficiency Check. With a tion catalyst. Make triplicate injections of clean condensate trap installed in the recov- the high level methane standard (Section ery system, replace the carrier gas cylinder 3.4.1). The oxidation catalyst operation is ac- with the high level methane standard gas ceptable if the FID response is less than 1 cylinder (Section 3.4.1). Set the four-port percent of the injected methane concentra- valve to the recovery position, and attach an tion. ICV to the recovery system. With the sample 5.2.2 Reduction Catalyst Efficiency recovery valve in vent position and the flow- Check. With the oxidation catalyst unheated control and ICV valves fully open, evacuate or bypassed and the heated reduction cata- the manometer or gauge, the connecting tub- lyst bypassed, make triplicate injections of ing, and the ICV to 10 mm Hg absolute pres- the high level methane standard (Section sure. Close the flow-control and vacuum 3.4.1). Repeat this procedure with both cata- pump valves. lysts operative. The reduction catalyst oper- After the NDIR response has stabilized, ation is acceptable if the response under switch the sample recovery valve from vent both conditions agree within 5 percent. to collect. When the manometer or pressure 5.2.3 Analyzer Linearity Check and NMO gauge begins to register a slight positive Calibration. While operating both the oxida- pressure, open the flow-control valve. Keep tion and reduction catalysts, conduct a lin- the flow adjusted so that atmospheric pres- earity check of the analyzer using the pro- sure is maintained in the system within 10 pane standards specified in Section 3.4.2. percent. Continue collecting the sample in a Make triplicate injections of each calibra- normal manner until the ICV is filled to a tion gas, and then calculate the average re- nominal gauge pressure of 300 mm Hg. Close sponse factor (area/ppm C) for each gas, as the ICV valve, and remove the ICV from the well as the overall mean of the response fac- system. Place the sample recovery valve in tor values. The instrument linearity is ac- the vent position, and return the recovery ceptable if the average response factor of system to its normal carrier gas and normal each calibration gas is within 2.5 percent of operating conditions. Analyze the ICV for the overall mean value and if the relative CO2 using the NMO analyzer; the catalyst ef- standard deviation (Section 6.9) for each set ficiency is acceptable if the CO2 concentra- of triplicate injections is less than 2 percent. tion is within 2 percent of the methane Record the overall mean of the propane re- standard concentration. sponse factor values as the NMO calibration 5.1.3 System Performance Check. Con- response factor (RFNMO). struct a liquid sample injection unit similar Repeat the linearity check using the CO2 in design to the unit shown in Figure 25–7. standards specified in Section 3.4.3. Make Insert this unit into the condensate recovery triplicate injections of each gas, and then and conditioning system in place of a con- calculate the average response factor (area/ densate trap, and set the carrier gas and aux- ppm C) for each gas, as well as the overall iliary O2 flow rates to normal operating lev- mean of the response factor values. Record els. Attach an evacuated ICV to the system, the overall mean of the response factor val- and switch from system vent to collect. With ues as the CO2 calibration response factor the carrier gas routed through the injection (RFCO2). Linearity is acceptable if the aver- unit and the oxidation catalyst, inject a liq- age response factor of each calibration gas is uid sample (See Sections 5.1.3.1 to 5.1.3.4) within 2.5 percent of the overall mean value into the injection port. Operate the trap re- and if the relative standard deviation for covery system as described in Section 4.3.3. each set of triplicate injections is less than Measure the final ICV pressure, and then 2 percent. The RFCO2 must be witnin 10 per- analyze the vessel to determine the CO2 con- cent of the RFNMO. centration. For each injection, calculate the 5.2.4 System Peformance Check. Check percent recovery using the equation in Sec- the column separation and overall perform- tion 6.6. ance of the analyzer by making triplicate in- The performance test is acceptable if the jections of the calibration gases listed in average percent recovery is 100±10 percent Section 3.4.4. The analyzer performance is with a relative standard deviation (Section acceptable if the measured NMO value for 6.9) of less than 5 percent for each set of trip- each gas (average of triplicate injections) is licate injections as follows: within 5 percent of the expected value.
954
VerDate 11
5.3 NMO Analyzer Daily Calibration. N=Carbon number of the liquid compound in-
5.3.1 CO2 Response Factor. Inject trip- jected (N=12 for decane, N=6 for hexane). licate samples of the high level CO2 calibra- Pf =Final pressure of the intermediate collec- tion gas (Section 3.4.3), and calculate the av- tion vessel, mm Hg absolute. erage response factor. The system operation Pb =Barometric pressure, cm Hg. is adequate if the calculated response factor Pti =Gas sample tank pressure before sam- is within 5 percent of the RFCO2 calculated pling, mm Hg absolute. during the initial performance test (Section Pt =Gas sample tank pressure after sampling, 5.2.3). Use the daily response factor (DRFCO2) but before pressurizing, mm Hg absolute. for analyzer calibration and the calculation Ptf =Final gas sample tank pressure after of measured CO2 concentrations in the ICV pressurizing, mm Hg absolute. samples. Tf =Final temperature of intermediate col- 5.3.2 NMO Response Factors. Inject trip- lection vessel, °K. licate samples of the mixed propane calibra- Tti =Sample tank temperature before sam- tion cylinder (Section 3.4.4.1), and calculate pling, °K. the average NMO response factor. The sys- Tt =Sample tank temperature at completion tem operation is adequate if the calculated of sampling, °K. response factor is within 5 percent of the Ttf =Sample tank temperature after pressur- ° RFNMO calculated during the initial perform- izing, K. ance test (Section 5.2.4). Use the daily re- V=Sample tank volume, m3. sponse factor (DRFNMO) for analyzer calibra- Vt =Sample train volume, cc. tion and calculation of NMO concentrations Vv =Intermediate collection vessel volume, in the sample tanks. m3. 3 5.4 Sample Tank and ICV Volume. The Vs =Gas volume sampled, dsm . volume of the gas sampling tanks used must n=Number of data points. be determined. Determine the tank and ICV q=Total number of analyzer injections of in- volumes by weighing them empty and then termediate collection vessel during anal- filled with deionized distilled water; weigh to ysis (where k=injection number, 1 . . . q). the nearest 5 g, and record the results. Alter- r=Total number of analyzer injections of natively, measure the volume of water used sample tank during analysis (where to fill them to the nearest 5 ml. j=injection number, 1 . . . r). xi =Individual measurements. 6. Calculations x¯ =Mean value. ρ=Density of liquid injected, g/cc. All equations are written using absolute Θ=Leak check period, min. pressure; absolute pressures are determined ∆Ρ=Allowable pressure change, cm Hg. by adding the measured barometric pressure 6.2 Allowable Pressure Change. For the to the measured gauge or manometer pres- pretest leak check, calculate the allowable sure. pressure change: 6.1 Nomenclature. C=TGNMO concentration of the effluent, ppm C equivalent.
Cc=Calculated condensible organic (conden- sate trap) concentration of the effluent, ppm C equivalent. 6.3 Sample Volume. For each test run, calculate the gas volume sampled: Ccm=Measured concentration (NMO analyzer) for the condensate trap ICV, ppm CO2. Ct=Calculated noncondensible organic con- centration (sample tank) of the effluent, ppm C equivalent.
Ctm=Measured concentration (NMO analyzer) for the sample tank, ppm NMO. F=Sampling flow rate, cc/min. L=Volume of liquid injected, µl. M=Molecular weight of the liquid injected, g/ g-mole. 6.4 Noncondensible Organics. For each mC =TGNMO mass concentration of the efflu- sample tank, determine the concentration of ent, mg C/dsm3. nonmethane organics (ppm C):
955
VerDate 11
6.5 Condensible Organics. For each con- densate trap determine the concentration of organics (ppm C):
6.6 TGNMO. To determine the TGNMO bon for each test run, use the following equa- concentration for each test run, use the fol- tion:
lowing equation: mc=0.4993 C Eq. 25–6 C=Ct+Cc Eq. 25–5 6.8 Percent Recovery. To calculate the 6.7 TGNMO Mass Concentration. To deter- percent recovery for the liquid injections to the condensate recovery and conditioning mine the TGNMO mass concentration as car- system use the following equation.
M V P C Percent recovery =1604 .v t cm Eq . 25 - 7 L P Tf N
6.9 Relative Standard Deviation. Concentrations by Total Combustion Analy- sis: A Comparison of Infrared with Flame Ionization Detectors. Paper No. 75–33.2. (Pre- sented at the 68th Annual Meeting of the Air Pollution Control Association. Boston, Mas- sachusetts. June 15–20, 1975.) 14 p. 2. Salo, Albert E., William L. Oaks, and Robert D. MacPhee. Measuring the Organic Carbon Content of Source Emissions for Air Pollution Control. Paper No. 74–190. (Pre- 7. Bibliography sented at the 67th Annual Meeting of the Air 1. Salo, Albert E., Samuel Witz, and Robert Pollution Control Association. Denver, Colo- D. MacPhee. Determination of Solvent Vapor rado. June 9–13, 1974.) 25 p.
956
VerDate 11
957
VerDate 11
958
VerDate 11
959
VerDate 11
960
VerDate 11
961
VerDate 11
962
VerDate 11
963
VerDate 11
964
VerDate 11
965
VerDate 11
METHOD 25A—DETERMINATION OF TOTAL GAS- 2. Definitions EOUS ORGANIC CONCENTRATION USING A 2.1 Measurement System. The total equip- FLAME IONIZATION ANALYZER ment required for the determination of the 1. Applicability and Principle gas concentration. The system consists of 1.1 Applicability. This method applies to the following major subsystems: the measurement of total gaseous organic 2.1.1 Sample Interface. That portion of concentration of vapors consisting primarily the system that is used for one or more of of alkanes, alkenes, and/or arenes (aromatic the following: sample acquisition, sample hydrocarbons). The concentration is ex- transportation, sample conditioning, or pro- pressed in terms of propane (or other appro- tection of the analyzer from the effects of priate organic calibration gas) or in terms of the stack effluent. carbon. 2.1.2 Organic Analyzer. That portion of 1.2 Principle. A gas sample is extracted the system that senses organic concentra- from the source through a heated sample tion and generates an output proportional to line, if necessary, and glass fiber filter to a the gas concentration. flame ionization analyzer (FIA). Results are 2.2 Span Value. The upper limit of a gas reported as volume concentration equiva- concentration measurement range that is lents of the calibration gas or as carbon specified for affected source categories in the equivalents. applicable part of the regulations. The span
966
VerDate 11
value is established in the applicable regula- riod of operation during which no unsched- tion and is usually 1.5 to 2.5 times the appli- uled maintenance, repair or adjustment took cable emission limit. If no span value is pro- place. vided, use a span value equivalent to 1.5 to 2.6 Response Time. The time interval 2.5 times the expected concentration. For from a step change in pollutant concentra- convenience, the span value should cor- tion at the inlet to the emission measure- respond to 100 percent of the recorder scale. ment system to the time at which 95 percent 2.3 Calibration Gas. A known concentra- of the corresponding final value is reached as tion of a gas in an appropriate diluent gas. displayed on the recorder. 2.4 Zero Drift. The difference in the meas- 2.7 Calibration Error. The difference be- urement system response to a zero level cali- tween the gas concentration indicated by the bration gas before and after a stated period measurement system and the known con- of operation during which no unscheduled centration of the calibration gas. maintenance, repair, or adjustment took 3. Apparatus place. A schematic of an acceptable measurement 2.5 Calibration Drift. The difference in the system is shown in Figure 25A–1. The essen- measurement system response to a mid-level tial components of the measurement system calibration gas before and after a stated pe- are described below:
3.1 Organic Concentration Analyzer. A 3.3 Sample Line. Stainless steel or flame ionization analyzer (FIA) capable of Teflon* tubing to transport the sample gas meeting or exceeding the specifications in to the analyzer. The sample line should be this method. heated, if necessary, to prevent condensation 3.2 Sample Probe. Stainless steel, or in the line. equivalent, three-hole rake type. Sample 3.4 Calibration Valve Assembly. A three- holes shall be 4 mm in diameter or smaller way valve assembly to direct the zero and and located at 16.7, 50, and 83.3 percent of the calibration gases to the analyzers is rec- equivalent stack diameter. Alternatively, a ommended. Other methods, such as quick- single opening probe may be used so that a connect lines, to route calibration gas to the gas sample is collected from the centrally lo- analyzers are applicable. cated 10 percent area of the stack cross-sec- tion. * Mention of trade names or specific prod- ucts does not constitute endorsement by the Environmental Protection Agency.
967
VerDate 11
3.5 Particulate Filter. An in-stack or an 5.3 Calibration Error. Less than ±5 per- out-of-stack glass fiber filter is rec- cent of the calibration gas value. ommended if exhaust gas particulate loading 6. Pretest Preparations is significant. An out-of-stack filter should 6.1 Selection of Sampling Site. The loca- be heated to prevent any condensation. tion of the sampling site is generally speci- 3.6 Recorder. A strip-chart recorder, ana- fied by the applicable regulation or purpose log computer, or digital recorder for record- of the test; i.e., exhaust stack, inlet line, etc. ing measurement data. The minimum data recording requirement is one measurement The sample port shall be located at least 1.5 value per minute. Note: This method is often meters or 2 equivalent diameters upstream applied in highly explosive areas. Caution of the gas discharge to the atmosphere. and care should be exercised in choice of 6.2 Location of Sample Probe. Install the equipment and installation. sample probe so that the probe is centrally located in the stack, pipe, or duct and is 4. Calibration and Other Gases sealed tightly at the stack port connection. Gases used for calibrations, fuel, and com- 6.3 Measurement System Preparation. bustion air (if required) are contained in Prior to the emission test, assemble the compressed gas cylinders. Preparation of measurement system following the manufac- calibration gases shall be done according to turer’s written instructions in preparing the the procedure in Protocol No. 1, listed in Ci- sample interface and the organic analyzer. tation 2 of Bibliography. Additionally, the Make the system operable. manufacturer of the cylinder should provide FIA equipment can be calibrated for al- a recommended shelf life for each calibration most any range of total organics concentra- gas cylinder over which the concentration tions. For high concentrations of organics does not change more than ±2 percent from (>1.0 percent by volume as propane) modi- the certified value. For calibration gas val- fications to most commonly available ana- ues not generally available (i.e., organics be- lyzers are necessary. One accepted method of tween 1 and 10 percent by volume), alter- equipment modification is to decrease the native methods for preparing calibration gas size of the sample to the analyzer through mixtures, such as dilution systems, may be the use of a smaller diameter sample cap- used with prior approval of the Adminis- illary. Direct and continuous measurement trator. of organic concentration is a necessary con- Calibration gases usually consist of pro- sideration when determining any modifica- pane in air or nitrogen and are determined in tion design. terms of the span value. Organic compounds 6.4 Calibration Error Test. Immediately other than propane can be used following the prior to the test series, (within 2 hours of the above guidelines and making the appropriate start of the test) introduce zero gas and corrections for response factor. high-level calibration gas at the calibration 4.1 Fuel. A 40 percent H2/60 percent He or valve assembly. Adjust the analyzer output 40 percent H2/60 percent N2 gas mixture is to the appropriate levels, if necessary. Cal- recommended to avoid an oxygen synergism culate the predicted response for the low- effect that reportedly occurs when oxygen level and mid-level gases based on a linear concentration varies significantly from a response line between the zero and high-level mean value. responses. Then introduce low-level and mid- 4.2 Zero Gas. High purity air with less level calibration gases successively to the than 0.1 parts per million by volume (ppmv) measurement system. Record the analyzer of organic material (propane or carbon responses for low-level and mid-level calibra- equivalent) or less than 0.1 percent of the tion gases and determine the differences be- span value, whichever is greater. tween the measurement system responses 4.3 Low-level Calibration Gas. An organic and the predicted responses. These dif- calibration gas with a concentration equiva- ferences must be less than 5 percent of the lent to 25 to 35 percent of the applicable span respective calibration gas value. If not, the value. measurement system is not acceptable and 4.4 Mid-level Calibration Gas. An organic must be replaced or repaired prior to testing. calibration gas with a concentration equiva- No adjustments to the measurement system lent to 45 to 55 percent of the applicable span shall be conducted after the calibration and value. before the drift check (Section 7.3). If adjust- 4.5 High-level Calibration Gas. An organic ments are necessary before the completion of calibration gas with a concentration equiva- the test series, perform the drift checks prior lent to 80 to 90 percent of the applicable span to the required adjustments and repeat the value. calibration following the adjustments. If 5. Measurement System Performance Specifica- multiple electronic ranges are to be used, tions each additional range must be checked with 5.1 Zero Drift. Less than ±3 percent of the a mid-level calibration gas to verify the mul- span value. tiplication factor. 5.2 Calibration Drift. Less than ±3 percent 6.5 Response Time Test. Introduce zero of span value. gas into the measurement system at the
968
VerDate 11
calibration valve assembly. When the system vironmental Protection Agency, Environ- output has stabilized, switch quickly to the mental Monitoring and Support Laboratory. high-level calibration gas. Record the time Research Triangle Park, NC. June 1978. from the concentration change to the meas- 3. Gasoline Vapor Emission Laboratory urement system response equivalent to 95 Evaluation—Part 2. U.S. Environmental Pro- percent of the step change. Repeat the test tection Agency, Office of Air Quality Plan- three times and average the results. ning and Standards. Research Triangle Park, 7. Emission Measurement Test Procedure NC. EMB Report No. 75–GAS–6. August 1975. 7.1 Organic Measurement. Begin sampling METHOD 25B—DETERMINATION OF TOTAL GAS- at the start of the test period, recording EOUS ORGANIC CONCENTRATION USING A NON- time and any required process information DISPERSIVE INFRARED ANALYZER as appropriate. In particular, note on the re- cording chart periods of process interruption 1. Applicability and Principle or cyclic operation. 1.1 Applicability. This method applies to 7.2 Drift Determination. Immediately fol- the measurement of total gaseous organic lowing the completion of the test period and concentration of vapors consisting primarily hourly during the test period, reintroduce of alkanes. (Other organic materials may be the zero and mid-level calibration gases, one measured using the general procedure in this at a time, to the measurement system at the method, the appropriate calibration gas, and calibration valve assembly. (Make no adjust- an analyzer set to the appropriate absorption ments to the measurement system until band.) The concentration is expressed in after both the zero and calibration drift terms of propane (or other appropriate or- checks are made.) Record the analyzer re- ganic calibration gas) or in terms of carbon. sponse. If the drift values exceed the speci- 1.2 Principle. A gas sample is extracted fied limits, invalidate the test results pre- from the source through a heated sample ceding the check and repeat the test follow- line, if necessary, and glass fiber filter to a ing corrections to the measurement system. nondispersive infrared analyzer (NDIR). Re- Alternatively, recalibrate the test measure- sults are reported as volume concentration ment system as in Section 6.4 and report the equivalents of the calibration gas or as car- results using both sets of calibration data bon equivalents. (i.e., data determined prior to the test period 2. Definitions and data determined following the test pe- The terms and definitions are the same as riod). for Method 25A. 8. Organic Concentration Calculations 3. Apparatus Determine the average organic concentra- The apparatus is the same as for Method tion in terms of ppmv as propane or other 25A with the exception of the following: calibration gas. The average shall be deter- 3.1 Organic Concentration Analyzer. A mined by the integration of the output re- nondispersive infrared analyzer designed to cording over the period specified in the ap- measure alkane organics and capable of plicable regulation. meeting or exceeding the specifications in If results are required in terms of ppmv as this method. carbon, adjust measured concentrations 4. Calibration Gases using Equation 25A–1. The calibration gases are the same as re- Cc=K Cmeas Eq. 25A–1 quired for Method 25A, Section 4. No fuel gas Where: is required for an NDIR. Cc=Organic concentration as carbon, ppmv. 5. Measurement System Performance Specifica- Cmeas=Organic concentration as measured, tions ppmv. 5.1 Zero Drift. Less than ±3 percent of the K=Carbon equivalent correction factor, span value. K=2 for ethane. 5.2 Calibration Drift. Less than ±3 percent K=3 for propane. of the span value. K=4 for butane. 5.3 Calibration Error. Less than ±5 per- K=Appropriate response factor for other cent of the calibration gas value. organic calibration gases. 6. Pretest Preparations 9. Bibliography 6.1 Selection of Sampling Site. Same as in 1. Measurement of Volatile Organic Com- Method 25A, Section 6.1. pounds—Guideline Series. U.S. Environ- 6.2 Location of Sample Probe. Same as in mental Protection Agency. Research Tri- Method 25A, Section 6.2. angle Park, NC. Publication No. EPA–450/2– 6.3 Measurement System Preparation. 78–041. June 1978. p. 46–54. Prior to the emission test, assemble the 2. Traceability Protocol for Establishing measurement system following the manufac- True Concentrations of Gases Used for Cali- turer’s written instructions in preparing the bration and Audits of Continuous Source sample interface and the organic analyzer. Emission Monitors (Protocol No. 1). U.S. En- Make the system operable.
969
VerDate 11
6.4 Calibration Error Test. Same as in pressure to within 1 mm Hg in the range of Method 25A, Section 6.4. 0 to 1,100 mm Hg. 6.5 Response Time Test Procedure. Same 2.2.4 Sample Tank. Stainless steel or alu- as in Method 25A, Section 6.5. minum cylinder, with a minimum volume of 7. Emission Measurement Test Procedure 4 liters and equipped with a stainless steel sample tank valve. Proceed with the emission measurement 2.3 Vacuum Pump. Capable of evacuating immediately upon satisfactory completion of to an absolute pressure of 10 mm Hg. the calibration. 2.4 Purging Pump. Portable, explosion 7.1 Organic Measurement. Same as in proof, and suitable for sampling NMOC. Method 25A, Section 7.1. 2.5 Pilot Probe Procedure. The following 7.2 Drift Determination. Same as in Meth- are needed only if the tester chooses to use od 25A, Section 7.2. the procedure described in section 4.2.1. 8. Organic Concentration Calculations 2.5.1 Pilot Probe. Tubing of sufficient The calculations are the same as in Meth- strength to withstand being driven into the od 25A, Section 8. landfill by a post driver and an outside diam- 9. Bibliography eter of at least 6.0 millimeters smaller than the sample probe. The pilot probe shall be The bibliography is the same as in Method capped on both ends and long enough to go 25A. through the landfill cover and extend no less METHOD 25C—DETERMINATION OF NONMETH- than 1.0 meter into the landfill. ANE ORGANIC COMPOUNDS (NMOC) IN MSW 2.5.2 Post Driver and Compressor. Capable LANDFILL GASES of driving the pilot probe and the sampling probe into the landfill. 1. Applicability and Principle 2.6 Auger Procedure. The following are needed only if the tester chooses to use the 1.1 Applicability. This method is applica- procedure described in section 4.2.2. ble to the sampling and measurement of non- 2.6.1 Auger. Capable of drilling through methane organic compounds (NMOC) as car- the landfill cover and to a depth of no less bon in MSW landfill gases. than 0.9 meters into the landfill. 1.2 Principle. A sample probe that has 2.6.2 Pea Gravel. been perforated at one end is driven or 2.6.3 Bentonite. augered to a depth of 1.0 meter below the bottom of the landfill cover. A sample of the 2.7 NMOC Analyzer, Barometer, Ther- landfill gas is extracted with an evacuated mometer, and Syringes. Same as in sections cylinder. The NMOC content of the gas is de- 2.3, 2.4.1, 2.4.2, 2.4.4, respectively, of Method termined by injecting a portion of the gas 25. into a gas chromatographic column to sepa- 3. Reagents rate the NMOC from carbon monoxide (CO), carbon dioxide (CO2), and methane (CH4); the 3.1 NMOC Analysis. Same as in Method 25, NMOC are oxidized to CO2, reduced to CH4, section 3.2. and measured by a flame ionization detector 3.2 Calibration. Same as in Method 25, (FID). In this manner, the variable response section 3.4, except omit section 3.4.3. of the FID associated with different types of organics is eliminated. 4. Procedure
2. Apparatus 4.1 Sample Tank Evacuation and Leak Check. Conduct the sample tank evacuation 2.1 Sample Probe. Stainless steel, with and leak check either in the laboratory or the bottom third perforated. The sample the field. Connect the pressure gauge and probe shall be capped at the bottom and sampling valve to the sample tank. Evacuate shall have a threaded cap with a sampling the sample tank to 10 mm Hg absolute pres- attachment at the top. The sample probe sure or less. Close the sampling valve, and shall be long enough to go through and ex- allow the tank to sit for 60 minutes. The tend no less than 1.0 meter below the landfill tank is acceptable if no change is noted. In- cover. If the sample probe is to be driven clude the results of the leak check in the into the landfill, the bottom cap should be test report. designed to facilitate driving the probe into 4.2 Sample Probe Installation. The tester the landfill. may use the procedure in sections 4.2.1 or 2.2 Sampling Train. 4.2.2. CAUTION: Since this method is com- 2.2.1 Rotameter with Flow Control Valve. plex, only experienced personnel should per- Capable of measuring a sample flow rate of form this test. LFG contains methane, there- 500 ml/min or less (30.5±3.1 m3/min). The con- fore explosive mixtures may exist on or near trol valve shall be made of stainless steel. the landfill. It is advisable to take appro- 2.2.2 Sampling Valve. Stainless steel. priate safety precautions when testing land- 2.2.3 Pressure Gauge. U-tube mercury ma- fills, such as refraining from smoking and in- nometer, or equivalent, capable of measuring stalling explosion-proof equipment.
970
VerDate 11
4.2.1 Pilot Probe Procedure. Use the post Place the sample probe in the hole and back- driver to drive the pilot probe at least 1.0 fill with pea gravel to a level 0.6 meters from meter below the landfill cover. Alternative the surface. The sample probe shall protrude procedures to drive the probe into the land- at least 0.3 meters above the landfill cover. fill may be used subject to the approval of Seal the remaining area around the probe the Administrator. with bentonite. Allow 24 hours for the land- Remove the pilot probe and drive the sam- fill gases to equilibrate inside the augered ple probe into the hole left by the pilot probe before sampling. probe. The sample probe shall extend not less 4.3 Sample Train Assembly. Prepare the than 1.0 meter below the landfill cover and sample by evacuating and filling the sample shall protrude about 0.3 meters above the tank with helium three times. After the landfill cover. Seal around the sampling third evacuation, charge the sample tank probe with bentonite and cap the sampling with helium to a pressure of approximately probe with the sampling probe cap. 325 mm Hg. Record the pressure, the ambient 4.2.2 Auger Procedure. Use an auger to temperature, and the barometric pressure. drill a hole through the landfill cover and to Assemble the sampling probe purging system at least 1.0 meter below the landfill cover. as shown in figure 1.
4.4 Sampling Procedure. Open the sam- use the carrier gas bypass valve to pressurize pling valve and use the purge pump and the the sample cylinder to approximately 1,060 flow control valve to evacuate at least two mm Hg absolute pressure with helium and sample probe volumes from the system at a record the final pressure. Alternatively, the flow rate of 500 ml/min or less (30.5±3.1 m3/ sample tank may be pressurized in the lab. If min). Close the sampling valve and replace not analyzing for N2, the sample cylinder the purge pump with the sample tank appa- may be pressurized with zero air. Use Method ratus as shown in figure 2. Open the sampling 3C to determine the percent N in the sam- valve and the sample tank valves and, using 2 ple. Presence of N indicates infiltration of the flow control valve, sample at a flow rate 2 ambient air into the gas sample. The landfill of 500 ml/min or less (30.5±3.1 m3/min) until the sample tank gauge pressure is zero. Dis- sample is acceptable if the concentration of connect the sampling tank apparatus and N2 is less than 20 percent.
971
VerDate 11
4.5 Analysis. The oxidation, reduction, quate. Record the value obtained for any and measurement of NMOC is similar to measured NMOC. Return the column oven Method 25. Before putting the NMOC ana- temperature to 85 °C in preparation for the lyzer into routine operation, conduct an ini- next analysis. Analyze each sample in trip- tial performance test. Start the analyzer, licate, and report the average as Ctm. and perform all the necessary functions to 4.6 Audit Samples. Same as in Method 25, put the analyzer into proper working order. section 4.5. Conduct the performance test according to 4.7 Deactivation of Sample Probe Holes. the procedures established in section 5.1. Once sampling has taken place, either plug Once the performance test has been success- the sampling probes with a cap or remove fully completed and the NMOC calibration the probes and refill the hole with cover ma- response factor has been determined, proceed terial. with sample analysis as follows: 4.5.1 Daily Operations and Calibration 5. Calibration and Operational Checks Checks. Before and immediately after the Maintain a record of performance of each analysis of each set of samples or on a daily item. basis (whichever occurs first), conduct a cali- 5.1 Initial NMOC Analyzer Performance bration test according to the procedures es- Test. Same as in Method 25, section 5.2, ex- tablished in section 5.2. If the criteria of the cept omit the linearity checks for CO2 stand- daily calibration test cannot be met, repeat ards. the NMOC analyzer performance test (sec- 5.2 NMOC Analyzer Daily Calibration. tion 5.1) before proceeding. NMOC response factors, same as in Method 4.5.2 Operating Conditions. Same as in 25, section 5.3.2. Method 25, section 4.4.2. 4.5.3 Analysis of Sample Tank. Purge the 6. Calculations sample loop with sample, and then inject the sample. Under the specified operating condi- All equations are written using absolute pressure; absolute pressures are determined tions, the CO2 in the sample will elute in ap- proximately 100 seconds. As soon as the de- by adding the measured barometric pressure tector response returns to baseline following to the measured gauge of manometer pres- sure. the CO2 peak, switch the carrier gas flow to backflush, and raise the column oven tem- 6.1 Nomenclature. perature to 195 °C as rapidly as possible. A Bw=moisture content in the sample, fraction rate of 30 °C/min has been shown to be ade- CN2=measured N2 concentration, fraction 972
VerDate 11
Ct=calculated NMOC concentration, ppmv C Vapor Concentrations by Total Combustion equivalent Analysis: A Comparison of Infrared with Ctm=measured NMOC concentration, ppmv C Flame Ionization Detectors. Paper No. 75– equivalent 33.2. (Presented at the 68th Annual Meeting Pb=barometric pressure, mm Hg of the Air Pollution Control Association. Pti=gas sample tank pressure before sam- Boston, Massachusetts. June 15–20, 1975.) p. pling, mm Hg absolute 14. Pt=gas sample tank pressure at completion 2. Salon, Albert E., William L. Oaks, and of sampling, but before pressurizing, mm Robert D. MacPhee. Measuring the Organic Hg absolute Carbon Content of Source Emissions for Air Ptf=final gas sample tank pressure after pres- Pollution Control. Paper No. 74–190. (Pre- surizing, mm Hg absolute sented at the 67th Annual Meeting of the Air Pw=vapor pressure of H2O (from table 25C–1), Pollution Control Association. Denver, Colo- mm Hg rado. June 9–13, 1974.) p. 25. Tti=sample tank temperature before sam- pling, °K METHOD 25D—DETERMINATION OF THE VOLA- TILE ORGANIC CONCENTRATION OF WASTE Tt=sample tank temperature at completion of sampling, but before pressuring, °K SAMPLES Ttf=sample tank temperature after pressur- izing, °K Introduction r=total number of analyzer injections of Performance of this method should not be sample tank during analysis (where attempted by persons unfamiliar with the j=injection number, 1. . .r) operation of a flame ionization detector 6.2 Water Correction. Use table 25C–1, the (FID) or an electrolytic conductivity detec- LFG temperature, and barometric pressure tor (ELCD) because knowledge beyond the at the sampling site to calculate Bw. scope of this presentation is required. P 1. Applicability and Principle = w Bw 1.1 Applicability. This method is applica- Pb ble for determining the volatile organic (VO) concentration of a waste sample. TABLE 25C±1.ÐMOISTURE CORRECTION 1.2 Principle. A sample of waste is ob- tained at a point which is most representa- Vapor Pres- tive of the unexposed waste (where the waste Temperature, °C sure of has had minimum opportunity to volatilize H2O, mm Hg to the atmosphere). The sample is suspended in an organic/aqueous matrix, then heated 4 ...... 6.1 and purged with nitrogen for 30 min in order 6 ...... 7.0 to separate certain organic compounds. Part 8 ...... 8.0 of the sample is analyzed for carbon con- 1 ...... 9.2 centration, as methane, with an FID, and 12 ...... 10.5 14 ...... 12.0 part of the sample is analyzed for chlorine 16 ...... 13.6 concentration, as chloride, with an ELCD. 18 ...... 15.5 The VO concentration is the sum of the car- 20 ...... 17.5 bon and chlorine content of the sample. 22 ...... 19.8 24 ...... 22.4 2. Apparatus 26 ...... 25.2 28 ...... 28.3 2.1 Sampling. The following equipment is 30 ...... 31.8 required: 2.1.1 Sampling Tube. Flexible Teflon, 0.25 6.3 NMOC Concentration. Use the follow- in. ID. ing equation to calculate the concentration NOTE: Mention of trade names or specific of NMOC for each sample tank. products does not constitute endorsement by the Environmental Protection Agency. Ptf 2.1.2 Sample Container. Borosilicate T 1 r glass, 40 mL, and a Teflon lined screw cap ca- C = tf ∑C ()j pable of forming an air tight seal. t tm 2.1.3 Cooling Coil. Fabricated from 0.25 in. Pt Pti − − − ()1 Bw C N2 r j=1 ID 304 stainless steel tubing with a thermo- T T couple at the coil outlet. t ti 2.2 Analysis. The following equipment is required: 7. Bibliography 2.2.1 Purging Apparatus. For separating 1. Salon, Albert E., Samuel Witz, and Rob- the VO from the waste sample. A schematic ert D. MacPhee. Determination of Solvent of the system is shown in Figure 25D–1. The
973
VerDate 11
purging apparatus consists of the following major components.
2.2.1.1 Purging Flask. A glass container to drical glass tube. One end of the tube is open hold the sample while it is heated and purged while the other end is sealed. Exact dimen- with dry nitrogen. The cap of the purging sions are shown in Figure 25D–2. flask is equipped with three fittings: one for 2.2.1.2 Purging Lance. Glass tube, 6-mm a purging lance (fitting with the #7 Ace- OD by 30 cm long. The purging end of the thread), one for the Teflon exit tubing (side tube is fitted with a four-arm bubbler with fitting, also a #7 Ace-thread), and a third (a each tip drawn to an opening 1 mm in diame- 50-mm Ace-thread) to attach the base of the ter. purging flask as shown in Figure 25D–2. The Details and exact dimensions are shown in base of the purging flask is a 50-mm ID cylin- Figure 25D–2.
974
VerDate 11
975
VerDate 11
2.2.1.3 Coalescing Filter. Porous fritted details of the design are shown in Figure disc incorporated into a container with the 25D–3. same dimensions as the purging flask. The
2.2.1.4 Constant Temperature Chamber. A 2.2.1.5 Three-way Valve. Manually oper- forced draft oven capable of maintaining a ated, stainless steel. To introduce calibra- uniform temperature around the purging tion gas into system. flask and coalescing filter of 75±2°C. 2.2.1.6 Flow Controllers. Two, adjustable. One capable of maintaining a purge gas flow
976
VerDate 11
rate of 6±.06 L/min. The other capable of ELCD to measure the chlorine concentra- maintaining a calibration gas flow rate of 1– tion. 100 mL/min. 2.2.2.1 FID. A heated FID meeting the fol- 2.2.1.7 Rotameter. For monitoring the air lowing specifications is required. flow through the purging system (0–10 L/ 2.2.2.1.1 Linearity. A linear response (+ 5 min). percent) over the operating range as dem- 2.2.1.8 Sample Splitters. Two heated flow onstrated by the procedures established in restrictors (placed inside oven or heated to Section 5.1.1. 120±10°C). At a purge rate of 6 L/min, one will 2.2.2.1.2 Range. A full scale range of 50 pg supply a constant flow to the first detector carbon/sec to 50 µKg carbon/sec. Signal at- (the rest of the flow will be directed to the tenuators shall be available to produce a second sample splitter). The second splitter minimum signal response of 10 percent of will split the analytical flow between the full scale. second detector and the flow restrictor. The 2.2.2.1.3 Data Recording System. A digital approximate flow to the FID will be 40 mL/ integration system compatible with the FID min and to the ELCD will be 15 mL/min, but for permanently recording the output of the the exact flow must be adjusted to be com- detector. The recorder shall have the capa- patible with the individual detector and to bility to start and stop integration at points meet its linearity requirement. The two selected by the operator or it shall be capa- sample splitters will be connected to each ble of the ‘‘integration by slices’’ technique other by 1⁄8″ OD stainless steel tubing. (this technique involves breaking down the 2.2.1.9 Flow Restrictor. Stainless steel chromatogram into smaller increments, in- tubing, 1⁄8″ OD, connecting the second sample tegrating the area under the curve for each splitter to the ice bath. Length is deter- portion, subtracting the background for each mined by the resulting pressure in the purg- portion, and then adding all of the areas to- ing flask (as measured by the pressure gether for the final area count). gauge). The resulting pressure from the use 2.2.2.2 ELCD. An ELCD meeting the fol- of the flow restrictor shall be 6–7 psiG. lowing specifications is required. The ELCD 2.2.1.10 Filter Flask. With one-hole stop- components shall consist of quartz reactor per. Used to hold ice bath. Excess purge gas tubing and 1-propanol as electrolyte. The is vented through the flask to prevent con- electrolyte flow through the conductivity densation in the flowmeter and to trap vola- cell shall be 1 to 2 mL/min. tile organic compounds. NOTE: A 1⁄4-in. ID quartz reactor tube is 2.2.1.11 Four-way Valve. Manually oper- recommended to reduce carbon buildup and ated, stainless steel. Placed inside oven, used the resulting detector maintenance. to bypass purging flask. 2.2.2.2.1 Linearity. A linear response (± 10 2.2.1.12 On/Off Valves. Two, stainless percent) over the response range as dem- steel. One heat resistant up to 130°C and onstrated by the procedures in Section 5.1.2. placed between oven and ELCD. The other a 2.2.2.2.2 Range. A full scale range of 5.0 pg/ toggle valve used to control purge gas flow. sec to 500 ng/sec chloride. Signal attenuators 2.2.1.13 Pressure Gauge. Range 0–40 psi. To shall be available to produce a minimum sig- monitor pressure in purging flask and coa- nal response of 10 percent of full scale. lescing filter. 2.2.2.2.3 Data Recording System. A digital 2.2.1.14 Sample Lines. Teflon, 1/4≤″ OD, integration system compatible with the out- used inside the oven to carry purge gas to put voltage range of the ELCD. The recorder and from purging chamber and to and from must have the capability to start and stop coalescing filter to four-way valve. Also used integration at points selected by the opera- to carry sample from four-way valve to first tor or it shall be capable of performing the sample splitter. ‘‘integration by slices’’ technique. 2.2.1.15 Detector Tubing. Stainless steel, 3. Reagents 1⁄8″ OD, heated to 120±10°C. Used to carry sample gas from each sample splitter to a de- 3.1 Sampling. tector. Each piece of tubing must be wrapped 3.1.1 Polyethylene Glycol (PEG). Ninety- with heat tape and insulating tape in order eight percent pure with an average molecu- to insure that no cold spots exist. The tubing lar weight of 400. Before using the PEG, re- leading to the ELCD will also contain a heat- move any organic compounds that might be resistant on-off valve (Section 2.2.1.12) which detected as volatile organics by heating it to shall also be wrapped with heat-tape and in- 120°C and purging it with nitrogen at a flow sulation. rate of 1 to 2 L/min for 2 hours. The cleaned 2.2.2 Volatile Organic Measurement Sys- PEG must be stored under a 1 to 2 L/min ni- tem. Consisting of an FID to measure the trogen purge until use. The purge apparatus carbon concentration of the sample and an is shown in Figure 25D–4.
977
VerDate 11
3.2 Analysis. percent 1,1-dichloroethylene by volume in ni- 3.2.1 Sample Separation. The following trogen. are required for the sample purging step. 3.2.2.4 Water. Deionized distilled water 3.2.1.1 PEG. Same as Section 3.1.1. that conforms to American Society for Test- 3.2.1.2 Purge Gas. Zero grade nitrogen ing and Materials Specification D 1193–77, (N2), containing less than 1 ppm carbon. Type 3 (incorporated by reference as speci- 3.2.2 Volatile Organics Measurement. The fied in § 60.17), is required for analysis. At the following are required for measuring the VO option of the analyst, the KMnO4 test for ox- concentration. idizable organic matter may be omitted
3.2.2.1 Hydrogen (H2). Zero grade H2, 99.999 when high concentrations are not expected percent pure. to be present. 3.2.2.2 Combustion Gas. Zero grade air or 3.2.2.5 1–Propanol. ACS grade or better. oxygen as required by the FID. Electrolyte Solution. For use in the ELCD. 3.2.2.3 Calibration Gas. Pressurized gas cylinder containing 10 percent propane and 1
978
VerDate 11
4. Procedure fers to turbulent flow which results in mul- 4.1 Sampling. tiple-phase waste in effect behaving as sin- 4.1.1 Sampling Plan Design and Develop- gle-phase waste due to good mixing. ment. Use the procedures in chapter nine of 4.1.2.1 Install a sampling tap to obtain the the Office of Solid Waste’s publication, Test sample at a point which is most representa- Methods for Evaluating Solid Waste, third edi- tive of the unexposed waste (where the waste tion (SW–846), as guidance in developing a has had minimum opportunity to volatilize sampling plan. to the atmosphere). Assemble the sampling 4.1.2 Single Phase or Well-mixed Waste. apparatus as shown in Figure 25D–5. Well-mixed in the context of this method re-
4.1.2.2 Prepare the sampling containers as 4.1.3 Multiple-phase Waste. Collect a 10 g follows: Pour 30 mL of clean PEG into the sample of each phase of waste generated container. PEG will reduce but not eliminate using the procedures described in Section the loss of organics during sample collection. 4.1.2 or 4.1.5. Each phase of the waste shall be Weigh the sample container with the screw analyzed as a separate sample. Calculate the cap, the PEG, and any labels to the nearest weighted average VO concentration of the 0.01 g and record the weight (mst). Store the waste using Equation 13 (Section 6.14). containers in an ice bath until 1 h before 4.1.4 Solid waste. Add approximately 10 g sampling (PEG will solidify at ice bath tem- of the solid waste to a container prepared in peratures; allow the containers to reach the manner described in Section 4.1.2.2, mini- room temperature before sampling). mizing headspace. Cap and chill imme- 4.1.2.3 Begin sampling by purging the diately. sample lines and cooling coil with at least 4.1.5 Alternative to Tap Installation. If four volumes of waste. Collect the purged tap installation is impractical or impossible, material in a separate container and dispose fill a large, clean, empty container by sub- of it properly. merging the container into the waste below 4.1.2.4 After purging, stop the sample flow the surface of the waste. Immediately fill a and direct the sampling tube to a preweighed container prepared in the manner described sample container, prepared as described in in Section 4.1.2.2 with approximately 10 g of Section 4.1.2.2. Keep the tip of the tube below the waste collected in the large container. the surface of the PEG during sampling to Minimize headspace, cap and chill imme- minimize contact with the atmosphere. Sam- diately. ple at a flow rate such that the temperature 4.1.6 Alternative sampling techniques of the waste is less than 10°C. Fill the sample may be used upon the approval of the Admin- container and immediately cap it (within 5 istrator. seconds) so that a minimum headspace exists 4.2 Sample Recovery. in the container. Store immediately in a 4.2.1 Assemble the purging apparatus as cooler and cover with ice. shown in Figures 25D–1 and 25D–2. The oven
979
VerDate 11
shall be heated to 75 ± 2°C. The sampling sponse of the FID and the ELCD. Monitor the lines leading from the oven to the detectors readings of the pressure gauge and the ro- shall be heated to 120 ± 10°C with no cold tameter. If the readings fall below estab- spots. The flame ionization detector shall be lished setpoints, stop the purging, determine operated with a heated block. Adjust the the source of the leak, and resolve the prob- purging lance so that it reaches the bottom lem before resuming. Leaks detected during of the chamber. a sampling period invalidate that sample. 4.2.2 Remove the sample container from 4.3.2 As the purging continues, monitor the cooler, and wipe the exterior of the con- the output of the detectors to make certain tainer to remove any extraneous ice, water, that the analysis is proceeding correctly and or other debris. Reweigh the sample con- that the results are being properly recorded. tainer to the nearest 0.01 g, and record the Every 10 minutes read and record the purge weight (msf). Pour the contents of the sample flow rate, the pressure and the chamber tem- container into the purging flask, rinse the perature. Continue the purging for 30 min- sample container three times with a total of utes. 20 mL of PEG (since the sample container 4.3.3 For each detector output, integrate originally held 30 mL of PEG, the total vol- over the entire area of the peak starting at ume of PEG added to the purging flask will 1 minute and continuing until the end of the be 50 mL), transferring the rinsings to the run. Subtract the established baseline area purging flask after each rinse. Cap purging from the peak area. Record the corrected flask between rinses. The total volume of area of the peak. See Figure 25D–6 for an ex- PEG in the purging flask shall be 50 mL. Add ample integration. 50 mL of water to the purging flask. 4.4 Water Blank. A water blank shall be 4.3 Sample Analysis. analyzed for each batch of cleaned PEG pre- 4.3.1 Turn on the constant temperature pared. Transfer about 60 mL of water into chamber and allow the temperature to the purging flask. Add 50 mL of the cleaned equilibrate at 75 ± 2°C. Turn the four-way PEG to the purging flask. Treat the blank as valve so that the purge gas bypasses the described in Sections 4.2 and 4.3, excluding purging flask, the purge gas flowing through Section 4.2.2. Calculate the concentration of the coalescing filter and to the detectors carbon and chlorine in the blank sample (as- (standby mode). Turn on the purge gas. sume 10 g of waste as the mass). A VO con- Allow both the FID and the ELCD to warm centration equivalent to ≤10 percent of the up until a stable baseline is achieved on each applicable standard may be subtracted from detector. Pack the filter flask with ice. Re- the measured VO concentration of the waste place ice after each run and dispose of the samples. Include all blank results and docu- waste water properly. When the temperature mentation in the test report. of the oven reaches 75±2°C, start both inte- 5. Operational Checks and Calibration. grators and record baseline. After 1 min, turn the four-way valve so that the purge gas Maintain a record of performance of each flows through the purging flask, to the coa- item. lescing filter and to the sample splitters 5.1 Initial Performance Check of Purging (purge mode). Continue recording the re- System.
980
VerDate 11
Before placing the system in operation, after valve. Set the four-way bypass valve to a shutdown of greater than six months, after standby position so that the calibration gas any major modifications, and at least once flows through the coalescing filter only. In- per month during continuous operation, con- ject the calibration gas by turning the cali- duct the linearity checks described in Sec- bration gas valve from vent position to in- tions 5.1.1 and 5.1.2. Install calibration gas at ject position. Continue the calibration gas the three-way calibration gas valve. See Fig- flow for the appropriate period of time before ure 25D–1. switching the calibration valve to vent posi- 5.1.1 Linearity Check Procedure. Using tion. Continue recording the response of the the calibration standard described in Section FID and the ELCD for 5 min after switching 3.2.2.3 and by varying the injection time, it is off calibration gas flow. Make triplicate in- possible to calibrate at multiple concentra- jections of all six levels of calibration. tion levels. Use Equation 3 to calculate three 5.1.2 Linearity Criteria. Calculate the av- sets of calibration gas flow rates and run erage response factor (Equations 5 and 6) and times needed to introduce a total methane the relative standard deviation (RSD) (Equa- mass (mco) of 1, 5, and 10 mg into the system tion 10) at each level of the calibration curve (low, medium and high FID calibration, re- for both detectors. Calculate the overall spectively). Use Equation 4 to calculate mean of the three response factor averages three sets of calibration gas flow rates and for each detector. The FID linearity is ac- run times needed to introduce a total chlo- ceptable if each response factor is within 5 ride mass (mch) of 1, 5, and 10 mg into the sys- percent of the overall mean and if the RSD tem (low, medium and high ELCD calibra- for each set of triplicate injections is less tion, respectively). With the system operat- than 5 percent. The ELCD linearity is ac- ing in standby mode, allow the FID and the ceptable if each response factor is within 10 ELCD to establish a stable baseline. Set the percent of the overall mean and if the RSD secondary pressure regulator of the calibra- for each set of triplicate injections is less tion gas cylinder to the same pressure as the than 10 percent. Record the overall mean purge gas cylinder and set the proper flow value of the response factors for the FID and rate with the calibration flow controller (see the ELCD. If the calibration for either the Figure 25D–1). The calibration gas flow rate FID or the ELCD does not meet the criteria, can be measured with a flowmeter attached correct the detector/system problem and re- to the vent position of the calibration gas peat Sections 5.1.1 and 5.1.2.
981
VerDate 11
5.2 Daily Calibrations. tests. The availability of audit samples may 5.2.1 Daily Linearity Check. Follow the be determined by writing: Source Test Audit procedures outlined in Section 5.1.1 to ana- Coordinator (MD–77B), Quality Assurance Di- lyze the medium level calibration for both vision, Atmospheric Research and Exposure the FID and the ELCD in duplicate at the Assessment Laboratory, U.S. Environmental start of the day. Calculate the response fac- Protection Agency, Research Triangle Park, tors and the RSDs for each detector. For the NC 27711 or by calling the Source Test Audit FID, the calibration is acceptable if the av- Coordinator (STAC) at (919) 541–7834. The re- erage response factor is within 5 percent of quest for the audit sample must be made at the overall mean response factor (Section least 30 days prior to the scheduled compli- 5.1.2) and if the RSD for the duplicate injec- ance sample analysis. If audit samples are tion is less than 5 percent. For the ELCD, not available, follow the quality control the calibration is acceptable if the average sample procedures in Section 5.7. response factor is within 10 percent of the overall mean response factor (Section 5.1.2) 5.6 Audit Results. Calculate the audit and if the RSD for the duplicate injection is sample concentration according to the cal- less than 10 percent. If the calibration for ei- culation procedure described in the audit in- ther the FID or the ELCD does not meet the structions included with the audit sample. criteria, correct the detector/system problem Fill in the audit sample concentration and and repeat Sections 5.1.1 and 5.1.2. the analyst’s name on the audit response 5.2.2 Calibration Range Check. form included with the audit instructions. 5.2.2.1 If the waste concentration for ei- Send one copy to the EPA Regional Office or ther detector falls below the range of cali- the appropriate enforcement agency and a bration for that detector, use the procedure second copy to the STAC. The EPA Regional outlined in Section 5.1.1 to choose 2 calibra- office or the appropriate enforcement agency tion points that bracket the new target con- will report the results of the audit to the centration. Analyze each of these points in laboratory being audited. Include this re- triplicate (as outlined in Section 5.1.1) and sponse with the results of the compliance use the criteria in Section 5.1.2 to determine samples in relevant reports to the EPA Re- the linearity of the detector in this ‘‘mini- gional Office or the appropriate enforcement calibration’’ range. agency. 5.2.2.2 After the initial linearity check of 5.7 Quality Control Samples. If audit sam- the minicalibration curve, it is only nec- ples are not available, prepare and analyze essary to test one of the points in duplicate the two types of quality control samples for the daily calibration check (in addition (QCS) listed in Sections 5.7.1 and 5.7.2. Before to the points specified in Section 5.2.1). The placing the system in operation, after a average daily mini-calibration point should shutdown of greater than six months, and fit the linearity criteria specified in Section after any major modifications, analyze each 5.2.1. If the calibration for either the FID or QCS in triplicate. For each detector, cal- the ELCD does not meet the criteria, correct culate the percent recovery by dividing the detector/system problem and repeat the measured concentration by theoretical con- calibration procedure mentioned in the first centration and multiplying by 100. Deter- paragraph of Section 5.2.2. A mini-calibra- mine the mean percent recovery for each de- tion curve for waste concentrations above tector for each QCS triplicate analysis. The the calibration curve for either detector is ≤ optional. RSD for any triplicate analysis shall be 10 5.3 Analytical Balance. Calibrate against percent. For QCS 1 (methylene chloride), the ≥ standard weights. percent recovery shall be 90 percent for car- ≥ 5.4 Audit Procedure. Concurrently ana- bon as methane, and 55 percent for chlorine lyze the audit sample and a set of compli- as chloride. For QCS 2 (1,3-dichloro-2-pro- ≤ ance samples in the same manner to evalu- panol), the percent recovery shall be 15 per- ate the technique of the analyst and the cent for carbon as methane, and ≤6 percent standards preparation. The same analyst, an- for chlorine as chloride. If the analytical sys- alytical reagents, and analytical system tem does not meet the above-mentioned cri- shall be used both for compliance samples teria for both detectors, check the system and the EPA audit sample. If this condition parameters (temperature, system pressure, is met, auditing of subsequent compliance purge rate, etc.), correct the problem, and re- analyses for the same enforcement agency peat the triplicate analysis of each QCS. within 30 days is not required. An audit sam- 5.7.1 QCS 1, Methylene Chloride. Prepare ple set may not be used to validate different a stock solution by weighing, to the nearest sets of compliance samples under the juris- 0.1 mg, 55 µL of HPLC grade methylene chlo- diction of different enforcement agencies, ride in a tared 5 mL volumetric flask. Record unless prior arrangements are made with the weight in milligrams, dilute to 5 mL both enforcement agencies. with cleaned PEG, and inject 100 µL of the 5.5 Audit Samples. Audit Sample Avail- stock solution into a sample prepared as a ability. Audit samples will be supplied only water blank (50 mL of cleaned PEG and 60 to enforcement agencies for compliance mL of water in the purging flask). Analyze
982
VerDate 11
the QCS according to the procedures de- Cj=VO concentration of phase j, ppmw. scribed in Sections 4.2 and 4.3, excluding Sec- DRt=Average daily response factor of the tion 4.2.2. To calculate the theoretical car- FID, mg CH4 counts. bon concentration (in mg) in QCS 1, multiply DRth=Average daily response factor of the mg of methylene chloride in the stock solu- ELCD, mg Cl¥ counts. ¥3 tion by 3.777 × 10 . To calculate the theo- Fj= Weight fraction of phase j present in the retical chlorine concentration (in mg) in waste. QCS 1, multiply mg of methylene chloride in mco=Mass of carbon, as methane, in a cali- the stock solution by 1.670 × 10 ¥2. bration run, mg. 5.7.2 QCS 2, 1,3-dichloro-2-propanol. Pre- mch=Mass of chloride in a calibration run, pare a stock solution by weighing, to the mg. µ nearest 0.1 mg, 60 L of high purity grade 1,3- ms=Mass of the waste sample, g. dichloro-2-propanol in a tared 5 mL volu- msc=Mass of carbon, as methane, in the sam- metric flask. Record the weight in milli- ple, mg. grams, dilute to 5 mL with cleaned PEG, and msf=Mass of sample container and waste inject 100 µL of the stock solution into a sample, g. sample prepared as a water blank (50 mL of msh=Mass of chloride in the sample, mg. cleaned PEG and 60 mL of water in the purg- mst=Mass of sample container prior to sam- ing flask). pling, g. Analyze the QCS according to the proce- mvo=Mass of volatile organics in the sample, dures described in Sections 4.2 and 4.3, ex- mg. cluding Section 4.2.2. To calculate the theo- n=Total number of phases present in the retical carbon concentration (in mg) in QCS waste. 2, multiply mg of 1,3-dichloro-2-propanol in Pp=Percent propane in calibration gas (L/L). × ¥3 the stock solution by 7.461 10 . To cal- Pvc=Percent 1,1-dichloroethylene in calibra- culate the theoretical chlorine concentra- tion gas (L/L). tion (in mg) in QCS 2, multiply mg of 1,3- Qc=Flow rate of calibration gas, L/min. dichloro-2-propanol in the stock solution by tc=Length of time standard gas is delivered 1.099 × 10 ¥2. to the analyzer, min. 5.7.3 Routine QCS Analysis. For each set W=Weighted average VO concentration, of compliance samples (in this context, set is ppmw. per facility, per compliance test), analyze 6.2 Concentration of Carbon, as Methane, one QCS 1 and one QCS 2 sample. The per- in the Calibration Gas. cent recovery for each sample for each detec- C =(19.681 × P ) + (13.121 × P ) Eq. 1 ± c p vc tor shall be 13 percent of the mean recov- 6.3 Concentration of Chloride in the Cali- ery established for the most recent set of bration Gas. QCS triplicate analysis (Section 5.7). If the × sample does not meet this criteria, check the Ch=28.998 Pvc Eq. 2 system components and analyze another QCS 6.4 Mass of Carbon, as Methane, in a Cali- 1 and 2 until a single set of QCS meet the ± bration Run. 13 percent criteria. mco=Cc × Qc × tc Eq. 3 6.5 Mass of Chloride in a Calibration Run. 6. Calculations mch=Cch × Qc × tc Eq. 4 6.1 Nomenclature. 6.6 FID Response Factor, mg/counts.
Ab=Area under the water blank response Rt=mco/Ac Eq. 5 curve, counts. 6.7 ELCD Response Factor, mg/counts. A =Area under the calibration response c Rth=mch/Ac Eq. 6 curve, counts. 6.8 Mass of Carbon in the Sample. A =Area under the sample response curve, s m =DR (A ¥A ) Eq. 7 counts. sc t s b 6.9 Mass of Chloride in the Sample. C=Concentration of volatile organics in the sample, ppmw. msh=DRth (As¥Ab) Eq. 8 6.10 Mass of Volatile Organics in the Sam- Cc=Concentration of carbon, as methane, in the calibration gas, mg/L. ple. Chh=Concentration of chloride in the calibra- mvo=msc + msh Eq. 9 tion gas, mg/L. 6.11 Relative Standard Deviation.
n − 2 ∑()XXi 100 RSD = i=1 Eq. 10 X n −1
983 Pt. 60, App. A, Meth. 25E 40 CFR Ch. I (7–1–98 Edition)
6.12 Mass of Sample. C=(mvo x 1000)/ms Eq. 12
ms=msf¥mst Eq. 11 6.14 Weighted Average VO Concentration 6.13 Concentration of Volatile Organics in of Multi-phase Waste. Waste.
n = × W∑ Fj Cj Eq. 13 j=1
METHOD 25E—DETERMINATION OF VAPOR 3.1.4.3 Cooling Coil. Stainless steel (304), PHASE ORGANIC CONCENTRATION IN WASTE 0.25 in.–ID, equipped with a thermocouple at SAMPLES the coil outlet. 3.2 Analysis. The following equipment is Introduction required: Performance of this method should not be 3.2.1 Balanced Pressure Headspace Sam- attempted by persons unfamiliar with the pler. Perkin-Elmer HS–6, HS–100, or equiva- operation of a flame ionization detector lent, equipped with a glass bead column in- (FID) nor by those who are unfamiliar with stead of a chromatographic column. source sampling because knowledge beyond 3.2.2 FID. An FID meeting the following the scope of this presentation is required. specifications is required: 3.2.2.1 Linearity. A linear response (±5 1. Applicability and Principle percent) over the operating range as dem- 1.1 Applicability. This method is applica- onstrated by the procedures established in ble for determining the vapor pressure of Section 6.1.2. waste samples which represent waste which 3.2.2.2 Range. A full scale range of 1 to is or will be managed in tanks. 10,000 ppm CH4. Signal attenuators shall be 1.2 Principle. The headspace vapor of the available to produce a minimum signal re- sample is analyzed for carbon content by a sponse of 10 percent of full scale. headspace analyzer, which uses an FID. 3.2.3 Data Recording System. Analog strip chart recorder or digital integration system 2. Interferences compatible with the FID for permanently re- 2.1 The analyst shall select the operating cording the output of the detector. parameters best suited to the requirements 3.2.4 Thermometer. Capable of reading for a particular analysis. The analyst shall temperatures in the range of 30° to 60°C with produce confirming data through an ade- an accuracy of ±0.1°C. quate supplemental analytical technique and have the data available for review by the Ad- 4. Reagents ministrator. 4.1 Analysis. The following items are re- 3. Apparatus quired for analysis: 4.1.1 Hydrogen (H2). Zero grade. 3.1 Sampling. The following equipment is 4.1.2 Carrier Gas. Zero grade nitrogen, required: containing less than 1 ppm carbon (C) and 3.1.1 Sample Containers. Vials, glass, with less than 1 ppm carbon dioxide. butyl rubber septa, Perkin-Elmer Corpora- 4.1.3 Combustion Gas. Zero grade air or tion Numbers 0105–0129 (glass vials), B001– oxygen as required by the FID. 0728 (gray butyl rubber septum, plug style), 0105–0131 (butyl rubber septa), or equivalent. 4.2 Calibration and Linearity Check. The seal must be made from butyl rubber. 4.2.1 Stock Cylinder Gas Standard. 100 Silicone rubber seals are not acceptable. percent propane. The manufacturer shall: 3.1.2 Vial Sealer. Perkin-Elmer Number (a) Certify the gas composition to be accu- 105–0106, or equivalent. rate to ±3 percent or better (see Section 3.1.3 Gas-Tight Syringe. Perkin-Elmer 4.2.1.1); Number 00230117, or equivalent. (b) Recommend a maximum shelf life over 3.1.4 The following equipment is required which the gas concentration does not change for sampling. by greater than ±5 percent from the certified 3.1.4.1 Tap. value; and 3.1.4.2 Tubing. Telfon, 0.25-in. ID. Note: (c) Affix the date of gas cylinder prepara- Mention of trade names or specific products tion, certified propane concentration, and does not constitute endorsement by the En- recommended maximum shelf life to the cyl- vironmental Protection Agency. inder before shipment to the buyer.
984
VerDate 11
4.2.1.1 Cylinder Standards Certification. Standards, if such SRM’s are available. The The manufacturer shall certify the con- agreement between the initially determined centration of the calibration gas in the cyl- concentration value and the verification inder by (a) directly analyzing the cylinder concentration value shall be within ±5 per- and (b) calibrating his analytical procedure cent. The manufacturer must reverify all on the day of cylinder analysis. To calibrate calibration standards on a time interval con- his analytical procedure, the manufacturer sistent with the shelf life of the cylinder shall use, as a minimum, a three-point cali- standards sold. bration curve. 4.2.1.2 Verification of Manufacturer’s 5. Procedure Calibration Standards. Before using, the manufacturer shall verify each calibration 5.1 Sampling. standard by (a) comparing it to gas mixtures 5.1.1 Install a sampling tap to obtain the prepared in accordance with the procedure sample at a point which is most representa- described in Section 7.1 of Method 106 of part tive of the unexposed waste (where the waste 61, appendix B, or by (b) calibrating it has had minimum opportunity to volatilize against Standard Reference Materials to the atmosphere). Assemble the sampling (SRM’s) prepared by the National Bureau of apparatus as shown in Figure 25E–1.
5.1.2 Begin sampling by purging the sam- 5.2.1 Allow one hour for the headspace ple lines and cooling coil with at least four vials to equilibrate at the temperature speci- volumes of waste. Collect the purged mate- fied in the regulation. Allow the FID to rial in a separate container and dispose of it warm up until a stable baseline is achieved properly. on the detector. 5.1.3 After purging, stop the sample flow 5.2.2 Check the calibration of the FID and transfer the Teflon sampling tube to a daily using the procedures in Section 6.1.2. sample container. Sample at a flow rate such 5.2.3 Follow the manufacturer’s rec- that the temperature of the waste is <10°C ommended procedures for the normal oper- (<50°F). Fill the sample container halfway (±5 percent) and cap it within 5 seconds. ation of the headspace sampler and FID. Store immediately in a cooler and cover 5.2.4 Use the procedures in Sections 7.4 with ice. and 7.5 to calculate the vapor phase organic 5.1.4 Alternative sampling techniques vapor pressure in the samples. may be used upon the approval of the Admin- 5.2.5 Monitor the output of the detector to istrator. make certain that the results are being prop- 5.2 Analysis. erly recorded.
985
VerDate 11
6. Operational Checks and Calibration Pbar = Atmospheric pressure at analysis con- ditions, mm Hg (in. Hg). Maintain a record of performance of each item. P* = Organic vapor pressure in the sample, 6.1 Use the procedures in Section 6.1.1 to kPa (psi). calibrate the headspace analyzer and FID β = 1.333 X 10¥7 kPa/[(mm Hg)(ppm)], (4.91 X and check for linearity before the system is 10¥7 psi/[(in. Hg)(ppm)]) first placed in operation, after any shutdown 7.2 Linearity. Use the following equation longer than 6 months, and after any modi- to calculate the measured standard con- fication of the system. centration for each standard vial. 6.1.1 Calibration and Linearity. Use the C = k A + b Eq. 25E–1 procedures in Section 6.2.1 of Method 18 of m 7.2.1 Calculate the average measured Part 60, Appendix A, to prepare the stand- standard concentration (C ) for each set of ards and calibrate the flowmeters, using pro- ma triplicate standards and use the following pane as the standard gas. Fill the calibration equation to calculate the percent difference standard vials halfway (±5 percent) with de- ionized water. Purge and fill the airspace (PD) between Cma and Cs. with calibration standard. Prepare a mini- CC− mum of three calibration standards in trip- = s ma × licate at concentrations that will bracket PD 100 Eq. 25E - 2 the applicable cutoff. For a cutoff of 5.2 kPa, Cs prepare nominal concentrations of 30,000, The instrument linearity is acceptable if 50,000, and 70,000 ppm as propane. For a cut- the percent difference is within five for each off of 27.6 kPa, prepare nominal concentra- standard. tions of 200,000, 300,000, and 400,000 ppm as 7.3 Relative Standard Deviation (RSD). propane. 6.1.1.1 Use the procedures in Section 5.2.3 Use the following equation to calculate the to measure the FID response of each stand- RSD for each triplicate set of standards. ard. Use a linear regression analysis to cal- 2 culate the values for the slope (k) and the y- − 100 ∑()CCm ma intercept (b). Use the procedures in Sections RSD = Eq. 25E - 3 7.2 and 7.3 to test the calibration and the lin- Cma 2 earity. The calibration is acceptable if the RSD is 6.1.2 Daily FID Calibration Check. Check within five for each standard concentration. the calibration at the beginning and at the 7.4 Concentration of organics in the end of the daily runs by using the following headspace. Use the following equation to cal- procedures. Prepare two calibration stand- culate the concentration of vapor phase ards at the nominal cutoff concentration organics in each sample. using the procedures in Section 6.1.1. Place one at the beginning and one at the end of Ca = k A + b Eq. 25E–4 the daily run. Measure the FID response of 7.5 Vapor Pressure of Organics in the the daily calibration standard and use the Headspace Sample. Use the following equa- values for k and b from the most recent cali- tion to calculate the vapor pressure of bration to calculate the concentration of the organics in the sample. daily standard. Use an equation similar to P* = β Pbar Ca Eq. 25E–5 25E–2 to calculate the percent difference be- tween the daily standard and Cs. If the dif- METHOD 26—DETERMINATION OF HYDROGEN ference is within 5 percent, then the previous CHLORIDE EMISSIONS FROM STATIONARY values for k and b may be used. Otherwise, SOURCES use the procedures in Section 6.1.1 to recali- brate the FID. 1. Applicability, Principle, Interferences, Precision, Bias, and Stability 7. Calculations 1.1 Applicability. This method is applica- 7.1 Nomenclature. ble for determining emissions of hydrogen A = Measurement of the area under the re- halides (HX) [hydrogen chloride (HCl), hy- sponse curve, counts. drogen bromide (HBr), and hydrogen fluoride b = y-intercept of the linear regression line. (HF)] and halogens (X2) [chlorine (Cl2) and C = Measured vapor phase organic con- a bromine (Br2)] from stationary sources. centration of sample, ppm as propane. Sources, such as those controlled by wet Cma = Average measured vapor phase organic scrubbers, that emit acid particulate matter concentration of standard, ppm as pro- must be sampled using Method 26A. pane. NOTE: Mention of trade names or specific Cm = Measured vapor phase organic con- centration of standard, ppm as propane. products does not constitute endorsement by the Environmental Protection Agency.] Cs = Calculated standard concentration, ppm as propane. 1.2 Principle. An integrated sample is ex- k = Slope of the linear regression line. tracted from the source and passed through
986 Environmental Protection Agency Pt. 60, App. A, Meth. 26
a prepurged heated probe and filter into di- trate (NO3¥) to interfere with measurements lute sulfuric acid and dilute sodium hydrox- of very low Br¥ levels. ide solutions which collect the gaseous hy- 1.4 Precision and Bias. The within-labora- drogen halides and halogens, respectively. tory relative standard deviations are 6.2 and The filter collects other particulate matter 3.2 percent at HCl concentrations of 3.9 and including halide salts. The hydrogen halides 15.3 ppm, respectively. The method does not are solubilized in the acidic solution and exhibit a bias to Cl2 when sampling at con- form chloride (Cl¥), bromide (Br¥), and fluo- centrations less than 50 ppm. ride (F¥) ions. The halogens have a very low 1.5 Sample Stability. The collected Cl¥ solubility in the acidic solution and pass samples can be stored for up to 4 weeks. through to the alkaline solution where they 1.6 Detection Limit. The analytical detec- ∂ are hydrolyzed to form a proton (H ), the tion limit for Cl¥ is 0.1 µg/ml. Detection lim- halide ion, and the hypohalous acid (HClO or its for the other analyses should be similar. HBrO). Sodium thiosulfate is added in excess to the alkaline solution to assure reaction 2. Apparatus with the hypohalous acid to form a second halide ion such that 2 halide ions are formed 2.1 Sampling. The sampling train is for each molecule of halogen gas. The halide shown in Figure 26–1, and component parts ions in the separate solutions are measured are discussed below. by ion chromatography (IC). 2.1.1 Probe. Borsilicate glass, approxi- 1.3 Interferences. Volatile materials, such mately 3/8-in. (9-mm) I.D. with a heating sys- as chlorine dioxide (ClO2) and ammonium tem to prevent moisture condensation. A chloride (NH4Cl), which produce halide ions Teflon-glass filter in a mat configuration upon dissolution during sampling are poten- shall be installed behind the probe to remove tial interferents. Interferents for the halide particulate matter from the gas stream (see measurements are the halogen gases which section 2.1.5). A glass wool plug should not be disproportionate to a hydrogen halide and a used to remove particulate matter since a hydrohalous acid upon dissolution in water. negative bias in the data could result. However, the use of acidic rather than neu- 2.1.2 Three-Way Stopcock. A borosilicate tral or basic solutions for collection of the glass three-way stopcock with a heating sys- hydrogen halides greatly reduces the dissolu- tem to prevent moisturecondensation. The tion of any halogens passing through this so- heated stopcock should connect to the outlet lution. The simultaneous presence of HBr of the heated filter and the inlet of the first and CL2 may cause a positive bias in the impinger. The heating system shall be capa- HCL result with a corresponding negative ble of preventing condensation up to the bias in the Cl2 result as well as affecting the inlet of the first impinger. Silicone grease HBr/Br2 split. High concentrations of nitro- may be used, if necessary, to prevent leak- gen oxides (NOX) may produce sufficient ni- age.
987
VerDate 11
2.1.3 Impingers. Four 30-ml midget 2.1.9 Purge Pump, Purge Line, Drying impingers with leak-free glass connectors. Tube, Needle Valve, and Rate Meter. Pump Silicone grease may be used, if necessary, to capable of purging the sampling probe at 2 li- prevent leakage. For sampling at high mois- ters/min, with drying tube, filled with silica ture sources or for sampling times greater gel or equivalent, to protect pump, and a than 1 hour, a midget impinger with a short- rate meter capable of measuring 0 to 5 liters/ ened stem (such that the gas sample does not min. bubble through the collected condensate) 2.1.10 Stopcock Grease, Valve, Pump, Vol- should be used in front of the first impinger. ume Meter, Barometer, and Vacuum Gauge. 2.1.4 Drying Tube or Impinger. Tube or Same as in Method 6, Sections 2.1.4, 2.1.7, impinger, of Mae West design, filled with 6- 2.1.8, 2.1.10, 2.1.11, and 2.1.12. to 16-mesh indicating type silica gel, or 2.1.11 Temperature Measuring Devices. equivalent, to dry the gas sample and to pro- Temperature measuring device to monitor tect the dry gas meter and pump. If the sili- the temperature of the probe and a ther- ca gel has been used previously, dry at 175 °C mometer or other temperature measuring (350 °F) for 2 hours. New silica gel may be device to monitor the temperature of the used as received. Alternatively, other types sampling system from the outlet of the probe of desiccants (equivalent or better) may be to the inlet of the first impinger. used. 2.1.12 Ice Water Bath. To minimize loss of 2.1.5 Filter. When the stack gas tempera- absorbing solution. ture exceeds 210°C (410°F) and the HCl con- 2.2 Sample Recovery. centration is greater than 20 ppm, a quartz- 2.2.1 Wash Bottles. Polyethylene or glass, fiber filter may be used. 2.1.6 Filter Holder and Support. The filter 500-ml or larger, two. holder should be made of Teflon or quartz. 2.2.2 Storage Bottles. 100- or 250-ml, high- The filter support shall be made of Teflon. density polyethylene bottles with Teflon All-Teflon filter holders and supports are screw cap liners to store impinger samples. available from Savillex Corp., 5325 Hwy 101, 2.3 Sample Preparation and Analysis. The Minnetonka, MN 55345. materials required for volumetric dilution 2.1.7 Sample Line. Leak-free, with com- and chromatographic analysis of samples are patible fittings to connect the last impinger described below. to the needle valve. 2.3.1 Volumetric Flasks. Class A, 100-ml 2.1.8 Rate Meter. Rotameter, or equiva- size. lent, capable of measuring flow rate to with- 2.3.2 Volumetric Pipets. Class A, assort- in 2 percent of the selected flow rate of 2 li- ment. To dilute samples into the calibration ters/min. range of the instrument.
988
VerDate 11
2.3.3 Ion Chromatograph. Suppressed or Eq. 26–3 nonsuppressed, with a conductivity detector Alternately, solutions containing a nominal and electronic integrator operating in the certified concentration of 1000 mg/l NaCl are peak area mode. Other detectors, strip chart commercially available as convenient stock recorders, and peak height measurements solutions from which standards can be made may be used. by appropriate volumetric dilution. Refrig- erate the stock standard solutions and store 3. Reagents no longer than one month. Unless otherwise indicated, all reagents 3.2.4 Chromatographic Eluent. Effective must conform to the specifications estab- eluents for nonsuppressed IC using a resin- lished by the Committee on Analytical Re- or silica-based weak ion exchange column agents of the American Chemical Society are a 4 mM potassium hydrogen phthalate (ACS reagent grade). When such specifica- solution, adjusted to pH 4.0 using a saturated tions are not available, the best available sodium borate solution, and a 4 mM 4-hy- grade shall be used. droxy benzoate solution, adjusted to pH 8.6 3.1 Sampling. using 1 N NaOH. An effective eluent for sup- 3.1.1 Water. Deionized, distilled water pressed ion chromatography is a solution that conforms to ASTM Specification D 1193– containing 3 mM sodium bicarbonate and 2.4 77, Type 3. mM sodium carbonate. Other dilute solu- 3.1.2 Acidic Absorbing solution, 0.1 N Sul- tions buffered to a similar pH and containing furic Acid (H2SO4). To prepare 100 ml of the no interfering ions may be used. When using absorbing solution for the front impinger suppressed ion chromatography, if the pair, slowly add 0.28 ml of concentrated ‘‘water dip’’ resulting from sample injection H2SO4 to about 90 ml of water while stirring, interferes with the chloride peak, use a 2 mM and adjust the final volume to 100 ml using NaOH/2.4 mM sodium bicarbonate eluent. additional water. Shake well to mix the solu- tion. 4. Procedure 3.1.3 Alkaline Absorbing Solution, 0.1 N 4.1 Sampling. Sodium Hydroxide (NaOH). To prepare 100 ml 4.1.1 Preparation of Collection Train. Pre- of the scrubber solution for the back pair of pare the sampling train as follows: Pour 15 impingers, dissolve 0.40 g of solid NaOH in ml of the acidic absorbing solution into each about 90 ml of water, and adjust the final so- one of the first pair of impingers, and 15 ml lution volume to 100 ml using additional of the alkaline absorbing solution into each water. Shake well to mix the solution. one of the second pair of impingers. Connect 3.1.4 Sodium Thiosulfate (Na2S2O3.5H2O) the impingers in series with the knockout 3.2 Sample Preparation and Analysis. impinger first, if used, followed by the two 3.2.1 Water. Same as in Section 3.1.1. impingers containing the acidic absorbing 3.2.2 Absorbing Solution Blanks. A sepa- solution and the two impingers containing rate blank solution of each absorbing rea- the alkaline absorbing solution. Place a gent should be prepared for analysis with the fresh charge of silica gel, or equivalent, in field samples. Dilute 30 ml of each absorbing the drying tube or impinger at the end of the solution to approximately the same final impinger train. volume as the field samples using the blank 4.1.2 Adjust the probe temperature and sample of rinse water. the temperature of the filter and the stop- 3.2.3 Halide Salt Stock Standard Solu- cock, i.e., the heated area in Figure 26–1 to a tions. Prepare concentrated stock solutions temperature sufficient to prevent water con- from reagent grade sodium chloride (NaCl), densation. This temperature should be at sodium bromide (NaBr), and sodium fluoride least 20°C above the source temperature, but (NaF). Each must be dried at 110°C for two or not greater than 120°C. The temperature more hours and then cooled to room tem- should be monitored throughout a sampling perature in a desiccator immediately before run to ensure that the desired temperature is weighing. Accurately weigh 1.6 to 1.7 g of the maintained. dried NaCl to within 0.1 mg, dissolve in 4.1.3 Leak-Check Procedure. A leak-check water, and dilute to 1 liter. Calculate the prior to the sampling run is optional; how- exact Cl- concentration using Equation 26–1. ever, a leak-check after the sampling run is µg Cl¥/ml = g of NaCl × 103 × 35.453/58.44 mandatory. The leak-check procedure is as follows: Temporarily attach a suitable (e.g., Eq. 26–1 0–40 cc/min) rotameter to the outlet of the In a similar manner, accurately weigh and dry gas meter and place a vacuum gauge at solubilize 1.2 to 1.3 g of dried NaBr and 2.2 to or near the probe inlet. Plug the probe inlet, 2.3 g of NaF to make 1-liter solutions. Use pull a vacuum of at least 250 mm Hg (10 in. Equations 26–2 and 26–3 to calculate the Br¥ Hg), and note the flow rate as indicated by and F¥ concentrations. the rotameter. A leakage rate not in excess µg Br¥/ml = g of NaBr × 103 × 79.904/102.90 of 2 percent of the average sampling rate is Eq. 26–2 acceptable. (NOTE: Carefully release the µg F¥/ml = g of NaF × 103 × 18.998/41.99 probe inlet plug before turning off the
989
VerDate 11
pump.) It is suggested (not mandatory) that amount used in the sampling train (these are the pump be leak-checked separately, either the absorbing solution blanks described in prior to or after the sampling run. If done Section 3.2.2); dilute to the approximate vol- prior to the sampling run, the pump leak- ume of the corresponding samples using check shall precede the leak-check of the rinse water directly from the wash bottle sampling train described immediately above; being used. Add the same amount of sodium if done after the sampling run, the pump thiosulfate solution to the 0.1 N NaOH ab- leak-check shall follow the train leak-check. sorbing solution blank. Also, save a portion To leak-check the pump, proceed as follows: of the rinse water used to rinse the sampling Disconnect the drying tube from the probe- impinger assembly. Place a vacuum gauge at train. Place each in a separate, prelabeled the inlet to either the drying tube or pump, storage bottle. The sample storage bottles pull a vacuum of 250 mm (10 in.) Hg, plug or should be sealed, shaken to mix, and labeled. pinch off the outlet of the flowmeter, and Mark the fluid level. then turn off the pump. The vacuum should 4.3 Sample Preparation for Analysis. Note remain stable for at least 30 sec. Other leak- the liquid levels in the storage bottles and check procedures may be used, subject to the confirm on the analysis sheet whether or not approval of the Administrator, U.S. Environ- leakage occurred during transport. If a no- mental Protection Agency. ticeable leakage has occurred, either void 4.1.4 Purge Procedure. Immediately be- the sample or use methods, subject to the ap- fore sampling, connect the purge line to the proval of the Administrator, to correct the stopcock, and turn the stopcock to permit final results. Quantitatively transfer the the purge pump to purge the probe (see Fig- sample solutions to 100-ml volumetric flasks, ure 1A of Figure 26–1). Turn on the purge and dilute to 100 ml with water. pump, and adjust the purge rate to 2 liters/ 4.4 Sample Analysis. min. Purge for at least 5 minutes before sam- 4.4.1 The IC conditions will depend upon pling. analytical column type and whether sup- 4.1.5 Sample Collection. Turn on the sam- pressed or nonsuppressed IC is used. An ex- pling pump, pull a slight vacuum of approxi- ample chromatogram from a nonsuppressed mately 25 mm Hg (1 in. Hg) on the impinger system using a 150-mm Hamilton PRP–X100 train, and turn the stopcock to permit stack gas to be pulled through the impinger train anion column, a 2 ml/min flow rate of 4 mM (see Figure 1C of Figure 26–1). Adjust the 4-hydroxy benzoate solution adjusted to a pH µ sampling rate to 2 liters/min, as indicated by of 8.6 using 1 N NaOH, a 50- l sample loop, the rate meter, and maintain this rate to and a conductivity detector set on 1.0 µS full within 10 percent during the entire sampling scale is shown in Figure 26–2. run. Take readings of the dry gas meter vol- 4.4.2 Before sample analysis, establish a ume and temperature, rate meter, and vacu- stable baseline. Next, inject a sample of um gauge at least once every 5 minutes dur- water, and determine if any Cl¥, Br¥, or F¥ ing the run. A sampling time of 1 hour is rec- appears in the chromatogram. If any of these ommended. Shorter sampling times may in- ions are present, repeat the load/injection troduce a significant negative bias in the procedure until they are no longer present. HCl concentration. At the conclusion of the Analysis of the acid and alkaline absorbing sampling run, remove the train from the solution samples requires separate standard stack, cool, and perform a leak-check as de- calibration curves; prepare each according to scribed in section 4.1.2. Section 5.2. Ensure adequate baseline separa- 4.2 Sample Recovery. Disconnect the tion of the analyses. impingers after sampling. Quantitatively 4.4.3 Between injections of the appro- transfer the contents of the acid impingers priate series of calibration standards, inject and the knockout impinger, if used, to a in duplicate the reagent blanks, quality con- leak-free storage bottle. Add the water trol sample, and the field samples. Measure rinses of each of these impingers and con- the areas or heights of the Cl¥, Br¥, and F¥ necting glassware to the storage bottle. Re- peaks. Use the mean response of the dupli- peat this procedure for the alkaline cate injections to determine the concentra- impingers and connecting glassware using a separate storage bottle. Add 25 mg sodium tions of the field samples and reagent blanks thiosulfate per the product of ppm of halogen using the linear calibration curve. The val- anticipated to be in the stack gas times the ues from duplicate injections should agree dscm stack gas sampled. [Note: This amount within 5 percent of their mean for the analy- of sodium thiosulfate includes a safety fac- sis to be valid. Dilute any sample and the tor of approximately 5 to assure complete re- blank with equal volumes of water if the action with the hypohalous acid to form a concentration exceeds that of the highest second Cl¥ ion in the alkaline solution.] standard. Save portions of the absorbing reagents (0.1 4.5 Audit Analysis. An audit sample must N H2SO4 and 0.1 N NaOH) equivalent to the be analyzed, subject to availability.
990
VerDate 11
at least four calibration standards for each absorbing reagent containing the appro- priate stock solutions such that they are within the linear range of the field samples. Using one of the standards in each series, en- sure adequate baseline separation for the peaks of interest. Inject the appropriate se- ries of calibration standards, starting with the lowest concentration standard first both before and after injection of the quality con- trol check sample, reagent blanks, and field samples. This allows compensation for any instrument drift occurring during sample analysis. Determine the peak areas, or heights, for the standards and plot individual values ver- sus halide ion concentrations in µg/ml. Draw a smooth curve through the points. Use lin- ear regression to calculate a formula de- scribing the resulting linear curve.
6. Quality Assurance 6.1 Applicability. When the method is used to analyze samples to demonstrate com- pliance with a source emission regulation, a set of two audit samples must be analyzed. 6.2 Audit Procedure. The audit sample are chloride solutions. Concurrently analyze the two audit samples and a set of compliance samples in the same manner to evaluate the technique of the analyst and the standards preparation. The same analyst, analytical reagents, and analytical system shall be used both for compliance samples and the EPA audit samples. If this condition is met, au- diting the subsequent compliance analyses for the same enforcement agency within 30 days is not required. An audit sample set may not be used to validate different sets of compliance samples under the jurisdiction of different enforcement agencies, unless prior arrangements are made with both enforce- ment agencies. 6.3 Audit Sample Availability. The audit samples may be obtained by writing or call- ing the EPA Regional Office or the appro- priate enforcement agency. The request for the audit samples must be made at least 30 days prior to the scheduled compliance sam- ple analyses. 6.4 Audit Results. 6.4.1 Calculate the concentrations in mg/ dscm using the specified sample volume in the audit instructions. NOTE: Indication of acceptable results may 5. Calibration be obtained immediately by reporting the audit results in mg/dscm and compliance re- 5.1 Dry Gas Metering System. Thermom- sults in total µg HCl/sample to the respon- eters, Rate Meter, and Barometer. Same as sible enforcement agency. Include the re- in Method 6, sections 5.1, 5.2, 5.3, and 5.4. sults of both audit samples, their identifica- 5.2 Ion Chromatograph. To prepare the tion numbers, and the analyst’s name with calibration standards, dilute given amounts the results of the compliance determination (1.0 ml or greater) of the stock standard solu- samples in appropriate reports to the EPA tions to convenient volumes, using 0.1 N Regional Office or the appropriate enforce- H2SO2 or 0.1 N NaOH, as appropriate. Prepare ment agency. Include this information with 991
VerDate 11
subsequent analyses for the same enforce- K=10¥3 mg/µg. ment agency during the 30-day period. 8. Bibliography 6.4.2 The concentrations of the audit sam- ples obtained by the analyst shall agree 1. Steinsberger, S.C. and J.H. Margeson, within 10 percent of the actual concentra- ‘‘Laboratory and Field Evaluation of a Meth- tions. If the 10 percent specification is not odology for Determination of Hydrogen Chlo- met, reanalyze the compliance samples and ride Emissions form Municipal and Hazard- audit samples, and include initial and rea- ous Waste Incinerators,’’ U.S. Environ- nalysis values in the test report. mental Protection Agency, Office of Re- 6.4.3 Failure to meet the 10 percent speci- search and Development, Report No. 600/3–89/ fication may require retests until the audit 064, April 1989. Available from the National problems are resolved. However, if the audit Technical Information Service, Springfield, results do not affect the compliance or non- VA 22161 as PB89220586/AS. compliance status of the affected facility, 2. State of California, Air Resources Board. the Administrator may waive the reanalysis Method 421. ‘‘Determination of Hydrochloric requirement, further audits, or retests and Acid Emissions from Stationary Sources.’’ accept the results of the compliance test. March 18, 1987. While steps are being taken to resolve audit 3. Cheney, J.L. and C.R. Fortune. Improve- analysis problems, the Administrator may ments in the Methodology for Measuring Hy- also choose to use the data to determine the drochloric Acid in Combustion Source Emis- compliance or noncompliance status of the sions. J. Environ. Sci. Health. A19(3): 337–350. affected facility. 1984. 4. Stern, D. A., B. M. Myatt, J. F. 7. Calculations Lachowski, and K. T. McGregor. Speciation of Halogen and Hydrogen Halide Compounds Retain at least one extra decimal figure in Gaseous Emissions. In: Incineration and beyond those contained in the available data Treatment of Hazardous Waste: Proceedings in intermediate calculations, and round off of the 9th Annual Research Symposium, Cin- only the final answer appropriately. cinnati, Ohio, May 2–4, 1983. Publication No. 7.1 Sample Volume, Dry Basis, Corrected 600/9–84–015. July 1984. Available from Na- to Standard Conditions. Calculate the sam- tional Technical Information Service, ple volume using Eq. 6–1 of Method 6. Springfield, VA 22161 as PB84–234525. 7.2 Total µg HCl, HBr, or HF Per Sample. 5. Holm, R. D. and S. A. Barksdale. Analy- ¥ ¥ mHX=K Vs (SX ¥BX ) Eq. 26–4 sis of Anions in Combustion Products. In: Ion where: Chromatographic Analysis of Environmental ¥ Pollutants. E. Sawicki, J. D. Mulik, and E. BX =Mass concentration of applicable ab- sorbing solution blank, µg halide ion Wittgenstein (eds.). Ann Arbor, Michigan, (Cl¥, Br¥, F¥)/ml, not to exceed 1 µg/ml Ann Arbor Science Publishers. 1978. pp. 99– which is 10 times the published analyt- 110. µ ical detection limit of 0.1 g/ml. Method 26A—Determination of Hydrogen Ha- mHX=Mass of HCl, HBr, or HF in sample, lide and Halogen Emissions from Station- µ g. ary Sources—Isokinetic Method SX¥=Analysis of sample, µg halide ion (Cl¥, Br¥, F¥)/ml. 1. Applicability, Principle, Interferences, Vs=Volume of filtered and diluted sample, Precision, Bias, and Stability ml. 1.1 Applicability. This method is applica- K =1.028 (µg HCl/µg¥mole)/(µg Cl¥/ HCl ble for determining emissions of hydrogen µg¥mole). halides (HX) [hydrogen chloride (HCl), hy- K =1.013 (µg HBr/µg¥mole)/(µg Br¥/ HBr drogen bromide (HBr), and hydrogen fluoride µg¥mole). (HF)] and halogens (X ) [chlorine (Cl ) and K =1.053 (µg HF/µg¥mole)/(µg F¥/ 2 2 HF bromine (Br )] from stationary sources. This µg¥mole). 2 method collects the emission sample 7.3 Total µg Cl or Br Per Sample. 2 2 isokinetically and is therefore particularly ¥ ¥ mX2=Vs (SX ¥BX ) Eq. 26–5 suited for sampling at sources, such as those where: controlled by wet scrubbers, emitting acid particulate matter (e.g., hydrogen halides mX2=Mass of Cl2 or Br2 in sample, µg. dissolved in water droplets). [Note: Mention 7.4 Concentration of Hydrogen Halide or of trade names or specific products does not Halogen in Flue Gas. constitute endorsement by the Environ- C=K m /V ( ) Eq. 26–6 HX,X2 m std mental Protection Agency.] where: 1.2 Principle. Gaseous and particulate pol- C=Concentration of hydrogen halide (HX) lutants are withdrawn isokinetically from or halogen (X2), dry basis, mg/dscm. the source and collected in an optional cy- Vm(std)= Dry gas volume measured by the clone, on a filter, and in absorbing solutions. dry gas meter, corrected to standard con- The cyclone collects any liquid droplets and ditions, dscm. is not necessary if the source emissions do
992
VerDate 11
not contain them; however, it is preferable chloride (NH4Cl), which produce halide ions to include the cyclone in the sampling train upon dissolution during sampling are poten- to protect the filter from any moisture tial interferents. Interferents for the halide present. The filter collects other particulate measurements are the halogen gases which matter including halide salts. Acidic and al- disproportionate to a hydrogen halide and an kaline absorbing solutions collect the gase- hypohalous acid upon dissolution in water. ous hydrogen halides and halogens, respec- The use of acidic rather than neutral or tively. Following sampling of emissions con- basic solutions for collection of the hydrogen taining liquid droplets, any halides/halogens halides greatly reduces the dissolution of dissolved in the liquid in the cyclone and on any halogens passing through this solution. the filter are vaporized to gas and collected The simultaneous presence of both HBr and in the impingers by pulling conditioned am- C1 may cause a positive bias in the HCl re- bient air through the sampling train. The 2 sult with a corresponding negative bias in hydrogen halides are solubilized in the acidic the C1 result as well as affecting the HBr/ solution and form chloride (Cl¥), bromide 2 Br split. High concentrations of nitrogen ox- (Br¥), and fluoride (F¥) ions. The halogens 2 have a very low solubility in the acidic solu- ides (NOx) may produce sufficient nitrate tion and pass through to the alkaline solu- (NO3¥) to interfere with measurements of tion where they are hydrolyzed to form a very low Br- levels. proton (H∂), the halide ion, and the 1.4 Precision and Bias. The method has a hypohalous acid (HClO or HBrO). Sodium possible measurable negative bias below 20 thiosulfate is added to the alkaline solution ppm HCl perhaps due to reaction with small to assure reaction with the hypohalous acid amounts of moisture in the probe and filter. to form a second halide ion such that 2 ha- Similar bias for the other hydrogen halides lide ions are formed for each molecule of is possible. halogen gas. The halide ions in the separate 1.5 Sample Stability. The collected Cl¥ solutions are measured by ion chroma- samples can be stored for up to 4 weeks for tography (IC). If desired, the particulate analysis for HCl and C12. matter recovered from the filter and the 1.6 Detection Limit. The in-stack detec- probe is analyzed following the procedures in tion limit for HCl is approximately 0.02µg per Method 5. [NOTE: If the tester intends to use liter of stack gas; the analytical detection this sampling arrangement to sample con- limit for HCl is 0.1 1µg/ml. Detection limits currently for particulate matter, the alter- for the other analyses should be similar. native TeflonR probe liner, cyclone, and fil- ter holder should not be used. The TeflonR 2. Apparatus filter support must be used. The tester must also meet the probe and filter temperature 2.1 Sampling. The sampling train is requirements of both sampling trains.] shown in Figure 26A–1; the apparatus is simi- 1.3 Interferences. Volatile materials, such lar to the Method 5 train where noted as fol- as chlorine dioxide (ClO2) and ammonium lows:
993
VerDate 11
2.1.1 Probe Nozzle. Borosilicate or quartz 2.1.5 Filter Holder. Borosilicate or quartz glass; constructed and calibrated according glass, or Teflon filter holder, with a Teflon to Method 5, Sections 2.1.1 and 5.1, and cou- filter support and a sealing gasket. The seal- pled to the probe liner using a Teflon ing gasket shall be constructed of Teflon or union; a stainless steel nut is recommended equivalent materials. The holder design shall for this union. When the stack temperature provide a positive seal against leakage at exceeds 210 °C (410 °F), a one-piece glass noz- any point along the filter circumference. The zle/liner assembly must be used. holder shall be attached immediately to the 2.1.2 Probe Liner. Same as Method 5, Sec- outlet of the cyclone. tion 2.1.2, except metal liners shall not be 2.1.6 Impinger Train. The following sys- used. Water-cooling of the stainless steel tem shall be used to determine the stack gas sheath is recommended at temperatures ex- moisture content and to collect the hydro- ceeding 500 °C. Teflon may be used in lim- gen halides and halogens: five or six ited applications where the minimum stack impingers connected in series with leak-free temperature exceeds 120 °C (250 °F) but never ground glass fittings or any similar leak-free exceeds the temperature where Teflon is es- noncontaminating fittings. The first im- timated to become unstable (approximately pinger shown in Figure 26A–1 (knockout or 210 °C). condensate impinger) is optional and is rec- 2.1.3 Pitot Tube, Differential Pressure ommended as a water knockout trap for use Gauge, Filter Heating System, Metering Sys- under high moisture conditions. If used, this tem, Barometer, Gas Density Determination impinger should be constructed as described Equipment. Same as Method 5, Sections 2.1.3, below for the alkaline impingers, but with a 2.1.4, 2.1.6, 2.1.8, 2.1.9, and 2.1.10. shortened stem, and should contain 50 ml of 2.1.4 Cyclone (Optional). Glass or 0.1 N H2SO4. The following two impingers Teflon . Use of the cyclone is required only (acid impingers which each contain 100 ml of when the sample gas stream is saturated 0.1 N H2SO4) shall be of the Greenburg-Smith with moisture; however, the cyclone is rec- design with the standard tip (Method 5, Sec- ommended to protect the filter from any tion 2.1.7). The next two impingers (alkaline moisture droplets present. impingers which each contain 100 ml of 0.1 N
994
VerDate 11
NaOH) and the last impinger (containing sili- 3.1.2 Acidic Absorbing Solution, 0.1 N Sul- ca gel) shall be of the modified Greenburg- furic Acid (H2SO4). To prepare 1 L, slowly Smith design (Method 5, Section 2.1.7). The add 2.80 ml of concentrated H2SO4 to about condensate, acid, and alkaline impingers 900 ml of water while stirring, and adjust the shall contain known quantities of the appro- final volume to 1 L using additional water. priate absorbing reagents. The last impinger Shake well to mix the solution. shall contain a known weight of silica gel or 3.1.3 Alkaline Absorbing Solution, 0.1 N equivalent desiccant. Teflon impingers are Sodium Hydroxide (NaOH). To prepare 1 L, an acceptable alternative. dissolve 4.00 g of solid NaOH in about 900 ml 2.1.7 Ambient Air Conditioning Tube (Op- of water and adjust the final volume to 1 L tional). Tube tightly packed with approxi- using additional water. Shake well to mix mately 150 g of fresh 8 to 20 mesh sodium hy- the solution. droxide-coated silica, or equivalent, 3.1.4 Filter. Teflon mat (e.g., Pallflex (Ascarite II has been found suitable) to dry TX40H145) filter. When the stack gas tem- and remove acid gases from the ambient air perature exceeds 210 °C (410 °F) a quartz fiber used to remove moisture from the filter and filter may be used. cyclone, when the cyclone is used. The inlet 3.1.5 Silica Gel, Crushed Ice, and Stopcock and outlet ends of the tube should be packed Grease. Same as Method 5, Sections 3.1.2, with at least 1-cm thickness of glass wool or 3.1.4, and 3.1.5, respectively. filter material suitable to prevent escape of 3.1.6 Sodium Thiosulfate, (Na2S2O33.5H2O). fines. Fit one end with flexible tubing, etc. to allow connection to probe nozzle following 3.2 Sample Recovery the test run. 3.2.1 Water. Same as Section 3.1.1. 2.2 Sample Recovery. The following items 3.2.2 Acetone. Same as Method 5, Section are needed: 3.2. 2.2.1 Probe-Liner and Probe-Nozzle Brush- 3.3 Sample Analysis. es, Wash Bottles, 3.3.1 Water. Same as Section 3.1.1. Glass Sample Storage Containers, Petri 3.3.2 Reagent Blanks. A separate blank Dishes, Graduated Cylinder or Balance, and solution of each absorbing reagent should be Rubber Policeman. Same as Method 5, Sec- prepared for analysis with the field samples. tions 2.2.1, 2.2.2, 2.2.3, 2.2.4, 2.2.5, and 2.2.7. Dilute 200 ml of each absorbing solution (250 2.2.2 Plastic Storage Containers. Screw- ml of the acidic absorbing solution, if a con- cap polypropylene or polyethylene contain- densate impinger is used) to the same final ers to store silica gel. High-density poly- volume as the field samples using the blank ethylene bottles with Teflon screw cap liners sample of rinse water. If a particulate deter- to store impinger reagents, 1-liter. mination is conducted, collect a blank sam- 2.2.3 Funnels. Glass or high-density poly- ple of acetone. ethylene, to aid in sample recovery. 3.3.3 Halide Salt Stock Standard Solu- 2.3 Analysis. For analysis, the following tions. Prepare concentrated stock solutions equipment is needed: from reagent grade sodium chloride (NaCl), 2.3.1 Volumetric Flasks. Class A, various sodium bromide (NaBr), and sodium fluoride sizes. (NaF). Each must be dried at 110 °C for 2 or 2.3.2 Volumetric Pipettes. Class A, assort- more hours and then cooled to room tem- ment, to dilute samples to calibration range perature in a desiccator immediately before of the ion chromatograph (IC). weighing. Accurately weigh 1.6 to 1.7 g of the 2.3.3 Ion Chromatograph. Suppressed or dried NaCl to within 0.1 mg, dissolve in nonsuppressed, with a conductivity detector water, and dilute to 1 liter. Calculate the and electronic integrator operating in the exact Cl¥ concentration using Equation 26A– peak area mode. Other detectors, a strip 1. chart recorder, and peak heights may be µg Cl¥/ml = g of NaCl × 103 × 35.453/58.44 used. Eq. 26A–1 3. Reagents In a similar manner, accurately weigh and solubilize 1.2 to 1.3 g of dried NaBr and 2.2 to Unless otherwise indicated, all reagents 2.3 g of NaF to make 1-liter solutions. Use must conform to the specifications of the Equations 26A–2 and 26A–3 to calculate the Committee on Analytical Reagents of the Br¥ and F¥ concentrations. American Chemical Society (ACS reagent µg Br¥/ml = g of NaBr × 103 × 79.904/102.90 grade). When such specifications are not Eq. 26A–2 available, the best available grade shall be used. µg F¥/ml = g of NaF × 103 × 18.998/41.99 3.1 Sampling. Eq. 26A–3 3.1.1 Water. Deionized, distilled water Alternately, solutions containing a nomi- that conforms to American Society of Test- nal certified concentration of 1000 mg/L NaCl ing and Materials (ASTM) Specification D are commercially available as convenient 1193–77, Type 3 (incorporated by reference as stock solutions from which standards can be specified in § 60.17). made by appropriate volumetric dilution.
995
VerDate 11
Refrigerate the stock standard solutions and (e.g., ∆ H=1 in. H2O) to vaporize any liquid store no longer than 1 month. and hydrogen halides in the cyclone or on 3.3.4 Chromatographic Eluent. Same as the filter and pull them through the train Method 26, Section 3.2.4. into the impingers. After 30 minutes, turn off the flow, remove the conditioning tube, and 4. Procedure examine the cyclone and filter for any visi- Because of the complexity of this method, ble moisture. If moisture is visible, repeat testers and analysts should be trained and this step for 15 minutes and observe again. experienced with the procedures to ensure Keep repeating until the cyclone is dry. reliable results. [Note: It is critical that this is repeated 4.1 Sampling. until the cyclone is completely dry.] 4.1.1 Pretest Preparation. Follow the gen- 4.2 Sample Recovery. Allow the probe to eral procedure given in Method 5, Section cool. When the probe can be handled safely, 4.1.1, except the filter need only be des- wipe off all the external surfaces of the tip of iccated and weighed if a particulate deter- the probe nozzle and place a cap loosely over mination will be conducted. the tip. Do not cap the probe tip tightly 4.1.2 Preliminary Determinations. Same while the sampling train is cooling down be- as Method 5, Section 4.1.2. cause this will create a vacuum in the filter 4.1.3 Preparation of Sampling Train. Fol- holder, drawing water from the impingers low the general procedure given in Method 5, into the holder. Before moving the sampling Section 4.1.3, except for the following vari- train to the cleanup site, remove the probe, ations: wipe off any silicone grease, and cap the Add 50 ml of 0.1 N H2SO4 to the condensate open outlet of the impinger train, being care- impinger, if used. Place 100 ml of 0.1 N H2SO4 ful not to lose any condensate that might be in each of the next two impingers. Place 100 present. Wipe off any silicone grease and cap ml of 0.1 N NaOH in each of the following the filter or cyclone inlet. Remove the um- two impingers. Finally, transfer approxi- bilical cord from the last impinger and cap mately 200–300 g of preweighed silica gel the impinger. If a flexible line is used be- from its container to the last impinger. Set tween the first impinger and the filter hold- up the train as in Figure 26A–1. When used, er, disconnect it at the filter holder and let the optional cyclone is inserted between the any condensed water drain into the first im- probe liner and filter holder and located in pinger. Wipe off any silicone grease and cap the heated filter box. the filter holder outlet and the impinger 4.1.4 Leak-Check Procedures. Follow the inlet. Ground glass stoppers, plastic caps, leak-check procedures given in Method 5, serum caps, Teflon tape, Parafilm, or alu- Sections 4.4.1 (Pretest Leak-Check), 4.1.4.2 minum foil may be used to close these open- (Leak-Checks During the Sample Run), and ings. Transfer the probe and filter/impinger 4.1.4.3 (Post-Test Leak-Check). assembly to the cleanup area. This area 4.1.5 Train Operation. Follow the general should be clean and protected from the procedure given in Method 5, Section 4.1.5. weather to minimize sample contamination Maintain a temperature around the filter or loss. Inspect the train prior to and during and (cyclone, if used) of greater than 120 °C disassembly and note any abnormal condi- (248 °F). tions. Treat samples as follows: For each run, record the data required on 4.2.1 Container No. 1 (Optional; Filter a data sheet such as the one shown in Meth- Catch for Particulate Determination). Same od 5, Figure 5–2. If the condensate impinger as Method 5, Section 4.2, Container No. 1. becomes too full, it may be emptied, re- 4.2.2 Container No. 2 (Optional; Front- charged with 50 ml of 0.1 N H2SO4, and re- Half Rinse for Particulate Determination). placed during the sample run. The conden- Same as Method 5, Section 4.2, Container No. sate emptied must be saved and included in 2. the measurement of the volume of moisture 4.2.3 Container No. 3 (Knockout and Acid collected and included in the sample for Impinger Catch for Moisture and Hydrogen analysis. The additional 50 ml of absorbing Halide Determination). Disconnect the reagent must also be considered in calculat- impingers. Measure the liquid in the acid and ing the moisture. After the impinger is re- knockout impingers to ±1 ml by using a installed in the train, conduct a leak-check graduated cylinder or by weighing it to ±0.5 as described in Method 5, Section 4.1.4.2. g by using a balance. Record the volume or 4.1.6 Post-Test Moisture Removal (Op- weight of liquid present. This information is tional). When the optional cyclone is in- required to calculate the moisture content of cluded in the sampling train or when mois- the effluent gas. Quantitatively transfer this ture is visible on the filter at the end of a liquid to a leak-free sample storage con- sample run even in the absence of a cyclone, tainer. Rinse these impingers and connecting perform the following procedure. Upon com- glassware including the back portion of the pletion of the test run, connect the ambient filter holder (and flexible tubing, if used) air conditioning tube at the probe inlet and with water and add these rinses to the stor- operate the train with the filter heating sys- age container. Seal the container, shake to tem at least 120 °C (248 °F) at a low flow rate mix, and label. The fluid level should be
996
VerDate 11
marked so that if any sample is lost during 4.3.3.1 The IC conditions will depend upon transport, a correction proportional to the analytical column type and whether sup- lost volume can be applied. Retain rinse pressed or nonsuppressed IC is used. Prior to water and acidic absorbing solution blanks calibration and sample analysis, establish a and analyze with the samples. stable baseline. Next, inject a sample of 4.2.4 Container No. 4 (Alkaline Impinger water, and determine if any Cl¥, Br¥, or F¥ Catch for Halogen and Moisture Determina- appears in the chromatogram. If any of these tion). Measure and record the liquid in the ions are present, repeat the load/injection alkaline impingers as described in Section procedure until they are no longer present. 4.2.3. Quantitatively transfer this liquid to a Analysis of the acid and alkaline absorbing leak-free sample storage container. Rinse solution samples requires separate standard these two impingers and connecting glass- calibration curves; prepare each according to ware with water and add these rinses to the container. Add 25 mg of sodium thiosulfate Section 5.2. Ensure adequate baseline separa- per ppm halogen-dscm of stack gas sampled. tion of the analyses. [Note: This amount of sodium thiosulfate in- 4.3.3.2 Between injections of the appro- cludes a safety factor of approximately 5 to priate series of calibration standards, inject assure complete reaction with the in duplicate the reagent blanks and the field hypohalous acid to form a second Cl¥ ion in samples. Measure the areas or heights of the the alkaline solution.] Seal the container, Cl¥, Br¥, and F¥ peaks. Use the average re- shake to mix, and label; mark the fluid level. sponse to determine the concentrations of Retain alkaline absorbing solution blank and the field samples and reagent blanks using analyze with the samples. the linear calibration curve. If the values 4.2.5 Container No. 5 (Silica Gel for Mois- from duplicate injections are not within 5 ture Determination). Same as Method 5, Sec- percent of their mean, the duplicate injec- tion 4.2, Container No. 3. tion shall be repeated and all four values 4.2.6 Container Nos. 6 through 9 (Reagent used to determine the average response. Di- Blanks). Save portions of the absorbing re- lute any sample and the blank with equal agents (0.1 N H2SO4 and 0.1 N NaOH) equiva- volumes of water if the concentration ex- lent to the amount used in the sampling ceeds that of the highest standard. train; dilute to the approximate volume of 4.4 Audit Sample Analysis. Audit samples the corresponding samples using rinse water must be analyzed subject to availability. directly from the wash bottle being used. Add the same ratio of sodium thiosulfate so- 5. Calibration lution used in container No. 4 to the 0.1 N NaOH absorbing reagent blank. Also, save a Maintain a laboratory log of all calibra- portion of the rinse water alone and a por- tions. tion of the acetone equivalent to the amount 5.1 Probe Nozzle, Pitot Tube, Dry Gas Me- used to rinse the front half of the sampling tering System, Probe Heater, Temperature train. Place each in a separate, prelabeled Gauges, Leak-Check of Metering System, sample container. and Barometer. Same as Method 5, Sections 4.2.7 Prior to shipment, recheck all sam- 5.1, 5.2, 5.3, 5.4, 5.5, 5.6, and 5.7, respectively. ple containers to ensure that the caps are 5.2 Ion Chromatograph. To prepare the well-secured. Seal the lids of all containers calibration standards, dilute given amounts around the circumference with Teflon tape. (1.0 ml or greater) of the stock standard solu- Ship all liquid samples upright and all par- tions to convenient volumes, using 0.1 N ticulate filters with the particulate catch H SO or 0.1 N NaOH, as appropriate. Prepare facing upward. 2 4 at least four calibration standards for each 4.3 Sample Preparation and Analysis. absorbing reagent containing the three stock Note the liquid levels in the sample contain- solutions such that they are within the lin- ers and confirm on the analysis sheet wheth- ear range of the field samples. Using one of er or not leakage occurred during transport. If a noticeable leakage has occurred, either the standards in each series, ensure adequate void the sample or use methods, subject to baseline separation for the peaks of interest. the approval of the Administrator, to correct Inject the appropriate series of calibration the final results. standards, starting with the lowest con- 4.3.1 Container Nos. 1 and 2 and Acetone centration standard first both before and Blank (Optional; Particulate Determina- after injection of the quality control check tion). Same as Method 5, Section 4.3. sample, reagent blanks, and field samples. 4.3.2 Container No. 5. Same as Method 5, This allows compensation for any instru- Section 4.3 for silica gel. ment drift occurring during sample analysis. 4.3.3 Container Nos. 3 and 4 and Absorbing Determine the peak areas, or height, of the Solution and Water Blanks. Quantitatively standards and plot individual values versus transfer each sample to a volumetric flask or halide ion concentrations in µg/ml. Draw a graduated cylinder and dilute with water to smooth curve through the points. Use linear a final volume within 50 ml of the largest regression to calculate a formula describing sample. the resulting linear curve.
997
VerDate 11
6. Quality Control also approximately 5 percent of the mass concentration anticipated to result from Same as Method 5, Section 4.4. a one hour sample at 10 ppmv HCl.) 7. Quality Assurance C=Concentration of hydrogen halide (HX) or halogen (X ), dry basis, mg/dscm. 7.1 Applicability. When the method is 2 m =Mass of HCl, HBr, or HF in sample, µg. used to demonstrate compliance with a regu- HX m =Mass of Cl or Br in sample, µg. lation, a set of two audit samples shall be X2 2 2 µ ¥ analyzed. SX¥-=Analysis of sample, g halide ion (Cl , ¥ ¥ 7.2 Audit Procedure. The currently avail- Br , F )/ml. able audit samples are chloride solutions. VS=Volume of filtered and diluted sample, Concurrently analyze the two audit samples ml. and a set of compliance samples in the same 8.2 Average Dry Gas Meter Temperature manner to evaluate the technique of the ana- and Average Orifice Pressure Drop. See data lyst and the standards preparation. The sheet (Figure 5–2 of Method 5). same analyst, analytical reagents, and ana- 8.3 Dry Gas Volume. Calculate Vm(std) and lytical system shall be used both for compli- adjust for leakage, if necessary, using the ance samples and the Environmental Protec- equation in Section 6.3 of Method 5. tion Agency (EPA) audit samples. 8.4 Volume of Water Vapor and Moisture 7.3 Audit Sample Availability. Audit sam- Content. Calculate the volume of water ples will be supplied only to enforcement vapor Vw(std) and moisture content Bws from agencies for compliance tests. Audit samples the data obtained in this method (Figure 5– may be obtained by writing the Source Test 2 of Method 5); use Equations 5–2 and 5–3 of Audit Coordinator (MD–77B), Quality Assur- Method 5. ance Division, Atmospheric Research and 8.5 Isokinetic Variation and Acceptable Exposure Assessment Laboratory, U.S. Envi- Results. Use Method 5, Sections 6.11 and 6.12. ronmental Protection Laboratory, Research 8.6 Acetone Blank Concentration, Ace- Triangle Park, NC 27711 or by calling the tone Wash Blank Residue Weight, Particu- Source Test Audit Coordinator (STAC) at late Weight, and Particulate Concentration. (919) 541–7834. The request for the audit sam- For particulate determination. ples should be made at least 30 days prior to 8.7 Total µg HCl, HBr, or HF Per Sample. the scheduled compliance sample analysis. 7.4 Audit Results. Calculate the con- mHX=K Vs (SX¥–BX¥) Eq. 26A–4 centrations in mg/dscm using the specified where: ¥ sample volume in the audit instructions. In- KHC1 = 1.028 (µg HCl/µg-mole)/(µg Cl /µg- clude the results of both audit samples, their mole). ¥ identification numbers, and the analyst’s KHBr=1.013 (µg HBr/µg-mole)/(µg Br /µg- name with the results of the compliance de- mole). ¥ termination samples in appropriate reports KHF=1.053 (µg HF/µg-mole)/(µg F /µg-mole). to the EPA regional office or the appropriate 8.8 Total µg Cl2 or Br2 Per Sample. enforcement agency. (NOTE: Acceptability mX2= Vs (SX¥–BX¥) Eq. 26A–5 of results may be obtained immediately by 8.9 Concentration of Hydrogen Halide or reporting the audit results in mg/dscm and Halogen in Flue Gas. compliance results in total µg HCl/sample to the responsible enforcement agency.) The C=K mHX,X2/Vm(std) Eq. 26A–6 concentrations of the audit samples obtained where: K=1010¥ mg/µg by the analyst shall agree within 10 percent 8.10 Stack Gas Velocity and Volumetric of the actual concentrations. If the 10 per- Flow Rate. Calculate the average stack gas cent specification is not met, reanalyze the velocity and volumetric flow rate, if needed, compliance samples and audit samples, and using data obtained in this method and the include initial and reanalysis values in the equations in Sections 5.2 and 5.3 of Method 2. test report. Failure to meet the 10 percent specification may require retests until the 9. Bibliography audit problems are resolved. 1. Steinsberger, S. C. and J. H. Margeson. Laboratory and Field Evaluation of a Meth- 8. Calculations odology for Determination of Hydrogen Chlo- Retain at least one extra decimal figure ride Emissions from Municipal and Hazard- beyond those contained in the available data ous Waste Incinerators. U.S. Environmental in intermediate calculations, and round off Protection Agency, Office of Research and only the final answer appropriately. Development. Publication No. 600/3–89/064. 8.1 Nomenclature. Same as Method 5, Sec- April 1989. Available from National Tech- tion 6.1. In addition: nical Information Service, Springfield, VA
1 BX¥=Mass concentration of applicable ab- 22161 as PB89220586/AS. sorbing solution blank, µg halide ion 2. State of California Air Resources Board. (Cl¥, Br¥, F¥)/ml, not to exceed 1 µg/ml Method 421—Determination of Hydrochloric which is 10 times the published analyt- Acid Emissions from Stationary Sources. ical detection limit of 0.1 µg/ml. (It is March 18, 1987.
998
VerDate 11
3. Cheney, J.L. and C.R. Fortune. Improve- 2.7 Initial Vacuum (Vi). The vacuum ap- ments in the Methodology for Measuring Hy- plied to the delivery tank at the beginning of drochloric Acid in Combustion Source Emis- the static vacuum test, as specified in the sions. J. Environ. Sci. Health. A19(3): 337–350. appropriate regulation, in mm H2O. 1984. 2.8 Allowable Pressure Change (∆p). The 4. Stern, D.A., B.M. Myatt, J.F. Lachowski, allowable amount of decrease in pressure and K.T. McGregor. Speciation of Halogen during the static pressure test, within the and Hydrogen Halide Compounds in Gaseous time period t, as specified in the appropriate Emissions. In: Incineration and Treatment of regulation, in mm H2O. Hazardous Waste: Proceedings of the 9th An- 2.9 Allowable Vacuum Change (∆v). The nual Research Symposium, Cincinnati, Ohio, allowable amount of decrease in vacuum dur- May 2–4, 1983. Publication No. 600/9–84–015. ing the static vacuum test, within the time July 1984. Available from National Technical period t, as specified in the appropriate regu- Information Service, Springfield, VA 22161 as lation, in mm H2O. PB84–234525. 3. Apparatus 5. Holm, R.D. and S.A. Barksdale. Analysis 3.1 Pressure Source. Pump or compressed of Anions in Combustion Products. In: Ion gas cylinder of air or inert gas sufficient to Chromatographic Analysis of Environmental pressurize the delivery tank to 500 mm H2O Pollutants, E. Sawicki, J.D. Mulik, and E. above atmospheric pressure. Wittgenstein (eds.). Ann Arbor, Michigan, 3.2 Regulator. Low pressure regulator for Ann Arbor Science Publishers. 1978. pp. 99– controlling pressurization of the delivery 110. tank. 3.3 Vacuum Source. Vacuum pump capa- METHOD 27—DETERMINATION OF VAPOR TIGHT- ble of evacuating the delivery tank to 250 NESS OF GASOLINE DELIVERY TANK USING mm H2O below atmospheric pressure. PRESSURE-VACUUM TEST 3.4 Pressure-Vacuum Supply Hose. 1. Applicability and Principle 3.5 Manometer. Liquid manometer, or equivalent instrument, capable of measuring 1.1 Applicability. This method is applica- up to 500 mm H O gauge pressure with ±2.5 ble for the determination of vapor tightness 2 mm H2O precision. of a gasoline delivery tank which is equipped 3.6 Pressure-Vacuum Relief Valves. The with vapor collection equipment. test apparatus shall be equipped with an in- 1.2 Principle. Pressure and vacuum are line pressure-vacuum relief valve set to acti- applied alternately to the compartments of a vate at 675 mm H2O above atmospheric pres- gasoline delivery tank and the change in sure or 250 mm H2O below atmospheric pres- pressure or vacuum is recorded after a speci- sure, with a capacity equal to the pressur- fied period of time. izing or evacuating pumps. 2. Definitions and Nomenclature 3.7 Test Cap for Vapor Recovery Hose. 2.1 Gasoline. Any petroleum distillate or This cap shall have a tap for manometer con- petroleum distillate/alcohol blend having a nection and a fitting with shut-off valve for Reid vapor pressure of 27.6 kilopascals or connection to the pressure-vacuum supply greater which is used as a fuel for internal hose. combustion engines. 3.8 Caps for Liquid Delivery Hoses. 2.2 Delivery Tank. Any container, includ- 4. Pretest Preparations ing associated pipes and fittings, that is at- 4.1 Summary. Testing problems may tached to or forms a part of any truck, trail- occur due to the presence of volatile vapors er, or railcar used for the transport of gaso- and/or temperature fluctuations inside the line. delivery tank. Under these conditions, it is 2.3 Compartment. A liquid-tight division often difficult to obtain a stable initial pres- of a delivery tank. sure at the beginning of a test, and erroneous 2.4 Delivery Tank Vapor Collection test results may occur. To help prevent this, Equipment. Any piping, hoses, and devices it is recommended that, prior to testing, on the delivery tank used to collect and volatile vapors be removed from the tank route gasoline vapors either from the tank and the temperature inside the tank be al- to a bulk terminal vapor control system or lowed to stabilize. Because it is not always from a bulk plant or service station into the possible to attain completely these pretest tank. conditions a provision to ensure reproducible 2.5 Time Period of the Pressure or Vacu- results is included. The difference in results um Test (t). The time period of the test, as for two consecutive runs must meet the cri- specified in the appropriate regulation, dur- terion in Sections 5.2.5 and 5.3.5. ing which the change in pressure or vacuum 4.2 Emptying of Tank. The delivery tank is monitored, in minutes. shall be emptied of all liquid. 2.6 Initial Pressure (Pi). The pressure ap- 4.3 Purging of Vapor. As much as possible, plied to the delivery tank at the beginning of the delivery tank shall be purged of all vola- the static pressure test, as specified in the tile vapors by any safe, acceptable method. appropriate regulation, in mm H2O. One method is to carry a load of non-volatile
999
VerDate 11
liquid fuel, such as diesel or heating oil, im- 5.2.6 Compare the average measured mediately prior to the test, thus flushing out change in pressure to the allowable pressure all the volatile gasoline vapors. A second change, > p, as specified in the regulation. If method is to remove the volatile vapors by the delivery tank does not satisfy the vapor blowing ambient air into each tank compart- tightness criterion specified in the regula- ment for at least 20 minutes. This second tion, repair the sources of leakage, and re- method is usually not as effective and often peat the pressure test until the criterion is causes stabilization problems, requiring a met. much longer time for stabilization during 5.2.7 Disconnect the pressure source from the testing. the pressure-vacuum supply hose, and slowly 4.4 Temperature Stabilization. As much open the shut-off valve to bring the tank to as possible, the test shall be conducted under atmospheric pressure. isothermal conditions. The temperature of 5.3 Vacuum Test. the delivery tank should be allowed to equilibrate in the test environment. During 5.3.1 Connect the vacuum source to the the test, the tank should be protected from pressure-vacuum supply hose. extreme environmental and temperature 5.3.2 Open the shut-off valve in the vapor variability, such as direct sunlight. recovery hose cap. Slowly evacuate the tank to V , the initial vacuum specified in the reg- 5. Test Procedure i ulation. 5.1 Preparations. 5.3.3 Close the shut-off valve and allow 5.1.1 Open and close each dome cover. the pressure in the tank to stabilize, adjust- 5.1.2 Connect static electrical ground con- ing the pressure if necessary to maintain a nections to tank. Attach the liquid delivery vacuum of V . When the pressure stabilizes, and vapor return hoses, remove the liquid de- i record the time and initial vacuum. livery elbows, and plug the liquid delivery fittings. 5.3.4 At the end of t minutes, record the time and final vacuum. NOTE: The purpose of testing the liquid de- livery hoses is to detect tears or holes that 5.3.5 Repeat steps 5.3.2 through 5.3.4 until would allow liquid leakage during a delivery. the change in vacuum for two consecutive ± Liquid delivery hoses are not considered to runs agrees within 12.5 mm H2O. Calculate be possible sources of vapor leakage, and the arithmetic average of the two results. thus, do not have to be attached for a vapor 5.3.6 Compare the average measured leakage test. Instead, a liquid delivery hose change in vacuum to the allowable vacuum could be either visually inspected, or filled change, > v, as specified in the regulation. If with water to detect any liquid leakage.) the delivery tank does not satisfy the vapor 5.1.3 Attach the test cap to the end of the tightness criterion specified in the regula- vapor recovery hose. tion, repair the sources of leakage, and re- 5.1.4 Connect the pressure-vacuum supply peat the vacuum test until the criterion is hose and the pressure-vacuum relief valve to met. the shut-off valve. Attach a manometer to 5.3.7 Disconnect the vacuum source from the pressure tap. the pressure-vacuum supply hose, and slowly 5.1.5 Connect compartments of the tank open the shut-off valve to bring the tank to internally to each other if possible. If not atmospheric pressure. possible, each compartment must be tested 5.4 Post-Test Clean-Up. Disconnect all separately, as if it were an individual deliv- test equipment and return the delivery tank ery tank. to its pretest condition. 5.2 Pressure Test. 6. Alternative Procedures 5.2.1 Connect the pressure source to the pressure-vacuum supply hose. 6.1 The pumping of water into the bottom 5.2.2 Open the shut-off valve in the vapor of a delivery tank is an acceptable alter- recovery hose cap. Applying air pressure native to the pressure source described above. Likewise, the draining of water out of slowly, pressurize the tank to Pi, the initial pressure specified in the regulation. the bottom of a delivery tank may be sub- 5.2.3 Close the shut-off valve and allow stituted for the vacuum source. Note that the pressure in the tank to stabilize, adjust- some of the specific step-by-step procedures ing the pressure if necessary to maintain in the method must be altered slightly to ac- pressure of Pi. When the pressure stabilizes, commodate these different pressure and vac- record the time and initial pressure. uum sources. 5.2.4 At the end of t minutes, record the 6.2 Techniques other than specified above time and final pressure. may be used for purging and pressurizing a 5.2.5 Repeat steps 5.2.2 through 5.2.4 until delivery tank, if prior approval is obtained the change in pressure for two consecutive from the Administrator. Such approval will runs agrees within ±12.5 mm H2O. Calculate be based upon demonstrated equivalency the arithmetic average of the two results. with the above method.
1000
VerDate 11
METHOD 28—CERTIFICATION AND AUDITING OF allows a 1-inch diameter piece of wood to WOOD HEATERS pass through the grate, or, if not, to the top of the grate. Firebox height is not nec- 1. Applicability and Principle essarily uniform but must account for vari- 1.1 Applicability. This method is applica- ations caused by internal baffles, air chan- ble for the certification and auditing of wood nels, or other permanent obstructions. heaters. This method describes the test facil- 2.11.2 Length. The longest horizontal fire ity, test fuel charge, and wood heater oper- chamber dimension that is parallel to a wall ation as well as procedures for determining of the chamber. burn rates and particulate emission rates 2.11.3 Width. The shortest horizontal fire and for reducing data. chamber dimension that is parallel to a wall 1.2 Principle. Particulate matter emis- of the chamber. sions are measured from a wood heater burn- 2.12 Wood Heater. An enclosed, wood- ing a prepared test fuel crib in a test facility burning appliance capable of and intended maintained at a set of prescribed conditions. for space heating or domestic water heating, as defined in the applicable regulation. 2. Definitions 2.13 Pellet Burning Wood Heater. A wood heater which meets the following criteria: (1) 2.1 Burn Rate. The rate at which test fuel The manufacturer makes no reference to is consumed in a wood heater. Measured in burning cord wood in advertising or other kilograms of wood (dry basis) per hour (kg/ literature, (2) the unit is safety listed for pel- hr). let fuel only, (3) the unit operating and in- 2.2 Certification or Audit Test. A series of struction manual must state that the use of at least four test runs conducted for certifi- cordwood is prohibited by law, and (4) the cation or audit purposes that meets the burn unit must be manufactured and sold includ- rate specifications in Section 5. ing the hopper and auger combination as in- 2.3 Firebox. The chamber in the wood tegral parts. heater in which the test fuel charge is placed and combusted. 3. Apparatus 2.4 Secondary Air Supply. An air supply 3.1 Insulated Solid Pack Chimney. For in- that introduces air to the wood heater such stallation of wood heaters. Solid pack insu- that the burn rate is not altered by more lated chimneys shall have a minimum of 2.5 than 25 percent when the secondary air sup- cm (1 in.) solid pack insulating material sur- ply is adjusted during the test run. The wood rounding the entire flue and possess a label heater manufacturer can document this demonstrating conformance to U.L. Stand- through design drawings that show the sec- ard 103 (incorporated by reference. See ondary air is introduced only into a mixing § 60.17). chamber or secondary chamber outside the 3.2 Platform Scale and Monitor. For mon- firebox. itoring of fuel load weight change. The scale 2.5 Test Facility. The area in which the shall be capable of measuring weight to wood heater is installed, operated, and sam- within 0.05 kg (0.1 lb) or 1 percent of the ini- pled for emissions. tial test fuel charge weight, whichever is 2.6 Test Fuel Charge. The collection of greater. test fuel pieces placed in the wood heater at 3.3 Wood Heater Temperature Monitors. the start of the emission test run. Seven, each capable of measuring tempera- 2.7 Test Fuel Crib. The arrangement of ture to within 1.5 percent of expected abso- the test fuel charge with the proper spacing lute temperatures. requirements between adjacent fuel pieces. 3.4 Test Facility Temperature Monitor. A 2.8 Test Fuel Loading Density. The thermocouple located centrally in a verti- weight of the as-fired test fuel charge per cally oriented 150 mm (6 in.) long, 50 mm (2 unit volume of usable firebox. in.) diameter pipe shield that is open at both 2.9 Test Fuel Piece. The 2 x 4 or 4 x 4 wood ends, capable of measuring temperature to piece cut to the length required for the test within 1.5 percent of expected temperatures. fuel charge and used to construct the test 3.5 Balance (optional). Balance capable of fuel crib. weighing the test fuel charge to within 0.05 2.10 Test Run. An individual emission test kg (0.1 1b). which encompasses the time required to con- 3.6 Moisture Meter. Calibrated electrical sume the mass of the test fuel charge. resistance meter for measuring test fuel 2.11 Usable Firebox Volume. The volume moisture to within 1 percent moisture con- of the firebox determined using the following tent. definitions: 3.7 Anemometer. Device capable of de- 2.11.1 Height. The vertical distance ex- tecting air velocities less than 0.10 m/sec (20 tending above the loading door, if fuel could ft/min), for measuring air velocities near the reasonably occupy that space, but not more test appliance. than 2 inches above the top (peak height) of 3.8 Barometer. Mercury, aneroid or other the loading door, to the floor of the firebox barometer capable of measuring atmospheric (i.e., below a permanent grate) if the grate pressure to within 2.5 mm Hg (0.1 in. Hg).
1001
VerDate 11
3.9 Draft Gauge. Electromanometer or Addition of moisture to previously dried other device for the determination of flue wood is not allowed. It is recommended that draft or static pressure readable to within the test fuel be stored in a temperature and 0.50 Pa (0.002 in. H2O). humidity-controlled room. 3.10 Humidity Gauge. Psychrometer or 4.2.3 Fuel Temperature. The test fuel hygrometer for measuring room humidity. shall be at the test facility temperature 18 to 3.11 Sampling Methods. Use particulate 32 °C (65 to 90 °F). emission measurement Method 5G or Method 4.3 Test Fuel Charge Specifications. 5H to determine particulate concentrations, 4.3.1 Fuel Dimensions. The dimensions of gas flow rates, and particulate emission each test fuel piece shall conform to the rates. nominal measurements of 2 x 4 and 4 x 4 lum- ber. Each piece of test fuel (not including 4. Test Facility. Test Fuel Properties, and Test spacers) shall be of equal length, except as Fuel Charge Specifications necessary to meet requirements in Section 5 4.1 Test Facility. 6.2.5, and shall closely approximate ⁄6 the di- 4.1.1 Wood Heater Flue. Steel flue pipe ex- mensions of the length of the usable firebox. tending to 2.6±0.15 m (8.5±0.5 ft) above the top The fuel piece dimensions shall be deter- of the platform scale, and above this level, mined in relation to the appliance’s firebox insulated solid pack type chimney extending volume according to guidelines listed below: to 4.6+0.3 m (15±1 ft) above the platform 4.3.1.1 If the usable firebox volume is less 3 3 scale, and of the size specified by the wood than or equal to 0.043 m (1.5 ft ), use 2 x 4 heater manufacturer. This applies to both lumber. freestanding and insert type wood heaters. 4.3.1.2 If the usable firebox volume is 3 3 Other chimney types (e.g., solid pack insu- greater than 0.043 m (1.5 ft ) and less than or 3 3 lated pipe) may be used in place of the steel equal to 0.085 m (3.0 ft ), use 2 x 4 and 4 x 4 flue pipe if the wood heater manufacturer’s lumber. About half the weight of the test written appliance specifications require such fuel charge shall be 2 x 4 lumber, and the re- chimney for home installation (e.g., zero mainder shall be 4 x 4 lumber. clearance wood heater inserts). Such alter- 4.3.1.3 If the usable firebox volume is 3 3 native chimney or flue pipe must remain and greater than 0.085 m (3.0 ft ), use 4 x 4 lum- be sealed with the wood heater following the ber. certification test. 4.3.2 Test Fuel Spacers. Air-dried, Doug- 4.1.2 Test Facility Conditions. The test fa- las fir lumber meeting the fuel properties in cility temperature shall be maintained be- Section 4.2. The spacers shall be 130 x 40 x 20 tween 18 and 32 °C (65 and 90 °F) during each mm (5 x 1.5 x 0.75 in.). 4.3.3 Test Fuel Charge Density. The test test run. fuel charge density shall be 112 ± 11.2 kg/m3 Air velocities within 0.6 m (2 ft) of the test (7 ± 0.7 lb/ft3) of usable firebox volume on a appliance and exhaust system shall be less wet basis. than 0.25 m/sec (50 ft/min) without fire in the 4.4 Wood Heater Thermal Equilibrium. unit. The average of the wood heater surface tem- The flue shall discharge into the same peratures at the end of the test run shall space or into a space freely communicating agree with the average surface temperature with the test facility. Any hood or similar at the start of the test run to within 70 °C device used to vent combustion products (125 °F). shall not induce a draft greater than 1.25 Pa (0.005 in. H2O) on the wood heater measured 5. Burn Rate Criteria when the wood heater is not operating. For test facilities with artificially induced 5.1 Burn Rate Categories. One emission barometric pressures (e.g., pressurized cham- test run is required in each of the following bers), the barometric pressure in the test fa- burn rate categories: cility shall not exceed 1,033 mb (30.5 in. Hg) BURN RATE CATEGORIES during any test run. 4.2 Test Fuel Properties. The test fuel (Average kg/hr, dry basis) shall conform to the following requirements: Category 1 Category 2 Category 3 Category 4 4.2.1 Fuel Species. Untreated, air-dried, Douglas fir lumber. Kiln-dried lumber is not <0.80 0.80 to 1.25 1.25 to 1.90 Maximum permitted. The lumber shall be certified C burn rate. grade (standard) or better Douglas fir by a lumber grader at the mill of origin as speci- 5.1.1 Maximum Burn Rate. For Category fied in the West Coast Lumber Inspection 4, the wood heater shall be operated with the Bureau standard No. 16 (incorporated by ref- primary air supply inlet controls fully open erence. See § 60.17). (or, if thermostatically controlled, the ther- 4.2.2 Fuel Moisture. The test fuel shall mostat shall be set at maximum heat out- have a moisture content range between 16 to put) during the entire test run, or the maxi- 20 percent on a wet basis (19 to 25 percent dry mum burn rate setting specified by the man- basis). ufacturer’s written instructions.
1002
VerDate 11
5.1.2 Other Burn Rate Categories. For in place and in operation for at least 50 burn rates in Categories 1 through 3, the hours. Record and report hourly catalyst wood heater shall be operated with the pri- exit temperature data (Section 6.2.2) and the mary air supply inlet control, or other me- hours of operation. chanical control device, set at a predeter- 6.1.2 Non-Catalyst Wood Heater. Operate mined position necessary to obtain the aver- the wood heater using the fuel described in age burn rate required for the category. Section 6.1.1 at a medium burn rate for at 5.2 Alternative Burn Rates for Burn Rate least 10 hours. Record and report the hours Categories 1 and 2. If a wood heater cannot of operation. be operated at a burn rate below 0.80 kg/hr, 6.2 Pretest Preparation. Record the test two test runs shall be conducted with burn fuel charge dimensions and weights, and rates within Category 2. If a wood heater wood heater and catalyst descriptions as cannot be operated at a burn rate below 1.25 shown in the example in Figure 28–3. kg/hr, the flue shall be dampered or the air 6.2.1 Wood Heater Installation. Assemble supply otherwise controlled in order to the wood heater appliance and parts in con- achieve two test runs within Category 2. formance with the manufacturer’s written Evidence that a wood heater cannot be op- installation instructions. Place the wood erated at a burn rate less than 0.80 kg/hr heater centrally on the platform scale and shall include documentation of two or more connect the wood heater to the flue de- attempts to operate the wood heater in burn scribed in Section 4.1.1. Clean the flue with rate Category 1 and fuel combustion has an appropriately sized, wire chimney brush stopped, or results of two or more test runs before each certification test. demonstrating that the burn rates were 6.2.2 Wood Heater Temperature Monitors. greater than 0.80 kg/hr when the air supply For catalyst-equipped wood heaters, locate a controls were adjusted to the lowest possible temperature monitor (optional) about 25 mm position or settings. Stopped fuel combus- (1 in.) upstream of the catalyst at the cen- tion is evidenced when an elapsed time of 30 troid of the catalyst face area, and locate a minutes or more has occurred without a temperature monitor (mandatory) that will measurable (< 0.05 kg (0.1 lb) or 1.0 percent, indicate the catalyst exhaust temperature. whichever is greater) weight change in the This temperature monitor is centrally lo- test fuel charge. See also Section 6.4.3. Re- cated within 25 mm (1 in.) downstream at the port the evidence and the reasoning used to centroid of catalyst face area. Record these determine that a test in burn rate Category locations. 1 cannot be achieved; for example, two at- Locate wood heater surface temperature tempts to operate at a burn rate of 0.4 kg/hr monitors at five locations on the wood heat- are not sufficient evidence that burn rate er firebox exterior surface. Position the tem- Category 1 cannot be achieved. perature monitors centrally on the top sur- face, on two sidewall surfaces, and on the NOTE: After July 1, 1990, if a wood heater bottom and back surfaces. Position the mon- cannot be operated at a burn rate less than itor sensing tip on the firebox exterior sur- 0.80 kg/hr, at least one test run with an aver- face inside of any heat shield, air circulation age burn rate of 1.00 kg/hr or less shall be walls, or other wall or shield separated from conducted. Additionally, if flue dampering the firebox exterior surface. Surface tem- must be used to achieve burn rates below 1.25 perature locations for unusual design shapes kg/hr (or 1.0 kg/hr), results from a test run (e.g., spherical, etc.) shall be positioned so conducted at burn rates below 0.90 kg/hr need that there are four surface temperature mon- not be reported or included in the test run itors in both the vertical and horizontal average provided that such results are re- planes passing at right angles through the placed with results from a test run meeting centroid of the firebox, not including the fuel the criteria above. loading door (total of five temperature mon- itors). 6. Procedures 6.2.3 Test Facility Conditions. Locate the 6.1 Catalytic Combustor and Wood Heater test facility temperature monitor on the Aging. The catalyst-equipped wood heater or horizontal plane that includes the primary a wood heater of any type shall be aged be- air intake opening for the wood heater. Lo- fore the certification test begins. The aging cate the temperature monitor 1 to 2 m (3 to procedure shall be conducted and docu- 6 ft) from the front of the wood heater in the mented by a testing laboratory accredited 90° sector in front of the wood heater. according to procedures in § 60.535 of 40 CFR Use an anemometer to measure the air ve- Part 60. locity. Measure and record the room air ve- 6.1.1 Catalyst-equipped Wood Heater. Op- locity before the pretest ignition period erate the catalyst-equipped wood heater (Section 6.3) and once immediately following using fuel described in Section 4.2 or cord- the test run completion. wood with a moisture content between 15 Measure and record the test facility’s am- and 25 percent on a wet basis. Operate the bient relative humidity, barometric pres- wood heater at a medium burn rate (Cat- sure, and temperature before and after each egory 2 or 3) with a new catalytic combustor test run.
1003
VerDate 11
Measure and record the flue draft or static spacers to the top of the test fuel piece(s) on pressure in the flue at a location no greater top of the test fuel charge is optional. than 0.3 m (1 ft) above the flue connector at To avoid stacking difficulties, or when a the wood heater exhaust during the test run whole number of test fuel pieces does not re- at the recording intervals (Section 6.4.2). sult, all piece lengths shall be adjusted uni- 6.2.4 Wood Heater Firebox Volume. Deter- formly to remain within the specified load- mine the firebox volume using the defini- ing density. The shape of the test fuel crib tions for height, width, and length in Section shall be geometrically similar to the shape 2. Volume adjustments due to presence of of the firebox volume without resorting to firebrick and other permanent fixtures may special angular or round cuts on the individ- be necessary. Adjust width and length di- ual fuel pieces. mensions to extend to the metal wall of the 6.2.6 Sampling Method. Prepare the sam- wood heater above the firebrick or perma- pling equipment as defined by the selected nent obstruction if the firebrick or obstruc- method. Collect one particulate emission tion extending the length of the side(s) or sample for each test run. back wall extends less than one-third of the 6.2.7 Secondary Air Adjustment Valida- usable firebox height. Use the width or tion. If design drawings do not show the in- length dimensions inside the firebrick if the troductions of secondary air into a chamber firebrick extends more than one-third of the outside the firebox (Section 2.4), conduct a usable firebox height. If a log retainer or separate test of the wood heater’s secondary grate is a permanent fixture and the manu- air supply. Operate the wood heater at a facturer recommends that no fuel be placed burn rate in Category 1 (Sections 5.1 or 5.2) outside the retainer, the area outside of the with the secondary air supply operated fol- retainer is excluded from the firebox volume lowing the manufacturer’s written instruc- calculations. tions. Start the secondary air validation test In general, exclude the area above the ash run as described in Section 6.4.1, except no lip if that area is less than 10 percent of the emission sampling is necessary and burn rate usable firebox volume. Otherwise, take into data shall be recorded at 5-minute intervals. account consumer loading practices. For in- After the start of the test run, operate the stance, if fuel is to be loaded front-to-back, wood heater with the secondary air supply an ash lip may be considered usable firebox set as per the manufacturer’s instructions, volume. but with no adjustments to this setting. Include areas adjacent to and above a baf- After 25 percent of the test fuel has been con- fle (up to two inches above the fuel loading sumed, adjust the secondary air supply con- opening) if four inches or more horizontal trols to another setting, as per the manufac- space exist between the edge of the baffle turer’s instructions. Record the burn rate and a vertical obstruction (e.g., sidewalls or data (5-minute intervals) for 20 minutes fol- air channels). lowing the air supply adjustment. 6.2.5 Test Fuel Charge. Prepare the test Adjust the air supply control(s) to the fuel pieces in accordance with the specifica- original position(s), operate at this condition tions in Section 4.3. Determine the test fuel for at least 20 minutes, and repeat the air moisture content with a calibrated electrical supply adjustment procedure above. Repeat resistance meter or other equivalent per- the procedure three times at equal intervals formance meter. (To convert moisture meter over the entire burn period as defined in Sec- readings from the dry basis to the wet basis: tion 6.4. If the secondary air adjustment re- (100)(percent dry reading) ÷ (100 + percent dry sults in a burn rate change of more than an reading) = percent moisture wet basis.) De- average of 25 percent between the 20-minute termine fuel moisture for each fuel piece periods before and after the secondary ad- (not including spacers) by averaging at least justments, the secondary air supply shall be three moisture meter readings, one from considered a primary air supply, and no ad- each of three sides, measured parallel to the justment to this air supply is allowed during wood grain. Average all the readings for all the test run. the fuel pieces in the test fuel charge. If an 6.3 Pretest Ignition. Build a fire in the electrical resistance type meter is used, pen- wood heater in accordance with the manu- etration of insulated electrodes shall be one- facturer’s written instructions. fourth the thickness of the test fuel piece or 6.3.1 Pretest Fuel Charge. Crumpled news- 19 mm (0.75 in.), whichever is greater. Meas- paper loaded with kindling may be used to ure the moisture content within a 4-hour pe- help ignite the pretest fuel. The pretest fuel, riod prior to the test run. Determine the fuel used to sustain the fire, shall meet the same temperature by measuring the temperature fuel requirements prescribed in Section 4.2. of the room where the wood has been stored The pretest fuel charge shall consist of whole for at least 24 hours prior to the moisture de- 2 x 4’s that are no less than 1/3 the length of termination. the test fuel pieces. Pieces of 4 x 4 lumber in Attach the spacers to the test fuel pieces approximately the same weight ratio as for with uncoated, ungalvanized nails or staples the test fuel charge may be added to the pre- as illustrated in Figure 28–1. Attachment of test fuel charge.
1004
VerDate 11
6.3.2 Wood Heater Operation and Adjust- weight of the fuel remaining and start the ments. Set the air inlet supply controls at test run. Record and report any other cri- any position that will maintain combustion teria, in addition to those specified in this of the pretest fuel load. At least one hour be- section, used to determine the moment of fore the start of the test run, set the air sup- the test run start (e.g., firebox or catalyst ply controls at the approximate positions temperature), whether such criteria are spec- necessary to achieve the burn rate desired ified by the wood heater manufacturer or the for the test run. Adjustment of the air sup- testing laboratory. Record all wood heater ply controls, fuel addition or subtractions, individual surface temperatures, catalyst and coalbed raking shall be kept to a mini- temperatures, any initial sampling method mum but are allowed up to 15 minutes prior measurement values, and begin the particu- to the start of the test run. For the purposes late emission sampling. Within 1 minute fol- of this method, coalbed raking is the use of lowing the start of the test run, open the a metal tool (poker) to stir coals, break wood heater door, load the test fuel charge, burning fuel into smaller pieces, dislodge and record the test fuel charge weight. Re- fuel pieces from positions of poor combus- cording of average, rather than individual, tion, and check for the condition of uniform surface temperatures is acceptable for tests charcoalization. Record all adjustments conducted in accordance with § 60.533(o)(3)(i) made to the air supply controls, adjustments of 40 CFR Part 60. to and additions or subtractions of fuel, and Position the fuel charge so that the spac- any other changes to wood heater operations ers are parallel to the floor of the firebox, that occur during pretest ignition period. with the spacer edges abutting each other. If Record fuel weight data and wood heater loading difficulties result, some fuel pieces temperature measurements at 10-minute in- may be placed on edge. If the usable firebox tervals during the hour of the pretest igni- volume is between 0.043 and 0.085 m3 (1.5 and tion period preceding the start of the test 3.0 ft3), alternate the piece sizes in vertical run. During the 15-minute period prior to the stacking layers to the extent possible. For start of the test run, the wood heater loading example, place 2 x 4’s on the bottom layer in door shall not be open more than a total of direct contact with the coal bed and 4 x 4’s 1 minute. Coalbed raking is the only adjust- on the next layer, etc. (See Figure 28–2). Po- ment allowed during this period. sition the fuel pieces parallel to each other and parallel to the longest wall of the firebox NOTE: One purpose of the pretest ignition to the extent possible within the specifica- period is to achieve uniform charcoalization tions in Section 6.2.5. of the test fuel bed prior to loading the test Load the test fuel in appliances having un- fuel charge. Uniform charcoalization is a usual or unconventional firebox design main- general condition of the test fuel bed evi- taining air space intervals between the test denced by an absence of large pieces of burn- fuel pieces and in conformance with the ing wood in the coal bed and the remaining manufacturer’s written instructions. For fuel pieces being brittle enough to be broken any appliance that will not accommodate into smaller charcoal pieces with a metal the loading arrangement specified in the poker. Manipulations to the fuel bed prior to paragraph above, the test facility personnel the start of the test run should be done to shall contact the Administrator for an alter- achieve uniform charcoalization while main- native loading arrangement. taining the desired burn rate. In addition, The wood heater door may remain open some wood heaters (e.g., high mass units) and the air supply controls adjusted up to may require extended pretest burn time and five minutes after the start of the test run in fuel additions to reach an initial average order to make adjustments to the test fuel surface temperature sufficient to meet the charge and to ensure ignition of the test fuel thermal equilibrium criteria in Section 4.4. charge has occurred. Within the five minutes The weight of pretest fuel remaining at the after the start of the test run, close the wood start of the test run is determined as the dif- heater door and adjust the air supply con- ference between the weight of the wood heat- trols to the position determined to produce er with the remaining pretest fuel and the the desired burn rate. No other adjustments tare weight of the cleaned, dry wood heater to the air supply controls or the test fuel with or without dry ash or sand added con- charge are allowed (except as specified in sistent with the manufacturer’s instructions Sections 6.4.3 and 6.4.4) after the first five and the owner’s manual. The tare weight of minutes of the test run. Record the length of the wood heater must be determined with time the wood heater door remains open, the the wood heater (and ash, if added) in a dry adjustments to the air supply controls, and condition. any other operational adjustments. 6.4 Test Run. Complete a test run in each 6.4.2 Data Recording. Record fuel weight burn rate category, as follows: data, wood heater individual surface and cat- 6.4.1 Test Run Start. When the kindling alyst temperature measurements, other and pretest fuel have been consumed to leave wood heater operational data (e.g., draft), a fuel weight between 20 and 25 percent of test facility temperature and sampling the weight of the test fuel charge, record the method data at 10-minute intervals (or more
1005
VerDate 11
frequently at the option of the tester) as runs in that burn rate category shall be used shown on example data sheet, Figure 28–4. in calculating the weighted average emission 6.4.3 Test Fuel Charge Adjustment. The rate (see Section 8.1). The measurement data test fuel charge may be adjusted (i.e., re-po- and results of all test runs shall be reported sitioned) once during a test run if more than regardless of which values are used in cal- 60 percent of the initial test fuel charge culating the weighted average emission rate weight has been consumed and more than 10 (see NOTE: in Section 5.2). minutes have elapsed without a measurable 6.7 Pellet Burning Heaters. Certification < ( 0.05 kg (0.1 lb) or 1.0 percent, whichever is testing procedures for pellet burning wood greater) weight change. The time used to heaters are based on the procedures in this make this adjustment shall be less than 15 method. The differences in the procedures seconds. from the sections in Method 28 are as fol- 6.4.4 Air Supply Adjustment. Secondary lows: air supply controls may be adjusted once during the test run following the manufac- 6.7.1 Test Fuel Properties. The test fuel turer’s written instructions (see Section shall be all wood pellets with a moisture con- 6.2.7). No other air supply adjustments are tent no greater than 20 percent on a wet allowed during the test run. basis (25 percent on a dry basis). Determine Recording of wood heater flue draft during the wood moisture content with either the test run is optional for tests conducted ASTM–D2016–74(82)(Method A) or ASTM in accordance with § 60.533(o)(3)(i) of 40 CFR D4442–84. (incorporated by reference. See Part 60. Section 60.17). 6.4.5 Auxiliary Wood Heater Equipment 6.7.2 Test Fuel Charge Specifications. The Operation. Heat exchange blowers sold with test fuel charge size shall be as per the man- the wood heater shall be operated during the ufacturer s written instructions for main- test run following the manufacturer’s writ- taining the desired burn rate. ten instructions. If no manufacturer’s writ- 6.7.3 Wood Heater Firebox Volume. The ten instructions are available, operate the firebox volume need not be measured or de- heat exchange blower in the ‘‘high’’ position. termined for establishing the test fuel (Automatically operated blowers shall be op- charge size. The firebox dimensions and erated as designed.) Shaker grates, by-pass other heater specifications needed to iden- controls, or other auxiliary equipment may tify the heater for certification purposes be adjusted only one time during the test shall be reported. run following the manufacturer’s written in- 6.7.4 Heater Installation. Arrange the structions. heater with the fuel supply hopper on the Record all adjustments on a wood heater platform scale as described in Section 6.2.1. operational written record. 6.7.5 Pretest Ignition. Start a fire in the NOTE: If the wood heater is sold with a heater as directed by the manufacturer’s heat exchange blower as an option, test the written instructions, and adjust the heater wood heater with the heat exchange blower controls to achieve the desired burn rate. Op- operating as described in Sections 5 and 6 erate the heater at the desired burn rate for and report the results. As an alternative to at least 1 hour before the start of the test repeating all test runs without the heat ex- run. change blower operating, the tester may con- 6.7.6 Sampling Method. Method 5G or 5H duct one test run without the blower operat- shall be used for the certification testing of ing as described in Section 6.4.5 at a burn pellet burners. Prepare the sampling equip- rate in Category 2 (Section 5.1). If the emis- ment as described in Method 5G or 5H. Col- sion rate resulting from this test run with- out the blower operating is equal to or less lect one particulate emission sample for than the emission rate plus 1.0 g/hr for the each test run. test run in burn rate Category 2 with the 6.7.7 Test Run. Complete a test run in blower operating, the wood heater may be each burn rate category as follows: considered to have the same average emis- 6.7.7.1 Test Run Start. When the wood sion rate with or without the blower operat- heater has operated for at least 1 hour at the ing. Additional test runs without the blower desired burn rate, add fuel to the supply hop- operating are unnecessary. per as necessary to complete the test run, 6.5 Consecutive Test Runs. Test runs on a record the weight of the fuel in the supply wood heater may be conducted consecutively hopper (the wood heater weight), and start provided that a minimum one-hour interval the test run. Add no additional fuel to the occurs between test runs. hopper during the test run. 6.6 Additional Test Runs. The testing lab- Record all the wood heater surface tem- oratory may conduct more than one test run peratures, the initial sampling method meas- in each of the burn rate categories specified urement values, the time at the start of the in Section 5.1. If more than one test run is test, and begin the emission sampling. Make conducted at a specified burn rate, the re- no adjustments to the wood heater air sup- sults from at least two-thirds of the test ply or wood supply rate during the test run.
1006
VerDate 11
6.7.7.2 Data Recording. Record the fuel (wood heater) weight data, wood heater tem- n perature and operational data, and emission ∑ sampling data as described in Section 6.4.2. ()KE = i=1 i i 6.7.7.3 Test Run Completion. Continue Ew n Eq. 28 - 1 emission sampling and wood heater oper- Ki ation for 2 hours. At the end of the test run, ∑ stop the particulate sampling, and record the i=1 final fuel weight, the run time, and all final where: measurement values. Ew = Weighted average emission rate, g/hr; 6.7.8 Calculations. Determine the burn Ei = Emission rate for test run, i, from Meth- rate using the difference between the initial od 5G or 5H, g/hr; and final fuel (wood heater) weights and the ki = Test run weighting factor = Pi∂1—Pi±1; procedures described in Section 8.3. Com- n = Total number of test runs; plete the other calculations as described in Pi = Probability for burn rate during test run, Section 8. i, obtained from Table 28–1. Use linear in- terpolation to determine probability val- 7. Calibrations ues for burn rates between those listed 7.1 Platform Scale. Perform a multipoint on the table. calibration (at least five points spanning the NOTE: Po always equals 0, P(n∂1) always operational range) of the platform scale be- equals 1, P1 corresponds to the probability of fore its initial use. The scale manufacturer’s the lowest recorded burn rate, P2 cor- calibration results are sufficient for this pur- responds to the probability of the next low- pose. Before each certification test, audit est burn rate, etc. An example calculation is the scale with the wood heater in place by shown on Figure 28–5. weighing at least one calibration weight 8.2 Average Wood Heater Surface Tem- (Class F) that corresponds to 20 percent to 80 peratures. Calculate the average of the wood percent of the expected test fuel charge heater surface temperatures for the start of weight. If the scale cannot reproduce the the test run (Section 6.3.1) and for the test value of the calibration weight within 0.05 kg run completion (Section 6.3.6). If the two av- (0.1 lbs) or 1 percent of the expected test fuel erage temperatures do not agree within 70 °C charge weight, whichever is greater, recali- (125 °F), report the test run results, but do brate the scale before use with at least five not include the test run results in the test calibration weights spanning the operational average. Replace such test run results with range of the scale. results from another test run in the same 7.2 Balance (optional). Calibrate as de- burn rate category. scribed in Section 7.1. 8.3 Burn Rate. 7.3 Temperature Monitor. Calibrate as in 60 WM 100 − % Method 2, Section 4.3, before the first certifi- BR = wd w Eq.-28 2 cation test and semiannually thereafter. θ 7.4 Moisture Meter. Calibrate as per the 100 manufacturer’s instructions before each cer- Where: tification test. BR = Dry wood burn rate, kg/hr (lb/hr) 7.5 Anemometer. Calibrate the anemom- Wwd = Total mass of wood burned during the eter as specified by the manufacturer’s in- test run, kg (lb) structions before the first certification test θ - Total time of test run, min. and semiannually thereafter. %Mw = Average moisture in test fuel charge, 7.6 Barometer. Calibrate against a mer- wet basis, percent. cury barometer before the first certification 8.4 Reporting Criteria. Submit both raw test and semiannually thereafter. and reduced test data for wood heater tests. 7.7 Draft Gauge. Calibrate as per the man- Specific reporting requirements are as fol- ufacturer’s instructions; a liquid manometer lows: does not require calibration. 8.4.1 Wood Heater Identification. Report wood heater identification information. An 7.8 Humidity Gauge. Calibrate as per the example data form is shown on Figure 28–4. manufacturer’s instructions before the first 8.4.2 Test Facility Information. Report certification test and semiannually there- test facility temperature, air velocity, and after. humidity information. An example data 8. Calculations and Reportinq form is shown on Figure 28–4. 8.4.3 Test Equipment Calibration and Carry out calculations retaining at least Audit Information. Report calibration and one extra decimal figure beyond that of the audit results for the platform scale, test fuel acquired data. Round off figures after the balance, test fuel moisture meter, and sam- final calculation. pling equipment including volume metering 8.1 Weighted Average Emission Rate. systems and gaseous analyzers.
1007
VerDate 11
8.4.4 Pretest Procedure Description. Re- 2. Analytical methods—brief description of port all pretest procedures including pretest sample recovery and analysis procedures. fuel weight, burn rates, wood heater tem- peratures, and air supply settings. An exam- f. Quality Control and Assurance Procedures ple data form is shown on Figure 28–4. and Results 8.4.5 Particulate Emission Data. Report a summary of test results for all test runs and 1. Calibration procedures and results—cer- the weighted average emission rate. Submit tification procedures, sampling and analysis copies of all data sheets and other records procedures. collected during the testing. Submit exam- 2. Test method quality control proce- ples of all calculations. dures—leak-checks, volume meter checks, 8.4.6 Suggested Test Report Format. stratification (velocity) checks, proportion- ality results. a. Introduction 1. Purpose of test—certification, audit, ef- APPENDICES ficiency, research and development. 1. Results and Example Calculations. Com- 2. Wood heater identification—manufac- plete summary tables and accompanying ex- turer, model number, catalytic/ noncata- amples of all calculations. lytic, options. 3. Laboratory—name, location (altitude), 2. Raw Data. Copies of all uncorrected data participants. sheets for sampling measurements, tempera- 4. Test information—date wood heater re- ture records and sample recovery data. Cop- ceived, date of tests, sampling methods used, ies of all pretest burn rate and wood heater number of test runs. temperature data. 3. Sampling and Analytical Procedures. b. Summary and Discussion of Results Detailed description of procedures followed 1. Table of results (in order of increasing by laboratory personnel in conducting the burn rate)—test run number, burn rate, par- certification test, emphasizing particularly ticulate emission rate, efficiency (if deter- parts of the procedures differing from the mined), averages (indicate which test runs methods (e.g., approved alternatives). are used). 4. Calibration Results. Summary of all 2. Summary of other data—test facility calibrations, checks, and audits pertinent to conditions, surface temperature averages, certification test results with dates. catalyst temperature averages, pretest fuel 5. Participants. Test personnel, manufac- weights, test fuel charge weights, run times. turer representatives, and regulatory observ- 3. Discussion—Burn rate categories ers. achieved, test run result selection, specific test run problems and solutions. 6. Sampling And Operation Records. Copies of uncorrected records of activities not in- c. Process Description cluded on raw data sheets (e.g., wood heater 1. Wood heater dimensions—volume, door open times and durations). height, width, lengths (or other linear di- 7. Additional Information. Wood heater mensions), weight, volume adjustments. manufacturer’s written instructions for op- 2. Firebox configuration—air supply loca- eration during the certification test. tions and operation, air supply introduction location, refractory location and dimensions, 9. Bibliography catalyst location, baffle and by-pass location 1. Oregon Department of Environmental and operation (include line drawings or pho- Quality Standard Method for Measuring the tographs). Emissions and Efficiencies of Woodstoves, 3. Process operation during test—air sup- June 8, 1984. Pursuant to Oregon Administra- ply settings and adjustments, fuel bed ad- justments, draft. tive Rules Chapter 340, Division 21. 4. Test fuel—test fuel properties (moisture 2. American Society for Testing Materials. and temperature), test fuel crib description Proposed Test Methods for Heating Perform- (include line drawing or photograph), test ance and Emissions of Residential Wood- fuel charge density. Fired Closed Combustion-Chamber Heating Appliances. E–6 Proposal P 180. August, 1986. d. Sampling Locations 3. Radian Corporation, OMNI Environ- Describe sampling location relative to mental Services, Inc., Cumulative Prob- wood heater. Include drawing or photograph. ability for a Given Burn Rate Based on Data Generated in the CONEG and BPA Studies. e. Sampling and Analytical Procedures Package of materials submitted to the Fifth 1. Sampling methods—brief reference to Session of the Regulatory Negotiation Com- operational and sampling procedures and op- mittee, July 16–17, 1986. tional and alternative procedures used.
1008
VerDate 11
1009
VerDate 11
1010
VerDate 11
1011
VerDate 11
1012
VerDate 11
FIGURE 28±5ÐEXAMPLE CALCULATION OF Burn Test number Pi Ei Ki WEIGHTED AVERAGE EMISSION RATE rate
Burn 1 ...... 0.65 0.121 5.0 0.300 Test rate Emis- 2 ...... 0.90 0.300 4.7 0.259 Burn rate category num- (Dry- sions 3 ...... 1.00 0.380 5.3 0.422 ber kg/ (g/hr) 4 ...... 1.45 0.722 3.8 0.532 hr) 5 ...... 2.00 0.912 5.1 0.278 1 ...... 1 0.65 5.0 1 2 ...... 2 0.85 6.7 K1=P2¥Po=0.300¥0=0.300 2 ...... 3 0.90 4.7 K =P P =0.380 0.121=0.259 2 ...... 4 1.00 5.3 2 3¥ 1 ¥ 3 ...... 5 1.45 3.8 K3=P4¥P2=0.722¥0.300=0.422 4 ...... 6 2.00 5.1 K4=P5¥P3=0.912¥0.380=0.532 1 As permitted in Section 6.6, this test run may be omitted K5=P6¥P4=1¥0.722=0.278 from the calculation of the weighted average emission rate because three runs were conducted for this burn rate category.
Ew equals (0.3)(5.0) + (0.259)(4.7) + (0.422)(5.3) + TABLE 28±1ÐBURN RATE WEIGHTED PROB- (0.532)(3.8) + (0.278)(5.1) divided by 1.791 ABILITIES FOR CALCULATING WEIGHTED AVER- E =4.69 g/hr. w AGE EMISSION RATESÐContinued TABLE 28±1ÐBURN RATE WEIGHTED PROB- Burn rate (kg/hr-dry) Cumulative Prob- ABILITIES FOR CALCULATING WEIGHTED AVER- ability (P) AGE EMISSION RATES 1.00 ...... 0.380 1.05 ...... 0.407 Cumulative Prob- 1.10 ...... 0.460 Burn rate (kg/hr-dry) ability (P) 1.15 ...... 0.490 0.00 ...... 0.000 1.20 ...... 0.550 0.05 ...... 0.002 1.25 ...... 0.572 0.10 ...... 0.007 1.30 ...... 0.620 0.15 ...... 0.012 1.35 ...... 0.654 0.20 ...... 0.016 1.40 ...... 0.695 0.25 ...... 0.021 1.45 ...... 0.722 0.30 ...... 0.028 1.50 ...... 0.750 0.35 ...... 0.033 1.55 ...... 0.779 0.40 ...... 0.041 1.60 ...... 0.800 0.45 ...... 0.054 1.65 ...... 0.825 0.50 ...... 0.065 1.70 ...... 0.840 0.55 ...... 0.086 1.75 ...... 0.857 0.60 ...... 0.100 1.80 ...... 0.875 0.65 ...... 0.121 1.85 ...... 0.882 0.70 ...... 0.150 1.90 ...... 0.895 0.75 ...... 0.185 1.95 ...... 0.906 0.80 ...... 0.220 2.00 ...... 0.912 0.85 ...... 0.254 2.05 ...... 0.920 0.90 ...... 0.300 2.10 ...... 0.925 0.95 ...... 0.328 2.15 ...... 0.932
1013
VerDate 11
TABLE 28±1ÐBURN RATE WEIGHTED PROB- affected facility, as specified in 40 CFR ABILITIES FOR CALCULATING WEIGHTED AVER- 60.530. 1.2 Principle. A gas sample is extracted AGE EMISSION RATESÐContinued from a location in the stack of a wood-fired Cumulative Prob- appliance while the appliance is operating at Burn rate (kg/hr-dry) ability (P) a prescribed set of conditions. The gas sam- ple is analyzed for percent carbon dioxide 2.20 ...... 0.936 (CO2), percent oxygen (O2), and percent car- 2.25 ...... 0.940 bon monoxide (CO). These stack gas compo- 2.30 ...... 0.945 2.35 ...... 0.951 nents are measured for determining dry mo- 2.40 ...... 0.956 lecular weight of exhaust gas. Total moles of 2.45 ...... 0.959 exhaust gas are determined stoichiomet- 2.50 ...... 0.964 rically. Air to fuel ratio is determined by re- 2.55 ...... 0.968 lating the mass of dry combustion air to the 2.60 ...... 0.972 mass of dry fuel consumed. 2.65 ...... 0.975 2.70 ...... 0.977 2. Definitions 2.75 ...... 0.979 2.80 ...... 0.980 2.1 Burn Rate, Firebox, Secondary Air Sup- 2.85 ...... 0.981 ply, Test Facility, Test Fuel Charge, Test 2.90 ...... 0.982 Fuel Crib, Test Fuel Loading Density, Test 2.95 ...... 0.984 3.00 ...... 0.984 Fuel Piece, Test Run, Usable Firebox Vol- 3.05 ...... 0.985 ume, and Wood Heater. Same as Method 28, 3.10 ...... 0.986 Sections 2.1 and 2.3 to 2.12. 3.15 ...... 0.987 2.2 Air to Fuel Ratio. Ratio of the mass of 3.20 ...... 0.987 dry combustion air introduced into the fire- 3.25 ...... 0.988 box, to the mass of dry fuel consumed (grams 3.30 ...... 0.988 3.35 ...... 0.989 of dry air per gram of dry wood burned). 3.40 ...... 0.989 3.45 ...... 0.989 3. Apparatus 3.50 ...... 0.990 3.1 Test Facility. Insulated Solid Pack 3.55 ...... 0.991 Chimney, Platform Scale and Monitor, Room 3.60 ...... 0.991 3.65 ...... 0.992 Temperature Monitor, Balance, Moisture 3.70 ...... 0.992 Meter, Anemometer, Barometer, Draft 3.75 ...... 0.992 Gauge, and Humidity Gauge. Same as Meth- 3.80 ...... 0.993 od 28, Sections 3.1, 3.2, and 3.4 to 3.10, respec- 3.85 ...... 0.994 tively. 3.90 ...... 0.994 3.2 Sampling System. Probe, Condenser, 3.95 ...... 0.994 Valve, Pump, Rate Meter, Flexible Bag, 4.00 ...... 0.994 4.05 ...... 0.995 Pressure Gauge, and Vacuum Gauge. Same 4.10 ...... 0.995 as Method 3, Sections 2.2.1 to 2.2.8, respec- 4.15 ...... 0.995 tively. The sampling systems described in 4.20 ...... 0.995 Method 5H, Sections 2.2.1, 2.2.2, and 2.2.3, 4.25 ...... 0.995 may be used. 4.30 ...... 0.996 3.3 Analysis. Orsat analyzer, same as Meth- 4.35 ...... 0.996 4.40 ...... 0.996 od 3, Section 2.3; or instrumental analyzers, 4.45 ...... 0.996 same as Method 5H, Sections 2.2.4 and 2.2.5, 4.50 ...... 0.996 for CO2 and CO analyzers, except use a CO 4.55 ...... 0.996 analyzer with a range of 0 to 5 percent and 4.60 ...... 0.996 use a CO2 analyzer with a range of 0 to 5 per- 4.65 ...... 0.996 cent. Use an O analyzer capable of providing 4.70 ...... 0.996 2 4.75 ...... 0.997 a measure of O2 in the range of 0 to 25 per- 4.80 ...... 0.997 cent by volume at least once every 10 min- 4.85 ...... 0.997 utes. Prepare cylinder gases for the three 4.90 ...... 0.997 analyzers as described in Method 5H, Section 4.95 ...... 0.997 3.3. ≥5.00 ...... 1.000 4. Test Preparation METHOD 28A—MEASUREMENT OF AIR TO FUEL 4.1 Test Facility, Wood Heater Appliance RATIO AND MINIMUM ACHIEVABLE BURN Installation, and Test Facility Conditions. RATES FOR WOOD-FIRED APPLIANCES Same as Method 28, Sections 4.1.1 and 4.1.2, respectively, with the exception that baro- 1. Applicability and Principle metric dampers or other devices designed to 1.1 Applicability. This method is applicable introduce dilution air downstream of the for the measurement of air to fuel ratios and firebox shall be sealed. minimum achievable burn rates, for deter- 4.2 Wood Heater Air Supply Adjustments. mining whether a wood-fired appliance is an This section describes how dampers are to be
1014
VerDate 11
set or adjusted and air inlet ports closed or household tools by the removal of parts or sealed during Method 28A tests. The speci- the addition of parts generally available at fications in this section are intended to en- retail stores (e.g., addition of a pipe cap or sure that affected facility determinations plug, addition of a small metal plate to an are made on the facility configurations that inlet hole on a nondecorative sheet metal could reasonably be expected to be employed surface, or removal of riveted or screwed by the user. They are also intended to pre- damper stops). vent circumvention of the standard through (d) Any flue damper, other adjustment the addition of an air port that would often mechanisms or other air inlet ports that are be blocked off in actual use. These specifica- found and documented in several (e.g., a tions are based on the assumption that con- number sufficient to reasonably conclude sumers will remove such items as dampers or that the practice is not unique or uncom- other closure mechanism stops if this can be mon) actual installations as having been ad- done readily with household tools; that con- justed to a more closed position, or closed by sumers will block air inlet passages not visi- consumers, installers, or dealers. ble during normal operation of the appliance 4.2.2 Air Supply Adjustments During Test. using aluminum tape or parts generally The test shall be performed with all air in- available at retail stores; and that consum- lets identified under this section in the ers will cap off any threaded or flanged air closed or most closed position or in the con- inlets. They also assume that air leakage figuration which otherwise achieves the low- around glass doors, sheet metal joints or est air inlet (e.g., greatest blockage). through inlet grilles visible during normal For the purposes of this section, air flow operation of the appliance would not be fur- shall not be minimized beyond the point nec- ther blocked or taped off by a consumer. essary to maintain combustion or beyond It is not the intention of this section to the point that forces smoke into the room. cause an appliance that is clearly designed, Notwithstanding Section 4.2.1, any flue intended, and, in most normal installations, damper, adjustment mechanism or air inlet used as a fireplace to be converted into a port (whether or not equipped with flue wood heater for purposes of applicability dampers or adjusting mechanisms) that is testing. Such a fireplace would be identifi- visible during normal operation of the appli- able by such features as large or multiple ance and which could not reasonably be glass doors or panels that are not gasketed, closed further or blocked except through relatively unrestricted air inlets intended, in means that would significantly degrade the large part, to limit smoking and fogging of aesthetics of the facility (e.g., through use of glass surfaces, and other aesthetic features duct tape) will not be closed further or not normally included in wood heaters. blocked. 4.2.1 Adjustable Air Supply Mechanisms. 4.3 Test Fuel Properties and Test Fuel Any commercially available flue damper, Charge Specifications. Same as Method 28, other adjustment mechanism or other air Sections 4.2 to 4.3, respectively. inlet port that is designed, intended or oth- 4.4 Sampling System. erwise reasonably expected to be adjusted or 4.4.1 Sampling Location. Same as Method closed by consumers, installers, or dealers 5H, Section 5.1.2. and which could restrict air into the firebox 4.4.2 Sampling System Set Up. Set up the shall be set so as to achieve minimum air sampling equipment as described in Method into the firebox, i.e., closed off or set in the 3, Section 3.2, or as in Method 3A, Section 7. most closed position. Flue dampers, mechanisms and air inlet 5. Procedures ports which could reasonably be expected to 5.1 Pretest Preparation. Same as Method be adjusted or closed would include: 28, Sections 6.2.1 and 6.2.3 to 6.2.5. (a) All internal or externally adjustable 5.2 Pretest Ignition. Same as Method 28, mechanisms (including adjustments that af- Section 6.3. Set the wood heater air supply fect the tightness of door fittings) that are settings to achieve a burn rate in Category 1 accessible either before and/or after installa- or the lowest achievable burn rate (see Sec- tion. tion 4.2). (b) All mechanisms, other inlet ports, or 5.3 Test Run. Same as Method 28, Section inlet port stops that are identified in the 6.4. Begin sample collection at the start of owner’s manual or in any dealer literature as the test run as defined in Method 28, Section being adjustable or alterable. For example, 6.4.1. If Method 3 is used, collect a minimum an inlet port that could be used to provide of two bag samples simultaneously at a con- access to an outside air duct but which is stant sampling rate for the duration of the identified as being closable through use of test run. A minimum sample volume of 30 1 additional materials whether or not they are per bag is recommended. If instrumental gas supplied with the facility. concentration measurement procedures are (c) Any combustion air inlet port or com- used, conduct the gas measurement system mercially available flue damper or mecha- performance specifications checks as de- nism stop, which would readily lend itself to scribed in Method 5H, Sections 6.7, 6.8, and closure by consumers who are handy with 6.9. The zero drift and calibration drift limits
1015
VerDate 11
for all three analyzers shall be 0.2 percent O2, CO2) values for the two analyses agree within CO2, or CO, as applicable, or less. Other 0.5 percent (e.g., 6.0 percent O2 for bag 1 and measurement system performance specifica- 6.5 percent O2 for bag 2, agree within 0.5 per- tions are as defined in Method 5H, Section 4. cent), the results of the bag analyses may be Sample at a constant rate for the duration of averaged for the calculations in Section 6. If the test run. the analysis results do not agree within 0.5 5.3.1 Data Recording. Record wood heater percent for each component, calculate the operational data, test facility temperature, air-to-fuel ratio using both sets of analyses sample train flow rate, and fuel weight data and report the results. at 10-minute intervals. 5.3.2 Test Run Completion. Same as Meth- 6. Calculations od 28, Section 6.4.6. 5.4 Analysis Procedure. Carry out calculations, retaining at least 5.4.1 Method 3 Integrated Bag Samples. one extra decimal figure beyond that of the Within 4 hours after the sample collection, acquired data. Round off figure after the analyze each bag sample for percent CO2, O2, final calculation. Other forms of the equa- and CO using an Orsat analyzer as described tions may be used as long as they give equiv- in Method 3, Sections 4.2.5 through 4.2.7. alent results. 5.4.2 Instrumental Analyzers. Average the 6.1 Nomenclature. percent CO , CO, and O values for the test 2 2 Md=Dry molecular weight, g/g-mole(lb/lb- run. mole). 5.5 Quality Control Procedures. %CO =Percent CO by volume (dry basis). 5.5.1 Data Validation. The following qual- 2 2 ity control procedure is suggested to provide %O2=Percent O2 by volume (dry basis). a check on the quality of the data. %CO=Percent CO by volume (dry basis). 5.5.1.1 Calculate a fuel factor, F0, using the %N2=Percent N2 by volume (dry basis). following equation: NT=Total gram-moles of dry exhaust gas per kg of wood burned (lb-moles/lb). 20.% 9 − O YCO2=Measured mole fraction of CO2 (e.g., 10 = 2 percent CO =.10 mole fraction), g/g-mole Fo Eq. 28 a - 4 2 (lb/lb-mole). %CO2 YCO=Measured mole fraction of CO (e.g., 1 where: percent CO=.01 mole fraction), g/g-mole %O2 Percent O2 by volume (dry basis). (lb/lb-mole). %CO Percent CO by volume (dry basis). 2 2 YHC=Assumed mole fraction of HC (dry as 20.9 Percent O by volume in ambient air. 2 CH4) If CO is present in quantities measurable by =0.0088 for catalytic wood heaters; this method, adjust the O2 and CO2 values be- =0.0132 for noncatalytic wood heaters. fore performing the calculation for F0 as fol- =0.0080 for pellet-fired wood heaters. lows: 0.280=Molecular weight of N2 or CO, divided %CO2 (adj) = %CO2 + %CO by 100. %O2 (adj) = %O2 ¥ 0.5 %CO 0.320=Molecular weight of O2 divided by 100. where: 0.440=Molecular weight of CO2 divided by 100. %CO = Percent CO by volume (dry basis). 42.5=Gram-moles of carbon in 1 kg of dry 5.5.1.2 Compare the calculated F0 factor wood assuming 51 percent carbon by with the expected F0 range for wood (1.000 - weight dry basis (.0425 lb/lb). 1.120). Calculated F0 values beyond this ac- 510=Grams of carbon in exhaust gas per kg of ceptable range should be investigated before wood burned. accepting the test results. For example, the 1,000=Grams in 1 kg. strength of the solutions in the gas analyzer 6.2 Dry Molecular Weight. Use Equation and the analyzing technique should be 28a–1 to calculate the dry molecular weight checked by sampling and analyzing a known of the stack gas. concentration, such as air. If no detectable or correctable measurement error can be Md=0.440(%CO2)+0.320(%O)2)+0.280(%N2+%CO) identified, the test should be repeated. Alter- Eq. 28a–1 natively, determine a range of air to fuel NOTE: The above equation does not con- ratio results that could include the correct sider argon in air (about 0.9 percent, molecu- lar weight of 37.7). A negative error of about value by using an F0 value of 1.05 and cal- 0.4 percent is introduced. The tester may opt culating a potential range of CO2 and O2 val- ues. Acceptance of such results will be based to include argon in the analysis using proce- on whether the calculated range includes the dures subject to approval of the Adminis- exemption limit and the judgment of the ad- trator. ministrator. 6.3 Dry Moles of Exhaust Gas. Use Equa- 5.5.1.3 Method 3 Analyses. Compare the tion 28a–2 to calculate the total moles of dry results of the analyses of the two bag sam- exhaust gas produced per kilogram of dry ples. If all the gas components (O2, CO, and wood burned. 1016
VerDate 11
Similarly, inductively coupled plasma-mass 42. 5 spectroscopy (ICP–MS) may be used for anal- N = Eq. 28 a - 2 ysis of Sb, As, Ba, Be, Cd, Cr, Co, Cu, Pb, Mn, T + + Ni, As, Tl and Zn. ()YYYCO2 CO HC 6.4 Air to Fuel Ratio. Use Equation 28a–3 2. Range, Detection Limits, Precision, and to calculate the air to fuel ratio on a dry Interferences mass basis. 2.1 Range. For the analysis described and for similar analyses, the ICAP response is × − () linear over several orders of magnitude. ()NMT d 510 AF/ = Eq . 28 a - 3 Samples containing metal concentrations in ()1000 the nanograms per ml (ng/ml) to micrograms per ml (µg/ml) range in the final analytical 6.5 Burn Rate. Calculate the fuel burn solution can be analyzed using this method. rate as in Method 28, Section 8.3. Samples containing greater than approxi- mately 50 µg/ml As, Cr, or Pb should be di- 7. Bibliography luted to that level or lower for final analysis. Same as Method 3, Section 7, and Method Samples containing greater than approxi- 5H, Section 7. mately 20 µg/ml of Cd should be diluted to that level before analysis. METHOD 29—DETERMINATION OF METALS 2.2 Analytical Detection Limits. (NOTE: EMISSIONS FROM STATIONARY SOURCES See section 2.3 for the description of in-stack detection limits.) 1. Applicability and Principle 2.2.1 ICAP analytical detection limits for 1.1 Applicability. This method is applica- the sample solutions (based on Method 6010 in EPA Publication SW–846, Third Edition ble to the determination of antimony (Sb), (November 1986) including updates I, II, IIA, arsenic (As), barium (Ba), beryllium (Be), and IIB, as incorporated by reference in cadmium (Cd), chromium (Cr), cobalt (Co), § 60.17(i)) are approximately as follows: Sb (32 copper (Cu), lead (Pb), manganese (Mn), mer- ng/ml), As (53 ng/ml), Ba (2 ng/ml), Be (0.3 ng/ cury (Hg), nickel (Ni), phosphorus (P), sele- ml), Cd (4 ng/ml), Cr (7 ng/ml), Co (7 ng/ml), nium (Se), silver (Ag), thallium (T1), and Cu (6 ng/ml), Pb (42 ng/ml), Mn (2 ng/ml), Ni zinc (Zn) emissions from stationary sources. (15 ng/ml), P (75 ng/ml), Se (75 ng/ml), Ag (7 This method may be used to determine par- ng/ml), Tl (40 ng/ml), and Zn (2 ng/ml). ICP– ticulate emissions in addition to the metals MS analytical detection limits (based on emissions if the prescribed procedures and based on Method 6020 in EPA Publication precautions are followed. SW–846, Third Edition (November 1986) as in- 1.1.1 Hg emissions can be measured, alter- corporated by reference in § 60.17(i)) are natively, using EPA Method 101A of Appen- lower generally by a factor of ten or more. dix B, 40 CFR Part 61. Method 101–A meas- Be is lower by a factor of three. The actual ures only Hg but it can be of special interest sample analytical detection limits are sam- to sources which need to measure both Hg ple dependent and may vary due to the sam- and Mn emissions. ple matrix. 1.2 Principle. A stack sample is with- 2.2.2 The analytical detection limits for drawn isokinetically from the source, partic- analysis by direct aspiration AAS are ap- ulate emissions are collected in the probe proximately as follow: Sb (200 ng/ml), As (2 and on a heated filter, and gaseous emissions ng/ml), Ba (100 ng/ml), Be (5 ng/ml), Cd (5 ng/ are then collected in an aqueous acidic solu- ml), Cr (50 ng/ml), Co (50 ng/ml), Cu (20 ng/ tion of hydrogen peroxide (analyzed for all ml), Pb (100 ng/ml), Mn (10 ng/ml), Ni (40 ng/ metals including Hg) and an aqueous acidic ml), Se (2 ng/ml), Ag (10 ng/ml), Tl (100 ng/ solution of potassium permanganate (ana- ml), and Zn (5 ng/ml). lyzed only for Hg). The recovered samples 2.2.3 The detection limit for Hg by are digested, and appropriate fractions are CVAAS (on the resultant volume of the analyzed for Hg by cold vapor atomic absorp- disgestion of the aliquots taken for Hg analy- tion spectroscopy (CVAAS) and for Sb, As, ses) can be approximately 0.02 to 0.2ng/ml, Ba, Be, Cd, Cr, Co, Cu, Pb, Mn, Ni, P, Se, Ag, depending upon the type of CVAAS analyt- Tl, and Zn by inductively coupled argon plas- ical instrument used. ma emission spectroscopy (ICAP) or atomic 2.2.4 The use of GFAAS can enhance the absorption spectroscopy (AAS). Graphite fur- detection limits compared to direct aspira- nace atomic absorption spectroscopy tion AAS as follows: Sb (3 ng/ml), As (1 ng/ (GFAAS) is used for analysis of Sb, As, Cd, ml), Be (0.2 ng/ml), Cd (0.1 ng/ml), Cr (1 ng/ Co, Pb, Se, and Tl if these elements require ml), Co (1 ng/ml), Pb (1 ng/ml), Se (2 ng/ml), greater analytical sensitivity than can be and T1 (ng/ml). obtained by ICAP. If one so chooses, AAS 2.3 In-stack Detection Limits. may be used for analysis of all listed metals 2.3.1 For test planning purposes in-stack if the resulting in-stack method detection detection limits can be developed by using limits meet the goal of the testing program. the following information (1) the procedures
1017
VerDate 11
described in this method, (2) the analytical od detection limits for the above set of con- detection limits described in Section 2.2 and ditions are presented in Table 29–1 and were in EPA Publication SW–846, Third Edition calculated by using Eq. 29–1. (November 1986) including updates I, II, IIA A×B/C=D Eq. 29–1 and IIB, as incorporated by reference in Where: § 60.17(i), (3) the normal volumes of 300 ml A=Analytical detectin limit, µg/ml. (Analytical Fraction 1) for the front-half and B=Liquid volume of digested sample prior to 150 ml (Analytical Fraction 2A) for the back- aliquotting for analysis, Ml. half samples, and (4) a stack gas sample vol- C=Stack sample gas volume, dsm3. 3 ume of 1.25 m . The resultant in-stack meth- D=In-stack detection limit, µg/m3.
TABLE 29±1.ÐIN-STACK METHOD DETECTION LIMITS (µg/m 3) FOR THE FRONT-HALF, THE BACK- HALF, AND THE TOTAL SAMPLING TRAIN USING ICAP AND AAS
Metal Front-half: Probe and Back-half: Impingers Back-half: Impingers Total train: filter 1±3 (4±6) a
Antimony ...... 1 7.7 (0.7) 1 3.8 (0.4) 1 11.5 (1.1) Arsenic ...... 1 12.7 (0.3) 1 6.4 (0.1) 1 19.1 (0.4) Barium ...... 0.5 0.3 0.8 Beryllium ...... 1 0.07 (0.05) 1 0.04 (0.03) 1 0.11 (0.08) Cadmium ...... 1 1.0 (0.02) 1 0.5 (0.01) 1 1.5 (0.03) Chromium ...... 1 1.7 (0.2) 1 0.8 (0.1) 1 2.5 (0.3) Cobalt ...... 1 1.7 (0.2) 1 0.8 (0.1) 1 2.5 (0.3) Copper ...... 1.4 0.7 2.1 Lead ...... 1 10.1 (0.2) 1 5.0 (0.1) 1 15.1 (0.3) Manganese ...... 1 0.5 (0.2) 1 0.2 (0.1) 1 0.7 (0.3) Mercury ...... 2 0.06 2 0.3 2 0.2 2 0.56 Nickel ...... 3.6 1.8 5.4 Phosphorus ...... 18 9 27 Selenium ...... 1 18 (0.5) 1 9 (0.3) 1 27 (0.8) Silver ...... 1.7 0.9 2.6 Thallium ...... 1 9.6 (0.2) 1 4.8 (0.1) 1 14.4 (0.3) Zinc ...... 0.5 0.3 0.8 a Mercury analysis only. 1 Detection limit when analyzed by GFAAS. 2 Detection limit when analyzed by CVAAS, estimated for Back-Half and Total Train. See Sections 2.2 and 5.4.3. Note: Actual method in-stack detection limits may vary from these values, as described in Section 2.3.3.
2.3.2 To ensure optimum precision/resolu- analysis can be increased to as much as 10 tion in the analyses, the target concentra- ml, thus improving the in-stack detection tions of metals in the analytical solutions limit by a factor of ten compared to a 1 ml should be at least ten times their respective aliquot size. analytical detection limits. Under certain 2.3.3.1 A nominal one hour sampling run conditions, and with greater care in the ana- will collect a stack gas sampling volume of lytical procedure, these concentrations can about 1.25 m3. If the sampling time is in- be as low as approximately three times the creased to four hours and 5 m3 are collected, respective analytical detection limits with- the in-stack method detection limits would out seriously impairing the precision of the be improved by a factor of four compared to analyses. On at least one sample run in the the values shown in Table source test, and for each metal analyzed, 29–1. perform either repetitive analyses, Method 2.3.3.2 The in-stack detection limits as- of Standard Additions, serial dilution, or ma- sume that all of the sample is digested and trix spike addition, etc., to document the the final liquid volumes for analysis are the quality of the data. normal values of 300 ml for Analytical Frac- 2.3.3 Actual in-stack method detection tion 1, and 150 ml for Analytical Fraction 2A. limits are based on actual source sampling If the volume of Analytical Fraction 1 is re- parameters and analytical results as de- duced from 300 to 30 ml, the in-stack detec- scribed above. If required, the method in- tion limits for that fraction of the sample stack detection limits can be improved over would be improved by a factor of ten. If the those shown in Table 29–1 for a specific test volume of Analytical Fraction 2A is reduced by either increasing the sampled stack gas from 150 to 25 ml, the in-stack detection lim- volume, reducing the total volume of the di- its for that fraction of the sample would be gested samples, improving the analytical de- improved by a factor of six. Matrix effect tection limits, or any combination of the checks are necessary on sample analyses and three. For extremely low levels of Hg only, typically are of much greater significance the aliquot size selected for digestion and for samples that have been concentrated to
1018
VerDate 11
less than the normal original sample vol- Ag were not detected in the tests. However, ume. Reduction of Analytical Fractions 1 based on the analytical detection limits of and 2A to volumes of less than 30 and 25 ml, the ICAP for these metals, their precisions respectively, could interfere with the re- could be similar to those for the other met- dissolving of the residue and could increase als when detected at similar levels. interference by other compounds to an intol- 2.5 Interferences. Iron (Fe) can be a spec- erable level. tral interference during the analysis of As, 2.3.3.3 When both of the modifications de- Cr, and Cd by ICAP. Aluminum (Al) can be a scribed in Sections 2.3.3.1 and 2.3.3.2 are used spectral interference during the analysis of simultaneously on one sample, the resultant As and Pb by ICAP. Generally, these inter- improvements are multiplicative. For exam- ferences can be reduced by diluting the ana- ple, an increase in stack gas volume by a fac- lytical sample, but such dilution raises the tor of four and a reduction in the total liquid in-stack detection limits. Background and sample digested volume of both Analytical overlap corrections may be used to adjust for Fractions 1 and 2A by a factor of six would spectral interferences. Refer to Method 6010 result in an improvement by a factor of in EPA Publication SW–846 Third Edition twenty-four of the in-stack method detection (November 1986) including updates I, II, IIA limit. and IIB, as incorporated by reference in 2.4 Precision. The precision (relative § 60.17(i) the other analytical methods used standard deviation) for each metal detected for details on potential interferences to this in a method development test performed at a method. For all GFAAS analyses, use matrix sewage sludge incinerator were found to be modifiers to limit interferences, and matrix as follows: Sb (12.7 percent), As (13.5 per- match all standards. cent), Ba (20.6 percent), Cd (11.5 percent), Cr (11.2 percent), Cu (11.5 percent), Pb (11.6 per- 3. Apparatus cent), P (14.6 percent), Se (15.3 percent), Tl (12.3 percent), and Zn (11.8 percent). The pre- 3.1 Sampling. A schematic of the sam- cision for Ni was 7.7 percent for another test pling train is shown in Figure 29–1. It has conducted at a source simulator. Be, Mn, and general similarities to the Method 5 train.
1019
VerDate 11
3.1.1 Probe Nozzle (Probe Tip) and alternate tips are constructed of materials Borosilicate or Quartz Glass Probe Liner. that are free from contamination and will Same as Method 5, Sections 2.1.1 and 2.1.2, not interfere with the sample. If a probe tip except that glass nozzles are required unless other than glass is used, no correction to the
1020
VerDate 11
sample test results to compensate for the ples and blanks. Glass or polyethylene bot- nozzle’s effect on the sample is allowed. tles may be used for other sample types. Probe fittings of plastic such as Teflon, poly- 3.2.3 Graduated Cylinder. Glass or equiva- propylene, etc. are recommended instead of lent. metal fittings to prevent contamination. If 3.2.4 Funnel. Glass or equivalent. one chooses to do so, a single glass piece con- 3.2.5 Labels. For identifying samples. sisting of a combined probe tip and probe 3.2.6 Polypropylene Tweezers and/or Plas- liner may be used. tic Gloves. For recovery of the filter from 3.1.2 Pitot Tube and Differential Pressure the sampling train filter holder. Gauge. Same as Method 2, Sections 2.1 and 3.3 Sample Preparation and Analysis. 2.2, respectively. 3.3.1 Volumetric Flasks, 100-ml, 250-ml, 3.1.3 Filter Holder. Glass, same as Method and 100-ml. For preparation of standards and 5, Section 2.1.5, except use a Teflon filter sample dilutions. support or other non-metallic, non-contami- 3.3.2 Graduated Cylinders. For prepara- nating support in place of the glass frit. tion of reagents. 3.1.4 Filter Heating System. Same as 3.3.3 ParrR Bombs or Microwave Pressure Method 5, Section 2.1.6. Relief Vessels with Capping Station (CEM 3.1.5 Condenser. Use the following system Corporation model or equivalent). For sam- for condensing and collecting gaseous metals ple digestion. and determining the moisture content of the 3.3.4 Beakers and Watch Glasses. 250-ml stack gas. The condensing system shall con- beakers, with watch glass covers, for sample sist of four to seven impingers connected in digestion. series with leak-free ground glass fittings or other leak-free, non-contaminating fittings. 3.3.5 Ring Stands and Clamps. For secur- Use the first impinger as a moisture trap. ing equipment such as filtration apparatus. The second impinger (which is the first 3.3.6 Filter Funnels. For holding filter paper. HNO3/H2O2 impinger) shall be identical to the first impinger in Method 5. The third im- 3.3.7 Disposable Pasteur Pipets and Bulbs. 3.3.8 Volumetric Pipets. pinger (which is the second HNO3/H2O2 im- pinger) shall be a Greenburg Smith impinger 3.3.9 Analytical Balance. Accurate to with the standard tip as described for the within .01 mg. second impinger in Method 5, Section 2.1.7. 3.3.10 Microwave or Conventional Oven. The fourth (empty) impinger and the fifth For heating samples at fixed power levels or temperatures, respectively. and sixth (both acidified KMnO4) impingers are the same as the first impinger in Method 3.3.11 Hot Plates. 5. Place a thermometer capable of measuring 3.3.12 Atomic Absorption Spectrometer to within 1°C (2°F) at the outlet of the last (AAS). Equipped with a background correc- impinger. If no Hg analysis is planned, then tor. the fourth, fifth, and sixth impingers are not 3.3.12.1 Graphite Furnace Attachment. used. With Sb, As, Cd, Co, Pb, Se, and Tl hollow 3.1.6 Metering System, Barometer, and cathode lamps (HCLs) or electrodeless dis- Gas Density Determination Equipment. charge lamps (EDLs). Same as Methods 7041 Same as Method 5, Sections 2.1.8 through (Sb), 7060 (As), 7131 (Cd), 7201 (Co), 7421 (Pb), 2.1.10, respectively. 7740 (Se), and 7841 (Tl) in EPA publication 3.1.7 Teflon Tape. For capping openings SW–846 Third Edition (November 1986) in- and sealing connections, if necessary, on the cluding updates I, II, IIA and IIB, as incor- sampling train. porated by reference in § 60.17(i). 3.2. Sample Recovery. Same as Method 5, 3.3.12.2 Cold Vapor Mercury Attachment. Sections 2.2.1 through 2.2.8 (Probe-Liner and With a mercury HCL or EDL, an air recir- Probe-Nozzle Brushes or Swabs, Wash Bot- culation pump, a quartz cell, an aerator ap- tles, Sample Storage Containers, Petri paratus, and a heat lamp or desiccator tube. Dishes, Glass Graduated Cylinder, Plastic The heat lamp shall be capable of raising the Storage Containers, Funnel and Rubber Po- temperature at the quartz cell by 10°C above liceman, and Glass Funnel), respectively, ambient, so that no condensation forms on with the following exceptions and additions: the wall of the quartz cell. Same as Method 3.2.1 Non-metallic Probe-Liner and Probe- 6020 in EPA publication SW–846 Third Edi- Nozzle Brushes or Swabs. Use non-metallic tion (November 1986) including updates I, II, probe-liner and probe-nozzle brushes or IIA and IIB, as incorporated by reference in swabs for quantitative recovery of materials § 60.17(i). See NOTE NO. 2: Section 5.4.3 for collected in the front-half of the sampling other acceptable approaches for analysis of train. Hg in which analytical detection limits of 3.2.2 Sample Storage Containers. Use 0.002 ng/ml were obtained. glass bottles (see the Precaution: in Section 3.3.13 Inductively Coupled Argon Plasma 4.3.2 of this Method) with Teflon-lined caps Spectrometer. With either a direct or se- that are non-reactive to the oxidizing solu- quential reader and an alumina torch. Same tions, with capacities of 1000- and 500-ml, for as EPA Method 6010 in EPA publication SW– storage of acidified KMnO4- containing sam- 846 Third Edition (November 1986) including 1021
VerDate 11
updates I, II, IIA and IIB, as incorporated by liter: this solution is 10 percent H2SO4 (V/V). reference in § 60.17(i). Dissolve, with stirring, 40 g of KMnO4 into 10 3.3.14 Inductively Coupled Plasma-Mass percent H2SO4 (V/V) and add 10 percent H2SO4 Spectrometer. Same as EPA Method 6020 in (V/V) with stirring to make a volume of 1 EPA publication SW–846 Third Edition (No- liter. Prepare and store in glass bottles to vember 1986) including updates I, II, IIA and prevent degradation. This reagent shall con- IIB, as incorporated by reference in § 60.17(i). tain less than 2 ng/ml of Hg. Precaution: To prevent autocatalytic decom- 4. Reagents position of the permanganate solution, filter 4.1 Unless otherwise indicated, it is in- the solution through Whatman 541 filter tended that all reagents conform to the spec- paper. Also, due to the potential reaction of ifications established by the Committee on the potassium permanganate with the acid, Analytical Reagents of the American Chemi- there could be pressure buildup in the solu- cal Society, where such specifications are tion storage bottle. Therefore these bottles available. Otherwise, use the best available shall not be fully filled and shall be vented grade. to relieve excess pressure and prevent explo- 4.2 Sampling Reagents. sion potentials. Venting is required, but not 4.2.1 Sample Filters. Without organic in a manner that will allow contamination of binders. The filters shall contain less than the solution. A No. 70–72 hole drilled in the 1.3 µg/in.2 of each of the metals to be meas- container cap and Teflon liner has been used. ured. Analytical results provided by filter 4.3.3 HNO3, 0.1 N. Add with stirring 6.3 ml manufacturers stating metals content of the of concentrated HNO3 (70 percent) to a flask filters are acceptable. However, if no such re- containing approximately 900 ml of water. sults are available, analyze filter blanks for Dilute to 1000 ml with water. Mix well. This each target metal prior to emission testing. reagent shall contain less than 2 ng/ml of Quartz fiber filters meeting these require- each target metal. ments are recommended. However, if glass 4.3.4 HCl, 8 N. Carefully add with stirring fiber filters become available which meet 690 ml of concentrated HCl to a flask con- these requirements, they may be used. Filter taining 250 ml of water. Dilute to 1000 ml efficiencies and unreactiveness to sulfur di- with water. Mix well. This reagent shall con- oxide (SO2) or sulfur trioxide (SO3) shall be tain less than 2 ng/ml of Hg. as described in Section 3.1.1 of Method 5. 4.4 Glassware Cleaning Reagents. 4.2.2 Water. To conform to ASTM Speci- 4.4.1 HNO3, Concentrated. Fisher ACS fication D1193–77, Type II (incorporated by grade or equivalent. reference—See § 60.17). If necessary, analyze 4.4.2 Water. To conform to ASTM Speci- the water for all target metals prior to field fication D1193–77, Type II (incorporated by use. All target metals should be less than 1 reference—See § 60.17). ng/ml. 4.4.3 HNO , 10 Percent (V/V). Add with 4.2.3 Nitric Acid (HNO ). Concentrated. 3 3 stirring 500 ml of concentrated HNO to a Baker Instra-analyzed or equivalent. 3 flask containing approximately 4000 ml of 4.2.4 Hydrochloric Acid (HCL). Con- water. Dilute to 5000 ml with water. Mix centrated. Baker Instra-analyzed or equiva- lent. well. This reagent shall contain less than 2 ng/ml of each target metal. 4.2.5 Hydrogen Peroxide (H2O2), 30 Percent (V/V). 4.5 Sample Digestion and Analysis Re- 4.2.6 Potassium Permanganate (KMnO ). agents. 4 The metals standards, except Hg, may also 4.2.7 Sulfuric Acid (H2SO4). Concentrated. 4.2.8 Silica Gel and Crushed Ice. Same as be made from solid chemicals as described in Method 5, Sections 3.1.2 and 3.1.4, respec- Citation 3 of the Bibliography. Refer to Cita- tively. tions 1, 2, or 5 of the Bibliography for addi- 4.3 Pretest Preparation of Sampling Re- tional information on Hg standards. The 1000 agents. µg/ml Hg stock solution standard may be 4.3.1 HNO3/H2O2 Absorbing Solution, 5 Per- made according to Section 6.2.5 of Method cent HNO3/10 Percent H2O2. Add carefully 101A. with stirring 50 ml of concentrated HNO3 to 4.5.1 HCL, Concentrated. a 1000-ml volumeric flask containing ap- 4.5.2 Hydrofluoric Acid (HF), Con- proximately 500 ml of water, and then add centrated. carefully with stirring 333 ml of 30 percent 4.5.3 HNO3, Concentrated. Baker Instra- H2O2. Dilute to volume with water. Mix well. analyzed or equivalent. This reagent shall contain less than 2 ng/ml 4.5.4 HNO3, 50 Percent (V/V). Add with of each target metal. stirring 125 ml of concentrated HNO3 to 100 4.3.2 Acidic KMnO4 Absorbing Solution, 4 ml of water. Dilute to 250 ml with water. Mix Percent KMnO4 (W/V), 10 Percent H2SO4 (V/ well. This reagent shall contain less than 2 V). Prepare fresh daily. Mix carefully, with ng/ml of each target metal. stirring, 100 ml of concentrated H2SO4 into 4.5.5 HNO3, 5 Percent (V/V). Add with stir- approximately 800 ml of water, and add ring 50 ml of concentrated HNO3 to 800 ml of water with stirring to make a volume of 1 water. Dilute to 1000 ml with water. Mix
1022
VerDate 11
well. This reagent shall contain less than 2 ml with 5 ml of 4 percent KMnO4, 5 ml of 15 ng/ml of each target metal. percent HNO3, and then water. Mix well. Use 4.5.6 Water. To conform to ASTM Speci- at least five separate aliquots of the working fication D1193–77, Type II (incorporated by Hg standard solution and a blank to prepare reference—See § 60.17). the standard curve. These aliquots and blank 4.5.7 Hydroxylamine Hydrochloride and shall contain 0.0, 1.0, 2.0, 3.0, 4.0, and 5.0 ml Sodium Chloride Solution. See Citation 2 of of the working standard solution containing the Bibliography for preparation. 0, 200, 400, 600, 800, and 1000 ng Hg, respec- 4.5.8 Stannous Chloride. See Citation 2 of tively. Prepare quality control samples by the Bibliography for preparation. making a separate 10 µg/ml standard and di- 4.5.9 KMnO4, 5 Percent (W/V). See Citation luting until in the calibration range. 2 of the Bibliography for preparation. 4.5.34 ICAP Standards and Quality Con- 4.5.10 H2SO4, Concentrated. trol Samples. Calibration standards for ICAP 4.5.11 Potassium Persulfate, 5 Percent (W/ analysis can be combined into four different V). See Citation 2 of the Bibliography for mixed standard solutions as follows: preparation. 4.5.12 Nickel Nitrate, Ni (NO3)2 6H2O. MIXED STANDARD SOLUTIONS FOR ICAP 4.5.13 Lanthanum Oxide, La2 O3. ANALYSIS 4.5.14 Hg Standard (AAS Grade), 1000 µg/ ml. Solution Elements 4.5.15 Pb Standard (AAS Grade), 1000 µg/ ml. I ...... As, Be, Cd, Mn, Pb, Se, Zn. µ II ...... Ba, Co, Cu, Fe. 4.5.16 As Standard (AAS Grade), 1000 g/ III ...... Al, Cr, Ni. ml. IV ...... Ag, P, Sb, Tl. 4.5.17 Cd Standard (AAS Grade), 1000 µg/ ml. Prepare these standards by combining and 4.5.18 Cr Standard (AAS Grade), 1000 µg/ diluting the appropriate volumes of the 1000 ml. µg/ml solutions with 5 percent HNO3. A mini- 4.5.19 Sb Standard (AAS Grade), 1000 µg/ mum of one standard and a blank can be ml. used to form each calibration curve. How- 4.5.20 Ba Standard (AAS Grade), 1000 µg/ ever, prepare a separate quality control sam- ml. ple spiked with known amounts of the target 4.5.21 Be Standard (AAS Grade), 1000 µg/ metals in quantities in the mid-range of the ml. calibration curve. Suggested standard levels 4.5.22 Co Standard (AAS Grade), 1000 µg/ are 25 µg/ml for Al, Cr and Pb, 15 µg/ml for ml. Fe, and 10 µg/ml for the remaining elements. 4.5.23 Cu Standard (AAS Grade), 1000 µg/ Prepare any standards containing less than 1 ml. µg/ml of metal on a daily basis. Standards 4.5.24 Mn Standard (AAS Grade), 1000 µg/ containing greater than 1 µg/ml of metal ml. should be stable for a minimum of 1 to 2 4.5.25 Ni Standard (AAS Grade), 1000 µg/ weeks. For ICP–MS, follow Method 6020 in ml. EPA Publication SW–846 Third Edition (No- 4.5.26 P Standard (AAS Grade), 1000 µg/ml. vember 1986) including updates I, II, IIA and 4.5.27 Se Standard (AAS Grade), 1000 µg/ IIB, as incorporated by reference in § 60.17(i). ml. 4.5.35 GFAAS Standards. Sb, As, Cd, Co, 4.5.28 Ag Standard (AAS Grade), 1000 µg/ Pb, Se, and Tl. Prepare a 10 µg/ml standard ml. by adding 1 ml of 1000 µg/ml standard to a 4.5.29 Tl Standard (AAS Grade), 1000 µg/ 100-ml volumetric flask. Dilute with stirring ml. to 100 ml with 10 percent HNO3. For GFAAS, 4.5.30 Zn Standard (AAS Grade), 1000 µg/ matrix match the standards. Prepare a 100 ml. ng/ml standard by adding 1 ml of the 10 µg/ml 4.5.31 Al Standard (AAS Grade), 1000 µg/ standard to a 100-ml volumetric flask, and ml. dilute to 100 ml with the appropriate matrix 4.5.32 Fe Standard (AAS Grade), 1000 µg/ solution. Prepare other standards by diluting ml. the 100 ng/ml standards. Use at least five 4.5.33 Hg Standards and Quality Control standards to make up the standard curve. Samples. Prepare fresh weekly a 10 µg/ml in- Suggested levels are 0, 10, 50, 75, and 100 ng/ termediate Hg standard by adding 5 ml of ml. Prepare quality control samples by mak- 1000 µg/ml Hg stock solution prepared ac- ing a separate 10 µg/ml standard and diluting cording to Method 101A to a 500-ml volu- until it is in the range of the samples. Pre- metric flask; dilute with stirring to 500 ml pare any standards containing less than 1 µg/ by first carefully adding 20 ml of 15 percent ml of metal on a daily basis. Standards con- HNO3 and then adding water to the 500-ml taining greater than 1 µg/ml of metal should volume. Mix well. Prepare a 200 ng/ml work- be stable for a minimum of 1 to 2 weeks. ing Hg standard solution fresh daily: add 5 4.5.36 Matrix Modifiers. ml of the 10 µg/ml intermediate standard to 4.5.36.1 Nickel Nitrate, 1 Percent (V/V). a 250-ml volumetric flask, and dilute to 250 Dissolve 4.956 g of Ni (NO3)2&διωιδε´6H2O or
1023
VerDate 11
other nickel compound suitable for prepara- weighed directly in the impinger just prior tion of this matrix modifier in approxi- to final train assembly. mately 50 ml of water in a 100-ml volumetric 5.1.3.2 Based on the specific source sam- flask. Dilute to 100 ml with water. pling conditions, the use of an empty first 4.5.36.2 Nickel Nitrate, 0.1 Percent (V/V). impinger can be eliminated if the moisture Dilute 10 ml of 1 percent nickel nitrate solu- to be collected in the impingers will be less tion to 100 ml with water. Inject an equal than approximately 100 ml. amount of sample and this modifier into the 5.1.3.3 If Hg analysis will not be per- graphite furnace during GFAAS analysis for formed, the fourth, fifth, and sixth impingers As. as shown in Figure 29–1 are not required. 4.5.36.3 Lanthanum. Carefully dissolve 5.1.3.4 To insure leak-free sampling train 0.5864 g of La2 O3 in 10 ml of concentrated connections and to prevent possible sample HNO3, and dilute the solution by adding it contamination problems, use Teflon tape or with stirring to approximately 50 ml of other non-contaminating material instead of water. Dilute to 100 ml with water, and mix silicone grease. well. Inject an equal amount of sample and Precaution: Exercise extreme care to pre- this modifier into the graphite furnace dur- vent contamination within the train. Pre- ing GFAAS analysis for Pb. vent the acidic KMnO4 from contacting any 4.5.37 Whatman 40 and 541 Filter Papers glassware that contains sample material to (or equivalent). For filtration of digested be analyzed for Mn. Prevent acidic H2O2 from samples. mixing with the acidic KMnO4. 5.1.4 Leak-Check Procedures. Follow the 5. Procedure leak-check procedures given in Method 5, 5.1 Sampling. The complexity of this Section 4.1.4.1 (Pretest Leak-Check), Section method is such that, to obtain reliable re- 4.1.4.2 (Leak-Checks During the Sample sults, both testers and analysts must be Run), and Section 4.1.4.3 (Post-Test Leak- trained and experienced with the test proce- Checks). dures, including source sampling; reagent 5.1.5 Sampling Train Operation. Follow preparation and handling; sample handling; the procedures given in Method 5, Section safety equipment and procedures; analytical 4.1.5. When sampling for Hg, use a procedure calculations; reporting; and the specific pro- analagous to that described in Section 7.1.1 cedural descriptions throughout this meth- of Method 101A, 40 CFR Part 61, Appendix B, od. if necessary to maintain the desired color in 5.1.1 Pretest Preparation. Follow the the last acidified permanganate impinger. same general procedure given in Method 5, For each run, record the data required on a Section 4.1.1, except that, unless particulate data sheet such as the one shown in Figure emissions are to be determined, the filter 5–2 of Method 5. need not be desiccated or weighed. First, 5.1.6 Calculation of Percent Isokinetic. rinse all sampling train glassware with hot Same as Method 5, Section 4.1.6. tap water and then wash in hot soapy water. 5.2 Sample Recovery. Next, rinse glassware three times with tap 5.2.1 Begin cleanup procedures as soon as water, followed by three additional rinses the probe is removed from the stack at the with water. Then soak all glassware in a 10 end of a sampling period. The probe should percent (V/V) nitric acid solution for a mini- be allowed to cool prior to sample recovery. mum of 4 hours, rinse three times with When it can be safely handled, wipe off all water, rinse a final time with acetone, and external particulate matter near the tip of allow to air dry. Cover all glassware open- the probe nozzle and place a rinsed, non-con- ings where contamination can occur until taminating cap over the probe nozzle to pre- the sampling train is assembled for sam- vent losing or gaining particulate matter. Do pling. not cap the probe tip tightly while the sam- 5.1.2 Preliminary Determinations. Same pling train is cooling; a vacuum can form in as Method 5, Section 4.1.2. the filter holder with the undesired result of 5.1.3 Preparation of Sampling Train. drawing liquid from the impingers onto the 5.1.3.1 Set up the sampling train as shown filter. in Figure 29–1. Follow the same general pro- 5.2.2 Before moving the sampling train to cedures given in Method 5, Section 4.1.3, ex- the cleanup site, remove the probe from the cept place 100 ml of the HNO3/H2O2 solution sampling train and cap the open outlet. Be (Section 4.3.1. of this method) in each of the careful not to lose any condensate that second and third impingers as shown in Fig- might be present. Cap the filter inlet where ure 29–1. Placee 100 ml of the acidic KMnO4 the probe was fastened. Remove the umbili- absorbing solution (Section 4.3.2 of this cal cord from the last impinger and cap the method) in each of the fifth and sixth impinger. Cap the filter holder outlet and impingers as shown in Figure 29–1, and trans- impinger inlet. Use non-contaminating caps, fer approximately 200 to 300 g of pre-weighed whether ground-glass stoppers, plastic caps, silica gel from its container to the last im- serum caps, or Teflon tape to close these pinger. Alternatively, the silica gel may be openings.
1024
VerDate 11
5.2.3 Alternatively, the following proce- 5.2.4 Transfer the probe and filter-im- dure may be used to disassemble the train pinger assembly to a cleanup area that is before the probe and filter holder/oven are clean and protected from the wind and other completely cooled: Initially disconnect the potential causes of contamination or loss of filter holder outlet/impinger inlet and loose- sample. Inspect the train before and during ly cap the open ends. Then disconnect the disassembly and note any abnormal condi- probe from the filter holder or cyclone inlet tions. Take special precautions to assure and loosely cap the open ends. Cap the probe that all the items necessary for recovery do tip and remove the umbilical cord as pre- not contaminate the samples. The sample is recovered and treated as follows (see sche- viously described. matic in Figures 29–2a and 29–2b):
1025
VerDate 11
1026
VerDate 11
5.2.5 Container No. 1 (Sample Filter). acid-washed polypropylene or Teflon coated Carefully remove the filter from the filter tweezers or clean, disposable surgical gloves holder and place it in its labeled petri dish rinsed with water and dried. If it is necessary container. To handle the filter, use either
1027
VerDate 11
to fold the filter, make certain the particu- move visible particulate. Make a final rinse late cake is inside the fold. Carefully trans- of the brush and filter holder. After all ace- fer the filter and any particulate matter or tone washings and particulate matter have filter fibers that adhere to the filter holder been collected in the sample container, gasket to the petri dish by using a dry (acid- tighten the lid so that acetone will not leak cleaned) nylon bristle brush. Do not use any out when shipped to the laboratory. Mark metal-containing materials when recovering the height of the fluid level to determine this train. Seal the labeled petri dish. whether or not leakage occurred during 5.2.6 Container No. 2. (Acetone Rinse). transport. Clearly label the container to Perform this procedure only if a determina- identify its contents. tion of particulate emissions is to be made. 5.2.7 Container No. 3 (Probe Rinse). Keep Quantitatively recover particulate matter the probe assembly clean and free from con- and any condensate from the probe nozzle, tamination during the probe rinse. Rinse the probe fitting, probe liner, and front half of probe nozzle and fitting, probe liner, and the filter holder by washing these compo- front-half of the filter holder thoroughly nents with a total of 100 ml of acetone, while with a total of 100 ml of 0.1 N HNO , and simultaneously taking great care to see that 3 place the wash into a sample storage con- no dust on the outside of the probe or other tainer. surfaces gets in the sample. The use of ex- actly 100 ml is necessary for the subsequent (NOTE: The use of a total of exactly 100 ml blank correction procedures. Distilled water is necessary for the subsequent blank correc- may be used instead of acetone when ap- tion procedures.) proved by the Administrator and shall be Perform the rinses as applicable and gen- used when specified by the Administrator; in erally as described in Method 12, Section these cases, save a water blank and follow 5.2.2. Record the volume of the rinses. Mark the Administrator’s directions on analysis. the height of the fluid level on the outside of 5.2.6.1 Carefully remove the probe nozzle, the storage container and use this mark to and clean the inside surface by rinsing with determine if leakage occurs during trans- acetone from a wash bottle while brushing port. Seal the container, and clearly label with a non-metallic brush. Brush until the the contents. Finally, rinse the nozzle, probe acetone rinse shows no visible particles, then liner, and front-half of the filter holder with make a final rinse of the inside surface with water followed by acetone, and discard these acetone. rinses. 5.2.6.2 Brush and rinse the sample exposed 5.2.8 Container No. 4 (Impingers 1 through inside parts of the probe fitting with acetone 3, Moisture Knockout Impinger, when used, in a similar way until no visible particles re- HNO3/H2O2 Impingers Contents and Rinses). main. Rinse the probe liner with acetone by Due to the potentially large quantity of liq- tilting and rotating the probe while squirt- uid involved, the tester may place the im- ing acetone into its upper end so that all in- pinger solutions from impingers 1 through 3 side surfaces will be wetted with acetone. in more than one container, if necessary. Allow the acetone to drain from the lower Measure the liquid in the first three end into the sample container. A funnel may impingers to within 0.5 ml using a graduated be used to aid in transferring liquid washings cylinder. Record the volume. This informa- to the container. Follow the acetone rinse tion is required to calculate the moisture with a non-metallic probe brush. Hold the content of the sampled flue gas. Clean each probe in an inclined position, squirt acetone of the first three impingers, the filter sup- into the upper end as the probe brush is port, the back half of the filter housing, and being pushed with a twisting action three connecting glassware by thoroughly rinsing times through the probe. Hold a sample con- with 100 ml of 0.1 N HNO3 using the proce- tainer underneath the lower end of the dure as applicable in Method 12, Section probe, and catch any acetone and particulate 5.2.4. matter which is brushed through the probe (NOTE: The use of exactly 100 ml of 0.1 N until no visible particulate matter is carried HNO rinse is necessary for the subsequent out with the acetone or until none remains 3 blank correction procedures. Combine the in the probe liner on visual inspection. Rinse rinses and impinger solutions, measure and the brush with acetone, and quantitatively record the final total volume. Mark the collect these washings in the sample con- height of the fluid level, seal the container, tainer. After the brushing, make a final ace- and clearly label the contents.) tone rinse of the probe as described above. 5.2.6.3 It is recommended that two people 5.2.9 Container Nos. 5A (0.1 N HNO3), 5B clean the probe to minimize sample losses. (KMnO4/H2SO4 absorbing solution), and 5C (8 Between sampling runs, keep brushes clean N HCl rinse and dilution). and protected from contamination. Clean the 5.2.9.1 When sampling for Hg, pour all the inside of the front-half of the filter holder by liquid from the impinger (normally impinger rubbing the surfaces with a non-metallic No. 4) that immediately preceded the two brush and rinsing with acetone. Rinse each permanganate impingers into a graduated surface three times or more if needed to re- cylinder and measure the volume to within
1028
VerDate 11
0.5 ml. This information is required to cal- since weight gained in the silica gel im- culate the moisture content of the sampled pinger is used for moisture calculations. Al- flue gas. Place the liquid in Container No. ternatively, if a balance is available in the 5A. Rinse the impinger with exactly 100 ml of field, record the weight of the spent silica gel 0.1 N HNO3 and place this rinse in Container (or silica gel plus impinger) to the nearest 0.5 No. 5A. g. 5.2.9.2 Pour all the liquid from the two per- 5.2.11 Container No. 7 (Acetone Blank). If manganate impingers into a graduated cyl- particulate emissions are to be determined, inder and measure the volume to within 0.5 at least once during each field test, place a ml. This information is required to calculate 100-ml portion of the acetone used in the the moisture content of the sampled flue gas. sample recovery process into a container la- Place this acidic KMnO4 solution into Con- beled No. 7. Seal the container. tainer No. 5B. Using a total of exactly 100 ml 5.2.12 Container No. 8A (0.1 N HNO3 of fresh acidified KMnO4 solution for all Blank). At least once during each field test, rinses (approximately 33 ml per rinse), rinse place 300 ml of the 0.1 N HNO3 solution used the two permanganate impingers and con- in the sample recovery process into a con- necting glassware a minimum of three times. tainer labeled No. 8A. Seal the container. Pour the rinses into Container No. 5B, care- 5.2.13 Container No. 8B (Water Blank). At fully assuring transfer of all loose precip- least once during each field test, place 100 ml itated materials from the two impingers. of the water used in the sample recovery Similarly, using 100 ml total of water, rinse process into a container labeled No. 8B. Seal the permanganate impingers and connecting the container. glass a minimum of three times, and pour 5.2.14 Container No. 9 (5 Percent HNO3/10 the rinses into Container 5B, carefully assur- Percent H2O2 Blank). At least once during ing transfer of any loose precipitated mate- each field test, place 200 ml of the 5 Percent rial. Mark the height of the fluid level, and HNO3/10 Percent H2O2 solution used as the ni- clearly label the contents. Read the Pre- tric acid impinger reagent into a container caution: in Section 4.3.2. NOTE: Due to the labeled No. 9. Seal the container. potential reaction of KMnO4 with acid, pres- 5.2.15 Container No. 10 (Acidified KMnO4 sure buildup can occur in the sample storage Blank). At least once during each field test, bottles. Do not fill these bottles completely place 100 ml of the acidified KMnO4 solution and take precautions to relieve excess pres- used as the impinger solution and in the sure. A No. 70–72 hole drilled in the container sample recovery process into a container la- cap and Teflon liner has been used success- beled No. 10. Prepare the container as de- fully. scribed in Section 5.2.9.2. Read the Pre- 5.2.9.3 If no visible deposits remain after caution: in Section 4.3.2. and read the NOTE in the water rinse, no further rinse is nec- Section 5.2.9.2. essary. However, if deposits remain on the 5.2.16 Container No. 11 (8 N HCl Blank). At impinger surfaces, wash them with 25 ml of least once during each field test, place 200 ml 8 N HCl, and place the wash in a separate of water into a sample container labeled No. sample container labeled No. 5C containing 11. Then carefully add with stirring 25 ml of 200 ml of water. First, place 200 ml of water 8 N HCl. Mix well and seal the container. in the container. Then wash the impinger 5.2.17 Container No. 12 (Sample Filter walls and stem with the HCl by turning the Blank). Once during each field test, place impinger on its side and rotating it so that into a petri dish labeled No. 12 three unused the HC1 contacts all inside surfaces. Use a blank filters from the same lot as the sam- total of only 25 ml of 8 N HCl for rinsing both pling filters. Seal the petri dish. permanganate impingers combined. Rinse 5.3 Sample Preparation. Note the level of the first impinger, then pour the actual rinse the liquid in each of the containers and de- used for the first impinger into the second termine if any sample was lost during ship- impinger for its rinse. Finally, pour the 25 ment. If a noticeable amount of leakage has ml of 8 N HCl rinse carefully into the con- occurred, either void the sample or use tainer. Mark the height of the fluid level on methods, subject to the approval of the Ad- the outside of the container to determine if ministrator, to correct the final results. A leakage occurs during transport. diagram illustrating sample preparation and 5.2.10 Container No. 6 (Silica Gel). Note analysis procedures for each of the sample the color of the indicating silica gel to deter- train components is shown in Figure 29–3. mine whether it has been completely spent 5.3.1 Container No. 1 (Sample Filter). and make a notation of its condition. Trans- 5.3.1.1 If particulate emissions are being fer the silica gel from its impinger to its determined, first desiccate the filter and fil- original container and seal it. The tester ter catch without added heat (do not heat may use a funnel to pour the silica gel and the filters to speed the drying) and weigh to a rubber policeman to remove the silica gel a constant weight as described in Section 4.3 from the impinger. The small amount of par- of Method 5. ticles that might adhere to the impinger 5.3.1.2 Following this procedure, or ini- wall need not be removed. Do not use water tially, if particulate emissions are not being or other liquids to transfer the silica gel determined in addition to metals analysis,
1029
VerDate 11
divide the filter with its filter catch into 5.3.1.3 If the sampling train includes an portions containing approximately 0.5 g optional glass cyclone in front of the filter, each. Place the pieces in the analyst’s choice prepare and digest the cyclone catch by the of either individual microwave pressure re- procedures described in section 5.3.1.2 and lief vessels or ParrR Bombs. Add 6 ml of con- then combine the digestate with the digested filter sample. centrated HNO3 and 4 ml of concentrated HF to each vessel. For microwave heating, 5.3.2 Container No. 2 (Acetone Rinse). microwave the samples for approximately 12 Note the level of liquid in the container and to 15 minutes total heating time as follows: confirm on the analysis sheet whether or not heat for 2 to 3 minutes, then turn off the leakage occurred during transport. If a no- ticeable amount of leakage has occurred, ei- microwave for 2 to 3 minutes, then heat for ther void the sample or use methods, subject 2 to 3 minutes, etc., continue this alter- to the approval of the Administrator, to cor- nation until the 12 to 15 minutes total heat- rect the final results. Measure the liquid in ing time are completed (this procedure this container either volumetrically within 1 should comprise approximately 24 to 30 min- ml or gravimetrically within 0.5 g. Transfer utes at 600 watts). Microwave heating times the contents to an acid-cleaned, tared 250-ml are approximate and are dependent upon the beaker and evaporate to dryness at ambient number of samples being digested simulta- temperature and pressure. If particulate neously. Sufficient heating is evidenced by emissions are being determined, desiccate sorbent reflux within the vessel. For conven- for 24 hours without added heat, weigh to a tional heating, heat the ParrR Bombs at 140 constant weight according to the procedures °C (285 °F) for 6 hours. Then cool the samples described in Section 4.3 of Method 5, and re- to room temperature, and combine with the port the results to the nearest 0.1 mg. Redis- acid digested probe rinse as required in Sec- solve the residue with 10 ml of concentrated tion 5.3.3. HNO3.
1030
VerDate 11
Quantitatively combine the resultant sam- 5.3.3 Container No. 3 (Probe Rinse). Verify ple, including all liquid and any particulate that the pH of this sample is 2 or lower. If it matter, with Container No. 3 before begin- is not, acidify the sample by careful addition
ning Section 5.3.3. with stirring of concentrated HNO3 to pH 2.
1031
VerDate 11
Use water to rinse the sample into a beaker, Measure and record the volume to within 0.1 and cover the beaker with a ribbed watch ml. glass. Reduce the sample volume to approxi- 5.3.4.2 Microwave Digestion Procedure. mately 20 ml by heating on a hot plate at a Add 10 ml of 50 percent HNO3 and heat for 6 temperature just below boiling. Digest the minutes total heating time in alternations of sample in microwave vessels or ParrR Bombs 1 to 2 minutes at 600 Watts followed by 1 to by quantitatively transferring the sample to 2 minutes with no power, etc., similar to the the vessel or bomb, carefully adding the 6 ml procedure described in Section 5.3.1. Allow of concentrated HNO3, 4 ml of concentrated the sample to cool. Add 10 ml of 3 percent HF, and then continuing to follow the proce- H2O2 and heat for 2 more minutes. Add 50 ml dures described in Section 5.3.1.2. Then com- of hot water, and heat for an additional 5 bine the resultant sample directly with the minutes. Cool, filter the sample, and dilute acid digested portions of the filter prepared to 150 ml (or the appropriate volume for the previously in Section 5.3.1.2. The resultant expected metals concentrations) with water. combined sample is referred to as ‘‘Sample This dilution produces Analytical Fraction Fraction 1’’. Filter the combined sample 2A. Measure and record the volume to within using Whatman 541 filter paper. Dilute to 300 0.1 ml. ml (or the appropriate volume for the ex- (NOTE: All microwave heating times given pected metals concentration) with water. are approximate and are dependent upon the This diluted sample is ‘‘Analytical Fraction number of samples being digested at a time. 1’’. Measure and record the volume of Ana- Heating times as given above have been lytical Fraction 1 to within 0.1 ml. Quan- found acceptable for simultaneous digestion titatively remove a 50-ml aliquot and label of up to 12 individual samples. Sufficient as ‘‘Analytical Fraction 1B’’. Label the re- heating is evidenced by solvent reflux within maining 250-ml portion as ‘‘Analytical Frac- the vessel.) tion 1A’’. Analytical Fraction 1A is used for 5.3.5 Container No. 5A (Impinger 4), Con- ICAP or AAS analysis for all desired metals tainer Nos. 5B and 5C (Impingers 5 and 6). except Hg. Analytical Fraction 1B is used for Keep the samples in Containers Nos. 5A, 5B, the determination of front-half Hg. and 5C separate from each other. Measure 5.3.4 Container No. 4 (Impingers 1–3). and record the volume of 5A to within 0.5 ml. Measure and record the total volume of this Label the contents of Container No. 5A to be sample to within 0.5 ml and label it ‘‘Sample Analytical Fraction 3A. To remove any Fraction 2’’. Remove a 75- to 100-ml aliquot brown MnO2 precipitate from the contents of for Hg analysis and label the aliquot ‘‘Ana- Container No. 5B, filter its contents through lytical Fraction 2B’’. Label the remaining por- Whatman 40 filter paper into a 500 ml volu- tion of Container No. 4 as ‘‘Sample Fraction metric flask and dilute to volume with 2A’’. Sample Fraction 2A defines the volume water. Save the filter for digestion of the of Analytical Fraction 2A prior to digestion. brown MnO2 precipitate. Label the 500 ml fil- All of Sample Fraction 2A is digested to trate from Container No. 5B to be Analytical produce ‘‘Analytical Fraction 2A’’. Analytical Fraction 3B. Analyze Analytical Fraction 3B Fraction 2A defines the volume of Sample for Hg within 48 hours of the filtration step. Fraction 2A after its digestion and the vol- Place the saved filter, which was used to re- ume of Analytical Fraction 2A is normally move the brown MnO2 precipitate, into an 150 ml. Analytical Fraction 2A is analyzed appropriately sized vented container, which for all metals except Hg. Verify that the pH will allow release of any gases including of Sample Fraction 2A is 2 or lower. If nec- chlorine formed when the filter is digested. essary, use concentrated HNO3 by careful ad- In a laboratory hood which will remove any dition and stirring to lower Sample Fraction gas produced by the digestion of the MnO2, 2A to pH 2. Use water to rinse Sample Frac- add 25 ml of 8 N HCl to the filter and allow tion 2A into a beaker and then cover the to digest for a minimum of 24 hours at room beaker with a ribbed watch glass. Reduce temperature. Filter the contents of Con- Sample Fraction 2A to approximately 20 ml tainer No. 5C through a Whatman 40 filter by heating on a hot plate at a temperature into a 500-ml volumetric flask. Then filter just below boiling. Then follow either of the the result of the digestion of the brown MnO2 digestion procedures described in Sections from Container No. 5B through a Whatman 5.3.4.1 or 5.3.4.2. 40 filter into the same 500-ml volumetric 5.3.4.1 Conventional Digestion Procedure. flask, and dilute and mix well to volume Add 30 ml of 50 percent HNO3, and heat for 30 with water. Discard the Whatman 40 filter. minutes on a hot plate to just below boiling. Mark this combined 500-ml dilute HCl solu- Add 10 ml of 3 percent H2O2 and heat for 10 tion as Analytical Fraction 3C. more minutes. Add 50 ml of hot water, and 5.3.6 Container No. 6 (Silica Gel). Weigh heat the sample for an additional 20 minutes. the spent silica gel (or silica gel plus im- Cool, filter the sample, and dilute to 150 ml pinger) to the nearest 0.5 g using a balance. (or the appropriate volume for the expected 5.4 Sample Analysis. For each sampling metals concentrations) with water. This di- train sample run, seven individual analytical lution produces Analytical Fraction 2A. samples are generated; two for all desired
1032
VerDate 11
metals except Hg, and five for Hg. A sche- Wave- matic identifying each sample container and Element length the prescribed analytical preparation and (nm) analysis scheme is shown in Figure 29–3. The Cadmium ...... 226.502 first two analytical samples, labeled Analyt- Chromium ...... 267.716 ical Fractions 1A and 1B, consist of the di- Cobalt ...... 228.616 gested samples from the front-half of the Copper ...... 324.754 train. Analytical Fraction 1A is for ICAP, Iron ...... 259.940 ICP–MS or AAS analysis as described in Sec- Lead ...... 220.353 tions 5.4.1 and 5.4.2, respectively. Analytical Manganese ...... 257.610 Fraction 1B is for front-half Hg analysis as Nickel ...... 231.604 described in Section 5.4.3. The contents of Phosphorous ...... 214.914 the back-half of the train are used to prepare Selenium ...... 196.026 the third through seventh analytical sam- Silver ...... 328.068 Thallium ...... 190.864 ples. The third and fourth analytical sam- Zinc ...... 213.856 ples, labeled Analytical Fractions 2A and 2B, contain the samples from the moisture re- These wavelengths represent the best com- moval impinger No. 1, if used, and HNO3 H2O2 bination of specificity and potential detec- impingers Nos. 2 and 3. Analytical Fraction tion limit. Other wavelengths may be sub- 2A is for ICAP, ICP–MS or AAS analysis for stituted if they can provide the needed speci- target metals, except Hg. Analytical Frac- ficity and detection limit, and are treated tion 2B is for analysis for Hg. The fifth with the same corrective techniques for through seventh analytical samples, labeled spectral interference. Initially, analyze all Analytical Fractions 3A, 3B, and 3C, consist samples for the target metals (except Hg) of the impinger contents and rinses from the plus Fe and Al. If Fe and Al are present, the empty impinger No. 4 and the H SO /KMnO 2 4 4 sample might have to be diluted so that each Impingers Nos. 5 and 6. These analytical of these elements is at a concentration of samples are for analysis for Hg as described less than 50 ppm so as to reduce their spec- in Section 5.4.3. The total back-half Hg catch is determined from the sum of Analytical tral interferences on As, Cd, Cr, and Pb. Per- Fractions 2B, 3A, 3B, and 3C. Analytical form ICP–MS analysis by following Method Fractions 1A and 2A can be combined propor- 6020 in EPA Publication SW–846 Third Edi- tionally prior to analysis. tion (November 1986) including updates I, II, 5.4.1 ICAP and ICP–MS Analysis. Analyze IIA, and IIB, as incorporated by reference in Analytical Fractions 1A and 2A by ICAP § 60.17(i). using Method 6010 or Method 200.7 (40 CFR (NOTE: When analyzing samples in a HF part 136, appendix C). Calibrate the ICAP, matrix, an alumina torch should be used; and set up an analysis program as described since all front-half samples will contain HF, in Method 6010 or Method 200.7. Follow the use an alumina torch.) quality control procedures described in Sec- 5.4.2. AAS by Direct Aspiration and/or tion 7.3.1. Recommended wavelengths for GFAAS. If analysis of metals in Analytical analysis are as follows: Fractions 1A and 2A by using GFAAS or di- rect aspiration AAS is needed, use Table 29– Wave- Element length 2 to determine which techniques and proce- (nm) dures to apply for each target metal. Use Table 29–2, if necessary, to determine tech- Aluminum ...... 308.215 niques for minimization of interferences. Antimony ...... 206.833 Arsenic ...... 193.696 Calibrate the instrument according to Sec- Barium ...... 455.403 tion 6.3 and follow the quality control proce- Beryllium ...... 313.042 dures specified in Section 7.3.2.
TABLE 29±2.ÐAPPLICABLE TECHNIQUES, METHODS AND MINIMIZATION OF INTERFERENCE FOR AAS ANALYSIS
Interferences 1 Wavelength Metal Technique SW±846 method No. (nm) Cause Minimization
Fe ...... Aspiration ...... 7380 248.3 Contamination ...... Great care taken to avoid contamination. Pb ...... Aspiration ...... 7420 283.3 217.0 nm alternate ...... Background correction re- quired. Pb ...... Furnace ...... 7421 283.3 Poor recoveries ...... Matrix modifier, add 10 ul of phosphorus acid to 1 ml of prepared sample in sampler cup. Mn ...... Aspiration ...... 7460 279.5 403.1 nm alternate ...... Background correction re- quired.
1033
VerDate 11
TABLE 29±2.ÐAPPLICABLE TECHNIQUES, METHODS AND MINIMIZATION OF INTERFERENCE FOR AAS ANALYSISÐContinued
1 Interferences Metal Technique SW±846 Wavelength method No. (nm) Cause Minimization
Ni ...... Aspiration ...... 7520 232.0 352.4 nm alternate Fe, Background correction re- Co, and Cr. quired. Matrix matching or nitrous- oxide/acetylene flame. Nonlinear response ...... sample dilution or use 352.3 nm line. Se ...... Furnace ...... 7740 196.0 Volatility ...... Spike samples and ref- erence materials and add nickel nitrate to minimize volatilization. Adsorption & scatter ...... Background correction is required and Zeeman background correction can be useful. Ag ...... Aspiration ...... 7760 328.1 Adsorption & Scatter AgCl Background correction is insoluble. required. Avoid Hydro- chloric acid unless silver is in solution as a chlo- ride complex Sample and standards mon- itored for aspiration rate. Tl ...... Aspiration ...... 7840 276.8 ...... Background correction is required. Hydrochloric acid should not be used. Tl ...... Furnace ...... 7841 276.8 Hydrochloric acid or chlo- Background correction is ride. required. Verify that losses are not occurring for volatization by spiked samples or standard addition; Palla- dium is a suitable matrix modifier. Zn ...... Aspiration ...... 7950 213.9 High Si, Cu, & P Contami- Strontium removes Cu nation. and phosphate, Great care taken to avoid con- tamination. Sb ...... Aspiration ...... 7040 217.6 1000 mg/ml Pb Ni, Cu, or Use secondary wave- acid. lengths of 231.1.nm; match sample & stand- ards acid concentration or use nitrous oxidefacetylene flame. Sb ...... Furnace ...... 7041 217.6 High Pb ...... Secondary Wavelength or Zeeman correction. As ...... Furnace ...... 7060 193.7 Arsenic volatilization ...... Spiked samples and add Aluminum ...... nickel nitrate solution to digestates prior to anal- ysis. Use Zeeman background correction. Ba ...... Aspiration 7080 ...... 7080 553.6 Calcium ...... High hollow cathode cur- Barium ionization ...... rent and narrow band set. 2 ml of KCl per 100 ml of sample. Be ...... Aspiration ...... 7090 234.9 500 ppm Al High Mg and Add 0.1% fluoride. Si. Use method of standard additions. Be ...... Furnace ...... 7091 234.9 Be in optical path ...... Optimize parameters to minimize effects. Cd ...... Aspiration ...... 7130 228.8 Absorption and light scat- Background correction is tering. required. Cd ...... Furnace ...... 7131 228.8 As above ...... As above. Excess Chloride ...... Ammonium phosphate Pipet tips ...... used as a matrix modi- fier. Use cadmiun-free tips.
1034
VerDate 11
TABLE 29±2.ÐAPPLICABLE TECHNIQUES, METHODS AND MINIMIZATION OF INTERFERENCE FOR AAS ANALYSISÐContinued
1 Interferences Metal Technique SW±846 Wavelength method No. (nm) Cause Minimization
Cr ...... Aspiration ...... 7190 357.9 Akali metal ...... KCl ionization suppressant in samples and stand- ardsÐConsult mfgs lit- erature. Co ...... Furnace ...... 7201 240.7 Excess chloride ...... Use Method of Standard Additions. Cr ...... Furnace ...... 7191 357.9 200 mg/L Ca and P ...... All calcium nitrate for a known constant effect and to eliminate effect of phosphate. Cu ...... Aspiration ...... 7210 324.7 Absorption & scatter ...... Consult manufacturer's manual. 1 Refer to EPA publication SW±846 Third Edition (November 1986) including updates I, II, IIA, and IIB, as incorporated by ref- erence in § 60.17(i).
5.4.3 CVAAS Hg analysis. Analyze Analyt- NOTE NO. 1 TO SECTION 5.4.3. When Hg levels ical Fractions 1B, 2B, 3A, 3B, and 3C sepa- in the sample fractions are below the in- rately for Hg using CVAAS following the stack detection limit given in Table 29–1, se- method outlined in Method 7470 in EPA Pub- lect a 10 ml aliquot for digestion and analy- lication SW–846 Third Edition (November sis as described. 1986) including updates I, II, IIA and IIB, as NOTE NO. 2 TO SECTION 5.4.3. Optionally, Hg incorporated by reference in § 60.17(i) or in can be analyzed by using the CVAAS analyt- Standard Methods for the Examination of Water ical procedures given by some instrument and Wastewater, 16th Edition, (1985), Method manufacturer’s directions. These include 303F, as incorporated by reference in § 60.17, calibration and quality control procedures or, optionally using NOTE No. 2 in this sec- for the Leeman Model PS200, the Perkin tion. Set up the calibration curve (zero to Elmer FIAS systems, and similar models, if 1000 ng) as described in Method 7470 or simi- available, of other instrument manufactur- lar to Method 303F using 300-ml BOD bottles ers. For digestion and analyses by these in- instead of Erlenmeyers. Perform the follow- struments, perform the following two steps: ing for each Hg analysis. From each original (1) Digest the sample aliquot through the sample, select and record an aliquot in the addition of the aqueous hydroxylamine hy- size range from 1 ml to 10 ml. If no prior drochloride/sodium chloride solution the knowledge of the expected amount of Hg in same as described in this Section 5.4.3.: (The the sample exists, a 5 ml aliquot is suggested Leeman, Perkin Elmer, and similar instruments for the first dilution to 100 ml (see NOTE No. described in this note add automatically the 1 in this Section). The total amount of Hg in necessary stannous chloride solution during the µ the aliquot shall be less than 1 g and with- automated analysis of Hg.) and in the range (zero to 1000 ng) of the calibra- (2) Upon completion of the digestion de- tion curve. Place the sample aliquot into a scribed in paragraph (1), of this note, analyze separate 300-ml BOD bottle, and add enough the sample according to the instrument water to make a total volume of 100 ml. Next manufacturer’s directions. This approach al- add to it sequentially the sample digestion lows multiple (including duplicate) auto- solutions and perform the sample prepara- mated analyses of a digested sample aliquot. tion described in the procedures of Method 7470 or Method 303F. (See NOTE No. 2 in this 6. Calibration Section). If the maximum readings are off- scale (because Hg in the aliquot exceeded the Maintain a laboratory log of all calibra- calibration range; including the situation tions. where only a 1-ml aliquot of the original 6.1 Sampling Train Calibration. Calibrate sample was digested), then dilute the origi- the sampling train components according to nal sample (or a portion of it) with 0.15 per- the indicated sections of Method 5: Probe cent HNO3 (1.5 ml concentrated HNO3 per Nozzle (Section 5.1); Pitot Tube (Section 5.2); liter aqueous solution) so that when a 1- to Metering System (Section 5.3); Probe Heater 10-ml aliquot of the ‘‘0.15 HNO3 percent dilu- (Section 5.4); Temperature Gauges (Section tion of the original sample’’ is digested and 5.5); Leake-Check of the Metering System analyzed by the procedures described above, (Section 5.6); and Barometer (Section 5.7). it will yield an analysis within the range of 6.2 Industively Coupled Argon Plasma the calibration curve. Spectrometer Calibration. Prepare standards
1035
VerDate 11
as outlined in Section 4.5. Profile and cali- for the similar sample filter. Filter the brate the instrument according to the manu- digestate and the contents of Container No. facturer’s recommended procedures using 11 through Whatman 40 paper into a 500-ml those standards. Check the calibration once volumetric flask, and dilute to volume with per hour. If the instrument does not repro- water. These steps produce a blank for Ana- duce the standard concentrations within 10 lytical Fraction 3C. percent, perform the complete calibration 7.1.6 Analyze the blanks for Analytical procedures. Perform ICP–MS analysis by fol- Fraction Blanks 1A and 2A per Section 5.4.1 lowing Method 6020 in EPA Publication SW– and/or Section 5.4.2. Analyze the blanks for 846 Third Edition (November 1986) including Analytical Fractions 1B, 2B, 3A, 3B, and 3C updates I, II, IIA and IIB, as incorporated by per Section 5.4.3. Analysis of the blank for reference in § 60.17(i). Analytical Fraction 1A produces the front- 6.3 Atomic Absorption Spectrometer—Di- half reagent blank correction values for the rect Aspiration AAS, GFAAS, and CVAAS desired metals except for Hg; Analysis of the analyses. Prepare the standards as outlined blank for Analytical Fraction 1B produces in Section 4.5 and use them to calibrate the the front-half reagent blank correction value spectrometer. Calibration procedures are for Hg. Analysis of the blank for Analytical also outlined in the EPA methods referred to Fraction 2A produces the back-half reagent in Table 29–2 and in Method 7470 in EPA Pub- blank correction values for all of the desired lication SW–846 Third Edition (November metals except for Hg, while separate analy- 1986) including updates I, II, IIA and IIB, as ses of the blanks for Analytical Fractions incorporated by reference in § 60.17(i) or in 2B, 3A, 3B, and 3C produce the back-half rea- Standard Methods for the Examination of Water gent blank correction value for Hg. and Wastewater, 16th Edition, (1985), Method 7.2 Quality Control Samples. Analyze the 303F (for Hg) as incorporated by reference in following quality control samples. § 60.17. Run each standard curve in duplicate 7.2.1 ICAP and ICP–MS Analysis. Follow and use the mean values to calculate the the respective quality control descriptions in calibration line. Recalibrate the instrument Section 8 of Methods 6010 and 6020 of EPA approximately once every 10 to 12 samples. Publication SW–846 Third Edition (November 1986) including updates I, II, IIA and IIB, as 7. Quality Control incorporated by reference in § 60.17(i). For 7.1 Field Reagent Blanks, if analyzed. the purposes of a source test that consists of Perform the digestion and analysis of the three sample runs, modify those require- blanks in Container Nos. 7 through 12 that ments to include the following: two instru- were produced in Sections 5.2.11 through ment check standard runs, two calibration 5.2.17, respectively. For Hg field reagent blank runs, one interference check sample at blanks, use a 10 ml aliquot for digestion and the beginning of the analysis (analyze by analysis. Method of Standard Additions unless within 7.1.1 Digest and analyze one of the filters 25 percent), one quality control sample to from Container No. 12 per Section 5.3.1, 100 check the accuracy of the calibration stand- ml from Container No. 7 per Section 5.3.2, ards (required to be within 25 percent of cali- and 100 ml from Container No. 8A per Sec- bration), and one duplicate analysis (re- tion 5.3.3. This step produces blanks for Ana- quired to be within 20 percent of average or lytical Fractions 1A and 1B. repeat all analyses). 7.1.2 Combine 100 ml of Container No. 8A 7.2.2. Direct Aspiration AAS and/or with 200 ml from Container No. 9, and digest GFAAS Analysis for Sb, As, Ba, Be, Cd, Cu, and analyze the resultant volume per Sec- Cr, Co, Pb, Ni, Mn, Hg, P, Se, Ag, Tl, and Zn. tion 5.3.4. This step produces blanks for Ana- Analyze all samples in duplicate. Perform a lytical Fractions 2A and 2B. matrix spike on at least one front-half sam- 7.1.3 Digest and analyze a 100-ml portion ple and one back-half sample, or one com- of Container No. 8A to produce a blank for bined sample. If recoveries of less than 75 Analytical Fraction 3A. percent or greater than 125 percent are ob- 7.1.4 Combine 100 ml from Container No. tained for the matrix spike, analyze each 10 with 33 ml from Container No. 8B to sample by the Method of Standard Additions. produce a blank for Analytical Fraction 3B. Analyze a quality control sample to check Filter the resultant 133 ml as described for the accuracy of the calibration standards. If Container No. 5B in Section 5.3.5, except do the results are not within 20 percent, repeat not dilute the 133ml. Analyze this blank for the calibration. Hg within 48 hrs. of the filtration step, and 7.2.3 CVAAS Analysis for Hg. Analyze all use 400 ml as the blank volume when cal- samples in duplicate. Analyze a quality con- culating the blank mass value. Use the ac- trol sample to check the accuracy of the tual volumes of the other analytical blanks calibration standards (if not within 15 per- when calculating their mass values. cent, repeat calibration). Perform a matrix 7.1.5 Digest the filter that was used to re- spike on one sample (if not within 25 percent, move any brown MnO2 precipitate from the analyze all samples by the Method of Stand- blank for Analytical Fraction 3B by the ard Additions). Additional information on same procedure as described in Section 5.3.5 quality control can be obtained from Method
1036
VerDate 11
7470 of EPA Publication SW–846 Third Edi- Va=Total volume of digested sample solution tion (November 1986) including updates I, II, (Analytical Fraction 2A), ml (see Section IIA and IIB, as incorporated by reference in 5.3.4.1 or 5.3.4.2, as applicable). § 60.17(i) or in Standard Methods for the Exam- 8.4.3 Total Train, Metals (except Hg). Cal- ination of Water and Wastewater, 16th Edition, culate the total amount of each of the quan- (1985), Method 303F as incorporated by ref- tified metals collected in the sampling train erence in § 60.17. as follows: M =(M M ) + (M M ) Eq. 29–3 8. Calculations t fh ¥ fhb bh ¥ bhb where: 8.1 Dry Gas Volume. Using the data from Mt=Total mass of each metal (separately this test, calculate Vm(std), the dry gas sample stated for each metal) collected in the volume at standard conditions as outlined in sampling train, µg. Section 6.3 of Method 5. Mfhb=Blank correction value for mass of 8.2 Volume of Water Vapor and Moisture metal detected in front-half field reagent Content. Using the total volume of conden- blank, µg. sate collected during the source sampling, Mbhb=Blank correction value for mass of calculate the volume of water vapor Vw(std) metal detected in back-half field reagent and the moisture content B of the stack ws blank, µg. gas. Use Equations 5–2 and 5–3 of Method 5. 8.4.3.1 If the measured blank value for the 8.3 Stack Gas Velocity. Using the data front half (M ) is in the range 0.0 to ‘‘A’’ µg from this test and Equation 2–9 of Method 2, fhb [where ‘‘A’’ µg equals the value determined calculate the average stack gas velocity. by multiplying 1.4 µg/in.2 times the actual 8.4 Metals (Except Hg) in Source Sample. area in in.2 of the sample filter], use M to 8.4.1 Analytical Fraction 1A, Front-Half, fhb correct the emission sample value (M ); if Metals (except Hg). Calculate separately the fh Mfhb exceeds ‘‘A’’ µg, use the greater of I or amount of each metal collected in Sample II: Fraction 1 of the sampling train using the I. ‘‘A’’ µg. following equation: II. the lesser of (a) Mfhb, or (b) 5 percent of M =C F V Eq. 29–1 fh a1 d soln,1 Mfh. where: If the measured blank value for the black- Mfh=Total mass of each metal (except Hg) half (Mbhb) is in the range 0.0 to 1 µg, use Mbhb collected in the front half of the sam- to correct the emission sample value (Mbh); if pling train (Sample Fraction 1), µg. Mbhb) exceeds 1 µg, use the greater of I or II: Ca1=Concentration of metal in Analytical I. 1 µg. Fraction 1A as read from the standard II. the lesser of (a) Mbhb or (b) 5 percent of µ curve, g/ml. Mbh. Fd=Dilution factor (Fd = the inverse of the 8.5 Hg in Source Sample. fractional portion of the concentrated 8.5.1 Analytical Fraction 1B; Front-Half sample in the solution actually used in Hg. Calculate the amount of Hg collected in the instrument to produce the reading the front-half, Sample Fraction 1, of the Ca1. For example, if a 2 ml aliquot of Ana- sampling train by using Equation 29–4: lytical Fraction 1A is diluted to 10 ml to place it in the calibration range, Fd = 5). Qfh Vsoln,1=Total volume of digested sample solu- Hg = ()V Eq. 29− 4 tion (Analytical Fraction 1), ml. fh soln,1 8.4.1.1 If Analytical Fractions 1A and 2A Vf1 B are combined, use proportional aliquots. where: Then make appropriate changes in Equations Hgfh=Total mass of Hg collected in the front- 29–1 through 29–3 to reflect this approach. half of the sampling train (Sample Frac- 8.4.2 Analytical Fraction 2A, Back-Half, tion 1), µg. Metals (except Hg). Calculate separately the Qfh=Quantity of Hg, µg, TOTAL in the ALI- amount of each metal collected in Fraction QUOT of Analytical Fraction 1B selected 2 of the sampling train using the following for digestion and analysis. equation. 8.5.1.1 For example, if a 10 ml aliquot of
Mbh=Ca2 Fa Va Eq. 29–2 Analytical Fraction 1B is taken and digested where: and analyzed (according to Section 5.4.3 and its NOTES Nos. 1 and 2), then calculate and Mbh=Total mass of each metal (except Hg) collected in the back-half of the sam- use the total amount of Hg in the 10 ml ali- pling train (Sample Fraction 2), µg. quot for Qfh. Ca2=Concentration of metal in Analytical Vsoln,1=Total volume of Analytical Fraction 1, Fraction 2A as read from the standard ml. curve, (µg/ml). Vf1B=Volume of aliquot of Analytical Frac- Fa=Aliquot factor, volume of Sample Frac- tion 1B analyzed, ml. tion 2 divided by volume of Sample Frac- 8.5.1.2 For example, if a 1 ml aliquot of tion 2A (see Section 5.3.4.) Analytical Fraction 1B was diluted to 50 ml
1037
VerDate 11
with 0.15 percent HNO3 as described in Sec- 8.5.2.1.1 For example, if a 10 ml aliquot of tion 5.4.3 to bring it into the proper analyt- Analytical Fraction 2B is taken and digested ical range, and then 1 ml of that 50-ml wa di- and analyzed (according to Section 5.4.3 and gested according to Section 5.4.3 and ana- its NOTES Nos. 1 and 2), then calculate and lyzed, Vf1B would be 0.02 ml. use the total amount of Hg in the 10 ml ali- 8.5.2 Analytical Fractions 2B, 3A, 3B, and quot for Q . 3C; Back Half Hg. bh2 8.5.2.1 Calculate the amount of Hg col- Vsoln,2=Total volume of Sample Fraction 2, lected in Sample Fraction 2 by using Equa- ml. tion 29–5: Vf2B=Volume of Analytical Fraction 2B ana- lyzed, ml. Q 8.5.2.1.2 For example, if 1 ml of Analytical = bh2 − Hgbh2 ()Vsoln,2 Eq. 29 5 Fraction 2B was diluted to 10 ml with 0.15 Vf2 B percent HNO3 as described in Section 5.4.3 to where: bring it into the proper analytical range, and then 5 ml of that 10-ml was analyzed, Vf2B Hgbh2=Total mass of Hg collected in Sample Fraction 2, µg. would be 0.5 ml. Qbh2=Quantity of Hg, µg, TOTAL in the ALI- 8.5.2.2 Calculate each of the back-half Hg QUOT of Analytical Fraction 2B selected values for Analytical Fractions 3A, 3B, and for digestion and analysis. 3C by using Equation 29–6:
Q = bh3() A, B,C − Hgbh3() A, B,C ()Vsoln,3() A , B,C Eq. 29 6 Vf3() A, B,C
where: Hgbhb=Blank correction value for mass of Hg Hgbh3(A,B,C)=Total mass of Hg collected sepa- detected in back-half field reagent rately in Fraction 3A, 3B, or 3C, µg. blanks, µg. Qbh3(A,B,C)=Quantity of Hg, µg, TOTAL, sepa- 8.5.4 If the total of the measured blank rately, in the ALIQUOT of Analytical Frac- values (Hgfhb+Hgbhb) is in the range of 0.0 to tion 3A, 3B, and 3C selected for digestion 0.6 µg, then use the total to correct the sam- and analysis, (see previous notes in Sec- ple value (Hgfh+Hgbh); if it exceeds 0.6 µg, use tions 8.5.1 and 8.5.2 describing the quan- the greater of I. or II: tity ‘‘Q’’ and calculate similarly). I. 0.6 µg. V ( )=Volume, separately, of Analytical f3 A,B,C II. the lesser of (a) (Hgfhb+Hgbhb), or (b) 5 Fraction 3A, 3B, or 3C analyzed, ml (see percent of the sample value (Hgfh+Hgbh). previous notes in Sections 8.5.1 and 8.5.2, 8.6 Individual Metal Concentrations in describing the quantity ‘‘V’’ and cal- Stack Gas. Calculate the concentration of culate similarly). each metal in the stack gas (dry basis, ad- Vsoln,3(A,B,C)=Total volume, separately, of Ana- justed to standard conditions) by using lytical Fraction 3A, 3B, or 3C, ml. Equation 29–9: 8.5.2.3 Calculate the total amount of Hg collected in the back-half of the sampling KM train by using Equation 29–7: =4 t − Cs Eq. 29 9 Hgbh=Hgbh2+Hgbh3A+Hgbh3B+Hgbh3C Eq. 29–7 Vm() std where: Hgbh=Total mass of Hg collected in the back- Cs=Concentration of a metal in the stack half of the sampling train, µg. gas, mg/dscm. ¥3 8.5.3 Total Train Hg Catch. Calculate the K4=10 mg/µg. total amount of Hg collected in the sampling Mt=Total mass of that metal collected in the train by using Equation 29–8: sampling train, µg; (substitute Hgt for Mt
Hgt=(Hgfh-Hgfhb)+(Hgbh-Hgbhb) Eq. 29–8 for the Hg calculation). where: Vm(std)=Volume of gas sample as measured by the dry gas meter, corrected to dry Hgt=Total mass of Hg collected in the sam- pling train, µg. standard conditions, dscm. Hgfhb=Blank correction value for mass of Hg 8.7 Isokinetic Variation and Acceptable detected in front-half field reagent Results. Same as Method 5, Sections 6.11 and blank, µg. 6.12, respectively.
1038
VerDate 11
9. Bibliography Performance Specification 8—Performance Specifications for Volatile Organic Com- 1. Method 303F in Standard Methods for the pound Continuous Emission Monitoring Examination of Water Wastewater, 16th Edi- Systems in Stationary Sources tion, 1985. Available from the American Pub- Performance Specification 9—Specifications lic Health Association, 1015 18th Street NW., and Test Procedures for Gas Washington, DC 20036. Chromatographic Countiuous Emission 2. EPA Methods 6010, 6020, 7000, 7041, 7060, Monitoring Systems in Stationary 7131, 7421, 7470, 7740, and 7841, Test Methods for Sources Evaluating Solid Waste: Physical/Chemical Methods. SW–846, Third Edition, September PERFORMANCE SPECIFICATION 1— 1986, with updates I, II, IIA and IIB. Office of SPECIFICATIONS AND TEST PROCEDURES FOR Solid Waste and Emergency Response, U.S. OPACITY CONTINUOUS EMISSION MONITORING Environmental Protection Agency, Washing- SYSTEMS IN STATIONARY SOURCES ton, DC 20460. 1. Applicability and Principle 3. EPA Method 200.7, Code of Federal Regu- lations, Title 40, Part 136, Appendix C. July 1, 1.1 Applicability. This specification con- 1987. tains requirements for the design, perform- 4. EPA Methods 1 through 5, Code of Federal ance, and installation of instruments for Regulations, Title 40, Part 60, Appendix A. opacity continuous emission monitoring sys- July 1, 1991. tems (CEMS’s) and data computation proce- 5. EPA Method 101A, Code of Federal Regu- dures for evaluating the acceptability of a lations, Title 40, Part 61, Appendix B. July 1, CEMS. Certain design requirements and test 1991. procedures established in this specification may not apply to all instrument designs. In [36 FR 24877, Dec. 23, 1971] such instances, equivalent design require- ments and test procedures may be used with EDITORIAL NOTE: For FEDERAL REGISTER ci- prior approval of the Administrator. tations affecting part 60, appendix A see the Performance Specification 1 (PS 1) applies List of CFR Sections in the Finding Aids sec- to opacity monitors installed after March 30, tion of this volume. 1983. Opacity monitors installed before March 30, 1983, are required to comply with APPENDIX B TO PART 60—PERFORMANCE the provisions and requirements of PS 1 ex- SPECIFICATIONS cept for the following: (a) Section 4. ‘‘Installation Specifica- Performance Specification 1—Specifications tions.’’ and test procedures for opacity continu- (b) Sections 5.1.4, 5.1.6, 5.1.7, and 5.1.8 of ous emission monitoring systems in sta- Section 5, ‘‘Design and Performance Speci- tionary sources fications.’’ Performance Specification 2—Specifications (c) Section 6.4 of Section 6 ‘‘Design Speci- and test procedures for SO2 and NOx con- fications Verification Procedure.’’ tinuous emission monitoring systems in An opacity monitor installed before March stationary sources 30, 1983, need not be tested to demonstrate Performance Specification 3—Specifications compliance with PS 1 unless required by reg- and test procedures for O2 and CO2 con- ulatory action other than the promulgation tinuous emission monitoring systems in of PS 1. If an existing monitor is replaced stationary sources with a new monitor, PS 1 shall apply except Performance Specification 4—Specifications that the new monitor may be located at the and test procedures for carbon monoxide old measurement location regardless of continuous emission monitoring systems whether the location meets the requirements in stationary sources of Section 4. If a new measurement location Performance Specification 4A—Specifica- is to be determined, the new location shall tions and test procedures for carbon meet the requirements of Section 4. monoxide continuous emission monitor- 1.2 Principle. The opacity of particulate ing systems in stationary sources matter in stack emissions is continuously Performance Specification 5—Specifications monitored by a measurement system based and test procedures for TRS continuous upon the principle of transmissometry. Light emission monitoring systems in station- having specific spectral characteristics is ary sources projected from a lamp through the effluent Performance Specification 6—Specifications in the stack or duct, and the intensity of the and test procedures for continuous emis- projected light is measured by a sensor. The sion rate monitoring systems in station- projected light is attenuated because of ab- ary sources sorption and scattered by the particulate Performance Specification 7—Specifications matter in the effluent; the percentage of and test procedures for hydrogen sulfide visible light attenuated is defined as the continuous emission monitoring systems opacity of the emission. Transparent stack in stationary sources emissions that do not attenuate light will
1039
VerDate 11