Stand Alone Report 12 Ambient Air Quality Modeling Protocol and Results Bear Lodge Project – Upton Hydrometallurgical Plant
Ambient Air Quality Modeling Protocol and Results Bear Lodge Project – Upton Hydrometallurgical Plant Rare Element Resources, Inc. Weston County, Wyoming
September, 2014
Prepared by:
IML Air Science a division of Inter-Mountain Laboratories, Inc.
555 Absaraka Sheridan, Wyoming 82801 (307) 674-7506 www.imlairscience.com TABLE OF CONTENTS SECTION PAGE
1 INTRODUCTION ...... 1 1.1. Project Overview ...... 2 1.2. Modeling Overview ...... 2 1.3. Pollutants of Concern...... 3 1.4. Regulatory Status ...... 4 2 EMISSION AND SOURCE DATA ...... 5 2.1. Facility Processes and Emission Controls Affected ...... 5 2.2. Emission Factors Used to Calculate Potential Emissions ...... 5 2.3. Schedule of Fugitive Particulate Emissions ...... 6 2.4. Schedule of Tailpipe Emissions ...... 7 2.5. Source Parameters ...... 8 2.6. Greenhouse Gas Emissions ...... 13 3 AMBIENT AIR QUALITY IMPACT MODELING METHODOLOGY ...... 14 3.1. Model Selection and Justification ...... 14 3.2. Model Options ...... 14 3.3. Averaging Periods ...... 15 3.4. Building Downwash ...... 15 3.5. Elevation Data ...... 15 3.6. Receptor Network ...... 16 3.6.1. Fenceline Receptors ...... 16 3.6.2. Public Road Corridor Receptors ...... 16 3.6.3. Fine Grid ...... 16 3.6.4. Intermediate Grid ...... 16 3.6.5. Coarse Grid ...... 17 3.7. Meteorological Data ...... 20 3.8. Background Concentrations ...... 20
3.9. Ambient Ratio Method for Modeling NO2 ...... 21
3.10. Dry Depletion Option for Modeling PM10 ...... 23
3.10.1. Rationale for Using Dry Depletion in Refined PM10 Analysis .... 24
3.10.2. Precedent for Using Dry Depletion in Refined PM10 Analysis ... 24
Upton Plant Modeling Protocol and Results Page ii 3.10.3. Input Parameters for Dry Depletion Option ...... 25 4 APPLICABLE REGULATORY LIMITS FOR CITERIA POLLUTANTS ...... 27 4.1. Methodology for Evaluation of Compliance with Standards ...... 27 4.2. NAAQS and PSD Increments ...... 27 4.3. Presentation of Modeling Results ...... 28 4.4. Summary ...... 29 5 AIR QUALITY RELATED VALUES (AQRV) MODELING METHODOLOGY .... 30 5.1. Introduction ...... 30 5.2. Model Selection and Justification ...... 33 5.2.1. CALMET ...... 34 5.2.2. CALPUFF ...... 34 5.2.3. CALPOST ...... 34 5.3. Meteorological, Terrain and Land Use Data ...... 34 5.3.1. Time Period ...... 35 5.3.2. Prognostic Meteorological Data ...... 35 5.3.3. CALMET Diagnostic Meteorological Data ...... 35 5.3.4. CALMET Approach ...... 36 5.3.5. CALMET Parameter Settings ...... 36 5.3.6. Terrain Data ...... 36 5.3.7. Land Use Data ...... 36 5.3.8. CALMET Switch Settings ...... 36 5.4. Modeling Domain, Sources and Receptors ...... 37 5.5. CALPUFF Model Inputs ...... 43 5.5.1. Background Concentrations ...... 43 5.5.2. Chemistry Modeling ...... 43 5.5.3. Particle Size Distribution ...... 43 5.5.4. CALPUFF Switch Settings ...... 44 5.6. CALPUFF Model Outputs, Calculations and Evaluation Methods ...... 45 5.6.1. CALPOST and POSTUTIL ...... 45 5.6.2. Visibility Impact Determination ...... 46 5.6.3. Comparison to Existing AQRV Status ...... 46 5.6.4. Calculation of Light Extinctions ...... 46 5.6.5. Deposition Analysis ...... 48
Upton Plant Modeling Protocol and Results Page iii 5.6.6. CALPOST Switch Settings ...... 48 5.7. Presentation of Modeling Results ...... 49 6 AERMOD MODELING RESULTS AND ANALYSIS ...... 51 6.1. Introduction ...... 51
6.2. PM10 Modeling Analysis ...... 54
6.2.1. Initial PM10 Modeling Results ...... 55
6.2.2. PM10 Modeling Over-Prediction Problems ...... 60
6.2.3. Refined PM10 Modeling Results ...... 60
6.3. PM2.5 Modeling Analysis ...... 63
6.3.1. PM2.5 Modeling Results ...... 63
6.4. NO2 Modeling Analysis ...... 68
6.5. SO2 Modeling Analysis ...... 73 6.6. CO Modeling Analysis...... 80 7 CALPUFF MODELING RESULTS AND ANALYSIS ...... 83 7.1. Introduction ...... 83 7.2. Visibility Analysis ...... 85 7.2.1. Basis for Analysis ...... 85 7.2.2. Preliminary Modeled Visibility Impacts ...... 86 7.2.3. Effect of Coarse Particulate on CALPUFF Visibility Assessment .. 88 7.2.4. Final Modeled Visibility Impacts ...... 88 7.3. Deposition Analysis ...... 89 7.3.1. Basis for Analysis ...... 89 7.3.2. Modeled Deposition Fluxes ...... 90 8 REFERENCES ...... 92 9 APPENDIX A ...... 96 10 APPENDIX B ...... 103
Upton Plant Modeling Protocol and Results Page iv LIST OF TABLES Table 2-1: Maximum Potential Fugitive Emissions (tons/year) ...... 7 Table 2-2: Maximum Engine, Boiler and Heater Emissions per Year (Tons) ...... 8 Table 2-3: Area Source Emission Totals (tons) ...... 9 Table 2-4: Natural Gas Point Source Emission Rates and Stack Parameters ...... 9 Table 2-5: Particulate Point Source Emission Rates and Stack Parameters ...... 10 Table 2-6: Maximum Area and Point Source Emission Rates (tons/yr) ...... 10 Table 3-1: Non-Default Settings in AERMOD ...... 15 Table 3-2: Assumed Background Concentrations for Modeling Analysis ...... 21
Table 3-3: Assumed PM10 Particle Size Distribution for Dry Depletion Option ...... 26 Table 4-1: National Ambient Air Quality Standards (µg/m3) ...... 28 Table 5-1: CALMET Switch Settings ...... 37
Table 5-2: Fugitive PM10 Particle Size Distribution ...... 44 Table 5-3: CALPUFF Switch Settings ...... 45 Table 5-4: CALPOST Switch Settings ...... 50 Table 6-1: Summary of Predicted Pollutant Concentrations (AERMOD) ...... 52 Table 6-2: Summary of PSD Increment Comparisons (AERMOD) ...... 53
Table 6-3: Top 20 Receptors, Annual Average PM10 Impacts (Initial Run) ...... 56 nd Table 6-4: Top 50 Receptors, 24-Hr 2 High PM10 Concentrations (Initial Run) ...... 57 nd Table 6-5: Top 20 Receptors, 24-Hr 2 High PM10 Values With Dry Depletion ...... 61
Table 6-6: Top 20 Receptors, Annual Average PM2.5 Values ...... 64 th Table 6-7: Top 20 Receptors, 98 percentile of 24-Hr Maximum PM2.5 Values ...... 65
Table 6-8: Top 20 Receptors, Annual Average NO2 (ARM) ...... 69 th Table 6-9: Top 20 Receptors, 98 percentile of Daily Max 1-Hr NO2 Values (ARM2) ... 70
Table 6-10: Top 20 Receptors, 3-Hr Maximum SO2 ...... 74 th Table 6-11: Top 20 Receptors, 99 percentile of Daily Maximum 1-Hr SO2 Values ...... 75 Table 7-1: Visibility Analysis Summary ...... 86 Table 7-2: Visibility Analysis Top 50 delta-dv Values (without coarse PM) ...... 87 Table 7-3: Current Acid Deposition at Wind Cave National Park (kg/ha/yr) ...... 91 Table 7-4: Acid Deposition Modeling Analysis at Devils Tower NM (Wet + Dry) ...... 91
Upton Plant Modeling Protocol and Results Page v LIST OF FIGURES
Figure 2-1: Upton Plant Emission Source Identification ...... 11 Figure 2-2: AERMOD Map of Upton Plant Emission Source Locations ...... 12 Figure 3-1: Upton Plant Permit Boundary and Fine Grid Receptors ...... 18 Figure 3-2: Upton Plant Receptors and Modeling Domain ...... 19
Figure 3-3: ARM2 Method Equation for Ambient NO2/NOx Ratio ...... 22 Figure 5-1: Bear Lodge Project and Nearest Sensitive Areas ...... 32 Figure 5-2: Bear Lodge Project CALPUFF Modeling Domain ...... 39 Figure 5-3. CALPUFF Model Receptors and Emission Sources ...... 40 Figure 5-4. Bull Hill Mine Emission Sources ...... 41 Figure 5-5. Upton Plant Emission Sources ...... 42
Figure 6-1. Initial Annual Average PM10 Concentrations (Without Background) ...... 58 nd Figure 6-2. Initial 2 High 24-Hr Average PM10 Concentrations (Without Background) 59
Figure 6-3. Receptors Exceeding the 24-Hour PM10 Standard...... 62
Figure 6-4. Annual PM2.5 Concentrations (Without Background) ...... 66 th Figure 6-5. 98 Pctile 24-Hour PM2.5 Concentrations (Without Background) ...... 67
Figure 6-6. Annual NO2 Concentrations (ARM, Without Background) ...... 71 th Figure 6-7. 98 Percentile 1-Hr NO2 Concentrations (ARM2, Without Background) ..... 72
Figure 6-8. Modeled Annual SO2 Concentrations (Without Background) ...... 76
Figure 6-9. Modeled Maximum 24-Hour SO2 Concentrations (Without Background) .... 77
Figure 6-10. Modeled Maximum 3-Hour SO2 Concentrations (Without Background) .... 78 Figure 6-11. Modeled 99th Percentile 1-Hour SO2 Concentrations (w/o Background) . 79 Figure 6-12. Modeled Maximum 8-Hr CO Concentrations (Without Background) ...... 81 Figure 6-13. Modeled Maximum 1-Hr CO Concentrations (Without Background) ...... 82
LIST OF APPENDICES Appendix A: Emission Inventory Calculations Appendix B: Source Apportionment and Timing
Upton Plant Modeling Protocol and Results Page vi 1 INTRODUCTION Rare Element Resources, Inc. (RER) proposes to construct the Bear Lodge Project, a rare earth element mine and separate processing facility in northeast Wyoming. The proposed processing facility will be a hydrometallurgical plant located in Weston County, at the outskirts of Upton, Wyoming. Rare earth ore will be delivered to the Upton plant from the companion Bull Hill Mine located approximately 30 miles north of Upton in Crook County, Wyoming. An assessment of the potential air quality impacts of these proposed facilities is required as part of the environmental review process under U.S. Nuclear Regulatory Commission Rules and Regulations. RER enlisted IML Air Science to develop a project emissions inventory and to model the potential impacts of these emissions on ambient air quality. IML was also asked to assess potential project impacts on Air Quality Related Values (AQRV’s) at Class I and sensitive Class II areas. The following modeling protocol and modeling report address the Upton plant portion of the Bear Lodge Project.
The Upton plant site and surrounding area lie within a Class II airshed under the Clean Air Act. There are no Class I areas within 50 kilometers of the Upton plant site. The nearest Class I area is Wind Cave National Park, approximately 70 miles southeast of the Upton site. The second nearest is Badlands National Park, 115 miles to the east. The Northern Cheyenne Indian Reservation is approximately 140 miles to the northwest. Sensitive Class II areas within 100 km of the Upton site include Devils Tower National Monument, Jewell Cave National Monument, Black Elk Wilderness Area and Mount Rushmore National Memorial (BLM 2014). An AQRV analysis will be conducted for a 200-km square modeling domain that includes Wind Cave National Park and the above-named sensitive Class II areas. Associated impacts at Badlands National Park are presumed to be less than those at Wind Cave since it lies farther to the east.
This air quality modeling protocol addresses the approach for assessing the ambient air quality impacts from the proposed Upton plant emissions for comparison with the National Ambient Air Quality Standards (NAAQS) for particulate matter less than 10
microns in diameter (PM10), particulate matter less than 2.5 microns in diameter (PM2.5),
carbon monoxide (CO), sulfur dioxide (SO2) and nitrogen dioxide (NO2). It also addresses the approach for comparing modeled project impacts to the Prevention of
Significant Deterioration (PSD) Class II increments for PM10, PM2.5, SO2 and NO2. Such comparisons are made for disclosure purposes only. Since the Upton plant does not
Upton Plant Modeling Protocol and Results Page 1
qualify as a major source of any criteria pollutant or hazardous air pollutant (HAP), it is not subject to PSD regulations. Project-related emissions of hazardous air pollutants
(HAP) and greenhouse gases (carbon dioxide or CO2) will be estimated and summarized, but not modeled.
A project emissions inventory is presented in Appendix A to this document. The spatial apportionment of those emissions is detailed in Appendix B.
1.1. Project Overview The proposed processing facility will consist of a hydrometallurgical plant (Hydromet) and tailings storage facility (TSF). Pre-concentrate from the Bull Hill Mine will be unloaded at the plant and processed to further concentrate the rare earth oxides. Processing will include hydrochloric acid leach, precipitation, dewatering and acid regeneration. Waste material will be hauled by truck from the Hydromet plant to the TSF. The TSF will undergo contemporaneous reclamation to maintain a maximum exposed area of 35 acres. Rare earth oxides will be shipped by over-the-road trucks to commercial refining facilities.
Construction of the plant will generate air emissions of a lesser magnitude than subsequent operation of these facilities. This analysis will focus on the ongoing operation phase to identify the anticipated maximum emissions year at the Upton plant and to analyze potential impacts on ambient air quality for that year.
Fugitive emission sources of particulate matter (PM10, PM2.5) include ore and waste haulage activities, wind erosion, product transport, pickup traffic, delivery trucks, support
vehicles and passenger vehicles. Particulates (PM10, PM2.5), carbon monoxide (CO),
carbon dioxide (CO2), oxides of nitrogen and sulfur (NOx and SO2), and a small amount of hazardous air pollutants (HAP) will be emitted by mobile equipment engine exhaust and by stationary sources such as boiler and scrubber stacks, space heaters, an emergency generator and fire water pumps.
1.2. Modeling Overview This modeling protocol addresses two separate modeling scenarios: (1) modeling for ambient air quality impacts (AAQI) at the Upton facility boundary, and at locations within 50 km of the project, and (2) modeling for AQRV impacts from the entire Bear Lodge Project, including visibility and deposition impacts, at Class I and sensitive Class II
Upton Plant Modeling Protocol and Results Page 2
areas in the general vicinity of the Upton plant and Bull Hill Mine. Since these two scenarios utilize a different scope of emission sources and different modeling assumptions, software models, and meteorological data sets, they are addressed separately.
The ambient air quality impact analysis will be performed using the AERMOD dispersion model. Sections 3 and 4 of this document apply to the AERMOD modeling protocol. AQRV impact analysis will be performed using the CALPUFF model. Section 5 applies to the CALMET/CALPUFF modeling protocol. Section 2 discusses project related emissions and modeled emission sources, which apply equally to AERMOD and CALPUFF. The modeling results for the AAQI and AQRV analyses appear in Section 6 and 7, respectively.
1.3. Pollutants of Concern Both combustion emissions and fugitive dust emissions will be modeled in the air quality analysis. The stationary and fugitive emission sources at the Upton plant will produce particulate matter smaller than ten microns in size (PM10) and particulate matter smaller
than 2.5 microns in size (PM2.5). Stationary and mobile sources will emit PM10, PM2.5,
carbon monoxide (CO), sulfur dioxide (SO2) and oxides of nitrogen (NOx). For the NO2 analysis in AERMOD, it is assumed that a portion of the NOx emissions will be
converted to NO2, using variations of the ambient ratio method. Thus, five criteria
pollutants (PM10, PM2.5, CO, SO2 and NO2) will be analyzed for compliance with the
NAAQS. Four of these pollutants, PM10, PM2.5, SO2 and NO2 will be further analyzed for comparison with the PSD increments in Class II areas. The comparisons to the PSD Class II increments will be made for disclosure purposes only, intended to evaluate a threshold of concern for potential impacts. This analysis does not represent a regulatory PSD increment comparison since the Upton plant does not qualify as a PSD source.
Both the NAAQS and the PSD analyses will be conducted using the AERMOD software. The modeling domain for the Upton plant will extend 50 km in all directions from the plant. Modeled impacts within this domain will be compared to the NAAQS and Class II PSD increments.
The principle form of HAP will be formaldehyde in diesel engine exhaust. Formaldehyde emissions from the Upton plant will be inventoried but not modeled. Diesel engines emit from 2% to 5% as much formaldehyde per unit of energy input as natural gas fired
Upton Plant Modeling Protocol and Results Page 3
engines (EPA 1995c). The latter are used extensively in the region for compressor stations, heaters, and other applications in the oil and gas industry. Appendix A shows maximum annual formaldehyde emissions of 0.4 tons at the Upton plant.
1.4. Regulatory Status The Upton plant will be a non-categorical stationary source. Criteria pollutant emissions from the facility will be below the New Source Review major source threshold of 250 tons/year. Therefore, the facility will not be subject to PSD permitting regulations. The potential to emit HAPs will be less than 10 tons/year for any individual HAP, and less than 25 tons/year for all HAPs combined. Therefore, the facility will not be a major HAP source. Point source emissions of criteria pollutants from the facility will be less than the Title V source threshold of 100 tons per year. As a minor source of criteria pollutant emissions and HAPs, the Upton plant will be required to obtain an air quality construction permit from the Wyoming Department of Environmental Quality, Air Quality Division (AQD). The emissions inventory and modeling results generated in this study will be submitted to AQD as part of the air quality construction permit application.
Upton Plant Modeling Protocol and Results Page 4
2 EMISSION AND SOURCE DATA
2.1. Facility Processes and Emission Controls Affected The proposed Upton plant is a hydrometallurgical plant which will process pre- concentrated rare earth oxide ore from the Bull Hill Mine to generate a rare earth element concentrate. Facility processes and emission controls planned for the Upton plant include paving of the plant area, the use of water and a chemical dust suppressant to control fugitive dust emissions from unpaved roads, baghouses and bin vents to minimize dust generation from material handling systems, and a wet scrubber to eliminate acid vapor emissions from the leaching and precipitation circuits.
2.2. Emission Factors Used to Calculate Potential Emissions The Upton plant will generate stationary source, fugitive dust and tailpipe emissions. In general, fugitive dust emissions will include traffic on unpaved roads (including haulage of ore, waste, product, and chemical supplies), traffic on paved roads, earthwork, road maintenance, topsoil activities, and wind erosion on disturbed areas. Emission factors for these sources are provided in EPA’s AP-42, Compilation of Air Pollutant Emission Factors as listed below (EPA 1995c):
Unpaved roads Chapter 13, Section 13.2.2 Paved roads Chapter 13, Section 13.2.1 Earth moving Chapter 13, Section 13.3, Table 13.3-1 Topsoil stripping and reclamation Chapter 11, Section 11.9, Table 11.9-4 Wind erosion Chapter 11, Section 11.9, Table 11.9-4
In some cases fugitive PM2.5 emission factors were not available in AP-42. For wind
erosion and earth moving activities, a PM2.5/ PM10 ratio of 15% was applied to the
respective PM10 emission factor. For unpaved road dust, a PM2.5/ PM10 ratio of 10% was
applied to the respective PM10 emission factor. These ratios follow recommendations in a study performed for the Western Regional Air Partnership (WRAP) by Midwest Research Institute (MRI 2006).
Published fugitive dust emission factors are modified by specific control measures. EPA guidance provided in AP-42 allows for natural mitigation of fugitive dust emissions based on days of precipitation per year (page 13.2.2-7, Equation 2). For the Upton plant area this value was determined to average 90 days per year, from a map presented in
Upton Plant Modeling Protocol and Results Page 5
AP-42 (Figure 13.2.2-1). A precipitation day is defined in Section 13.2 of AP-42 as any day measuring precipitation of 0.01” or more. The emission factor correction for precipitation days applies to all unpaved roads. Guidance also typically allows for 50% control efficiency on project roads with the use of water trucks. A recent federal EIS used 85% control for the combination of water trucks and seasonal application of chemical dust suppressant on unpaved roads (TRC 2005). The Upton plant emissions inventory assumes 85% control on unpaved project roads. Paved road emission factors account for control inherently.
Gasoline and diesel equipment tailpipe emissions were calculated using emission factors from several sources. THC (total hydrocarbon), SO2, CO2 and aldehyde emission factors were taken from AP-42 Chapter 3, Table 3.3-1. NOx, CO, and PM10 emission factors for diesel engines are based on EPA standards for various engine tier ratings and horsepower (EPA 1998). Mobile diesel equipment and emergency generators were assumed to conform to Tier 3 diesel engine standards. PM2.5 emissions from equipment tailpipes were assumed to be 97% of PM10 emissions (EPA 2004a). Emission factors for natural gas fired equipment were obtained from AP-42, Table 1.4-1 (EPA 1995c). Conversion factors for greenhouse gas emissions associated with electricity consumption, stated in terms of carbon dioxide equivalent (CO2e), were obtained from EPA’s Clean Energy website (EPA 2013a).
In most cases, equipment activity levels used to calculate emissions were based on scheduled hours, equipment availability and typical load factors. Load factors for pickup trucks, passenger vehicles, and certain support equipment were assumed to be 20% to 25%, due to low throttle settings and/or intermittent operation. The load factor for light plants was assumed to be 100% due to continuous operation when scheduled to run. Most other load factors, including those for haul trucks and loaders, were assumed to be 40%. In its Nonroad Model, EPA uses 40% for much of the diesel powered equipment (EPA, 2010).
2.3. Schedule of Fugitive Particulate Emissions The potential fugitive emission rates from the Upton plant are summarized in Table 2-1. Detailed emission calculations for the proposed project have been provided in Appendix A. The basis for source apportionment of equipment-generated fugitive emissions is presented in Appendix B. The scenario with the maximum yearly equipment activity will be modeled since it corresponds to the highest total for fugitive dust emissions, at 17
Upton Plant Modeling Protocol and Results Page 6
tons of PM10 (Table 2-1). This scenario occurs at full production, and also corresponds
to the year of maximum PM2.5 and gaseous emissions.
Table 2-1: Maximum Potential Fugitive Emissions (tons/year)
2 Activity Area (ft ) PM10 tons/yr PM2.5 tons/yr Wind Erosion on Disturbed Areas 3,943,922 10.32 1.55
Equipment Item Quantity PM10 tons/yr PM2.5 tons/yr Front End Loader 1 0.79 0.12 End Dump Truck 1 1.95 0.19 Track Dozer 1 0.01 0.00 Motor Grader 1 0.96 0.10 Water Truck 1 2.60 0.26 Passenger Vehicle 15 0.04 0.01 Pickup Truck 2 0.04 0.01 Preconcentrate Transport Truck 9 0.28 0.07 Commercial Delivery Truck 1 0.00 0.00 TOTAL FUGITIVE DUST EMISSIONS (TONS/YEAR) 17.00 2.31
2.4. Schedule of Tailpipe Emissions Table 2-2 summarizes potential combustion emissions from equipment tailpipes and stationary sources. All stationary source emissions are assumed to be from natural gas combustion. With the exception of facilities construction, these emissions are assumed to be constant from year to year. The highest annual tailpipe and stationary source emissions of PM10, PM2.5, CO, SO2 and NOx will coincide with the year of maximum
fugitive dust emissions. NOx emissions are projected to peak at 37.57 tons per year. Detailed emission calculations for the proposed project have been provided in Appendix A. The apportionment of tailpipe emissions to modeled area sources is presented in Appendix B.
For purposes of modeling in AERMOD using the Tier 2 ARM option, NOx emissions will
be multiplied by 0.75 to estimate annual NO2 impacts. To estimate 1-hour NO2 impacts,
the ARM2 option will be used, since it accounts for variable ambient NO2/ NOx ratios
based on absolute NOx concentration (see Section 3.2 below). NO2 is the regulated pollutant, with associated NAAQS and PSD increments, per Section 6.2.3 of EPA’s Guideline on Air Quality Models (40 CFR 51 Appendix W).
Upton Plant Modeling Protocol and Results Page 7
Table 2-2: Maximum Engine, Boiler and Heater Emissions per Year (Tons)
Total Tons/Yr by Equipment THC NOx CO SO2 CO2 PM10 PM2.5 HAP Front End Loader 0.07 0.19 0.17 0.06 34 0.01 0.01 0.01 End Dump Truck 0.13 0.36 0.31 0.11 63 0.02 0.02 0.03 Track Dozer 0.030.070.090.02130.000.000.01 Motor Grader 0.17 0.46 0.40 0.14 80 0.02 0.02 0.03 Water Truck 0.22 0.58 0.50 0.18 100 0.03 0.03 0.04 Portable Light Plant 0.02 0.30 0.07 0.02 11 0.02 0.02 0.00 Passenger Vehicle 0.11 0.29 0.25 0.09 50 0.01 0.01 0.02 Pickup Truck 0.040.040.580.05940.060.060.04 Preconcentrate Transport Truck 0.07 0.10 7.97 0.48 268 0.01 0.00 0.11 Commercial Delivery Truck 0.00 0.00 0.33 0.02 11 0.00 0.00 0.00 Emergency Generator 0.10 0.26 0.23 0.01 46 0.01 0.01 0.02 Firewater Pump 0.00 0.03 0.01 0.00 1 0.00 0.00 0.00 Process Boiler 11.23 31.25 85.78 0.61 122,537 7.76 7.76 0.08 Dryer Oil Heaters 0.35 3.03 1.29 0.02 3,865 0.24 0.24 0.00 Plant Heater 0.05 0.40 0.17 0.00 515 0.03 0.03 0.00 Shop Heater 0.01 0.12 0.05 0.00 155 0.01 0.01 0.00 Office Heater 0.01 0.08 0.03 0.00 103 0.01 0.01 0.00
FACILITY TOTALS (tons per year) 12.62 37.57 98.21 1.83 127,946 8.26 8.25 0.40
2.5. Source Parameters The 25 modeled emission sources in AERMOD will include area sources and point sources. Area sources include disturbed acreage, overburden and tailings storage areas, ponds, project roads, the access road to the project, and plant facilities. AERMOD release heights for area sources of fugitive dust will follow recent EPA guidance (EPA 2012) assuming average vehicle heights are 3.0 meters for heavy equipment, and 2.0 meters for lighter vehicles. Based on this guidance, release heights for 3-meter and 2-meter vehicle heights are 2.55 and 1.70 meters, respectively. Corresponding sigma-Z values are 2.37 and 1.58 meters, respectively. For those sources dominated by wind erosion (e.g. facilities areas), release heights are assumed to be 1 foot and sigma-Z is assumed to be zero. Release heights for equipment tailpipe emissions are assumed to be 1 meter for heavy equipment and 0.5 meter for light vehicles, with a sigma-Z of zero.
Appendix B details the apportionment of equipment and fugitive emissions among these sources. Based on this apportionment process, Table 2-3 summarizes area source emissions (tons/year) of modeled pollutants for the modeled year. For particulate sources, these emissions include both fugitive and tailpipe emissions.
Upton Plant Modeling Protocol and Results Page 8
Table 2-3: Area Source Emission Totals (tons)
Area Source Category PM10 PM2.5 NOx SO2 CO
TSF 10.07 1.43 0.85 0.22 0.76 OB Storage 2.37 0.36 0.07 0.00 0.01 Pond Area 0.41 0.06 0.01 0.00 0.00 Haulroad 3.11 0.37 0.85 0.27 0.81 Access Road 0.30 0.10 0.33 0.46 6.62 Delivery Road 0.09 0.03 0.03 0.13 2.11 Pickup Road 0.02 0.01 0.06 0.03 0.15 Hydromet Plant 0.02 0.01 0.29 0.02 0.24 Tailings Stockpile 0.80 0.13 0.19 0.06 0.17
TOTAL 17.20 2.51 2.69 1.19 10.89
Table 2-4 summarizes combustion point source emission rates (g/sec) and associated stack parameters for the modeled year. All modeled point sources have a vertical discharge.
Table 2-4: Natural Gas Point Source Emission Rates and Stack Parameters
Oil Plant Office Point Sources Boiler 1 Shop Heater Heaters Heater Heater No. of Units 1 3 1 1 1 MMBtu/hr per Unit 38.80 2.50 2.00 0.60 0.40 Stack Diameter (meters) 0.6096 0.440 0.2286 0.2032 0.1524 Stack Height (meters) 30 20 25 10 10 Temp °K 422 366 344 344 344 Flow Rate (m3/sec) 6.899 1.065 0.290 0.087 0.058 Velocity (m/sec) 23.64 21.02 7.07 2.68 3.18
Emission Rate PM10 (g/sec) 3.646E-02 2.114E-02 1.879E-03 5.638E-04 3.759E-04
Emission Rate PM2.5 (g/sec) 3.646E-02 2.114E-02 1.879E-03 5.638E-04 3.759E-04
Emission Rate NOx (g/sec) 4.797E-01 2.615E-01 2.324E-02 6.973E-03 4.649E-03
Emission Rate SO2 (g/sec) 2.878E-03 1.669E-03 1.484E-04 4.451E-05 2.967E-05 Emission Rate CO (g/sec) 4.030E-01 1.113E-01 9.891E-03 2.967E-03 1.978E-03
Note 1: stack diameter is effective combined diameter; emission rates are for 3 oil heaters combined
Upton Plant Modeling Protocol and Results Page 9
Table 2-5 summarizes particulate (dust) point source emission rates (g/sec) and associated stack parameters for the modeled year. All modeled point sources have a vertical discharge.
Table 2-5: Particulate Point Source Emission Rates and Stack Parameters
PM10 Emission Emission 1 1 1 Stack1 Stack Flow Em ission Rate Rate No. of Diameter Height Rate Velocity Rate PM10 PM2.5 Baghouse Source Units (meters) (meters) Temp °FTemp °K (m3/sec) (m/sec) (gr/dscf) (g/sec) (g/sec) Chemical Unloading Bin Vents 4 0.4064 10 80 299.7 0.8825 6.80 0.005 0.0346 0.0052 PC Feed Chute Bin Vent 1 0.2032 10 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 PC Discharge Chute Baghouse 1 0.4572 10 80 299.7 2.7578 16.80 0.005 0.0270 0.0041 Silo Feed Bin Vent 1 0.2032 20 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 Product Transfer Bin Vents 3 0.3520 10 80 299.7 0.6619 6.80 0.005 0.0195 0.0029 Screw Dryer Scrubber 2 0.4311 10 160 344.1 1.0134 6.94 0.005 0.0173 0.0026 Ammonium Nitrate Scrubber 1 0.2032 10 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 Ventilation Scrubber 1 0.3048 20 80 299.7 0.5516 7.56 0.005 0.0054 0.0008 Kiln Effluent Gas Scrubber 1 0.381 20 200 366.3 1.3485 11.83 0.005 0.0108 0.0016
Note 1: stack diameter is the effective diameter for multiple units; flow rates and emission rates are combined for multiple units
Table 2-6 summarizes maximum emission rates from all source types at the Upton facility.
Table 2-6: Maximum Area and Point Source Emission Rates (tons/yr)
Source Type PM10 PM2.5 NOx SO2 CO
Area Sources 17.20 2.51 2.69 1.19 10.89 Combustion Point Sources 8.05 8.05 34.88 0.64 87.32 Particulate Point Sources 4.20 0.63 0.00 0.00 0.00
TOTAL 29.46 11.19 37.57 1.83 98.21
Figure 2-1 shows the Upton Plant permit boundary along with the locations and orientations of all modeled area sources. Modeled point sources reside within the main facilities area in the northwestern portion of the project area (Figure 2-2). Area sources were digitized as rectangles or polygons to reduce model complexity and execution time. Line area sources correspond to project and access roads. AERMOD accepts the digitized centerline of line area sources and generates rectangles with user-specified dimensions paralleling this centerline. Figure 2-1 identifies the modeled area sources by the name identifying a particular area source group. Figure 2-2 confirms the area source locations and adds point sources on a map produced by AERMOD.
Upton Plant Modeling Protocol and Results Page 10
Source emission rates will be assumed to be uniform since the Upton Plant will operate continuously. For point sources, average emission rates in tons/year will be converted to grams/second. For area sources, average emission rates of tons/year will be converted to lbs./hour/ft2 for the area over which the source emissions are distributed. Appendix B details the emission source apportionment among these modeled areas.
Figure 2-1: Upton Plant Emission Source Identification
4884400
4884000
4883600 Permit Boundary TSF
OB2 4883200 PLANT
OB1
POND 4882800 ACCESS_RD
DELIV_RD
HAUL_RD 4882400
4882000 524800 525200 525600 526000 526400 526800 527200 527600 528000
Upton Plant Modeling Protocol and Results Page 11
Figure 2-2: AERMOD Map of Upton Plant Emission Source Locations
Upton Plant Modeling Protocol and Results Page 12
2.6. Greenhouse Gas Emissions
Greenhouse gas (GHG) emissions are inventoried but not modeled. There are no NAAQS associated with GHG concentrations in the atmosphere. The significant sources of GHG associated with the Upton plant are combustion emissions, process
emissions and equivalent emissions from electricity consumption in the form of CO2. Combustion emissions from equipment engine exhaust and natural gas-powered, stationary equipment are estimated using emission factors from AP-42. Table 2-2 shows the estimated CO2 totals from combustion, with a maximum of 127,946 tons per
year (tpy). Process emissions of CO2 (from chemical reactions) are expected to add an estimated 20,498 tpy.
The principal uses of electricity for the Upton Plant are the material classification equipment, the HCL recovery unit, and the utilities required to operate the plant. The
estimated indirect GHG emissions, or CO2e, from electricity consumption will reach a maximum of 30,797 tpy.
Minor amounts of methane and nitrous oxides, both of which are considered greenhouse gases, will be emitted from natural gas combustion. The GHG potential or
CO2 equivalent of these emissions is a small fraction of one percent of the estimated
total CO2 emissions from the Upton plant. Therefore, the above figures for maximum
direct and indirect CO2 emissions are representative of GHG emissions, yielding a maximum combined GHG emissions of 179,241 tpy.
Upton Plant Modeling Protocol and Results Page 13
3 AMBIENT AIR QUALITY IMPACT MODELING METHODOLOGY
3.1. Model Selection and Justification The proposed facility includes point sources, a tailings storage area, overburden storage areas, a pond, project roads and disturbed areas. The area sources have a wide range of parameters that are too complex to merge into a single emission point. Therefore, criteria pollutant emissions will be modeled with the American Meteorological Society (AMS) and EPA Regulatory model (AERMOD) Version 13350 to evaluate air dispersion from multiple sources. AERMOD was chosen over the Industrial Source Complex (ISC3) model since it has been promulgated by the EPA as the preferred air dispersion model in the Agency's "Guideline on Air Quality Models" (40 CFR 51 Appendix W). AERMOD officially replaced the ISC3 air dispersion model effective December 9, 2006 (one year after rule promulgation) as published in the Federal Register on November 9, 2005. The Lakes Environmental software will be used to implement the AERMOD model (Lakes AERMOD View Version 8.5).
3.2. Model Options The AERMOD regulatory settings will be left in the default settings with two exceptions. First, while the regulatory default Tier 2 ambient ratio (ARM) method will be used to
model annual average NO2 concentrations, the non-default ARM2 method will be used
to model 1-hour NO2 concentrations. For ARM, EPA recommends an ambient NO2/NOx ratio of 75% for the annual average, and 80% for the 1-hour values (EPA 2011). ARM2 estimates the conversion of NOx to NO2 from empirical data and is considered more
realistic for modeling high NOx concentrations on a short-term basis. Section 3.9 describes these conversion options in greater detail and offers a justification for the
superiority of ARM2 for modeling 1-hour NO2 impacts.
Second, for modeling short-term PM10 impacts, the dry depletion option will be evaluated and compared to the default setting (no dry depletion). The default setting will be used for a screening model run, followed by a refined model run for select receptors using the non-default setting. Section 3.10 below discusses the basis for modeling
fugitive PM10 from dust emissions using dry depletion. Table 3-1 summarizes the non- default settings used for AERMOD.
Upton Plant Modeling Protocol and Results Page 14
Table 3-1: Non-Default Settings in AERMOD
NON‐DEFAULT OPTION PURPOSE MODELING SCENARIO
Ambient Ratio Method ARM2 Modeling NO2 conversion 1‐hr averaging interval for NO2
Dry Depletion Account for particle deposition Refined PM10 analysis only
The US EPA 1-hour NO2 NAAQS option in AERMOD View will be exercised to output th both the annual average and the 98 percentile 1-hour NO2 concentrations for each receptor. The 98th percentile 1-hour concentration is in keeping with the format of the NAAQS standard.
3.3. Averaging Periods For the purpose of this modeling analysis, the annual and 24-hour averaging periods
will be utilized for PM10 and PM2.5 modeling. The 8-hour and 1-hour averaging periods will be used for CO modeling. The annual and 1-hour averaging periods will be used for
NO2 while the annual, 24-hour, 3-hour and 1-hour averaging periods will be used for
SO2 modeling. These averaging periods are consistent with the NAAQS primary and secondary standards and the PSD increments. All short-term model results will be presented in the format of the appropriate NAAQS standard, adjusted to the modeling nd period of one year. These include: (a) 2 high 24-hour PM10 value over one year, (b) th th th one-year 98 percentile, or 8 high 24-hour PM2.5 value over one year, (c) one-year 98 th th th percentile, or 8 high 1-hour NO2 value, (d) one-year 99 percentile, or 4 high 1-hour
SO2 value.
3.4. Building Downwash Based on the proposed facility design, buildings and/or structures will cause negligible influences on normal atmospheric flow in the immediate vicinity of the emission sources. Therefore building downwash will not be modeled.
3.5. Elevation Data The terrain surrounding the Upton site is somewhat hilly. In addition, the terrain encompassing model receptors includes hills and valleys. Therefore, the Elevated Terrain mode will be used. Receptor elevations will be obtained from USGS digital elevation model (DEM) files.
Upton Plant Modeling Protocol and Results Page 15
3.6. Receptor Network Figure 3-1 shows the Upton permit boundary and close-in receptors. Figure 3-2 displays the entire modeling domain and overall AERMOD receptor placement. The model domain includes a total of 3,783 receptors, including public road corridor, fenceline, fine grid, intermediate grid and coarse grid receptors. The receptor grid extends in all directions from the project site to at least 50 km from the project center. The receptor network is described below.
3.6.1. Fenceline Receptors Discrete receptors will be placed along the project boundary at least every 100 meters in linear fenceline distance, with a receptor placed at each boundary corner. Areas inside the project boundary will not be analyzed.
3.6.2. Public Road Corridor Receptors Discrete receptors will be placed along the short public access road leading to the plant site. These receptors will be spaced 100 meters apart and positioned 150 meters on either side of the centerline of the roadway. The purpose of the road corridor receptors is to assess potential impacts on ambient air due to fugitive dust emissions from project- related traffic.
3.6.3. Fine Grid A fine grid of receptors will be placed at 100-meter spacing within a 1,000-meter-wide corridor around the project boundary (Figure 3-1). The placement of these fine-grid receptors is intended to identify the highest impacts, particularly 24-hour PM10 and 1-
hour NO2, which would be expected to occur either along the fenceline or within this 1,000-meter-wide corridor.
3.6.4. Intermediate Grid In addition to the fine grid, an intermediate grid of receptors will be placed at 500-meter spacing, from the outer edge of the fine grid outward to a distance 10 kilometers (km) in all directions from the project center. A second intermediate grid will be placed at 1-km spacing, from the outer edge of the first intermediate grid outward in all directions to a distance 15 km from the project center (Figure 3-2).
Upton Plant Modeling Protocol and Results Page 16
3.6.5. Coarse Grid A coarse grid will be placed at 5-km spacing, from the outer edge of the second intermediate grid outward in all directions to a distance of 50 km from the project center. (Figure 3-2).
Upton Plant Modeling Protocol and Results Page 17
Figure 3-1: Upton Plant Permit Boundary and Fine Grid Receptors
Upton Plant Modeling Protocol and Results Page 18
Figure 3-2: Upton Plant Receptors and Modeling Domain
Upton Plant Modeling Protocol and Results Page 19
3.7. Meteorological Data The baseline meteorological data collected from the Upton Plant weather station represents one year (August 1, 2012 through July 31, 2013). EPA requires that AERMOD be run with a minimum of five years of National Weather Service data, or a minimum of one year of on-site meteorological data (EPA 2005). Therefore the model will use the hourly data collected on-site for surface meteorology. The Upton Plant meteorological station meets EPA’s Meteorological Monitoring Guidance for Regulatory Modeling Applications (EPA 2000).
No upper air data are available at the Upton site. The upper air data will be obtained from the nearest available (and most representative) source, the Rapid City, South Dakota National Weather Service upper air site. This data set will be processed using the AERMET program. The surface characteristics (albedo, bowen ratio and roughness) representative of the land type surrounding the meteorological station location are required by the AERMET data processing procedures. AERSURFACE will be used to estimate the surface characteristics at the site based on land use/type files generated by the USGS. The AERMET program will combine the on-site meteorological data with the upper air data to create the AERMOD meteorological data files.
3.8. Background Concentrations For this ambient air quality impact analysis, only the project impacts will be modeled. Background concentrations for each pollutant and averaging interval will be added to the modeled impacts to estimate maximum, total ambient concentrations for comparison to the NAAQS. The Air Quality Division (AQD) of the Wyoming Department of Environmental Quality has recommended background concentrations for modeled pollutants in eastern Wyoming. The data sources for assumed background concentrations are as follows:
PM10 – Antelope Coal Mine background monitor (AQD)
PM2.5 – Newcastle Refining air quality permit modeling (AQD)
NO2 – Newcastle Refining air quality permit modeling (AQD)
SO2 – Newcastle Refining air quality permit modeling (AQD) CO – Newcastle Refining air quality permit modeling (AQD) Table 3-2 lists the background concentrations to be used for this modeling analysis. The
Antelope Mine PM10 monitor, with multiple years of hourly data collected, is located approximately 60 miles southwest of the Upton Plant site. Antelope Mine represents a
Upton Plant Modeling Protocol and Results Page 20
conservative background site given the extent of surface coal mine and oil and gas
development in the region. Background concentrations provided by AQD for PM2.5, NO2,
SO2 and CO were used in modeling for a recent project near Newcastle, and are deemed representative of northeastern Wyoming (WDEQ 2013).
Table 3-2: Assumed Background Concentrations for Modeling Analysis Back- Averaging Interval ground NAAQS Limit Pollutant and Statistic (µg/m3) (µg/m3) Annual Average 15 -- PM10 4th High 24-Hr 40 150 Maximum Annual Average 3.4 12 PM2.5 24-Hr High 8 35 Annual Average 6 100 th NO2 98 Percentile of 21 187 Daily 1-Hr Highs Annual Average 1.3 -- 24-Hr 16.3 -- SO2 3-Hr 124.7 1300 th 99 Percentile of 43.2 200 Daily 1-Hr Highs 8-Hr High 378 10000 CO 1-Hr High 680 40000
3.9. Ambient Ratio Method for Modeling NO2
The Upton plant will be modeled for NO2 impacts using the Tier 2 ARM option for
modeling annual average NO2 concentrations. The more recently developed ARM2
option will be used to model 1-hour NO2 concentrations for comparison to the 1-hour NAAQS, for which peak concentrations can be critical. The ARM2 option is believed to
be more realistic since it adjusts the ambient NO2/NOx ratios according to modeled NOx concentrations. Higher concentrations yield lower ratios due to the depletion of ozone within the high-NOx impact region (Trinity 2014). ARM2 adjusts the NO2/NOx ratios based on an empirical fit to tens of thousands of monitored ratios across the country (Figure 3-3). The upper bound of the observed ambient ratios in Figure 3-3 was th estimated for each NOx concentration bin by calculating the 98 percentile of the observed ratios in that bin (RTP 2013). This results in a conservatively high estimate of
Upton Plant Modeling Protocol and Results Page 21
the ambient ratio. Like other conversion options in AERMOD, ARM2 has still been found to over-predict ambient NO2/NOx ratios at high modeled NOx concentrations (API 2013).
In addition to the empirical fit shown in Figure 3-3, the ambient ratio established by the ARM2 method is bounded above by a maximum of 0.90. It is bounded below by a minimum of 0.20, which corresponds to the assumed in-stack NO2/NOx ratio. The assumption of 0.20 is conservative, as one recommended in-stack ratio for natural gas fired boilers is 0.10 (CAPCOA 2011). The boiler will be the chief source of NOx at the Upton Plant.
Figure 3-3: ARM2 Method Equation for Ambient NO2/NOx Ratio
Source: RTP Env. Assoc. Inc. 2013 R/S/L Modelers Workshop, April 2013
The ARM option is even more conservative than ARM2. It assumes fixed ambient ratios (0.75 for annual and 0.80 for 1-hour concentrations) that do not account for lower rates of conversion at high-end NOx concentrations. ARM2’s implementation in AERMOD represents a more refined approach than ARM to generating the appropriate ambient ratios (Trinity 2014).
Upton Plant Modeling Protocol and Results Page 22
3.10. Dry Depletion Option for Modeling PM10 Fugitive dust emissions from mobile equipment and wind erosion are the principal contributors to near-field PM10 impacts at the Upton plant. EPA studies have established the tendency for ground-level, fugitive dust emissions to partially settle out within a short distance of the emission source (EPA 1994a) (EPA 1995a). This deposition includes a portion of the PM10 fraction (Countess 2001). Conservation of mass requires that deposition be accompanied by plume depletion. This is the purpose of the dry depletion option in AERMOD and its predecessor model, ISC3 (EPA 1995b). Dry depletion accounts for the partial settling and deposition of PM10 particles as the dust plume disperses away from the source. The mechanisms for particle deposition and settling include gravity, diffusion, impaction and others. Failure to account for deposition and depletion can lead dispersion models such as AERMOD to significantly over-predict maximum 24-hour PM10 concentrations.
Several studies have cited the tendency of ISC3, the predecessor to AERMOD, to over-
predict maximum 24-hour PM10 concentrations by a factor of four (Cliff 2011, Sullivan 2006, Pace 2005). Moreover, a study by McVehil-Monnett demonstrated AERMOD to be equivalent to, or more conservative than ISC3 in predicting short-term impacts from fugitive dust emissions (MMA 2011). EPA scientist Thompson Pace recently proposed a conceptual model “to approximate the dust removal near the source that is not accounted for in either the current emissions inventories or commonly used regional scale air quality models” (Pace 2005).
EPA guidance emphasizes the need to coordinate the use of deposition modeling options with the appropriate reviewing authority (EPA 2005). For the Upton plant, the AERMOD dry depletion option will not be used in the initial modeling analysis. The model execution times with dry depletion enabled are an order of magnitude longer, making it impractical to use for the entire modeling domain. The dry deposition option
will, however, be considered in a refined analysis of 24-hour PM10 impacts. Modeling only those receptors from the initial modeling analysis that exceed the NAAQS will reduce total execution time with the dry depletion option to a reasonable level. This strategy is influenced by guidance from the New Mexico Air Quality Bureau (New Mexico 2006): “Because of the length of time to run a model with plume depletion, the Bureau recommends only applying plume depletion to receptors that are modeled to be above standards when the model is run without plume depletion.”
Upton Plant Modeling Protocol and Results Page 23
3.10.1. Rationale for Using Dry Depletion in Refined PM10 Analysis The Upton plant meets EPA’s dry deposition criteria of multiple, quantifiable sources of fugitive emissions where a refined modeling analysis is being conducted and deposition is likely to occur (Trinity 2007). While these criteria were originally associated with ISC3, EPA guidance for AERMOD is similar (EPA 2005). Fugitive dust will be the dominant
source of air particulates at the Upton Plant. Historically, short-term modeling of PM10 impacts at receptors close to fugitive dust sources has been shown to over-predict ambient concentrations (Cliffs 2011) (MMA 2011). The results of a study posted by EPA
“suggest that rapid deposition of PM10 particles, and the relatively long residence time of the optical plume associated with small particles (<2µm), may have led to overestimates of airborne particle mass in plumes” (Fitz 2002).
The likelihood of deposition of particles in the PM10 size range is large for this application. In addition to gravity settling, high modeled concentrations at receptors within a few hundred meters of the fugitive emission sources suggest the likelihood of high concentration gradients. These gradients are expected to produce significant diffusion-based settling. The Fugitive Dust Model (FDM) was developed two decades ago to compute concentration and deposition impacts from fugitive dust sources. A key feature of FDM was the improved gradient-transfer deposition algorithm, which is significant for particles in the PM10 size class (EPA 1992).
3.10.2. Precedent for Using Dry Depletion in Refined PM10 Analysis Precedent has been established by state and federal agencies for using the dry depletion option in AERMOD to model short-term impacts from fugitive dust emissions.
For example, a coal lease application in Utah triggered PM10 modeling that included a refined analysis using deposition and plume depletion (BLM 2010). Page 9 of Appendix K in the Alton Coal Lease DEIS states, “deposition was only considered for assessing the final PM10 modeled ambient air impacts.” Page 10 states, “the primary pollutants of concern are fugitive dust.”
The Colorado Department of Public Health and Environment (CDPHE) uses dry
depletion to model PM10 impacts from fugitive dust sources at mining facilities seeking air quality construction permits (Majano 2013). Recent projects for which this option was used include the Lafarge Gypsum Ranch Pit, Oxbow Mining’s Elk Creek Mine, and Bowie Resources’ Bowie N.2 Mine.
Upton Plant Modeling Protocol and Results Page 24
An Environmental Impact Statement (EIS) for the Dewey-Burdock In-Situ Uranium Project in western South Dakota shows the use of AERMOD’s dry depletion option to predict impacts from project-related fugitive dust emissions (NRC 2014). While the EIS
presented 24-hour PM10 results for select model receptors with and without this option, NRC based its impact assessment on the results of the dry depletion modeling run.
EPA cited dry deposition in a study conducted using ISC3 at a Wyoming surface coal mine (EPA 1995b). “In order to appropriately model the particulate emission scenarios, the depletion of dispersed particles from the plume due to gravitational settling and other dry deposition factors were considered.”
A recent modeling analysis was triggered by high fugitive dust impacts in the Salt River
area of Arizona. Maricopa County was reclassified as a serious PM10 nonattainment area on June 10, 1996. The primary sources of particulate pollution in this area are “fugitive dust from construction sites, agricultural fields, unpaved parking lots and roads, disturbed vacant lots and paved roads” (Maricopa 2006). Cited among the “general characteristics that make AERMOD suitable for application in the Salt River Study area” is the claim that “gravitational settling and dry deposition are handled well.”
3.10.3. Input Parameters for Dry Depletion Option AERMOD provides two methods for specifying particle characteristics under the dry depletion option. Method 1, used for this analysis, requires the user to input particle size distribution and particle density. The latter, not to be confused with bulk density, is commonly cited in the literature as 2.65 g/cm3 for soil particles. The Environmental Science Division of Argonne National Lab states, “A typical value of 2.65 g/cm3 has been suggested to characterize the soil particle density of a general mineral soil (Freeze and Cherry 1979). Aluminosilicate clay minerals have particle density variations in the same range” (ANL 2013). A study of fugitive dust from unpaved road surfaces also cites 2.65 g/cm3 for soil particle density (Watson 1996).
The PM10 particle size distribution for fugitive dust was obtained from the modeling protocol for a mine in Arizona (Rosemont 2009). The modelers for the Rosemont project acquired this distribution from AP-42 Section 13.2.4 and applied it to fugitive dust emissions from haul roads. Because Section 13.2.4 applies to aggregate handling and storage piles, another source was consulted to validate the use of this particle size
Upton Plant Modeling Protocol and Results Page 25
distribution for haul road dust. A study by Watson, Chow and Pace referenced in a New Jersey Department of Environmental Protection report (NJDEP 2005) found that 52.3% of the particulate from road and soil dust is less than 10 µm in diameter. Of this particulate 10.7% was found to be smaller than 2.5 µm in diameter and the remaining 41.6% fell between 10 and 2.5 µm. Assuming that fugitive dust particle sizes follow a lognormal distribution (EPA 2013b), these two data points were transformed into a multi-point particle size distribution for comparison to the original particle size distribution. The geometric mass mean diameter for the original distribution is 6.47 µm, while the mean diameter for the lognormal distribution is 5.76 µm. Since these values
are very similar, the Rosemont PM10 size distribution will be used for AERMOD dry
deposition modeling (Table 3-3) of fugitive dust sources. PM10 emissions from stationary combustion sources were modeled as uniform particles 2.5 µm in diameter.
Table 3-3: Assumed PM10 Particle Size Distribution for Dry Depletion Option
Particle Size (µm) Fraction 2.2 0.069 3.17 0.128 6.1 0.385 7.82 0.224 9.32 0.194
Upton Plant Modeling Protocol and Results Page 26
4 APPLICABLE REGULATORY LIMITS FOR CITERIA POLLUTANTS
4.1. Methodology for Evaluation of Compliance with Standards The modeled concentration of the five criteria pollutants will be compared to the
National Ambient Air Quality Standards. Predicted PM10, PM2.5, SO2, and NO2 concentrations will also be compared to the allowable Prevention of Significant Deterioration (PSD) increments for Class II airsheds. The Upton Plant is not subject to a regulatory PSD increment analysis since it is not a major emission source. The PSD increments and modeled concentrations are provided for disclosure purposes only.
4.2. NAAQS and PSD Increments The applicable standards and associated averaging intervals to be used in the modeling analysis are summarized in Table 4-1. Primary standards provide public health protection. Secondary standards provide public welfare protection, including protection against decreased visibility and damage to animals, crops, vegetation, and buildings. PSD increments protect air quality in Class I and Class II areas from significant deterioration.
The purpose of PSD increments is to protect public health and welfare, and to preserve, protect, and enhance the air quality in national parks, national wilderness areas, national monuments, national seashores, and other areas of special national or regional natural, recreational, scenic, or historic value. The goal of this program is to prevent significant deterioration of air quality in areas that meet the NAAQS. Areas in the U.S. have been classified in two categories for the purpose of this program. Class I areas include national wilderness areas, parks and memorial parks of a certain size, and international parks. Class II areas include most of the remaining parts of the country.
Upton Plant Modeling Protocol and Results Page 27
Table 4-1: National Ambient Air Quality Standards (µg/m3) Criteria Averaging Primary Secondary PSD Class I PSD Class II Pollutant Time NAAQS NAAQS Increments Increments Nitrogen Annual 100 100 2.5 25 Dioxide 1-hour 187 ------
PM10 24-hour 150 150 8 30 Annual ------4 17
PM2.5 24-hour 35 35 2 9 Annual 12 15 1 4
SO2 1-hour 200 ------3-hour --- 1,300 25 512 24-hour ------5 91 Annual ------2 20
CO 1-hour 40,000 ------8-hour 10,000 ------
4.3. Presentation of Modeling Results The purpose of the AERMOD dispersion modeling outlined in this protocol is to predict ambient air quality impacts from maximum emission rates at the Upton plant. These predictions will be compared to relevant NAAQS and PSD increments in the Class II area surrounding the Upton site (CALPUFF will be used to model AQRV impacts from the Bear Lodge Project, which includes the Upton Plant and Bull Hill Mine, at sensitive Class II areas within 100 miles of the project, and at the Class I Wind Cave National Park). The final impact analysis will include all the information necessary for this comparison. It will include: (a) maximum impacts for each pollutant in the format of the applicable standard for each averaging period; (b) locations of the model receptors where these impacts are predicted to occur; (c) an emission source location map; (d) a complete list of source parameters; (e) complete modeling input and output files; (f) tabular output of top receptor concentrations for each modeled pollutant and averaging period; and (g) graphic presentations of the modeling results for each pollutant, with isopleth maps based on predicted project impacts.
Upton Plant Modeling Protocol and Results Page 28
4.4. Summary The AERMOD model with on-site meteorological data and maximum project emissions will be used to assess the ambient air quality impact of the criteria pollutants associated with the proposed Upton Plant. The model will be initially run with regulatory default options, with two exceptions. The ARM2 option will be used for modeling NOx
conversion to NO2 in the assessment of 1-hour NO2 impacts. A refined model run will
be conducted for 24-hour PM10 impacts using the dry depletion option in AERMOD.
Emissions of PM10, PM2.5, CO, SO2 and NOx associated with the proposed emission
sources will be modeled. NOx impacts will be converted to NO2 impacts and maximum modeled concentrations of all five pollutants will be compared to NAAQS and (where applicable) PSD Class II increments.
Upton Plant Modeling Protocol and Results Page 29
5 AIR QUALITY RELATED VALUES (AQRV) MODELING METHODOLOGY
5.1. Introduction The purpose of AQRV modeling is to identify and disclose impacts on Class I area resources (i.e., visibility, flora, fauna, etc.) by the projected emissions from a proposed project. AQRVs are resources which may be adversely affected by a change in air quality. The Bull Hill Mine and Upton Plant will be modeled to determine their combined potential AQRV impacts at the nearest federal Class I area, Wind Cave National Park, which lies approximately 80 miles southeast of the BHM site and 70 miles southeast of the Upton site. AQRV impacts will also be modeled at sensitive Class II areas within 100 km of the project sites, including Devils Tower National Monument, Jewell Cave National Monument, Black Elk Wilderness Area and Mount Rushmore National
Memorial (BLM 2014). Species to be modeled are PM10, PM2.5, SO2, SO4, NOx, NHNO3
and NO3. Elemental carbon (EC) and secondary organic aerosol (SOA) will also be enabled in the model, but with zero project-related emissions. This is needed for background visibility calculations and to comply with the latest Federal Land Manager protocol (FLAG 2010).
Figure 5-1 depicts the Bear Lodge Project boundaries and the boundaries of nearby Class I and sensitive Class II areas to be modeled. Based on relative distances and prevailing wind directions, the Bear Lodge Project is expected to have less impact on AQRVs at Badlands National Park than at Wind Cave National Park. Therefore, Badlands National Park is not included in the AQRV modeling (i.e., potential impacts at Badlands would be bounded by modeled impacts at Wind Cave).
This protocol has been developed following applicable portions of the U.S. Environmental Protection Agency (EPA) guidance document: Interagency Workgroup on Air Quality Modeling (IWAQM) Phase 2 Summary Report And Recommendations for Modeling Long Range Transport Impacts, December 1998 (IWAQM 1998). It makes adjustments based on the findings of EPA’s draft Reassessment of the Phase 2 Summary Report published in May 2009 (EPA 2009). It also reflects certain elements of the Federal Land Manager protocol (FLAG 2010) and the Western Regional Air Partnership BART protocol (WRAP 2006).
AQRVs that are generally evaluated for the federal mandatory Class I areas include:
Upton Plant Modeling Protocol and Results Page 30
o Visibility – Visual Plume o Visibility – Regional Haze o Acid Deposition
Visibility can be affected by plume impairment or regional haze. Plume impairment results from a contrast or color difference between a plume and a viewed background such as the sky or a terrain feature. Regional haze occurs at distances where the plume has become evenly dispersed in the atmosphere and is not definable. The primary
causes of regional haze are sulfates and nitrates, which are formed from SO2 and NOX through chemical reactions in the atmosphere. Impacts at distances greater than 30 to 50 km are generally referred to as regional haze. Devils Tower is roughly 30 km from the BHM site, Wind Cave National Park is roughly 130 km from the BHM site, and the distances from Bear Lodge Project sources to all other sensitive Class II areas fall somewhere in between these extremes. Given these distances and the fact that the project will not generate a singular plume of emissions, it is assumed that any visibility impacts at Wind Cave National Park and sensitive Class II areas will be in the form of regional haze.
Upton Plant Modeling Protocol and Results Page 31
Figure 5-1: Bear Lodge Project and Nearest Sensitive Areas
Upton Plant Modeling Protocol and Results Page 32
5.2. Model Selection and Justification Evaluation of the impacts on Air Quality Related Values (AQRVs) from the proposed Bear Lodge Project at Wind Cave and sensitive Class II areas will be conducted using CALPUFF, which is the recommended model for long range transport applications (EPA 2005). CALPUFF is also recommended by the Federal Land Managers (FLM) for AQRV analyses, to simulate visibility and deposition impacts on a Class I area (FLAG 2010). The most recent, EPA-approved version of CALPUFF is Version 5.8.4 – Level 130731. IML Air Science will use a commercial version of CALPUFF 5.8.4 and CALMET 5.8.4 from Lakes Environmental, supplemented with CALPOST Version 6.4 to take advantage of recent visibility post-processing improvements. With its latest release (CALPUFF View Version 7.2.0), Lakes Environmental provides the option to combine CALPOST 6.4 (TRC Version 6.221) with CALPUFF Version 5.8.4 in order to conform to FLAG 2010 post-processing guidelines. The version of CALPOST is not tied to the version of CALPUFF.
CALPUFF is a non-steady-state puff dispersion model that simulates the effects of time- and space-varying meteorological conditions on pollution transport, transformation, and removal. CALPUFF can be applied for long-range transport and for complex terrain. The CALPUFF model calculates the change in light extinction caused by a source (or group of sources) as part of the regional haze calculations. The EPA has proposed the use of CALPUFF for applications involving long-range transport, which is typically defined as transport over distances beyond 50 km (IWAQM 1998).
The CALPUFF model accounts for chemical transformations that occur during plume
transport using algorithms to calculate the conversion of SO2 to sulfates and NOx to nitrates. The IWAQM Phase 2 report (IWAQM 1998) recommended the use of the MESOPUFF II scheme, which requires the user to select additional species to be modeled, e.g., sulfates (SO4), nitrates (NO3) and nitric acid (HNO3). It also requires the input of background ozone and ammonia concentrations. Although the CALPUFF model provides default values for background concentrations, values specific to the Class I area being modeled are recommended given the sensitivity of the model to these parameters (see Section 5.5.1 below). For visibility calculations, site-specific relative humidity data are also recommended in the post processing step. Monthly average relative humidity values from the BHM and Upton sites will be used for this AQRV modeling exercise.
Upton Plant Modeling Protocol and Results Page 33
The CALPUFF Modeling System includes three main components: CALMET, CALPUFF, CALPOST, and a large set of preprocessing and postprocessing programs designed to interface the model with standard, routinely available meteorological and geophysical datasets.
5.2.1. CALMET CALMET is a meteorological model that develops hourly wind and temperature fields on a three-dimensional gridded modeling domain. Associated two-dimensional fields such as mixing heights, surface characteristics, and dispersion properties are also included in the file produced by CALMET.
5.2.2. CALPUFF CALPUFF is a transport and dispersion model that advects “puffs” of material emitted from modeled sources, simulating dispersion and transformation processes along the way. In doing so it typically uses the fields generated by CALMET, or as an option, it may use simpler non-gridded meteorological fields explicitly incorporated in the resulting distribution of puffs throughout a simulation period. In this case it will use CALMET- generated meteorological data. The primary output files from CALPUFF contain either hourly concentrations or hourly deposition fluxes evaluated at selected receptor locations.
5.2.3. CALPOST CALPOST is used to process these files, producing tabulations that summarize the results of the simulation (concentrations at each receptor, for example). When performing visibility related modeling, CALPOST uses concentrations from CALPUFF to compute extinction coefficients and related measures of visibility, reporting these for selected averaging times and locations.
5.3. Meteorological, Terrain and Land Use Data Preprocessed data will be acquired for incorporation into CALMET. This will include three dimensional mesoscale data (MM5), hourly surface observations from Bear Lodge Project weather stations, upper air data from the National Weather Service (NWS) station at Rapid City, precipitation data, terrain elevations, and land use classifications.
Upton Plant Modeling Protocol and Results Page 34
5.3.1. Time Period According to 40 CFR Part 51 Appendix W, the length of the modeled meteorological period should be long enough to ensure that the all meteorological conditions are adequately represented in the model results. EPA recommends that consecutive years from the most recent, readily available 5-year period are preferred, but when mesoscale meteorological data are used (i.e., MM5) in conjunction with on-site surface data, a minimum of one year of modeling will generally meet regulatory requirements (Earth Tech 2001). The Bear Lodge Project has collected on-site, hourly meteorological data for slightly less than two years at the time of this report. Therefore the modeling analysis will be conducted using one year (August 2012 through July 2013) of on-site data coupled with concurrent mesoscale data and upper air data.
5.3.2. Prognostic Meteorological Data The CALMET/CALPUFF modeling system currently includes the capability to incorporate 3-dimensional prognostic meteorological data from a mesoscale wind field model (MM5) into the processing of meteorological data through the CALMET Diagnostic Wind Model (DWM). This is most commonly accomplished by using the MM5 data as the initial guess for the wind field in CALMET. The MM5 data used in this modeling effort will span the entire 200 km by 200 km modeling domain, with 12-km horizontal resolution and 18 vertical layers. This data set will be obtained from Lakes Environmental.
5.3.3. CALMET Diagnostic Meteorological Data EPA recommends using a “hybrid” CALMET, to include MM5 and weather station data (EPA 2009). EPA recommends against the use of the “no-observation” methods for CALMET (NOOBS=1, 2). The CALMET NOOBS mode is less conservative; therefore meteorological observations will be blended with the MM5 data as input to the CALMET/CALPUFF modeling system. These will include one year of hourly on-site meteorological data from the BHM and Upton sites and one year of upper air data from Rapid City, the only upper air station in the region. Precipitation records will be supplied by the on-site data sets. Traditionally, the FLMs have recommended a CALMET grid resolution of approximately 4 km. There is concern that the increased structural detail in the horizontal wind fields resulting from application of CALMET at higher grid resolutions may lead to spurious effects on plume dispersion which may not be obvious (WRAP 2006). EPA studies show little, if any, sensitivity to the increase in grid resolution within CALMET relative to the MM5 grid resolution (EPA 2009). Therefore, a 4 km grid resolution will be used for CALMET.
Upton Plant Modeling Protocol and Results Page 35
5.3.4. CALMET Approach CALMET uses a two-step approach to calculate wind fields. In the first step, an initial guess field is adjusted for slope flows and terrain blocking effects, for example, to produce a step 1 wind field. In the second step, an objective analysis is performed to introduce observational data into the Step 1 wind field. EPA recommends elimination of CALMET diagnostic adjustments to first-guess wind field (EPA 2009). EPA recommends continuation of incorporation of surface observations for radii of influence (RMAX1, RMAX2, RMAX3, R1, R2, R3) set to minimal values to preserve the integrity of prognostic meteorological data used as the first-guess wind field. These recommendations will be followed in modeling the Bear Lodge Project.
5.3.5. CALMET Parameter Settings The maximum mixing height (ZIMAX) has an EPA default value of 3000 m AGL. All the other parameters are set on a case by case basis taking the terrain surrounding the observation stations into consideration.
5.3.6. Terrain Data Gridded terrain elevations for the modeling domain are derived from 3 arc-second digital elevation models (DEMs) produced by the United States Geological Survey (USGS). The files cover 1-degree by 1-degree blocks of latitude and longitude. The elevations are in meters relative to mean sea level and have a resolution of about 90 meters. These data will be processed to generate 4 km average terrain heights that will be input into CALMET.
5.3.7. Land Use Data Surface properties such as albedo, Bowen ratio, roughness length and leaf area index are computed proportionally to the fractional land use. The land use data are based on the Composite Theme Grid format (CTG) using Level I USGS land use categories. The 4 km land use grid will be mapped into the 14 primary CALMET land use categories.
5.3.8. CALMET Switch Settings Most of the default switch settings for CALMET will be used. Several parameters do not have default values. Table 5-1 lists some of these key parameter settings as proposed, and as implemented in the WRAP Protocol (WRAP 2006).
Upton Plant Modeling Protocol and Results Page 36
Table 5-1: CALMET Switch Settings
Parameter WRAP Setting Proposed Setting RMAX1 50 KM 60 KM RMAX2 100 KM 100 KM RMAX3 100 KM 100 KM R1 100 KM 30 KM R2 200 KM 50 KM ZIMAX 4500 m AGL 3000 m AGL TERRAD 10 KM 16 KM
5.4. Modeling Domain, Sources and Receptors The CALPUFF modeling domain will be selected to include the project area, Wind Cave National Park, and four sensitive Class II areas. In order to adequately characterize potential AQRV impacts to Wind Cave National Park and sensitive Class II areas, the modeling domain will extend 100 km in all directions from a point in the southern Black Hills that is central to both project sources and these sensitive areas (200 km by 200 km grid). IWAQM recommends modeling 50 km beyond the relevant Class I boundary to provide meteorological continuity and to account for any potential wind circulation. For the Bear Lodge Project, the proposed buffer width meets this criterion.
Receptor locations and elevations for the Wind Cave National Park Class I area will be obtained from the National Park Service database in order to generate visibility data compatible with and comparable to previous modeling exercises. Additional model receptors will be placed around the boundaries of the following sensitive Class II areas:
1. Devils Tower National Monument 2. Mount Rushmore National Memorial 3. Black Elk Wilderness Area 4. Jewell Cave National Monument A total of 261 model receptors will be used. 192 receptor locations will be obtained for Wind Cave. An additional 69 receptors will be placed around the boundaries of the four sensitive Class II areas listed above. Figure 5-2 shows the proposed AQRV modeling domain, along with source and receptor locations. Figure 5-3 presents a closer view of
Upton Plant Modeling Protocol and Results Page 37
sources and receptors. Modeled emission sources and emission rates will be identical to those configured in the AERMOD model. Bull Hill Mine sources are displayed in Figure 5-4, with Upton Plant sources in Figure 5-5.
Upton Plant Modeling Protocol and Results Page 38
Figure 5-2: Bear Lodge Project CALPUFF Modeling Domain
Upton Plant Modeling Protocol and Results Page 39
Figure 5-3. CALPUFF Model Receptors and Emission Sources
Upton Plant Modeling Protocol and Results Page 40
Figure 5-4. Bull Hill Mine Emission Sources
Upton Plant Modeling Protocol and Results Page 41
Figure 5-5. Upton Plant Emission Sources
Upton Plant Modeling Protocol and Results Page 42
5.5. CALPUFF Model Inputs 5.5.1. Background Concentrations CALPUFF requires ozone and ammonia background concentrations in order to characterize atmospheric chemistry. These species influence the rates of formation of sulfates and nitrates, aerosols that affect visibility.
This modeling exercise will incorporate a uniform monthly background ozone concentration of 60 ppb. This is conservative since monthly ozone concentrations at Wind Cave National Park averaged 43 ppb from 2009 through 2011, with a maximum monthly average of 50 ppb (EPA 2013c).
For ammonia background, IWAQM recommends 1 ppb for forested lands, 10 ppb for grasslands, and 0.5 ppb for arid lands (IWAQM 1998). The relevant ammonia background is at Devils Tower National Monument, Wind Cave National Park and the Black Hills National Forest. Since the predominant land use at these locations is forest, a value of 1 ppb will be used in the model.
5.5.2. Chemistry Modeling The MESOPUFF II pseudo-first-order chemical reaction mechanism (MCHEM=1) will be used for the conversion of SO2 to sulfate (SO4) and NOx to nitrate (NO3) as recommended by EPA (WRAP 2006). MESOPUFF II is a 5-species scheme in which all emissions of nitrogen oxides are simply input as NOx. In the MESOPUFF II scheme, the conversion of SO2 to sulfates and NOx to nitrates is dependent on relative humidity (RH), with an enhanced conversion rate at high RH. This modeling exercise will therefore incorporate an adjustment factor for RH. Aqueous phase oxidation is currently not modeled, leading to an underestimation of sulfate formation in clouds or fog.
5.5.3. Particle Size Distribution The dominant pollutant emitted from ground-level sources at the Bear Lodge Project will be fugitive PM10. Calpuff models the atmospheric dispersion and attempts to model the settling of particulate matter based on an input particle size distribution. This modeling exercise will use a PM10 size distribution for haul road dust taken from the Rosemont Copper Project protocol (Rosemont 2009) and based on AP-42 Section 13.2.4 (EPA 1995c). Table 5-2 lists the corresponding size distribution.
Upton Plant Modeling Protocol and Results Page 43
Table 5-2: Fugitive PM10 Particle Size Distribution
Particle Size (µm) Fraction 2.2 0.069 3.17 0.128 6.1 0.385 7.82 0.224 9.32 0.194
All tailpipe particulate emissions will be modeled as PM2.5.
5.5.4. CALPUFF Switch Settings Most of the default switch settings for CALPUFF will be used, with the exception of the number of pollutants emitted and the number of chemical species modeled. Table 5-3 lists the default values and proposed values for some of the key parameter settings.
The increase in number of species emitted accounts for NOx, SO2, PM10 and PM2.5 emissions.
Upton Plant Modeling Protocol and Results Page 44
Table 5-3: CALPUFF Switch Settings Parameter Description Default Value Proposed Notes Value Group 1 – General Options
NSPEC Number of chemical 5 9 species NSE Number of species 3 4 emitted METFM Meteorological data 1 1 1 = CALMET file format PGTIME Pasquill-Gifford 60 60 Minutes (PG) MGAUSS Near-field vertical 1 1 1 = Gaussian distribution MCTADJ Terrain adjustments 3 3 3 = Partial plume path to plume path adjustment MCHEM Chemical 1 1 1 = MESOPUFF II mechanism chemistry MDISP Method for 3 3 3 = PG for rural and dispersion McElroy-Pooler (MP) coefficients for urban MREG Regulatory default 1 1 1 = Technical options checks must conform to EPA Long Range Transport guidance SYTDEP Equations used to 550 550 Puff size (m) beyond determine sigma-y which equations and (Heffter) are used to -z determine sigma y and z MHFTSZ Heffter equation for 0 0 0 = Not use Heffter sigma z
5.6. CALPUFF Model Outputs, Calculations and Evaluation Methods 5.6.1. CALPOST and POSTUTIL The CALPUFF results will be post-processed using the CALPOST and POSTUTIL processors. POSTUTIL is a post processing program used to process the concentrations generated by CALPUFF. POSTUTIL occurs prior to the visibility processing in CALPOST and allows the user to sum the contributions of sources from
Upton Plant Modeling Protocol and Results Page 45
different CALPUFF simulations into a total concentration file. Monthly RH adjustment factors will be applied directly to the background and modeled sulfate and nitrate concentrations in CALPOST.
5.6.2. Visibility Impact Determination The general theory for performing visibility calculations with the CALPUFF modeling system is described in the Interagency Workgroup on Air Quality Modeling Phase 2 Summary Report and Recommendations for Modeling Long Range Transport Impacts (IWAQM 1998). The theory is also summarized in Section 5.6.4 below. Change of light extinction is the preferred metric for assessing visibility impairment. Light extinction can in turn be converted to a haze index, measured in deciviews (dv). Visibility impact on a Class I area is considered significant if the source’s contribution to visibility impairment, modeled as the 98th percentile of the daily (24-hour) changes in deciviews, is equal to or greater than the contribution threshold of 0.5 dv (FLAG 2010). Stated differently, a source can be reasonably anticipated to cause or contribute to an impairment of visibility if the 98th percentile of the distribution of modeled changes in light extinction is greater than 0.5 dv. Changes in visibility at the above-referenced Class I and sensitive Class II areas will be calculated from the Bear Lodge Project model outputs and reported in terms of the 98th percentile change in dv at each modeled receptor, as well as the total light extinction at each receptor.
5.6.3. Comparison to Existing AQRV Status Assessing some Air Quality Related Values (e.g., crop injury, or visibility effects) is fundamentally tied to knowing the current stress being exerted on the system. This is reflected in the current background visibility. Assessing the response of a resource is related to the cumulative effects of all the current existing stresses (IWAQM 1998). The evaluation of the Bear Lodge modeling results will therefore consider the current visual resource and visibility impairment at Wind Cave National Park. Studies conducted by the National Park Service and the Western Regional Air Partnership (WRAP) will provide references for current conditions.
5.6.4. Calculation of Light Extinctions The calculation of regional visibility impacts in CALPUFF takes into account the scattering of light caused by several particulate matter (PM) constituents in the atmosphere. This scattering of light is referred to as extinction. The PM constituents that are accounted for in the visibility calculations include ammonium sulfate, ammonium nitrate, organic carbon, elemental carbon, soil, and coarse and fine PM. The CALPUFF model calculates the light extinction attributable to a source's emissions and compares it
Upton Plant Modeling Protocol and Results Page 46
to the extinction caused by the background constituents to estimate a change in extinction.
The extinction caused by a source's emissions is affected by several factors. One such factor is the formation of light scattering constituents by chemical transformation during
plume transport, e.g., conversion of SO2 to sulfates and NOx to nitrates. These chemical transformations are dependent on the level of available gaseous ammonia and ozone in the atmosphere, i.e., the higher the ammonia and ozone concentration in the air, the greater the transformation, and hence the greater the light extinction. Since sulfates and nitrates are hygroscopic in nature, the light extinction caused by these constituents is also affected by relative humidity (RH). The other PM constituents are considered to be non-hygroscopic. The visibility analysis will be conducted using monthly average relative humidity adjustment factors, or f(RH) values.
The CALPOST postprocessor will be used for the calculation of the impact from the modeled source’s primary and secondary particulate matter concentrations on light extinction. The formula that is used is the existing IMPROVE/EPA formula, which is applied to determine a change in light extinction due to increases in the particulate matter component concentrations. Using the notation of CALPOST, the formula is the following:
Bext = 2.2 x fS(RH) x [Small Sulfates] + 4.8 x fL(RH) x [Large Sulfate] + 2.4 x fS(RH) x [Small Nitrates] + 5.1 x fL(RH) x [Large Nitrates] + 2.8 x [Small Organic Mass] + 6.1 x [Large Organic Mass] + 10 x [Elemental Carbon] + 1 x [Fine Soil] + 0.6 x [Coarse Mass] + 1.7 x fSS(RH) x [Sea Salt] + [Rayleigh Scattering] + 0.33 x [NO2 (ppb)]
3 The concentrations, in square brackets, are in μg/m and bext is in units of inverse megameters or Mm-1. The Rayleigh scattering term will be set to the value of 10 Mm-1, the default value recommended in EPA guidance for tracking reasonable progress (WRAP 2006).
Upton Plant Modeling Protocol and Results Page 47
Each hour’s source-caused extinction is calculated by first using the hygroscopic components of the source caused concentrations, due to ammonium sulfate and nitrate, and monthly f(RH) values specific to Wind Cave National Park. The contribution to the total source-caused extinction from ammonium sulfate and nitrate is then added to the other, non-hygroscopic components of the particulate concentration to yield the total hourly source caused extinction. The terms fS(RH), fL(RH) and fSS(RH) are relative humidity adjustment factors for small particles, large particles and sea salts respectively. These values will be taken from the Federal Land Managers Air Quality Related Values Workgroup Phase 1 Report Revised Draft Table V.1-2, V.1-3 and V1.-4 (FLAG 2008) which list f(RH) values for each Class I area.
5.6.5. Deposition Analysis Atmospheric deposition includes wet and dry fluxes of the pollutants modeled (g/m2/sec), represented as sulfur and nitrogen calculated in pollutant-specific runs of CALPOST. Modeled fluxes are for the modeled species and do not directly represent the mass flux of either sulfur or nitrogen. Adjustments are therefore made for the ratio of
molecular weight of S and N vs. the molecular weight of the species modeled (SO2,
SO4, NOx, HNO3, NO3). The deposition flux of sulfur includes contributions from any modeled sulfur compounds. The deposition flux of nitrogen includes contributions from any modeled nitrogen compounds.
The CALPUFF output files will contain the wet and dry deposition fluxes of both primary and secondary species. The wet and dry fluxes must be added to obtain the total flux of each species, at each receptor, each hour. The POSTUTIL processor will be configured to sum the wet and dry fluxes, and to compute the total sulfur and nitrogen contributed by the modeled species for subsequent CALPOST processing.
5.6.6. CALPOST Switch Settings Table 5-4 lists default and proposed values for key parameters for CALPOST. The maximum relative humidity will be lowered from 98% to 95% based on recent FLM guidance (FLAG 2008). The default value for LVPMC is “True,” indicating that coarse particulate matter (PM10-2.5) is included in the visibility model. CALPOST will also be run with LVPMC set to “False.” Both sets of results will be presented.
There is evidence and precedent that supports excluding ground-level, fugitive PM10 emissions from the assessment of project impacts on far-field visibility conditions. A recent EIS for a gas development in southern Wyoming discussed the exclusion of
Upton Plant Modeling Protocol and Results Page 48
fugitive PM10 emissions from visibility assessment (TRC 2006). Appendix F to the EIS
states, “In post-processing the PM10 impacts at all far-field receptor locations, the PM10 impacts from Project alternative traffic emissions (production and construction) were not included in the total estimated impacts, only the PM2.5 impacts were considered. This assumption was based on supporting documentation from the Western Regional Air Partnership (WRAP) analyses of mechanically generated fugitive dust emissions that
suggest that particles larger than PM2.5 tend to deposit out rapidly near the emissions source and do not transport over long distances (Countess et al. 2001). This phenomenon is not modeled adequately in CALPUFF; therefore, to avoid overestimates
of PM10 impacts at far-field locations, these sources were not considered in the total
modeled impacts. However, the total PM10 impacts from traffic emissions were included in all in-field concentration estimates.”
5.7. Presentation of Modeling Results The purpose of the AQRV modeling outlined in this protocol is to disclose impacts from emissions at the Bear Lodge Project to Air Quality Related Values (AQRV) at Wind Cave National Park and sensitive Class II areas in the project region. The final impact analysis will present all predicted impacts from the project, and compare these predictions to background conditions. The visibility impact analysis will include the 98th percentile of the 24-hour changes in haze index (deciviews). It will also include a table comparing modeled deposition rates to monitored conditions, significance thresholds and critical loads.
Upton Plant Modeling Protocol and Results Page 49
Table 5-4: CALPOST Switch Settings Parameter Description Default Value Proposed Notes Value Group 1 ASPEC Species to process No Default VISIB Visibility processing
Group 2 MFRH Particle growth 4 4 4 = IMPROVE (2006) f(RH) curve f(RH) tabulations for sea salt and for sulfate and nitrate RHMAX Maximum relative 98 95 FLAG (2008) guidance humidity (%) in growth curve Modeled Species
LVSO4 Include sulfate T T LVNO3 Include nitrate T T LVNO2 Include nitrogen T T dioxide absorption LVOC Include organic T T carbon LVPMC Include coarse T T particulates LVPMF Include fine T T particulates LVEC Include elemental T T carbon Extinction Efficiency
EEPMC Particulate matter 0.6 0.6 coarse EEPMF Particulate matter 1.0 1.0 fine EEPMCBK Particulate matter 0.6 0.6 Background particulate coarse background species
EESO4 Ammonium sulfate 3.0 3.0 EENO3 Ammonium nitrate 3.0 3.0 EEOC Organic carbon 4.0 4.0 EESOIL Soil 1.0 1.0 EEEC Elemental carbon 10.0 10.0
Upton Plant Modeling Protocol and Results Page 50
6 AERMOD MODELING RESULTS AND ANALYSIS
6.1. Introduction The stationary and fugitive emission sources at the Upton Plant will produce particulate
matter smaller than ten microns in size (PM10) and particulate matter smaller than 2.5
microns in size (PM2.5). Stationary and mobile sources will emit PM10, PM2.5, carbon monoxide (CO), sulfur dioxide (SO2) and oxides of nitrogen (NOx). For the NO2 analysis
in AERMOD, it is assumed that a portion of the NOx emissions will be converted to NO2,
using variations of the ambient ratio method. Thus, five criteria pollutants (PM10, PM2.5,
CO, SO2 and NO2) were analyzed for compliance with the NAAQS using the AERMOD
dispersion modeling software. For disclosure purposes four of these pollutants, PM10,
PM2.5, SO2 and NO2 were further analyzed for comparison to the allowable PSD increments in Class II areas. For each scenario, emissions from all 25 emission sources identified and quantified in the Upton Plant emissions inventory (Figures 2-1 and 2-2), were modeled. Each model run, with the exception of a “dry depletion” run discussed in Section 6.2 below, produced maximum pollutant concentrations and related statistics at all 3,583 receptors in the 100-km by 100-km modeling domain (Figure 3-2).
Table 6-1 summarizes the results of the AERMOD model runs for all pollutants and relevant averaging intervals. All results are presented in the format of the applicable NAAQS, referred to as design values. Predicted total ambient concentrations are computed as the sum of the design-value project impacts and the background concentrations. For each pollutant, this sum is given as a percentage of the NAAQS.
Separate modeling results are shown in Table 6-1 for PM10 based on two scenarios:
1. Initial modeling or screening analysis (regulatory default settings) 2. Refined analysis for 50 highest receptors (dry depletion option)
The second scenario included one receptor that (with background added) exceeded the 24-hour NAAQS in the initial modeling.
Sections 6.2 through 6.6 discuss results in detail for each of the five criteria pollutants listed in Table 6-1. All receptors were predicted to be in compliance with all NAAQS as
reflected in Table 6-1, with the exception of PM10. Initial modeling showed 24-hour PM10 impacts plus background, exceeding the NAAQS at a single receptor situated on the project boundary close to project fugitive dust emission sources (i.e., roads and active
Upton Plant Modeling Protocol and Results Page 51
areas). Refined modeling with the dry depletion option reduced this predicted exceedance from 116% to 105% of the NAAQS. Initial modeling also showed one receptor approaching the annual Wyoming standard for PM10. Refined modeling with the dry depletion option reduced the predicted concentration at this receptor from 97% to 87% of the Wyoming standard. NO2 modeling showed a maximum 1-hour NO2 impact plus background, at 88% of the NAAQS. The highest receptor is located on the project boundary, a few hundred meters from the Tailings Storage Facility.
Table 6-1: Summary of Predicted Pollutant Concentrations (AERMOD) Averaging Ambient Back- Total Ambient NAAQS % of Interval and Impact ground Concentration Limit NAAQS Receptor (UTM Pollutant Statistic (µg/m3) (µg/m3) (µg/m3) (µg/m3) Limit Easting, Northing)
PM10 Initial Run Annual Average 33.3 15 48.34 -- 526962, 4883832 (No Dry 2nd High 24-Hr 133.6 40.0 173.6 150 115.7% 526962, 4883832 Depletion) Maxim um
PM10 Final Run Annual Average 28.6 15 43.61 -- (Top 179 2nd High 24-Hr Receptors With 117.1 40.0 157.1 150 104.8% 526962, 4883832 Maxim um Dry Depletion) Annual Average 4.8 3.4 8.2 12 68.3% 526962, 4883832 PM2.5 24-Hr 8th High 13.4 8.0 21.4 35 61.3% 526962, 4883832 Annual Average 16.0 6.0 22.0 100 22.0% 526962, 4883832
th NO2 98 Percentile of 141.9 21.0 162.9 187 87.1% 527062, 4883832 Daily 1-Hr Highs Annual Average 4.8 1.3 6.1 -- 527262, 4883832 24-Hr High 25.7 16.3 42.0 -- 527262, 4883832
SO2 3-Hr High 64.6 124.7 189.3 1300 14.6% 527262, 4883832 99th Percentile of 111.0 43.2 154.2 200 77.1% 527262, 4883832 Daily 1-Hr Highs 8-Hr High 611.9 378.0 989.9 10000 9.9% 527262, 4883832 CO 1-Hr High 2905.0 680.0 3585.0 40000 9.0% 527850, 4884480
Upton Plant Modeling Protocol and Results Page 52
Table 6-2 compares model predictions to PSD Class II increments. Class I increments were not evaluated since there are no Class I areas within 50 kilometers of the Bull Hill Mine. Although the Bull Hill Mine is not a major source and does not meet the criteria for PSD regulation, the results in Table 6-2 are presented for disclosure purposes.
Comparisons between modeled concentrations and PSD increments rely on EPA’s definition: for any period other than annual, the allowable increment for a given pollutant may be exceeded during one such period per year at any one location (EPA 1986).
Therefore, the relevant 24-hour PM10 and PM2.5 concentrations are the highest yearly 2nd high values at any one receptor. There are no PSD increments associated with the
1-hour NO2 and 1-hour SO2 concentrations. For the 3-hour and 24-hour SO2 increments, the highest concentration at any receptor serves as a surrogate (since it will always be higher than the highest yearly 2nd high).
Table 6-2: Summary of PSD Increment Comparisons (AERMOD)
Allowable % of Class Averaging Interval Class II Class II PSD II PSD Pollutant and Statistic Impact Increment Increment Annual Average 33.3 17 196% PM10 Initial Run (No Dry Highest 24-hr 2nd Depletion) 133.6 30 445% High
PM10 Final Run (Top 20 Annual Average 28.6 17 168% Receptors With Dry Highest 24-hr 2nd Depletion) 117.1 30 390% High Annual Average 4.8 4 120% PM2.5 Highest 24-hr 2nd 17.6 9 196% High
Annual Average 16.0 25 64% NO th 2 98 Percentile of 141.9 -- -- Daily 1-Hr Highs Annual Average 4.8 20 24%
24-Hr 25.7 91 28% SO2 3-Hr 64.6 512 13% 99th Percentile of 158.9 -- -- Daily 1-Hr Highs
8-Hr High 611.9 -- -- CO 1-Hr High 2905.0 -- --
Upton Plant Modeling Protocol and Results Page 53
It can be seen from Table 6-2 that all potential Class II impacts from NO2 and SO2 emissions fell below the associated PSD increment. The highest second-high, 24-hour
PM10 concentration (modeled with dry depletion enabled) was 390% of the PSD
increment. The second high annual average PM10 concentration (modeled with dry depletion enabled) was 168% of the PSD increment. The highest second-high, 24-hour
PM2.5 concentration was 196% of the PSD increment and the second-high annual average was 120% of the PSD increment. For both pollutants, receptors with predicted values above the increment were confined to locations on or near the Upton Plant project boundary or the public access road. These receptors are in close proximity to project-related sources of fugitive dust. It is not unusual for ground-level sources of fugitive dust to cause PM impacts in excess of the PSD Class II increments at monitors located near the sources. Monitored values at coal mine boundaries in the Powder
River Basin routinely exceed the PM10 and PM2.5 increments. Like the Bear Lodge Project and its Upton Plant facility, however, these mines are not subject to PSD regulation.
6.2. PM10 Modeling Analysis
The primary source of PM10 emissions will be fugitive dust generated by traffic on unpaved roads, road maintenance, drilling and construction activities, and wind erosion
on disturbed areas. A small fraction of the total PM10 emissions will be generated by fuel
combustion. Nearly all of these combustion emissions will also qualify as PM2.5 (particles with aerodynamic diameter less than 2.5 microns). Accordingly, the outcome
of this PM10 modeling study is driven by ground-level sources of fugitive dust.
The maximum yearly PM10 emissions from the Upton Plant were modeled for potential impacts on ambient air quality at all receptors in the modeling domain. The model produced maximum receptor concentrations for any calendar day (24-hour average) and for the entire modeling period (annual average). In order to characterize maximum, short-term impacts, the modeling period spanned one year of hourly meteorological conditions. Section 6.2.1 discusses initial modeling results for the 3,783 receptors located along the project boundary and between the boundary and the edge of the modeling domain. Section 6.2.2 discusses the issue of model over-prediction.
Upton Plant Modeling Protocol and Results Page 54
6.2.1. Initial PM10 Modeling Results
PM10 results from the initial AERMOD run are presented below. Table 6-3 lists the top 20 receptors ranked by annual average concentrations. Table 6-4 lists the top 50 receptors ranked by 2nd high, 24-hour concentrations (consistent with the NAAQS format). Figure 6-1 is an isopleth, or contour plot of the predicted annual concentrations attributable solely to the Upton Plant. Because the significant impacts (greater than 1 µg/m3) are confined to receptors within approximately 5 km of the project center, Figure 6-1 zooms in on this area. Figure 6-2 is a close-up isopleth map of the predicted maximum 24-hour concentrations attributable to the Upton Plant.
Table 6-3 shows the top 20 modeled annual average PM10 concentrations. AERMOD predicted compliance with the annual NAAQS at all receptors. Seven of the 3,783 receptors initially exceeded the annual, Class II PSD increment of 17 µg/m3. Table 6-4 shows the top 50 receptors. One receptor, with a background of 40 µg/m3 added to modeled impacts, was above the 24-hour NAAQS of 150 µg/m3. Figure 6-2 shows the highest receptor concentrations are confined to a small area in the corner of the project boundary. The single receptor above exceeding the NAAQS (with the background of 40 µg/m3 added) falls on the corner of the project boundary (Figure 6-3).
It is not unusual for dispersion modeling of ground-level fugitive dust sources to predict high short-term values near the emission sources. The Buffalo RMP cited modeled
exceedances of 24-hour PM10 and PM2.5 standards in Wyoming’s PRB (BLM 2013). Phase I of the PRB Coal Review states that for near-field receptors, the predicted 24-
hour PM10 and PM2.5 concentrations show localized exceedences of the NAAQS for the base year of 2004, as well as for future years (BLM 2005). An updated study relates these high values to coal bed methane and coal mining activities, stating these exceedances are limited to small individual receptor areas in the near-field (BLM 2009). Table 6-5 shows that a refined analysis of the highest receptors using AERMOD’s dry depletion option (see Section 6.2.3), lowered this non-compliant receptor from 173.57 µg/m3 to 157.14 µg/m3.
Upton Plant Modeling Protocol and Results Page 55
Table 6-3: Top 20 Receptors, Annual Average PM10 Impacts (Initial Run)
UTM UTM Maximum Modeled PSD Class II Increment Easting Northing Concentration (µg/m3) (µg/m3) 526962 4883832 33.34 17 526962 4883913 24.10 17 526720 4884209 18.45 17 526712 4884244 17.62 17 526972 4882614 17.51 17 526972 4882613 17.48 17 526820 4884211 17.20 17 526321 4884221 16.96 17 526221 4884218 16.47 17 526812 4884244 16.46 17 526675 4884333 15.89 17 526421 4884223 15.85 17 526620 4884208 15.79 17 526977 4882513 15.78 17 526612 4884244 15.78 17 526974 4882514 15.45 17 526121 4884216 15.30 17 526318 4884321 15.15 17 526520 4884206 15.03 17 526775 4884333 14.90 17
Upton Plant Modeling Protocol and Results Page 56
nd Table 6-4: Top 50 Receptors, 24-Hr 2 High PM10 Concentrations (Initial Run)
Maximum Modeled Maximum UTM UTM NAAQS Concentration Concentration Concentration with 3 Easting Northing 3 3 (µg/m ) (µg/m ) Background (µg/m ) 526962 4883832 133.57 173.57 150 526962 4883913 107.83 147.83 150 527294 4882218 96.30 136.30 150 527194 4882217 96.22 136.22 150 526820 4884211 88.99 128.99 150 526812 4884244 86.55 126.55 150 526720 4884209 85.43 125.43 150 527094 4882215 83.50 123.50 150 526712 4884244 82.40 122.40 150 526775 4884333 80.03 120.03 150 526675 4884333 74.29 114.29 150 526612 4884244 72.89 112.89 150 526620 4884208 71.84 111.84 150 527196 4882117 71.64 111.64 150 526321 4884221 71.55 111.55 150 527389 4882333 71.12 111.12 150 527394 4882220 70.00 110.00 150 527393 4882233 69.83 109.83 150 526520 4884226 69.23 109.23 150 526520 4884206 68.97 108.97 150 527296 4882118 68.89 108.89 150 526221 4884218 68.61 108.61 150 526318 4884321 67.00 107.00 150 527385 4882433 65.99 105.99 150 526517 4884326 65.98 105.98 150 526421 4884223 65.23 105.23 150 526737 4884423 64.19 104.19 150 526657 4884469 63.70 103.70 150 526168 4884317 63.64 103.64 150 526216 4884418 63.61 103.61 150 526972 4882614 63.29 103.29 150 526121 4884216 63.24 103.24 150 526972 4882613 63.18 103.18 150 526418 4884323 62.63 102.63 150 526512 4884526 62.53 102.53 150 526962 4884013 61.29 101.29 150 526416 4884423 60.82 100.82 150 526316 4884421 60.80 100.80 150 527198 4882017 60.62 100.62 150 526510 4884626 60.59 100.59 150 526500 4884500 60.49 100.49 150 526213 4884518 60.37 100.37 150 527096 4882115 60.26 100.26 150 527297 4882018 59.83 99.83 150 526515 4884426 59.76 99.76 150 526116 4884416 59.60 99.60 150 526113 4884516 59.20 99.20 150 526068 4884314 57.38 97.38 150 526413 4884523 57.18 97.18 150 526837 4884423 57.13 97.13 150
Upton Plant Modeling Protocol and Results Page 57
Figure 6-1. Initial Annual Average PM10 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 58
nd Figure 6-2. Initial 2 High 24-Hr Average PM10 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 59
6.2.2. PM10 Modeling Over-Prediction Problems
The PM10 modeling results in Section 6.2.1 may reflect AERMOD’s tendency to over- predict the transportability and the resultant air quality impacts of fugitive dust emissions (Cliffs 2011). Among several possible causes, predicted concentrations do not account for particle electrostatic agglomeration, enhanced gravitational settling and deposition near the point of release (AECOM 2012).
This tendency was exposed in ISCST3, the regulatory model that preceded AERMOD. Although AERMOD improved on many of ISCST3’s features, these improvements were confined primarily to stationary sources and buoyant plumes. Even with the improvements to AERMOD, the problem of over-predicting 24-hour PM10 impacts from fugitive dust persists (Sullivan 2006). For low-level emission plumes, AERMOD has not been evaluated extensively by EPA for performance against measured data. In 2011 MMA conducted a modeling analysis to determine whether AERMOD would yield significant improvements over the ISC3 Short Term model in the prediction of short-term particulate concentrations for surface mining operations. The study found that AERMOD
still over-predicts short-term PM10 concentrations, and even exceeds the predictions of ISCST3 at model receptors positioned from 100 to 500 meters from the sources of fugitive emissions (MMA 2011). The study concludes that AERMOD “consistently predicts concentrations higher than ISCST in the range of concentrations that would be critical decision points in the permitting process.”
6.2.3. Refined PM10 Modeling Results
In an attempt to address the problem of over-predicting impacts from fugitive dust at the Upton Plant, AERMOD was re-run for impacts at select receptors using the dry depletion option. This option, also available with ISCST3, seeks to account for particulate deposition near the source. It requires the user to input particle densities and size distributions (see Section 3.10.3). The receptors modeled with dry depletion included the 50 highest-concentration receptors from the initial model run. It was not realistic to use this option for the initial run, as modeling impacts on all receptors in the modeling domain would have required several hundred hours to execute.
Upton Plant Modeling Protocol and Results Page 60
With the dry depletion option enabled, AERMOD predicted lower 24-hour PM10 impacts as summarized in Table 6-5. Taking into account the 24-hour background, the model predicted a highest yearly 2nd high value of 157.1 µg/m3, slightly exceeding the NAAQS. As seen in Figure 6-3, this receptor falls on the project boundary.
nd Table 6-5: Top 20 Receptors, 24-Hr 2 High PM10 Values With Dry Depletion
Maximum Modeled Maximum NAAQS UTM UTM Concentration Concentration with Concentration Easting Northing 3 3 3 (µg/m ) Background (µg/m ) (µg/m ) 526962 4883832 117.14 157.14 150 526962 4883913 73.33 113.33 150 527294 4882218 45.69 85.69 150
527194 4882217 44.09 84.09 150 526820 4884211 43.47 83.47 150 526812 4884244 38.97 78.97 150 527094 4882215 37.28 77.28 150 526321 4884221 36.96 76.96 150 526720 4884209 36.03 76.03 150 526221 4884218 35.44 75.44 150 527389 4882333 35.40 75.40 150 526712 4884244 35.33 75.33 150 527385 4882433 35.15 75.15 150 526962 4884013 33.91 73.91 150 526775 4884333 33.61 73.61 150 526121 4884216 32.15 72.15 150 527393 4882233 31.72 71.72 150 527394 4882220 31.23 71.23 150 526421 4884223 31.06 71.06 150 526972 4882614 31.06 71.06 150
Upton Plant Modeling Protocol and Results Page 61
Figure 6-3. Receptors Exceeding the 24-Hour PM10 Standard
4886000
Compliant Receptors
> NAAQS w/ DD
4884000
4882000
4880000 523000 524000 525000 526000 527000 528000 529000
Upton Plant Modeling Protocol and Results Page 62
6.3. PM2.5 Modeling Analysis
Particulate matter in the form of PM2.5 emissions were modeled in a similar fashion to
PM10 emissions, except that the dry depletion option was not used. Other modeling exercises have demonstrated that accounting for particulate deposition does not affect modeled PM2.5 concentrations nearly to the degree than it affects PM10. The primary
source of PM2.5 emissions will be fuel combustion (primarily natural gas and diesel).
PM2.5 will also include the smaller fugitive dust particles generated by traffic on unpaved roads, road maintenance, drilling and construction activities, and wind erosion on disturbed areas.
The maximum yearly PM2.5 emissions from the Upton Plant were modeled for potential impacts on ambient air quality at all receptors in the modeling domain. The model produced maximum receptor concentrations for any calendar day (24-hour average) and for the entire modeling period (annual average). The 24-hour design value was computed for each receptor as the 8th high (98th percentile) concentration.
6.3.1. PM2.5 Modeling Results
Results from the AERMOD model run are presented below. The model predicted NAAQS compliance for all receptors and averaging intervals. One receptor showed an annual average concentration above the applicable Class II PSD increment of 4 µg/m3. This receptor falls on the project boundary. The maximum yearly 2nd high 24-hour concentration was predicted to exceed the Class II PSD increment at less than 1% of the receptors, all within 200 meters of the project boundary. Table 6-6 lists the top 20 receptors ranked by predicted annual average concentrations. Table 6-7 lists the top 20 receptors ranked by the predicted 98th percentile of 24-hour maximum concentrations. Figure 6-4 is an isopleth, or contour plot of the predicted annual concentrations attributable solely to the Upton Plant. Figure 6-5 is an isopleth map of the predicted, maximum 24-hour impacts attributable to the Upton Plant.
Upton Plant Modeling Protocol and Results Page 63
Table 6-6: Top 20 Receptors, Annual Average PM2.5 Values
Maximum UTM UTM Maximum Modeled NAAQS Concentration 3 Concentration with 3 Easting Northing Concentration (µg/m ) 3 (µg/m ) Background (µg/m ) 526962 4883832 4.80 6.03 12 526962 4883913 3.93 6.01 12 526720 4884209 2.63 5.98 12 526820 4884211 2.61 5.97 12 526972 4882614 2.58 5.92 12 526972 4882613 2.57 5.90 12 526712 4884244 2.52 5.84 12 526812 4884244 2.50 5.76 12 526321 4884221 2.44 5.73 12 526221 4884218 2.36 5.70 12 526977 4882513 2.33 5.68 12 526675 4884333 2.30 5.67 12 526974 4882514 2.28 5.66 12 526421 4884223 2.27 5.66 12 526775 4884333 2.26 5.63 12 526962 4884013 2.26 5.63 12 526620 4884208 2.23 5.60 12 526612 4884244 2.23 5.59 12 526121 4884216 2.20 5.58 12 527294 4882218 2.19 5.54 12 Table 6-6 shows that all receptor concentrations are predicted to comply with the annual NAAQS (12 µg/m3) and the top modeled concentration of 4.80 µg/m3 is slightly above the PSD Class II increment. The highest predicted receptor concentration, with background added, is about 50% of the NAAQS. Modeled concentrations are shown in Figure 6-4.
Upton Plant Modeling Protocol and Results Page 64
th Table 6-7: Top 20 Receptors, 98 percentile of 24-Hr Maximum PM2.5 Values
Maximum UTM UTM Maximum Modeled NAAQS Concentration 3 Concentration with 3 Easting Northing Concentration (µg/m ) 3 (µg/m ) Background (µg/m ) 526962 4883832 13.44 21.44 35 526962 4883913 11.39 19.39 35 526720 4884209 9.68 17.68 35 526712 4884244 9.36 17.36 35 526820 4884211 9.20 17.20 35 526812 4884244 8.91 16.91 35 526321 4884221 8.75 16.75 35 526675 4884333 8.70 16.70 35 526221 4884218 8.51 16.51 35 527194 4882217 8.23 16.23 35 526775 4884333 8.15 16.15 35 527294 4882218 8.10 16.10 35 526421 4884223 7.80 15.80 35 526620 4884208 7.71 15.71 35 526612 4884244 7.69 15.69 35 526121 4884216 7.64 15.64 35 526318 4884321 7.58 15.58 35 526168 4884317 7.54 15.54 35 526657 4884469 7.43 15.43 35 526520 4884206 7.38 15.38 35
Table 6-7 shows that all receptor concentrations are predicted to comply with the 24- hour NAAQS (35 µg/m3) and the top 5 design values are slightly greater than the PSD Class II increment (9 µg/m3). These receptors lie on the project boundary or within a hundred meters of the project boundary. The highest predicted 24-hour receptor concentration, with background added, is 61% of the NAAQS. This is confirmed graphically in Figure 6-5.
Upton Plant Modeling Protocol and Results Page 65
Figure 6-4. Annual PM2.5 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 66
th Figure 6-5. 98 Pctile 24-Hour PM2.5 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 67
6.4. NO2 Modeling Analysis
NO2 emissions are derived from oxides of nitrogen (NOx), at an assumed conversion ratio of 75% for annual impacts and a variable rate from 20% to 90% for 1-hour impacts.
This rate depends on absolute NOx concentrations. The primary source of NOx emissions will be fuel combustion from mobile and stationary sources.
Section 3.9 above discusses the AERMOD options employed for modeling NO2 impacts. The model predicted maximum hourly receptor concentrations by calendar day and the 98th percentile of the hourly maxima for the modeled year. It also predicted the average receptor concentrations for the entire year (annual average). Based on the rationale presented in Section 3.9, annual average NO2 concentrations were modeled using the Tier 2 Ambient Ratio Method (ARM). The ARM2 option was selected to model
1-hour NO2 concentrations due to its superior handling of NOx to NO2 conversion at
peak NOx concentrations. The ARM2 option accounts for lower NOx-NO2 conversion
rates at the higher NOx concentrations typically associated with short averaging intervals and receptors located near the emission sources.
Results from the NO2 AERMOD model run are presented below. The model predicted NAAQS compliance for all receptors and averaging intervals. It also predicted that all receptor concentrations will be below the annual PSD increment of 25 µg/m3. Table 6-8 lists the top 20 receptors ranked by annual average concentrations. The highest receptor concentration, with background added, was predicted to be 22% of the annual NAAQS. Table 6-9 lists the top 20 receptors ranked according to the 1-hour design value (98th percentile). The highest receptor concentration, with background added, was predicted to be 87% of the 1-hour NAAQS. Figure 6-6 is an isopleth, or contour plot of the predicted annual concentrations attributable solely to the Upton Plant. Figure 6-7 is an isopleth map of the predicted, 98th percentile 1-hour concentrations attributable to the Upton Plant.
Upton Plant Modeling Protocol and Results Page 68
Table 6-8: Top 20 Receptors, Annual Average NO2 (ARM)
Maximum UTM UTM Maximum Modeled NAAQS Concentration 3 Concentration with 3 Easting Northing Concentration (µg/m ) 3 (µg/m ) Background (µg/m ) 526962 4883832 16.00 22.00 100 526962 4883913 11.13 17.13 100 527062 4883832 9.40 15.40 100 527162 4883832 6.80 12.80 100 526820 4884211 5.90 11.90 100 526812 4884244 5.69 11.69 100 526720 4884209 5.37 11.37 100 527062 4883913 5.22 11.22 100 526712 4884244 5.07 11.07 100 526962 4884013 4.91 10.91 100 526775 4884333 4.88 10.88 100 527262 4883832 4.74 10.74 100 527162 4883913 4.70 10.70 100 526675 4884333 4.45 10.45 100 527342 4883432 4.31 10.31 100 527346 4883332 4.28 10.28 100 526912 4884244 4.02 10.02 100 526962 4884113 3.86 9.86 100 526920 4884212 3.74 9.74 100 527351 4883233 3.73 9.73 100
Upton Plant Modeling Protocol and Results Page 69
th Table 6-9: Top 20 Receptors, 98 percentile of Daily Max 1-Hr NO2 Values (ARM2)
Maximum NAAQS UTM UTM Maximum Modeled 3 Concentration with Concentration Easting Northing Concentration (µg/m ) 3 3 Background (µg/m ) (µg/m ) 527062 4883832 141.91 162.91 188 526962 4883832 131.97 152.97 188 527162 4883913 127.13 148.13 188 526962 4883913 124.00 145.00 188 527162 4883832 123.94 144.94 188 527062 4883913 123.00 144.00 188 527262 4883832 117.78 138.78 188 527346 4883332 104.30 125.30 188 527342 4883432 103.68 124.68 188 526720 4884209 101.70 122.70 188 526712 4884244 94.26 115.26 188 526812 4884244 93.72 114.72 188 526820 4884211 93.54 114.54 188 527329 4883732 89.54 110.54 188 526612 4884244 88.32 109.32 188 527162 4884013 87.93 108.93 188 527538 4883541 87.13 108.13 188 526675 4884333 87.03 108.03 188 527262 4883913 86.14 107.14 188 527536 4883614 85.89 106.89 188
Upton Plant Modeling Protocol and Results Page 70
Figure 6-6. Annual NO2 Concentrations (ARM, Without Background)
Upton Plant Modeling Protocol and Results Page 71
th Figure 6-7. 98 Percentile 1-Hr NO2 Concentrations (ARM2, Without Background)
Upton Plant Modeling Protocol and Results Page 72
6.5. SO2 Modeling Analysis
The primary source of SO2 emissions from the Upton Plant will be fuel combustion from mobile and stationary sources.
The maximum yearly SO2 emissions from the Upton Plant were modeled for potential impacts on ambient air quality at all receptors in the modeling domain. The model produced maximum hourly receptor concentrations by calendar day and the 99th percentile of these hourly maxima for the modeled year. It also produced 3-hour maxima, 24-hour maxima, and the annual average receptor concentrations for the modeled year.
Results from the SO2 AERMOD model run are presented below and summarized in Table 6-1. All receptor concentrations were predicted to comply with the appropriate NAAQS. The annual average and maximum 24-hour average values were all very low (no NAAQS exists for these intervals). Table 6-10 lists the top 20 receptors ranked by 3- hour average concentrations. The highest receptor concentration, with background added, was predicted to be 15% of the 3-hour NAAQS. Table 6-11 lists the top 20 receptors ranked by the 4th-high 1-hour maximum (99th percentile) concentrations. The highest receptor concentration, with background added, was predicted to be 77% of the 1-hour NAAQS. Figure 6-8 is an isopleth, or contour plot of the predicted annual concentrations attributable solely to the Upton Plant. Figure 6-9 is an isopleth map of the predicted maximum 24-hour concentrations attributable to the Upton Plant. Figure 6- 10 is an isopleth map of the predicted maximum 3-hour concentrations attributable to the Upton Plant. Figure 6-11 is an isopleth map of the predicted, 99th percentile 1-hour concentrations attributable to the Upton Plant. AERMOD predicted that all receptor concentrations will be less than the PSD increments for all relevant averaging intervals (see Table 6-2).
Upton Plant Modeling Protocol and Results Page 73
Table 6-10: Top 20 Receptors, 3-Hr Maximum SO2 Maximum UTM UTM Maximum Modeled NAAQS Concentration 3 Concentration with 3 Easting Northing Concentration (µg/m ) 3 (µg/m ) Background (µg/m ) 527262 4883832 64.58 189.28 1300 527850 4884480 52.98 177.68 1300 527513 4883590 51.52 176.22 1300 527262 4883913 46.18 170.88 1300 527536 4883614 45.54 170.24 1300 527162 4883913 44.97 169.67 1300 526962 4883832 44.13 168.83 1300 527062 4883832 43.73 168.43 1300 527359 4883664 41.74 166.44 1300 527747 4884488 40.93 165.63 1300 527607 4883685 40.61 165.31 1300 527342 4883432 38.64 163.34 1300 527769 4883880 37.20 161.90 1300 526962 4883913 36.92 161.62 1300 527346 4883332 36.66 161.36 1300 527678 4883756 36.30 161.00 1300 527334 4883632 35.03 159.73 1300 527727 4883799 34.71 159.41 1300 527355 4883133 34.67 159.37 1300 527293 4883956 34.54 159.24 1300
Upton Plant Modeling Protocol and Results Page 74
th Table 6-11: Top 20 Receptors, 99 percentile of Daily Maximum 1-Hr SO2 Values
Maximum UTM UTM Maximum Modeled NAAQS Concentration 3 Concentration with 3 Easting Northing Concentration (µg/m ) 3 (µg/m ) Background (µg/m ) 527262 4883832 111.04 154.24 200 527062 4883832 109.75 152.95 200 527850 4884480 103.05 146.25 200 527162 4883913 85.13 128.33 200 527607 4883685 76.14 119.34 200 527536 4883614 74.36 117.56 200 527262 4883913 74.20 117.40 200 527887 4884391 71.48 114.68 200 527329 4883732 70.25 113.45 200 527813 4884569 70.16 113.36 200 527971 4884573 69.45 112.65 200 527359 4883664 68.68 111.88 200 527513 4883590 68.66 111.86 200 527162 4883832 64.32 107.52 200 527862 4884113 63.11 106.31 200 527293 4883956 59.77 102.97 200 527293 4883956 59.77 102.97 200 526962 4883832 58.12 101.32 200 527950 4884346 58.02 101.22 200
Upton Plant Modeling Protocol and Results Page 75
Figure 6-8. Modeled Annual SO2 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 76
Figure 6-9. Modeled Maximum 24-Hour SO2 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 77
Figure 6-10. Modeled Maximum 3-Hour SO2 Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 78
Figure 6-11. Modeled 99th Percentile 1-Hour SO2 Concentrations (w/o Background)
Upton Plant Modeling Protocol and Results Page 79
6.6. CO Modeling Analysis The primary source of CO emissions from the Upton Plant will be fuel combustion from mobile and stationary sources.
The maximum yearly CO emissions from the Upton Plant were modeled for potential impacts on ambient air quality at all receptors in the modeling domain. The model produced maximum 1-hour and 8-hour receptor concentrations over the 1-year modeling period.
Results from the CO AERMOD model run are illustrated below. Modeled concentrations at all receptors were predicted to be below the applicable NAAQS. As shown in Table 6- 1, all modeled concentrations of CO (with background added) constituted less than 10% of the NAAQS, and are therefore not tabulated separately. Figure 6-12 is an isopleth, or contour plot of the predicted maximum 8-hour concentrations attributable to the Upton Plant. Figure 6-13 is an isopleth map of the predicted maximum 1-hour concentrations attributable to the Upton Plant.
Upton Plant Modeling Protocol and Results Page 80
Figure 6-12. Modeled Maximum 8-Hr CO Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 81
Figure 6-13. Modeled Maximum 1-Hr CO Concentrations (Without Background)
Upton Plant Modeling Protocol and Results Page 82
7 CALPUFF MODELING RESULTS AND ANALYSIS
7.1. Introduction The purpose of AQRV modeling is to identify and disclose impacts on Class I and sensitive Class II area resources (i.e., visibility, flora, fauna, etc.) by the projected emissions from a proposed project. AQRVs are resources which may be adversely affected by a change in air quality. Based on its proximity to the Wind Cave National Park, a federally mandated Class I area, and to four sensitive Class II areas, the Bear Lodge Project was modeled to determine its potential AQRV impacts. Species modeled
included PM10, PM2.5, SO2, NOx, SO4, NHNO3 and NO3. The first four of these would be emitted by the project, while the other three may form in the atmosphere.
The model selected for AQRV impact analysis (recommended by EPA and the Federal Land Managers) is CALPUFF, along with its companion models CALMET and CALPOST. In addition to the above seven species, elemental carbon (EC) and secondary organic aerosol (SOA) were enabled in the model to accommodate Visibility Method 8.1. Visibility model outputs included daily background light extinction at receptors in Wind Cave National Park, to which the project impacts were added. By contrast, the modeled atmospheric deposition rates were attributable only to project emissions. Background deposition rates and significance thresholds were obtained from sources outside the model.
Visibility impacts from the Bear Lodge Project at Wind Cave and other model receptors were modeled under two scenarios. The first one included coarse particulate matter th (PM10) in computing total light extinction, which resulted in a 98 percentile of 24-hour changes in visibility (relative to background) of 0.7%. This level of change in visibility is well below the 5% change considered barely perceptible by 50% of the viewers. The th second scenario excluded PM10 (but not PM2.5) from this computation, resulting in a 98 percentile of 24-hour changes in visibility of 0.3%. Section 5.6.6 presents evidence and precedent for the validity of the second scenario, due to CALPUFF’s lack of accounting for deposition of most PM10-2.5 particles within a short distance of ground-level, fugitive emission sources.
Atmospheric deposition (also known as acid deposition), another measure of AQRV impact, is modeled by CALPUFF as the deposition of a variety of species containing
Upton Plant Modeling Protocol and Results Page 83
nitrogen and sulfur. SO2 and NOx emissions from the Bear Lodge Project constitute potential sources of acid deposition at Wind Cave National Park and other sensitive areas. The modeled deposition rates predicted by CALPUFF were first compared to measured deposition rates at Wind Cave. Second, the modeled deposition rates were compared to estimated critical loads at Wind Cave, below which no harmful impacts to the ecosystem would be expected to occur. Third, the modeled deposition rates were compared to the deposition analysis thresholds established by the U.S. Forest Service, below which deposition impacts are considered negligible. Section 7.3 presents these comparisons and predicts that annual deposition impacts from the Bear Lodge Project will be less than the deposition analysis thresholds for nitrogen and sulfur. This section also shows that historical deposition rates are substantially lower than the estimated critical loads for both sulfur and nitrogen.
In summary, atmospheric deposition and visibility model results predict impacts below the AQRV standards. The model predicts no significant impacts to any portions of the Class I area at Wind Cave National Park, or to any of the sensitive Class II areas included in the model.
Upton Plant Modeling Protocol and Results Page 84
7.2. Visibility Analysis
7.2.1. Basis for Analysis
In August 1977, the federal Clean Air Act was amended by Congress to establish the following national goal for visibility protection:
“Congress hereby declares as a national goal the prevention of any future, and the remedying of any existing, impairment of visibility in mandatory class I Federal areas which impairment results from man-made air pollution.”
To address this goal for each of the 156 mandatory federal Class I areas across the nation, the U.S. Environmental Protection Agency (EPA) developed regulations to reduce the impact of large industrial sources on nearby Class I areas.
The 1977 Clean Air Act Amendments also established the Prevention of Significant Deterioration (PSD) permit program, which included consultation with federal land managers on visibility impacts and public participation in permitting decisions. The PSD permit program was delegated to South Dakota on July 6, 1994, and later approved in South Dakota’s State Implementation Plan on January 22, 2008.
In 1980, EPA adopted regulations to address “reasonably attributable visibility impairment”, or visibility impairment caused by one or a small group of man-made sources generally located in close proximity to a specific Class I area. Most visibility impairment occurs when pollution in the form of small particles scatters or absorbs light. Air pollutants are emitted from a variety of natural and anthropogenic sources. Natural sources can include windblown dust and smoke from wildfires. Anthropogenic sources can include motor vehicles, electric utility and industrial fuel burning, prescribed burning, and mining operations. More pollutants mean more absorption and scattering of light, which reduce the clarity and color of scenery. Some types of particles such as sulfates and nitrates scatter more light, particularly during humid conditions. Other particles like elemental carbon from combustion processes are highly efficient at absorbing light.
Commonly, visibility is observed by the human eye and the object may be a single viewing target or scenery. A common measure of visual resources is the haze index,
Upton Plant Modeling Protocol and Results Page 85
expressed in deciviews (dv). The deciview is a metric used to represent normalized light extinction attributable to visibility-affecting pollutants.
The visibility threshold of concern is not exceeded if the 98th percentile change in light extinction is less than 5% for each year modeled, when compared to the annual average natural condition value for that Class I area (FLAG 2010). A 5% change in light extinction is equivalent to a 0.5 dv change in visibility. When assessing visibility impairment from regional haze, EPA guidelines indicate that for a source whose 98th percentile value of the haze index, evaluated on a 24-hour average basis, is greater than 0.5 dv is considered to contribute to regional haze visibility impairment.
7.2.2. Preliminary Modeled Visibility Impacts
Wind Cave National Park, located approximately 130 km southeast of the proposed Bear Lodge Project, is the nearest Class I area and the only one in the modeling domain. The maximum potential air emissions from the project were modeled for impacts on visibility at Wind Cave and four sensitive Class II areas within 100 km of the project, using the CALPUFF software and modeling protocol discussed in Section 5 of this report. The modeling results, with and without consideration of coarse particulate matter (PM10) emissions from the Bear Lodge Project, are summarized in Table 7-1.
Project emissions of fine particulate matter (PM2.5) were included in both model runs, along with oxides of nitrogen and sulfur. These three species, along with organic carbon, are the primary contributors to visibility impairment in the Wind Cave region (DENR 2010).
Table 7-1: Visibility Analysis Summary Scenario Statistic 1‐Year Significance Threshold 98th pctile Δdv 0.07 0.50 Modeled With Coarse #Days > 0.5 ∆dv 0 ‐‐ Particulate #Days > 1.0 ∆dv 0 ‐‐ Maximum Δdv 0.22 ‐‐
98th pctile Δdv 0.03 0.50 Modeled Without #Days > 0.5 ∆dv 0 ‐‐ Coarse #Days > 1.0 ∆dv 0 ‐‐ Particulate Maximum Δdv 0.11 ‐‐
Upton Plant Modeling Protocol and Results Page 86
Table 7-2: Visibility Analysis Top 50 delta-dv Values (without coarse PM) Easting Northing Delta dv Date Area 618.794 4856.994 0.113 1/12/13 Black Elk WA 521.580 4936.500 0.063 5/4/13 Devils Tower NM 521.580 4938.500 0.057 10/27/12 Devils Tower NM 521.580 4938.500 0.052 2/20/13 Devils Tower NM 521.580 4936.500 0.046 3/7/13 Devils Tower NM 521.580 4938.500 0.036 12/25/12 Devils Tower NM 618.794 4856.994 0.033 3/29/13 Black Elk WA 618.794 4856.994 0.031 10/26/12 Black Elk WA 620.166 4822.506 0.031 12/31/12 Wind Cave NP 614.290 4856.837 0.027 11/2/12 Black Elk WA 626.116 4860.832 0.026 2/21/13 Mt Rushmore NM 523.580 4938.500 0.026 3/15/13 Devils Tower NM 523.580 4938.500 0.025 9/22/12 Devils Tower NM 592.912 4841.820 0.024 1/13/13 Jewell Cave NM 618.794 4856.994 0.024 1/28/13 Black Elk WA 618.794 4856.994 0.024 2/28/13 Black Elk WA 592.912 4841.820 0.023 12/27/12 Jewell Cave NM 521.580 4936.500 0.023 5/1/13 Devils Tower NM 521.580 4936.500 0.022 5/3/13 Devils Tower NM 523.580 4936.500 0.021 10/11/12 Devils Tower NM 523.580 4938.500 0.021 11/10/12 Devils Tower NM 521.580 4938.500 0.021 5/2/13 Devils Tower NM 523.580 4938.500 0.02 12/24/12 Devils Tower NM 592.912 4841.820 0.019 1/11/13 Jewell Cave NM 592.912 4841.820 0.019 1/19/13 Jewell Cave NM 618.794 4856.994 0.018 11/23/12 Black Elk WA 521.580 4936.500 0.018 3/19/13 Devils Tower NM 521.580 4938.500 0.017 3/20/13 Devils Tower NM 615.470 4860.225 0.016 11/25/12 Black Elk WA 521.580 4936.500 0.016 12/22/12 Devils Tower NM 624.933 4860.839 0.016 1/4/13 Mt Rushmore NM 521.580 4938.500 0.016 4/10/13 Devils Tower NM 521.580 4936.500 0.016 6/10/13 Devils Tower NM 523.580 4936.500 0.015 10/28/12 Devils Tower NM 618.794 4856.994 0.015 11/3/12 Black Elk WA 592.844 4843.307 0.015 1/3/13 Jewell Cave NM 626.116 4860.832 0.015 1/22/13 Mt Rushmore NM 618.794 4856.994 0.015 1/29/13 Black Elk WA 592.844 4843.307 0.015 3/4/13 Jewell Cave NM 621.928 4861.272 0.015 4/23/13 Mt Rushmore NM 618.794 4856.994 0.014 11/15/12 Black Elk WA 523.580 4936.500 0.014 12/6/12 Devils Tower NM 620.215 4819.730 0.014 3/21/13 Wind Cave NP 521.580 4938.500 0.014 4/1/13 Devils Tower NM 523.580 4938.500 0.014 7/26/13 Devils Tower NM 523.580 4936.500 0.013 12/14/12 Devils Tower NM 634.136 4831.104 0.013 1/6/13 Wind Cave NP 595.297 4843.736 0.013 2/26/13 Jewell Cave NM 614.290 4856.837 0.013 4/11/13 Black Elk WA
Table 7-2 lists the top 50 24-hour visibility impacts, the associated receptors, and dates.
Upton Plant Modeling Protocol and Results Page 87
7.2.3. Effect of Coarse Particulate on CALPUFF Visibility Assessment
Even without the exclusion of PM10-2.5, Table 7-1 shows the 98th percentile of the annual, 24-hour average changes in haze index to be far less than the contribution
threshold of 0.5 dv. With the PM10-2.5 exclusion, the modeled ∆dv values are even lower. Deposition is recognized as an important effect that can lead to rapid concentration
depletion in a fugitive PM10 emissions plume generated at or near ground level. Physical measurements reported by the South Dakota Department of Natural Resources (DENR) and the Western Regional Air Partnership (WRAP) conclude that coarse mass
particulates (i.e., PM10 and larger) contribute a small fraction toward visibility impairment at Wind Cave. DENR’s Regional Haze State Implementation Plan states, “In the 1st quarter, ammonia sulfate and ammonia nitrate have the greatest impact on visibility impairment in the Wind Cave National Park. In the 2nd quarter, ammonia sulfate has the greatest impact on visibility impairment in the Wind Cave National Park in the last five years. In the 3rd quarter, organic carbon mass has the greatest impact on visibility impairment followed by ammonia sulfate. In the 4th quarter, ammonia sulfates and ammonia nitrate continue to contribute the greatest with one exception in 2005” (DENR 2010). In 2005, organic carbon dominated due to wild fires.
Despite the above findings and the fact that virtually all of the PM10-2.5 emissions from the Bear Lodge Project would be ground-level fugitive dust, initial CALPUFF modeling
results showed PM10-2.5 emissions to be dominant in predicting changes in visibility at Wind Cave and sensitive Class II area receptors. On days with non-zero ∆dv values,
CALPUFF attributed on average about 56% of the change in visibility to PM10-2.5 emissions. Removing coarse particulates from the visibility analysis, as allowed for in the CALPUFF post-processor CALPOST, lowered these ∆dv values proportionately.
7.2.4. Final Modeled Visibility Impacts
The deciview haze index is derived from calculated light extinction measurements so that uniform changes in haziness correspond to uniform incremental changes in perception across the entire range of conditions, from pristine to highly impaired. The deciview haze index is calculated directly from the total light extinction coefficient (bext expressed in inverse megameters [Mm-1]) as follows:
Upton Plant Modeling Protocol and Results Page 88
-1 dv = 10 ln (bext/10 Mm )
CALPOST produced maximum 24-hour light extinction values for each model receptor at Wind Cave National Park and the four sensitive Class II areas. With coarse -1 particulate matter included in the model, the highest 24-hr total bext was 16.04 Mm . The corresponding background extinction on that day (without Bear Lodge Project impacts) was 15.694 Mm-1, providing a basis for the change in the haze index reported above. Including coarse PM, CALPUFF predicted the highest ∆dv value of 0.22 dv at a receptor along the Black Elk Wilderness Area boundary. The second highest ∆dv value of 0.15 dv occurred at a receptor along the Devils Tower boundary.
These top receptors did not change with coarse particulate matter excluded from the model. The highest 24-hr ∆dv value of 0.11 dv occurred at the same Black Elk Wilderness Area receptor, and the second highest ∆dv value of 0.06 dv occurred at the same Devils Tower receptor.
7.3. Deposition Analysis 7.3.1. Basis for Analysis
Air pollution emitted from a variety of sources is deposited from the air into ecosystems. Of particular concern are compounds containing sulfur and nitrogen that deposit from the air into the soil or surface waters. These pollutants may cause ecological changes, such as long-term acidification, soil nutrient imbalances affecting plant growth, and loss of biodiversity.
The term critical load is used to describe the threshold of air pollution deposition that causes harm to sensitive resources in an ecosystem. A critical load is technically defined by the National Atmospheric Deposition Program as “the quantitative estimate of an exposure to one or more pollutants below which significant harmful effects on specified sensitive elements of the environment are not expected to occur according to present knowledge.” Critical loads are typically expressed in terms of kilograms per hectare per year (kg/ha/yr) of wet or total (wet + dry) deposition. Critical loads are widely used to set policy for resource protection in Europe and Canada. They are presently emerging as guidelines to help in the protection of Class I areas in the United States. Recommended critical loads for nitrogen alone range from 1.5 kg/ha/yr at sensitive
Upton Plant Modeling Protocol and Results Page 89
alpine regions such as Rocky Mountain National Park (Fenn 2003), to 8 kg/ha/yr at Mt. Rainier, to 10-25 kg/ha/yr in mixed and short-grass prairie systems (USFS 2010).
Due to the lower elevation and absence of lakes with low acid buffering capacity at Wind Cave and throughout the northern Great Plains, it is believed that conditions in Wisconsin and Minnesota are more representative than conditions in the Rocky Mountains. Based on the Acid Deposition Control Act passed by Minnesota, the sulfur (S) deposition limit that would protect the most sensitive lakes and streams from acidification was set at 11 kg/ha/yr for the Class I Boundary Waters Canoe Area Wilderness (USFS 2013). Total S plus 20% of nitrogen (N) deposition was set at 12 kg/ha/yr, implying a critical load for N of 5 kg/ha/yr. The Forest Service shows similar thresholds for the Rainbow Lake Wilderness in Wisconsin (7.5 kg/ha/yr each, for S and N). The combined critical loads (S + N) of 17 kg/ha/yr in Minnesota and 15 kg/ha/yr in Wisconsin are consistent with the 10-to-25 kg/hr/yr range cited above for N in mixed and short-grass prairie systems.
Another measure often applied to sulfur and nitrogen deposition is the Deposition Analysis Threshold, or concern threshold, below which estimated impacts from a source are considered negligible. In the Class I areas of Colorado, Wyoming and Montana where high mountain lakes often exhibit low acid neutralization capacity, this threshold has been set by the U.S. Forest Service at 0.005 kg/ha/yr for sulfur and the same for nitrogen. In the eastern U.S., including Wisconsin and Minnesota, the Class I thresholds are 0.010 kg/ha/yr (FLM 2011). To date, no concern threshold has been published for Class I areas in South Dakota. For conservatisim, the modeling results are compared to the 0.005 kg/ha/yr value.
7.3.2. Modeled Deposition Fluxes
In order to assess potential impacts of the Bear Lodge Project on atmospheric deposition at Wind Cave National Park, it is necessary to examine current conditions. Table 7-3 summarizes current atmospheric conditions at Wind Cave for the modeled years. Samples were collected and analyzed under the National Acid Deposition Program (NADP 2012). The combined (S + N) deposition rate or flux averaged just over 4 kg/ha/yr during the three-year period.
Upton Plant Modeling Protocol and Results Page 90
Table 7-3: Current Acid Deposition at Wind Cave National Park (kg/ha/yr) Year NH4 NO3 SO4 S (inferred) N (inferred) S + N 2009 2.14 4.68 3.00 1.00 2.72 3.72
2010 3.04 5.29 3.48 1.16 3.56 4.72
2011 2.30 4.78 2.70 0.90 2.87 3.77
Average 1.02 3.05 4.07 Source: National Atmospheric Deposition Program/National Trends Network, 2012
Table 7-4 presents the results of wet and dry deposition modeling of the Bear Lodge Project emissions using CALPUFF. Maximum nitrogen deposition was predicted at Devils Tower National Monument. Maximum sulfur deposition was predicted at Black Elk Wilderness Area. The highest combination of sulfur and nitrogen was predicted at Devils Tower. Table 7-4 compares these results to deposition analysis thresholds and critical loads.
Table 7-4: Acid Deposition Modeling Analysis at Devils Tower NM (Wet + Dry) Parameter Sulfur Nitrogen Sulfur+Nitrogen Modeled 1-yr average µg/m2/sec 0.00000045 0.00000203 0.00000248
Modeled 1-yr average kg/ha/yr 0.000141 0.000641 0.000782
Concern threshold (kg/ha/yr) 0.005 0.005 0.010
Wind Cave 3-yr Monitored Avg (kg/ha/yr) 1.02 3.05 4.07
Estimated critical load (kg/ha/yr) 12 5 17
The results of the deposition analysis predict that impacts from the Bear Lodge Project on Wind Cave National Park will be insignificant. First, Table 7-4 shows that measured deposition flux for S and N are less than the estimated critical loads, by a significant margin. Second, Table 7-4 predicts that potential annual deposition impacts from the Bear Lodge Project will be less than the concern thresholds.
Upton Plant Modeling Protocol and Results Page 91
8 REFERENCES
1. ANL 2013, Argonne National Laboratory, Soil Density, http://web.ead.anl.gov/resrad/datacoll/soildens.htm 2. API 2013, Prepared by RTP Environmental Associates for the American Petroleum Institute, Ambient Ratio Method Version 2 (ARM2) for Use With AERMOD for 1-hr
NO2 Modeling – Development and Evaluation Report, September 20, 2013. 3. BLM 2014, Task 3A Report for the Powder River Basin Coal Review, Cumulative Air Quality Effects, Prepared for the Bureau of Land Management by AECOM, February 2014. 4. BLM 2010, Draft Environmental Impact Statement, Appendix K. Air Resources Impact Assessment Technical Report For The Alton Coal Lease By Application, September 2010. 5. CAPCOA 2011, California Air Pollution Control Officers Association, Modeling
Compliance of the Federal 1-Hour NO2 NAAQS, October 2011. 6. Cliffs 2011, Michael E. Long, Director Environmental Strategy and Programs, Cliffs Natural Resources, Air Quality Modeling and Impacts on the Mining Industry: An Overview, September 26, 2011. 7. Countess 2001, Methodology For Estimating Fugitive Windblown And Mechanically Resuspended Road Dust Emissions Applicable For Regional Scale Air Quality Modeling; April 2001. 8. EPA 2013a, Calculations and References, EPA Clean Energy Website, http://www.epa.gov/cleanenergy/energy-resources/refs.html, see also NRC reference to EPA Clean Energy website at http://pbadupws.nrc.gov/docs/ML1320/ML13205A377.pdf 9. EPA 2013b, Ron Meyers, Office of Air Quality Planning and Standards, Sector Policy and Programs Division, Measurement Policy Group, Personal Communication; May 2, 2013. 10. EPA 2013c, EPA AirData Website, AQS ID#46-033-0132, 2009-2011, Daily maximum 8-hr ozone concentration (ppm); accessed May 2, 2013. 11. EPA 2012, Haul Road Workgroup Final Report Submission to EPA-QAQPS, March 2, 2012. 12. EPA 2011, Additional Clarification Regarding Application of Appendix W Modeling
Guidance for the 1-hour NO2 National Ambient Air Quality Standard, EPA Air Quality Modeling Group Memorandum, C439-01, March 1, 2011.
Upton Plant Modeling Protocol and Results Page 92
13. EPA 2010, Exhaust and Crankcase Emission Factors for Nonroad Engine Modeling, Compression-Ignition, July 2010. 14. EPA 2005, USEPA, 40 CFR Part 51, Revision to the Guideline on Air Quality Models: Adoption of a Preferred General Purpose (Flat and Complex Terrain) Dispersion Model and Other Revisions, November 9, 2005. 15. EPA 2004a, EPA, Exhaust and Crankcase Emission Factors for Non-Road Engine Modeling - Compression Ignition, April 2004. 16. EPA 2004b, Addendum, User’s Guide for the AMS/EPA Regulatory Model – AERMOD, September 2004. 17. EPA 2000, EPA, Meteorological Monitoring Guidance for Regulatory Modeling Applications, February 2000. 18. EPA 1998, EPA, Control of Emissions of Air Pollution from Non-Road Diesel Engines; Final Rule, Subpart 89.112, October 1998. 19. EPA 1996, EPA, Report on Revisions to 5th Edition AP-42 Section 3.3 Gasoline and Diesel Industrial Engines, September 1996. 20. EPA 1994a, Development and Testing of a Dry Deposition Algorithm (Revised), Publication #EPA-454/R-94-015; April, 1994. 21. EPA 1995a, User’s Guide for the ISC3 Dispersion Models, Volume II; September 1995. 22. EPA 1995b, Modeling Fugitive Dust Impacts from Surface Coal Mining Operations, Phase III, Evaluating Model Performance; December, 1995. 23. EPA 1995c, AP 42, Fifth Edition, Compilation of Air Pollutant Emission Factors, Volume 1: Stationary Point and Area Sources, 1995 (multiple updates through 2012), http://www.epa.gov/ttn/chief/ap42/ 24. EPA 1994a, Development and Testing of a Dry Deposition Algorithm (Revised), Publication #EPA-454/R-94-015; April, 1994. 25. EPA 1992, EPA Region 10, User’s Guide for the Fugitive Dust Model (FDM) (Revised), Volume I: User’s Instructions; Sept 1992. 26. Earth Tech 2001, State of Alaska Department of Environmental Conservation, CALPUFF/MM5 Study Report, June 2001. 27. Fitz 2002, Fitz, D., D. Pankratz, R. Philbrick and G. Li, Evaluation of the Transport and Deposition of Fugitive Dust using Lidar, Proceedings, EPA’s 11th Annual Emission Inventory Conference, 2002. 28. Majano 2013, Rosendo Majano, Air Quality Modeler, Colorado Department of Public Health and Environment, Personal Communication, May 2013.
Upton Plant Modeling Protocol and Results Page 93
29. Maricopa 2006, Maricopa Association of Governments, Modeling Protocol in Support of a 5 Percent Plan for PM-10 for the Maricopa County Nonattainment Area, September 2006. 30. MMA 2011, McVehil-Monnett Associates, Inc., Draft “White Paper,” Status of CAAA Section 234, Regulatory and Technical Issues Update, May 2011. 31. MRI 2006, Midwest Research Institute. Background Document for Revisions to Fine Fraction Ratios Used for AP-42 Fugitive Dust Emission Factors, February 2006. 32. New Mexico 2006, New Mexico Air Quality Bureau Air Dispersion Modeling Guidelines, February 2006 33. NJDEP 2005, New Jersey Department of Environmental Protection, Chromium Work Group, Air Transport Subgroup, Chapter 5, Public Comment Draft; March 2005. 34. NRC 2014, U.S. Nuclear Regulatory Commission, Environmental Impact Statement for the Dewey-Burdock Project in Custer and Fall River Counties, South Dakota, Supplement to the Generic Environmental Impact Statement for In-Situ Leach Uranium Milling Facilities, January 2014. 35. Pace 2005, Thompson G. Pace, US EPA, Methodology to Estimate the Transportable Fraction (TF) of Fugitive Dust Emissions for Regional and Urban Scale Air Quality Analyses, 8/3/2005 Revision. 36. Rosemont 2009, Calpuff Modeling Protocol For Rosemont Copper Project To Assess Impacts On Class I Areas, October 30, 2009. 37. RTP 2013, Mark Podrez – RTP Env. Assoc. Inc., 2013 R/S/L Modelers Workshop, April 23, 2013. 38. Sullivan 2006, Westbrook, J.A., and Sullivan, P.S., Fugitive Dust Modeling with
AERMOD for PM10 Emissions from a Municipal Waste Landfill, Specialty Conference, 2006. 39. TCEQ 2012, Chapter 106, Subchapter W, Stationary Engines and Turbines, §106.512 (6)(A), Effective August 16, 2012. 40. TRC 2005, Air Quality Technical Support Document, Jonah Infill Drilling Project DEIS, Appendix B: Project Emissions Inventories, TRC Environmental Corporation, March 24, 2004. 41. Trinity 2014, Qiguo Jing, PhD and George J. Schewe, CCM, QEP, Hourly NO to NO2 Conversion Methods in AERMOD, May 14, 2014. 42. Trinity 2007, Arron J. Heinerikson, Abby C. Goodman, Divya Harrison, Mary Pham, Trinity Consultants, Modeling Fugitive Dust Sources With Aermod, Revised January 2007.
Upton Plant Modeling Protocol and Results Page 94
43. Watson 1996, J.G. Watson, Desert Research Institute, Effectiveness Demonstration of Fugitive Dust Control Methods for Public Unpaved Roads and Unpaved Shoulders on Paved Roads, August 1996. 44. WDEQ 2013, Wyoming Department of Environmental Quality – Air Quality Division, Electronic Communication from Don Watzel, November 2013.
Upton Plant Modeling Protocol and Results Page 95
9 APPENDIX A
EMISSION INVENTORY CALCULATIONS
Upton Plant Modeling Protocol and Results Page 96
Upton Plant Tailpipe Emission Factors Engine Parameters Emission Factors (lb/hp‐hr) Equipment Schedule Basis For Equipment Tailpipe Emissions Horse‐ Load Hours Days per Weeks Number Avail- Operated (Assumes Tier 3 Diesel Engines) THC NOx CO SO2 CO2 PM10 PM2.5 HAP power Factor Fuel per Day Week per Year of Units ability Hours per Yr Front End Loader 501 40% Diesel 0.00247 0.00661 0.00573 0.00205 1.15000 0.00033 0.00032 0.00046 1 7 52 1 80% 291 End Dump Truck 375 40% Diesel 0.00247 0.00661 0.00573 0.00205 1.15000 0.00033 0.00032 0.00046 2.50 7 52 1 80% 728 Track Dozer 125 40% Diesel 0.00247 0.00661 0.00822 0.00205 1.15000 0.00033 0.00032 0.00046 1.50 7 52 1 80% 437 Motor Grader 297 40% Diesel 0.00247 0.00661 0.00573 0.00205 1.15000 0.00033 0.00032 0.00046 4 7 52 1 80% 1,165 Water Truck 375 40% Diesel 0.00247 0.00661 0.00573 0.00205 1.15000 0.00033 0.00032 0.00046 4 7 52 1 80% 1,165 Portable Light Plant 7 100% Diesel 0.00251 0.03100 0.00668 0.00205 1.15000 0.00220 0.00213 0.00046 10 7 52 1 80% 2,912 Passenger Vehicle 200 20% Diesel 0.00247 0.00661 0.00573 0.00205 1.15000 0.00033 0.00032 0.00046 0.50 7 52 15 80% 2,184 Pickup Truck 300 25% Gasoline 0.00042 0.00048 0.00660 0.00059 1.08000 0.00072 0.00070 0.00049 4 7 52 2 80% 2,330 Preconcentrate Transport Truck 475 40% Diesel 0.00031 0.00044 0.03414 0.00205 1.15000 0.00002 0.00002 0.00046 0.75 7 52 9 100% 2,457 Commercial Delivery Truck 350 20% Diesel 0.00031 0.00044 0.03414 0.00205 1.15000 0.00002 0.00002 0.00046 0.75 7 52 1 100% 273 Emergency Generator 400 100% Diesel 0.00247 0.00661 0.00573 0.00033 1.15000 0.00033 0.00032 0.00046 0.55 7 52 1 100% 200 Firewater Pump 100 100% Diesel 0.00251 0.03100 0.00668 0.00205 1.15000 0.00220 0.00213 0.00046 0.05 7 52 1 100% 18
Emission Factor Sources:
1. AP‐42 Table 3.3‐1, Emission Factors for Uncontrolled Gasoline and Diesel Industrial Engines (Diesel THC, SO2, CO2, Aldehydes)
2. EPA, Exhaust and Crankcase Emission Factors for Non‐Road Engine Modeling ‐ Compression Ignition, April 2004 (PM2.5)
3. EPA, Control of Emissions of Air Pollution from Non‐Road Diesel Engines; Final Rule, Subpart 89.112, October 1998 (all Tiers: NOx, CO, PM10) 4. EPA, AP‐42, Volume II, Appendix H, June 30, 1995 (THC, NOx, CO) Exhaust Emission Rates For High Altitude Light Duty Gasoline Powered Trucks I, Assumed Fleet Miles = 50,000 5. EPA, CFR 86.007‐11 Emission standards and Supplemental Requirements for 2007 and Later Model Year Diesel Heavy‐Duty Engines and Vehicles.
NOx Emission Factor Details: Tier 3 off‐road and emergency generator diesel engines: EF = 3.0 g/hp‐hr from Reference #2 above Pumps, forklifts, light plants: EF = 0.031 lb/hp‐hr from AP‐42 Table 3.3‐1 Diesel Engines Pickup trucks: EF = 0.000479 lb/hp‐hr from Reference #4 above Highway trucks: EF = 0.00044 lb/hp‐hr from Reference #5 above
Upton Plant Modeling Protocol and Results Page 97
Upton Plant Combustion Emissions
Total Tons/Yr by Equipment THC NOx CO SO2 CO2 PM10 PM2.5 HAP Front End Loader 0.07 0.19 0.17 0.06 34 0.01 0.01 0.01 End Dump Truck 0.13 0.36 0.31 0.11 63 0.02 0.02 0.03 Track Dozer 0.03 0.07 0.09 0.02 13 0.00 0.00 0.01 Motor Grader 0.17 0.46 0.40 0.14 80 0.02 0.02 0.03 Water Truck 0.22 0.58 0.50 0.18 100 0.03 0.03 0.04 Portable Light Plant 0.02 0.30 0.07 0.02 11 0.02 0.02 0.00 Passenger Vehicle 0.11 0.29 0.25 0.09 50 0.01 0.01 0.02 Pickup Truck 0.04 0.04 0.58 0.05 94 0.06 0.06 0.04 Preconcentrate Transport Truck 0.07 0.10 7.97 0.48 268 0.01 0.00 0.11 Commercial Delivery Truck 0.00 0.00 0.33 0.02 11 0.00 0.00 0.00 Emergency Generator 0.10 0.26 0.23 0.01 46 0.01 0.01 0.02 Firewater Pump 0.00 0.03 0.01 0.00 1 0.00 0.00 0.00 Process Boiler 11.23 31.25 85.78 0.61 122,537 7.76 7.76 0.08 Dryer Oil Heaters 0.35 3.03 1.29 0.02 3,865 0.24 0.24 0.00 Plant Heater 0.05 0.40 0.17 0.00 515 0.03 0.03 0.00 Shop Heater 0.01 0.12 0.05 0.00 155 0.01 0.01 0.00 Office Heater 0.01 0.08 0.03 0.00 103 0.01 0.01 0.00
FACILITY TOTALS (tons per year) 12.62 37.57 98.21 1.83 127,946 8.26 8.25 0.40
Upton Plant Modeling Protocol and Results Page 98
Upton Plant Mobile Fugitive Emissions
Fleet Transit Speed Weight Control PM10 PM2.5 Emission Factor Equipment Item Quantity Hours Duty2 or Rate1 (tons) lb/VMT VMT lb/hr Efficiency tons/yr tons/yr Reference Front End Loader 1 291 491 5.45 0% 0.79 0.12 AP‐42 Table 11.9‐4 End Dump Truck 1 728 63% 10 219.8 5.66 4,586 85% 1.95 0.19 AP‐42 Section 13.2.2 Track Dozer 1 437 0.03 0% 0.01 0.00 AP‐42 Table 11.9‐1 Motor Grader 1 1,165 90% 5 33.7 2.43 5,242 85% 0.96 0.10 AP‐42 Section 13.2.2 Water Truck 1 1,165 75% 10 100 3.97 8,736 85% 2.60 0.26 AP‐42 Section 13.2.2 Passenger Vehicle 15 2,184 20% 20 8.5 0.01 8,736 0% 0.04 0.01 AP‐42 Section 13.2.1 Pickup Truck 2 2,330 60% 20 3 0.00 27,955 0% 0.04 0.01 AP‐42 Section 13.2.1 Preconcentrate Transport Truck 9 2,457 25% 20 41 0.05 12,285 0% 0.28 0.07 AP‐42 Section 13.2.1 Commercial Delivery Truck 1 273 5% 20 13 0.01 273 0% 0.00 0.00 AP‐42 Section 13.2.1
MOBILE EQUIPMENT FUGITIVE PM10 EMISSIONS (TONS/YEAR) 6.68 0.76
Constants for PM10 Calculations Notes:
AP‐42 Industrial Unpaved Roads: k for PM10 1.5 1. For mobile equipment, miles/hr; for drills, average holes per hour; for loaders, tons/hr AP‐42 Industrial Unpaved Roads: a0.92. Transit duty accounts for idle time, wait time, load time, dump time, etc. AP‐42 Industrial Unpaved Roads: b0.454. For industrial unpaved roads, emission factor E = k(s/12)a(W/3)b Average silt content haul roads (% per AP‐42 Table 13.2.2‐ 1, western surface mine haul roads): s8.45. For shovels and loaders, used 0.037 lb TSP/ton; a density of 2.11 ton/cy was assumed (s.g. = 2.5, TR Table 14.16) Average silt content public roads (% per AP‐42 Revised 1.5 1.4 Table 13.2.2‐1, western surface mine haul roads): s 6.4 6. Bulldozing overburden, PM10 emission factor E = 0.75(1.0)(s) /(M) Average moisture content (%, Table 11.9‐3, OB): M7.97. Hours for vans, pickups and OTR trucks are based on project road travel only
Average number wet days/year (AP‐42 Figure 13.2.2‐1) 90 8. PM2.5 emissions calculated as 15% of PM10 for earth moving activities and wind erosion; 10% of PM10 for unpaved
Where separate factors were not given, PM10 was assumed to be 30% of TSP (AP‐42 Section 13.2.2, at 12%
silt, KPM10/KTSP = 1.5/4.9 = 0.306) 0.30 9. Preconcentrate transport truck avg wt = (23.5+58.5)/2 = 41 tons, payload = 35 tons, trips/day = 425/35 = 13
Paved Roads: k for PM10 0.0022 10. Control Efficiency taken from EPA‐450/3‐88‐908 (see Bear Lodge Plan of Operations, App. C, Access Road Design)
Paved Roads: k for PM2.5 0.00054 11. Emissions from passenger vehicles, pickups, delivery and product transport trucks assume paved surfaces 2 Paved Roads: silt loading AP‐42 Table 13.2.1‐2: (sL in g/m )0.6
Upton Plant Modeling Protocol and Results Page 99
Upton Plant Other Fugitive Emissions
TSP Emission Control PM10 PM2.5 Emission Factor Activity Area (ft2) Factor (tpy/acre) Efficiency tons/yr tons/yr Reference Wind Erosion on Disturbed Areas 3,943,922 0.38 0% 10.32 1.55 AP‐42 Table 11.9‐4
TOTAL OTHER FUGITIVE PM10 EMISSIONS (TONS/YEAR) 10.32 1.55
Constants for PM10 Calculations Notes:
Where separate factors were not given, PM10 was 1. PM2.5 emissions calculated as 15% of PM10 for earth moving activities and assumed to be 30% of TSP (AP‐42 Section 13.2.2, at 12% wind erosion (AP‐42, Section 13.2.2, Background Document for Revisions to Fine Fraction Ratios) silt, KPM10/KTSP = 1.5/4.9 = 0.306) 0.30
Upton Plant Modeling Protocol and Results Page 100
Upton Plant Solids Handling Emissions and Source Parameters
PM10 Emission Emission 1 1 1 1 Stack Stack Flow Emission Rate Rate PM10 PM2.5 No. of Diameter Height Rate Velocity Rate PM10 PM2.5 Hours Tons per Tons per Baghouse Source Units (meters) (meters) Temp °FTemp °K (m3/sec) (m/sec) (gr/dscf) (g/sec) (g/sec) per Year Year Year Chemical Unloading Bin Vents 4 0.4064 10 80 299.7 0.8825 6.80 0.005 0.0346 0.0052 8,760 1.20 0.18 PC Feed Chute Bin Vent 1 0.2032 10 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 8,760 0.08 0.01 PC Discharge Chute Baghouse 1 0.4572 10 80 299.7 2.7578 16.80 0.005 0.0270 0.0041 8,760 0.94 0.14 Silo Feed Bin Vent 1 0.2032 20 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 8,760 0.08 0.01 Product Transfer Bin Vents 3 0.3520 10 80 299.7 0.6619 6.80 0.005 0.0195 0.0029 8,760 0.68 0.10 Screw Dryer Scrubber 2 0.4311 10 160 344.1 1.0134 6.94 0.005 0.0173 0.0026 8,760 0.60 0.09 Ammonium Nitrate Scrubber 1 0.2032 10 80 299.7 0.2206 6.80 0.005 0.0022 0.0003 8,760 0.08 0.01 Ventilation Scrubber 1 0.3048 20 80 299.7 0.5516 7.56 0.005 0.0054 0.0008 8,760 0.19 0.03 Kiln Effluent Gas Scrubber 1 0.381 20 200 366.3 1.3485 11.83 0.005 0.0108 0.0016 8,760 0.38 0.06
Note 1: Stack diameter is the effective diameter for multiple units; flow rates and emission rates are combined for multiple units Note 2: Assumed barometric pressure = 25.74 in Hg (12.64 psia)
Upton Plant Modeling Protocol and Results Page 101
Upton Plant Gas Fired Equipment Emissions
Gas-Fired Small Equipment Item Plant Heater Shop Heater Office Heater Dryer Oil Heater Process Boiler Number of Units 1 1 1 3 1 Operating hours/yr 4,380 4,380 4,380 8,760 8,760 Maximum duty (MMBtu/hr) 2.00 0.60 0.40 2.50 237.80 Heating value (Btu/scf) 1,020 1,020 1,020 1,020 1,020 Years in operation 43 43 43 43 43
Natural Gas Emission Factors (lb/106 btu) - Sources: AP-42 Table 1.4-1, Table 1.4-3
Pollutant Plant Heater Shop Heater Office Heater Dryer Oil Heater Process Boiler
NOx 0.092 0.092 0.092 0.092 0.030 CO 0.039 0.039 0.039 0.039 0.082
PM10/PM2.5 0.007 0.007 0.007 0.007 0.007
SO2 0.001 0.001 0.001 0.001 0.001 TOC 0.011 0.011 0.011 0.011 0.011 VOC 0.005 0.005 0.005 0.005 0.005
CO2 118 118 118 118 118 HAP 0.00008 0.00008 0.00008 0.00008 0.00008
Emissions (tons/yr) Pollutant Plant Heater Shop Heater Office Heater Dryer Oil Heater Process Boiler Total
NOx 0.40 0.12 0.08 3.03 31.25 34.88 CO 0.17 0.05 0.03 1.29 85.78 87.32
PM10/PM2.5 0.03 0.01 0.01 0.24 7.76 8.05
SO2 0.00 0.00 0.00 0.02 0.61 0.64 TOC 0.05 0.01 0.01 0.35 11.23 11.66 VOC 0.02 0.01 0.00 0.18 5.62 5.83
CO2 515 155 103 3,865 122,537 127,175 HAP 0.00 0.00 0.00 0.00 0.08 0.08
Upton Plant Modeling Protocol and Results Page 102
10 APPENDIX B
SOURCE APPORTIONMENT
Upton Plant Modeling Protocol and Results Page 103
Upton Plant Emission Source Apportionment Summary (tons/yr)
Area Source Category PM10 PM2.5 NOx SO2 CO TSF 10.07 1.43 0.85 0.22 0.76 OB Storage 2.37 0.36 0.07 0.00 0.01 Pond Area 0.41 0.06 0.01 0.00 0.00 Haulroad 3.11 0.37 0.85 0.27 0.81 Access Road 0.30 0.10 0.33 0.46 6.62 Delivery Road 0.09 0.03 0.03 0.13 2.11 Pickup Road 0.02 0.01 0.06 0.03 0.15 Hydromet Plant 0.02 0.01 0.29 0.02 0.24 Tailings Stockpile 0.80 0.13 0.19 0.06 0.17 TOTAL AREA SOURCES 17.20 2.51 2.69 1.19 10.89
Upton Plant Modeling Protocol and Results Page 104
Upton Plant Particulate Emissions Source Apportionment
Emitter/Area Source PM10 tons/yr PM2.5 tons/yr TSF OB Storage Pond Area Haulroad Access Road Delivery Road Pickup Road Hydromet Plant Tailings Stockpile Front End Loader 0.80 0.13 100.00% End Dump Truck 1.97 0.21 75.00% 25.00% Track Dozer 0.01 0.00 100.00% Motor Grader 0.98 0.12 100.00% Water Truck 2.63 0.29 50.00% 50.00% Passenger Vehicle 0.05 0.02 80.00% 20.00% Pickup Truck 0.11 0.07 25.00% 15.00% 35.00% 12.50% 12.50% Preconcentrate Transport Truck 0.29 0.07 75.00% 25.00% Commercial Delivery Truck 0.00 0.00 75.00% 15.00% 10.00% Wind Erosion on Disturbed Areas 10.32 1.55 70.00% 23.00% 4.00% 3.00% Portable Light Plant 0.02 0.02 100.00% Emergency Generator 0.01 0.01 100.00% Firewater Pump 0.00 0.00 100.00%
TOTAL FUGITIVE PM10 17.20 2.51 10.07 2.37 0.41 3.11 0.30 0.09 0.02 0.02 0.80
TOTAL FUGITIVE PM2.5 1.43 0.36 0.06 0.37 0.10 0.03 0.01 0.01 0.13 Modeled Area (acres) 216.56 71.20 11.15 1.55 6.59 2.47 2.27 24.55 0.04 2 PM10 Emission Rate (lb/hr/ft ) 2.438E‐07 1.748E‐07 1.941E‐07 1.053E‐05 2.377E‐07 1.818E‐07 5.686E‐08 3.252E‐09 1.053E‐04 2 PM2.5 Emission Rate (lb/hr/ft ) 3.462E‐08 2.621E‐08 2.911E‐08 1.259E‐06 8.000E‐08 5.875E‐08 3.199E‐08 3.154E‐09 1.683E‐05
Upton Plant Modeling Protocol and Results Page 105
Upton Plant Gaseous Emissions Source Apportionment
Emitter/Area Source NOx tons/yr CO tons/yr SO2 tons/yr TSF OB Storage Pond Area Haulroad Access Road Delivery Road Pickup Road Hydromet Plant Tailings Stockpile Front End Loader 0.19 0.17 0.06 100.00% End Dump Truck 0.36 0.31 0.11 75.00% 25.00% Track Dozer 0.07 0.09 0.02 100.00% Motor Grader 0.46 0.40 0.14 100.00% Water Truck 0.58 0.50 0.18 50.00% 50.00% Passenger Vehicle 0.29 0.25 0.09 80.00% 20.00% Pickup Truck 0.04 0.58 0.05 25.00% 15.00% 35.00% 12.50% 12.50% Preconcentrate Transport Truck 0.10 7.97 0.48 75.00% 25.00% Commercial Delivery Truck 0.00 0.33 0.02 75.00% 15.00% 10.00% Portable Light Plant 0.30 0.07 0.02 70.00% 23.00% 4.00% 3.00% Emergency Generator 0.26 0.23 0.01 100.00% Firewater Pump 0.03 0.01 0.00 100.00%
TOTAL FUGITIVE NOx 2.69 10.89 1.19 0.85 0.07 0.01 0.85 0.33 0.03 0.06 0.29 0.19 TOTAL FUGITIVE CO 0.76 0.01 0.00 0.81 6.62 2.11 0.15 0.24 0.17
TOTAL FUGITIVE SO2 0.22 0.00 0.00 0.27 0.46 0.13 0.03 0.02 0.06 Modeled Area (acres) 216.56 71.20 11.15 1.55 6.59 2.47 2.27 24.55 0.04 2 NOx Emission Rate (lb/hr/ft ) 2.066E‐08 5.122E‐09 5.688E‐09 2.879E‐06 2.592E‐07 6.699E‐08 1.463E‐07 6.251E‐08 2.526E‐05 CO Emission Rate (lb/hr/ft2) 1.849E‐08 1.104E‐09 1.226E‐09 2.749E‐06 5.268E‐06 4.484E‐06 3.572E‐07 5.025E‐08 2.190E‐05 2 SO2 Emission Rate (lb/hr/ft ) 5.392E‐09 3.387E‐10 3.762E‐10 9.053E‐07 3.685E‐07 2.738E‐07 6.077E‐08 3.223E‐09 7.838E‐06
Upton Plant Modeling Protocol and Results Page 106