Locating and Estimating Air Emissions From Sources of and Cadmium Compounds

U. S. Environmental Protection Agency Office of Air and Radiation Office of Air Quality Planning and Standards Locating and Estimating Air Emissions From Sources of Cadmium and Cadmium Compounds

Prepared for: Anne Pope Office of Air Quality Planning and Standards Technical Support Division U. S. Environmental Protection Agency Research Triangle Park, N.C. 27711

Prepared by: Midwest Research Institute 401 Harrison Oaks Boulevard, Suite 350 Cary, North Carolina 27513 TABLE OF CONTENTS Section Page 1 PURPOSE OF DOCUMENT ...... 1-1 REFERENCES ...... 1-4 2 OVERVIEW OF DOCUMENT CONTENTS ...... 2-1 3 BACKGROUND...... 3-1 NATURE OF POLLUTANT ...... 3-1 OVERVIEW OF PRODUCTION, USE, AND EMISSIONS ....3-4 Production...... 3-4 Use...... 3-6 Emissions ...... 3-6 REFERENCES ...... 3-13 4 EMISSIONS FROM CADMIUM PRODUCTION ...... 4-1 CADMIUM AND CADMIUM PRODUCTION ...4-1 Process Description ...... 4-3 Emissions and Controls ...... 4-6 CADMIUM PRODUCTION ...... 4-11 Process Description ...... 4-14 Emissions and Controls ...... 4-19 CADMIUM STABILIZERS PRODUCTION ...... 4-20 Process Description ...... 4-22 Emission and Controls ...... 4-24 OTHER CADMIUM COMPOUND PRODUCTION ...... 4-25 Process Descriptions ...... 4-25 Emissions and Controls ...... 4-29 REFERENCES ...... 4-32 5 EMISSIONS FROM MAJOR USES OF CADMIUM ...... 5-1 CADMIUM ...... 5-1 Process Description ...... 5-2 Emission Control ...... 5-6 Emissions ...... 5-7 SECONDARY BATTERY MANUFACTURE ...... 5-7 Process Description ...... 5-9 Emission Control Measures ...... 5-13 Emissions ...... 5-13 CADMIUM STABILIZERS FOR ...... 5-14 Process Description ...... 5-16 Emission Control Measures ...... 5-17 Emissions ...... 5-18 IN PLASTICS ...... 5-18 Process Description ...... 5-20 Emission Control Measures ...... 5-22 Emissions ...... 5-22 REFERENCES ...... 5-23

iii TABLE OF CONTENTS (continued) 6 EMISSIONS FROM COMBUSTION SOURCES ...... 6-1 COMBUSTION...... 6-5 Coal Characteristics ...... 6-6 Process Description ...... 6-10 Emission Control Measures ...... 6-15 Emissions ...... 6-16 FUEL OIL COMBUSTION ...... 6-20 Fuel Oil Characteristics ...... 6-22 Process Description ...... 6-23 Emission Control Measures ...... 6-27 Emissions ...... 6-28 NATURAL GAS COMBUSTION ...... 6-33 Natural Gas Characteristics ...... 6-33 Process Description ...... 6-34 Emission Control Measures ...... 6-34 Emissions ...... 6-34 WOOD COMBUSTION...... 6-35 Process Description ...... 6-35 Emission Control Measures ...... 6-37 Emissions ...... 6-39 MUNICIPAL WASTE COMBUSTION ...... 6-40 Municipal Waste Characteristics . . . 6-42 Process Description ...... 6-42 Emission Control Measures ...... 6-48 Emissions ...... 6-50 INCINERATORS ...... 6-51 Process Description ...... 6-54 Emission Control Measures ...... 6-58 Emissions ...... 6-59 MEDICAL WASTE ...... 6-61 Process Description ...... 6-62 Emission Control Measures ...... 6-67 Emissions ...... 6-68 REFERENCES ...... 6-75 7 EMISSIONS FROM NONFERROUS /REFINING ...... 7-1 PRIMARY SMELTING ...... 7-1 Process Description ...... 7-2 Emission Control Measures ...... 7-4 Emissions ...... 7-5 PRIMARY SMELTING ...... 7-7 Process Description ...... 7-8 Emission Control Measures ...... 7-13 Emissions ...... 7-14 PRIMARY SMELTING AND REFINING ...... 7-15 Process Description - Electrolytic ..... 7-17 Process Description--Pyrometallurgical (Electrothermic) ...... 7-22 Emission Control Measures ...... 7-23 Emissions ...... 7-24

iv TABLE OF CONTENTS (continued) SECONDARY COPPER SMELTING AND REFINING ..... 7-25 Process Description ...... 7-26 Emission Control Measures ...... 7-31 Emissions ...... 7-33 SECONDARY ZINC RECOVERY FROM METALLIC . . . 7-34 Process Description ...... 7-34 Emission Control Measures ...... 7-39 Emissions ...... 7-41 SECONDARY ZINC RECOVERY FROM PRODUCTION . . 7-41 Process Description ...... 7-43 Emission Control Measures ...... 7-48 Emissions ...... 7-48 REFERENCES ...... 7-49 8 EMISSIONS FROM MISCELLANEOUS SOURCES ...... 8-1 AND STEEL PRODUCTION ...... 8-1 Process Description ...... 8-4 Emission Control Measures ...... 8-10 Emissions ...... 8-12 PORTLAND MANUFACTURING ...... 8-12 Process Description ...... 8-15 Emission Control Measures ...... 8-18 Emissions ...... 8-20 ROCK PROCESSING ...... 8-20 Process Description ...... 8-23 Emission Control Measures ...... 8-25 Emissions ...... 8-26 BLACK PRODUCTION ...... 8-27 Process Description ...... 8-29 Emission Control Measures ...... 8-31 Emissions ...... 8-31 MOBILE SOURCES ...... 8-32 REFERENCES ...... 8-33 9 SOURCE TEST PROCEDURES ...... 9-1 INTRODUCTION ...... 9-1 MULTIPLE SAMPLING TRAINS ...... 9-2 Method 0012-Methodology for the Determin- ation of Metals Emissions in Exhaust Gases from Hazardous Waste Incineration and Similar Combustion Sources ...... 9-2 Methodology for the Determination of Metals Emissions in Exhaust Gases from Hazardous Waste Incineration and Similar Combustion Sources ...... 9-4 CARB Method 436-Determination of Multiple Metals Emissions from Stationary Sources . . 9-4 EPA Method 29-Methodology for the

v TABLE OF CONTENTS (continued) Determination of Metals Emissions in Exhaust Gases from Incineration and Similar Combustion Sources () .....9-5 ANALYTICAL METHODS FOR DETERMINATION OF CADMIUM . . 9-5 Flame AAS ...... 9-6 Graphite AAS ...... 9-6 Hydride AA...... 9-6 Cold-Vapor AA ...... 9-7 SUMMARY...... 9-7 REFERENCES ...... 9-9

APPENDIX A - NATIONWIDE EMISSION ESTIMATES ...... A-1 APPENDIX B - SUMMARY OF COMBUSTION SOURCE CADMIUM EMISSION DATA ...... B-1 APPENDIX C - PLANT LOCATIONS AND ANNUAL CAPACITIES FOR MISCELLANEOUS EMISSIONS SOURCES ...... C-1

vi LIST OF FIGURES Figure Page

3-1 1991 supply and demand for cadmium ...... 3-5 3-2 End use of cadmium ...... 3-7 4-1 Flow diagram for cadmium refining ...... 4-4 4-2 Process flowsheet for the production of cadmium pigments ...... 4-17 4-3 General flowsheet for the production of powdered cadmium stabilizers...... 4-23 5-1 Cadmium electroplating process...... 5-4 5-2 Simplified flow diagram for production of sintered plate -cadmium batteries...... 5-11 6-1 Distribution of sewage sludge incinerators in the United States ...... 6-53 6-2 Process flow diagram for sludge incineration ..... 6-55 6-3 Major components of an incineration system...... 6-64 7-1 Typical primary lead-processing scheme ...... 7-3 7-2 Typical primary copper-smelting process...... 7-9 7-3 Typical primary zinc-smelting process...... 7-18 7-4 Process flow diagram for second-grade copper recovery...... 7-28 7-5 Process flow diagram for high-grade and alloying...... 7-29 7-6 Process flow diagram for secondary zinc processing . . 7-36 7-7 Process flow diagram for Waelz kiln process...... 7-45 7-8 Process flow diagram for zinc calcine formation. . . . 7-47 8-1 General flow diagram for the iron and steel industry . . 8-5

8-2 Process flow diagram of Portland cement manufacturing process ...... 8-16

8-3 Typical flowsheet for processing phosphate rock . . . 8-24

vii LIST OF FIGURES (continued) 8-4 Process flow diagram for carbon black manufacturing process ...... 8-30 9-1. Typical multiple metals sampling train ...... 9-3

viii LIST OF TABLES Table Page 3-1 PHYSICAL PROPERTIES OF CADMIUM ...... 3-2 3-2 SIC CODES OF INDUSTRIES ASSOCIATED WITH CADMIUM EMISSIONS ...... 3-10 3-3 ESTIMATED 1990 NATIONWIDE CADMIUM EMISSIONS FOR SELECTED SOURCE CATEGORIES ...... 3-12 4-1 CADMIUM AND PRODUCERS ...... 4-2 4-2 INVENTORY OF CADMIUM EMISSION SOURCES AND CONTROLS FOR CADMIUM REFINING PLANTS ...... 4-7 4-3 PRIMARY ZINC AND CADMIUM PRODUCERS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXIC CHEMICALS RELEASE INVENTORY ...... 4-10 4-4 CADMIUM EMISSION FACTORS FOR CADMIUM REFINING PLANT USING LEAD DUST ...... 4-12 4-5 EMISSION FACTORS FOR CADMIUM AND CADMIUM OXIDE PRODUCTION ...... 4-13 4-6 COMMON CADMIUM PIGMENTS PRODUCED IN 1991 ...... 4-15 4-7 CURRENT CADMIUM PRODUCERS ...... 4-16 4-8 INORGANIC PIGMENTS MANUFACTURERS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXIC CHEMICALS RELEASE INVENTORY ...... 4-21 4-9 MANUFACTURERS OF ORGANIC COMPOUNDS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXIC CHEMICALS RELEASE INVENTORY ...... 4-26 4-10 CADMIUM COMPOUND MANUFACTURERS (OTHER THAN CADMIUM OXIDE, PIGMENTS, AND STABILIZERS) ..... 4-27 4-11 CADMIUM COMPOUND MANUFACTURERS (OTHER THAN CADMIUM OXIDE, PIGMENTS, AND STABILIZERS ...... 4-30 4-12 MANUFACTURERS OF INORGANIC COMPOUNDS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXIC CHEMICALS RELEASE INVENTORY ...... 4-31 5-1 MARKET AREAS FOR CADMIUM COATINGS ...... 5-3 5-2 COMPOSITION AND OPERATING PARAMETERS OF BATH ...... 5-6

ix LIST OF TABLES (continued) 5-3 NICKEL-CADMIUM BATTERY PRODUCERS--1990 ...... 5-9 5-4 REPORTED CADMIUM EMISSIONS BY MANUFACTURERS OF FORMULATED RESINS AND PRODUCTS ...... 5-19 5-5 REPORTED CADMIUM EMISSIONS BY PRODUCERS OF CUSTOM COMPOUNDED RESINS ...... 5-21 6-1 DISTRIBUTION OF FOSSIL FUEL CONSUMPTION IN THE UNITED STATES ...... 6-3 6-2 COAL HEATING VALUES ...... 6-7 6-3 EXAMPLES OF COAL HEAT CONTENT VARIABILITY ...... 6-8 6-4 CADMIUM CONCENTRATION IN COAL BY COAL TYPE ...... 6-11 6-5 CADMIUM CONCENTRATION IN COAL BY REGION ...... 6-11 6-6 CALCULATED UNCONTROLLED CADMIUM EMISSION FACTORS FOR COAL COMBUSTION ...... 6-18 6-7 SUMMARY OF CADMIUM EMISSION FACTORS FOR COAL COMBUSTION ...... 6-19 6-8 BEST TYPICAL CADMIUM EMISSION FACTORS FOR COAL COMBUSTION ...... 6-21 6-9 TYPICAL HEATING VALUES OF FUEL OILS ...... 6-24 6-10 TYPICAL FUEL OIL HEATING VALUES FOR SPECIFIC REGIONS ...... 6-25 6-11 CADMIUM CONCENTRATION IN OIL BY OIL TYPE ...... 6-26 6-12 CALCULATED UNCONTROLLED CADMIUM EMISSION FACTORS FOR FUEL OIL COMBUSTION ...... 6-29 6-13 MEASURED CADMIUM EMISSION FACTORS FOR FUEL OIL COMBUSTION...... 6-30 6-14 CADMIUM EMISSION FACTORS FOR FUEL OIL COMBUSTION GENERATED FROM "HOT SPOTS" TESTS .... 6-32 6-15 BEST TYPICAL CADMIUM EMISSION FACTORS FOR FUEL OIL COMBUSTION...... 6-33 6-16 SUMMARY OF CADMIUM EMISSION FACTORS FOR WOOD COMBUSTION ...... 6-39

x LIST OF TABLES (continued) 6-17 SUMMARY OF GEOGRAPHICAL DISTRIBUTION OF MWC FACILITIES...... 6-41 6-18 CURRENT AND FORECAST COMPOSITION OF DISPOSED RESIDENTIAL AND COMMERCIAL WASTE (WEIGHT PERCENT) ...... 6-43 6-19 BEST TYPICAL CADMIUM EMISSION FACTORS FOR MUNICIPAL WASTE COMBUSTORS ...... 6-52 6-20 SUMMARY OF CADMIUM EMISSION FACTORS FOR SEWAGE SLUDGE INCINERATION ...... 6-60 6-21 UNCONTROLLED CADMIUM EMISSION FACTORS FOR MEDICAL WASTE INCINERATORS ...... 6-70 6-22 SUMMARY OF CONTROLLED CADMIUM EMISSION FACTORS AND CONTROL EFFICIENCIES FOR MEDICAL WASTE INCINERATORS ...... 6-73 6-23 BEST TYPICAL UNCONTROLLED CADMIUM EMISSION FACTORS FOR MEDICAL WASTE INCINERATORS ...... 6-74 7-1 DOMESTIC PRIMARY LEAD SMELTERS AND REFINERIES .....7-2 7-2 PRIMARY LEAD PRODUCERS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXICS RELEASE INVENTORY ...... 7-5 7-3 CADMIUM EMISSION FACTORS FOR LEAD-SMELTING FACILITIES...... 7-6 7-4 DOMESTIC PRIMARY COPPER SMELTERS AND REFINERIES ....7-7 7-5 PRIMARY COPPER PRODUCERS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXICS RELEASE INVENTORY ...... 7-14 7-6 CADMIUM EMISSIONS FROM PRIMARY COPPER PRODUCTION . . . 7-16 7-7 DOMESTIC PRIMARY ZINC PRODUCERS ...... 7-17 7-8 PRIMARY ZINC PRODUCERS REPORTING CADMIUM EMISSIONS IN THE 1990 TOXICS RELEASE INVENTORY ...... 7-24 7-9 CADMIUM EMISSIONS FROM PRIMARY ZINC PRODUCTION .... 7-25 7-10 DOMESTIC SECONDARY COPPER PRODUCERS ...... 7-26 7-11 SECONDARY COPPER PRODUCERS REPORTING CADMIUM EMISSIONS IN THE 1989 AND 1990 TOXICS RELEASE INVENTORY ...... 7-34

xi LIST OF TABLES (continued) 7-12 DOMESTIC PRODUCERS OF SECONDARY ZINC FROM METALLIC SCRAP ...... 7-35 7-13 CADMIUM EMISSIONS FROM SECONDARY ZINC RECOVERY FROM SCRAP...... 7-42 7-14 DOMESTIC PRODUCERS OF SECONDARY ZINC FROM EAF DUST . 7-44 8-1 INTEGRATED IRON AND STEEL PLANTS ...... 8-3 8-2 PRODUCTION CAPACITY FOR INTEGRATED IRON AND STEEL FACILITIES IN THE UNITED STATES IN 1991 ...... 8-6 8-3 CADMIUM RELEASES REPORTED BY IRON AND STEEL FACILITIES IN 1990 TRI ...... 8-13 8-4 CADMIUM EMISSIONS REPORTED FROM BETHLEHEM STEEL, SPARROWS POINT, MARYLAND ...... 8-14 8-5 CADMIUM EMISSION FACTORS FROM PORTLAND CEMENT FACILITIES ...... 8-21 8-6 CARBON BLACK PRODUCTION FACILITIES ...... 8-28 9-1 CADMIUM SAMPLING METHODS ...... 9-8

xii SECTION1 PURPOSEOFDOCUMENT

TheU.S.EnvironmentalProtectionAgency(EPA),State,and localairpollutioncontrolagenciesarebecomingincreasingly awareofthepresenceofsubstancesintheambientairthatmay betoxicatcertainconcentrations. Thisawareness,inturn,has ledtoattemptstoidentifysource/receptorrelationshipsfor thesesubstancesandtodevelopcontrolprogramstoregulate emissions. Unfortunately,littleinformationexistsonthe ambientairconcentrationofthesesubstancesoraboutthe sourcesthatmaybedischargingthemtotheatmosphere.

Toassistgroupsinterestedininventoryingairemissionsof variouspotentiallytoxicsubstances,EPAispreparingaseries ofdocumentssuchasthisthatcompilesavailableinformationon sourcesandemissionsofthesesubstances. Priordocumentsin theseriesarelistedbelow:

EPA Substance PublicationNumber Acrylonitrile EPA-450/4-84-007a CarbonTetrachloride EPA-450/4-84-007b Chloroform EPA-450/4-84-007c EthyleneDichloride EPA-450/4-84-007d Formaldehyde EPA-450/4-91-012 Nickel EPA-450/4-84-007f EPA-450/4-84-007g EPA-450/4-84-007h EPA-450/4-84-007i Epichlorohydrin EPA-450/4-84-007j VinylideneChloride EPA-450/4-84-007k EthyleneOxide EPA-450/4-84-007l

1-1 Chlorobenzene EPA-450/4-84-007m PolychlorinatedBiphenyls(PCB’s) EPA-450/4-84-007n PolycyclicOrganicMatter(POM) EPA-450/4-84-007p Benzene EPA-450/4-84-007q Perchloroethyleneand EPA-450/2-89-013 Trichloroethylene MunicipalWasteCombustion EPA-450/2-89-006 CoalandOilCombustion EPA-450/2-89-001 1,3-Butadiene EPA-450/2-89-021 Chromium(Supplement) EPA-450/2-89-002 SewageSludge EPA-450/2-90-009 Styrene EPA-450/4-91-029

Thisdocumentdealsspecificallywithcadmiumandcadmium compounds;however,themajorityoftheinformationcontainedin thisdocumentconcernscadmium. Sourcesofcadmiumemissions evaluatedinthisdocumentinclude:(1)cadmiumproductionand useprocesses;(2)emissionsfromcombustionsources;(3) productionofothernonferrousmetalswherecadmiumemissions resultasinadvertentbyproductsoftheprocess;(4)production processesforselectedmaterialsotherthannonferrousmetals; and(5)mobilesources.

Inadditiontotheinformationpresentedinthisdocument, anotherpotentialsourceofemissionsdataforcadmiumand cadmiumcompoundsistheToxicChemicalReleaseInventory(TRI) formrequiredbySection313ofTitleIIIofthe1986Superfund AmendmentsandReauthorizationAct(SARA313).1 SARA313 requiresownersandoperatorsoffacilitiesincertainStandard IndustrialClassificationCodesthatmanufacture,import,process orotherwiseusetoxicchemicals(aslistedinSection313)to reportannuallytheirreleasesofthesechemicalstoall environmentalmedia. AspartofSARA313,EPAprovidespublic accesstotheannualemissionsdata. TheTRIdatainclude generalfacilityinformation,chemicalinformation,andemissions data. Airemissionsdataarereportedastotalfacilityrelease

1-2 estimatesforfugitiveemissionsandpointsourceemissions. No individualprocessorstackdataareprovidedtoEPAunderthe program. TheTRIrequiressourcestousestackmonitoringdata forreporting,ifavailable,buttheruledoesnotrequirestack monitoringorothermeasurementofemissionsifitis unavailable. Ifmonitoringdataareunavailable,emissionsare tobequantifiedbasedonbestestimatesofreleasestothe environment.

ThereaderiscautionedthattheTRIwillnotlikelyprovide facility,emissions,andchemicalreleasedatasufficientfor conductingdetailedexposuremodelingandriskassessment. In manycases,theTRIdataarebasedonannualestimatesof emissions(i.e.,onemissionfactors,materialbalance calculations,andengineeringjudgment). Werecommendtheuseof TRIdatainconjunctionwiththeinformationprovidedinthis documenttolocatepotentialemittersofcadmiumandtomake preliminaryestimatesofairemissionsfromthesefacilities.

CadmiumisofparticularimportanceasaresultoftheClean AirActAmendmentsof1990. Cadmiumanditscompoundsare includedintheTitleIIIlistofhazardousairpollutantsand willbesubjecttostandardsestablishedunderSection112, includingmaximumachievablecontroltechnology(MACT). Also, Section112(c)(6)ofthe1990Amendmentsmandatethatcadmium (amongothers)besubjecttostandardsthatallowforthemaximum degreeofreductionofemissions. Thesestandardsaretobe promulgatednolaterthan10yearsfollowingthedateof enactment.

Thedataoncadmiumemissionsarebased,wheneverpossible, ontheresultsofactualtestprocedures. Datapresentedinthis documentaretotalcadmiumemissionsanddonotdifferentiatethe metallicandionicformsofcadmium. Thesamplingandanalysis

1-3 proceduresemployedforthedeterminationofthecadmium concentrationsfromvarioussourcesarepresentedinSection9, SourceTestMethod.

1-4 REFERENCESFORSECTION1 1. ToxicChemicalReleaseReporting: CommunityRight-To-Know. FederalRegister52(107): 21152-21208. June4,1987.

1-5 SECTION2

OVERVIEWOFDOCUMENTCONTENTS

AsnotedinSection1,thepurposeofthisdocumentisto assistFederal,State,andlocalairpollutionagenciesand otherswhoareinterestedinlocatingpotentialairemittersof cadmiumandcadmiumcompoundsandestimatingairemissionsfrom thesesources. Becauseofthelimitedbackgrounddataavailable, theinformationsummarizedinthisdocumentdoesnotandshould notbeassumedtorepresentthesourceconfigurationoremissions associatedwithanyparticularfacility.

Thissectionprovidesanoverviewofthecontentsofthis document. Itbrieflyoutlinesthenature,extent,andformatof thematerialpresentedintheremainingsectionsofthis document.

Section3ofthisdocumentprovidesabriefsummaryofthe physicalandchemicalcharacteristicsofcadmiumandcadmium compoundsandanoverviewoftheirproductionanduses. A chemicalusetreesummarizesthequantitiesofcadmiumproduced aswellastherelativeamountsconsumedbyvariousenduses. Thisbackgroundsectionmaybeusefultosomeonewhowantsto developageneralperspectiveonthenatureofthesubstanceand whereitismanufacturedandconsumed.

Sections4to7ofthisdocumentfocusonthemajor industrialsourcecategoriesthatmaydischargecadmium- containingairemissions. Section4discussestheproductionof

2-1 cadmiumandcadmiumcompounds. Section5discussesthedifferent majorusesofcadmiumasanindustrialfeedstock. Section6 discussesemissionsfromcombustionsources. Section7discusses emissionsfromselectednonferroussmelting/refiningprocesses, andSection8discussesemissionsfrommiscellaneousproduction processesandmobilesources. Foreachmajorindustrialsource categorydescribed,processdescriptionsandflowdiagramsare givenwhereverpossible,potentialemissionpointsare identified,andavailableemissionfactorestimatesarepresented thatshowthepotentialforcadmiumemissionsbeforeandafter controlsareemployedbyindustry. Individualcompaniesare namedthatarereportedtobeinvolvedwiththeproductionand/or useofcadmiumbasedonindustrycontacts,theToxicRelease Inventory(TRI),andavailabletradepublications.

Thefinalsectionofthisdocumentsummarizesavailable proceduresforsourcesamplingandanalysisofcadmium. Details arenotprescribednorisanyEPAendorsementgivenorimplied foranyofthesesamplingandanalysisprocedures. AppendixA presentscalculationsusedtoderivetheestimated1990 nationwidecadmiumemissions. AppendixBpresentsasummaryof thecombustionsourcetestdata. AppendixClistsnamesand locationsofelectricarcfurnaces,U.S.Portlandcement manufacturers,phosphaterockprocessors,andelemental phosphorusproducers.

Thisdocumentdoesnotcontainanydiscussionofhealthor otherenvironmentaleffectsofcadmium,nordoesitincludeany discussionofambientairlevelsorambientairmonitoring techniques.

Commentsonthecontentorusefulnessofthisdocumentare welcome,asisanyinformationonprocessdescriptions,operating

2-2 practices,controlmeasures,andemissionsthatwouldenableEPA toimproveitscontents. Allcommentsshouldbesentto: Chief,EmissionFactorandMethodologySection(MD-14) EmissionInventoryBranch U.S.EnvironmentalProtectionAgency ResearchTrianglePark,NC 27711

2-3 SECTION3 BACKGROUND

Thissectiondiscussescadmiumanditscompoundsandalloys, theirchemicalandphysicalproperties,andtheircommercial uses. Thesectionalsoprovidesstatisticsoncadmiumproduction anduse. Finally,thesectionpresentsnationwideestimatesof cadmiumemissionsfromthesourcesdiscussedintheother sectionsofthisdocument.

NATUREOFPOLLUTANT

Cadmiumisasoft,ductile,silvery-whitemetal. Itwas discoveredbyStromeyerin1817asanimpurityinzinccarbonate. Table3-1summarizescadmium’schemicalandphysicalproperties.

Whenheatedinair,cadmiumformsafumeofbrown-colored cadmiumoxide,CdO. Otherelementswhichreactreadilywith cadmiummetaluponheatingincludethehalogens,, ,andtellurium. Themetalisnotattackedbyaqueous solutionsofalkalihydroxides.

Cadmiumisslowlyattackedbywarmdilutehydrochloricor sulfuricacidwiththeevolutionofhydrogenbutisrapidly oxidizedtothecadmiumionbyhotdilutenitricacidwith

evolutionofvariousoxidesofnitrogen(NOx). Cadmiumis displacedfromsolutionbymoreelectropositivemetalssuchas

3-1 TABLE3-1.PHYSICALPROPERTIESOFCADMIUM

Property Value

Atomicweight 112.41 Crystalstructure Hexagonal CASregistrynumber 7440-43-9 Atomicnumber 48 2

Outerelectronconfiguration 4d105s2 Metallicradius,Å 1.54 Covalentradium,Å 1.48 Electrodepotential,normal,V CdCd2+ +2e -0.4013

Meltingpoint,°C 321.1 Boiling,point,°C 767 Latentheatoffusion,J/g(cal/g)a 55.2(13.2) Latentheatofvaporization,J/g(cal/g)a 886.9(212) Specificheat,J/mol·K(cal/mol·K)a Solid,20°C 25.9(6.19) Liquid,321°to700°C 29.7(7.10)

Electricalresistivity,µΩ-cm at22°C 7.27 at400°C 34.1 at600°C 34.8 at700°C 35.8 ,kg/m3 at26°C 8,624 atmeltingpoint 8,020 at400°C 7,930 at600°C 7,720

Thermalconductivity,W(m·K),at0°C 98 Vaporpressure,mmHg at382°C 0.7598 at478°C 7.598 at595°C 75.98 at767°C 759.8

Source:Reference1. aToconvertJtocal,divideby4.184.

3-2 zincoraluminum. Thehydroxideofcadmium,Cd(OH)2,is virtuallyinsolubleinalkalinemedia. Thecadmiumionforms stablecomplexeswithammonia,aswellascyanideandhalide .

Elementalcadmiumisusedprimarilyasanelectroplated, corrosionresistantcoatingappliedtoiron,steel,brass, copper,andaluminum. Cadmiumcoatingsareespeciallyusefulfor protectingsurfacesexposedtocorrosivemarineenvironments. An addedadvantageofusingcadmiumsurfacecoatingsisthatcadmium ispreferentiallyattackedbythecorrosiveenvironmentand protectsthebasemetalfromcorrosion. Evenifthecadmium coatingisslightlydamaged,itcontinuestoprovideprotection tothebasemetal.

Elementalcadmiumcoatingsalsohavealowcoefficientof ,goodelectricalconductivity,areeasilysoldered,and havelowvolumecorrosionproducts. Thecoatingsreducegalvanic corrosionbetweensteelandothermetals,particularlyaluminum.

Technicallyandcommerciallyimportantcadmiumcompounds includetheoxide,,selenide,,sulfate,, ,andvariousorganiccadmiumsaltsoffattyacids,such asthepalmitateandstearate. Theonlynaturallyoccurring compoundisthesulfide,CdS(),whichisanaccessory mineralinsulfideoresoflead,zinc,andcopperandinsulfur- bearingcoals.1,2

Cadmiumformsalloyswithmanymetals;thesealloysfall intotwomajorgroups: thoseinwhichcadmiumhelpsreducethe meltingpointandthoseinwhichcadmiumimprovesmechanical properties.1,3,4

3-3 OVERVIEWOFPRODUCTION,USE,ANDEMISSIONS

Thissubsectionsummarizescadmiumproductionstatistics, identifiesindustrialcategoriesusingcadmium,andprovides estimatesofnationwidecadmiumemissions.

Production

Primaryproductionofcadmiumoccursasabyproductof smeltingdomesticandimportedzincconcentratesandofrecovery fromleadsmelterbaghousedust. Therearefourmajorproducing companiesintheU.S.,andthreeproducethecadmiumfrom smeltingzincconcentrates. Onlyonecompanyrecoverscadmium fromtheleadsmelterbaghousedust.5

Figure3-1presentsthe1991supply-and-demanddiagramfor cadmium. Theinformationinthisfigurewasobtainedfromthe U.S.BureauofMines,DivisionofMineralCommodities.5 Asshown inFigure3-1,thetotalU.S.supplyofcadmiumwas4,368Mg (4,805tons). Anestimated38percentofthetotalsupply resultedfromU.S.primaryandsecondaryproductionprocesses, and47percentwastheresultofimports. Theremaining 15percentcamefromproducerstockpiles. Figure3-1alsoshows thatofthetotal1991U.S.cadmiumsupply,74percentwasused tomeetdomesticdemands,while4percentmetexportdemands,and 22percentsuppliedindustrystocks. Exportsofcadmiumarein theformofcadmiummetalandcadmiuminalloys,,flue dust,residues,andscrap.5

TheBureauofMinesreportedU.S.productionof329Mg (362tons)ofcadmiumsulfide(includinglithoponeandcadmium sulfoselenide)and1,089Mg(1,198tons)ofplatingsalts, cadmiumoxide,andothercompounds. Theremaining1,820Mg (2,002tons)ofU.S.demandin1991apparentlywascomprisedof

3-4

cadmiummetal,alloys,andimportedcompounds.5 The1991demand of3,238Mg(3,562tons)showninFigure3-1representsaslight increaseoverthe1990demandlevelof3,107Mg(3,418tons),but lessthanthe1989demandlevelsof4,096Mg(4,506tons).

Use

TheBureauofMinesestimatesthatU.S.consumptionof cadmiumandcadmiumcompoundsoccursprincipallyinthefollowing fiveareas:

1. Batteryproduction; 2. Coatingsandplatings; 3. Pigments; 4. Plasticandsyntheticproducts(primarilyas stabilizers);and 5. Alloysandotherproducts.

Theestimatedpercentageofthetotal1991U.S.cadmium supplythatwasconsumedbyeachend-usecategoryisshownin Figure3-2. Batteryproduction,at45percent,accountsforthe largestpercentageofcadmiumconsumption(1,457Mg/1,603tons). Coatingandplatingoperationswerethenextlargestconsumerat 20percent(648Mg/713tons). Thethirdandfourthlargest consumercategorieswerepigmentsat16percent(518Mg/570tons) andplasticandsyntheticproducts(presumedtobeprimarily stabilizers)at12percent(389Mg/428tons). Thesmallestend- usecategorywasalloysandotherusesat7percent (227Mg/250tons).

Emissions

Twodistinctmethodswereusedtodevelopnationwide emissionestimatesforspecificsourcecategories. Thefirst

3-6

methodinvolveddevelopingsource-specificemissionfactorsand applyingthoseemissionfactorstoestimatesofnationwidesource activitytocalculatenationwidemercuryemissionestimates. The secondmethodreliedonextrapolatingemissionestimatesfromthe ToxicChemicalsReleaseInventorySystem(TRI).6

Cadmiumisemittedfromanumberofindustrialprocesses (e.g.,fossilfuelcombustion,wasteincineration,andmineral processingoperations)becauseitispresentasacontaminantin theprocessfeed. Forthoseprocesses,anemissionfactor-based approachwasusedtoestimatenationwidecadmiumemissions. A comprehensivereviewandanalysisofbothinformationoncadmium contentinthefeedmaterialandemissiontestdatawas conducted. Primarysourcesofinformation,whichwereused includedongoingEPAregulatorydevelopmentactivities, informationthatisbeingcollectedbyEPAtodeveloptoxicair pollutantemissionfactorsinAP-427,andanEPAdatabaseon toxicairpollutantemissionfactors.8 Uponcompletionofthe review,a"besttypical"emissionfactorwasselected. This informationwascombinedwithreadilyavailablepublisheddataon sourcecategoryactivitytocalculatenationwideemission estimates.

Thesourceofemissionsinformationusedforsource categoriesthatinvolvecadmiumusewastheTRIform,requiredby Section313ofTitleIIIofthe1986SuperfundAmendmentsand ReauthorizationAct(SARA313).6 Thissectionrequiresowners andoperatorsoffacilitiesinStandardIndustrialClassification (SIC)codes20-39thatmanufacture,import,process,orotherwise usetoxicchemicalstoreportannualairreleasesofthese chemicals. Theemissionsmaytobebasedonsourcetests(if available);otherwise,emissionsmaybebasedonemission factors,massbalances,orotherapproaches.

3-8 Inselectedcases,facilitiesreportedtoTRIundermultiple SICcodes. Asaresult,itwasdifficulttoassignemissionsto aspecificSICcode. Inthosecases,effortsweremadeto determinetheappropriateSICcodesassociatedwiththe emissions. IfappropriateSICcodescouldnotbeexplicitly identified,thedatawerenotusedintheanalysis.

Table3-2presentsacompilationofSICcodesthathavebeen associatedwithcadmiumemissions.7,8 ThistableliststheSIC codesthatwereidentifiedasapotentialsourceofcadmium emissions,providesadescriptionoftheSICcode,andidentifies otheremissionsourcesthatdonothaveanassignedSICcode.7,8

Table3-3providesasummaryoftheestimated1990 nationwidecadmiumemissionsforthosesourcecategorieswhere adequateinformationwasavailable(i.e.,emissionfactorsand productiondata). AppendixApresentsthedatausedforeachof theseestimates,assumptions,andemissioncalculationsforeach ofthesesourcecategories. Theestimatedemissionswerebased onemissionfactorsprovidedinthisdocumentorcalculatedfrom sourcetestdataandappropriateprocessinformation,if available.

Ofthefivemajorsourcecategories,cadmiumemissions resultingfromcombustionsourcesaccountedforatotalof266Mg (293tons)orapproximately82percentofthetotalestimated emissions. Withinthecombustionsourcecategory,themajor contributortocadmiumemissionswasfromthecombustionofcoal, followedbyoilcombustion,sewagesludge,andmunicipalwaste. Thenonferroussmeltingandrefiningsourcecategoryaccountsfor about38Mg(42tons)orapproximately12percentofthetotal estimatedemissions.

3-9 TABLE3-2.SICCODESOFINDUSTRIESASSOCIATEDWITHCADMIUMEMISSIONS

SICcode Industry 0711 SoilPreparationServices(fertilizerapplication) 266 WovenFabricFinishing 2611 PulpMills 2621 PaperMills 2816 InorganicPigmentsManufacture 2819 IndustrialInorganicCompounds,NotElsewhereClassified(nec) 2851 PaintandAlliedProducts 2869 IndustrialOrganicChemicals,nec(plasticsstabilizers) 2874 PhosphateFertilizers 2879 PesticidesandAgriculturalChemicals,nec(traceelements) 2895 CarbonBlack 2911 PetroleumRefining 2951 AsphaltPavingMixturesandBlocks 3053 Gaskets,Packing,andSealingDevices 3081 UnsupportedPlastic,FilmandSheet 3083 LaminatedPlastics,Plate,Sheet,andProfileShapes 3087 CustomCompoundingofPurchasedPlasticsResins(withCdpigments) 3089 PlasticsProducts,nec 3229 PressedandBlownGlassandGlassware,nec 3241 Cement,Hydraulic(dryandwetprocess) 3264 PorcelainElectricalSupplies 3312 BlastFurnacesandSteelMills 3313 FerroalloyProduction 332 IronandSteelFoundries 3321 GrayandDuctileIronFoundries 3331 PrimaryCopperSmeltingandRefining 3339 PrimarySmeltingandRefiningofNonferrousMetals(zinc,lead,cadmium) 3341 SecondarySmeltingandRefiningofNonferrousMetals(zinc,lead,copper) 3351 CopperRolling,Drawing,andExtruding 3356 NonferrousRollingandDrawing,ExceptCopperandAluminum 3357 NonferrousWireDrawingandInsulating 3362 Brass,Bronze,Copper,Copper-BaseAlloyFoundries 3369 NonferrousFoundries,nec 3365 AluminumFoundries 3399 PrimaryMetalProducts,nec 3431 EnameledIronandMetalSanitaryWare 3340 FabricatedStructuralMetalProducts(diecasting) 3452 Bolts,Nuts,Screws,Rivets,andWashers 3471 PlatingandPolishing(cadmiumelectroplating) 3492 FluidPowerValuesandHoseFittings 3494 ValvesandPipeFittings,nec

3-10 TABLE3-2.(continued)

SICcode Industry 3585 RefrigerationandHeatingEquipment 3691 StorageBatteries 3692 PrimaryBatteries,DryandWet 3694 InternalCombustionEngineElectricalEquipment 3714 MotorVehiclePartsandAccessories 3721 Aircraft 3728 AircraftPartsandAuxiliaryEquipment,nec 3952 LeadPencils,Crayons,andArtists’Materials 4953 RefuseSystems(municipalwastecombustion) 9661 SpaceResearchandTechnology -- CoalCombustion -- GeneralLaboratoryUse -- OilCombustion -- WoodCombustion -- NaturalGasCombustion

Source:References7and8.

3-11 TABLE3-3.ESTIMATED1990NATIONWIDECADMIUMEMISSIONS FORSELECTEDSOURCECATEGORIES

CadmiumEmissions SourceCategory Mg Tons Basis CadmiumProduction CadmiumRefining 4.2 4.6 AppendixA CadmiumPigmentProduction 1.6 1.8 AppendixA CadmiumStabilizerProduction NA NA Noemissionfactors OtherCadmiumCompoundProduction NA NA Noemissionfactors MajorUsesofCadmium CadmiumElectroplating NA NA Noemissionfactors SecondaryBatteryManufacture 0.32 0.35 AppendixA CadmiumStabilizers(Plastics) NA NA Noemissionfactors CadmiumPigments(Plastics) NA NA Noemissionfactors

CombustionSources CoalCombustion 220.2 242.7 AppendixA OilCombustion 23.5 25.9 AppendixA NaturalGasCombustion NA NA Noemissionfactors MunicipalWasteCombustion 8.4 9.2 AppendixA SewageSludgeCombustion 9.9 10.9 AppendixA MedicalWasteCombustion 3.6 3.9 AppendixA WoodCombustion 0.38 0.41 AppendixA NonferrousSmeltingandRefining PrimaryLeadSmelting 14.3 15.8 AppendixA PrimaryCopperSmelting 5.6 6.2 AppendixA PrimaryZincSmelting 5.7 6.3 AppendixA SecondaryCopperSmelting 10.8 11.9 AppendixA SecondaryZincSmelting(scrap) 1.5 1.7 AppendixA SecondaryZincSmelting(EAF) NA NA Noemissionfactors MiscellaneousSources IronandSteel NA NA Noemissionfactors PortlandCementProduction 13.1 14.4 AppendixA PhosphateRockProcessing NA NA Noemissionfactors CarbonBlackProduction 0.07 0.08 AppendixA MobileSources NA NA Noemissionfactors TOTAL 323 356

NA=notavailable

3-12 REFERENCES 1. Hollander,M.L.,andS.C.Carapella,Jr. Cadmiumand CadmiumAlloys. (In)Kirk-OthmerEncyclopediaofChemical Technology. Volume4. M.Grayson,ed. AWiley-Interscience Publication,JohnWileyandSons,NewYork,NY. 1978. 2. Parker,P.D. CadmiumCompounds. (In)Kirk-Othmer EncyclopediaofChemicalTechnology. Volume4. 3rded. M.Grayson,ed. AWiley-IntersciencePublication,JohnWiley andSons,NewYork,NY. 1978. 3. CadmiumAssociation/CadmiumCouncil. TechnicalNoteson Cadmium. CadmiuminAlloys. CadmiumAssociation,London; CadmiumCouncil,NewYork,NY. 1978. 4. CadmiumAssociation. Cadmium. TheQualityMetal. Cadmium Association,London. undated. 5. Llewellyn,T.O. Cadmiumin1991. BranchofIndustrial ,DivisionofMineralCommodities,BureauofMines, U.S.DepartmentoftheInterior,Washington,DC. January1993. 6. U.S.EnvironmentalProtectionAgency. 1990ToxicRelease Inventory,OfficeofToxicSubstances. Washington,DC. January1993. 7. U.S.EnvironmentalProtectionAgency. CompilationofAir PollutionEmissionFactors,AP-42,FourthEdition, SupplementE. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. October1992. 8. XATEF. Crosswalk/AirToxicEmissionFactorDataBase. Version2for1992update. OfficeofAirQualityPlanning andStandards,U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. September1992.

3-13 SECTION4

EMISSIONSFROMCADMIUMPRODUCTION

Thissectiondescribesthepotentialsourcesofcadmium emissionsfromtheproductionofcadmiumandcadmiumcompounds. Thefollowingsubsections,coveringcadmiumrefiningandcadmium oxideproduction,cadmiumpigmentsproduction,cadmiumstabilizer production,andothercadmiumcompounds,presentprocess descriptions,identifypotentialcadmiumemissionsourcesand controls,andquantifycadmiumemissions.

CADMIUMREFININGANDCADMIUMOXIDEPRODUCTION

Cadmiummineralsdonotoccurinconcentrationsand quantitiessufficientenoughtojustifyminingthemintheirown right,buttheyarepresentinmostzincoresascadmiumsulfide (themineralgreenockite)andareconcentratedduringzincore processing.1 Theresultingzincoreconcentratesfromore processingcontainfrom0.1to0.8percentcadmiumbyweight.1 Cadmiummetalisrecoveredaseither:(1)abyproductofthe extractionandrefiningofzincmetalfromzincsulfideore concentratesinelectrolyticzincsmelters;or(2)themain productintheprocessingofleadblastfurnacedusts. Cadmium oxideisproducedinasecondaryprocessusingcadmiummetalas thefeedmaterial.

Table4-1liststhecadmiummetalandcadmiumoxideproducers alongwiththeirlocations,processfeedmaterials,andprocesses

4-1 TABLE 4-1. CADMIUM AND CADMIUM OXIDE PRODUCERS

Cadmium Refining Plant Location Feed Material Process Cadmium Metal ASARCO, Inc. Denver, CO Crude CdO in blast Electromotive Cd furnace baghouse dust from East Helena lead smelter. Big River Zinc Corp. Sauget, IL Zinc roaster calcine. Electromotive Cd Jersey Miniere Zinc (JMZ) Clarksville, TN Zinc roaster calcine. Electrolytic Cd Zinc Corp. of America (ZCA) Bartlesville, OK Zinc roaster calcine.a Electromotive Cd Cadmium Oxide ASARCO, Inc. Denver, CO Cd metal. Air oxidation from retort furnace 4-2 Big River Zinc Corp. Sauget, IL Cd metal. Air oxidation from retort furnace Proctor and Gamble Co. Phillipsburg, NJ Not available. Not available Witco Chemical Co. Brooklyn, NY Cd metal. Air oxidation from retort furnace

aZCA also has an electrothermic primary zinc smelter in Monaca, Pennsylvania. That smelter also uses a crude zinc calcine recovered from (EAF) dusts. The calcine from the EAF dusts also contains cadmium.

Source: References 1-3 used. Currently,therearefourplantsthatproducecadmium metalintheUnitedStates;twoofthesefourplantsalsoproduce cadmiumoxide. Threeofthefourcadmiummetalrefiningplants arecolocatedatelectrolyticzincsmeltersandinclude:Big RiverZinc(BRZ)CorporationlocatedinSauget,Illinois;Jersey MiniereZinc(JMZ)locatedinClarksville,Tennessee;andZinc CorporationofAmerica(ZCA)locatedinBartlesville,Oklahoma. Thefourthcadmiummetalrefiningplant,ASARCO,Inc.,islocated inDenver,Coloradoandprocessesleadblastfurnacedust. CadmiumoxideisproducedatbothBRZandASARCO. (AnotherZCA primaryzincsmelter,inMonaca,Pennsylvania,doesnothavean associatedcadmiumrefinery).2

Reference3listsProctorandGambleCo.asaproducerof cadmiumoxide. However,atthepresenttime,itisnotclear whethercadmiumoxideismanufacturedatthislocationorwhether thecompanyonlydistributescadmiumoxidefromthislocation. CadmiumoxideisalsoproducedatWitcoChemicalCompanylocated inBrooklyn,NewYork. Thisplantproducescadmiumoxideforits ownuseintheproductionofcadmiumstabilizers.

ProcessDescription1

Figure4-1isageneralprocessflowdiagramforthe productionofcadmiummetalandcadmiumoxideatelectrolyticand electromotivecadmiumrefiningplants. Atthethreeelectrolytic zincsmelters,cadmiumisremovedasanimpurityfromthe leachatesolutionoftheroastedzincoreconcentrateorcalcine. Cadmiumisalsorecoveredfromsolutionsobtainedbyleaching leadblastfurnacebaghousedustscontainingimpurecadmiumoxide withaweaksulfuricacidsolution. In1986,thesourceofthe duststreatedbytheASARCOcadmiumrefinerywasASARCO’sEast Helenaleadsmelter. Currently,ASARCOusesGodfreyroaster baghousedustfromtheASARCOElPasoleadsmelter.4

4-3

Thecadmium-bearingfeed(StreamA)isleachedordissolved insulfuricacidinStep1. Next,thesulfuricacidsolution (StreamB)istreatedbyvarioussolutionpurificationsteps (Step2). Thepurifiedsolution(StreamC)istreatedwithzinc dusttoprecipitateametalliccadmium"sponge"(StreamD)(Step 3). Thecadmiumspongeisredissolvedinsulfuricacid;the solutionundergoesadditionalpurificationstepstoproducea purifiedsolution(StreamEorE’;Step4). TheJMZcadmium refineryinClarksville,Tennesseeusestheelectrolyticprocess torecovermetalliccadmiumfromthepurifiedcadmiumsulfate solution(StreamE). Theothercadmiumrefineriesusethe electromotiveprocess.3

Intheelectrolyticcadmiumrefiningprocess(Step5), electrolysisofthepurifiedcadmiumsulfatesolution(StreamE) depositscadmiumoncathodes. Thecadmiummetal(StreamF)is strippedfromtheelectrodesandtransferredtoacadmiummelting furnace(Step7).Themoltencadmium(StreamG)iscastinto ballsandsheetsforcadmiumelectroplatinganodesorcastinto slabs,ingots,andsticksforalloying,pigmentproduction,and cadmiumoxideproduction(Step8).

Intheelectromotivecadmiumrefiningprocess,zincdustis addedtothepurifiedcadmiumsulfatesolution(StreamE’)to displacecadmiumas"sponge"metal(StreamF)inStep6. The spongeisbriquetted,melted(Step7),andcast(Step8)into productsforsaleorfurtherprocessing.

JMZandZCAproduceonlycastcadmiummetalproducts. ASARCO andBRZalsoproducepowderedcadmiummetal,cadmiumoxide,or both. Cadmiumfromthemeltingfurnace(StreamG)istransferred toaretortfurnace(Step9or11). Inpowderedcadmium production(Step9),cadmium(StreamG)isroutedtoasealed

4-5 retortthathasbeenpurgedofoxygenwithcarbondioxide. Cadmiumvaporizesandcondensesasapowder(StreamI)during retortingintheabsenceofoxygen. Thecondensedpowderis packagedinStep10. Incadmiumoxideproduction(Step11), retortinginairoxidizescadmiumtocadmiumoxide,whichis collectedinabaghouse(Step12)andpackaged(Step13).

EmissionsandControls

Duringcadmiumandcadmiumoxideproduction,cadmiumis emittedfrommeltingfurnaces(Step7),retorting(Steps9and 11),castingandtapping(Step8),andpackaging(Steps10 and13). ChargingtheleachtanksforStep1withleadblast furnacedusts(StreamA)andsolutionsheatingtanks(Step4)at theDenverrefinerywereadditionalsources.

In1986,theEPAinventoriedcadmiumemissionsourcesat cadmiumrefiningplantsbasedonSection114responses,emission testreports,tripreports,andapreviouscadmiumsourcesurvey publishedin1985.5 Initiallygeneratedemissionsestimateswere revisedbasedonindustrycomments. Table4-2showsthecadmium emissionratesdevelopedin1986fornormalandmaximumoperation atthreeofthefourcadmiumrefiningplants. Revisedestimates fornormaloperationsusingthesamegeneralemissionestimation methodologyattheBRZplantweremadein19896 andarepresented inthetableinplaceofthe1986data. However,thedatashown formaximumoperationarefromthe1986study. Revisedemission estimatesfortheASARCOplantweredevelopedin1992forthe StateofColoradobyJACACorporation.4 TheASARCOfacilityhad undergonesubstantialmodificationssincethe1986study. The JACAstudydevelopedemissionestimatesformaximumoperation only. ThesedataarereportedinTable4-2. Emissionestimates fornormaloperationfromthe1986studyarenotreportedin Table4-2becauseoftheprocessandcontrolmodifications

4-6 TABLE4-2.INVENTORYOFCADMIUMEMISSIONSOURCESAND CONTROLSFORCADMIUMREFININGPLANTS

Emissions,kg/yr Maximum Normal opera- opera- Control Plant Source Typea tion tion deviceb BigRiverZinc (7)Cadmiummeltingfurnace H 70 0UNC,L Company,Sauget,IL *Cadmiumholdingfurnace H 93 76UNC (8)Cadmiumcastingfurnace+ F68<1BH tapping/casting (11)Cadmiumoxidefurnace H 814 661 BH TotalSauget 1,045 737

JerseyMiniereZinc, (7-8)Cadmiummelting/casting H <1 <1BH,L Clarksville,TN furnace

ZincCorp.of (7-8)Cadmiummelting/casting H<1<1WS America,Bartlesville, furnace OK (8)Cadmiumtapping/casting H <1 <1UNC

ASARCOGlobe, (1)Hotwaterleaching Denver,CO Dustcharging H 8.8 --- UNC Reaction/Filtration H 6.1 --- UNC (1)Acidleaching H 4.3 ---UNC (4)Purification H 6.1 ---UNC (4)Solutionsheating Chargingandpumpout H 1.7 --- UNC Heating H 158.5 --- UNC

(4)SpongeProduction H 2.1 --- UNC *Premelt H 12.7 ---BH RetortBuilding (12)CdOproduction H 56.6 ---BH (10)CdOpackaging H 11.6 ---BH (10)CdOpackaging H 23.2 ---BH (10)Cdpackaging H 11.2 ---BH *,hoods,Cdmelting,Cd H 70.8 ---BH condenser *Fugitiveemissions(roadways,by- F<3 ---DS productstoragepiles) TotalDenver 377 aH=pointsource;F=fugitivesource. bUNC=uncontrolled;BH=baghouse;WS=wetscrubber;DS=dust suppressant;L=layerofcausticonmoltencadmium. cBasedonAP-42methodologydevelopedforaggregatematerials.

*SourcenotdepictedinFigure4-1. Source:References1,4,and6V

4-7 implementedatASARCOsincethen,becauseoftheuseofnew emissiontestdatafromtestsconductedsince1986,andbecause theemissionestimationmethodologiesusedin1986versusthe 1992JACAstudyaresubstantiallydifferent. Theemission sourcesarenumberedusingthesamenumberingschemeasin Figure4-1. Cadmiummeltingfurnacesandcadmiumretortfurnaceswere identifiedasthetwotypesofprocessemissionsourcesat cadmiumrefiningplants. Cadmiummeltingfurnacesareusedto melteithercadmiumspongeorsheets.1 Alayerofcausticonthe moltenmetalsurfaceisusedtopreventoxidationofthemetal, tohelpremoveimpurities,andtoprovidesomecontrolof particulatematteratthreecadmiumrefiningplants.1 Theother plant(JMZ)usesalayerofresintoachievethesameresults.1 Processcadmiumemissionsfromthemeltingfurnacearecontrolled byabaghouseatJMZandbyawetscrubberatZCA.1 Ahooding systemductsfugitiveemissionsfromthecharging/drossingport andfromthetapping/castingareatothebaghouseatJMZ.1 At ASARCO,forcedventilationisinplaceduringfurnaceoperation andduringchargingandtapping/casting.1 Sincethe1986study, BRZmadeseveralimprovementsintheircadmiumrefiningprocess. Amongtheseimprovements,wastheductingofthecadmium tapping/castingareaofthecadmiummeltingfurnacetothe existingleadanodefurnaces’processfugitiveemissions baghouse.6

CadmiumretortfurnacesareusedonlyatBRZandASARCOin theproductionofcadmiumoxideand/orcadmiummetal.1 Emissions fromthesesourceswereestimatedusingemissiontestdatafrom theseplants.1 InOctober1988,BRZalsomadeimprovementsto thecadmiumoxideproductcollectionsystem.6 Thechanges includedanewproductcollectionbaghouse,anewventilation systemandfugitiveemissionsbaghouse,andanenclosedand automatedcadmiumoxidepackagingoperation. Whilethesechanges

4-8 enhancedtheoperationofthecadmiumoxidesystem,itwas assumedthatnoreductionsinemissionswererealized. Thenew baghousehasthesameoperatingparametersastheonethatwas replaced. Therefore,thetestconductedin1986wasstill consideredtobevalidandwasusedtodevelopemission estimates. Additionally,fugitiveemissionsfromcadmiumoxide productionandpackaginghadbeenassumedtobenegligiblein 1986becauseatthattime,thecadmiumoxideproductionand packagingoperationswerehousedinaseparateroomwithinthe cadmiumbuilding;thenewventilationsystemprobablyimproves workingconditionsinsidethecadmiumoxideproductionand packagingareas. ASARCOhasalsoimprovedoperationsatthe Denverlocationwiththeadditionofabaghousetocontrol emissionsfrompremeltoperationsandanotherbaghousetocontrol fugitiveemissionsfromtheCdfurnaces,hoods,Cdmelting operations,andtheCdcondenserintheretortdepartment.4

Becausethreeofthefourcadmiumrefineriesareoperatedin conjunctionwithzincsmelters,theannualemissionsreportedby theseplantsinthe1990ToxicChemicalsReleaseInventory(see Table4-3)comprisethesumofbothsources.7 The860kg (1,896lb)cadmiumemissionreportedbyBigRiverZinc Corporationisslightlymorethanthesumoftheestimatesin Reference4forthecadmiumrefiningoperations(737kg)and primaryzincsmelting(100kg). TheJMZplantreported227kg (500lb)cadmiumemissionsin1990comparedtothe1986estimates of<1kgforthecadmiumrefineryand28kgforthezincsmelter. TheZCAplantreportedatotalof2,936kg(6,472lb)cadmium emittedin1990,whereasthe1986estimateswere<1kgforthe refineryand12kgforthesmelter. TheASARCOcadmiumrefining plantreported180kg(396lb)cadmiumemissions(177kg[391lb] frompointsources)in1990.7 Thedifferencesbetweenthese estimates(1990TRIandreferences1,4,and6)arelikelythe resultofdifferencesinproductionandintheassumptionsused

4-9 TABLE4-3.PRIMARYZINCANDCADMIUMPRODUCERSREPORTINGCADMIUM EMISSIONSINTHE1990TOXICCHEMICALSRELEASEINVENTORY

Emissions,kg(lb) Monitoring Plant Nonpoint Point Total Data

AsarcoInc,GlobePlant, 2 177 180 no Denver,Colorado (5) (391) (396) (Cadmiumrefineryfromleadsmelter dusts) BigRiverZincCorp., 113 747 860 no Sauget,Illinois (250) (1,646) (1,896) JerseyMiniereZinc, 113 113 226 yes Clarksville,Tennessee (250) (250) (500) (point) ZincCorporationofAmerica, 2 2,934 2,936 no Bartlesville,Oklahoma (4) (6,468) (6,472)

TOTAL 230 3,971 4,202 (509) (8,755) (9,264) aCurrentlynotacadmiumrefiner.

Source:Reference7.

4-10 todeveloptheemissionestimates. Forexample,atBRZ,the plantpersonnelstillusetheemissionfactorsdevelopedfromthe 1986studytodeveloptheiremissionestimatesandmultiplythese factorsbytheproductionlevelsfortheparticularyear. Also, becausethelevelsofcadmiuminthezincsulfideoreresidues varybyalmostanorderofmagnitude,theresultingemission estimatescouldalsovarysignificantly.

Table4-4providesemissionfactorsforthecadmiumrefining plantusingleadblastfurnacedusts. Theseemissionfactors weredevelopedfromthe1992JACAstudybyusingthemaximum annualemissionsfromTable4-2anddividingbythemaximum productionratedataasnotedinthefootnotesinTable4-4.

Emissionfactorsforthecadmiummetalandcadmiumoxide productionprocessesarepresentedinTable4-5. Theseemission factorsweredevelopedbasedonemissiontestsconductedattwo ofthecadmiumrefiningplants.

CADMIUMPIGMENTSPRODUCTION

Cadmiumisemittedduringthemanufactureofcadmium pigments. Thissubsectionwilldescribethemanufacturing process,emissions,andcontrols. Mostoftheinformationherein isfroma1988EmissionStandardsDivisionreportoncadmium emissionsfrompigmentandstabilizermanufactureandthe1985 BackgroundInformationDocumentforCadmiumEmissionSources.5,8

Cadmiumpigmentsarestableinorganiccoloringagentsthat providearangeofbrilliantshadesofyellow,,,and maroon. Thepigmentsarebasedoncadmiumsulfide(CdS),which yieldsagoldenyellowpigment. Partialsubstitutionofcadmium inthecrystallatticebyzincormercury,andofsulfurby selenium,producesaseriesofintercrystallinecompoundsmaking

4-11 TABLE4-4.CADMIUMEMISSIONFACTORSFORCADMIUMREFINING PLANTUSINGLEADBLASTFURNACEDUST

Emissionfactor Processstep(emissiontypea) lb/tonCd kg/MgCd produced produced Controlb (1)Hotwaterleaching Dustcharging(H) 0.0297c 0.0149c UNC Reaction/filtration(H) 0.0207c 0.0103c UNC (1)Acidleaching(H) 0.0146c 0.0073c UNC (4)Purification(H) 0.0207c 0.0103c UNC (4)SolutionHeating Charging&pumpout(H) 0.0029c 0.0057c UNC Heating(H) 0.535c 0.268c UNC (4)SpongeProduction(H) 0.007c 0.0035c UNC *Premelt(H) 0.0429c 0.0214c BH RetortBuilding (12)CdOproduction(H) 0.1333d 0.0667d BH (10)CdOpackaging(H) 0.0274d 0.0137d BH (10)CdOpackaging(H) 0.0546d 0.0273d BH (10)Cdpackaging(H) 0.071e 0.0355e BH *Furnaces,hoods,Cdmelting,cdcondenser 0.1216f 0.0608f BH (H) *Fugitiveemissions(roadways,by-product 0.0051f 0.0026f DS storagepiles)(F) aH=pointsource,F=fugitivesource b UNC=uncontrolled,BH=baghouse,WS=wetscrubber,DS=dustsuppressant cEmissionfactorscalculatedbydividingtheannualemissionsfromTable4-2bythemaximum annualthroughputthroughtheleaching,solutions,andpremeltdepartmentof592.2MgCd/yr (652.8tonsCd/yr). dEmissionfactorscalculatedbydividingtheannualemissionsfromTable4-2bythemaximum annualthroughputofCdOof849.1MgCd/yr(936tonsCd/yr). eEmissionfactorscalculatedbydividingtheannualemissionsfromTable4-2bythemaximum annualthroughputofCdpowderof315.7MgCd/yr(348tonsCd/yr). fEmissionfactorscalculatedbydividingtheannualemissionsfromTable4-2bythemaximum plantthroughputof1,164.8MgCd/yr(1,284tonsCd/yr).

*StepnotdepictedinFigure4-1.

Source:DevelopedfrominformationfromReference4.

4-12 TABLE4-5.EMISSIONFACTORSFORCADMIUMANDCADMIUMOXIDEPRODUCTION

EmissionFactor

b b Control/Basis Processa lb/tonCd kg/MgCd (7)Cadmium 0.149c 0.075c Uncontrolled.Emission meltingfurnace factorbasedonsource test (8)Cadmium 0.00149 0.00075 Partlyenclosedhood tapping/casting ductedtobaghouseor wetscrubber.Emission factor *Cadmium 0.149 0.075 Uncontrolled,Emission holdingfurnace factor (11)Cadmium 1.9x10-4 9.8x10-5 Baghouse.Basedon oxidefurnace workroomairsampling. fugitive emissions (includes packaging fugitives) (11)Cadmium 1.30 0.651 Estimatebasedon oxidefurnace productcollection baghousetest.Method 5samplingJune1986 (n=3).Estimatedfor 1988prodn. aNumberscorrespondtothestepsinFigure4-1.Anasteriskdenotesastepnotshowninthe figure. bBasedontheamountofcadmiumprocessed.Forexample,forBigRiverZincCorp.,1,016Mg (1,120tons)cadmiumwasprocessedin1988toproduce1,110Mg(1,220tons)cadmiumoxide. (About5percentoftheCdOproducedwasrejected.)Allcadmiummetalproducedwasusedto produceCdO. cDidnotuseinestimatingBRZemissionsfromthissource.Assumednegligiblebecauseoflayer ofcaustic.

Source:Reference6.

4-13 uptheintermediatecolorsinthelemon-yellowtomaroonrangeof colors. Table4-6liststhemostcommoncadmiumpigments producedin1991. Table4-7liststhecurrentcadmiumpigment producers.

Cadmiumpigmentshaveexcellentthermalstabilitywhichmakes themessentialforuseinhigh-temperatureprocessing,orwhere highservicetemperaturesareencountered. Mostcadmiumpigments areusedinplastics,buttheyarealsousedinpaints,coatings, ceramics,,andtoalesserdegreeinrubber,paper,and inks.9

Therearetwobasictypesofcadmiumpigmentsproducedinthe UnitedStates: purepigments,basedoncadmiumsulfideor cadmiumselenide;andlithophonepigments,whicharepurecadmium pigmentsthathavebeendilutedwithbariumsulfate. Thepure pigmentsareusedundilutedwhenlowpigmentloadingsaredesired asincolorconcentratesforplastics. Lithoponeshaveonly one-halfthetintingpowerofpurepigments,butwhenhigh pigmentloadingscanbetolerated,lithoponesoffertinting strengthandhidingpowercomparable,onanequalcostbasis, withthepurepigments. Theirgreatestuseisinthecoloringof plasticswithdryblends.8

ProcessDescription5,8,9

Cadmiumpigmentproductionisbasedonagenericprocess, illustratedinFigure4-2. However,cadmiumpigment manufacturershavedevelopeddifferingproprietaryproceduresfor creatingpigments,withspecifichuesandproperties. These proprietaryproceduresincludevaryingtypesandpercentagesof ingredients,alteringthecalcinationtime,andaddingor deletingfiltration,washing,drying,blending,orgrinding operations.

4-14 TABLE4-6.COMMONCADMIUMPIGMENTSPRODUCEDIN1991

C.I.Pigment Name

Orange20 Cadmiumsulfoselenideorange Orange20:1 Cadmiumsulfoselenidelithophone orange Red108 Cadmiumsulfoselenidered Red108:1 Cadmiumsulfoselenidelithophonered Yellow35 Cadmiumzincsulfideyellow Yellow35:1 Cadmiumzincsulfidelithophone Yellow37 Cadmiumsulfideyellow Yellow37:1 Cadmiumsulfidelithophoneyellow

Note:EachpigmentismanufacturedbyallcompanieslistedinTable4-7,exceptOrange20:1, whichisnotproducedbytheNewJerseyplant.

4-15 TABLE4-7.CURRENTCADMIUMPIGMENTPRODUCERS

Companyname Location

EngelhardCorporation, Louisville,KY PigmentsandAdditivesDivision (FormerlyHarshaw/FiltrolPartnership) FerroCorporation,Coatings,Colors, Cleveland,OH andElectronicMaterialsGroup,ColorDivision HansonIndustries,SCMChemicals,Inc., Baltimore,MD Subsidiary UniversalFoodsCorporation,Warner-JenkinsonCompany, SouthPlainfield,NJ H.K.ColorGroup, (FormerlyH.KohnstammandCompany)

Source:Reference3.

4-16

Thesourceofcadmiumforcadmiumpigmentproductionisa

puresolutionofeithercadmiumsulfate,CdSO4,orcadmium nitrate,Cd(NO3)2. Cadmiumsulfateismorecommonlyused. These solutionsareeitherpurchasedinbulkorproducedon-siteby dissolvingcadmiumoxide,cadmiummetal,orcadmiumsponge(a porous,high-surface-areaformofcadmiummetal)(StreamA)in theappropriateacid(StreamB). Zincsalts(StreamC)maybe

addedtothedissolver(Step1). TheCdSO4 solution(StreamD) isthenroutedtoaprecipitationreactor(Step2or2’)and mixedwithvaryingquantitiesofanaqueoussolutionofsodium sulfide(orotheralkalisulfide,dependingonthedesiredcolor) (StreamE). ThisprecipitatesCdSincrystallographicform. To formpigmentswitharedshade(cadmiumsulfoselenides),the cadmiumsulfatesolution(StreamE)isreactedwithanalkali sulfide-selenide(StreamG). Redscanalsobeproducedbyadding mercuricsulfide(StreamF)totheprecipitationreactor.

Cadmiumpuretonepigmentproduction(noBaSO4)isdepicted inthefigurebythepathincorporatingsteps1,2’-5’,6’’,and 7’’. LithoponeproductionisrepresentedinFigure5bytwo paths,Steps1-7andSteps1,2’-8’. Toformlithopones, iseitheraddedtotheprecipitationReactor2asbariumsulfide (StreamH)oraddedtothemechanicalblender(Step6’)asbarium

sulfate(streamH’). BaSO4 precipitatesalongwithCdSorCd(S, Se)inReactor2.

Whenthebatchprocessprecipitationreactioniscomplete,

theCdSorCd(S,Se)precipitates(StreamIwithBaSO4;StreamI’

withoutBaSO4)arefilteredfromthesolution,washed,anddried inSteps3or3’and4or4’. Theveryfine,colored particulates(StreamJorJ’)donotyetpossesspigment properties. Thecolorsandpropertiesofthepigmentsdevelop duringtheircalcination,orroasting. Inthecalcination

4-18 process(Step5or5’),thedriedpigmentprecipitatematerial (StreamJorJ’)istransferredtoafurnaceandheatedto between550°and650°C(1022°to1202°F). Thisconvertsthe pigmentmaterialfromacubictoamorestable,hexagonalcrystal structure. Inanalternativeroutetolithoponepigments,the cadmiumpuretonepigmentproducedbyStep5’maybeblendedin Step6’withbariumsulfate(StreamH’). Thecalcinedpigment (StreamK,K’,orK’’)isthenwashedwithhydrochloricacidto removeanyremainingsolublecadmiumparticles. Theproductis thenwashedwithwater,filtered,anddried(Step6or7’for lithopones;Step6’’forpuretonepigments). Thecadmium pigmentemergesasafiltercake,whichiseithergroundand packagedasthefinalproduct,orfurtherprocessedbeforefinal packaging. Thefine,discretepigmentparticleshavediameters ofabout1µm(range0.1to3.5µm).

EmissionsandControls

Cadmiumispotentiallyemittedfromthedissolver(Step1), theprecipitationreactor(Step2,2’),thedryer(Step4,4’) thecalciningfurnace(Step5or5’),theblender(Step6’),and finalproductpackaging(Step7,8’,7"). Calciningemissions arethelargestsourceofcadmiumintheformofCdS,Cd(S,Se), orCdpigment(25percentCdinlithopone;65percentinpureCd pigment). Standardparticulatematteremissioncontrolsare used.

ReactorchargingforCdSO4 production(Step1)(attwo plants)istypicallyuncontrolled,thoughitiscontrolledbya low-energywetscrubberattheLouisville,Kentucky,plant. Calciningoperationsaregenerallycontrolledbywetscrubbers. Dryingoperationsaremostoftenuncontrolled;theLouisville plantcontrolleddryeremissionswithalow-energywetscrubber. Traydesigndryershavelowcadmiumemissions. Grinding,

4-19 blending,andpackagingoperationsaregenerallycontrolledby baghouses. Fugitiveemissionsoccurringinsidebuildingsduring thetransferandhandling(loading,unloading)ofcadmium- containingmaterialsaretypicallycapturedbyhoodsandducted toacontroldevice. Packagingemissionsarelow.5,8

Table4-8listscadmiumemissionsreportedbytheinorganic pigmentplantsinthe1990ToxicChemicalsReleaseInventory.7 Theestimatesfromthe1990Inventoryforthreeofthefour currentcadmiumpigmentproducers(denotedbyfootnote"a"in Table4-8)addupto0.83Mg(1,838lb).7

Theresultsoftwostudiesofthecadmiumpigmentindustry werepublishedinreferences4and6. Bothofthesestudies estimatedcadmiumemissionsanddevelopedcadmiumemission factorsfromindividualsourcesateachofthefourplants identifiedtobeproducingcadmiumpigments. Eachofthefour plantsprovidedinformationthroughSection114information requests. Additionally,sitevisitsweremadetothreeofthese plantsandemissiontestswereconductedattwooftheplants. Throughoutthesestudies,theindividualplantsclaimedprocess descriptionsandallprocessdatatobeconfidentialbusiness information(CBI). Asaresult,emissionsfromeachplantare presentedinthesereferencesastotalcadmiumemissionsinstead ofbyindividualemissionsource. Becausetheproductiondatais CBI,emissionfactorscannotbedeterminedforanysourcesin thissourcecategory.

CADMIUMSTABILIZERSPRODUCTION

Cadmiumisemittedduringthemanufactureofcadmium stabilizers. Thissubsectionwilldescribethemanufacturing process,emissionsources,andemissioncontrols.

4-20 TABLE4-8.INORGANICPIGMENTSMANUFACTURERSREPORTINGCADMIUM EMISSIONSINTHE1990TOXICCHEMICALSRELEASEINVENTORY

Emissions,kg(lb) Monitoring Plant Nonpoint Point Total Data CPChemicalsInc. 113 113 226 no Sumter,SouthCarolina (250) (250) (500) DrakenfeldColors, 10 136 146 yes Ciba-GeigyPigmentsDivision, (23) (300) (323) (nonpoint) Washington,Pennsylvania FerroCorp., 113 45 158 no Cleveland,Ohioa (250) (98) (348) FerroCorp., 113 5 118 no Pittsburgh,Pennsylvania (250) (10) (260) EngelhardCorp. 7 590 597 no Louisville,Kentuckya (15) (1,300) (1,315) JohnsonMattheyInc. 113 113 226 no WestChester,Pennsylvania(alsoreported (250) (250) (500) underseveralotherpossibleSICs,including 3341,secondarynonferrousmetals) SCMGlidcoOrganicsCorp. 07979 no (HansonIndustries), (175) (175) Baltimore,Marylanda TOTAL 470 1,081 1,551 (1,038) (2,383) (3,421) aCurrentcadmiumpigmentmanufacturerinTable4-7.

Source:Reference7.

4-21 Cadmium-containingstabilizersareusedtoarrestthe degradationprocessesthatoccurinpolyvinylchloride(PVC)and relatedpolymerswhenexposedtoheatandultravioletlight (sunlight). Cadmium-basedstabilizersareusuallypreparedby mixingbarium,lead,orzincorganicsaltswithcadmiumorganic salts. Theproductsarehighlyeffective,long-lifestabilizers withnoadverseeffectonPVCprocessing. Cadmiumstabilizers alsoensurethatPVCdevelopsgoodinitialcolorandclarity, allowhighprocessingtemperatures,andensurealongerservice lifeforthePVC. Thestabilizerscontain1to15percent cadmium;thestabilizedPVCcontainsabout0.5to2.5percent cadmium.9

ProcessDescription

Cadmiumstabilizerproductioncanbeahighlyvariable processbecausemanyofthestabilizersarecustomblendedfor specificapplications.

Liquidstabilizers(1to4percentcadmium)areproducedby dissolvingcadmiumoxideinaheatedsolutionoftheappropriate, long-chainfattyacid(e.g.,2-ethylhexanoic[forcadmium octoate]ordecanoic)andaninertorganicsolvent. Afterthe slowacid-basereaction,thesolutionisheatedtodriveoffthe waterproduced. Theremainingproductisfilteredandthe cadmiumsoapsolutionispackagedindrumsforsale. In1983, liquidstabilizersrepresentedabout67percentofthecadmium stabilizermarket. However,powderedstabilizerproduction offersmoreopportunitiesforcadmiumemissions.5,8

Aprocessflowdiagramformanufactureofpowderedcadmium stabilizersisillustratedinFigure4-3. Thereactantsare preparedbytreatingtheappropriateorganicacid(e.g.stearic orlauricacid)withcausticsoda(Na2CO3)toproduceasoluble

4-22 sodiumsoap(StreamA). Acadmiumchloridesolution(StreamB) ispreparedbydissolvingcadmiummetalorCdOinhydrochloric . Thesodiumsaltoftheorganicacid(thesolublesoap) (StreamA)isaddedtothecadmiumchloridesolution(StreamB) inthecadmiumreactor(Step1)atanelevatedtemperature (providedbyadditionofsteam)inthepresenceofacatalystto precipitatethecadmiumsoap(StreamC). Step2probably involvestheadditionofthebariumorganicsaltoritsprecursor reactants. Theresultantslurryisroutedtoacentrifugeand dewatered(Step3). Thesolidsoapiswashed,dried,possibly blended,andpackaged(Steps4-7). Thefinalpowderedstabilizer productcontains7to15percentcadmium. Additivesand moisteningagentscanbeblendedwiththesoapasnecessaryto produceaparticularendproduct. Thenumberandsequenceof blending,grinding,andpackagingoperationsvarywiththefinal product.8

EmissionandControls

Cadmiumemissionsourcesandcontrolsduringcadmium stabilizermanufacturearesummarizedinthissection. The chargingofpowderedcadmiumoxidetotheorganicacidsolution isapotentialcadmiumemissionsourcefromliquidcadmium stabilizerproduction. Thisprocessistypicallycontrolledbya wetscrubber. Fugitiveemissionsarecapturedbyhooding,which iseitherductedtobaghousesorventedtotheair.5,8

Potentialcadmiumemissionsourcesduringpowderedstabilizer productionincludecadmiumoxideproduction,chargingcadmium oxidetothereactor(Step1),drying(Step5),blending(Step 6),grinding(atonefacility),weighing,andpackaging(Step7) ofthefinalproduct. Thecadmiumoxideproductionprocess(one facility)iscontrolledbyabaghouse. Reactorsarecontrolled bywetscrubbers. Dryingoperationsaregenerallyuncontrolled.

4-24 Grinding,blending,weighing,andpackagingoperationsare controlledbyhoodingandbaghouses.5,8

Table4-9liststhemanufacturersoforganicchemicalswho reportedcadmiumemissionsinthe1990ToxicChemicalsRelease Inventory. Thetotalforallplantsproducingstabilizersin Table4-9is3.33Mg(3.67tons).

Forsimilarreasonsdescribedaboveforcadmiumpigment manufacturers,emissionfactorscouldnotbecalculatedfor individualprocesssteps.

OTHERCADMIUMCOMPOUNDPRODUCTION

Theproductionprocessesusedtoproducecadmiumpigmentsand stabilizers,CdS,CdSO4,andcadmiumoxidehavebeendescribed above. Ratherthandescribetheproductionofalargenumberof othercadmiumcompounds,productionprocessesaredescribedonly forafewoftheothercompoundswhosemanufacturersreported cadmiumemissionsinthe1990ToxicChemicalsReleaseInventory.7 Thecadmiumcompoundsdescribedandtheirusesarelistedin Table4-10.

ProcessDescriptions

Cadmiumhydroxide,Cd(OH2),isproducedbyaddingasolution

ofcadmiumnitrate,Cd(NO3)2,toaboilingsolutionofsodiumor potassiumhydroxide(NaOHorKOH). Thecadmiummetal,oxide, hydroxide,orcarbonateisdigestedwithnitricacidfollowedby crystallizationtoproducecadmiumnitrate.12

Hydratedamorphouscadmiumcarbonate,CdCO3,isprecipitated fromcadmiumsaltsolutionsbyaddingsodiumorpotassium . Heatingamorphouscadmiumcarbonatewithammonium

4-25 TABLE4-9.MANUFACTURERSOFORGANICCOMPOUNDSREPORTINGCADMIUM EMISSIONSINTHE1990TOXICCHEMICALSRELEASEINVENTORY

Emissions,kg(lb)

Monitoringdata Plant Nonpoint Point Total

*AkzoChemicalInc.,InterstabDiv., 0 226 226 no NewBrunswick,NewJersey(compounds (500) (500) BaCdstabilizers) *ArgusDivision,Witco 2,180 342 2,522 no Corp.,Brooklyn,NewYork (4,805) (755) (5,560) (BaCdvinylheatstabilizers) *FerroCorporation, 113 113 226 no Bedford,Ohio (250) (250) (500) (Cadmiumoctoate,PVCstabilizers) Rohm&HaasDelaware 2 02yes Valley,Inc.,Bristol,Pennsylvania(also (5) (5) (nonpoint) reportedunderSIC2821,Plastics materials,andresins) *SyntheticProductsCo., 1 113 114 no Stratford,Connecticut (1) (250) (251) (compoundsBaCd stabilizers) *SyntheticProductsCo., 113 113 225 no Cleveland,Ohio (250) (250) (500) (Cadmiumstearate) *VanderbiltChemicalCorp., 1 13 14 no Bethel,Connecticut (3) (28) (31) (Cadmiumdiethyldithio-carbamate)

TOTAL 2,410 920 3,330 (5,314) (2,033) (7,347)

*Assumedtobecadmiumstabilizermanufacturers.

Source:References7and10.

4-26 TABLE4-10.OTHERCADMIUMCOMPOUNDSANDTHEIRUSES

Compound Uses

Cd(OH)2 Usedtopreparenegativeelectrodesfornickel-cadmiumbatteries.

Cd(NO3)2 Impartsareddish-yellowlustertoglassandporcelainware.

CdCO3 Startingcompoundtoproduceothercadmiumsalts.

Cd(CN)2 Usedincopperbrightelectroplating;byproductofcadmiumelectroplating.

CdCl2 Usedinphotography,dyeing,calicoprinting,andsolutionstoprecipitate .

CdI2 Usedinphotographyandprocessengraving. CdTe Usedforsemiconductorsandphotoconductorsinsolarcells,andinfrared, nuclear-radiation,andgamma-raydetectors.

CdSe Usedforsemiconductorsandphotoconductors;notedforfastresponsetime andhighsensitivitytolongerwavelengthsoflight.

Source:Reference11

4-27 chlorideat150°to180°C(302°Fto356°F)intheabsenceof oxygengivesthecrystallineform. Anhydrouscadmiumcarbonate ispreparedbyaddingexcessammoniumcarbonatetoacadmium chloridesolutionfollowedbydryingtheprecipitateat100°C (212°F).12

Cadmiumcyanidecanbeformedin-situ asaby-productby dissolvingcadmiumoxideinexcesssodiumcyanideelectroplating solution.12

Hydratedcadmiumchloride,CdCl2.5H2Ocanbepreparedby reactionsinaqueoussolutionbetweenhydrogenchloride,HCl,and

cadmiummetaloracompoundsuchasCdCO3,CdS,CdO,orCd(OH)2. Thereactionsolutionisthenevaporatedtorecovercrystalsof thehydratedsalt.12

Anhydrouscadmiumchloride,CdCl2,maybepreparedbyseveral methods:12

1.Refluxingthehydratewiththionylchloride,SOCl2.

2.Calciningthehydrate(removesH2O)inanatmosphereofHCl gas.

3.Chlorinatingdrycadmiumacetate,CdO2CCH3,withacetyl chloride,CH3COCl,inglacialaceticacid(CH3CO2H)

4.Mixinghydratedcadmiumnitrate,Cd(NO3)2.4H2O,withhot

concentratedhydrochloricacidandremovingCdCl2 bydistilling thesolution. 5. Treatingcadmiummetalwithchlorinegas. 6. Treatingcadmiummetalwithhydrogenchloridegas.

Cadmiumiodide,CdI2,ispreparedbydissolvingcadmiummetal

oracompoundsuchasCdO,Cd(OH)2,orCdCO3 inhydroiodicacid

4-28 (HI). Thebetaformisrecoveredbyslowcrystallizationfrom solutionsorfromfusedsaltmixtures.12

Cadmiumtelluride,CdTe,maybeproducedbyoneofthree processes:12

1.Combiningelementalcadmiumandtelluriumathigh temperatures 2.Treatingsolutionsofcadmiumsaltswithhydrogentelluride gas(H2Te) 3.Treatingsolutionsofcadmiumsaltswithanalkali (e.g.,withNa2TeorK2Te).

Producersofcadmiumchemicalsnotdiscussedinother subsectionsarelistedinTable4-12.

EmissionsandControls

Informationonemissionsandcontrolsinplaceatthese plantswasnotreadilyavailable. Cadmiumemissionsmightbe expectedfromprocessesinwhichsolutionsareheatedandin dryingoperations. However,onlythreeoftheproducerslisted inTable4-11werefoundinthe1990ToxicChemicalsRelease Inventory,andonlyoneofthemreportedcadmiumemissions(see Table4-12).7 Emissionfactorswerenotfoundforproductionof thecadmiumchemicalsdiscussedinthissubsection.

4-29 TABLE4-11.CADMIUMCOMPOUNDMANUFACTURERS(OTHERTHAN CADMIUMOXIDE,PIGMENTS,ANDSTABILIZERS)

Producer/location Compounds

AirProductsandChemicals, (C2H5)2Cd(diethylcadmium) Inc.,IndustrialGasesDivisionSpecialty

GasDept.,Hometown,PA (CH3)2Cd() ASARCO,Inc. CdS Denver,CO

J.T.Baker,Inc. CdCl2,CdSO4 Phillipsburg,NJ

ChemtechIndustries,Inc. CdCl2 Cd(BF4)2,CdSO4 HarstanDivision,St.Louis,MO

W.A.ClearyCorp.,Somerset, CdCl2 NJ

Deepwater,Inc.,Carson,CA CdI2 EaglePicher,Miami,OK CdS,CdTe

EngelhardCorp.,Catalystsand CdCl2,Cd(BF4)2,CdSO4,CdWO4 ChemicalsDivision,Cleveland,OH GELightingComponents CdSe,CdS MarketingandSalesOperation,Cleveland, OH

HallChemicalCompany, Cd(NO3)2 Arab,AL JohnsonMatthey,Danvers,MA CdTe McKenzieChemicalWorks,Inc. Cdacetylacetonate Bush,LA

MortonInternational,Inc.,SpecialtyChemicals CdS,(CH3)2Cd ,AdvancedMaterials,CVD,Inc. subsidiary,Woburn,MA PhilippBrothersChemicals, Inc.,C.P.ChemicalsSubsidiary

Sewaren,NJ Cd(BF4)2 Sumter,NC Cd(NO3)2

ShepherdChemicalCompany, Cd(CO3)2,Cd(OH)2,Cd(NO3)2 Cincinnati,OH R.T.Vanderbilt,Inc., Cddiethyldithiocarbamate VanderbiltChemicalCorp.Subsidiary, Bethel,CT

Source:Reference3.

4-30 TABLE4-12.MANUFACTURERSOFINORGANICCOMPOUNDSREPORTINGCADMIUM EMISSIONSINTHE1990TOXICCHEMICALSRELEASEINVENTORY

Emissions,kg(lb)

MonitoringData Plant Nonpoint Point Total

AmericanMicrotrace 0.5 2 2.5 no Corp.,Fairbury,Nebraska(micronutrients (1) (4) (5) foragriculture)a CPChemicalsInc. 000no Sumter,SouthCarolina (Cadmiumnitrate) HallChemicalCompany,Arab 000no Plant,Arab,Alabama(Cadmiumnitrate) ShepherdChemicalCompany, 000no Cincinnati,Ohio(Cadmiumcarbonate, hydroxide,nitrate)

TOTAL 0.5 2 2.5 (1) (4) (5) aMayproducezincsulfatefromslabzinc,galvanizer’sdross,orbaghousedustfromthebrass industrycontaminatedwithtracesofcadmium.2,13 Zincsulfatecapacityoftheplantis20,000Mg (36percentZn).3

Source:References3,6,and9.

4-31 REFERENCES 1.EmissionStandardsandEngineeringDivision. Cadmium EmissionsfromCadmiumRefiningandPrimaryZinc/ZincOxide Smelting--PhaseITechnicalReport. EPA-450/3-87-011. OfficeofAirQualityPlanningandStandards,U.S. EnvironmentalProtectionAgency,ResearchTrianglePark, NorthCarolina. June1987. 2.Jolly,J.H. Zinc. (In)MineralsYearbook1989. VolumeI. MetalsandMinerals. BureauofMines,U.S.Departmentof theInterior,U.S.GovernmentPrintingOffice,Washington, D.C. 1991. pp.1145-1174. 3.SRIInternational. 1991DirectoryofChemicalProducers: UnitedStatesofAmerica. SRIInternational,MenloPark, California. 1991. pp.516,571,582,869-871,946. 4.JACACorporation. AnalysisandAlternatives-AirEmissions ControlStudy-ASARCOGlobePlant,Denver,Colorado. Final Report. PreparedfortheColoradoDepartmentofHealth. August6,1992. 5.RadianCorporation. BackgroundInformationDocumentfor CadmiumEmissionSources. DCNNo.85-203-012-23-06,EPA ContractNo.68-02-3818,Assignment23. PollutantAssessment Branch,U.S.EnvironmentalProtectionAgency,Research TrianglePark,NorthCarolina. May1985. pp.39-56,174, 197-212. 6.MidwestResearchInstitute. BigRiverZincCorporation. AssessmentofCadmiumandArsenicEmissions. EPAContract No.68-02-4395(MRIProjectNo.8986-28). PreparedforAir andRadiationBranch,U.S.EnvironmentalProtectionAgency, RegionV,Chicago,Illinois,andDivisionofAirPollution Control,IllinoisEnvironmentalProtectionAgency, Springfield,Illinois. September1989. 38pp. 7.U.S.EnvironmentalProtectionAgency. 1990ToxicChemicals ReleaseInventory. OfficeofToxicSubstances,Washington, DC. January1993. 8.EmissionStandardsDivision. CadmiumEmissionsfromPigment andStabilizerManufacturing--PhaseITechnicalReport. OfficeofAirQualityPlanningandStandards,U.S. EnvironmentalProtectionAgency,ResearchTrianglePark, NorthCarolina. June1988.

4-32 9.CadmiumAssociation/CadmiumCouncil. TechnicalNoteson Cadmium. CadmiumProduction,PropertiesandUses. Cadmium Association,London;CadmiumCouncil,NewYork,NewYork. 1991. 8pp. 10.1991ThomasRegisterofAmericanManufacturers. Thomas PublishingCompany,NewYork,NewYork. 1991. 11.McGraw-HillDictionaryofScientificandTechnicalTerms, ThirdEdition. SybilP.Parker,EditorinChief. McGraw-Hill,NewYork,NewYork. 1984. 12.Parker,P.D. CadmiumCompounds. (In)Kirk-Othmer EncyclopediaofChemicalTechnology. Volume4. 3rdEd. M.Grayson,ed. AWiley-IntersciencePublication,JohnWiley andSons,NewYork,NewYork. 1978. pp.397-411. 13.Mumma,C.,K.Bohannon,andF.Hopkins. ChemicalTechnology andEconomicsinEnvironmentalPerspectives. TaskVI- CadmiuminPhosphateFertilizerProduction. EPA560/2-77-006. OfficeofToxicSubstances,U.S. EnvironmentalProtectionAgency. Washington,D.C. November 1977. 35pp.

4-33 SECTION5

EMISSIONSFROMMAJORUSESOFCADMIUM

Emissionsfromindustrialprocessesthatusecadmiumare discussedinthissection. Basedonthe1991U.S.industrial demandfigurespresentedinFigure3-1,Section3,cadmiumand cadmiumcompoundshavefourmajorcommercialuses. Theseare: (1)electroplating,(2)secondary(i.e.,rechargeable)battery manufacture,(3)heatstabilizersforsyntheticmaterialsand plasticresins,and(4)pigmentsforplasticproducts. This sectionisdividedintofoursubsections,onedevotedtoeach majoruse. Eachsubsectionpresentsabriefintroductiontothe industryandageneraldiscussionoftheproductionprocess. Wherecadmiumisusedintheprocess,descriptionsofexisting cadmiumemissioncontrolmeasuresandestimatesofcadmium emissionfactorsaregiven. Thelevelofdetailwillvary accordingtotheavailabilityofinformation,particularlyfor emissionswheredatamaybeincompleteorabsent.

CADMIUMELECTROPLATING

In1991,cadmiumelectroplatingapplicationsaccountedfor approximately20percentofthetotaldemandforcadmium.1 In cadmiumelectroplating,athinlayerofcadmiumisdeposited directlyoverabasemetal(usuallysteel)toprovidecorrosion protection,alowcoefficientoffriction,andalowelectrical contactresistance. Inaddition,cadmiumcoatingsalsoareused intheelectricalindustrybecausecadmiumiseasilysoldered. Cadmiumelectroplatingisperformedonsuchitemsasaircraft

5-1 fasteners,cableconnectorsforcomputers,shipcomponents,and automobileenginecomponents. Table5-1presentsthemajor marketareasforcadmiumcoatings.2

Anestimated1,200metalfinishingjobshopsthatperform cadmiumelectroplatingoperateintheUnitedStates.3 Metal finishingshopsaretypicallylocatedatorneartheindustries theyserve. Therefore,thegeographicaldistributionofthe metalfinishingshopscloselyfollowsthatofthemanufacturing baseintheU.S.3

ProcessDescription

Aflowdiagramforatypicalcadmiumelectroplatingprocess ispresentedinFigure5-1. Priortoplating,thepartsundergo aseriesofpretreatmentstepstosmooththesurfaceofthepart andtoremoveanysurfacesoil,grease,oroil. Pretreatment stepsincludepolishing,grinding,and/ordegreasingofthepart toprepareforplating. Thepartbeingplatedisrinsedafter eachstepintheprocesstopreventcarry-overofsolutionthat maycontaminatethebathsusedinsuccessiveprocesssteps.

Polishingandgrindingareperformedtosmooththesurfaceof thepart. Degreasingisperformedeitherbydippingthepartin organicsolventsorbyvapordegreasingthepartusingorganic solvents. Vapordegreasingistypicallyusedwhenthesurface loadingofoilorgreaseisexcessive. Thetwoorganicsolvents mostcommonlyusedforcleaningapplicationsare trichloroethyleneandperchloroethylene.

Alkalinecleaningissometimesusedtodislodgesurfacesoil andpreventitfromsettlingbackontothemetal. Thesecleaning solutionsaretypicallymadeupofcompounds,suchassodium carbonate,sodiumphosphate,andsodiumhydroxide;theyusually

5-2 TABLE5-1.MARKETAREASFORCADMIUMCOATINGSa

Percentageofallcadmium Marketarea coatingproducts,%

Electronicsandcommunications 22.5

Automotiveparts 30.0

Aircraft/aerospacefasteners 12.5

Industrialfasteners 17.5

Ordinance 6.0

Shipbuilding 5.0

Hardware(hinges,etc.) 2.5

Railroadandother 2.5

Householdappliances 1.5

Source:Reference2. aBasedon1989data.

5-3

containasurfactant. Alkalinecleaningtechniquesinclude soakingandcathodicandanodiccleaning.

Aciddipsmaybeusedtoremoveanytarnishoroxidefilms formedinthealkalinecleaningstepandtoneutralizethe alkalinefilm. Aciddipsolutionstypicallycontainfrom10to 30percentbyvolumehydrochloricorsulfuricacidinwater.

Theexactpretreatmentstepsuseddependupontheamountof ,grease,oroilontheparts. Followingpretreatment,the partsaretransferredtotheplatingtank.

Severalcadmiumplatingbathformulationsareusedtodeposit cadmiumonthebasemetalorpart;however,thecadmiumcyanide bathisthepredominantformulationusedtodepositcadmium. Otherbathformulationsusedincludeaneutralsulfate,anacid fluorborate,oranacidsulfatebath. Currently,theuseof theseotherbathformulationsisnotappreciablebecausethe cadmiumdepositsformedfromthesebathsarenotofsufficient quality(i.e.,donotdisplaythedesiredphysicalproperties)to gainwidespreadacceptance. Therefore,thefollowingdiscussion willfocusonthecadmiumcyanideplatingbath.

Table5-2presentsthebathcompositionandoperating parametersofthecadmiumcyanidebath.4 Incadmiumplating,the part(s)isplacedinatankandconnectedintotheelectrical circuitasthecathode. Ifsmallpartsaretobeplated,the partsarefirstplacedinaplatingbarreloronaplatingrack. Thebarrelorplatingrackisthenplacedinthetankand connectedintotheelectricalcircuit. Ascurrentisapplied, cadmiumionsinthesolutionaredrawntothenegatively-charged cathodewheretheyundergoreduction,resultinginthecadmium beingdepositedonthepart. Theefficiencyoftheplatingbath isbasedontheamountofcurrentthatisconsumedinthe

5-5 TABLE5-2.COMPOSITIONANDOPERATINGPARAMETERSOFCADMIUMCYANIDE PLATINGBATH

Component Operatingrange

Compositionofbath,g/L(oz/gal) Cadmium 20(2.7) Cadmiumoxide 22(3.0) Sodiumcarbonate 30-60(4.0-8.0) Sodiumcyanide 101(13.5) Sodiumhydroxide 14(1.9)

Operatingparameters Currentdensity,A/m2 (A/ft2) 54-970(5-90) Temperature,°C(°F) 15-38(60-100) Cathodeefficiency,% 90-95 Typeofanodesused Cadmium Anodeefficiency,% 100

Source:Reference4. depositionreactionversustheamountofcurrentthatisconsumed byothersidereactions. Forcadmiumplatingbaths,thecathode efficiencytypicallyrangesbetween90to95percent;therefore, 90to95percentofthecurrentsuppliedtothetankisconsumed inthedepositionreaction. Theremaining5to10percentis consumedbyothersidereactions,suchastheevolutionof hydrogengasatthecathodeandtheevolutionofoxygengasat theanode.

Followingplating,thepartisthoroughlyrinsed. Most cadmiumplatedproductsdonotrequireanyfurthertreatment; however,somepartsareoftenpost-treatedwithabrightdip. Thisdipisachromateconversioncoating,whichiscolored, painted,orlacquered,dependinguponthepartspecifications.

EmissionControl

Noairpollutioncontrolmeasuresarecurrentlybeingusedon cadmiumelectroplatingtanks. Localexhaustventilationis

5-6 sometimesusedonthesetanksasaprecautionarymeasureagainst workerexposure.

Emissions

BasedontheventilationguidelinespublishedbytheAmerican NationalStandardsInstitute(ANSI),theemissionpotentialfrom cadmiumelectroplatingtanksisextremelylow. Cadmiumcyanide electroplatingtanksaregivenahazardousclassificationofD-4, thelowestpossiblerating.5 Inthe1990ToxicChemicalRelease Inventory(TRI),41facilitiesreportedcadmiumemissionsunder SIC3471,PlatingandPolishing.6 Totalcadmiumemissions reportedfromthesefacilitiestotaled3,324Mg(3,656tons). However,itshouldbenotedthat25percentofthefacilities accountfor98percentoftheemissions. Fiftypercentofthe facilitiesreportedzeroemissions,and25percentreportedless than10lb/yrofcadmiumemissions. Areviewofthefacilities withthehigheremissionestimatesrevealedthatsomeofthe facilitiesweremanufacturersofplatingbathchemicalsandnot cadmiumplatingfacilities. Noadditionaldataareavailable regardingcadmiumemissionsfromcadmiumelectroplatingtanks.

SECONDARYBATTERYMANUFACTURE

Cadmiumisusedintheproductionofseveraltypesof secondary(rechargeable)batteries. In1991,thisareaaccounted forapproximately45percentofthetotaldemandforcadmium.1 Thissubsectionfocusesonemissionsandcontrolsduring productionofnickel-cadmiumbatteries,thelargestsegmentof thecadmiumbatteryindustry. Otherbatterytypesthatuse cadmiumincludesilver-cadmiumbatteries,whichhaveaerospace applications,andmercuricoxide-cadmiumbuttoncells. Informationwasnotavailableonthepotentialforcadmium emissionsfromtheseotherbatterytypes.

5-7 Nickel-cadmiumcellsaremanufacturedinavarietyofforms andsizesforprincipallytwoapplications: industrialand portablebatteries. Nickel-cadmiumbatteriesforindustrialuse areusuallythevented(oropen)orsemi-sealedtypeandmaybe eitherpocketplate,sinteredplate,orfiberstructured construction. Vented(open)celldesignsarecurrentlyusedfor larger-sizedcellsdesignedforindustrialorotherheavyduty applications. Intheseapplications,thebatteriesaresubject tofrequentchargingandrequireadditionofelectrolytesafter longperiodsofoperation. Applicationsfortheindustrial batteriesincludeseveralrailwayuses(e.g.,locomotive starting,emergencybraking,signalsandwarninglights),standby powerforalarmsystems,emergencylighting,military communications,solarenergystorage,navigationequipment, hospitaloperatingrooms,andaeronauticalapplications.7

Sealed-cellnickel-cadmiumbatteriesdesignedforportable applications(e.g.,toys,camcorders,portabletools,and cellulartelephones)usuallyareconstructedusingsinteredplate . Thesecellsaremanufacturedincylindrical,button, andprismaticshapes;theymayberechargedupto2,000times, andrequirenomaintenance.7

Eightprimaryproducers,theirplantlocations,batterytype, andprocessesusedwereidentifiedinanEPAreport.8 The informationoncompanynameandplantlocationswasupdatedusing emissionreportsfromthe1990TRIandispresentedin Table5-3.6 Inadditiontotheseprimaryproducers,some companiesmayassemblenickel-cadmiumbatteriesusingimported components.

5-8 TABLE5-3.NICKEL-CADMIUMBATTERYPRODUCERS--1990

Batterytype Process

Company Sealed Vented Sintered Pocket

Eagle-PicherIndustries,Inc., x Wet ColoradoSprings,CO

EvereadyBatteryCompany,Inc. Cleveland,OH x Dry Greenville,NC x x

GatesEnergyProducts,Inc. x x Wet Gainesville,FL

GNBIndustrialBatteryCompany Ft.Smith,AR x x x Kankakee,IL x x x

MarathonPowerTechnologies x x Assemblyonly Waco,TX

SaftAmerica,Inc. x x Wet Valdosta,GA

Source:References6and8. ProcessDescription

Nickel-cadmiumcellsutilizeareversibleelectrochemical reactionbetweencadmiumandnickelelectrodespackedinan alkalineelectrolyte(potassiumorlithiumhydroxide). The electrolytedoesnottakepartinthecharge/dischargereactions; itactsonlyasachargecarrier. Duringdischarge,thecadmium isoxidizedtocadmiumhydroxideatthecathode,andhydrated nickel(III)oxideisreducedtonickel(II)hydroxideatthe anode. Theprincipaldifferencebetweenthevarioustypesof nickel-cadmiumcellsisthenatureofthecellelectrodes. Three typesofpositiveelectrodes(anodes)areused: pocketplate, sinteredplate,orfiberplate. Thehydratednickeloxideatthe anodeisusuallyinpowderformandisheldinpocketplatesor suspendedinagelorpasteandplacedinsinteredorfiber electrodes. Negativeelectrodes(cathodes)usepocketplate,

5-9 sinteredpowder,fiberplate,foamorplasticbandedsupportsto holdthecadmiumhydroxideinplace. Graphiteorironoxideis commonlyaddedtoimprovetheconductivityofboththenickeland cadmiumhydroxide.7

Adescriptionofthesinteredplatewetprocessfornickel- cadmiumbatteryproductionispresentedinthissubsection. A flowdiagramfortheprocessisshowninFigure5-2. This processappearstohavethegreatestpotentialforcadmium emissionsasreportedbytheindustryinthe1990TRIsurvey.6 Descriptionswerenotavailablefortheotherproduction processes.

Insintered-plateformation,nickelpowderisheatedona nickel-platedsteelstriptogiveaporousmediumboundtoa . Heatingthenickelpowderathightemperatureswelds togetherthecontactpointsofthenickelpowdergrains. During theimpregnationsteps,solutionsofnickelorcadmiumimpregnate thevoidspacesofthesinterednickel. Duringthenickel impregnation,thesinteredplateissoakedwithasaturated solutionofnickelnitrateinnitricacid. Thecadmium impregnationstepissimilar,exceptthatthesaturatedsolution containscadmiumnitrate. Thecadmiumnitratesolutionmaybe preparedonsitefromcadmiumoxideorpurchased.9

Theimpregnatedplatesaredriedandthenimmersedina potassiumhydroxidesolutiontoconvertthenickelandcadmium saltstotheirrespectivehydroxides. Theanodes(withnickel hydroxide)andcathodes(withcadmiumhydroxide)undergoaseries ofstepsbeforebeingassembledintocellsandthenbatteries: washingandovendrying,finalcausticsoak,hotdeionizedwater rinse,formingincaustic,andfinalbrushandrinse.8,9

5-10

Twoalternativemethodsforimpregnatingthecathodeare used. Inonemethod,thecadmiumiselectrolyticallydeposited fromastandardcadmiumelectroplatingsolutionontothesintered plate. Thecadmium-platedsinteredstripisthenrinsedandis readyforassembly. Inanothermethodforcathodeproduction, drycadmiumpowderandabinderarepressedonwiremeshina moldandtransferredtotheassemblysteps.8,9 Sincethe individualcellsareprecycledbeforeassemblingintobatteries, itisnotimportantwhetherthecathodesareoriginally impregnatedwithCd(OH)2 (theproductofdischargereactions)or Cd(theproductofchargingreactions). Thereactionsareas follows:

discharge β ______2 -NiOOH+Cd+2H2O >Ni(OH)2 +Cd(OH)2 <______charge discharge - ______- Cd+2OH > Cd(OH)2 +2e <______charge Duringassembly,thenickel-containinganodeandthecadmium- containingcathodesareassembledalternatelyintocellswith feltednyloncelluloseseparators,andthecellsareassembled intobatterycasesofplasticornickel-platedsteel. The electrolytecontainingpotassiumhydroxideandlithiumhydroxide isaddedtotheassembledcomponentsinthebatterycase. The separatormaterialholdstheelectrolyte,aswellasseparates thenegativeandpositiveelectrodes. Thebatteriesfinally undergotestingandpacking;failedbatteriesarerejected.8,9

EmissionControlMeasures

Inanickel-cadmiumbatteryplant,themostcommonformsof cadmiumemittedarecadmiumnitrate,cadmiumhydroxide,and

5-12 possiblycadmiumoxide. Allairemissionsofcadmiumcompounds willoccurasparticulatematter--andprimarilyasfugitive emissionsduetomaterialhandlingandtransferprocedures,oven dryingoperations,andcellassembly.

Thepredominantcontrolmethodsusedintheindustryare: (1)hoodsandvacuumsystemsductedtodustcollectorsand (2)fabricfiltersincadmiumhandlingareas. Fabricfiltersare knowntobehighlyeffectiveparticulateremovaldevices, especiallyforthelowertemperatureemissionsanticipatedfor thisindustry. Atmostfacilities,fugitiveemissionsare containedwithintheplantandarecapturedandsenttoacontrol device.

Emissions

Cadmiumispotentiallyemittedfromseveralstepsinthe manufactureofnickel-cadmiumbatteries. Potentialemission sourceswerenotedwithasolidcircleinFigure5-2. Operations involvingthehandlingofdrycadmiumsaltsandpowders,oven drying,andcellassemblyarethelikeliestsources.

Solutionpreparationisalsoapotentialsourceofcadmium emissions. Ifthesinteredplatesaretobeimpregnatedwith cadmiumnitratesolution,cadmiummaybeemittedbythehandling ofdrysaltsduringsolutionpreparation. Preparationofthe cadmium-containingelectrolytefortheelectrolyticdeposition alsowouldemitcadmiumifdrymaterialisused. Ifthecathode ispreparedbythedry-pressingprocess,handlingofdrycadmium powdersandpressingthecadmiumpowderintothegridare potentialemissionsources. Ovendryingofcathodeplate materialandthecellassemblystepalsoarepotentialcadmium emissionsources.

5-13 Nocadmiumemissionfactorshavebeenpublishedforthe nickel-cadmiumbatteryproductionprocess,norareanyemission testdataavailablethatwouldallowthecalculationofemission factors. A1985backgrounddocumentoncadmiumemissionsources estimatednationwideemissionsfrombatterymanufacturingat 100kg/yr(220lb/yr).8 Inthe1990TRI,theeightbattery productionsitesshowninTable5-3reportedatotalannual cadmiumreleaseof316kg(697lb).6 Basedonthesedata,the productionofnickel-cadmiumbatteriesdoesnotappeartobea majorsourceofcadmiumemissions.

CADMIUMSTABILIZERSFORPLASTICS

Cadmiumcompounds,inconjunctionwithbariumcompounds,have beenwidelyusedasaneffectiveheatstabilizersystemfor polyvinylchloride(PVC)andrelatedpolymers. Polyvinyl chlorideisgenerallyregardedasoneofthemostversatileof polymersbecauseofitscompatibilitywithmanyothermaterials, suchasplasticizers,fillers,andotherpolymers. Themajor disadvantageisitspoorthermalstability. Thephysical appearanceandperformancepropertiesofPVCcanbemodifiedby theincorporatingadditives,butnothingcanbedoneto completelypreventpolymerdecompositionbyphysicalorchemical means. Additivesclassifiedasstabilizerscaneffectively hinderandreducethedegradationprocessuntilitessentially ceases,butabreakdownundertheactionofphysicalandchemical agentsisalwayspresenttosomedegree. Severalmechanismshave beenproposedasroutesforPVCdestruction. Thesemechanisms arequitesimilarchemicallyandcanbedirectlyrelatedtothe physicalstateofthePVC. Dehydrochlorinationisthemost significantcauseofdegradationinPVC. Theprocesscanbe initiatedeitherbylossofalabile-chlorineatomorthrougha freeradicalreactionwiththeresultantformationofadouble bond. Asdehydrochlorinationcontinues,conjugateddoublebonds

5-14 areformed,resultinginashiftinthewavelengthoflight absorbedbythepolymer. Thewavelengthofabsorbedlight changesaccordingtothenumberofconjugateddoublebondsystems thatarepresent,andthecolorofthepolymerchangesfromlight yellowtodarkyellowtoambertoreddish-brownandfinallyto black.10

Stabilizersareusuallyinorganicororganometallic compounds,whosenamesreflectthecationsinvolved. Themajor classesofstabilizersarebasedontin,lead,andamixtureof GroupIImetals,suchasbarium,cadmium,andzinc. TheGroupII metal(mixedmetal)stabilizershaveprogressedovertheyears fromsimpleadditionsofbariumsuccinateandcadmiumpalmitate tocomplexblendsofbarium/cadmium/zincsoaps,organophosphites, antioxidants,solvents,extruders,peptizers,colorants, (UV)absorbers,andmanyotherconstituents. Cadmium stabilizerswereinitiallyusedbecausetheyimpartclarityand retentionofinitialcolortoaPVCformulation. Thelong-term heatstabilitysuppliedbycadmiumandzincismuchlessthan thatofferedbybariumcompounds. Cadmiumstabilizersare functionallydependentupontheanions,andtheanionsarea majorfactorthataffectsproperties,suchaslubricity,plate- out,clarity,colordrift,andheatstability. Commonanionsfor cadmiumarethe2-ethylhexoate(octoate),phenate,benzoate,and stearate.10

Cadmium/bariumstabilizersarecommerciallyavailablein liquidorsolidform. Liquidstabilizersystemsareeasierto handleanddonotresultinplate-outproblems,whichmayoccur withthepowderedsystems. Theliquidstabilizersusuallyhavea lowercadmiumcontent(1to4percent)andarecheaperona weightbasis. Solidstabilizershaveahighercadmiumcontent(7 to15percent)andaremoreeffectivethanliquidstabilizerson aweightbasis.13

5-15 Inthesemixedmetalstabilizersystems,thecadmiumcontent rangesfrom1to15percent,andthestabilizersystem constitutesbetween0.5to2.5percentofthefinalPVC compoundedresin.7 Mostcadmium-containingstabilizersystems arebarium/cadmium-zincbasedmixtures;thesesystemsarebeing replacedwithbarium/zincproducts. Thesuccessfulreplacement ofcadmium-containingproductsdependsprincipallyontheuseof higherbarium-to-zincratiosthanbarium-to-cadmiumratiosand theanionchemistry,whichcompensatesforthesmallersizeof thezincatomcomparedtothecadmiumatom.11 Anestimated30 to35percentofthecadmium-containingstabilizerusageinthe U.S.haschangedtononcadmiumproducts,andthispercentageis expectedtoincreasetomorethan50percentbytheendof 1992.12

ProcessDescription

Theadditionofheatstabilizeradditivesoccursaspartof theoverallproductionoftheformulatedPVCresins. Formulation oftheresinnormallyusesablendersystemand,dependingupon theparticularPVCproduct,maybeabatchorcontinuous operation. Solidcadmiumstabilizersystemsmaybeadded directlytothedryPVCresinandthenthoroughlymixedwiththe resinparticles. Liquidcadmiumstabilizersmaybeadded directlytotheresinormixedwithaliquidplasticizerpriorto additiontotheresin. Theparticularsequenceofstabilizer additiondependsupontheprocessingmethodtobeused(e.g., calendering,extrusion,dipping). Afteralladditives,including thestabilizerhavebeenincorporated,theformulatedresinis usuallyafree-flowingpowderorgranulewiththeliquids adsorbedontheresinparticles.

Themostcommonusageofcadmium-basedstabilizersisfor flexibleandsemi-rigidPVCapplications.11 Ingeneral,cadmium-

5-16 basedstabilizersareusedintheproductionofflexibleand semi-rigidPVCproducts. ThesePVCproductsareprocessedby calendering,extrusion,orinjectionmoldingtechniques.10 Cadmium-basedstabilizersfindlimiteduseinrigidPVCproducts orfilmsforelectricaluses. Liquidcadmiumstabilizersmaybe usedinproductionofthefollowingtypesofPVCproducts: 1. Flexibleorsemi-rigidinjectionmolded; 2. Clearplastisols; 3. Thingaugeorlightlyfilledcalenderedfilms; 4. Clearandlightlyfilledextrudedfilmsorsheets;and 5. Dippingoperations. Solidcadmium-basedstabilizersmaybeusedinhighlyfilled calenderedsheet(e.g.,floortile)orothercalendering, injectionmolding,orextrusionprocessestomanufacturefilled (i.e.,nonclear)PVCproducts.

EmissionControlMeasures

Noinformationisavailableforthespecifictypesof emissioncontroldevicesusedtocontrolcadmiumemissions resultingfromproductionofPVCproducts. Onemanufacturing facilityusingcadmiumstabilizersindicatedthatthemajor emissionsourcewouldbeduetomaterialshandling.16 This sourcepresumablywouldbeintheresinformulationareaandifa smallbatchoperationwereused,duringtransferofthe formulatedresin. Mostsolidcadmiumstabilizersareproducedin forms(e.g.,flaxes,pellets)toreducedustemissionsduring handling.

Cadmiumemissionsfromtheprocessesofextruding,molding, andcalenderingareprobablyminimalsincethetemperatures necessarytovolatilizesignificantquantitiesofcadmium compoundswouldthermallydestroytheresinandotherorganic constituents.

5-17 Emissions

Cadmiumemissionsmayoccurwhencadmium-containing stabilizerssystemsareaddedtoPVCresinsduringformulation; priortoprocessingthePVCresin. Althoughuseofcadmiumin theproductionofstabilizersconstitutesabout12percentofthe totaldemandforcadmium,theemissionofcadmiumresultingfrom theuseofthestabilizersduringresinformulationhadnotbeen consideredapotentialsource. Table5-4presentscadmium emissionsbyseveralmanufacturersofformulatedresinsand plasticsreportedinthe1990TRI.6 Someofthesefacilitiesare probablyalsousingcadmium-basedpigmentsintheresins,butthe reportingsystemintheTRIdoesnoteasilydistinguishbetween thetwocadmiumproducts. Thus,someofthecadmiumemissions mayresultfrompigmentusage.

Noemissionfactorsarepublishedforthisprocess,andno testdataareavailabletoallowcalculationofanemission factor.

CADMIUMPIGMENTSINPLASTICS

About80percentofallcadmium-basedpigmentsisusedinthe plasticsindustry. Theother20percentisusedmostlyforthe colorationofpaints,coatings,ceramics,andglasses.15 Inthe plasticsindustry,pigmentsandotheradditivesareblendedwith theresinsbeforetheplasticsproductsaremanufactured. This blendingstepcanbedoneinconjunctionwithothermanufacturing stepsattheproductionsite. Alternatively,custom-blended resinscanbepurchasedfromanothercompanyandtransportedto theproductionsite. Thisisacommonpracticeforsmaller companiesorforspecialtyproducts. Table5-5lists manufacturersofcustomcompoundpurchasedresinswhoreported emissionsofcadmiuminthe1990TRI.6 Mostofthisgroupblends

5-18 TABLE5-4.REPORTEDCADMIUMEMISSIONSBYMANUFACTURERSOF FORMULATEDRESINSANDPLASTICPRODUCTS

Reportedemissions

Company Location kg lb Rohm&Haas,Inc. Bristol,PA 2 5 GencorpPolymerProducts Newcomerstown,OH 5 10 GeneralElectricPlastics Selkirk,NY 8 16 SyntheticProductsCompany Stratford,CT(2) 118 261 Cleveland,OH 227 500 MonsantoCompany Addyson,OH 5 11 GeneralElectricChemicals Washington,WV 1 2 HulsAmerica,Inc. MountainTop,PA 113 250 FranklinBurlingtonPlastics Burlington,NJ 227 500 O’SullivanCorp. Lebannon,PA 5 10 Winchester,VA 5 10 Yerington,NV 5 10 GaryCorp. Leominster,MA 43 94 NorthAmericanPlastics,Inc. Prairie,MS 5 10 VytronCorp. Loudon,TN 113 250 StandardProductsCompany Winnsboro,SC 2 5 RJFInternationalCorp. Marietta,OH 5 10 AchillesUSA,Inc. Everett,WA 16 35 IPCCorinthDivision,Inc. Corinth,MS 113 250 B.F.Goodrich Pedricktown,NJ 24 52 RegalitePlasticsCorp. NewtonUpperFalls,MA 5 10 TOTAL 1,047 2,301

Source:Reference6.

Note:Inadditiontothecompaniesandlocationsshowninthetable,16additionalcompaniesor locationsreportedzerocadmiumemissions.

5-19 TABLE5-5.REPORTEDCADMIUMEMISSIONSBYPRODUCERSOF CUSTOMCOMPOUNDEDRESINS

Emissions

Company Location kg lb PlasticsColorChip,Inc. Ashboro,NC 227 500 CalumetCity,IL 116 255 VistaChemicalCompany Jeffersontown,KY 116 255 ReedPlasticsCorp. Albion,MI 2 5 Holden,MA 2 5 GeneralColorandChemical Minerva,OH 113 250 Company A.Schulman,Inc. Akron,OH 5 10 PMSConsolidated Norwalk,OH 227 500 Ft.Worth,TX 116 255 Florence,KY 116 255 Gastonia,NC 227 500 St.Peters,MO 227 500 Somerset,NJ 116 255 ElkGroveVillage,IL 116 255 TeknorApexCompany Pawtucket,RI 5 10 HoechstCelanese Florence,KY 5 10 QuantumChemicalCorp. FairportHarbor,OH 227 500 TOTAL 1,963 4,320

Source:Reference6.

5-20 pigmentswithresins.

Thecostsofpurchasingcustomcompoundedresinshaverisen toalevelwheresomeproducersofplasticgoodshavechangedto blendingtheirownresins. Thisshiftinproductionlocalemay beparticularlytrueforusersofcadmiumpigments,becausethese pigmentsareexpensiveandhaveadvantagesofeasymixingand rapid,evenspreading.

ProcessDescription16

Mostcommercialpigmentshaveanaverageparticlesizeinthe rangeof10-3 to10-5 mm(0.01to1.0µ). Thedrypigmentpowders areusuallyagglomeratedbeforesaleinordertoreducematerial lossduringtransport. Theseagglomeratesmustbedispersedby thecompoundresinmanufacturer,eitherbeforeorduring processing. Theinitialstepindrypigmentdispersionis wettingofthepigmentsurface. Subsequentstepsarebreaking downofagglomerates,distributionoftheparticlesintheresin, andstabilizationofthedispersion.17 Sincecadmiumpigment losswouldbeminimalafterthedrypigmentiswet,this discussionfocusesonmaterialshandling.

Bulkmaterialscanbestoredinoutdoorsilos,boxes,bags, ordrums. Largevacuumpumpstransportmaterialsfromrailcaror trucktosilos. Smallervacuumpumpstransportmaterialsfrom onsitestorageinbags,drums,andboxestothehopperloadersof processmachinery,ifnottothemachinesthemselves. Vacuum linesenterhopperstangentiallysothatthematerialcanbe separatedfromtheconveyingairstream. Anexternalratio mixingvalveisusuallylocatedattheinletofeachvacuum hoppertoallowregrindsandotherrecycledmaterialtobe proportionallymixedwithvirginmaterialpriortoprocessing.17

5-21 EmissionControlMeasures

Accordingtoconversationswithcompanyofficialsat productionplants,itwasdeterminedthatcadmiumemissions originateprimarilyfrommaterialshandling.18,19

Handmethodsofblendingmaterialscanwasteupto25percent ofpurchasedcolorants. Automaticmethods,suchasmetering, mixing,andvacuumtransport,substantiallyreducewasteand emissions. Emissionsofpowderedmaterialsfromvacuumhoppers areusuallycontrolledwithfiltersandfloor-mounteddust collectors. Allcadmiumemissions,ascadmiumpigments,wouldbe inaparticulateformsotheuseofdustfiltersanddust collectorsshouldbeaneffectiveemissioncontrolmeasure. However,therearenotestdataavailabletosubstantiatethe effectivenessofthesecontrolsforthecadmiumpigment particulatefoundatthesesources.

Emissions

Emissionfactorsarenotavailableforpigmentblending operations,whichhavenotbeenrecognizedasapotentialsource ofcadmiumemissionsinprevioussurveys. Notestdataare availablethatcanbeusedtocalculateemissionfactors.

5-22 REFERENCES 1.1991MineralIndustrySurveys--Cadmium. AnnualReview. BureauofMines,U.S.DepartmentofInterior,Washington,DC. January1993. 2.Kraft,G.G. "TheFutureofCadmiumElectroplating,"Metal Finishing. Volume88,No.7. July1990. 3.Finishers’ManagementMedia/MarketBulletin. MetalFinishing JobShopIndustrialProfile...1985/86. Glenview,IL. 1986. 4.MetalFinishing. GuidebookandDirectoryIssue. January1993. Volume91,Number1A. NewYork,NY. 5.AmericanNationalStandard,PracticesforVentilationand OperationofOpen-SurfaceTanks,AmericanNationalStandards Institute,NewYork,NY. 6.U.S.EnvironmentalProtectionAgency. 1990ToxicsRelease Inventory. OfficeofToxicSubstances,Washington,DC. January1993. 7.CadmiumAssociation/CadmiumCouncil. TechnicalNoteson Cadmium. CadmiumProduction,Properties,andUses. Cadmium Association,London;CadmiumCouncil,NewYork,NY. 1991. 8.RadianCorporation. BackgroundInformationDocumentfor CadmiumEmissionSources. DCNNo.85-203-012-23-06,EPA ContractNo.68-02-3818,Assignment23. PollutantAssessment Branch,U.S.EnvironmentalProtectionAgency,Research TrianglePark,NC. May1985. 9.RadianCorporation. LocatingandEstimatingAirEmissions fromSourcesofNickel. EPA-450/4-84-007f. OfficeofAir QualityPlanningandStandards,U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. March1984. 10.ModernPlastics. Encyclopedia’92. "Chemicalsand Additivies: Stabilizers". Vol.68(11),pp.190-192. McGrawHillPublishing,NewYork,NY. October1991. 11.ModernPlastics. SpecialBuyersGuideandEncyclopedia,1993 Issue. "Additives: HeatStabilizers." Vol.69(13), pp.313-315. McGrawHillPublishing,NewYork,NY. December1992. 12.ModernPlastics. "SpecialReport: ChemicalsandAdditives." Vol.69(9),pp.61-63. McGrawHillPublishing,NewYork, NY. September1992.

5-23 13.CadmiumAssociation/CadmiumCouncil. TechnicalNoteson Cadmium: CadmiuminStabilizersforPlastics. Cadmium Association,London;CadmiumCouncil,NewYork,NY. 1978. 14.Telecon. Gupta,R., MidwestResearchInstitutewith P.Vorhees,RohmandHaas,Philadelphia,PA. Particulate emissionsfromtheuseofcadmiumstabilizersinPVC. April1992. 15.Lynch,R.F. ColorItCadmiumWhenYouNeedtheBest. PlasticsEng.,April1985. 16.Witlock,D.C. MaterialsHandling: InModernPlastics Encyclopedia,Vol.63(10A),October1986. 17.Schiek,R.C. Colorants: InModernPlasticsEncyclopedia; Vol.63(10A). October1986. 18.Telecon. Camp,J.,MidwestResearchInstitute,with V.Karagan,Avecor,Inc.,SanFernando,CA. April9,1992. Cadmiumpigmentsmaterialshandlingduringcompoundingwith resins. 19.Telecon. Camp,J.,MidwestResearchInstitute,with D.Frost,PMSConsolidated,Toledo,OH. April16,1992. Cadmiumpigmentmaterialshandlingduringcompoundingwith resins.

5-24 SECTION6

EMISSIONSFROMCOMBUSTIONSOURCES

Cadmiumisoftenfoundasatracecontaminantinfossilfuels orwastematerials. Whenthesematerialsarefedtocombustion processes,thecombinationoftheelevatedtemperatureofthe processandtherelativevolatilityofcadmiumresultsincadmium beingpartitionedbetweentheashandthecombustiongasexhaust stream. Thissectionaddressescadmiumemissionsfromseven stationarysourcecombustionprocesses: -Coalcombustion -Oilcombustion -Naturalgascombustion -Woodcombustion -Municipalwastecombustion -Sewagesludgeincineration -Medicalwasteincineration Thesesevenprocessesfallintotwogeneralcategories. The firstfourinvolvefossilfuelcombustionforenergygeneration, whilethelastthreeareprimarilywastedisposalprocesses, althoughsomeenergymayberecoveredfromtheseprocesses. The paragraphsbelowprovideageneralintroductiontothetwo combustioncategories. Aspartofthisintroduction,asummary ofnationwidefuelusageispresentedindetail. This informationwasusedinSection3todevelopnationwideemissions ofcadmiumfordifferentsectorsandfuels. Suchinformationis alsoneededtodevelopcadmiumemissioninventoriesforspecific

6-1 areas. Itisincludedintheintroductionratherthanin individualsectionsbecause(1)theindividualsectionsare organizedbyfueltyperatherthanbyusesectorand(2)fossil fuelusepatternsdiffergeographicallybyindustrysector. The introductionalsobrieflydescribesthewastecombustion category. Specificdiscussionsforthesevensourcecategories followtheseintroductoryparagraphs.

In1990,thetotalannualnationwideenergyconsumptionin theUnitedStateswas85.53x1012 megajoules(MJ) (81.15x1015 Britishthermalunits[Btu]).1 Ofthistotal,about 52.01x1012 MJ(49.35x1015 Btu)or61percentinvolved consumptionofcoal,petroleumproducts,andnaturalgasin nontransportationcombustionprocesses. (Nodatawereavailable onenergyconsumptionforwoodcombustionfromtheUnitedStates DepartmentofEnergy.) Table6-1summarizesthe1990United Statesdistributionoffossilfuelcombustionasafunctionof fuelintheutility,industrial,commercial,andresidential sectors. Theparagraphsbelowprovidebriefsummariesoffuel usepatterns;additionaldetailsonfuelconsumptionbysector foreachStatecanbefoundinReference1.

AsshowninTable6-1,theutilitysectoristhelargest fossilfuelenergyconsumerattherateofabout21x1012 MJ (20x1015 Btu)peryear. About80percentofthisenergywas generatedfromcoalcombustion,withbituminousandlignitecoal contributingsubstantiallygreaterquantitiesthananthracite coal. Infact,PennsylvaniaistheonlyStateinwhich anthracitecoalisusedforelectricpowergeneration. Although mostStatesreliedprimarilyoncoalforpowergeneration,the distributionamongfossilfuelsvariedfromStatetoState,and severalStatesreliedheavilyonnaturalgasandfueloilfor powergeneration. InCalifornia,naturalgasprovidesabout 90percentofthefossil-fuelbasedelectricityproduction,and

6-2 TABLE6-1.DISTRIBUTIONOFFOSSILFUELCONSUMPTION INTHEUNITEDSTATES

Annual energy consumption 1012 MJ (1015 Btu) Fuel Utilities Industrial Commercial Residential Total

Bituminous/ 16.939 2.892 0.085 0.045 19.961 lignite coal (16.071) (2.744) (0.081) (0.043) (18.939)

Anthracite coal 0.018 0.009 0.013 0.019 0.059 (0.017) (0.009) (0.012) (0.018) (0.056)

Distillate oil 1.201 1.245 0.513 0.882 3.900 (1.139) (1.181) (0.487) (0.837) (3.700)

Residual oil 0.091 0.436 0.255 - 0.782 (0.086) (0.414) (0.242) - (0.742)

Other petroleum 0.026 7.083 0.197 0.452 8.540 fuels (0.025) (6.720) (0.187) (0.429) (7.361)

Natural gas 3.015 8.925 2.907 4.762 19.609 (2.861) (8.468) (2.758) (4.518) (18.605)

Total 21.290 20.590 3.970 6.160 52.01 (20.199) (19.536) (3.767) (5.845) (49.35)

Source: Reference 1.

6-3 nocoalisused. InHawaii,fueloilisusedexclusively,while inOklahomaandTexas,amixtureofcoalandfueloilareused. InFlorida,Louisiana,Massachusetts,andNewYork,coal,fuel oil,andnaturalgaseachrepresentasubstantialfractionofthe powergeneration. TheStatesofIdaho,Maine,RhodeIsland,and Vermonthadnocoalconsumption. Idahoreliesexclusivelyon hydroelectricpower,whiletheNewEnglandStatesuseamixture offueloil,naturalgas,nuclear,andhydroelectricsources.

At20.6x1012 MJ(19.5x1015 Btu)peryear,theindustrial sectoristhesecondlargestconsumeroffossilfuels. This sectorusesamixtureofnaturalgas(43percent),fueloil (8percent),otherpetroleumfuels(34percent),andcoal (14percent). Theotherpetroleumfuelsthatareusedinclude primarilyliquifiedpetroleumgas,asphaltandroadoil,and othernonclassifiedfuels. Again,thedistributionamongthe threefueltypesvariessubstantiallyfromStatetoState,with eachofthethreecontributingsignificantfractionsinmost States. NotableexceptionsareHawaii,whichreliesalmost exclusivelyonpetroleumfuels;Alaska,whichreliesprimarilyon naturalgas;andthenortheasternStatesofConnecticut,New Hampshire,RhodeIsland,andVermont,whichusealmostnocoal.

AsshowninTable6-1,substantiallysmallerquantitiesof fossilfuelareusedinthecommercialandresidentialsectors thanareusedintheutilityandindustrialsectors. Thefuels usedareprimarilynaturalgas,fueloil,andliquifiedpetroleum gas(the"otherpetroleumfuels"intheresidentialcategory). AlmostallStatesuseamixtureofthefuels,butthe distributionsvarysubstantially,withsomeStateslike CaliforniaandLouisianausingprimarilynaturalgasandothers likeNewHampshireandVermontusingamuchgreaterfractionof fueloil. OneuniquecaseisPennsylvaniawhereanthracitecoal isusedinboththeresidentialandcommercialsectors.

6-4 Intheindividualsectionsbelow,additionalinformationwill bepresentedonthecadmiumcontentofthedifferentfuelsandon therelationshipbetweenfueltypeandemissions. However,for anygeographicarea,thecontributionofenergygeneration sourcestocadmiumemissionswillbeafunctionofthe distributionoffuelsusedinthedifferentsectorswithinthe area.

Thesourceswithinthesecondcombustioncategoryareengaged primarilyinwastedisposal. Cadmiumemissionsfromthese processesarerelatedtothecadmiumcontaminantlevelsinthe waste. Thedifferentwastetypesaregenerallycharacterized withdistinctsourcecategories.Furthermore,thesewaste disposalpracticesarenotstronglyrelated. Consequently,each ofthesecategorieswillbecharacterizedindividuallywithinthe sectionsbelowratherthaninageneraldiscussionhere. The eightsectionsbelowhaveaconsistentorganization. First,the characteristicsofthefuelorwastearedescribedand,inthe caseofthewastecombustionprocesses,thegeneralsource categoryisalsodescribed. Second,processdescriptionsare presentedandemissionpointsareidentified. Third,available emissioncontrolmeasuresareidentifiedanddescribed. Finally, emissionfactorsarepresented.

COALCOMBUSTION

AspresentedinTable6-1,mostcoalcombustionintheUnited Statesoccursintheutilityandindustrialsectors,withabout 85percentbeingbituminousandlignitecombustionwithinthe utilitysectorandabout14percentbeingbituminousandlignite combustionintheindustrialsector. Consequently,thefocusof thediscussionbelowwillbeonbituminousandlignitecoal combustioninutilityandindustrialboilers. However,limited informationonanthracitecoalcombustionwillalsobepresented.

6-5 CoalCharacteristics

Thecoalcharacteristicsofgreatestinterestinevaluating cadmiumemissionsfromcoalcombustionarecoalheatingvalues andcoalcadmiumcontent. Cadmiumemissionsareadirect functionofthecadmiumcontent,whileheatingvaluesareusedto convertemissionfactorsbetweenmassinput-basedandheat input-basedactivitylevels. Thissectionbrieflysummarizesthe informationaboutcoalheatinglevelsandcadmiumcontent containedinReferences2through4. Morecompletesummariescan befoundinReference2,anddetailedanalysesofcoalcadmium contentasafunctionofcoaltypeandgeographicregioncanbe foundinReferences3and4.

Coalisacomplexcombinationoforganicmatterandinorganic ashformedingeologicformationsfromsuccessivelayersof fallenvegetation. Coaltypesarebroadlyclassifiedas anthracite,bituminous,subbituminous,orlignite,and classificationismadebyheatingvaluesandamountsoffixed carbon,volatilematter,ash,,andmoisture.5 Formulas fordifferentiatingcoalsbasedonthesepropertiesaregivenin Reference6. Thesefourcoaltypesarefurthersubdivided into13componentgroups. Table6-2summarizesinformationabout theheatingvaluesforthesecomponentgroups.2

Theheatingvalueofcoalvariesbetweencoalregions, betweenmineswithinaregion,betweenseamswithinamine,and withinaseam. Thevariabilityisminimalcomparedtothatfound withtracemetallevelsdescribedbelow,butitmaybeimportant whenfuelheatcontentisusedastheactivitylevelmeasurefor sourceemissioncalculations. DatapresentedinTable6-3 illustratetheregionalvariabilityofcoalheatcontent. Heat contentamongcoalsfromseveraldifferentmineswithinaregion appearstoexhibitgreatervariabilitythaneithervariability

6-6 TABLE6-2.COALHEATINGVALUES

Heating value, kJ/kg (Btu/lb)

Component a Coal class group Definition Source Rangea Averagea

Anthracite A1 Meta-anthracite PA,RI 21,580-29,530 25,560 (9,310-12,740) (11,030)

A2 Anthracite CO,PA,NM 27,700-31,800 30,270 (11,950-13,720) (13,001)

A3 Semianthracite AR,PA,VA 27,460-31,750 29,800 (11,850-13,700) (12,860)

Bituminous B1 Low volatile AR,MD,OK,PA, 30,640-34,140 32,400 bituminous WV (13,220-14,730) (13,980)

B2 Medium volatile AL,PA,VA 31,360-33,170 32,170 bituminous (13,530-14,310) (13,880)

B3 High volatile AL,CO,KS,KY, 28,340-35,710 31,170 A bituminous MO,NM,PA, (12,230-14,510) (13,450) TN,TX,UT,VA, WV

B4 High volatile IL,KY,MO,OH,U 26,190-30-480 28,480 B bituminous T,WY (11,300-13,150) (12,290)

B5 High volatile IL,IN,IA,MI 24,450-27,490 26,030 C bituminous (10,550-11,860) (11,230)

Subbituminous S1 Subbituminous A MT,WA 23,940-25,820 24,890 (10,330-11,140) (10,740)

S2 Subbituminous B WY 21,650-22,270 21,970 (9,340-9,610) (9,480)

S3 Subbituminous C CO,WY 19,280-19,890 19,580 (8,320-8,580) (8,450)

Lignite L1 Lignite A ND,TX 16,130-17,030 16,660 (6,960-7,350) (7,190)

L2 Lignite B NA NA NA

Source: Reference 2. aNA = Not available.

6-7 TABLE 6-3. EXAMPLES OF COAL HEAT CONTENT VARIABILITY

Coal heat content, Btu/lb Percent variation about Variability Coal source Mean Range the mean

Intermine Eastern U.S. 12,320 10,750 - 13,891 variability Central U.S. 10,772 9,147 - 12,397 12.7

Western U.S. 11,227 9,317 - 13,134 15

17

Eastern U.S. 12,950 NAa 10,008 9,182 - 10,834 12,000 11,335 - 12,665 Intramine 4.8b variability Central U.S. 12,480 8.0 NAa 5.5 10,975 9,667 - 12,284 Source: Reference 2. Western U.S. 10,351 9,791 - 10,911 5.7c a NA = not available. 12.0 5.4 6-8 b Based on a standard deviation of 624.

c Based on a standard deviation of 708. Eastern U.S. 12,230 NAa 3.0d Intraseam variabilityd Central U.S. 10,709 10,304 - 11,113 Based on a standard deviation of 371. 3.7 Western U.S. 11,540 e Based on a standard deviation of 291. NAa 2.5e withinamineorwithinaseam. Forthesamplepointspresented inTable6-3,interminevariabilityaveraged15percent, intraminevariability7percent,andintraseamvariability 3percent. Becausefewcombustionsourcesburncoalfromjust oneseamoronemine,coalheatcontentvariabilitymay significantlyaffectemissionestimatesthatarebeingcalculated usingemissionfactors,coalusedata,andcoalheatcontent data,evenifthesourcegetsallitscoalfromthesameareaof thecountry.2

Toanevengreaterextentthantheheatingvalue,thecadmium contentofcoalvariessubstantiallyamongcoaltypes,at differentlocationsinthesamemine,andacrossgeographic regions. Themostcomprehensivesourceofinformationoncoal compositionistheUnitedStatesGeologicalSurvey(USGS) NationalCoalResourcesDataSystem(NCRDS). Geochemicaland traceelementdataarestoredwithintheUSCHEMfileofNCRDS. AsofOctober1982,thefilecontainedinformationon7,533coal samplesrepresentingallUnitedStatescoalprovinces. Trace elementanalysisforabout4,400coalsampleswereincludedin thedatabase. Thiscomputerizeddatasystemwasnotaccessed duringthecurrentstudyduetotimeandbudgetaryconstraints andinformationfromUSGSthatindicatedthatfewdatahadbeen addedtothesystemsince1972;however,asummaryofthedata presentedinReference2wasreviewed. Themostextensivesource ofpublishedtraceelementdatawasproducedbySwansonetal.of theUSGS.4 Thisreportcontainsdatafor799coalsamplestaken from150producingminesandincludesthemostimportantUnited Statescoalseams. DatafromtheSwansonstudywastheinitial inputintotheUSCHEMfileofNCRDS. Theinformationpresented heresummarizesBrooks’reviewoftheresultspublishedbyWhite andSwanson.2-4 Notethattheseresultsareconsistentwith unpublishedanalysesconductedbyUSGSonthedatacontainedin NCRDSasof1989.7

6-9

Table6-4presentsinformationonthemeanconcentrationof cadmiumincoalandonthedistributionsofcadmium concentrationsbycoaltype. Bituminouscoalshavethehighest meancadmiumconcentration,0.91partspermillionby weight(ppmwt). Thestandarddeviationofthemean,7.3ppmwt, exceedsthemean,indicatingsubstantialvariationwithinthe data. Bituminouscoalshavethegreatestreportedrangeof cadmiumconcentrations(<0.02to100ppmwt).2 Basedon conversationswithUSGSpersonnel,themeansreportedin Table6-4areregardedastypicalvaluesforin-groundcadmium concentrationincoalsintheUnitedStates.

Theconcentrationofcadmiumincoalalsovariesby geographicregionfromwhichthecoalismined. Basedonthe "besttypical"valuesforeachregion,whicharefootnotedin Table6-5,coalsfromtheInteriorProvincehavethehighestmean cadmiumconcentration,5.47ppmwt. Thatstudyalsoshowedthat thegreatestrangeofconcentrationsisfoundincoalsfromthe InteriorProvince,withareportedrangeof<0.02to100ppmwt. Also,basedonthebestavailabledata,thelowestmean concentrationisfoundincoalsfromtheAppalachianregion (0.13ppmwt).2 ThemeansreportedinTable6-5mayberegarded astypicalin-groundconcentrationsofcadmiumincoalsfromeach geographicregion.

ProcessDescription2,5,8

AsshowninTable6-1,almostallcoalcombustionoccursin utilityandindustrialboilers. Almostallofthecoalburnedis bituminousandsubbituminous(95percent)andlignite (4percent).2 However,theprocessesusedforthedifferent coalsarecomparable. Theparagraphsbelowfirstdescribethe boilersusedforbituminouscoalcombustion. Then,ligniteand

6-10 TABLE6-4.CADMIUMCONCENTRATIONINCOALBYCOALTYPE

Cadmium concentration, ppmwt

Cadmium concentration Number of Coal type range, ppmwt Mean Standard deviation samples

Bituminous <0.02 to 100 0.91 7.3 3,527 Subbituminous 0.04 to 3.7 0.38 0.47 640 Anthracite 0.1 to 0.3 0.22 0.30 52 Lignite <0.11 to 5.5 0.55 0.61 183

Source: Reference 2.

TABLE6-5.CADMIUMCONCENTRATIONINCOALBYREGION

Cadmium concentration, ppmwt Number of Region samples Range Arithmetic mean Standard deviation

Appalachian 2,749 --- 0.13a 0.21 331 0.03-6.8 0.7 ---

Interior 592 --- 5.47a 18.5 155 <0.02-100 7.1 ---

Illinois Basin 82 0.1-65 2.89 ---

Gulf Province 38 --- 0.50a 0.49 34 <0.11-5.5 1.3 ---

Northern Plains 371 --- 0.30a 0.48 490 0.02-2.7 0.08 ---

Rocky Mountains 512 --- 0.35a 0.38 124 <0.03-0.5 --- <0.5

Alaska 107 --- 0.28a 0.59 18 <0.1-0.7 <0.2 ---

Source: Reference 2. aValues are based on the most comprehensive data set currently available and may be used as typical values for cadmium in coal from these regions.

6-11 anthracitecombustionaredescribedbriefly. References5 and8offeradditionaldetailsontheseprocesses.

Thetwomajorcoalcombustiontechniquesusedtofire bituminousandsubbituminouscoalsaresuspensionfiringand gratefiring. Suspensionfiringistheprimarycombustion mechanisminpulverizedcoalandcyclonesystems. Gratefiring istheprimarymechanisminunderfeedandoverfeedstokers. Both mechanismsareemployedinspreaderstokers.

Pulverizedcoalfurnacesareusedprimarilyinutilityand largeindustrialboilers. Inthesesystems,thecoalis pulverizedinamilltotheconsistencyoftalcumpower(i.e.,at least70percentoftheparticleswillpassthrougha200-mesh sieve). Thepulverizedcoalisgenerallyentrainedinprimary airandsuspension-firedthroughtheburnerstothecombustion chamber. Pulverizedcoalfurnacesareclassifiedaseitherdry orwetbottom,dependingontheashremovaltechnique. Dry bottomfurnacesfirecoalswithhighashfusiontemperatures,and dryashremovaltechniquesareused. Inwetbottom(slagtap) furnaces,coalswithlowashfusiontemperaturesareused,and moltenashisdrainedfromthebottomofthefurnace.

Cyclonefurnacesburnlowashfusiontemperaturecoalcrushed toa4-meshsize. Thecoalisfedtangentially,withprimary air,toahorizontalcylindricalcombustionchamber. Smallcoal particlesareburnedinsuspension,whilethelargerparticles areforcedagainsttheouterwall. Becauseofthehigh temperaturesdevelopedintherelativelysmallfurnacevolume, andbecauseofthelowfusiontemperatureofthecoalash,much oftheashformsaliquidslagthatisdrainedfromthebottomof thefurnacethroughaslagtapopening. Cyclonefurnacesare usedmostlyinutilityandlargeindustrialapplications.

6-12 Inspreaderstokers,aflippingmechanismthrowsthecoal intothefurnaceandontoamovingfuelbed.Combustionoccurs partiallyinsuspensionandpartiallyonthegrate. Becausethe entrainedparticlesinthefurnaceexhausthavesubstantial carbon,flyashreinjectionfrommechanicalcollectorsiscommonly usedtoimproveboilerefficiency. Ashresidueinthefuelbed isdepositedinareceivingpitattheendofthegrate.

Inoverfeedstokers,coalisfedontoatravelingor vibratinggrateandburnsonthefuelbedasitprogresses throughthefurnace. Ashparticlesfallintoanashpitatthe rearofthestoker. "Overfeed"appliesbecausethecoalisfed ontothemovinggrateunderanadjustablegate. Conversely,in "underfeed"stokers,coalisfedintothefiringzonefrom underneathbymechanicalramsorscrewconveyers. Thecoalmoves inachannel,knownasaretort,fromwhichitisforcedupward, spillingoverthetopofeachsidetofeedthefuelbed. Combustioniscompletedbythetimethebedreachesthesidedump grates,fromwhichtheashisdischargedtoshallowpits.

ThenextmostcommoncoalusedintheUnitedStatesis lignite. Ligniteisarelativelyyoungcoalwithproperties intermediatetothoseofbituminouscoalandpeat. Because lignitehasahighmoisturecontent(35to40weightpercent)and alowwetbasisheatingvalue(3,900kcal/kg[7,200Btu/lb]),it generallyisusedasafuelonlyinareasinwhichitismined. Ligniteisusedmainlyforsteam/electricproductioninpower plantsandtypicallyisfiredinlargerpulverizedcoal-firedor cyclone-firedboilers.

Anthracitecoalisahigh-rankcoalwithmorefixedcarbon andlessvolatilematterthaneitherbituminouscoalorlignite. Becauseofitslowvolatilemattercontentandslightclinkering, anthraciteismostcommonlyfiredinmedium-sizedtravelinggrate

6-13 stokersandsmallhand-firedunits. Someanthracite (occasionallywithpetroleumcoke)isusedinpulverizedcoal- firedboilers,anditmaybeblendedwithbituminouscoal. Becauseofitslowsulfurcontent(typicallylessthan0.8weight percent)andminimalsmokingtendencies,anthraciteisconsidered adesirablefuelinareaswhereitisreadilyavailable. Inthe UnitedStates,anthraciteisminedinnortheasternPennsylvania andisconsumedmostlyinPennsylvaniaandsurroundingStates. Thelargestuseofanthraciteisforspaceheating. Lesser amountsareemployedforsteam/electricproduction,typicallyin underfeedstokerandpulverizedcoaldry-bottomboilers.

Althoughsmallquantitiesofcadmiummaybeemittedas fugitiveparticulatematterfromcoalstorageandhandling operations,theprimarysourceofcadmiumemissionsfromcoal combustionisthecombustionstack. Becausethecombustionzone inboilersoperatesattemperaturesinexcessof1100°C(2000°F), thecadmiuminthecoalisvolatilized. Asthefluegascoolsin theconvectiveheattransfersectionandfurtherintheair preheater,thevolatilizedcadmiumcondenses. Thecadmiummay condenseoradsorbontoexistingparticlesaccordingtothe availablesurfaceareaoritmaycondensehomogeneously,forming fineparticles. Thecadmiumthusvolatilizedwouldbedepleted inthebottomashandconcentratedintheflyashsincethefly ashhasmorerelativesurfaceareathanthebottomashandsince thebottomashdoesnotcomeincontactwiththevolatilized cadmiumlongenoughforthecadmiumtocondenseonthebottom ash.

EmissionControlMeasures6

Emissioncontrolmeasuresforcoal-firedboilersinclude controlsbasedoncombustordesignandoperatingpracticesthat aredirectedprimarilyatnitrogenoxides(NOx)andparticulate

6-14 matter(PM)controlandadd-onairpollutioncontroldevicesthat aredesignedforacidgasandPMcontrol. Thosemeasuresthat aremostlikelytoaffectcadmiumcontrolareadd-onPMandacid gascontroldevices. TheprimarytypesofPMcontroldevices usedforcoalcombustionincludemultiplecyclones,electrostatic precipitators,fabricfilters(baghouses),andscrubbers,while bothwetanddryfluegasdesulfurization(FGD)systemsareused forsulfurdioxide(SO2). SomemeasureofPMcontrolisalso obtainedfromashsettlinginboiler/airheater/economizerdust hoppers,largebreeches,andchimneybases,butthesemechanisms willnotsignificantlyreducecadmiumemissions.

Electrostaticprecipitators(ESP)arethemostcommonhigh efficiencycontroldeviceusedonpulverizedcoalandcyclone units. Thesedevicesarealsobeingusedincreasinglyon stokers. Generally,PMcollectionefficienciesareafunctionof thespecificcollectionarea(i.e.,theratioofthecollection plateareatothevolumetricflowrateoffluegasthroughthe device),andPMefficienciesof99.9weightpercenthavebeen measuredwithESP’s. Fabricfiltershaverecentlyseenincreased useinbothutilityandindustrialapplicationsbothasaPM controlmeasureandasthecollectionmechanismindryFGD systems,generallyeffectingabout99.8percentPMcontrol efficiency. ScrubbersarealsousedtocontrolPMemissions, althoughtheirprimaryuseistocontrolemissionsofsulfur .

Mechanicalcollectors,generallymultiplecyclones,arethe primarymeansofcontrolonmanystokersandaresometimes installedupstreamofhighefficiencycontroldevicesinorderto reducetheashcollectionburden. Dependingonapplicationand design,multiplecyclonePMefficienciescanvarytremendously. However,thesesystemsarerelativelyinefficientforfine particlesandarenotlikelytoprovidemeasurablecontrolof

6-15 cadmiumemissions,whichareprimarilyinthefineparticle fractionsoftheexhaust.

Dataontheperformanceofcoalcombustionemissioncontrol measures,relativetocadmium,arequitesparse. Furthermore, manyofthedatathatareavailablearesomewhatdatedandareof questionablereliability. Thesectiononemissionfactorsbelow presentstheavailabledataonemissioncontrolsystem performance. However,inevaluatingthepotentialemissionsfrom afacilityorgroupoffacilities,anyassumptionsaboutcontrol systemperformance,includingthosebasedonthedatapresented herein,shouldbeexaminedcarefullytoassurethattheyare supportedbyreliabletestdataobtainedviamethodscomparable tothosedescribedinSection9. Also,performanceestimates mustbeconsistentwiththephysicalandchemicalpropertiesof thecompoundsbeingemittedandwiththeoperating characteristicsofthesystemsbeingevaluated.

Emissions

Twodistinctsourcesofinformationwereusedtodevelopand evaluatecadmiumemissionfactorsforcoalcombustion. First, thedatapresentedaboveoncadmiumconcentrationsincoaland coalheatingvalueswereusedtodevelopmassbalance-based emissionfactorsundertheassumptionthatallcadmiumcharged withthecoalisemittedasfinePMinthestackgas. Second, theemissionfactorspresentedinthecoalandoilLocatingand Estimating(L&E)documentwerereviewedandsummarized.2 No attemptwasmadetoverifythesourcesofdatausedinthecoal andoilL&Edocumentortoratetheemissionfactorsthatwere developedtherein. Theresultsobtainedfromeachofthethese methodsarediscussedseparatelyintheparagraphsbelow. Then therelativemeritsoftheemissionfactorsobtainedbythe

6-16 differentmethodsareexamined,andthebesttypicalemission factorsareidentified.

Theinformationpresentedintheliteratureindicatesthat virtuallyallofthecadmiumcontainedinthecoalisemitted fromthefurnaceasfinePM. Consequently,thecoalheating valuespresentedinTable6-2andthecoalcadmiumconcentrations presentedinTable6-4canbeusedtodevelopemissionfactors formajorcoaltypesundertheassumptionthatallcadmiuminthe coalisemitted. Furthermore,notethatthecoalcomposition datainTable6-2arebasedonin-groundcadmiumconcentrations. ThecalculatedemissionfactorsshowninTable6-6arebasedon theassumptionthatas-firedcoalcontainsequivalent concentrations. Ifcadmiumconcentrationsarereducedduring coalcleaningoperations,theseestimateswillbebiasedhigh. PreliminarydatafromtheUnitedStatesDepartmentofEnergy indicatesthatthereissomereductionincadmiumconcentrations (25to50percent)fromcoalcleaning.9 Thecadmiumemission factorsderivedfromthesereducedcadmiumconcentrationsare alsoshowninTable6-6.

6-17 TABLE6-6.CALCULATEDUNCONTROLLEDCADMIUMEMISSIONFACTORS FORCOALCOMBUSTION

Calculatedcadmiumemissionfactorsa

15 12 Coaltype kg/10 J lb/10 Btu

Bituminousb 30(15-23) 70(35-53)

Subbituminousc 17(8.6-13) 40(20-30)

Anthracited 7.3(3.6-5.5) 17(8.4-13)

Lignite 33(17-25) 76(38-57) aValuesinparenthesesarebasedona25to50percentreductionincadmiumconcentrationsfromcoal cleaning. bBasedonarithmeticaverageofthefiveaverageheatingvaluesinTable6-2. cBasedonarithmeticaverageofthethreeaverageheatingvaluesinTable6-2. dBasedonaverageheatingvalueforcoalcategoryA2inTable6-2.

Acomprehensivesummaryofthetestdatageneratedpriorto 1989forcoal-firedboilersandfurnaceswaspresentedin Reference2. Thedatafromindividualteststhatwerepresented inthatreportarecompiledinTableB-1inAppendixB. Table6-7summarizesthesedataasafunctionofcoaltypeand controlstatus. Notethewiderangeofemissionfactorsforeach coaltype. Thisrangereflectsthesubstantialvariationincoal cadmiumcontentandhighlightstheneedtoobtaincoal-specific cadmiumdatatocalculateemissionestimateswheneverpossible. Alsonotethatthedataarecombinedacrossindustrysectorand boilertypebecausetheseparametersarenotexpectedtohavea substantialeffectonemissionfactors.

AsshowninReference2,theavailabletestdataoncadmium emissioncontrolarequitelimited. ExceptfortheESP,control efficienciesarebasedononlyafewdatapointsand,therefore, maynotbeveryreliable. AccordingtoReference2, approximately75percentcontrolofcadmiumemissionsisachieved

6-18 TABLE6-7.SUMMARYOFCADMIUMEMISSIONFACTORSFORCOALCOMBUSTION

Cadmium emission factors No. of Coal No. of data kg/1015 J lb/1012 Btu typea Control statusb boilers points Average Range Average Range

B UN 19 40 34 1.8-130 80 4.1-300

B MP or MC 11 20 38 0.082-210 88 0.19-490

B ESP or MP/ESP 30 45 2.5 0.0040-23 5.8 0.0090-53

B ESP-2 stage 1 1 19.8 -- 46 --

B WS or MC/WS 6 6 42 0.037-210 98 0.086-490

B FF 1 1 0.14 -- 0.33 --

SB UN 3 5 970 2.1-1,900 2,200 4.9-4,400

SB ESP or MP/ESP 4 5 1.6 0.17-6.0 3.7 0.39-14

SB WS 2 2 110 1.7-210 250 4.0-490

L MC 4 4 5.3 2.2-11 12 5.1-26

L ESP 3 3 0.95 0.52-1.5 2.2 1.2-3.5

L ESP/WS 1 2 13 0.77-25 31 1.8-59

A UN 3 3 1.0 0.60-1.5 2.4 1.4-3.5

Source: Reference 2. aB = bituminous, SB = subbituminous, L = lignite, A = anthracite. bUN = uncontrolled, MP = mechanical precipitation system, MC = multiclone, ESP = electrostatic precipitator, WS = wet scrubber.

6-19 byESP’s. Atleast75percentcontrolshouldbeachievedbya combinationofanESPandawetscrubberorbytwoESP’sin series. Insufficientdataareavailableontheperformanceof wetscrubbersrelativetocadmiumemissions,but,accordingto theliterature,wetscrubberscanachieveover99percentremoval ofPM. Cadmiumreductionwithawetscrubberisexpectedtobe lessthanthatsincecadmiumpartitionswiththefinePM,andwet scrubbersaremuchlesseffectiveinreducingemissionsoffine PM.10 Aconservativeestimateofcadmiumreductionwithawet scrubberwouldbe75percentsincePMcontrolwithascrubberis atleastasgoodasPMcontrolwithanESPsystem. Thereported 29percentcontrolachievedbymulticlonesisconsistentwiththe inefficiencyofthesesystemsinreducingcadmiumemissions.

Basedonreviewoftheavailabledata,thebestestimates foruncontrolledemissionfactorsfortypicalcoalcombustion facilitiesarethoseobtainedfromamassbalanceusingcoal compositiondata. Thisapproachwasselectedbecausethe availabletestdataareofuncertainquality,andthecoal concentrationdataarerepresentativeofamuchlargerindustry segment. Controlledemissionfactorswereobtainedbyapplying 75percentcontrolforESP’s,greaterthan75percentcontrolfor acombinationofESP’sandwetscrubbers,andgreaterthan 75percentcontrolfortwoESP’sinseries. Datawereinadequate toestimateefficienciesforsystemsequippedwithmechanical collectors,wetscrubbers,orfabricfilters. Theresultantbest typicalemissionfactorsareshowninTable6-8.

FUELOILCOMBUSTION

AsshowninTable6-1,fueloilusespansthefoursectors ofenergyusers. Distillatefueloilisusedextensivelyinall sectorswiththelargestuseintheutility(31percent)andthe industrial(32percent)sectors,butwithsubstantialamounts

6-20 TABLE6-8.BESTTYPICALCADMIUMEMISSIONFACTORSFORCOALCOMBUSTION

Cadmium emission factors Control a b Coal type status kg/1015 J lb/1012 Btu

B Uncontrolled 30 70

B ESP 7.7 18

B ESP/wet scrubber <7.7 <18

B ESP-2 stage <7.7 <18

SB Uncontrolled 17 40

SB ESP 4.4 10

SB ESP/wet scrubber <4.4 <10

SB ESP-2 stage <4.4 <10

A Uncontrolled 7.3 17

A ESP 1.8 4.3

A ESP/wet scrubber <1.8 <4.3

A ESP-2 stage <1.8 <4.3

L Uncontrolled 33 76

L ESP 8.4 19

L ESP/wet scrubber <8.4 <19

L ESP-2 stage <8.4 <19 aB = bituminous, SB = subbituminous, A = anthracite, L = lignite. bESP = electrostatic precipitator.

6-21 usedinboththecommercial(13percent)andresidential (23percent)sectors. Residualoilisusedprimarilyinthe industrial(56percent)andcommercial(33percent)sectors. Becausetheoilcombustionprocessisnotcomplex,andcontrol systemsarenotwidelyappliedtooil-firedunits,thediscussion belowwillfocusonfueloilcharacteristicsandonemissions fromoil-firedunits.1

FuelOilCharacteristics2

Thefueloilcharacteristicsofgreatestimportancefor characterizingcadmiumemissionsfromfueloilcombustionarethe heatingvalueandthecadmiumcontentoftheoil. Theheating valueisusedforconvertingfromemissionfactorswithmass-or volume-basedactivitylevelstothosewithactivitylevelsbased onheatinput.

Thetermfueloilcoversavarietyofpetroleumproducts, includingcrudepetroleum,lighterpetroleumfractionssuchas kerosene,andheavierresidualfractionsleftafterdistillation. Toprovidestandardizationandmeansforcomparison, specificationshavebeenestablishedthatseparatefueloilsinto variousgrades. Fueloilsaregradedaccordingtospecific gravityandviscosity,withNo.1Gradebeingthelightestand No.6theheaviest. Theheatingvalueoffueloilsisexpressed intermsofkJ/L(Btu/gal)ofoilat16°C(60°F)orkJ/kg (Btu/lb)ofoilat16°C(60°F). Theheatingvaluepergallon increaseswithspecificgravitybecausethereismoreweightper gallon. Theheatingvaluepermassofoilvariesinverselywith specificgravitybecauselighteroilcontainsmorehydrogen. For anuncrackeddistillateorresidualoil,heatingvaluecanbe approximatedbythefollowingequation:

Btu/lb =17,660+(69xAPIgravity)

6-22 Foracrackeddistillate,therelationshipbecomes:

Btu/lb=17,780+(54xAPIgravity) Table6-9providesanoverallsummaryoftheheatingvalues oftypicalfueloilsusedintheUnitedStates,andTable6-10 showsthevariabilityinfueloilheatingvaluesusedinvarious regionsofthecountry. AppendixBofReference2provides additionaldetails.

Thedatabaseforcadmiumcontentinfueloilsismuchmore limitedthanwasthecoalcadmiumcontentdatabase. Nosingle centralizeddatabaseisavailable,andtheinformationpresented belowisbasedonlimiteddatafromindividualstudies.

Concentrationsofcadmiuminfueloildependuponthetype ofoilused. Nocomprehensiveoilcharacterizationstudieshave beendone,butdataintheliteraturereportsimilarcadmium concentrationmeansandrangesinresidualanddistillateoils. Thesuggestedtypicalcadmiumcontentofresidualoilis 0.30ppmwt,andthatofdistillateoilis0.21ppmwt. The typicalvalueforcadmiumincrudeoilis0.03ppmwt. Table6-11 liststhetypicalvaluesforcadmiuminoils. Thetypicalvalues fordistillateandcrudeoilwereobtainedbytakingtheaverage ofthemeanvaluesfoundintheliterature. Thevaluefor residualoilwasbasedonreportedconcentrationswithoutusing thetwohighvaluesof2.27and2.02ppmwt.

ProcessDescription5,8

Fueloilsarebroadlyclassifiedintotwomajortypes: distillateandresidual. Distillateoils(fueloilgrade Nos.1and2)areusedmainlyindomesticandsmallcommercial applicationsinwhicheasyfuelburningisrequired. Distillates aremorevolatileandlessviscousthanresidualoils,having

6-23 TABLE 6-9. TYPICAL HEATING VALUES OF FUEL OILS

FUEL OIL GRADES No.1 No.2 No.4 No.5 No.6 −− b Type Distillate Distillate Very light residual Light residual Residual Crude Color Light Amber Black Black Black −− Heating valuea

kJ/L (Btu/gal) 38,200 40,900 40,700 41,200 41,800 39,900−42,200 (137,000) (141,000) (146,000) (148,000) (150,000) (143,000−152,000) kJ/kg (Btu/lb) 45,590−46,030 44,430−45,770 42,370−44,960 41,950−44,080 40,350−43,800 40,700−43,200 (19,670−19,860) (19,170−19,750) (18,280−19,400) (18,100−19,020) (17,410−18,900) (17,500−18,600)

Source: References 2 and 11. 6-24 aThe distillate and residual oil samples analyzed for Btu/gal and Btu/lb heating values are different; therefore, the heating values presented do not directly correspond to one another.

bThese crude oil values are based on a limited number of samples from West Coast field sites presented in Reference 11 and may not be representative of the distribution of crude oils processed in the United States. TABLE 6-10. TYPICAL FUEL OIL HEATING VALUES FOR SPECIFIC REGIONS2

No. 1 fuel oil No. 2 fuel oil No. 4 fuel oil

Heating value, kJ/L (Btu/gal) Heating value, kJ/L (Btu/gal) Heating value, kJ/L (Btu/gal)

No. of No. of No. of Region samples Range Average samples Range Average samples Range Average

Eastern 33 36,900-37,800 37,400 56 37,100-40,800 38,800 1 -- 40,700 (132,500-135,700) (134,200) (133,100-146,600) (139,500) -- (146,000)

Southern 13 37,000-37,700 37,400 19 38,000-39,400 38,800 0 -- -- (132,900-135,400) (134,300) (136,400-141,500) (139,400) -- --

Central 27 36,900-37,800 37,300 35 37,800-40,800 38,800 2 40-700-41,800 41,200 (132,500-135,700) (134,000) (135,900-146,600) (139,200) (146,000-150,100) (148,000)

Rocky 14 37,100-37,600 37,400 17 37,900-39,100 38,700 2 41,800-41,900 41,900 Mountain (133,100-135,100) (134,200) (136,100-140,400) (139,000) (150,100-150,500) (150,300)

Western 16 36,700-37,900 37,500 18 37,900-39,100 38,700 0 -- -- (131,700-136,200) (134,600) (136,100-140,500) (139,000) -- -- 6-25

No. 5 fuel oil No. 6 fuel oil

Heating value, kJ/L (Btu/gal) Heating value, kJ/L (Btu/gal)

No. of No. of samples Region samples Range Average Range Average

Eastern 1 -- 41,300 17 40,800-43,900 43,300 -- (148,400) (147,000-167,600) 161,900

Southern 0 -- -- 14 41,900-43,600 42,600 -- -- (150,500-156,500) (152,900)

Central 4 41,300-42,200 41,700 10 41,900-44,200 42,600 (148,400-151,500) (149,900) (160,600-158,900) (152,900)

Rocky 2 42,900-43,600 43,200 7 42,300-44,300 43,000 Mountain (153,900-156,500) (155,200) (161,900-159,200) (154,600)

Western 0 -- -- 12 41,700-45,500 43,000 -- -- (149,900-163,500) (154,400)

Source: Reference 2. negligibleashandnitrogencontentsandusuallycontainingless than0.3weightpercentsulfur. Residualoils(gradeNos.4,5, and6),ontheotherhand,areusedmainlyinutility,

TABLE6-11.CADMIUMCONCENTRATIONINOILBYOILTYPE

Cadmiumconcentration,ppmwt

Fueloiltype Range Typicalvalue

ResidualNo.6 0.010-2.3 0.30a

DistillateNo.2 0.010-0.95 0.21b

Crude 0.010-0.05 0.030c

Source:Reference2. aBasedonreportedconcentrationswithoutusingthetwohighvalues,2.3and2.0ppmwt. bAverageoftwostudies. cAverageofthreestudies. industrialandlargecommercialapplicationswithsophisticated combustionequipment. No.4oilissometimesclassifiedasa distillate;No.6issometimesreferredtoasBunkerC. Being moreviscousandlessvolatilethandistillateoils,theheavier residualoils(Nos.5and6)mustbeheatedtofacilitate handlingandproperatomization. Becauseresidualoilsare producedfromtheresidueafterlighterfractions(gasoline, kerosene,anddistillateoils)havebeenremovedfromthecrude oil,theycontainsignificantquantitiesofash,,and sulfur.

Oil-firedboilersandfurnacesaresimplerandhavemuch lessvariationindesignthanthecoal-firedsystemsdescribed earlier. Theprimarycomponentsofthesystemaretheburnerand thefurnace. Theburneratomizesthefuelandintroducesit alongwiththecombustionairintotheunitviatheflame. The

6-26 furnaceprovidestheresidencetimeandmixingneededtocomplete combustionofthefuel. Theprimarydifferenceinsystemsthat firedistillateoilandresidualoilisthattheresidualoil systemsmusthaveanoilpreheatertoreducetheviscosityofthe oilsothatitcanbeatomizedproperlyintheburner.

Theonlysourceofcadmiumemissionsfromoil-firedboilers andfurnacesisthecombustionstack. Becausetheentirefuel supplyisexposedtohighflametemperatures,essentiallyallof thecadmiumcontainedinthefueloilwillbevolatilized,with mostcondensingontosmallparticlesandthenexhaustedfromthe combustionchamberwiththecombustiongasexhaust. Unlessthe combustiongasesareexposedtohigh-efficiencyPMcontrol systems,whichtypicallyarenotfoundonoil-firedunits,the cadmiumwillbeexhaustedasfinePMthroughthecombustion stack.

EmissionControlMeasures8

Thethreetypesofcontrolmeasuresappliedtooil-fired boilersandfurnacesareboilermodifications,fuelsubstitution, andfluegascleaningsystems. Onlyfuelsubstitutionandflue gascleaningsystemswillaffectcadmiumemissions. Fuel

substitutionisusedprimarilytoreduceSO2 andNOx emissions. However,ifthesubstitutedfuelshavelowercadmium concentrations,thesubstitutionwillalsoreducecadmium emissions. BecausePMemissionsfromoil-firedunitsare generallymuchlowerthanthosefromcoal-firedunits, high-efficiencyPMcontrolsystemsaregenerallynotemployedon oil-firedsystems. Consequently,thesefluegascleaningsystems arenotlikelytoachievesubstantialcadmiumcontrol. However, thefluegassystemsthataretypicallyusedonoil-firedunits aredescribedbrieflybelow.

6-27 Fluegascleaningequipmentgenerallyisemployedonlyon largeroil-firedboilers. Mechanicalcollectors,aprevalent typeofcontroldevice,areprimarilyusefulincontrollingPM generatedduringsootblowing,duringupsetconditions,orwhena verydirtyheavyoilisfired. Duringthesesituations,high efficiencycycloniccollectorscanaffectupto85percent controlofPM,butlesscontrolofcadmiumisexpectedwith mechanicalcollectorsbecausecadmiumisenrichedontofinePM, whichisnotaseasilycapturedbycontroldevices.

Electrostaticprecipitatorsarecommonlyusedinoil-fired powerplants. OlderESP’smayremove40to60percentofthePM, butlowercadmiumcontrolisexpectedbecauseofthereasoncited above. Scrubbingsystemshavebeeninstalledonoil-fired boilerstocontrolbothsulfuroxidesandPM. Thesesystemscan achievePMcontrolefficienciesof50to60percent. Because theyprovidegascoolingbelowthecondensationpointofcadmium, somecadmiumcontrolmaybeobtained,butnodataareavailable ontheirperformance.

Emissions

Threetypesofinformationwereusedtodevelopemission factorsforoilcombustion. First,thedatadescribedaboveon fueloilheatingvalueandcadmiumcontentoffueloilswereused todevelopemissionfactorsbymassbalance,assumingthatall cadmiumfiredwiththefueloilisemittedthroughthestack. Second,theemissionfactorsfromthecoalandoilL&Edocument wereevaluatedandsummarized,butnoattemptwasmadetoverify originalreferencesortoratethesedata. Finally,rated emissiontestdatadevelopedinpreparationofthisdocumentwere evaluatedandsummarized. Theparagraphsbelowfirstpresentthe resultsgeneratedfromeachofthethreesources. Then,the relativemeritsoftheemissionfactorsgeneratedviaeachofthe

6-28 proceduresarediscussed,andthebest"typical"emissionfactors areidentified.

Theliteratureonfueloilcombustionsuggeststhat essentiallyallcadmiuminthefueloilisvolatilizedinthe combustionzone,withmostcondensingontosmallparticlesand exhaustedasfinePMinthecombustiongasstream. Usingthe assumptionthat100percentofthecadmiuminfueloilleavesthe boilerorfurnaceintheexhaustgases,thedatainTables6-9 and6-11canbeusedtocalculateuncontrolledemissionfactors forNo.2distillateandNo.6residualoil. Datapresentedin Reference8,whichshowanaveragecrudeoilheatingvalueof 42,500kJ/kg(18,000Btu/lb),canbecombinedwiththecadmium contentdatainTable6-11tocalculateanuncontrolledemission factorforcrudeoilcombustion. Theresultsofthese calculationsarepresentedinTable6-12.

TABLE6-12.CALCULATEDUNCONTROLLEDCADMIUMEMISSIONFACTORS FORFUELOILCOMBUSTION

Cadmiumemissionfactors

Fueloiltype kg/1015 J lb/1012 Btu

Crude 0.71 1.7

No.6Residual 7.1 17

No.2Distillate 4.7 11

Acomprehensivesummaryoftheemissiondatageneratedprior to1989waspreparedbyBrooks.2 Theseresultsaretabulatedin Table6-13. Themeasuredcadmiumemissionfactorsrangefrom 0.021kg/1015J(0.048lb/1012Btu)to91kg/1015J(212lb/1012Btu).

6-29 TABLE6-13.MEASUREDCADMIUMEMISSIONFACTORSFORFUELOILCOMBUSTION

Cadmiumemissionfactors FuelCd Industry Control content, kg/1015 J lb/1012 Btu Fueloiltype sectora statusb ppmwt Date

No.6oil I MC/WS 3.5 21 49c 1979

No.6oil I MC 3.5 91 210d 1979

No.6oil I UN 3.0 0.30 0.69 1978

No.6oil I UN 3.0 1.3 3.0 1978

No.6oil I UN 3.0 3.7 8.6 1978

1:1Residual/Crude U UN 0.70 0.021 0.048 1981

1:1Residual/Crude U UN 0.70 3.5 8.2 1981

1:1Residual/Crude U UN 0.50 14 33 1981

Distillatee R UN --- 11 26f 1979

Distillatee R UN 0.10 2.1 4.9 1981

Distillatee R UN 0.10 3.2 7.5 1981

Distillatee R UN 0.19 bdg bdg 1982

Source:Reference2. aU=utility,I=industrial,R=residential. bMC=multiclone,WS=wetscrubber,UN=uncontrolled. cOutletfromawetscrubber. dInlettoawetscrubber. eTypeofdistillateoilnotspecified. fAverageofeighttestsonsevenunits. gbd=belowdetectionlimit.

6-30 Theaveragedistillatevalueis5.4kg/1015J(13lb/1012Btu), similartothecalculatedvalueof4.7kg/1015J(11lb/1012Btu). ThevaluesforNo.6residualoilfromthe1979studyarehigher thanvaluesreportedintheotherstudiesdespitethepresenceof PMcontroldevices. Thecausesofthelargevariationin measuredcadmiumemissionfactorsareunknown.2 Consequently, thetestdatainTable6-13shouldbeusedcautiously.

Asapartofthisstudy,threetestreportspreparedasa partoftheCalifornia"HotSpots"programwerereviewed.11-13 The emissionfactorsgeneratedfromthesethreereportsare summarizedinTable6-14. Eachofthereportscontainedthedata onfueloilcharacteristicsneededtocalculatecadmiuminput rates,soTable6-14containsbothcalculatedemissionfactors basedoncadmiuminputlevelsandmeasuredemissionfactorsbased onstacktests. Becausecadmiumlevelsinallofthefueloils testedwerebelowdetectionlimits,allcalculatedemission factorsarereportedas"lessthan"values. Notethatallthree stacktestsshowedcadmiumemissionlevelsabovethedetection limitinthestackbutsubstantiallybelowthedetectionlimit forcadmiuminfueloil. Ifcadmiumlevelsinfueloilareclose tothedetectionlimit,thenthetestsshowedmeasuredemissions tobesubstantiallylessthancadmiuminputtotheprocess. On balance,thesedataprovidelittleinformationforemission factordevelopment.

6-31 TABLE6-14.CADMIUMEMISSIONFACTORSFORFUELOILCOMBUSTION GENERATEDFROMCALIFORNIA"HOTSPOTS"TESTS

Calculatedcadmiumemission Measuredcadmiumemission factorsa factorsa

Processtype Fueloiltype kg/1015 J lb/1012 Btu kg/1015 J lb/1012 Btu

Pipeline/ Crude <12 <27 2.5 5.8 processheater

Generator Crude <7.3 <17 0.54 1.3

Powerboiler Residual <4.7 <11 0.31 0.72

Source:References11through13. aBasedonassumedcrudeoilheatingvalueof42,500kJ/kg(18,300Btu/lb)andanassumedresidual oilheatingvalueof43,600kJ/kg(18,800Btu/lb).

Giventhelimitedemissiontestdataavailableandthe concernsaboutpossiblebiasesinthosedata,themassbalance approachwasusedtoestimatethebest"typical"emissionfactor fordistillateandresidualfueloilcombustion.

Theavailableinformationonuncontrolledcadmiumemissions fromcrudeoilcombustionisambiguous. Thelimitedtestdata presentedinTables6-13and6-14showmeasuredemissionfactors thatrangefrom0.02to14kg/1015 J(0.05to33lb/1012 Btu),a rangeofalmostthreeordersofmagnitude. Becausethesedata arequitesparseandtherelativequalityofthedatais uncertain,themidpointoftherangewasselectedasthebest "typical"emissionfactor.

Theuncontrolledemissionfactorsfordistillate,residual, andcrudeoilarepresentedinTable6-15. Dataareinsufficient todevelopcontrolledemissionfactorsforfueloilcombustion.

6-32 TABLE6-15.BESTTYPICALCADMIUMEMISSIONFACTORSFORFUELOILCOMBUSTION

Cadmiumemissionfactors

Fueloiltype kg/1015 J lb/1012 Btu

Crude 7.0 16

No.6Residual 7.1 17

No.2Distillate 4.7 11

NATURALGASCOMBUSTION

Naturalgasisoneofthemajorfuelsusedthroughoutthe country. AsshowninTable6-1,naturalgasisusedasanenergy sourceinallfoursectors,butthegreatestusesareinthe industrial(46percent)andresidential(15percent)sectors. Naturalgasisusedasanenergysourcethroughoutthecountry. ThefiveStatesthatconsumethelargestquantitiesofnatural gasareTexas,California,Louisiana,Illinois,andNewYork. However,onlyLouisianaandOklahomaconsumemoreenergyvia naturalgascombustionthanbyeithercoalorpetroleumproducts combustion.1

NaturalGasCharacteristics8,14

Naturalgasisconsideredtobeacleanfuel. Itconsists ofprimarilymethane(generally80percentorgreaterbymass), alongwithvaryingamountsofethane,propane,butane,andinert material(typicallynitrogen,carbondioxide,andhelium). The averageheatingvalueofnaturalgasisabout8,900kilocalories perstandardcubicmeter(kcal/scm)(1,000Britishthermalunits perstandardcubicfoot[Btu/scf]),withlevelsrangingfrom 8,000to9,000kcal/scm(900to1,100Btu/scf). Nodataare availableonthecadmiumcontentofnaturalgas. However, concentrationsareexpectedtobequitelow. Littlecadmiumis

6-33 expectedtobefoundinrawgas,andtheprocessingstepsusedto recoverliquidconstituentsandtoremovehydrogensulfidefrom therawgasshouldalsoremoveanycadmiumthatiscontainedin therawgas.

ProcessDescription14

Naturalgascombustionsourcescanbedividedintofour categories: utility/largeindustrialboilers,smallindustry boilers,commercialboilers,andresidentialfurnaces. These systemsareconfigureddifferently,butthecombustionprocesses arecomparableforallcategories. Thenaturalgasand combustionairaremixedinaburnerandintroducedtoa combustionchamberviaaflame. Thenaturalgasflame temperature,whichexceeds1000°C(1800°F),willvolatilizeany cadmiuminthefuel. Mostofthecadmiumwillthencondenseonto smallparticlesandbeexhaustedasfinePMfromtheboileror furnacewiththecombustiongasstream. Thisexhauststreamis theonlysourceofcadmiumemissionsfromnaturalgascombustion.

EmissionControlMeasures

Nocontrolmeasuresapplytonaturalgas-firedboilers,and furnacesareexpectedtoaffectcadmiumemissions.

Emissions

Nodataareavailableoncadmiumemissionsfromnaturalgas combustion,butemissionsareexpectedtobequitelow. As statedearlier,littlecadmiumisexpectedtobefoundinraw gas,and,giventheprocessingstepsthatnaturalgasundergoes, anycadmiumthatispresentwouldberemovedfromtherawgas. Consequently,nocadmiumemissionfactorispresentedfornatural gascombustion.

6-34 WOODCOMBUSTION

Woodandwoodwastesareusedasfuelinboththeindustrial andresidentialsectors. Intheindustrialsector,woodwasteis firedtoindustrialboilerstoprovideprocessheat,whilewood isfiredtofireplacesandwoodstovesintheresidential sectors. Nodataareavailableonthecadmiumcontentofwood andwoodwastes. Consequently,theinformationbelowincludes processdescriptionsforthethreecombustionprocesses(boilers, fireplaces,andwoodstoves),descriptionsofthecontrol measuresusedforwood-firedprocesses,andemissionfactors.

ProcessDescription14

Woodwastecombustioninboilersismostlyconfinedtothose industriesforwhichitisavailableasabyproduct. These boilersgenerateenergyandalleviatepossiblesolidwaste disposalproblems. Inboilers,woodwasteisnormallyburnedin theformofhoggedwood,sawdust,shavings,chips,sanderdust,or woodtrim. Heatingvaluesforthiswasterangefromabout2,200 to2,700kcal/kg(4,000to5,000Btu/lb)offuelonawet,as- firedbasis. Themoisturecontentistypicallynear50weight percentbutmayvaryfrom5to75weightpercent,dependingon thewastetypeandstorageoperations. Generally,barkisthe majortypeofwasteburnedinpulpmills;eitheramixtureof woodandbarkwasteorwoodwastealoneisburnedmostfrequently inthelumber,furniture,andplywoodindustries. Asof1980, approximately1,600wood-firedboilerswereoperatinginthe UnitedStates,withatotalcapacityofover30gigawatts(GW) (1.0x1011 Btu/hr). Nospecificdataonthedistributionof theseboilerswereidentified,butmostarelikelytobelocated intheSoutheast,thePacificNorthwestStates,Wisconsin, Michigan,andMaine.

6-35 Themostcommonfiringmethodemployedforlargerwood-fired boilersisthespreaderstoker. Woodentersthefurnacethrough afuelchuteandisspreadeitherpneumaticallyormechanically acrossthefurnace,wheresmallpiecesofthefuelburnwhilein suspension. Simultaneously,largerpiecesoffuelarespreadin athin,evenbedonastationaryormovinggrate. Naturalgasor oilisoftenfiredinspreaderstokerboilersasauxiliaryfuel tomaintainaconstantsteamsupplywhenthewoodwastesupply fluctuates. Auxiliaryfuelcanalsoprovidemoresteamthancan begeneratedfromthewastesupplyalone.

Anotherboilertypesometimesusedforwoodcombustionis thesuspension-firingboiler. Thisboilerdiffersfroma spreaderstokerinthatsmall-sizedfuel(normallylessthan 2mm)isblownintotheboilerandcombustedbysuspensionfiring inairratherthanonfixedgrates. Rapidchangesincombustion rateand,therefore,steamgenerationratearepossiblebecause thefinelydividedfuelparticlesburnveryquickly.

Woodstovesarecommonlyusedinresidencesasspace heaters,bothastheprimarysourceofresidentialheatandto supplementconventionalheatingsystems. Thethreedifferent categoriesofwoodstovesare:

· Theconventionalwoodstove; · Thenoncatalyticwoodstove;and · Thecatalyticwoodstove.

Theconventionalstovecategorycomprisesallstoveswithout catalyticcombustorsnotincludedintheothernoncatalytic categories(i.e.,noncatalyticandpellet). Conventionalstoves donothaveanyemissionsreductiontechnologyordesignfeatures and,inmostcases,weremanufacturedbeforeJuly1,1986.

6-36 Stovesofmanydifferentairflowdesignsmaybeinthiscategory, suchasupdraft,downdraft,crossdraft,andS-flow.

Noncatalyticwoodstovesarethoseunitsthatdonotemploy catalystsbutdohaveemissionsreductiontechnologyorfeatures. Typicalnoncatalyticdesignincludesbafflesandsecondary combustionchambers.

Catalyticstovesareequippedwithaceramicormetal honeycombdevice(calledacombustororconverter)thatiscoated withanoblemetalsuchasplatinumorpalladium. Thecatalyst materialreducestheignitiontemperatureoftheunburned volatileorganiccompounds(VOC’s)andcarbonmonoxide(CO)in theexhaustgases,thusaugmentingtheirignitionandcombustion atnormalstoveoperatingtemperatures.

Fireplacesareusedprimarilyforaestheticeffectsand secondarilyasasupplementalheatingsourceinhousesandother dwellings. Woodisthemostcommonfuelforfireplaces,butcoal anddensifiedwood"logs"mayalsobeburned. Theuser intermittentlyaddsfueltothefirebyhand.

Allofthesystemsdescribedaboveoperateattemperatures thatareabovetheboilingpointofcadmium. Consequently,any cadmiumcontainedinthefuelwillbeemittedwiththecombustion gasesasenrichedfinePM. Thecombustionexhauststackisthe onlysourceofcadmiumemissionsfromtheseprocesses.

EmissionControlMeasures14

AlthoughsomewoodstovesusecontrolmeasurestoreduceVOC andCOemissions,thesetechniquesarenotexpectedtoaffect cadmiumemissions. However,woodwasteboilersdoemployPM

6-37 controlequipment,whichmayprovidesomereduction. These systemsaredescribedbrieflybelow.

Currently,thefourmostcommoncontroldevicesusedto reducePMemissionsfromwood-firedboilersaremechanical collectors,wetscrubbers,ESP’s,andfabricfilters. Ofthese controls,onlythelastthreehavethepotentialforsignificant cadmiumreduction.

Themostwidelyusedwetscrubbersforwood-firedboilers areventuriscrubbers. Withgas-sidepressuredropsexceeding 4kilopascals(15inchesofwater),PMcollectionefficienciesof 90percentorgreaterhavebeenreportedforventuriscrubbers operatingonwood-firedboilers. Nodatawerelocatedonthe performanceofthesesystemsrelativetocadmiumemissions,but itisexpectedtobesomewhatlessbecausecadmiumislikelyto beconcentratedinthefinePM,whichislessreadilycollected bycontroldevices.

Fabricfilters(i.e.,baghouses)andESP’sareemployedwhen PMcollectionefficienciesabove95percentarerequired. Collectionefficienciesof93to99.8percentforPMhavebeen observedforESP’soperatingonwood-firedboilers,butcadmium efficienciesarelikelytobesomewhatlessbecauseofthereason citedabove. Fabricfiltershavehadlimitedapplicationsto wood-firedboilersbecauseoffirehazards. Despite complications,fabricfiltersaregenerallypreferredforboilers firingsalt-ladenwood. ThisfuelproducesfinePMwithahigh saltcontentforwhichfabricfilterscanachievehighcollection efficiencies. Intwotestsoffabricfiltersoperatingon -ladenwood-firedboilers,PMcollectionefficiencieswere above98percent. Nodataareavailableoncadmiumemission reductionforfabricfilters,butbecausecadmiumisenriched

6-38 ontofinePM,whichislessreadilycollectedthanPMasawhole, itisexpectedthatefficiencieswillbesomewhatlower.

Emissions

Thedataoncadmiumemissionsfromwoodcombustionarequite limited. Arecentstudytoupdatethewoodwastecombustion sectionofAP-42providedarangeandaveragetypicalemission factorforwoodwastecombustioninboilersbasedontheresults ofseventests. Table6-16presentstherangeandaverage obtainedfromthosetestsandtherangeandaveragefroma California"HotSpots"testofafluidized-bedwood-firedboiler notincludedintheAP-42update.15,16

TABLE6-16.SUMMARYOFCADMIUMEMISSIONFACTORSFORWOODCOMBUSTION

Cadmiumemissionfactors

10-5 kg/Mgwoodburned 10-5 lb/tonwoodburned

Operation Range Average Range Average

Woodwasteboilera 0.13-27 0.85 0.27-54 1.7

Woodwasteboiler-"HotSpots"b 0.44-1.0 0.74 0.88-2.0 1.5

Residentialwoodstove- -- 1.1 -- 2.2 conventional -- 3.6 -- 7.2

Residentialwoodstove- -- 1.0 -- 2.0 noncatalytic

Residentialwoodstove-catalytic -- 2.3 -- 4.6

Source:References15through17. aBasedonanassumedheatingvalueof10,460kJ/kg(4,500Btu/lb)andPMcontrol. bBasedonaheatingvalueof19,220kJ/kg(8,270Btu/lb)andPMcontrolwithmulticlonesandESP. Areviewoftheliteratureproducedfouremissionfactors forresidentialwoodstovecombustion,whicharealsopresented inTable6.16.15,17 Threeofthefouremissionfactorswere

6-39 providedbythesectiononresidentialwoodcombustioninthe recentAP-42andincludedemissionfactorsforconventional, noncatalytic,andcatalyticwoodstovecombustion. However,the datausedtodeveloptheseemissionfactorsshowedahighdegree ofvariabilitywithinthesourcepopulation. Thefourthemission factorfromtheliteraturewasbasedononlyasingletestatone locationofanuncontrolledconventionalwoodstoveandmaynot berepresentativeofconditionsacrosstheUnitedStates. Therefore,theseemissionfactorsshouldbeusedcautiously.

MUNICIPALWASTECOMBUSTION

Refuseormunicipalsolidwaste(MSW)consistsprimarilyof householdgarbageandothernonhazardouscommercial, institutional,andindustrialsolidwaste. Municipalwaste combustors(MWC’s)areusedtoreducethemassandvolumeofMSW thatultimatelymustbelandfilled.

Currently,over160MWCplantsareinoperationinthe UnitedStateswithcapacitiesgreaterthan36megagramsperday (Mg/d)(40tonsperday[ton/d])andatotalcapacityof approximately100,000Mg/d(110,000ton/d)ofMSW. Itis predictedthatby1997,thetotalMWCcapacitywillapproach 150,000Mg/d(165,000ton/d),whichrepresentsover28percentof theestimatedtotalamountofMSWgeneratedintheUnitedStates bytheyear2000. Table6-17showsthegeographicdistribution ofMWCunitsandcapacitiesbyStates.18

Inadditiontotheselargeunits,anumberofsmaller, specializedfacilitiesaroundtheUnitedStatesalsoburnMSW. However,thetotalnationwidecapacityofthosesmallerunitsis onlyasmallfractionofthetotalcapacityoftheunitswith individualcapacitiesof36Mg/d(40ton/d)andlarger.

6-40 TABLE6-17.SUMMARYOFGEOGRAPHICALDISTRIBUTIONOFMWCFACILITIES

Percentage of total Number of MWC State MWC capacity MWC capacity in the State facilities Mg/d (ton/d) United States

AK 2 150(170) <1 AL 2 900(990) 1 AR 5 350(380) <1 CA 3 2,330(2,560) 2 CT 9 6,050(6,660) 6 DC 1 910(1,000) 1 DE 1 550(600) <1 FL 14 15,770(17,350) 16 GA 1 450(500) <1 HI 1 2,510(2,760) 2 IA 1 180(200) <1 ID 1 45(50) <1 IL 1 1,450(1,600) 1 IN 1 2,150(2,360) 2 MA 10 9,400(10,340) 9 MD 3 3,460(3,810) 3 ME 4 1,700(1,870) 2 MI 5 4,380(4,820) 4 MN 13 4,850(5,330) 5 MO 1 71(78) <1 MS 1 140(150) <1 MT 1 65(72) <1 NC 4 710(780) 1 NH 4 780(860) 1 NJ 6 5,290(5,820) 5 NY 15 11,370(12,510) 11 OH 4 4,360(4,800) 4 OK 2 1,120(1,230) 1 OR 3 740(810) 1 PA 6 6,550(7,200) 6 PR 1 950(1,040) 1 SC 2 760(840) 1 TN 4 1,350(1,480) 1 TX 4 220(240) <1 UT 1 360(400) <1 VA 9 6,220(6,840) 6 WA 5 1,360(1,500) 1 WI 9 1,240(1,360) 1 TOTAL 160 101,200 (111,400) 100

Source: Reference 18.

6-41 MunicipalSolidWasteCharacteristics19

Municipalsolidwasteisaheterogeneousmixtureofthe variousmaterialsfoundinhousehold,commercial,andindustrial wastes. Majorconstituentsintypicalmunicipalwastearelisted inTable6-18. NodataontheconcentrationofcadmiuminMSW streamswerelocated,butknownsourcesofcadmiuminMSWare batteries,discardedelectricalequipmentandwiring,and plastics.

ProcessDescription8,18,20,21

ThethreeprincipalMWCclassesaremassburn,refuse- derivedfuel(RDF),andmodularcombustors. Theparagraphsbelow brieflydescribesomeofthekeydesignandoperating characteristicsofthesedifferentcombustortypes. References8,18,and20providemoredetailedprocess descriptionsandprocessdiagramsforeachofthesystems describedbelow.

Inmassburnunits,theMSWiscombustedwithoutany preprocessingotherthanremovalofitemstoolargetogothrough thefeedsystem. Inatypicalmassburncombustor,refuseis placedonagratethatmovesthroughthecombustor. Combustion airinexcessofstoichiometricamountsissuppliedbothbelow (underfireair)andabove(overfireair)thegrate. Massburn combustorsareusuallyerectedatthesite(asopposedtobeing prefabricatedatanotherlocation)andrangeinsizefrom46to 900Mg/d(50to1,000tons/d)ofMSWthroughputperunit. The massburncombustorcategorycanbedividedintomassburn refractorywall(MB/REF),massburn/waterwall(MB/WW),andmass burn/rotarywaterwall(MB/RC)designs. Thetwomostcommon, MB/REFandMB/WW,aredescribedbelow.

6-42 TABLE6-18.CURRENTANDFORECASTCOMPOSITIONOFDISPOSEDRESIDENTIAL ANDCOMMERCIALWASTE(WEIGHTPERCENT)

Year

Component 1980 1990

PaperandPaperboard 33.6 38.3

YardWastes 18.2 17.0

FoodWastes 9.2 7.7

Glass 11.3 8.8

Metals 10.3 9.4

Plastics 6.0 8.3

Wood 3.9 3.7

Textiles 2.3 2.2

RubberandLeather 3.3 2.5

Miscellaneous 1.9 2.1 TOTAL 100.0 100.0

Source:Reference19.

6-43 TheMB/REFcombustorsareolderfacilitiesthatcomprise severaldesigns. Onedesigninvolvesabatch-fedupright combustor,whichmaybecylindricalorrectangularinshape. A seconddesignconsistsofrectangularcombustionchamberswith traveling,rocking,orreciprocatinggrates. Thistypeof combustoriscontinuouslyfedandoperatesinanexcessairmode withbothunderfireandoverfireairprovided. Thewasteis movedonatravelinggrateandisnotmixedasitadvances throughthecombustor. Asaresult,wasteburnoutorcomplete combustionisinhibitedbyfuelbedthickness,andthereis considerablepotentialforunburnedwastetobedischargedinto thebottomashpit. Rockingandreciprocatinggratesystemsmix andaeratethewastebedasitadvancesthroughthecombustion chamber,therebyimprovingcontactbetweenthewasteand combustionairandincreasingtheburnoutofcombustibles. The systemgenerallydischargestheashattheendofthegratestoa waterquenchpitforcollectionanddisposalinalandfill. The MB/REFcombustorstypicallyoperateatrelativelyhighexcessair ratestopreventexcessivetemperatures,whichcanresultin refractorydamage,slagging,fouling,andcorrosionproblems. Oneadverseeffectofhigherexcessairlevelsisthepotential forincreasedcarryoverofPMfromthecombustionchamber,and, ultimately,increasedstackemissionrates.

Becauseoftheiroperatingcharacteristics,thetraveling gratesystemsmayhavecoolashpocketsinwhichcadmiumisnot exposedtohightemperaturesandistherebyretainedintheash, ratherthanbeingexhaustedwiththecombustiongasstream. Consequently,cadmiummaybeemittedasfugitiveemissionsfrom ashhandling. However,thecombustionstackistheprimary sourceofcadmiumemissions. Intherockingandreciprocating gratesystems,essentiallyallcadmiumwillbeexhaustedwiththe combustiongas.

6-44 TheMB/WWdesignrepresentsthepredominanttechnologyin theexistingpopulationoflargeMWC’s,anditisexpectedthat over50percentofnewunitswillbeMB/WWdesigns. InMB/WW units,thecombustorwallsareconstructedofmetaltubesthat containpressurizedwaterandrecoverradiantenergyfromthe combustionchamber. Withthistypeofsystem,unprocessedwaste (afterremovaloflarge,bulkyitemsandnoncombustibles)is deliveredbyanoverheadcranetoafeedhopperthatconveysthe wasteintothecombustionchamber. NearlyallmodernMB/WW facilitiesutilizereciprocatinggratesorrollergratestomove thewastethroughthecombustionchamber. Thegratestypically includetwoorthreeseparatesectionswheredesignatedstagesin thecombustionprocessoccur. Ontheinitialgratesection, referredtoasthedryinggrate,themoisturecontentofthe wasteisreducedpriortoignition. Inthesecondgratesection, theburninggrate,themajorityofactiveburningtakesplace. Thethirdgratesection,referredtoastheburnoutorfinishing grate,iswhereremainingcombustiblesinthewasteareburned. Bottomashisdischargedfromthefinishinggrateintoawater- filledashquenchpitorramdischarger. Fromthere,themoist ashisdischargedtoaconveyorsystemandtransportedtoanash loadingareaorstorageareapriortodisposal. Becausethe wastebedisexposedtofairlyuniformhighcombustion temperatures,cadmiumwillvolatilizeandcondenseonsmall particles. MostcadmiumwillbeexhaustedasfinePMwiththe combustiongases,althoughsomemaybepartitionedwiththeash.

Refuse-derivedfuelcombustorsburnMSWthathasbeen processedtovaryingdegrees,fromsimpleremovalofbulkyand noncombustibleitemsaccompaniedbyshredding,toextensive processingtoproduceafinelydividedfuelsuitableforco- firinginpulverizedcoal-fireboilers. ProcessingMSWtoRDF generallyraisestheheatingvalueofthewastebecausemanyof thenoncombustibleitemshavebeenremoved.

6-45 AsetofstandardsforclassifyingRDFtypeshasbeen establishedbytheAmericanSocietyforTestingandMaterials ThetypeofRDFusedisdependentontheboilerdesign. Boilers thataredesignedtoburnRDFastheprimaryfuelusuallyutilize spreaderstokersandfirefluffRDFinasemi-suspensionmode. Thismodeoffeedingisaccomplishedbyusinganairswept distributor,whichallowsaportionofthefeedtoburnin suspensionandtheremaindertobeburnedoutafterfallingona horizontaltravelinggrate. ThenumberofRDFdistributorsina singleunitvariesdirectlywithunitcapacity. Thedistributors arenormallyadjustablesothatthetrajectoryofthewastefeed canbevaried. Becausethetravelinggratemovesfromtherear tothefrontofthefurnace,distributorsettingsareadjustedso thatmostofthewastelandsonthereartwo-thirdsofthegrate toallowmoretimeforcombustiontobecompletedonthegrate. Bottomashdropsintoawater-filledquenchchamber. Underfire airisnormallypreheatedandintroducedbeneaththegratebya singleplenum. Overfireairisinjectedthroughrowsofhigh pressurenozzles,providingazoneformixingandcompletionof thecombustionprocess. Becauseessentiallyallofthewasteis exposedtohighcombustiontemperaturesonthegrate,mostofthe cadmiumintheRDFwillbedischargedwiththecombustiongas exhaustasfinePM.

Inafluidized-bedcombustor(FBC),flufforpelletizedRDF iscombustedonaturbulentbedofnoncombustiblematerial,such aslimestone,sand,orsilica. Initssimplestform,theFBC consistsofacombustorvesselequippedwithagasdistribution plateandunderfireairwindboxatthebottom. Thecombustion bedoverliesthegasdistributionplate. TheRDFmaybeinjected intoorabovethebedthroughportsinthecombustorwall. The combustorbedissuspendedor"fluidized"throughthe introductionofunderfireairatahighflowrate. Overfireair isusedtocompletethecombustionprocess.

6-46 GoodmixingisinherentintheFBCdesign. Fluidized-bed combustorshaveuniformgastemperaturesandmasscompositionsin boththebedandintheupperregionofthecombustor. This uniformityallowstheFBC’stooperateatlowerexcessairand temperaturelevelsthanconventionalcombustionsystems. Waste- firedFBC’stypicallyoperateatexcessairlevelsbetween30and 100percentandatbedtemperaturesaround815°C(1500°F). At thistemperature,mostofthecadmiumwillbevolatilized. The cadmiumthencondensesontosmallparticlesandisexhaustedwith thecombustiongasstreamasfinePM.

Intermsofnumberoffacilities,modularstarved-(or controlled-)air(MOD/SA)combustorsrepresentalargesegmentof theexistingMWCpopulation. However,becauseoftheirsmall sizes,theyaccountforonlyasmallpercentageofthetotal capacity. ThebasicdesignofaMOD/SAcombustorconsistsoftwo separatecombustionchambers,referredtoasthe"primary"and "secondary"chambers. Wasteisbatch-fedintermittentlytothe primarychamberbyahydraulicallyactivatedram. Thecharging binisfilledbyafront-endloaderorbyothermechanical systems. Wasteisfedautomaticallyonasetfrequency,with generally6to10minutesbetweencharges.

Wasteismovedthroughtheprimarycombustionchamberby eitherhydraulictransferramsorreciprocatinggrates. Combustorsusingtransferramshaveindividualhearthsuponwhich combustiontakesplace. Gratesystemsgenerallyincludetwo separategratesections. Ineithercase,wasteretentiontimes intheprimarychamberarelengthy,lastingupto12hours. Bottomashisusuallydischargedtoawetquenchpit.

Thequantityofairintroducedintheprimarychamber definestherateatwhichwasteburns. Combustionairis introducedintheprimarychamberatsubstoichiometriclevels,

6-47 resultinginafluegasrichinunburnedhydrocarbons. The combustionairflowratetotheprimarychamberiscontrolledto maintainanexhaustgastemperaturesetpoint[generally650°to 980°C(1200°to1800°F)],whichcorrespondstoabout40to 60percenttheoreticalair. Asthehot,fuel-richfluegases flowtothesecondarychamber,theyaremixedwithexcessairto completetheburningprocess. Thetemperatureoftheexhaust gasesfromtheprimarychamberisabovetheautoignitionpoint. Thus,completingcombustionissimplyamatterofintroducingair tothefuel-richgases. Theamountofairaddedtothesecondary chamberiscontrolledtomaintainadesiredfluegasexit temperature,typically980°to1200°C(1800°to2200°F). At theseprimarychamberandsecondarychambertemperatures, essentiallyallofthecadmiumcontainedinthewasteisexpected tobevolatilized,condenseontosmallparticles,andbeemitted asfinePMfromthesecondarychamberwiththecombustiongas stream.21

EmissionControlMeasures18

CadmiumemissionsfromMWCunitsaregenerallycontrolledby condensingthecadmiumvaporsfromthecombustionchamberto particleformandthenremovingtheparticle-phasecadmiumwitha high-efficiencyPMcontroldevice. ThePMcontroldevicesmost frequentlyusedintheUnitedStatesareESP’sandfabric filters. Typically,newerMWCsystemsuseacombinationofgas coolingandductsorbentinjection(DSI)orspraydryer(SD) systemsupstreamofthePMdevicetoreducetemperaturesand provideamechanismforacidgascontrol. Theparagraphsbelow brieflydescribetheDSIandSDprocesses. BecausetheESP’sand FF’susedonMWC’sarecomparabletothoseusedonother combustionsystems,theyarenotdescribed. Reference18 providesmoredetaileddescriptionsofthecontrolsystemsand additionalinformationontheperformanceofthesesystems.

6-48 Spraydryingincombinationwitheitherfabricfiltrationor anESPisthemostfrequentlyusedacidgascontroltechnology forMWC’sintheUnitedStates. Spraydryer/fabricfilter systemsaremorecommonthanSD/ESPsystemsandareusedmoston new,largeMWC’s. Inthespraydryingprocess,limeisslurried andtheninjectedintotheSDthrougheitherarotaryatomizeror dual-fluidnozzles. Thekeydesignandoperatingparametersthat significantlyaffectSDperformancearetheSD’soutletapproach tosaturationtemperatureandlime-to-acidgasstoichiometric ratio. TheSDoutletapproachtosaturationtemperatureis controlledbytheamountofwaterinthelimeslurry. More effectiveacidgasandmetalsremovaloccursatlower temperatures,butthegastemperaturemustbekepthighenoughto ensuretheslurryandreactionproductsareadequatelydried priortocollectioninthePMcontroldevice.

WithDSI,powderedsorbentispneumaticallyinjectedinto eitheraseparatereactionvesselorasectionoffluegasduct locateddownstreamofthecombustoreconomizer. Alkaliinthe

sorbent(generallycalcium)reactswithHClandSO2 toform alkalisalts(e.g.,calciumchloride[CaCl2]andcalciumsulfite

[CaSO3]). Reactionproducts,flyash,andunreactedsorbentare collectedwitheitheranESPorfabricfilter.

BasedonasummaryofMWCcadmiumemissiondatain Reference18,substantialcadmiumremovalcanbeachievedusing spraydryingorductsorbentinjectionincombinationwithfabric filtrationoranESP. Acadmiumremovalefficiencyof98percent canbeachievedwithanSD/ESPsystem,slightlylowerthanthe cadmiumcontrolwithanSD/FFsystem(99percent)becauseofthe metalsenrichmentofthefineparticles. Iftheremoval efficiencyofPMwithanESPis98percentorgreater,the removalefficiencyofcadmiumwithanESPwillgenerallybeat least95percent. Removalefficienciesgreaterthan95percent

6-49 cangenerallybeachievedbyDSI/ESPsystems. ADSI/FFsystem canachieve99percentremovalofcadmium.

Emissions18

Arecentstudyconductedtoupdatethemunicipalwaste combustionsectionofAP-42providedacomprehensivereviewof theavailableMWCcadmiumemissiondata,whicharesummarizedin TableB-2ofAppendixB. Theemissiondatathatarepresentedin AppendixBareinconcentrationunitsratherthanemission factorsbecausethestudyfoundthatmostofthetestreports containedinsufficientprocessdatatogenerateemissionfactors.

Afterreviewingthetestdata,theauthorsconcludedthat thedevelopmentofemissionfactorsforMWC’s,usingonlythe testreportswhichestimatedfeedrates,wouldeliminatedata fromsomanyfacilities,especiallykeyfacilities,thatthe valuesderivedwerenotlikelytoberepresentativeoftheentire MWCpopulation. Inaddition,thesubjectivenatureoftherefuse feedratescalledintoquestionthevalidityofthelimiteddata. Consequently,emissionfactorsweredevelopedusingtheF-factor, whichistheratioofthegasvolumeoftheproductsof combustiontotheheatingvalueofthefueldevelopedbytheEPA (EPAMethod19). ThisapproachrequiresanF-factorandan estimateofthefuelheatingvalue. ForMWC’s,anF-factorof 0.257dscm/MJ(9,570dscf/106 Btu)(atzeropercentoxygen)is assigned. Forallcombustortypes,exceptRDFcombustors,an assumedheatingvalueof10,500kJ/kg(4,500Btu/lb)refusewas alsoused. ForRDFcombustorunits,theprocessedrefusehasa higherheatingvalue,andtheassumedheatingvaluewas 12,800kJ/kg(5,500Btu/lb). Overall,thesedataare representativeofaveragevaluesforMWC’s.

6-50 Theresultantbesttypicalemissionfactorsfordifferent combinationsofcombustorandcontroldevicearepresentedin Table6-19. Whilethisproceduredoesprovidegoodaverage emissionfactorsthatrepresentanindustrycrosssection,it shouldnotbeusedtoconvertindividualdatapointsin AppendixB. TheassumedF-factorandwasteheatingvaluesabove maynotbeappropriateforspecificfacilities.

SEWAGESLUDGEINCINERATORS

Currently,about200sewagesludgeincinerators(SSI’s) operateintheUnitedStatesusingoneofthreetechnologies: multiplehearth,fluidized-bed,andelectricinfrared. Multiple hearthunitspredominate,withover80percentoftheidentified, operatingSSI’sbeingofthattype. About15percentofthe SSI’sarefluidized-bedcombustors;3percentareelectric ;andtheremaindercofiresewagesludgewithmunicipal solidwaste.22

Figure6-1showsthedistributionofsewagesludge incineratorsintheUnitedStates23 Mostfacilitiesarelocated intheEasternUnitedStates,butasubstantialnumberarealso locatedontheWestCoast. NewYorkhasthelargestnumberof SSIfacilitieswith33,followedbyPennsylvaniaandMichigan with21and19,respectively. About1.5x106 Mg (1.6x106 tons)ofsewagesludgeonadrybasisareestimatedto beincineratedannually.22

Nodatahavebeenlocatedonthecadmiumcontentofsewage sludge.

ThesectionsbelowprovideSSIprocessdescriptions,a discussionofcontrolmeasures,andasummaryofcadmiumemission factors.

6-51 TABLE6-19.BESTTYPICALCADMIUMEMISSIONFACTORSFORMUNICIPAL WASTECOMBUSTORS

Cadmiumemissionfactors Combustor Control -3 type statusa g/Mgwaste 10 lb/tonwaste

MassBurn/Waterwall UN 4.8 9.7

SD/FF 0.015 0.029

SD/ESP 0.056 0.11

ESP 0.55 1.1

DSI/FF 0.016 0.032

MassBurn/Rotary DSI/FF 0.012 0.024 Waterwall

MassBurn/RefractoryWall UN 5.7 1.1

ESP 0.10 0.20

DSI/ESP 0.044 0.089

Refuse-DerivedFuel-Fired UN 4.4 8.7

SD/FF 0.0 0.0

SD/ESP 0.042 0.084

ESP 0.083 0.17

Modular/ExcessAir DSI/FF 0.0081 0.016

Modular/StarvedAir UN 1.2 2.4

ESP 0.23 0.46

Source:Reference18. aUN=uncontrolled,SD=spraydryer,FF=fabricfilter,ESP=electrostaticprecipitator, DSI=ductsorbentinjection.

6-52

ProcessDescription8,22

Figure6-2presentsasimplifieddiagramofthesewage sludgeincinerationprocess,whichinvolvestwoprimarysteps. Thefirststepintheprocessofsewagesludgeincinerationis thedewateringofthesludge. Sludgeisgenerallydewatered untilitisabout15to30percentsolids. Whenitismorethan 25percentsolids,thesludgewillusuallyburnwithoutauxiliary fuel. Afterdewatering,thesludgeissenttotheincinerator, andthermaloxidationoccurs. Theunburnedresidualashis removedfromtheincinerator,usuallyonacontinuousbasis,and isdisposed. Aportionofthenoncombustiblewaste,aswellas unburnedvolatileorganiccompounds,iscarriedoutofthe combustorthroughentrainmentintheexhaustgasstream. Air pollutioncontroldevices,primarilywetscrubbers,areusedto removetheentrainedpollutantsfromtheexhaustgasstream. The gasstreamisthenexhausted,andthecollectedpollutantsare sentbacktotheheadofthewastewatertreatmentplantinthe scrubbereffluent. AsshowninFigure6-2,theprimarysourceof cadmiumemissionsfromtheSSIprocessisthecombustionstack. Somefugitiveemissionsmaybegeneratedfromashhandling,but thequantitiesareexpectedtobesmall. Becausecadmiumis relativelyvolatile,mostcadmiumwillleavethecombustion chamberasfinePMintheexhaustgas,althoughsomecadmiummay befoundintheashresidue.

TheparagraphsbelowbrieflydescribethethreeprimarySSI processesusedintheUnitedStates. References8and22provide moredetaileddescriptionsandprocessdiagrams.

Thebasicmultiplehearthfurnaceiscylindricalinshape andisorientedvertically. Theoutershellisconstructedof steel,linedwithrefractory,andsurroundsaseriesof horizontalrefractoryhearths. Ahollowcastironrotatingshaft

6-54 runsthroughthecenterofthehearths. Coolingairforthe centershaftandrabblearmsisintroducedintotheshaftbya fanlocatedatitsbase. Attachedtothecentralshaftarethe rabblearmswithteethshapedtorakethesludgeinaspiral motion,alternatingindirectionfromtheoutsidein,theninside out,betweenhearths. Typically,theupperandlowerhearthsare fittedwithfourrabblearms,andthemiddlehearthsarefitted withtwo. Burnersthatprovideauxiliaryheatarelocatedinthe sidewallsofthehearths.

Partiallydewateredsludgeistypicallyfedontothe perimeterofthetophearth. Typically,therabblearmsmovethe sludgethroughtheincineratorasthemotionoftherabblearms rakesthesludgetowardthecentershaft,whereitdropsthrough holeslocatedatthecenterofthehearth. Thisprocessis repeatedinallofthesubsequenthearths,withthesludgemoving inoppositedirectionsinadjacenthearths. Theeffectofthe rabblemotionistobreakupsolidmaterialtoallowbetter surfacecontactwithheatandoxygen.

Ambientairisfirstductedthroughthecentralshaftand itsassociatedrabblearms. Thisairisthentakenfromthetop oftheshaftandrecirculatedontothelowermosthearthas preheatedcombustionair. Thecombustionairflowsupward throughthedropholesinthehearths,countercurrenttotheflow ofthesludge,beforebeingexhaustedfromthetophearth.

Multiplehearthfurnacescanbedividedintothreezones. Theupperhearthscomprisethedryingzonewheremostofthe moistureinthesludgeisevaporated. Thetemperatureinthe dryingzoneistypicallybetween425°and760°C(800°and 1400°F). Sludgecombustionoccursinthemiddlehearths(second zone)asthetemperatureisincreasedbetween815°and925°C (1500°and1700°F). Whenexposedtothetemperaturesinboth

6-56 upperzones,mostcadmiumwillbevolatilized,condenseonsmall particles,andthenbedischargedasfinePMintheexhaustgas. Someofthecadmiummaybepartitionedwiththeash. Thethird zone,madeupofthelowermosthearth(s),isthecoolingzone. Inthiszone,theashiscooledasitsheatistransferredtothe incomingcombustionair.

Fluidized-bedcombustors(FBC’s)arecylindricallyshaped andorientedvertically. Theoutershellisconstructedofsteel andislinedwithrefractory. (nozzlesdesignedto deliverblastsofair)arelocatedatthebaseofthefurnace withinarefractory-linedgrid. Abedofsandrestsuponthe grid. Partiallydewateredsludgeisfedintothebedofthe furnace. Airinjectedthroughthetuyeres,atpressuresfrom20 to35kPa(3to5psig),simultaneouslyfluidizesthebedofhot sandandtheincomingsludge. Temperaturesof725°to825°C (1350°to1500°F),whicharesufficienttovaporizemostcadmium containedinthesludge,aremaintainedinthebed. Asthe sludgeburns,fineashparticles,includingcadmium,arecarried outthetopofthefurnacewiththeexhaustgas.

Anelectricincineratorconsistsofahorizontallyoriented, insulatedfurnace. Awovenwirebeltconveyorextendsthelength ofthefurnace,andinfraredheatingelementsarelocatedinthe roofabovetheconveyorbelt. Combustionairispreheatedbythe fluegasesandisinjectedintothedischargeendofthefurnace. Electricincineratorsconsistofanumberofprefabricated moduleswhichcanbelinkedtogethertoprovidethenecessary furnacelength. Thedewateredsludgecakeisconveyedintoone endoftheincinerator. Aninternalrollermechanismlevelsthe sludgeintoacontinuouslayerapproximately2.5centimeters(cm) (1inch[in.])thickacrossthewidthofthebelt. Thesludgeis sequentiallydriedandthenburnedasitmovesbeneaththe infraredheatingelements. Ashisdischargedintoahopperat

6-57 theoppositeendofthefurnace. Thepreheatedcombustionair entersthefurnaceabovetheashhopperandisfurtherheatedby theoutgoingash. Thedirectionofairflowiscountercurrentto themovementofthesludgealongtheconveyor.

EmissionControlMeasures22,24

MostSSI’sareequippedwithsometypeofwetscrubbing systemforPMcontrol. Becausethesesystemsprovidegascooling aswellasPMremoval,theycanprovidesomecadmiumcontrol. Theparagraphsbelowbrieflydescribethewetscrubbingsystems typicallyusedonexistingSSI’s.

WetscrubbercontrolsonSSI’srangefromlowpressuredrop spraytowersandwetcyclonestohigherpressuredropventuri scrubbersandventuri/impingementtrayscrubbercombinations. Themostwidelyusedcontroldeviceappliedtoamultiplehearth incineratoristheimpingementtrayscrubber. Olderunitsuse thetrayscrubberalonewhilecombinationventuri/impingement trayscrubbersarewidelyappliedtonewermultiplehearth incineratorsandtofluidized-bedincinerators. Mostelectric incineratorsandsomefluidized-bedincineratorsuseventuri scrubbersonly.

Inatypicalcombinationventuri/impingementtrayscrubber, hotgasexitstheincineratorandenterstheprecoolingorquench sectionofthescrubber. Spraynozzlesinthequenchsection cooltheincominggas,andthequenchedgasthenentersthe venturisectionofthecontroldevice. Venturiwaterisusually pumpedintoaninletweirabovethequencher. Theventuriwater entersthescrubberabovethethroatandfloodsthethroat completely. Mostventurisectionscomeequippedwithvariable throatstoallowthepressuredroptobeincreased,thereby increasingPMefficiency. Atthebaseofthefloodedelbow,the

6-58 gasstreampassesthroughaconnectingducttothebaseofthe impingementtraytower. Gasvelocityisfurtherreducedupon entrytothetowerasthegasstreampassesupwardthroughthe perforatedimpingementtrays. Waterusuallyentersthetrays frominletportsonoppositesidesandflowsacrossthetray. As gaspassesthrougheachperforationinthetray,itcreatesajet thatbubblesupthewaterandfurtherentrainssolidparticles. Atthetopofthetowerisamisteliminatortoreducethe carryoverofwaterdropletsinthestackeffluentgas.

Accordingtotheliterature,thecontrolofcadmium emissionswithwetscrubbercontrolsisexpectedtobelessthan thecontroloftotalPMemissions. Inastudyofemissionsfrom fourmunicipalwastewatersludgeincinerator,threemultiple hearthandonefluidized-bed,cadmiumandotherheavymetalswere foundtobeenrichedinthefineparticles,whicharenotas efficientlyremovedbyscrubbersaslargerparticles.24 The efficiencydataforcadmiumemissionscontrolforsewagesludge incinerationareverylimitedand,therefore,arenotcompletely reliable. BasedontwotestsreportedinReference22,an averagecadmiumcontrolefficiencyof75percentcanbeachieved withacombinationofventuriscrubberandimpingementscrubber systems. Bycontrast,basedontheresultsofthreeothertests, anaverageof95percentcontrolofPMcanbeachievedwithan impingementscrubberalone. About86percentPMcontrolwas achievedwithaventuriscrubberinanothertest.

Emissions

AsapartoftherecentupdateofAP-42,datahavebeen developedoncadmiumemissionsfromSSI’s.22 Thesedataare tabulatedinAppendixB,TableB-3andsummarizedinTable6-20. Becausenodataareavailableoncadmiumconcentrationsin

6-59 TABLE6-20.SUMMARYOFCADMIUMEMISSIONFACTORS FORSEWAGESLUDGEINCINERATION

Cadmiumemissionfactors

g/Mgdrysolids 10-3 lb/tondrysolids Incinerator Control No.data typea statusb points Range Average Range Average

MH UN 4 0.0010-49 26 0.0020-98 53

MH VS 2 0.17-0.65 0.41 0.34-1.3 0.82

MH IS 2 1.2-1.5 1.4 2.4-3.0 2.7

MH VS/ISor 5 0.32-7.8 3.0 0.64-16 5.9 VS/IS/AB

MH CY 3 0.86-32 12 1.7-65 25

MH CY/VSor 2 8.1-25 17 16-50 33 CY/VS/IS

MH ESP 1 --- 0.17 --- 0.35

MH FF 1 --- 0.014 --- 0.028

FB ISorVS/IS 5 0.0030-1.4 0.48 0.0060-2.9 0.97

Source:Reference22. aMH=multiplehearth,FB=fluidized-bed. bUN=uncontrolled,VS=venturiscrubber,IS=impingementscrubber,AB=afterburner, CY=cyclone,ESP=electrostaticprecipitator,FF=fabricfilter.

6-60 sludge,thetestdatainTable6-20representthebesttypical emissionfactorsforsewagesludgeincineration.

MEDICALWASTEINCINERATION

Medicalwasteincludesinfectiousandnoninfectiouswastes generatedbyavarietyoffacilitiesengagedinmedicalcare, veterinarycare,orresearchactivitiessuchashospitals, clinics,doctors’anddentists’offices,nursinghomes, veterinaryclinicsandhospitals,medicallaboratories,and medicalandveterinaryschoolsandtheirresearchunits. Medical wasteisdefinedbytheU.S.EPAas"anysolidwastewhichis generatedinthediagnosis,treatment,orimmunizationofhuman beingsoranimals,inresearchpertainingthereto,orinthe productionortestingofbiologicals." Amedicalwaste incinerator(MWI)isanydevicethatburnssuchmedicalwaste.25

RecentestimatesdevelopedbyEPAsuggestthatabout 3.06millionMg(3.36milliontons)ofmedicalwasteareproduced annuallyintheUnitedStates. Approximately5,000MWI’s,which aredistributedgeographicallythroughouttheUnitedStates,are usedtotreatthiswaste. Ofthese5,000units,about3,000are locatedathospitals;about150arelargercommercialfacilities; andtheremainderaredistributedamongveterinaryfacilities, nursinghomes,laboratories,andothermiscellaneous facilities.26

AvailableinformationindicatesthattheseMWIsystemscan besignificantsourcesofcadmiumemissions. Cadmiumemissions resultfromcadmium-bearingmaterialscontainedinthewaste. Althoughconcentrationsofspecificmetalsinthewastehavenot beenfullycharacterized,knowncadmiumsourcesinmedicalwaste includebatteries,pigments,andplastics. Batteries,primarily nickel-cadmiumandmercury-cadmiumbatteries,areamajorcadmium

6-61 source. -cadmiumbatteriesareusedintransistorized equipment,hearingaids,watches,calculators,computers,smoke detectors,taperecorders,regulatedpowersupplies,radiation detectionmeters,scientificequipment,pagers,oxygenandmetal monitors,andportableelectrocardiogrammonitors. Thenickel- cadmiumbatteryisthemostwidelyusedrechargeablehousehold batteryandisusedincomputers,hearingaids,andpocket calculators. Cadmiumpigmentsareprimarilyusedinplasticsbut arealsousedinpaints,enamels,printinginks,rubber,paper, andpaintedtextiles.27 Plasticsareusedindisposable instruments,syringes,petridishes,plasticcontainers, packaging,bedpans,urinebags,respiratorydevices,dialysis equipment,etc.28 Allofthesematerialscanberoutedtoan MWI,therebycontributingtocadmiumemissionsfromthissource category.

ProcessDescription

Althoughtheultimatedestinationofalmostallmedical wasteproducedintheUnitedStatesisasolidwastelandfill, thewastegenerallymustbetreatedbeforeitcanbelandfilled. TheprimaryfunctionsofMWIfacilitiesaretorenderthewaste biologicallyinnocuousandtoreducethevolumeandmassof solidsthatmustbelandfilledbycombustingtheorganicmaterial containedinthewaste. Overtheyears,awidevarietyofMWI systemdesignsandoperatingpracticeshavebeenusedto accomplishthesefunctions. Toaccountforthesesystem differences,anumberofMWIclassificationschemeshavebeen usedinpaststudies,includingclassificationbywastetype (pathological,mixedmedicalwaste,redbagwaste,etc.), classificationbyoperatingmode(continuous,intermittent, batch),andclassificationbycombustordesign(retort, fixed-hearth,pulsed-hearth,rotarykiln,etc.). Someinsight intoMWIprocesses,emissions,andemissionscontrolisprovided

6-62 byeachoftheseschemes. However,becausetheavailable evidencesuggeststhatcadmiumemissionsareaffectedprimarily bywastecharacteristics,thecharacterizationandcontrolof cadmiumemissionsfromMWI’scanbediscussedwithoutconsidering otherMWIdesignandoperatingpracticesindetail. The paragraphsbelowprovideagenericMWIprocessdescriptionand identifypotentialsourcesofcadmiumemissions. Moredetailed descriptionsofspecificMWIdesignandoperatingpracticescan befoundinReferences29through31.

AschematicofagenericMWIsystemthatidentifiesthe majorcomponentsofthesystemisshowninFigure6-3. As indicatedintheschematic,mostMWI’saremultiple-chamber combustionsystemsthatcompriseprimary,secondary,andpossibly tertiarychambers. TheprimarycomponentsoftheMWIprocessare thewaste-chargingsystem,theprimarychamber,theashhandling system,thesecondarychamber,andthecombustiongassystem, whicharediscussedbrieflybelow.

Medicalwasteisintroducedtotheprimarychamberviathe waste-chargingsystem. Thewastecanbechargedeithermanually ormechanically. Withmanualcharging,whichisusedonlyon batchandsmaller(generallyolder)intermittentunits,the operatoropensachargedooronthesideoftheprimarychamber andtossesbagsorboxesofwasteintotheunit. Whenmechanical feedsystemsareemployed,sometypeofmechanicaldeviceisused tochargethewastetotheincinerator. Themostcommon mechanicalfeedsystemisthehopper/ramassembly. Ina mechanicalhopper/ramfeedsystem,thefollowingstepstake place: (1)wasteisplacedintoacharginghoppermanually,and thehoppercoverisclosed;(2)afiredoorisolatingthehopper fromtheincineratoropens;(3)therammovesforwardtopushthe wasteintotheincinerator;(4)theramreversestoalocation behindthefiredoor;(5)afterthefiredoorcloses,awater

6-63 spraycoolstheram,andtheramretractstothestarting position;and(6)thesystemisreadytoacceptanothercharge. Theentirehopper/ramchargingsequencenormallyfunctionsasa controlled,automatically-timedsequencetoeliminate overcharging. Thesequencecanbeactivatedbytheoperatoror forlarger,fullyautomatedincinerators,itmaybeactivatedat presetintervalsbyanautomatictimer.30,31

Thepotentialforcadmiumemissionsfromthewaste-charging systemsislow. Mechanicalsystemsaregenerallyoperatedwitha double-doorsystemtominimizefugitiveemissions. Small quantitiesoffugitiveemissionsmaybegeneratedwhilethe chamberdoorisopenduringmanualcharging,butnodataare availableonthemagnitudeoftheseemissions.

Theprimarychamber(sometimescalledthe"ignition" chamber)acceptsthewasteandbeginsthecombustionprocess. MostmodernMWI’soperatethischamberina"controlled-air"mode tomaintaincombustionairlevelsatorbelowstoichiometric requirements. Theobjectivesofthiscontrolled-airoperation aretoprovideamoreuniformreleaseofvolatileorganic materialstothesecondarychamberandtominimizeentrainmentof solidsintheseoff-gases. Threeprocessesoccurintheprimary chamber. First,themoistureinthewasteisvolatilized. Second,thevolatilefractionofthewasteisvaporized,andthe volatilegasesaredirectedtothesecondarychamber. Third,the fixedcarbonremaininginthewasteiscombusted.

Theprimarychambergeneratestwoexhauststreams--the combustiongasesthatpasstothesecondarychamberandthesolid ashstreamthatisdischarged. Anymetalcompoundsinthewaste, includingcadmium,arepartitionedtothesetwostreamsinoneof threeways. Themetalsmayberetainedintheprimarychamber bottomashanddischargedassolidwaste;theymaybeentrained

6-65 asPMinthecombustiongases;ortheymaybevolatilizedand dischargedasavaporwiththecombustiongases. Becausethe primarychambertypicallyoperatesintherangeof650°to820°C (1200°to1500°F),mostofthecadmiuminthewastestreamwill bevolatilizedanddischargedtothesecondarychamber. Atthe lowerexhausttemperatures,thecadmiumcondensesontosmall particlesandisexhaustedasfinePMtothesecondarychamber. Asmallfractionmayberetainedintheprimarychamberbottom ash.

Theprimarychamberbottomashisdischargedviaanash removalsystemandtransportedtoalandfillfordisposal. The ashremovalsystemmaybeeithermanualormechanical. Typically,batchunitsandsmallerintermittentunitsemploy manualashremoval. Afterthesystemhasshutdownandtheash hascooled,theoperatorusesarakeorshoveltoremovetheash andplaceitinadrumordumpster. Someintermittent-dutyMWI’s andallcontinuouslyoperatedMWI’suseamechanicalashremoval system. Themechanicalsystemincludesthreemajorcomponents: (1)ameansofmovingtheashtotheendoftheincinerator hearth--usuallyanashtransferramorseriesoftransferrams, (2)acollectiondeviceorcontainerfortheashasitis dischargedfromthehearth,and(3)atransfersystemtomovethe ashfromthecollectionpoint. Generally,theseautomatic systemsaredesignedtominimizefugitiveemissions. For example,onetypeofcollectionsystemusesanashbinsealed directlytothedischargechuteorpositionedwithinanair- sealedchamberbelowthehearth. Adoororgatethatsealsthe chuteisopenedatregularintervalstoallowtheashtodrop intothecollectionbin. Whenthebinisfilled,theseal-gate isclosed,andthebinisremovedandreplacedwithanemptybin. Inanothersystem,theashisdischargedintoawaterpit. The ashdischargechuteisextendedintothewaterpitsothatanair sealismaintained. Thewaterbathquenchestheashastheash

6-66 iscollected. Amechanicaldevice,eitherarakeordrag conveyorsystem,isusedtointermittentlyorcontinuouslyremove theashfromthequenchpit. Theexcesswaterisallowedto drainfromtheashasitisremovedfromthepit,andthewetted ashisdischargedintoacollectioncontainer.

Thepotentialforcadmiumemissionsfrombothmechanicaland manualashdischargesystemsisminimal. Asdescribedabove, mostmechanicalsystemshavesealsandprovideashwettingas describedabovetominimizefugitivePMemissions. Whilemanual systemscangeneratesubstantialfugitivePM,theconcentrations ofcadmiumhavegenerallybeenshowntobequitelow.32 Consequently,fugitivecadmiumemissionsarenegligible.

Almostallthecadmiumthatenterstheprimarychamberis exhaustedtothesecondarychamberasfinePM,althoughasmall fractionmaybepartitionedwiththeash. Theprimaryfunction ofthesecondarychamberistocompletethecombustionofthe volatileorganiccompoundsthatwasinitiatedintheprimary chamber. Becausethetemperaturesinthesecondarychamberare typically980°C(1800°F)orgreater,essentiallyallofthe cadmiumthatentersthesecondarychamberwillbeexhaustedas finePM. Thehotexhaustgasesfromthesecondarychambermay passthroughanenergyrecoverydevice(wasteheatboilerorair- to-airheatexchanger)andanairpollutioncontrolsystembefore theyaredischargedtotheatmospherethroughthecombustion stack. Thiscombustionstackisthemajorrouteofcadmium emissionsfromMWI’s.

EmissionControlMeasures32

Anumberofairpollutioncontrolsystemconfigurationshave beenusedtocontrolPMandgaseousemissionsfromtheMWI combustionstacks. Mostoftheseconfigurationsfallwithinthe

6-67 generalclassesofwetsystemsanddrysystems. Wetsystems typicallycompriseawetscrubberdesignedforPMcontrol (venturiscrubberorrotaryatomizingscrubber)inserieswitha packed-bedscrubberforacidgasremovalandahigh-efficiency misteliminationsystem. Mostdrysystemsuseafabricfilter forPMremoval,butESP’shavebeeninstalledonsomelarger MWI’s. Thesedrysystemsmayusesorbentinjectionviaeither dryinjectionorspraydryersupstreamfromthePMdeviceto enhanceacidgascontrol. MoredetaileddescriptionsofMWIair pollutioncontrolsystemscanbefoundinReference32. The emissiondatapresentedinthesectionbelowprovideinformation ontheperformanceofsomeofthemorecommonsystems.

Emissions33-50

Overthepast5years,cadmiumemissionshavebeenmeasured atseveralMWI’sthroughouttheU.S.EPA’sregulatory developmentprogram,theMWIemissioncharacterizationstudies conductedbytheStateofCalifornia,andcompliancetests conductedinresponsetoStateairtoxicrequirements. Emission datafrom25MWI’swereidentifiedindevelopingthisL&E document. However,thedatafromonly18facilitieswere consideredadequateforemissionfactordevelopment. Forthe remainingfacilities,eitherprocessdatawereinsufficientto developemissionfactorsorthetestmethodologieswere consideredunacceptable. Emissiondataforthe18facilitiesare tabulatedinAppendixB,TableB-4. Theparagraphsbelow summarizetheinformationonuncontrolledemissionsandonthe performanceofemissioncontrolsystemscollectedfromthese 18facilities.

Theuncontrolledemissiondatacollectedat13facilities showedsubstantialvariability,withcadmiumemissionfactors rangingfrom0.12to22g/Mgofwastecharged(2.4x10-4 to

6-68 4.4x10-2 lb/ton).33-41,44,47-50 Thisrangeofemissionfactors representsavarietyofwastetypes(mixedmedicalwaste,redbag [infectious]wasteonly,andpathologicalwaste)andavarietyof incineratortypes(continuousandintermittentunitswithvaried operatingpractices). Whilethedataareinsufficientto demonstrateunequivocallyadirectrelationshipbetweenwaste characteristicsandemissions,thedatastronglysuggestthat mostofthisvariabilityisrelatedtodifferencesinthecadmium contentofthewaste. First,characterizationofthebottomash atseveralfacilitiesshowedlittlecadmiumintheash, indicatingthatmostofthecadmiuminthewasteisdischarged withthecombustiongases. Second,aspartofanEPAstudy, wastesfromtwodifferenthospitalswerefiredtothesame incineratorundercomparableoperatingconditions. Therewasan almostthreefolddifferenceintheaverageemissionfactorsfor thetwowastes,withwastesfromthesmallerhospitalyieldingan emissionfactorof1.6g/Mg(3.1x10-3 lb/ton)andthosefrom thelargerhospitalyieldingafactorof4.4g/Mg (8.8x10-3 lb/ton),againprovidingevidenceofwaste-related variation. Althoughtherehasbeensomespeculationthatthe higheremissionfactorsresultfromhavingcadmium-bearingitems, suchasbatteries,pigments,andplasticsinthewastestream, insufficientinformationisavailabletodefineconclusively thosewasteattributesthataffectcadmiumemissions.

Becauseemissionsarestronglyrelatedtowaste characteristics,separateuncontrolledemissionfactorswere developedforthedifferentwastetypes. Theseemissionfactors aresummarizedinTable6-21.Substantiallygreaterinformation isavailableformixedmedicalwasteincinerationthanforeither redbagorpathologicalwasteincineration. Consequently,the mixedwasteresultsareconsideredtobeamorereliable indicatoroftherangeofemissionfactorslikelytobefound acrosstheMWIpopulationthanaretheredbagorpathological

6-69 TABLE6-21.UNCONTROLLEDCADMIUMEMISSIONFACTORS FORMEDICALWASTEINCINERATORS

Cadmiumemissionfactors, g/Mg(10-3 lb/ton) No.of No.of Wastetype facilities testruns Range Average

Mixeda 13 72 0.12-22 2.5 (0.24-44) (5.0)

Redbagb 1 9 0.72-2.5 1.6 (1.4-5.0) (3.3)

Pathologicalc 1 6 <0.0-4.7 0.90 (<0.0-9.3) (1.8)

Source:References33-41,44,47-50. aBasedontherangeoffacilityaverages.Numberofrunsforeachfacilityrangedfrom 2to16. bBasedontherangespannedbythreetestaverages(eachtestcomprisedthreeruns)atone facility. cThisemissionfactorisstronglyinfluencedbyasinglelargevalue.Abetterestimateofemissions froma"typical"facilityisthetrimmedmean,whichis0.18g/Mg(0.37x10-3 lb/ton).

6-70 results. However,becausetherangeinemissionfactorsisso large,eventhemixedwasteemissionfactorsshouldbeappliedto individualMWI’swithcaution.

Theemissionfactorsfortheredbagandpathologicalwaste shouldbeusedwithextremecautionbecauseeachfactorisbased onresultsfromwastefiredatonlyonefacility. Two observationsarenoteworthyininterpretingthesedata. First, theredbagemissionfactorof1.6g/Mg(3.3x10-3 lb/ton)is withintherangeoftheemissionfactorsformixedmedicalwaste. However,thewastesweregeneratedbythesamefacilitythathad oneofthelargestmixedwasteemissionfactors,sotheredbag emissionfactormaybemisleading. Second,theemissionfactor forpathologicalwasteof0.90g/Mg(1.8x10-3 lb/ton)isnear thebottomendofthemixedwasterangeandcouldbeevenlower becauseitisstronglyinfluencedbyasinglelargevalue (4.7g/Mg[9.3x10-3 lb/ton]). Thisvalueisafactorof 20largerthanthesecondlargestvalue. Ifthelargestand smallestvaluesareremoved,thetrimmedmeanis0.18g/Mg (0.37x10-3 lb/ton),whichissimilartothemedianofthedata. Hence,theemissionfactorof0.18g/Mg(0.37x10-3 lb/ton)is recommendedasthebestemissionfactorforatypicalMWIfiring pathologicalwaste. However,thislowemissionfactoralsomay bemisleadingbecausetestsatthesamefacilityproducedthe lowestmixedwasteemissionfactor. Asevidencedbythese observations,theredbagandpathologicalemissiondataaretoo sparsetodifferentiateeffectivelybetweentheeffectsofwaste typeandfacility-specificwastepracticesoncadmiumemissions.

Substantiallyfewerdataareavailableoncontrolled emissionsthanonuncontrolledemissions.35,36,40-46,48,50 Thebest dataavailablearethosewhichcharacterizetheperformanceof sevenMWIairpollutioncontrolsystems--awetscrubbersystem,a venturiscrubbersystem,aventuriscrubber/packed-bedsystem,a

6-71 ductsorbentinjection/electrostaticprecipitatorsystem,a fabricfiltersystem,adryinjection/fabricfiltersystem,anda spraydryer/fabricfiltersystem. Table6-22presentscontrolled emissionfactorsandcadmiumemissioncontrolefficienciesfor theseairpollutioncontrolsystems. Becausecontrolledemission factorscouldonlybedevelopedforafewfacilities,theyare notlikelytorepresentthevariabilityacrosstheincinerator population. Therefore,itisrecommendedthatcontrolled emissionfactorsbedevelopedbyapplyingtheaveragecontrol efficienciestouncontrolledemissionfactorsoremissionrates ratherthanusingthecontrolledemissionfactorspresentedin Table6-22.

Theperformancesoftwoofthedrysystems(dry injection/fabricfilterandspraydryer/fabricfilter)were examinedwithandwithoutcarboninjection. Theresultsfrom thesetestsandfromthetestofthefabricfilterwithnocarbon injectionarepresentedinTableB-4,AppendixB. Theseresults indicatethatthedrysystemswithoutcarboninjectionprovided greaterthan99percentcontrolofcadmium. Forthesesystems, theoutletcadmiumemissionsrangefrom99.1percentto 99.9percentlowerthantheinletemissions. Thisvariabilityis consideredtobewellwithinthenormalrangeofprocessand emissiontestmethodvariabilityasdescribedinSection8. Consequently,theresultsareconsistentwithessentially completeremovalbythecontrolsystem. Thedrysystemswith carboninjectionachieveessentiallythesamecadmiumremovalas thosesystemswithoutcarboninjection,withcontrolefficiencies rangingfrom97.3percentto99.9percent.

Theemissiontestresultsforthewetsystemsarealso presentedinTableB-4,AppendixB. AsshowninTable6-22,the performanceofthewetsystemsincontrollingcadmiumemissions wasnotaseffectiveasthatachievedbythedrysystemswithor

6-72 d -- NA 34-45 -62-49 Range 99.4-99.8 97.3-99.9 99.6-99.8 99.8-99.9 98.6-99.5 >99.8->99.9 Control efficiency (%) d e -- 7.7 38 NA 99.1 99.6 99.1 99.7 99.8 >99.9 Average 2.1-11 1.4-20 Range 1.1-1.9 0.53-0.72 0.013-0.39 0.018-0.022 0.021-0.059 0.019-0.029 0.019-0.031 <0.0049-<0.0053 lb/ton waste -3 10 c 5.1 1.5 8.5 0.59 0.019 0.088 0.043 0.024 0.026 <0.0052 Average Cadmium emission factors Range 1.1-5.3 0.71-10 0.56-0.93 0.26-0.36 0.0064-0.20 0.010-0.029 0.0096-0.016 0.0088-0.011 0.0096-0.015 <0.0025-<0.0027 g/Mg waste

EFFICIENCIES FOR MEDICAL WASTE INCINERATORS 2.6 4.3 0.75 0.30 0.044 0.021 0.012 0.013 0.0097 <0.0026 Average

TABLE 6-22. SUMMARY OF CONTROLLED CADMIUM EMISSION FACTORS AND CONTROL 3 3 3 3 9 5 3 3 9 11 runs No. of , C = carbon injection. 1 1 4 1 1 1 1 1 1 1 No. of facilities b Control status WS VS VS/PB DI/ESP FF DI/FF DI/FF+C SD/FF SD/FF+C DI/FF a M RB type Waste NA = not available. Efficiency data were available for only one facility. M = mixed medical waste, RB = red bag waste. WS = wet scrubber, VS = venturi scrubber, PB = packed bed scrubber, DI = dry injection Ranges are for individual runs. Averages were obtained by taking the arithmetic mean of facility averages. ESP = electrostatic precipitator, FF = fabric filter, SD = spray dryer c d e Source: References 35, 36, 40-46, 48, 50. a b 6-68 withoutcarboninjection. Forthewetsystems,theoutlet cadmiumemissionsrangefrom160percenthigherto49percent lowerthantheinletemissions. Table6-23presentsthebest typicaluncontrolledemissionfactorsforMWI’s. Toobtainbest typicalcontrolledemissionfactorsforventuriscrubber/packed bedsystems,applya38-percentefficiencytotheseuncontrolled emissionfactors. Fordrysystemswithorwithoutcarbon injection,applya97-percentefficiencytotheseuncontrolled emissionfactors.

TABLE6-23.BESTTYPICALUNCONTROLLEDCADMIUMEMISSIONFACTORSFORMEDICAL WASTEINCINERATORS

Cadmiumemissionfactors

-3 Wastetype g/MgofWaste 10 lb/tonofWaste

Mixed 2.5 5.0

RedBag 1.6 3.3

Pathological 0.18 0.37

6-74 REFERENCES 1. UnitedStatesDepartmentofEnergy. StateEnergyData Report,ConsumptionEstimates,1960-1990. Report No.DOE/EIA-0214(90). EnergyInformationAdministration, OfficeofEnergyMarketsandEndUse,UnitedStates DepartmentofEnergy,Washington,DC. May1992. 2. Brooks,Garry. EstimatingAirToxicEmissionsfromCoaland OilCombustionSources. EPA-450/2-89-001. U.S. EnvironmentalProtectionAgency,ResearchTrianglePark,NC. April1989. 3. White,D.M.,L.O.Edwards,A.G.Eklund,D.A.DuBose,and F.D.Skinner. CorrelationofCoalPropertieswith EnvironmentalControlTechnologyNeedsforSulfurandTrace Elements. EPA-600/7-84-066. U.S.EnvironmentalProtection Agency,ResearchTrianglePark,NC. 1984. 4. Swanson,V.E.,J.H.Medlin,J.R.Hatch,S.L.Coleman, G.H.Wood,S.D.Woodruff,andR.T.Hildebrand. Collection, ChemicalAnalysis,andEvaluationofCoalSamplesin1975. USGSReportNo.76-468. UnitedStatesDepartmentofthe Interior,GeologicalSurvey.Washington,DC. 1976. 5. U.S.EnvironmentalProtectionAgency, CompilationofAir PollutantEmissionFactors,AP-42,FourthEdition, SupplementB. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. September1988. 6. BabcockandWilcoxCompany. Steam: ItsGenerationandUse. BabcockandWilcox,NewYork,NY. 1978. 7. LetterfromR.B.Finkelman,UnitedStatesDepartmentofthe Interior,toD.D.Wallace,MidwestResearchInstitute. February8,1993. Metalconcentrationsincoal. 8. Buonicore,A.J.andW.T.Davis,eds. AirPollution EngineeringManual. VanNostrandReinhold,NewYork, NY,1992. 9. ,D.J. ControllingToxicEmissions. Coal. June1990. 10. Ondov,J.M.,R.C.Ragaini,andA.H.Biermann. Elemental EmissionsfromaCoal-FiredPowerPlant. Comparisonofa VenturiWetScrubberSystemwithaCold-SideElectrostatic Precipitator. EnvironmentalScienceandTechnology. Volume13,Number5. May1979.

6-75 11. Osborne,W.E.andM.D.McDannel. EmissionsofAirToxic Species: TestConductedUnderAB2588fortheWesternStates PetroleumAssociation. ReportNo.CA72600-2601. CARNOT, Tustin,CA. May1990. 12. Pape&SteinerEnvironmentalServices. AB2588Testingat TexacoTradingandTransportation,Inc. Panoche Station-Heater,June18throughJune29,1990. Report PS-90-2187. Pape&SteinerEnvironmentalServices Bakersfield,CA. September1990. 13. Hopkins,K.C.andL.AGreen. AirToxicsTestingatMorro BayUnit3. March1-5,1990CARNOT,Tustin,CA. May1990. 14. U.S.EnvironmentalProtectionAgency, CompilationofAir PollutantEmissionFactors,AP-42,FourthEdition, SupplementE. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. October1992. 15. EmissionFactorDocumentationforAP-42Section1.6,Wood WasteCombustioninBoilers. U.S.EnvironmentalProtection Agency,ResearchTrianglePark,NC. 1992. 16. Schultz,S. AuberryEnergy,Inc.ToxicHotSpotsReport. YankeEnergy. October1990. 17. DeAngelis,D.G.,D.S.Ruffin,J.A.Peters,andR.B.Reznik. SourceAssessment: ResidentialWoodCombustion. EPA-600/2-80-426. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. 1980. 18. U.S.EnvironmentalProtectionAgency, EmissionFactor DocumentationforAP-42Section2.1,Refuse Combustion.U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. September1992. 19. RadianCorporation. MunicipalWasteCombustionStudy. CharacterizationoftheMunicipalWasteCombustionIndustry. EPA-530/SW-87-021h. U.S.EnvironmentalProtectionAgency, Washington,DC. June1987. 20. RadianCorporation. LocatingandEstimatingAirToxics EmissionsfromMunicipalWasteCombustors. EPA-450/2-89-006. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. April1989. 21. Greenberg,R.R.,W.H.Zoller,andG.E.Gordon. Composition andSizeDistributionsofParticlesReleasedinRefuse Incineration. EnvironmentalScienceandTechnology Volume12,Number5. May1978.

6-76 22. U.S.EnvironmentalProtectionAgency, EmissionFactor DocumentationforAP-42Section2.5,SewageSludge Incineration. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. September1992. 23. RadianCorporation. LocatingandEstimatingAirToxics EmissionsfromSewageSludgeIncinerators. EPA-450/2-90-009. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. May1990. 24. Bennett,R.L.andK.T.Knapp. Characterizationof ParticulateEmissionsfromMunicipalWastewaterSludge Incinerators. EnvironmentalScienceandTechnology. Volume16,Number12. December1982. 25. MidwestResearchInstitute. MedicalWasteIncinerators- BackgroundInformationforProposedStandardsand Guidelines: IndustryProfileReportforNewandExisting Facilities. DraftReport. Cary,NC. April1992. 26. U.S.EnvironmentalProtectionAgency, MedicalWaste Incinerators-BackgroundPaperforNewandExisting Facilities,Draft. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. June1992. 27. FranklinAssociates,Ltd. CharacterizationofProducts ContainingLeadandCadmiuminMunicipalSolidWasteinthe UnitedStates,1970to2000. EPA-530/SW-89-015A. U.S. EnvironmentalProtectionAgency,Washington,DC. January1989. 28. Doyle,B.,D.Drum,andJ.Lauber. TheSmolderingQuestion ofHospitalWastes. PollutionEngineeringMagazine. July1985. 29. RadianCorporation. HospitalWasteCombustionStudy: Data GatheringPhase. EPA-450/3-88-017. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. December1988. 30. Neulicht,R.,M.Turner,L.Chaput,D.Wallace,and S.Smith, OperationandMaintenanceofHospitalWaste Incinerators. EPA-450/3-89-002. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. March1989. 31. Neulicht,R.,L.Chaput,D.Wallace,M.Turner,and S.Smith, HospitalIncineratorOperatorTrainingCourse: VolumeIStudentHandbook. EPA-450/3-89-003. U.S.EnvironmentalProtectionAgency,ResearchTriangle Park,NC. March1989.

6-77 32. MidwestResearchInstitute. MedicalWasteIncinerators- BackgroundInformationforProposedStandardsand Guidelines: ControlTechnologyPerformanceReportforNew andExistingFacilities. Cary,NC. July1992. 33. Bumbaco,M.J. ReportonStackSamplingProgramtoMeasure theEmissionsofSelectedTraceOrganicCompounds, Particulates,HeavyMetalsandHClfromtheRoyalJubilee HospitalIncinerator,Victoria,BC. EnvironmentCanada. April1983. 34. McCormack,J.,P.Ouchida,andG.Lew. EvaluationTestona SmallHospitalRefuseIncinerator,SaintBernadines Hospital,SanBernadino,California,ARB/ML-89-028. CaliforniaAirResourcesBoard,Sacramento,CA. July1989. 35. England,G.,D.Hansell,J.Newhall,andN.Soelberg. MichiganHospitalIncineratorEmissionsTestProgram, VolumeII: SiteSummaryReportBorgessMedicalCenter Incinerator. EnergyandEnvironmentalResearchCorporation, Irvine,CA. August1991. 36. England,G.,D.Hansell,J.Newhall,andN.Soelberg. MichiganHospitalIncineratorEmissionsTestProgram, VolumeIII: SiteSummaryReportUniversityofMichigan MedicalCenterIncinerator,EnergyandEnvironmental ResearchCorporation,Irvine,CA. August1991. 37. RadianCorporation. MedicalWasteIncinerationEmission TestReport,LenoirMemorialHospital,Kinston,North Carolina,EMBReport90-MWI-3. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. May1990. 38. RadianCorporation. MedicalWasteIncinerationEmission TestReport,CapeFearMemorialHospital,Wilmington,North Carolina. EMBReport90-MWI-4. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. November1990. 39. RadianCorporation. MedicalWasteIncinerationEmission TestReport,AMICentralCarolinaHospital,Sanford,North Carolina. EMBReport90-MWI-5. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. December1990. 40. RadianCorporation. MedicalWasteIncinerationEmission TestReport,BorgessMedicalCenter,Kalamazoo,Michigan. EMBReport91-MWI-9. U.S.EnvironmentalProtectionAgency, ResearchTrianglePark,NC. December1991.

6-78 41. RadianCorporation. MedicalWasteIncinerationEmission TestReport,MorristownMemorialHospital,Morristown,New Jersey. EMBReport91-MWI-8. U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. December1991. 42. ETS,Inc. ComplianceTestingforSouthlandExchangeJoint Venture,Hampton,SouthCarolina. ETS,Inc.,Roanoke,VA. July1989. 43. SourceTestReport,MEGAofKentucky. August1988. 44. U.S.EnvironmentalProtectionAgency. EmissionFactor DocumentationforAP-42Section2.6,MedicalWaste Incineration. U.S.EnvironmentalAgency,ResearchTriangle Park,NC. September1992. 45. AirManagementServices. EmissionsTestReportfor Fox-ChaseCancerCenter,Philadelphia,PA. February1989. 46. AirNova,Inc. EmissionComplianceTestProgramforNazareth Hospital,Philadelphia,PA. September1989. 47. Jenkins,A.,P.Ouchida,andG.Lew. EvaluationRetestona HospitalRefuseIncineratoratSutterGeneralHospital, Sacramento,California,ARB/ML-88-026. CaliforniaAir ResourcesBoard,Sacramento,CA. April1988. 48. Jenkins,A.,P.Ouchida,andG.Lew. EvaluationTestona RefuseIncineratoratStanfordUniversityEnvironmental SafetyFacility,Stanford,California,ARB/ML-88-025. CaliforniaAirResourcesBoard,Sacramento,CA. April1988. 49. Jenkins,A.,P.Ouchida,andD.Simeroth. EvaluationTest onaHospitalRefuseIncineratoratSaintAgnesMedical Center,Fresno,California,ARB/SS-87-01. CaliforniaAir ResourcesBoard,Sacramento,CA. January1987. 50. Jenkins,A.,C.Castronovo,G.Lindner,P.Ouchida,and D.Simeroth. EvaluationTestonaHospitalRefuse IncineratoratCedarsSinaiMedicalCenter,LosAngeles, California,ARB/SS-87-11. CaliforniaAirResourcesBoard, Sacramento,CA. April1987.

6-79 SECTION7 EMISSIONSFROMNONFERROUSSMELTING/REFINING

Cadmiumisemittedfromvariousnonferroussmeltingand refiningoperationsincludingthefollowing:

1. Primaryleadsmelting; 2. Primarycoppersmelting; 3. Primaryzincsmeltingandrefining--electrolytic process; 4. Primaryzincsmeltingandrefining--electrothermic process; 5. Secondarycoppersmeltingandrefining; 6. Secondaryzincrecoveryfrommetallicscrap;and 7. Secondaryzincrecoveryfromsteelproduction.

Rawmaterialsprocessedatthefacilitieslistedabove includemineralsandoresextractedfromtheearth,aswellas metallicwastefrommiscellaneousproducts. Theserawmaterials containcadmium. Atvariousstagesofmanufacturing,theraw materialsareprocessedatelevatedtemperatures,therefore resultingincadmiumbeingemittedfromtherawmaterials. This sectionpresentsprocessinformation,airpollutioncontrol measures,andestimatesofcadmiumemissionsfromthesesources.

PRIMARYLEADSMELTING

Leadisrecoveredfromasulfideore,primarilygalena(lead sulfide[PbS]),whichalsocontainssmallamountsofcopper, iron,zinc,andothertraceelements. Cadmiumalsocanbe expectedtobepresentintheleadore. Onedatasourcehas

7-1 reportedthatthecadmiumcontentinleadoreisapproximately 0.02percent.1

Alistofprimaryleadsmelterscurrentlyinoperation withintheUnitedStates(U.S.)isgiveninTable7-1.2

TABLE7-1.DOMESTICPRIMARYLEADSMELTERSANDREFINERIES

Smelter Refinery 1990Production,Mg(tons)

ASARCO,EastHelena,MT ASARCO,Omaha,NE 61,000(67,000)

ASARCO,Glover,MO Samesite 112,000(123,200)

DoeRun(formerlySt.Joe), Samesite 231,000(254,100) Herculaneum,MO

Source:Reference2.

Adescriptionoftheprocessusedtomanufactureleadanda discussionoftheemissionsresultingfromthevariousoperations arepresentedbelow.

ProcessDescription3

Figure7-1containsaprocessflowdiagramforprimarylead smelting. Therecoveryofleadfromtheleadoreconsistsof threemainsteps: sintering,reduction,andrefining.

Sinteringiscarriedoutinasinteringmachine,whichisa continuoussteelpalletconveyorbelt. Eachpalletconsistsof perforatedgrates,andbeneaththegratesarewindboxes,which areconnectedtofanstoprovideadraftthroughthemoving sintercharge. Thesinteringreactionstakeplaceatabout 1000°C(1832°F)duringwhichleadsulfideisconvertedtolead

7-2 oxideandleadsulfate. Sincecadmiumboilsatapproximately 767°C(1415°F),mostofthecadmiumintheorecanbeexpectedto beemittedduringsintering.

Reductionofthesinteredleadiscarriedoutinablast furnaceatatemperatureof1600°C(2912°F). Thefurnaceis chargedwithamixtureofsinter(80to90percentofcharge); metallurgicalcoke(8to14percentofcharge);andother materials,suchaslimestone,silica,litharge,andunspecified -formingconstituents. Intheblastfurnace,thesinteris reducedtolead. Thelevelofheatneededtocreatethereaction issuppliedbycokecombustion. Slag,consistingofimpurities, iscontinuouslycollectedfromthefurnaceandiseither processedatthesmelter(foritsmetalcontent)orshippedto treatmentfacilities. Theimpuritiesincludearsenic,, coppersulfideandothermetalsulfides,andsilicates. Lead bullion,whichistheprimaryproduct,undergoesapreliminary treatmenttoremoveimpurities,suchascopper,sulfur,, antimony,andnickel,beforefurtherrefiningoccurs. Any residualcadmiumleftintheorecanbeexpectedtobeemitted duringthereductionstep. Refiningoftheleadbullionis carriedoutincastironkettles. Refinedlead,whichis99.99 to99.999percentpure,iscastintopigsforshipment.

EmissionControlMeasures3

CadmiumemissionsourcesareindicatedinFigure7-1by solidcircles. Emissioncontrolsonleadsmelteroperationsare usedforcontrollingparticulatematter(PM)andsulfurdioxide

(SO2)emissionsresultingfromtheblastfurnaceandsintering machines. Centrifugalcollectors(cyclones)areusedin conjunctionwithfabricfiltersorelectrostaticprecipitators (ESP’s)forPMcontrol. Becausecadmiumhasaboilingpointof approximately767°C(1415°F),mostofthecadmiumwill

7-4 potentiallycondenseinthecyclone. Thus,ahighdegreeof controlofcadmiumemissionsmaybeachievedinthefabricfilter orESP. However,nodataontheeffectivenessoffabricfilters andESP’sincontrollingcadmiumemissionsareavailable.

ControlofSO2 isachievedbyabsorptiontoformsulfuric acidinthesulfuricacidplants,whicharecommonlypartof lead-smeltingplants.

Emissions

Cadmium,whichmayexistintheoreata0.02-percent concentration,canpotentiallybeemittedwhentemperaturesreach highlevelsinthesinteringandreducingsteps. Becausethe sinteringstepiscarriedoutattemperaturesmuchhigherthan theboilingpointofcadmium,thesinteringstepisconsideredto betheprimarysourceofcadmiumemissions. Table7-2presents estimatesofcadmiumemissionsreportedbythreefacilities during1990,asrequiredunderSuperfundAmendmentsand ReauthorizationAct(SARA)TitleIIIregulations.4 TABLE7-2.PRIMARYLEADPRODUCERSREPORTINGCADMIUMEMISSIONSINTHE 1990TOXICSRELEASEINVENTORY

Emissions,kg(lbs)

Smelter Nonpoint Point Total

ASARCOInc.,Glover,MO 111(245) 415(914) 526(1,159)

ASARCOInc.,EastHelena,MT 3,175(6,985) 4,990(10,978) 8,165(17,963)

DoeRun,Herculaneum,MO 47(104) 5,526(12,157) 5,573(12,261)

Source:Reference4.

7-5 Testdatapertainingtocadmiumemissionsformthevarious operationsarenotavailable. Table7-3presentscadmium emissionfactorsreportedintheUSEPAdatabase,SPECIATE,for variousoperationsduringprimaryleadsmelting.5 Becausethe validityoftheseemissionfactorscannotbeverified,extreme cautionshouldbeexercisedwhenusingthesefactors.

PRIMARYCOPPERSMELTING

Theprincipalmethodforrecoveringcopperfromsulfideore ispyrometallurgicalsmelting. Copperorescontainsignificant quantitiesofarsenic,cadmium,lead,antimony,andotherheavy metals. Onedatasourcehasreportedthatcadmiumcontentin leadoreisapproximately0.01percent.6

Alistofprimarycoppersmelterscurrentlyoperatingwithin theU.S.isgiveninTable7-4.7 Adescriptionoftheprocess usedtomanufacturecopperandadiscussionoftheemissions resultingfromthevariousoperationsarepresentedbelow.

ProcessDescription8

Aconventionalpyrometallurgicalcopper-smeltingprocessis illustratedinFigure7-2.Thisprocessincludesroastingore concentratestoproducecalcine,smeltingofroasted(calcine feed)orunroasted(greenfeed)oreconcentratestoproduce matte,andconvertingthemattetoyieldblistercopperproduct (about99percentpure). Typically,theblistercopperis refinedinananodefurnace,castinto"anodes",andthensentto anelectrolyticrefineryforfurtherimpurityelimination.

Inroasting,chargematerialofcopperconcentrate,mixed withasiliceousflux(oftenalowgradeore),isheatedinair toabout650°C(1200°F),eliminating20to50percentofthe

7-6 TABLE7-3.CADMIUMEMISSIONFACTORSFORLEAD-SMELTING FACILITIES

Source Cadmiumemissionfactora classification Emissionsource code(SCC) lb/ton kg/Mg Sintering:singlestream 30301001 1.39941b 0.7b Blastfurnaceoperation 30301002 41.74965b 20.9b Drossreverberatoryfurnace 30301003 0.2438b 0.1219b Orecrushing 30301004 0.01668c 0.00834c Sintering:dualstreamfeedend 30301006 9.67987b 4.84b Slagfumefurnace 30301008 0.00359d 0.0018d Leaddrossing 30301009 0.00203d 0.001d Rawmaterialcrushingandgrinding 30301010 0.04553d 0.023d Rawmaterialunloading 30301011 0.00334e 0.00167e Rawmaterialstoragepiles 30301012 0.0025e 0.00125e Rawmaterialtransfer 30301013 0.00417e 0.00209e Sinteringchargemixing 30301014 0.01885e 0.00943e Sintercrushing/screening 30301015 0.06829f 0.03415f Sintertransfer 30301016 0.00911f 0.00456f Sinterfinesreturnhandling 30301017 0.40977f 0.2049f Blastfurnacetapping(metalandslag) 30301019 0.00728d 0.00364d Blastfurnaceleadpouring 30301020 0.04234d 0.02117d Blastfurnaceslagpouring 30301021 0.00075d 0.00038d Leadrefining/silverretort 30301022 0.08195d 0.04098d Leadcasting 30301023 0.03961d 0.0198d Reverberatoryorkettlesoftening 30301024 0.13659d 0.0683d Sintermachineleakage 30301025 0.02519f 0.0126f Sinterdumparea 30301026 0.00046f 0.00023f

Source:Reference5. aAllemissionfactorsarereportedasfoundintheSPECIATEdatabasewithoutroundingoff (Reference5).EmissionfactorsinSPECIATEdatabasearereportedinlb/tonofprocess activity. blb/ton(kg/Mg)ofconcentratedore. clb/ton(kg/Mg)oforecrushed. dlb/ton(kg/Mg)ofleadproduct. elb/ton(kg/Mg)ofrawmaterial. flb/ton(kg/Mg)ofsinter.

7-7 TABLE7-4.DOMESTICPRIMARYCOPPERSMELTERSANDREFINERIES

Smelter 1992Capacity,Mg(tons)

ASARCOInc.,Hayden,AZ 170,000(187,000)

CyprusMiamiMiningCo.,Globe,AZ 180,000(198,000)

MAGMACopperCo.,SanManuel,AZ 290,000(319,000)

CopperRangeCo.,WhitePine,MI 60,000(66,000)

PhelpsDodge,Hidalgo,NM 190,000(209,000)

ChinoMinesCo.,Hurley,NM 170,000(187,000)

ASARCOInc.,ElPaso,TX 104,000(114,400)

Kennecott,Garfield,UT 210,000(231,000)

Source:Reference7.

7-8 sulfurasSO2. Portionsofsuchimpuritiesasantimony,arsenic, andleadaredrivenoff,andsomeironisconvertedtooxide. Theroastedproduct,calcine,servesasadriedandheatedcharge forthesmeltingfurnace. Eithermultiple-hearthorfluidized- bedroastersareusedforroastingcopperconcentrate. Multiple-hearthroastersacceptmoistconcentrate,whereas fluidized-bedroastersarefedfinelygroundmaterial(60percent minus200mesh). Withbothofthesetypes,theroastingis autogenous. Becausethereislessairdilution,higherSO2 concentrationsarepresentinfluidized-bedroastergasesthanin multiple-hearthroastergases. Becausecadmiumhasaboiling pointof767°C(1415°F),mostofthecadmiumintheoremay remaininthecalcine,insteadofbeingemittedasanair pollutantduringroasting.

Inthesmeltingprocess,eitherhotcalcinesfromthe roasterorrawunroastedconcentratesaremeltedwithsiliceous fluxinasmeltingfurnacetoproducecoppermatte,amolten mixtureofcuproussulfide(Cu2S),ferroussulfide(FeS)andsome heavymetals. Therequiredheatcomesfrompartialoxidationof thesulfidechargeandfromburningexternalfuel. Mostofthe ironandsomeoftheimpuritiesinthechargeoxidizewiththe fluxestoformaslagontopofthemoltenbath;thisslagis periodicallyremovedanddiscarded. Coppermatteremainsinthe furnaceuntiltapped. Mattesproducedbythedomesticindustry rangefrom35to65percentcopper,with45percentbeingthe mostcommon. Thecoppercontentpercentageisreferredtoasthe mattegrade. Currently,fivesmeltingfurnacetechnologiesare usedintheU.S.: reverberatory,electric,Noranda,Outokumpu (flash),andInco(flash). Reverberatoryfurnacesmayoperateat temperaturesashighas1500°C(2732°F),whileflashfurnacesmay operateattemperaturesof1000°F(1832°F). Eventhoughthe exacttemperaturesatwhichtheothertwofurnacetechnologies (electricandNoranda)operatearenotknown,itisprobablethat

7-10 theyoperateattemperatureshigherthantheboilingpointof cadmium. Therefore,mostofthecadmiumthatremainsinthe calcinemaybeemittedasanairpollutantduringthesmelting step.

Reverberatoryfurnaceoperationisacontinuousprocess, withfrequentchargingofinputmaterialsandperiodictappingof matteandskimmingofslag. Heatissuppliedbycombustionof oil,gasorpulverizedcoal,andfurnacetemperaturesmayexceed 1500°C(2732°F).

Forsmeltinginelectricarcfurnaces,heatisgeneratedby theflowofanelectriccurrentincarbonelectrodes. These electrodesareloweredthroughthefurnaceroofandsubmergedin theslaglayerofthemoltenbath. Thefeedgenerallyconsists ofdriedconcentratesorcalcines;chargingwetconcentratesis avoided. Thechemicalandphysicalchangesoccurringinthe moltenbatharesimilartothoseoccurringinthemoltenbathof areverberatoryfurnace. Also,thematteandslag-tapping practicesaresimilaratbothfurnaces. Electricfurnacesdonot producefuelcombustiongases,soflowratesarelowerandSO2 concentrationsarehigherintheeffluentgasofelectric furnacesthanintheeffluentgasofreverberatoryfurnaces.

Flashfurnacesmeltingcombinestheoperationsofroasting andsmeltingtoproduceahigh-gradecoppermattefrom concentratesandflux. Inflashsmelting,driedoreconcentrates andfinelygroundfluxesareinjected,togetherwithoxygen, preheatedair,oramixtureofboth,intoafurnaceofspecial designwheretemperatureismaintainedatapproximately1000°C (1832°F). Incontrasttoreverberatoryandelectricfurnaces, flashfurnacesusetheheatgeneratedfrompartialoxidationof theirsulfidechargetoprovidemuchoralloftheenergy(heat)

7-11 requiredforsmelting. Theyalsoproduceoffgasstreams containinghighconcentrationsofSO2.

Slagproducedbyflashfurnaceoperationscontains significantlyhigheramountsofcopperthanisfoundin reverberatoryorelectricfurnaceoperations. Asaresult,the flashfurnaceandconverterslagsaretreatedinaslag-cleaning furnacetorecoverthecopper. Slag-cleaningfurnacesusually aresmallelectricfurnaces. Theflashfurnaceandconverter slagsarechargedtoaslag-cleaningfurnaceandareallowedto settleunderreducingconditions,withtheadditionofcokeor ironsulfide. Thecopper,whichisintheoxideforminthe slag,isconvertedtocoppersulfide. Thecoppersulfideis subsequentlyremovedfromthefurnaceandischargedtoa converterwithregularmatte. Iftheslag’scoppercontentis low,theslagisdiscarded.

TheNorandaprocess,asoriginallydesigned,allowedthe continuousproductionofblistercopperinasinglevesselby effectivelycombiningroasting,smeltingandconvertingintoone operation. Metallurgicalproblems,however,ledtotheoperation ofthesereactorsfortheproductionofcoppermatte. Asin flashsmelting,theNorandaprocesstakesadvantageoftheheat energyavailablefromthecopperore. Theremainingthermal energyrequirementissuppliedbyoilburners,orbycoalmixed withtheoreconcentrates.

Thefinalstepinblistercopperproductionisconversion. Thisstepeliminatestheremainingironandsulfurpresentinthe matteandleavesbehindonlythemolten"blister"copper. All butoneU.S.smelterusesPierce-Smithconverters,whichare refractory-linedcylindricalsteelshellsmountedontrunnionsat eitherend,androtatedaboutthemajoraxisforchargingand pouring. Anopeninginthecenteroftheconverterfunctionsas

7-12 amouththroughwhichmoltenmatte,siliceousflux,andscrap copperarecharged,andgaseousproductsarevented. Airor -richairisblownthroughthemoltenmatte. Ironsulfide

(FeS)isoxidizedtoironoxide(FeO)andSO2,andtheblowing andslag-skimmingstepsarerepeateduntilanadequateamountof relativelypureCu2S,called"whitemetal,"accumulatesinthe bottomoftheconverter. Arenewedairblastthenoxidizesthe coppersulfidesulfurtoSO2,leavingblistercopperinthe converter. Theblistercopperissubsequentlyremovedand transferredtorefiningfacilities. Thissegmentofconverter operationistermedthefinishblow. TheSO2 producedthroughout theoperationisventedtopollutioncontroldevices.

OnedomesticsmelterusesHobokenconverters,whichhasthe advantageofemissioncontrols. TheHobokenconverteris essentiallylikeaconventionalPierce-Smithconverter,except thatthisvesselisfittedwithasideflueatoneend,whichis shapedasaninvertedU. Thisfluearrangementpermitssiphoning ofgasesfromtheinterioroftheconverterdirectlyintothe offgascollectionsystem. Thisleavestheconvertermouthunder aslightvacuum.

Blistercopperusuallycontainsfrom98.5to99.5percent purecopper. Impuritiesmayincludegold,,antimony, arsenic,,iron,lead,nickel,selenium,sulfur, ,andzinc. Topurifyblistercopperfurther,fire refiningandelectrolyticrefiningareused. Infirerefining, blistercopperisplacedinafire-refiningfurnace;afluxis usuallyadded,andairisblownthroughthemoltenmixtureto oxidizeremainingimpurities,whichareremovedasaslag. The remainingmetalbathissubjectedtoareducingatmosphere,which reconvertscuprousoxidetocopper. Thetemperatureinthe furnaceisaround1100°C(2012°F). Thefire-refinedcopperis thencastintoanodes,andthenfurtherelectrolyticrefining

7-13 separatescopperfromimpuritiesbyelectrolysisinasolution containingcoppersulfateandsulfuricacid. Metallicimpurities precipitatefromthesolutionandformasludgethatisremoved andtreatedtorecoverpreciousmetals. Copperisdissolvedfrom theanodeanddepositedatthecathode. Cathodecopperis remeltedandmadeintobars,ingotsorslabsformarketing purpose. Thecopperproducedis99.95to99.97percentpure. Anyresidualcadmiumthathasnotbeenemittedduringthe smeltingstepmaybeemittedduringtherefiningstep.

EmissionControlMeasures8

CadmiumemissionsourcesareindicatedinFigure7-2by solidcircles. Emissioncontrolsoncoppersmeltersareusedfor

controllingPMandSO2 emissionsresultingfromroasters, smeltingfurnaces,andconvertors. Electrostaticprecipitators arethecommonPMcontroltechniquesemployedatcopper-smelting facilities.

ControlofSO2 emissionsisachievedbyabsorptionto sulfuricacidinthesulfuricacidplants,whicharecommonly partofcopper-smeltingplants.

Emissions

Cadmium,whichispresentintheore,canpotentiallybe emittedfromsmeltingfurnacesandconvertors. Table7-5 presentsestimatesofcadmiumemissionsreportedbythree facilitiesduring19904. Testdatapertainingtocadmiumemissionsfromprimary copperfacilitiesarelimited. Oneemissiontestreportat CopperRangeCompany,locatedinWhitePine,MI,containsresults ofmetalsanalysisandwasreviewedduringthisstudy.9 This facilityoperatesareverberatoryfurnacethatiscontrolledby

7-14 TABLE7-5.PRIMARYCOPPERPRODUCERSREPORTINGCADMIUMEMISSIONS INTHE1990TOXICSRELEASEINVENTORY

Emissions,kg(lbs)

Smelter Nonpoint Point Total

ASARCOInc.,ElPaso,TX 771(1,696) 3,048(6,706) 3,819(8,402)

ASARCOInc.,Hayden,AR 113(249) 1,203(2,647) 1,316(2,895)

Kennecott,Garfield,UT 113(249) 340(748) 453,(997)

Note: CypressMiamiMiningCo.,Globe,AZ,reportedzerocadimumemissionsinthe1990TRI.

Source:Reference4.

anESP. Theexhauststreamfromtheconvertor(whichis uncontrolled)ismixedwiththeexhaustfromtheESPoutletand isroutedthroughthemainstackanddischargedintothe atmosphere. Testingformetalswasperformedatthemainstack aftertwoexhauststreams(fromtheESPoutletandtheconvertor) weremixed. Cadmiumemissionsweremeasuredforthreemodesof convertoroperation: slag-blow,copper-blow,andconvertoridle (noblow)cycles. Thecadmiumlevelduringtheslag-blowcycle wasmeasuredtobethehighest,correspondingtoacadmium emissionrateof2.3509lb/hr. Additionally,theplantcapacity wasreportedtobeapproximately42tons/hroffeed,which consistsofmillconcentrate,limestone,ironpyrites,and recycledmaterial. Theactualprocessrateduringthetestis notknown. Sincethefeedmixvariesfromfacilitytofacility, thecadmiumemissionsmeasuredatCopperRangeCo.cannotbeused toestimateageneralcadmiumemissionfactorthatwouldbevalid onanindustrywidebasis. Additionally,CopperRangeCo.,isthe onlyfacilityintheU.S.thatoperatesareverberatoryfurnace. Allothercopper-smeltingfacilitiesuseflashfurnaceswhich inherentlyproducelessemissions.

7-15 Theonlyavailableemissionfactordataarefromthe SPECIATEdatabase. Table7-6presentscadmiumemissionfactors forvariousemissionpointsatprimarycopper-smeltingfacilities asreportedintheSPECIATEdatabase.5 Becausethevalidityof theseemissionfactorscannotbeverified,extremecautionshould beexercisedwhenusingthesefactors.

PRIMARYZINCSMELTINGANDREFINING

Zincisfoundprimarilyasthesulfideore, (ZnS). Itscommoncoproductoresareleadandcopper. Metal impuritiescommonlyassociatedwithZnSarecadmium(upto2 percent)andminorquantitiesofgermanium,,,and . Zincorestypicallycontainfrom3to11percentzinc. Someores,containingaslittleas2percent,arerecovered. Concentrationattheminebringsthisto49to54percentzinc, withapproximately31percentfreeanduncombinedsulfur.10

Alistofprimaryzincsmelterscurrentlyinoperation withintheU.S.isgiveninTable7-7.11 Zincoresareprocessed intometallicslabzincbytwobasicprocesses. Threeofthe fourdomesticU.S.zinc-smeltingfacilitiesusetheelectrolytic process,andoneplantusesapyrometallurgicalsmeltingprocess, whichistypicaloftheprimarynonferroussmeltingindustry. Theplantthatusesthepyrometallurgicalprocessprovidesenergy byelectricresistanceheating. Therefore,inthiscase,the pyrometallurgicalprocessisreferredtoastheelectrothermic process. Adescriptionoftheprocessusedtomanufacturezinc bytheelectrolyticandelectrothermicprocessesandadiscussion oftheemissionsresultingfromthevariousoperationsare presentedbelow.

7-16 TABLE7-6.CADMIUMEMISSIONSFROMPRIMARYCOPPERPRODUCTION

Cadmium Source emissionfactora classification lb/ton kg/Mg Emissionsource code(SCC) Reverberatorysmeltingfurnaceafterroaster--ESP 30300503 0.005 0.0025 Convertor(allconfigurations)--ESP 30300504 0.0036 0.0018 Fire(furnace)refining--ESP 30300505 0.001 0.0005 Oreconcentratedryer--ESP 30300506 0.001 0.0005 Reverberatorysmeltingfurnacewithorecharging 30300507 0.005 0.0025 (withoutroasting)--ESP Fluidized-bedroaster--ESP 30300509 0.0055 0.00275 Electricsmeltingfurnace--ESP 30300510 0.01 0.005 Flashsmelting 30300512 2.3128 1.1564 :fugitiveemissions--ESP 30300513 0.00026 0.00013 Reverberatoryfurnace:fugitiveemissions--ESP 30300514 0.00716 0.00358 Convertor:fugitiveemissions 30300515 0.02829 0.014145 Anoderefiningfurnace:fugitiveemissions--ESP 30300516 0.00005 0.000025 Slag-cleaningfurnace:fugitiveemissions--ESP 30300517 0.0008 0.0004 Slag-cleaningfurnace--ESP 30300522 0.001 0.0005 AFTMHR+RF/FBR+EF 30300524 0.18 0.09 Fluidized-bedroasterwithreverberatoryfurnace+ 30300525 0.0055 0.00275 convertor--ESP Concentratedryerwithelectricfurnace,cleaning 30300526 0.001 0.0005 furnaceandconvertor--ESP Concentratedryerwithflashfurnaceand 30300527 0.001 0.0005 convertor--ESP

Source:Reference5. aAllemissionfactorsarereportedasfoundintheSPECIATEdatabasewithoutroundingoff (Reference5).Allemissionfactorsreportedaboveareintheunitsoflb/tonofconcentratedore.

7-17 TABLE7-7.DOMESTICPRIMARYZINCPRODUCERS

1992slabzincproduction Company Typeofprocess capacity,Mg(tons)

BigRiverZincCo.,Sauget,IL Electrolytic 82,000(90,200)

JerseyMiniersZincCo., Electrolytic 98,000(107,800) Clarksville,TN

ZincCorporationofAmerica, Electrolytic 51,000(56,100) Bartlesville,OK

ZincCorporationofAmerica, Electrothermic 123,000(135,300) Monaca,PA

Source:Reference11.

ProcessDescription-Electrolytic10

Ageneraldiagramoftheelectrolyticandelectrothermic processesispresentedinFigure7-3. Electrolyticprocessing involvesfourmajorsteps: roasting,leaching,purification,and .

Roastingisaprocesscommontobothelectrolyticand pyrometallurgicalprocessing. Calcineisproducedbythe roastingreactionsinanyoneofthreedifferenttypesof roasters: multiple-hearth,suspension,orfluidized-bed. Multiple-hearthroastersaretheoldesttypeusedintheUnited States,whilefluidized-bedroastersarethemostmodern. The primaryzinc-roastingreactionoccursbetween640°and1000°C (1184°and1832°F),dependingonthetypeofroasterused. The reactionis:

2ZnS + 3O2 >2ZnO + 2SO2 (1)

Inamultiple-hearthroaster,theconcentrateisblown throughaseriesofnineormorehearthsstackedinsideabrick-

7-18

linedcylindricalcolumn. Asthefeedconcentratedropsthrough thefurnace,itisfirstdriedbythehotgasespassingthrough thehearthsandthenoxidizedtoproducecalcine. Thereactions areslowandcanonlybesustainedbytheadditionoffuel.

Inasuspensionroaster,thefeedisblownintoacombustion chamber,whichisverysimilartothatofapulverizedcoal furnace. Additionalgrinding,beyondthatrequiredfora multiple-hearthfurnace,isnormallyrequiredtoassurethatheat transfertothematerialisfastenoughtoinitiate desulfurizationandoxidationreactionsinthefurnacechamber. Hearthsatthebottomoftheroastercapturethelarger particles,whichneedmoretimeinthefurnacetocompletethe desulfurizationreaction.

Inafluidized-bedroaster,finelygroundsulfide concentratesaresuspendedandoxidizedwithinapneumatically supportedfeedstockbed. Thistechniqueachievesthelowest sulfurcontentcalcineofthethreeroasterdesigns.

Suspensionandfluidized-bedroastersaresuperiortothe multiplehearthforseveralreasons. Althoughtheyemitmore uncontrolledparticulate,theirreactionratesaremuchfaster,

allowinggreaterprocessrates. Also,theSO2 contentofthe effluentstreamsofthesetworoastersissignificantlyhigher, permittingmoreefficientandeconomicaluseofacidplantsto

controlSO2 emissions.

Cadmiumhasaboilingpointofapproximately767°C(1413°F). Therefore,mostofthecadmiumcanbeexpectedtobeemitted duringtheroastingstep,irrespectiveofwhichprocessis followed.

7-20 Leachingisthefirststepofelectrolyticreduction. In thisstep,thezincoxidereactswithsulfuricacidtoform aqueouszincsulfateinanelectrolytesolution.

+2 -2 ZnO+H2SO4 >Zn(aq)+SO4 (aq)+H2O (2)

Singleanddoubleleachmethodscanbeused,althoughthe formerexhibitsexcessivesulfuricacidlossesandpoorzinc recovery. Indoubleleaching,thecalcineisfirstleachedina neutralorslightlyalkalinesolution. Thereadilysoluble sulfatesfromthecalcinedissolve,butonlyaportionofthe zincoxideentersthesolution. Thecalcineisthenleachedin theacidicelectrolysisrecycleelectrolyte. Thezincoxideis dissolvedasshowninreaction2,asaremanyoftheimpurities, especiallyiron. Theelectrolyteisneutralizedbythisprocess, anditservesastheleachsolutionforthefirststageofthe calcineleaching. Thisrecyclingalsoservesasthefirststage ofrefining,sincemuchofthedissolvedironprecipitatesoutof thesolution. Variationsonthisbasicprocedureincludetheuse ofprogressivelystrongerandhotteracidbathstobringasmuch ofthezincintosolutionaspossible.

Purificationisaprocessinwhichavarietyofreagentsare addedtothezinc-ladenelectrolytetoforceimpuritiesto precipitate. Thesolidprecipitatesareseparatedfromthe solutionbyfiltration. Thetechniquesthatareusedareamong themostadvancedindustrialapplicationsofinorganicsolution chemistry. Processesvaryfromsmeltertosmelter,andthe detailsareproprietaryandoftenpatented. Metallicimpurities, suchasarsenic,antimony,,,nickel,and thallium,interfereseverelywiththeelectrolytedepositionof zincandtheirfinalconcentrationsarelimitedtolessthan 0.05milligramsperliter(4x10-7 poundspergallon).

7-21 Electrolysistakesplaceintanks,orcells,containinga numberofcloselyspacedrectangularmetalplates,whichactas anodes(madeofleadwith0.75to1.0percentsilver)andas cathodes(madeofaluminum). Aseriesofthreemajorreactions occurswithintheelectrolysiscells:

H2SO4 +- 2H2O>4H(aq)+4e +O2 (3) anode cathode 2Zn+2 +4e- > Zn° (4)

+ -2 4H (aq)+2SO4 (aq) > 2H2SO4 (5)

Oxygengasisreleasedattheanode;metalliczincis depositedatthecathode,andsulfuricacidisregeneratedwithin theelectrolyte.

Electrolyticzincsmelterscontainalargenumberof cells,oftenseveralhundred. Aportionoftheelectricalenergy releasedinthesecellsdissipatesasheat. Theelectrolyteis continuouslycirculatedthroughcoolingtowers,bothtolowerits temperatureandtoconcentratetheelectrolytethroughthe evaporationofwater. Periodically,eachcellisshutdown,and thezincisremovedfromtheplates.

Thefinalstageofelectrolyticzincsmeltingisthe meltingandcastingofthecathodezincintosmallslabs, 27kilograms(59pounds),orlargeslabs,640to1,100kilograms (1,408to2,420pounds).

7-22 ProcessDescription--Pyrometallurgical(Electrothermic)10

Sinteringisthefirststageofthepyrometallurgical reductionofzincoxidetoslabzinc. Sinteringremovesleadand cadmiumimpuritiesbyvolatilizationandproducesanagglomerated permeablemasssuitableforfeedtoretortingfurnaces. DowndraftsinteringmachinesoftheDwight-Lloydtypeareusedin theindustry. Gratepalletsarejoinedtogetherforacontinuous conveyorsystem. Combustionairisdrawndownthroughthegrate palletsandisexhaustedtoaparticulatecontrolsystem. The feedisamixtureofcalcine,recycledsinter,andcokeorcoal fuel. Havingalowboilingpoint,oxidesofleadandcadmiumare volatilizedfromthesinterbedandarerecoveredinthe particulatecontrolsystem. Asdescribedearlier,mostofthe cadmiumcanbeexpectedtobeemittedduringtheroastingstep. Anyresidualcadmiumwouldbeemittedduringthesinteringstep.

Inretorting,becauseofthelowboilingpointofmetallic zinc,906°C(1663°F),reductionandpurificationofzinc-bearing mineralscanbeaccomplishedtoagreaterextentthanwithmost minerals. Thesinteredzincoxidefeedisbroughtintoahigh temperaturereducingatmosphereof900°to1499°C(1652°to 2730°F). Undertheseconditions,thezincoxideis simultaneouslyreducedandvolatilizedtogaseouszinc:

ZnO + CO > Zn(vapor) + CO2 (6)

Carbonmonoxideregenerationalsooccurs:

CO2 +C >2CO (7)

Thezincvaporandcarbonmonoxidethatareproducedpass fromthemainfurnacetoacondenserwherezincrecoveryis

7-23 accomplishedbybubblingthegasmixturethroughamoltenzinc bath.

Retortingfurnacescanbeheatedeitherexternallyby combustionflamesorinternallybyelectricresistanceheating. Thelatterapproach,electrothermicreduction,istheonlymethod currentlypracticedintheUnitedStates,andithasgreater thermalefficiencythandoexternalheatingmethods. Inaretort furnace,preheatedcokeandsinter,silicaandmiscellaneous zinc-bearingmaterialsarefedcontinuouslyintothetopofthe furnace. Feedcokeservesastheprincipalelectricalconductor, producingheat;italsoprovidesthecarbonmonoxiderequiredfor zincoxidereduction. Furtherpurificationstepscanbe performedonthemoltenmetalcollectedinthecondenser. The moltenzincfinallyiscastintosmallslabs,27kilograms (59pounds),orthelargeslabs,640to1,100kilograms(1,408to 2,420pounds). Anycadmiumthatisleftoverfromtheroasting andsinteringstepsmaybeemittedduringtheretortingstep.

EmissionControlMeasures10,12

CadmiumemissionsourcesareindicatedinFigure7-3by solidcircles. Emissioncontrolsusedatelectrolyticzinc smeltersincludefabricfiltersforcontrollingPMfromore storageandhandlingoperations,andzinc-smeltingoperations. Emissioncontrolsemployedatelectrothermiczincsmelters includefabricfiltersforcontrollingPMfromsintermachines, sintersizingandcrushingoperations,electrothermicfurnace preheaters,electrothermicfurnaces,zinc-holdingfurnaces,and zinc-refiningcolumns.

ControlofSO2 emissionsatbothelectrolyticand electrothermiczincsmeltersisachievedbyabsorptionto

7-24 sulfuricacidinthesulfuricacidplants,whicharecommonly partofzinc-smeltingplants.

Emissions

Cadmium,whichispresentintheore,canpotentiallybe emittedfromroasters(inbothelectrolyticandelectrothermic processes)andfromsinteringmachinesandretortingstepsofthe electrothermicprocess. Table7-8presentsestimatesofcadmium emissionsreportedbythefourfacilitiesduring1990.4 The onlyavailableemissionfactordataarefromtheSPECIATEdata base. Table7-9presentscadmiumemissionfactorsforvarious emissionpointsatprimaryzinc-smeltingfacilitiesasreported intheSPECIATEdatabase.5 Becausethevalidityofthese emissionfactorscannotbeverified,extremecautionshouldbe exercisedwhenusingthem. TABLE7-8.PRIMARYZINCPRODUCERSREPORTINGCADMIUMEMISSIONS INTHE1990TOXICSRELEASEINVENTORY

Company Typeofprocess Emissions,kg(lb)

BigRiverZincCo.,Sauget,IL Electrolytic 860(1,892)

JerseyMiniereZincCo., Electrolytic 227(499) Clarksville,TN

ZincCorporationofAmerica, Electrolytic 2,936(6,459) Bartlesville,OK

ZincCorporationofAmerica, Electrothermic 1,724(3,793) Monaca,PA

Source:Reference4.

7-25 TABLE7-9.CADMIUMEMISSIONSFROMPRIMARYZINCPRODUCTION

Cadmium emissionfactora Sourceclassificationcode Emissionsource (SCC) lb/ton kg/Mg

Multiple-hearthroaster 30303002 4.98492b 2.49246

Sinterstrand 30303003 1.9764b 0.9882

Verticalretort/electrothermal 30303005 0.00008b 0.00004 furnace

Electrolyticprocessor 30303006 0.06588b 0.03294

Flashroaster 30303007 43.92b 21.96

Fluidized-bedroaster 30303008 47.5873b 23.79365

Rawmaterialhandlingandtransfer 30303009 0.08784c 0.04392

Sinterbreakingandcooling 30303010 0.03294d 0.01647

Zinccasting 30303011 0.0549e 0.02745

Rawmaterialunloading 30303012 0.00878c 0.00439

Source:Reference5.

aAllemissionfactorsarereportedasfoundintheSPECIATEdatabasewithoutroundingoff (Reference5).EmissionfactorsinSPECIATEdatabasearereportedinlb/tonofprocess activity.Itappearsthattheemissionfactorsreportedaboveareuncontrolledfactors.

blb/ton(kg/Mg)ofconcentratedore.

clb/ton(kg/Mg)ofrawmaterialprocessed.

dlb/ton(kg/Mg)ofsinterprocessed.

elb/ton(kg/Mg)ofzincproduced.

SECONDARYCOPPERSMELTINGANDREFINING

Thesecondarycopperindustryprocessesscrapmetalsfor therecoveryofcopper. Productsincluderefinedcopperor copperalloysinformssuchasingots,wirebar,anodes,andshot. Copperalloysarecombinationsofcopperwithothermaterials,

7-26 notably,,zinc,andlead. Also,forspecialapplications, combinationsincludesuchmetalsascobalt,manganese,iron, nickel,cadmium,andberyllium,andnonmetals,suchasarsenic andsilicon.13

Alistofsecondarycoppersmelterscurrentlyoperating withintheUnitedStatesisgiveninTable7-10.7 Adescription oftheprocessusedtomanufacturesecondarycopperanda discussionoftheemissionsresultingfromthevariousoperations arepresentedbelow. TABLE7-10.DOMESTICSECONDARYCOPPERPRODUCERS

Smelter 1992Capacity,Mg(tons)

CerroCopperProducts,Sauget,IL 70,000(77,000)

Chemetco(ConcordeMetals),Alton,IL 135,000(148,500)

FranklinSmelting&Refining,Philadelphia,PA 16,000(17,600)

GastonRecyclingIndustries,Gaston,SC 110,000(121,000)

SouthwireCo.,Carrolton,GA 105,000(115,500)

CyprusCasaGrandeCorp.,Lakeshore,AZ 45,000(49,500)

Source:Reference7.

ProcessDescription13

Theprincipalprocessesinvolvedincopperrecoveryare scrapmetalpretreatmentandsmelting. Pretreatmentincludes cleaningandconcentrationtopreparethematerialforthe smeltingfurnace. Smeltinginvolvesheatingandtreatingthe scraptoachieveseparationandpurificationofspecificmetals.

Thefeedmaterialusedintherecoveryprocesscanbeany metallicscrapcontainingausefulamountofcopper,bronze

7-27 (copperandtin),orbrass(copperandzinc). Traditionalforms arepunchings;turningsandborings;defectiveorsurplusgoods; metallurgicalresiduessuchasslags,skimmings,anddrosses;and obsolete,worn-out,ordamagedarticles,includingautomobile radiators,pipe,wire,bushings,andbearings.

Thetypeandqualityofthefeedmaterialdeterminesthe processesthesmelterwilluse. Duetothelargevarietyof possiblefeedmaterialsavailable,themethodofoperationvaries greatlybetweenplants. Generally,asecondarycopperfacility dealswithlesspurerawmaterialsandproducesamorerefined product,whereasbrassandbronzealloyprocessorstakecleaner scrapanddolesspurificationandrefining. Figure7-4isa flowsheetdepictingthemajorprocessesthatcanbeexpectedina secondarycopper-smeltingoperation. Abrassandbronzealloying operationisshowninFigure7-5.

Pretreatmentofthefeedmaterialcanbeaccomplished usingseveraldifferentprocedures,eitherseparatelyorin combination. Feedscrapisconcentratedbymanualandmechanical methods,suchassorting,stripping,shredding,andmagnetic separation. Feedscrapissometimesbriquettedinahydraulic press. Pyrometallurgicalpretreatmentmayincludesweating, burningofinsulation(especiallyfromwirescrap),anddrying (burningoffoilandvolatiles)inrotarykilns. Hydrometallurgicalmethodsincludeflotationandleaching,with chemicalrecovery.

Insmelting,low-gradescrapismeltedinacupola furnace,producing"blackcopper"(70to80percentCu)andslag; theseareoftenseparatedinareverberatoryfurnace. Fromhere, themeltistransferredtoaconverterorelectricfurnaceto produce"blister"copper,whichis90to99percentCu. The actualtemperatureatwhichthesmeltingtakesplaceisnot

7-28

known. However,itisbelievedthattheoperatingtemperatures arenotsignificantlydifferentfromthatofprimarycopper- smeltingoperations. Thetemperatureinthecupolafurnaceis believedtoexceedtheboilingpointofcadmium,whichis approximately767°C(1413°F). Therefore,mostofthecadmium potentiallywillbeemittedfromthecupolafurnace.

Blistercoppermaybepouredtoproduceshotorcastings, butisoftenfurtherrefinedelectrolyticallyorbyfire refining. Thefire-refiningprocessisessentiallythesameas thatdescribedfortheprimarycopper-smeltingindustry. The sequenceofeventsinfirerefiningis: (1)charging, (2)meltinginanoxidizingatmosphere,(3)skimmingtheslag, (4)blowingwithairoroxygen,(5)addingfluxes,(6)"poling" orotherwiseprovidingareducingatmosphere,(7)reskimming,and (8)pouring.

Toproducebronzeorbrass,ratherthancopper,an alloyingoperationisrequired. Clean,selectedbronzeandbrass scrapischargedtoameltingfurnacewithalloystobringthe resultingmixturetothedesiredfinalcomposition. Fluxesare addedtoremoveimpuritiesandtoprotectthemeltagainst oxidationbyair. Airoroxygenmaybeblownthroughthemeltto adjustthecompositionbyoxidizingexcesszinc.

Withzinc-richfeed,suchasbrass,thezincoxide concentrationintheexhaustgasissometimeshighenoughtomake recoveryforitsmetalvaluedesirable. Thisprocessis accomplishedbyvaporizingthezincfromthemeltathigh temperaturesandthencapturingtheoxidedownstreaminaprocess fabricfilter.

Thefinalstepisalwayscastingofthesuitablyalloyed orrefinedmetalintoadesiredform,i.e,shot,wirebar,anodes,

7-31 cathodes,ingots,orothercastshapes. Themetalfromthemelt isusuallypouredintoaladleorasmallpot(whichservesasa surgehopperandaflowregulator)thencontinuesintoamold.

EmissionControlMeasures13

Theprincipalpollutantsemittedfromsecondarycopper- smeltingactivitiesareparticulatematterinvariousforms. Removalofinsulationfromwirebyburningcausesparticulate emissionsofmetaloxidesandunburnedinsulation. Thedryingof chipsandboringstoremoveexcessoilsandcuttingfluidscan resultinlargeamountsofdensesmoke,consistingofsootand unburnedhydrocarbons,beingdischarged. Particulateemissions fromthetopofacupolafurnaceconsistofmetaloxidefumes, dirt,anddustfromlimestoneandcoke.

Thesmeltingprocessutilizeslargevolumesofairto oxidizesulfides,zinc,andotherundesirableconstituentsofthe feed. Thisproceduregeneratesconsiderableparticulatematter intheexitgasstream. Thewidevariationamongfurnacetypes, chargetypes,quality,extentofpretreatment,andsizeofthe chargeisreflectedinabroadspectrumofparticlesizesand variablegrainloadingsintheescapinggases. Onemajorfactor contributingtodifferencesinemissionratesistheamountof zincpresentinscrapfeedmaterials;thelow-boilingzinc evaporatesandcombineswithairoxygen,producingzincoxide fumes.

Metaloxidefumesfromfurnacesusedinsecondarysmelters havebeencontrolledbyfabricfilters,ESP’s,orwetscrubbers. Controlefficiencybyfabricfiltersmaybebetterthan 99percent,butcoolingsystemsareneededtopreventthehot exhaustgasesfromdamagingordestroyingthebagfilters. A two-stagesystemusingbothwaterjacketingandradiantcooling

7-32 iscommon. Electrostaticprecipitatorsarenotaswellsuitedto thisapplication,havingalow-collectionefficiencyfordense particulates,suchasoxidesofleadandzinc. Wetscrubber installationsalsoarerelativelyineffectiveinthesecondary copperindustry. Scrubbersareuseful,mainlyforparticles largerthan1micron,butthemetaloxidefumesgeneratedare generallysubmicroninsize.

Particulateemissionsassociatedwithdryingkilnscanbe similarlycontrolled. Dryingtemperaturesupto150°C(302°F) producerelativelycoolexhaustgases,requiringnoprecooling forcontrolbyfabricfilters.

Wireburninggenerateslargeamountsofparticulate matter,largelyunburnedcombustibles. Theseemissionscanbe effectivelycontrolledbydirect-flameafterburners,withan efficiencyof90percentorbetteriftheafterburnercombustion temperatureismaintainedabove1000°C(1832°F). Ifthe insulationcontainschlorinatedorganics,suchaspolyvinyl chloride,hydrogenchloridegaswillbegeneratedandwillnotbe controlledbytheafterburner.

Onesourceoffugitiveemissionsinsecondarysmelter operationsischargingofscrapintofurnacescontainingmolten metals. Thisoftenoccurswhenthescrapbeingprocessedisnot sufficientlycompactedtoallowafullchargetofitintothe furnacepriortoheating. Theintroductionofadditional materialontotheliquidmetalsurfaceproducessignificant amountsofvolatileandcombustiblematerialsandsmoke,which canescapethroughthechargingdoor. Briquettingthecharge offersawaytoavoidfractionalcharges. Whenfractional chargingcannotbeeliminated,fugitiveemissionsarereducedby turningoffthefurnaceburnersduringcharging. Thisreduces

7-33 theflowofexhaustgasesandenhancestheabilityoftheexhaust controlsystemtohandletheemissions.

Metaloxidefumesaregeneratednotonlyduringmelting, butalsoduringpouringofthemoltenmetalintothemolds. Otherdustsmaybegeneratedbythecharcoal,orotherlining, usedinassociationwiththemold. Coveringthemetalsurface withgroundcharcoalisamethodusedtomake"smooth-top" ingots. Thisprocesscreatesashowerorsparks,releasing emissionsintotheplantnearthefurnacetopandthemoldsbeing filled.

Emissions

Cadmiummaybeexpectedtobepresentinthescrapmetals thatareprocessedtorecoversecondarycopper. Therefore, cadmiumemissionscanbeexpectedfromsecondarycopper-smelting operations. Table7-11presentsestimatesofcadmiumemissions reportedbyfourfacilitiesduring1989and1990asrequired underSARATitleIIIregulations.4,14 However,notestdataare availablepertainingtocadmiumemissionsfromsecondarycopper- smeltingoperations. Theonlyavailableemissionfactordataare fromtheSPECIATEdatabase. Thesedatashowthecadmium emissionfactorforelectricinductionfurnaceatsecondary copper-smeltingfacilitiestobe0.0012kg/Mg(0.0024lb/ton)of materialchargedintothefurnace.15 Becausethevalidityof theseemissionfactorscannotbeverified,extremecautionshould beexercisedwhenusingthem.

7-34 TABLE7-11.SECONDARYCOPPERPRODUCERSREPORTINGCADMIUM EMISSIONSINTHE1989AND1990TOXICSRELEASEINVENTORY

Company Location Emissions,kg(lb) SouthwireCo. Carrolton,GA 57(125)a FranklinSmelting&Refining Philadelphia,PA 454(1,000)b GastonCopperRecycling Gaston,SC 4.5(10)b Industries Sauget,IL 680(1,496)b

SOURCE:References4and14

aTheemissionratereportedisfor1989.

bTheemissionratereportedisfor1990.

SECONDARYZINCRECOVERYFROMMETALLICSCRAP

Thesecondaryzincindustryprocessesobsoleteandscrap materialstorecoverzincasslabs,dust,andzincoxide.15 Table7-12presentsalistofU.S.facilitieswherezinc currentlyisrecoveredfrommetalscrap. Cadmiumcanbeexpected tobepresentinthemetallicscrapthatisprocessedtorecover zinc,andthereforebeemittedasanairpollutantduring differentprocessingsteps. Adescriptionoftheprocessusedto manufacturezincfrommetallicscrapispresentedbelow.

ProcessDescription15

Processinginvolvesthreeoperations: scrappretreatment, melting,andrefining. Processestypicallyusedineach operationareshowninFigure7-6. Moltenproductzincmaybe usedinzincgalvanizing.

ScrapPretreatment-- Pretreatmentisthepartialremovalofmetalandother contaminantsfromscrapcontainingzinc. Sweatingseparateszinc fromhigh-meltingmetalsandcontaminantsbymeltingthezincin

7-35 TABLE7-12.DOMESTICPRODUCERSOFSECONDARYZINC FROMMETALLICSCRAP

1990Productioncapacity,Mg Smelter Location (tons)

ArcoAlloys,Inc. Detroit,MI Seefootnotea

W.J.Bullock,Inc. Fairfield,AL Seefootnotea

T.L.Diamond&Co.,Inc. Spelter,WV Seefootnotea

FloridaSteelCo. Jackson,TN Seefootnotea

GulfReductionCorp. Houston,TX Seefootnotea

HugoNeu-ProlerCo. TerminalIsland,CA Seefootnotea

HuronValleySteelCorp. Belleville,MI Seefootnotea

IndianaSteel&WireCo.,Inc. Muncie,IN Seefootnotea

InteramericanZincInc. Adrian,MI Seefootnotea

NewEnglandSmeltingWorks, WestSpringfield,MA Seefootnotea Inc.

NucorYamatoSteelCo. Blytheville,AR Seefootnotea

TheRiverSmelting&RFGCo. Cleveland,OH Seefootnotea

ZincCorp.ofAmerica Palmerton,PA Seefootnotea

Source:Reference11. aThetotalzincproductioncapacityforall13plantsis58,000Mg(63,800tons). Individualcapacitydataarenotavailable.

7-36 kettle,rotary,reverberatory,muffleorelectricresistance furnaces. Usually,theproductzincthenisdirectlyusedin melting,refiningoralloyingprocesses. Thehigh-melting residueisperiodicallyrakedfromthefurnaceandfurther processedtorecoverzinc. Theseresiduesmaybeprocessedby crushing/screeningtorecoverimpurezincorbysodiumcarbonate leachingtoproducezincoxide. Thetemperatureatwhichthe pretreatmenttakesplacemaybehighlyvariable,dependingonthe typeofscrap. Itisbelievedthatthetemperatureatwhichthe pretreatmenttakesplacemaybecomparabletothatofprimary zincoperations. Therefore,thepretreatmentstepmayresultin cadmiumemissions.

Incrushing/screening,zinc-bearingresiduesare pulverizedorcrushedtobreakthephysicalbondsbetween metalliczincandcontaminants. Theimpurezincisthen separatedinascreeningorpneumaticclassificationstep.

Insodiumcarbonateleaching,thezinc-bearingresidues areconvertedtozincoxide,whichcanbereducedtozincmetal. Theyarecrushedandwashedtoleachoutzincfromcontaminants. Theaqueousstreamisthentreatedwithsodiumcarbonate, precipitatingzincasthehydroxideorcarbonate. The precipitateisthendriedandcalcinedtoconvertzinchydroxide intocrudezincoxide. TheZnOproductisusuallyrefinedto zincatprimaryzincsmelters.

Melting-- Zincismeltedat425-590°C(797-1094°F)inkettle, crucible,reverberatory,andelectricinductionfurnaces. Zinc tobemeltedmaybeintheformofingots,rejectcastings, flashingorscrap. Ingots,rejects,andheavyscrapare generallymeltedfirsttoprovideamoltenbathtowhichlight scrapandflashingareadded. Beforepouring,afluxisadded

7-38 andthebatchagitatedtoseparatethedrossthataccumulates duringthemeltingoperation. Thefluxfloatsthedrossand conditionsitsoitcanbeskimmedfromthesurface. After skimming,themeltcanbepouredintomoldsorladles. Any residualcadmiumleftoverfromthepretreatmentstepmaynotbe emittedduringthemeltingstagebecausethemeltingtakesplace atamuchlowertemperaturethantheboilingpointofcadmium.

Refining/Alloying-- Additionalprocessingstepsmayinvolvealloying, distillation,distillationandoxidation,orreduction. Alloying producesmainlyzincalloysfrompretreatedscrap. Oftenthe alloyingoperationiscombinedwithsweatingormelting.

Distillationretortsandfurnacesareusedtoreclaimzinc fromalloysortorefinecrudezinc. Retortdistillationisthe vaporizationat980-1250°C(1796-2282°F)ofelementalzincwith itssubsequentcondensationaszincdustorliquidzinc. Rapid coolingofthevaporstreambelowthezincmeltingpointproduces zincdust,whichcanberemovedfromthecondenserandpackaged. Ifslabzincisthedesiredproduct,thevaporsarecondensed slowlyatahighertemperature. Theresultantmeltiscastinto ingotsorslabs. Muffledistillationfurnacesproduce principallyzincingots,andgraphiterodresistancedistillation produceszincdust. Becausethedistillationtakesplaceat temperatureshigherthantheboilingpointofcadmium(767°C [1413°F]),cadmiummaypotentiallybeemittedduringthe distillationstep.

Retortandmufflefurnacedistillationandoxidation processesproducezincoxidedust. Theseprocessesaresimilar todistillationthroughthevaporizationstep. Incontrast,for distillation/oxidation,thecondenserisomitted,andthezinc vaporisdischargeddirectlyintoanairstreamleadingtoa

7-39 refractory-linedcombustionchamber. Excessairisaddedto completeoxidationandtocooltheproduct. Thezincoxide productisusuallycollectedinafabricfilter.

Inretortreduction,zincmetalisproducedbythe reactionofcarbonmonoxideandzincoxidetoyieldzincand carbondioxide. Carbonmonoxideissuppliedbythepartial oxidationofthecoke. Thezincisrecoveredbycondensation.

ZincGalvanizing-- Zincgalvanizingisthecoatingofcleanoxidefreeiron orsteelwithathinlayerofzincbyimmersioninmoltenzinc. Thegalvanizingoccursinavatorindiptankscontainingmolten zincandcoverflux.

EmissionControlMeasures15

Emissionsfromseatingandmeltingoperationsconsist principallyofparticulates,zincfumes,othervolatilemetals, fluxfumes,andsmokegeneratedbytheincompletecombustionof grease,rubberandplasticsinthezinc-bearingfeedmaterial. Zincfumesarenegligibleatlowfurnacetemperatures,forthey havealowvaporpressureevenat480°C(896°F). Withelevated temperatures,however,heavyfumingcanresult. Fluxemissions areminimizedbytheuseofanonfumingflux. Substantial emissionsmayarisefromincompletecombustionofcarbonaceous materialinthezincscrap. Thesecontaminantsareusually controlledbyafterburners. Furtheremissionsaretheproducts ofcombustionofthefurnacefuel. Sincethefurnacefuelis usuallynaturalgas,theseemissionsareminor. Inreverberatory furnaces,theproductsoffuelcombustionareemittedwiththe otheremissions. Otherfurnacesemitthefuelcombustion productsasaseparateemissionstream.

7-40 Particulatesfromsweatingandmeltingaremainlyhydrated

zincchloride(ZnCl2)andZnO,withsmallamountsofcarbonaceous material. TheseparticulatesalsocontainCu,Cd,maganese(Mn), andchromium(Cr).

Fabricfiltersaremostcommonlyusedtorecover particulateemissionsfromsweatingandmelting. Inone applicationonamuffle-sweatingfurnace,acycloneandfabric filterachievedparticulaterecoveryefficienciesinexcessof 99.7percent. Inanotherapplicationonareverberatorysweating furnace,afabricfilterremoved96.3percentofthe particulates,reducingthedustloadingfrom0.513g/Nm3 to 0.02g/Nm3. Fabricfiltersshowsimilarefficienciesinremoving particulatesfromexhaustgasesofmeltingfurnaces.

Crushingandscreeningoperationsalsoaresourcesofdust emissions. TheseparticulatesarecomposedofZn,Al,Cu,Fe, lead(Pb),Cd,tin(Sn),andCr,andtheycanberecoveredfrom hoodedexhaustsbyfabricfilters.

Thesodiumcarbonateleachingprocessproducesparticulate emissionsofZnOdustduringthecalciningoperation. Thisdust

canberecoveredinfabricfilters,althoughZnCl2 inthedust maycausepluggingproblems.

Emissionsfromrefiningoperationsaremainlymetallic fumes. Thesefumeanddustparticlesarequitesmall,withsizes rangingfrom0.05-1µ. Distillation/oxidationoperationsemit theirentireZnOproductintheexhaustgas. TheZnOhasavery smallparticlesize(0.03to0.5µ)andisrecoveredinfabric filterswithtypicalcollectionefficienciesof98to99percent.

7-41 Someemissionsofzincoxideoccurduringgalvanizing,but theseemissionsaresmallbecauseofthebathfluxcoverandthe relativelylowtemperaturemaintainedinthebath.

Emissions

Cadmiummaybeexpectedtobepresentinthescrapmetals thatareprocessedtorecoverzinc. Therefore,cadmiumemissions canbeexpectedfromsecondaryzincrecoveryoperations. One secondarysmelterreportedacadmiumemissionrateof588lb/yr duringtheyear1989.14 However,notestdataareavailable pertainingtocadmiumemissionsfromsecondaryzincrecovery operations. Theonlyavailableemissionfactordataarefromthe SPECIATEdatabase. Table7-13presentscadmiumemissionfactors forvariousemissionpointsatsecondaryzinc-smeltingfacilities asreportedintheSPECIATEdatabase.5 Becausethevalidityof theseemissionfactorscannotbeverified,extremecautionshould beexercisedwhenusingthem.

SECONDARYZINCRECOVERYFROMSTEELPRODUCTION

Zincalsoisrecoveredfromelectricarcfurnace(EAF) dustgeneratedatironandsteelmanufacturingfacilities. A studycarriedoutinMay1985bytheCenterforMetalsProduction estimatedthatapproximately14.5percentofEAFdustis processedforzincrecovery. Onedatasourcehasreportedthat cadmiumcontentintheEAFdustisapproximately0.05percent. Thus,cadmiumcanpotentiallybeemittedduringtheprocessof recoveringzincfromEAFdust.16

Table7-14containsalistoffacilitiesintheU.S.that arecapableofproducingEAFdust.11 Ofthesefacilities,only twoactuallyrecoverzinc.17 Adescriptionoftheprocessesused torecoverzincfromEAFdustispresentedbelow.

7-42 TABLE7-13.CADMIUMEMISSIONSFROMSECONDARYZINCRECOVERY FROMMETALSCRAP

Cadmium Source emission factora classification Emission source code (SCC) lb/ton kg/Mg

Retort furnace 30400801 0.03619b 0.018095 Horizontal 30400802 b 0.017325 Pot furnace 30400803 0.03465 0.00004 Galvanizing kettle 30400805 0.00008b 0.001925 Calcining kiln 30400806 0.034265 0.00385c 0.06853b

Rotary-sweat furnace 30400809 0.01386b 0.00693 Muffle-sweat furnace 30400810 b 0.00824 Electric resistance sweat furnace 30400811 0.01648 0.00385 Crushing/screening of zinc residues 30400812 0.0077b 0.001635 Kettle-sweat furnace (general metallic scrap) 30400824 0.004235 0.00327d 0.00847b

Reverberatory sweat furnace (general metallic scrap) 30400828 0.01001b 0.005005 Kettle-sweat furnace (general metallic scrap) 30400834 b 0.009625 Reverberatory sweat furnace (general metallic scrap) 30400838 0.01925 0.01232 Retort and muffle distillation: Pouring 30400851 0.02464b 0.00023 Retort and muffle distillation: Casting 30400852 0.000115 0.00046b 0.00023b

Retort distillation/oxidation 30400854 0.0231e 0.01155 Muffle distillation/oxidation 30400855 e 0.01155 Rotary sweating 30400862 0.0231 0.000345 Muffle sweating 30400863 0.00069b 0.00041 Kettle (pot) sweating 30400864 0.000215 0.00082b 0.00043c

Electric resistance sweating 30400865 0.00039f 0.000195 Retort and muffle distillation 30400872 b 0.00091 Casting 30400873 0.00182 0.000005 0.00001b

Source: Reference 5. aAll emission factors are reported as found in the SPECIATE data base without rounding off (Reference 5). Emission factors in SPECIATE data base are reported in lb/ton of process activity. It appears that the emission factors reported above are uncontrolled factors. blb/ton (kg/Mg) of zinc produced. clb/ton (kg/Mg) of zinc used. dlb/ton (kg/Mg) of residue. elb/ton (kg/Mg) of produced. flb/ton (kg/Mg) of scrap processed.

7-43 TABLE7-14.DOMESTICPRODUCERSOFSECONDARYZINC FROMEAFDUST

Zincrecovering EAFprocessing capacity,Mg(tons)a Company Location capacity,Mg(tons)

FloridaSteelCo. Jackson,TN 7,200(7,920) 1,400(1,540)

Horsehead CalumetCity,IL 72,000(79,200) 75,000(82,500)b Development ResourceCo.,Inc.

Horsehead Monaca,PA 18,000(19,800) Seefootnoteb Development ResourceCo.,Inc.

Horsehead Palmerton,PA 245,000(269,500) Seefootnoteb Development ResourceCo.,Inc.

Horsehead Rockwood,TN 90,000(99,000) Seefootnoteb Development ResourceCo.,Inc.

LacledeSteelCo. St.Louis,MO 36,000(39,600) 6,000(6,600)

NorthStarSteelCorp. Beaumont,TX 27,000(29,700) 5,000(5,500)

Nucor-YamamotoSteel Blytheville,AR 11,000(12,100) 1,800(1,980) Co.

ZiaTechnologyof Caldwell,TX 27,000(29,700) 4,500(4,950) TexasInc.

Source: Reference11. aEventhoughthereareninefacilities,whichhavethecapabilitytoprocessEAF,onlytwofacilities, FloridaSteelandNucor-YamamotoSteelCo.,actuallyrecoverzincfromEAFdust.17 bThecombinedzinccapacityforallfourlocationsofHorseheadDevelopmentCo.,is75,000Mg (82,500tons).Datapertainingtoindividualcapacitiesarenotavailable.

7-44 ProcessDescription16,18

TheprocessofrecoveringzincfromEAFdustiscarried outintwosteps. Inthefirststep,nonferrousingredientsare volatilizedfromtheEAFdust. Thesecondstepconsistsof processingthevolatilizednonferrousingredientsinarotary furnace(calciningkiln)toproduceazincoxidecalcine.

Twoprocessesareavailabletocarryoutthefirststepof volatilizingnonferrouscomponentsfromEAFdust. Thesearethe Waelzkilnandflamereactorprocesses.

Waelzkilnprocess-- Figure7-7presentsatypicalprocessflowdiagramforthe Waelzkilnprocess. Inthisprocess,EAFdustisfedalongwith anthracitecoalandsilicafluxesintoarotarykiln. Thekiln containstworeactionzones: thesolidmaterialchargezoneand thegaseouszoneabovethesolidzone. Coalinthesolidzone combuststoformcarbonmonoxide(CO)whichreducesthemetal oxides,thusvolatilizingthemetals. Nonferrousmetals volatilizewhenthetemperatureinthekilnreaches1000°C (1832°F). Whenthenonferrousmetalsreachthegaseouszoneupon volatilization,theyareagainoxidizedtoformcorresponding metaloxides. Theaboveprocessesaredescribedbythefollowing chemicalreactionsthattakeplaceinthekiln:

Combustionreactions

______C+½O2 >CO ______C+CO2 >2CO

7-45

Reductionreactions ______Fe3O4 +CO >3FeO+CO2 ______ZnO+CO >Zn+CO2 ______CdO+CO >Cd+CO2 Oxidationreactions ______CO+½O2 >CO2 ______Zn+½O2 >ZnO ______Cd+½O2 >CdO ZnOCalcineFormation-- ThenonferrousmetaloxideproductformedbyZnOcalcine formationiscapturedinafabricfilterandsubsequently processedinanaturalgas-firedcalciningkilntoformzinc calcine. Figure7-7alsopresentsatypicalprocessflowdiagram forthecalciningkilnoperation. Inthecalciningkiln,the nonferrousmetaloxidesareintroducedwithoutanyadditives,and selectivevolatilizationiscarriedouttoseparatecadmium, lead,,andfluorine. Thevolatilizedmaterialis collectedinafabricfilter. Zincoriginallypresentinthe formofZnOisrecoveredunalteredasresiduefromthecalcining kiln. TheZnOcansubsequentlybeprocessedbyelectrothermic processtorecoverzinc.

FlameReactorProcess-- Figure7-8presentsaflowdiagramfortheflamereactor processusedtovolatilizenonferrousmetalsfromdust. This processisanalternativemethodfortheWaelzkilnprocess. In theflamereactorprocess,EAFdustisfedintoawater-cooled, naturalgas-firedreactor. Thecombustionairisenrichedwith oxygenraisingtheoxygencontenttoalevelbetween40and 80percent. Flametemperaturesupto2200°C(3992°F)canbe obtainedthisway. AstheEAFdustisintroducedintothehot

7-47 gasesandtemperaturesreach1600°C(2912°F),refractory compoundspresentinthedustfusetoformmoltenslag. The moltenslagandthecombustiongasesareconveyedintoaslag convertorwhereiron-richslagiscontinuouslytapped. The offgasesexitingtheseparatorareoxidizedfurthertoform metallic(nonferrous)oxides,whicharerecoveredinafabric filterandshippedasfeedstocktozinc-smeltingfacilities.

EmissionControlMeasures16,18

EmissionsresultingfromprocessingofEAFdustchiefly consistofparticulatemattermadeupofnonferrousmetallic oxides(includingcadmiumoxide),whicharerecoveredinafabric filter. Thefabricfilterisusedmoreasaproductrecovery devicethanacontroldevice.

Emissions18

DatapertainingtocadmiumemissionsfromEAFdust processingarelimited. Onesourcehasreportedacadmium emissionrateof5.04x10-4 lb/hrmeasuredattheoutletofa fabricfilterservicingaflamereactor,whichcorrespondstoa EAFdustprocessingrateof40lb/min(1.2tons/hr). This resultsinacadmiumemissionfactorof2.1x10-4 kg/Mg (4.2x10-4 lb/ton)ofEAFdustprocessed. Nootherdetails pertainingtothetestareavailable. Therefore,thevalidityof thiscadmiumemissionfactorcannotbeverified. Thecadmium emissionfactormustbeusedwithextremecaution.

7-49 REFERENCES 1. EmissionStandardsDivision. CadmiumEmissionsfrom PrimaryLeadandPrimaryCopperSmelting--PhaseI TechnicalReport. EPA-450/3-88-006. OfficeofAir QualityPlanningandStandards,U.S.Environmental ProtectionAgency,ResearchTrianglePark,NC. June1988. p.5. 2. Woodbury,W.D.Lead--AnnualReport1990. BureauofMines, U.S.DepartmentoftheInterior,U.S.GovernmentPrinting Office,Washington,DC. April1992. 3. CompilationofAirPollutionEmissionFactors,VolumeI: StationaryPointandAreaSources(AP-42),FourthEdition. U.S.EnvironmentalProtectionAgency. ResearchTriangle Park,NC. Pp.7.6-1-7.6-15. 4. U.S.EnvironmentalProtectionAgency. 1990Toxic ChemicalReleaseInventory. OfficeofToxicSubstances, Washington,DC.,January1993. 5. SPECIATE. VolatileOrganicCompound(VOC)/Particulate Matter(PM)SpeciationDataSystem,Version1.4. Office ofAirandRadiation. OfficeofAirQualityPlanningand Standards,U.S.EnvironmentalProtectionAgency,Research TrianglePark,NC. October1991. 6. Reference1,p.9. 7. TelefaxfromJolly,J. BureauofMines,U.S.Department oftheInterior,Washington,DC. January23,1993. CapacitiesofU.S.coppersmelters. 8. Reference5,Pp.7.3-1-7.3-18. 9. TRCEnvironmentalCorporation. EmissionCharacterization Program. PreparedforCopperRangeCompany,WhitePine, Michigan. October15,1992. 10. Reference3,Pp.7.7-1-7.7-8. 11. TelefaxfromJolly,J. BureauofMines,U.S.Department oftheInterior,Washington,DC. February10,1993. CapacitiesofU.S.zincsmelters. 12. Zinc/ZincOxidePreliminarySourceAssessment. EPA-450/3-87-008. OfficeofAirQualityPlanningand Standards,U.S.EnvironmentalProtectionAgency,Research TrianglePark,NC. April1987. Pp.1-10-1-11.

7-50 13. Reference3,Pp.7.9-1-7.9-6. 14. U.S.EnvironmentalProtectionAgency. 1989Toxic ChemicalReleaseInventory. OfficeofToxicSubstances, Washington,DC. January1993. 15. Reference3.Pp.7.14-1-7.14-8. 16. James,S.E.,andC.O.Bounds. RecyclingLeadand Cadmium,AsWellAsZinc,fromEAFDust. (In)Lead-Zinc, ’90,Proc.WorldSymp.Metall.Environ.Control119th TMS Annu.Meet. T.S.MackeyandR.D.Prengaman,eds. Miner.Met.Mater.Soc.,Warrendale,PA. 1990. Pp.477-495. 17. Telecon. Kalagnanam,R.S.,MidwestResearchInstitute, withJ.H.Jolly.BureauofMines,U.S.Departmentofthe Interior.February10,1993. U.S.processorsofEAFdust. 18. Bounds,C.O.,andJ.F.Pusateri. EAFDustProcessingin theGas-FiredFlameReactorProcess. (In)Lead-Zinc,’90, Proc.WorldSymp.Metall.Environ.Control119th TMSAnnu. Meet. T.S.MackeyandR.D.Prengaman,eds. Miner.Met. Mater.Soc.,Warrendale,PA. 1990. Pp.511-528.

7-51 SECTION8

EMISSIONSFROMMISCELLANEOUSSOURCES

Cadmiumispresentinminerals,ores,andcrudesextracted fromtheearthandthesematerialsareusedinseveral manufacturingprocesses. Thesemanufacturingprocessescanbe potentialcadmiumemissionsourcesiftheyusecadmiumcontaining materialsinthermaltreatmentsteps. Themanufacturing processesdescribedinthissectionare:ironandsteel production,Portlandcementmanufacture,phosphaterock processing,carbonblackproduction,andmobilesources. This sectionpresentsprocessinformationandemissioncontrol measures,andestimatesofcadmiumemissionsfromeachofthese sources.

IRONANDSTEELPRODUCTION1

Twotypesofironandsteelplantswillbediscussedin thissection: integratedandnonintegrated. Becausecadmiumcan bepresentasatracecontaminantinprocessfeedmaterialssuch ascoal,ironore,andscrapmetal,processoperationsinboth planttypesarepotentialcadmiumemissionsources.

Integratedironandsteelplantsarethoseironand steelmakingfacilitiesthatarecapableofstartingwithironore asarawmaterialfeedandproducingfinishedsteelproducts. At aminimum,thesefacilitieshaveblastfurnacefacilitiesforpig ironproduction;steelmakingfurnaces(generallyoneormore basicoxygenfurnaces),andsteelfinishingoperations. Many

8-1 facilitiesalsohavecokemakingoperations,sinterplants,and electricarcfurnaceshopsformeltingscrap. Initssimplest form,theintegratedironandsteelprocessbeginswithpigiron productionfromironoreorpelletsintheblastfurnace. The moltenironistransferredfromtheblastfurnacetothebasic oxygenfurnace,wherethehotpigironandscrapmetalareheated andtransformedmetallurgicallytocarbonsteel. Thiscarbon steelisthencastandrolledintoafinalproject. Table8-1 providesalistingofintegratedironandsteelplantlocations.

Nonintegratedplantsconsistof"minimills"orspecialty millsthatproducecarbonsteel,stainlesssteelandothersteel alloysfromscrap. Typicaloperationsatthesefacilities includeelectricarcfurnacesforsteelmakingandsteelcasting andfinishingoperations,aswellasalloyingoperations. TableC-1liststhosefacilitiesthatuseelectricarcfurnaces.

Totalsteel(carbonandalloy)productionfor1991was 79.7x106 Megagrams(Mg)(87.8x106 tons). Ofthistotal, 70.7x106 Mg(77.9x106 tons)wascarbonsteel. In1991, nonintegratedplantsproduced9.1x106 Mg(10.0x106 tons)of stainlessandalloysteel.1 Inordertodistributethis productionamongfurnacetypes,basicoxygenandopenhearth furnaceswereassumedtoproduceonlycarbonsteel. Usingthis assumptionestimated1991productionlevelsforcarbonsteelwere 47.8x106 Mg(52.7x106 tons)inbasicoxygenfurnaces, 1.3x106 Mg(1.4x106 tons)inopenhearthfurnaces,and 16.7x106 Mg(18.4x106 tons)inelectricarcfurnaces. The totalstainlessandalloysteelproductionof9.1x106 Mg (10.0x106 tons/yr)isassumedtobefromelectricarcfurnaces. Intermediatesproducedbyintegratedironandsteelplantsin 1991include44.1x106 Mg(48.6x106 tons)ofpigironand 21.8x106 Mg(24.0x106 tons)ofcoke(includingbothfurnace andmerchantcokeplantsbutexcludingcokebreeze).

8-2 TABLE8-1.INTEGRATEDIRONANDSTEELPLANTS

No. of basic No. of oxygen blast Company Facility City and State furnaces furnaces

Acme Steel Company Acme Steel Company Chicago Plant Chicago, IL 2 2 Allegheny Ludlum Corp. Riverdale Works Riverdale, IL 2 1 Armco Steel Company, L.P.a Brankenridge Works Natrona, PA 2 1 a Ashland Works Ashland, KY 2 2 Armco Steel Company, L.P. Hamilton Plant Hamilton, OH - 2

Armco Steel Company, L.P. Bethlehem Steel Corp. a Middletown Works Middletown, OH 2 2 Bethlehem Steel Corp. Bethlehem Plant Bethlehem, PA 2 4 Bethlehem Steel Corp.a Burns Harbor Plant Burns Harbor, IN 3 2 a Sparrows Pt. Plant Sparrows Pt., MD 2 3 Geneva Steel Company Geneva Works Orem, UT 1 3

Gulf States Steela Inland Steel Companyb Gadsen, AL 2 1 LTV Steel Flat Rolled & Bar Co. Gadsen Works East Chicago, IN 4 9 a Cleveland Works Cleveland, OH 4 5 LTV Steel Flat Rolled & Bar Co. Indiana Harbor Works East Chicago, IN 2 2 LTV Steel LTV Stills Chicago, IL 1 1

McLouth Steel Trenton Works Trenton, MI 5 2 National Steel Corporation Granite City Steel Granite City, IL 2 2 National Steel Corporation Great Lakes Steel Ecorse, MI 2 3 Rouge Steel Company Rouge Works Dearborn, MI 2 2 Sharon Steel, Incorporated Farrell Works Farrell, PA 2 1

Sharon Steel, Incorporated U.S. Steel Corporation (USX) Monessen, Inc. Sharon, PA - 1 U.S. Steel Corporation (USX) Edgar Thomson Plant Braddock, PA 2 2 U.S. Steel Corporation (USX) Fairfield Works Fairfield, AL 3 1 a Fairless Works Fairless Hills, PA 1 2 U.S. Steel Corporation (USX) Gary Works Gary, IN 6 5

USS/Kobe Steel Company Warren Consolidated Inc. Lorain, OH 2 4 Weirton Steel Companya Warren Facility Warren, OH 2 1 Wheeling-Pittsburgh Steel Corp. Weirton Works Weirton, WV 2 4 c Steubenville Plant, N Steubenville, OH 2 2 Wheeling-Pittsburgh Steel Corp. Steubenville Plant, S Mingo Junction, OH 1 2

TOTAL 65 74

Source: Reference 2. aAlso has a sinter plant at this location.

bHas a sinter plant at the Indiana Harbor Works facility.

cHas a sinter plant at the Follansbee Plant, Follansbee, WV.

8-3 ProcessDescription

Figure8-1providesaflowdiagramforironandsteel production. Theprincipalcomponentsoftheprocessareiron production,steelmaking,andsteelfinishing. However,two importantancillarycomponentsarecokemakingandsinter production. Theprocessstepsdiscussedbelowapplytoan integratedplant. Processdifferenceswillbenotedfor nonintegratedplants.

Frequently,thefirststepintheprocessforan integratedplantistoproducemetallurgicalcoke(elemental carbon)fortheblastfurnace. Cokeisusedto: (1)providea substrateforrawmaterialsintheblastfurnace,(2)functionas fuelforthehotblastair,and(3)removeironoreoxides. Table8-2liststhecokeproductioncapacitiesforbatteries locatedatorassociatedcircleintegratedplants. Nonintegrated plantsdonotuseblastfurnacesand,therefore,donotneed coke. Thecokeismadefromcoalthatispulverizedandthen heatedinacokeovenwithoutoxygenat1050°C(1925°F)for12to 20hours. Volatilesaredrivenoff,andelementalcarbon(coke) andashareformed. Becausecadmiumisatracecontaminantin coal,thereisapotentialforemissionswhenthecoalisheated.

Twotypesofovens(arrangedinbatteries)canbeused: a slotovenandanonrecoveryoven. Theslotovenprocessrecovers volatilesthataredrivenoffduringtheheatingprocess,and thesevolatilesarerefinedtoproducecoke-ovengas,tar, sulfur,ammoniumsulfateandlightoil. Becausemostofthe volatilematerialsgeneratedbytheovenarecycledthrough recoveryprocesswithorganiccondensationsteps,cadmium emissionsgeneratedbyby-productcokemakingfacilitiesis

8-4

TABLE8-2.COKEPRODUCTIONCAPACITYFORINTEGRATEDIRONANDSTEELFACILITIES INTHEUNITEDSTATESIN1991

Totalcapacity

Facility No.of TotalNo.of Megagrams Tonsper batteries ovens perday day

AcmeSteel,Chicago,IL 2 100 1,626 1,600

Armco,Inc.,Ashland,KY 2 146 2,743 2,700

Armco,Inc.,Middleton,OH 3 203 4,608 4,535

BethlehemSteel,Bethlehem,PA 3 284 4,007 3,944

BethlehemSteel,BurnsHarbor,IN 2 164 4,400 4,380

BethlehemSteel,Lackawanna,NY 2 152 1,902 1,872

BethlehemSteel,SparrowsPoint,MD 3 210 4,134 4,069

GenevaSteel,Orem,UT 1 208 2,323 2,250

GulfStatesSteel,Gadsden,AL 2 130 2,845 2,800

InlandSteel,EastChicago,IN 6 446 5,868 5,775

LTVSteel,Pittsburgh,PA 5 315 5,491 5,404

LTVSteel,Chicago,IL 1 60 1,626 1,600

LTVSteel,Cleveland,OH 2 126 3,251 3,200

LTVSteel,Warren,OH 1 85 1,524 1,500

NationalSteel,GraniteCity,IL 2 90 1,544 1,520

NationalSteel,Ecorse,MI 1 78 940 925

USS,Div.ofUSXCorp.,Clairton,PA 12 816 12,843 12,640

USS,Div.ofUSXCorp.,Gary,IN 6 422 7,249 7,135

Wheeling-PittsburghSteel,EastSteubenville, WV 4 224 3,861 3,800

Total 60 4,259 72,835 71,649

SOURCE:Reference5.

8-6 expectedtobenegligible. However,nonrecoveryovensdonot recovervolatilesandinsteadcombustthem. Theseovensmaybea sourceofcadmiumemissions.

Afterthecokingcycleiscomplete,thecokeisquickly cooleddownbyaquenchingtower,topreventproductlossvia combustion. Furthercoolingoccursafterthewaterisdrained fromthecoke,itissizedtoremovetheundersizedmaterial (cokebreeze)andtransferredtostoragepiles. Becausecadmium mayremainasatracecontaminantinthecoke,smallquantities ofcadmiummaybereleasedasfugitiveemissionsduringthese handling,transfer,andstorageoperations.

Asecondancillaryprocessfoundatmanyintegratedplants isthesinteringoperation. Thesinteringprocessisa materials-recoveryprocess,whichconvertsfine-sizedraw materials,includingironore,cokebreeze,limestone,mill scale,andfluedust,intoanagglomeratedproductcalled "sinter."4 Cadmiumcanbecontainedintheironoreusedto producesinterandmaybeemittedwhenthesintermixtureis combusted. Therawmaterialsareplacedonacontinuous, travellinggratecalledthesinterstrand. Aburnerhood,atthe beginningofthesinterstrand,ignitesthecokeinthemixture. Combustionairisdrawndownthroughthematerialbedandintoa commonductleadingtoagascleaningdevice. Thefusedsinter isdischargedfromthesinterstrandwhereitiscrushedand screened. Undersizedsinterisrecycledtothemixingmilland backtothestrand. Theremainingsinterproductiscooledin theopenairorinacircularcoolerwithwaterspraysor mechanicalfans.

Theinitialprocesscommontoallintegratedplantsisthe blastfurnace,whichisusedtoproducemolteniron("pigiron"). Ironore,coke,limestonefluxandsinterareintroduced

8-7 ("charged")intothetopofthefurnace. Heatedairisinjected throughthebottomofthefurnace. Thisblastaircombuststhe cokecontainedinthebreezetomeltthesinter,andfluxwith theironoxidesintheoreandformmolteniron,slag,andcarbon monoxide(CO). Themoltenironandtheslagcollectinthe hearthatthebaseofthefurnaceandareperiodicallytapped. TheCOiscollectedthroughofftakesatthetopofthefurnace. Becausetheironorecontainstracesofcadmium,emissionsare possibleasparticulatematter(PM)entrainedintheCO. However,thisCOwillbeusedasfuelwithintheplantafterit iscleanedofPM;consequently,cadmiumemissionsareexpectedto benegligible.

Themoltenironfromtheblastfurnaceundergoes desulfurization,afterwhichitisintroducedtoabasicoxygen furnace(BOF)oropenhearthfurnacetomakesteel. These furnacesuseoxygenasarefiningagent. InaBOF,theraw materialistypically70-percentmoltenmetaland30-percent scrapmetal. Again,cadmiummaybepresentasacontaminantin thescrapmetal. Theoxygenreactswithcarbonandother impuritiesintherawmaterialandremovesthemfromthemolten metal. Largequantitiesofcarbonmonoxide(CO)areproducedby theexothermicreactionsintheBOF. Thesegases,whichmay containcadmium,frequentlyareburnedatthemouthofthe furnacetooxidizeCO. Thegasesarethenventedtogascleaning devicesbeforebeingdischargedtotheatmosphere.

TherearetwotypesofBOF’s. ConventionalBOF’sblow oxygenintothetopofthefurnacethroughawater-cooledlance. ThenewerQuelleBasicOxygenprocess(Q-BOP)furnacesinject oxygenthroughtuyeres,whicharelocatedinthebottomofthe furnace.Typicalcyclesinvolvescrapcharge,hotmetalcharge, oxygenblow(refining),testingfortemperatureand chemicalcompositionofthesteel,alloyadditionsandreblows

8-8 (ifnecessary),tapping,andslagging. Thefullfurnacecycle typicallyrangesfrom25to45minutes.

Theopen-hearthfurnace(OHF)isashallow, refractory-linedbasininwhichscrapandmoltenironaremelted togetherandthenrefinedintosteel. Themixtureofscrapand hotmetalmayconsistofeitherallscraporallhotmetal,buta half-and-halfmixtureofscrapandmetaliscommon. Because scrapisused,thereisapotentialforcadmiumemissions; however,noinformationwasavailabletoevaluateemissions potential. Becausethesefurnaceshavelimiteduseandarebeing eliminated,theyarenotconsideredasignificantcadmium emissionsources.

Nonintegratedplantsuseelectricalarcfurnaces(EAF’s) toproducecarbonandalloysteels. TherawmaterialforanEAF istypically100percentscrap,whichisasourceforcadmium emissions. Thesefurnacesarecylindricalandrefractory-lined. Theyhavecarbonelectrodesthatareraisedorloweredthrough thefurnaceroof. Withelectrodesraised,thefurnaceroofcan bemovedasidetoallowscrapsteeltobeintroducedbyan overheadcrane. Alloyingagentsandfluxingmaterialsare usuallyaddedthroughthedoorsonthesideofthefurnace. Heat generatedbytheelectricalcurrentpassingbetweenthe electrodesisusedtomeltthescrap. Aftermeltingandrefining periods,theslagandsteelarepouredfromthefurnaceby tilting. TheproductionofsteelinanEAFisabatchprocess. Cycles,or"heats,"rangefromabout11/2to5hoursforcarbon steelproductionandfrom5to10hours,ormore,foralloysteel production. Becausecadmiummaybecontainedinthescrapmetal usedtofeedEAF’s,theyareconsideredtobeapotentialsource ofcadmiumemissions.

8-9 AfterproductionineitheraBOF,OHF,orEAF,themolten steeliscastintomoldsoriscontinuouslycasttoforma finishedproduct. Thisfinalproductconsistsofshapescalled blooms,slabs,andbillets. Nocadmiumemissionsareexpected fromthisfinalprocessstepfromthecoldsteel,becausecadmium isnotexpectedtobepresentinthemoltensteelproduct.

EmissionControlMeasures2

Cadmiumisusuallyemittedasfineparticulatematterfrom hightemperatureoperationssuchasthefurnacesfoundinthe ironandsteelplants.6 Cadmiumemissionsfromironandsteel productionareexpectedfromfurnaces,BOF’s,OHF’s,EAF’s, sinteringoperations,andpossibly,chargingofby-productovens andnonrecoverycokeovenoperations. Nospecificdatawere availableoncadmiumcontrolfromironandsteeloperations. However,becausetheseprocessesgenerateprimarilyfine particles,thecadmiumcontrolefficienciesareassumedtobe equivalenttotheoverallPMcontrolefficienciesfor electrostaticprecipitators,highenergyscrubbers,andfabric filtersappliedtohightemperatureprocesses.

Informationavailableononenonrecoveryovenshowsthat chargingemissionsarecontrolledwithatravellinghoodthat ventstoabaghouse.7 Totalcaptureefficiencywasestimatedat 70percent,andcadmiumwasnotdetectedinthebaghousesamples. Thedetectionlevelwas0.5milligramperkilogram(mg/kg)[parts permillion(ppm)].7 Therefore,itcanbeassumedthereareno significantcadmiumemissionsfromnonrecoverycokeovens.

EstimatesoncontroldeviceefficiencyforPMare availableforESP’s,scrubbers,andbaghousesusedinsintering operations. Theseefficienciesrangefrom93percent(fora

8-10 cyclone,ESP,scrubberconfiguration)to99.9percent(fora baghouse)andarebasedonStatepermittinginformation.

Controlsusedforblastfurnacesincludecyclonesor gravitycollectorsincombinationwithscrubbers,baghouses,and ESP’s. Efficienciesforbaghousesrangedfrom98to 99.25percent. Iffollowedbyascrubber,theycouldbeashigh as99.9percent. Gravitycollectorsusedwithscrubberscould achieve99to99.9percentefficiency. Scrubbersusedwithan ESPhaveremovalefficienciesof99.3to99.9percent.

ControldevicesusedforBOF’sincludescrubbers,fabric filtersandESP’s. Controlefficiencyestimatesare: (1)99.4 to99.7percentforESP’s,(2)98.5to99.9percentfor scrubbers,(3)99percentforbaghouses,and(4)96percentfora cyclonewithscrubber. Noinformationwasavailableoncontrol devicesusedforopenhearthfurnaces.

BecauseasubstantialportionoftheemissionsfromanEAF arefugitiveinnature,emissioncontrolsystemsincludecapture andcollectionsystems. Thecapturesystemsincludeaprimary system,whichisdesignedtocaptureemissionsduringthemelt, andasecondarysystem,whichisdesignedtocapturefugitive emissionsfromcharging,tapping,andfurnaceleakage. Typically,primarycollectionsystemsareeitherdirectshell evacuation(DSE)systems(alsocalledfourthholesystems)or sidedrafthoodsystems,andsecondaryemissionsarecollected viacanopyhoods,close-capturelocalhoods,orfurnace enclosures. Gasescollectedbythesesystemsaregenerally directedtoafabricfilterforPMcollection. Available informationindicatesthatproperlydesignedandoperated secondarysystemsareexpectedtocapture75to95percentofthe emissionsfromEAF’s,whileDSEsystemscollect99percentofthe emissionsgeneratedduringthemelt. Performancedataforfabric

8-11 filtersaregenerallypresentedasachievableoutlet concentrationlevelsbutwell-operatedunitsareexpectedto achieveefficienciesthatexceed99percent.

Emissions1,8

Table8-3summarizescadmiumreleasesreportedinthe 1990ToxicChemicalReleaseInventory(TRI). Testdatafor cadmiumemissionsareonlyavailablefromoneplant,Bethlehem SteelatSparrowsPoint,Maryland. Becausethesedataarenot completeandcouldnotbevalidated,theyshouldbeused cautiously. Itshouldbenotedthatthisparticularplant reportednocadmiumemissionsinthe1990TRI. Thetestdatafor thisonefacilityaresummarizedinTable8-4,butnoinformation isavailableontheirrepresentativeness.

NospecificdataforcadmiumemissionsfromEAF’swere foundintheliterature,andnoemissiontestdatawereavailable topermitthecalculationofcadmiumemissions.

PORTLANDCEMENTMANUFACTURING

Morethan30rawmaterialsareusedtomanufacture Portlandcement. Thesematerialscanbeclassifiedintofour basicclassesofrawmaterials: lime,silica,alumina,andiron. Twoprocesses,thewetanddryprocesses,canbeusedto manufacturePortlandcement. In1990,therewereatotalof212 U.S.cementkilnswithacombinedtotalclinkercapacityof 73,518x103 Mg(81,056x103 tons). Ofthistotal,11kilns withacombinedcapacityof1,767x103 Mg(1,948x103 tons) wereinactive. Thetotalnumberofactivekilnswas201witha clinkercapacityof71,751x103 Mg(79,108x103 tons).11 The name,location,andclinkercapacityofeachkilnispresentedin AppendixC. Basedon1990U.S.cementkilncapacitydata,an

8-12 TABLE8-3.CADMIUMRELEASESREPORTEDBYIRON ANDSTEELFACILITIESIN1990TRI

Totalreleases

Facility/location kg/yr lb/yr ARMCO,KansasCity,MO 115.7 255 ASARCO,EastHelena,MT 8,164.7 18,000 BarbaryCoastSteelCorp.,Kent,WA 25.4 56 BethlehemSteel,Bethlehem,PA 176.9 390 BirminghamBolt,Birmingham,AL 46.3 102 BirminghamBolt,Bourbonnais,IL 50.3 111 BloomfieldFoundryInc.,Bloomfield,IA 4.54 10 BSCSteel,Jackson,MS 21.3 47 CascadeSteelRollingMills,McMinnville,OR 113.4 250 CitisteelUSA,Clairmont,DE 113.4 250 DanaCorp.,Richmond,IN 4.54 10 EastJordanIronWorks,EastJordan,MI 2.27 5 GMCSaginawGreyIronPlant,Saginaw,MI 115.7 255 JohnDeereFoundry,Waterloo,IA .45 1 LukensSteel,Coatesville,PA 39.9 88 NewportSteelCorp.,Newport,KY 2.27 5 NorfolkSteelCorp.,Chesapeake,VA 4.54 10 NucorSteel,Polymouth,UT 82.1 181 SalmonBaySteelCorp.,Kent,WA 9.98 22 SeattleSteel,Seattle,WA 226.8 500 SheffieldSteelCorp.,SandSprings,OK 2.27 5 RoanokeElectricSteel,Roanoke,VA 207.3 457 RougeSteel,Dearborn,MI 16.33 36 TOTAL 9,546.4 21,046

Source:Reference9.

8-13 TABLE8-4.CADMIUMEMISSIONSREPORTEDFROMBETHLEHEMSTEEL, SPARROWSPOINT,MARYLAND

Emissions

Operationsource kg/yr lb/yr

Sintering

--Transferpointscontrolledbyscrubber 0.59 <1.3

Limestonestockpile(fugitives) 0.45 1

Cokebreezeunloading(fugitives) 1.8 4

StockHouse(fugitives) 39.9 88

--Windboxanddischargecontrolledbybaghouse 0.27 <0.6

Blastfurnace

--Uncontrolledcasthouseroofmonitor 1.36 3.0

--Tapholeandtrough(notrunners) 0.91 2.0

Basicoxygenfurnace(BOF)

--Uncontrolled 0.45 1

--Fugitives 1.36 3

BOFcharging

--Atsource 0.45 1

--Controlledbybaghouse 0.045 0.1

--Hotmetaltransferatsource 0.91 2.0

BOFtapping 4.99 11.00

Source:References2and3. aTypicalofolderfurnaceswithnocontrols,orforcanopyhoodsortotalcasthouseevacuation.

8-14 estimated68percentofthePortlandcementismanufacturedusing thedryprocess,andtheremaining32percentisbasedonthewet process. Adescriptionoftheprocessesusedtomanufacture Portlandcementandadiscussionoftheemissionsresultingfrom thevariousoperationsarepresentedbelow.

ProcessDescription3,10

Figure8-2presentsabasicflowdiagramofthePortland cementmanufacturingprocess. Theprocesscanbedividedinto fourmajorsteps: rawmaterialacquisitionandhandling,kiln feedpreparation,pyroprocessing,andproductfinishing.

TheinitialstepintheproductionofPortlandcement manufacturingisrawmaterialsacquisition. ,whichis theelementofhighestconcentrationinPortlandcement,is obtainedfromavarietyofcalcareousrawmaterials,including limestone,chalk,marl,seashells,aragonite,andanimpure limestoneknownas"naturalcementrock." Theotherraw materials--,aluminum,andiron--areobtainedfromores andminerals,suchassand,shale,clay,andironore. Cadmium isexpectedtobepresentintheores(primarilyiron,zinc,lead andcopper)andmineralsextractedfromtheearth.6 Theonly potentialsourceofcadmiumemissionsfromrawmaterial acquisitionwouldbeduetowindblowncadmium-containing particulatefromthequarryoperations.

Thesecondstepinvolvespreparationoftherawmaterials forpyroprocessing. Rawmaterialpreparationincludesavariety ofblendingandsizingoperationsdesignedtoprovideafeedwith appropriatechemicalandphysicalproperties. Therawmaterial processingdifferssomewhatforwetanddryprocess. At facilitieswherethedryprocessisused,themoisturecontentin therawmaterial,whichcanrangebetween2and35percent,is

8-15 reducedtolessthan1percent. Cadmiumemissionscanoccur duringthisdryingprocess. Theleveloftheemissionswill dependuponthechemicalformofthemercuryintherawmaterial andthetemperatureanddurationofthedryingprocess. Heatfor dryingisprovidedbytheexhaustgasesfromthepyroprocessor. Atfacilitieswherethewetprocessisused,waterisaddedto therawmaterialduringthegrindingstep,therebyproducinga pumpableslurrycontainingapproximately65percentsolids.

Pyroprocessing(thermaltreatment)oftherawmaterialis carriedoutinthekiln,whichistheheartofthePortland cementmanufacturingprocess. Duringpyroprocessing,theraw materialistransformedintoclinkers,whicharegray,- hard,spherically-shapednodulesthatrangefrom0.32to5.1cm (0.125to2.0in.)indiameter. Thechemicalreactionsand physicalprocessesthattakeplaceduringpyroprocessinginclude:

1. Evaporationofuncombinedwaterfromrawmaterialsas materialtemperatureincreasesto100°C(212°F),

2. Dehydrationandprecalcinationasthematerial temperatureincreasesfrom100°Ctoapproximately430°C(806°F),

3. Calcinationduringwhichcarbondioxide(CO2)is evolved,between430°Cand900°C(1652°F),

4. Sinteringoftheoxidesintheburningzoneofthe rotarykilnattemperaturesupto1510°C(2750°F),and

5. Productionofcementclinkerasthetemperature decreasesfrom1510°Cto1370°C(2498°F).

Therotarykilnisalong,cylindrical,slightlyinclined, refractory-linedfurnace. Therawmaterialmixisintroducedin

8-17 thekilnattheelevatedend,andthecombustionfuelsare introducedintothekilnatthelowerend,inacountercurrent manner. Therotarymotionofthekilntransportstheraw materialfromtheelevatedendtothelowerend. Fuelsuchas coal,oil,ornaturalgasisusedtoprovideenergyfor calcination. Useofotherfuels,suchasshreddedmunicipal garbage,chippedrubber,coke,andwastesolvents,isbecoming increasinglypopular.

Pyroprocessingcanbecarriedoutusingoneoffour differentprocesses: wetprocess,dryprocess,dryprocesswith apreheater,anddryprocesswithapreheater/precalciner. These processesessentiallyaccomplishthesamephysicalandchemical stepsdescribedabove. Dependingontheprevalenceofpreheaters andprecalcinersatfacilitieswherePortlandcementis manufacturedbythedryprocess,thesesegmentsoftheprocess canbetheprimarysourcesofcadmiumemissions. Thisisbecause cadmiumpresentintherawmaterialcanevaporatereadilyduring thepreheatingandprecalciningsteps. Thelaststepinthe pyroprocessingisthecoolingoftheclinker. Thisprocessstep recoupsupto30percentoftheheatinputtothekilnsystem, locksindesirableproductqualitiesbyfreezingmineralogy,and makesitpossibletohandlethecooledclinkerwithconventional conveyingequipment. Finally,afterthecementclinkeris cooled,asequenceofblendingandgrindingoperationsiscarried outtotransformtheclinkerintofinishedPortlandcement.

EmissionControlMeasures

Withtheexceptionofthepyroprocessingoperations,the emissionsourcesinthePortlandcementindustrycanbe classifiedaseitherprocessfugitiveoropendustemissions. Theprimarypollutantsresultingfromthesefugitivesourcesare PM. Thecontrolmeasuresusedforthesefugitivedustsources

8-18 arecomparabletothoseusedthroughoutthemineralproducts industries.

Methodsusedtoreduceparticulatelevelsintheambient airduetovehiculartrafficincludepavingandroadwetting. Additionalmethodsthatareappliedtootheropendustsources includewetsprays,withandwithoutsurfactants,chemicaldust suppressants,windscreens,andprocessmodificationstoreduce dropheightsorenclosestorageoperations.

Processfugitiveemissionsourcesincludematerials handlingandtransfer,rawmillingoperationsindryprocess facilities,andfinishmillingoperations. Cadmiumemission sourcesareindicatedinFigure8-2bysolidcircles. Typically, particulateemissionsfromtheseprocessesarecapturedbya ventilationsystemcomprisingoneormoremechanicalcollectors withafabricfilterinaseries. Becausethedustfromthese unitsisreturnedtotheprocess,theyareconsideredtobe processunitsaswellasairpollutioncontroldevices. The industryusesshaker,reverseair,andpulsejetfilters,aswell assomecartridgeunits,butmostnewerfacilitiesusepulsejet filters. Forprocessfugitiveoperations,thedifferentsystems arereportedtoachievetypicaloutletPMloadingsof45 milligramspercubicmetermg/m3 (0.02grainsperactualcubic foot[gr/acf]). Becausethecadmiumisinparticleform,the performanceofthesesystemsrelativetocadmiumcontrolis expectedtobeequivalenttothisoverallparticulate performance. However,nodataareavailableoncadmium performanceoffugitivecontrolmeasures.

Inthepyroprocessingunits,PMemissionsarecontrolled byfabricfilters(reverseair,pulsejet,orpulseplenum), electrostaticprecipitators(ESP’s),andelectrifiedgravelbed (EGB)filters. ThereverseairfabricfiltersandESP’s

8-19 typicallyusedtocontrolkilnexhaustsarereportedtoachieve outletPMloadingsof45mg/m3 (0.02gr/acf). Clinkercooler systemsarecontrolledmostfrequentlywithpulsejetorpulse plenumfabricfilters,butreverseairfabricfilters,ESP’s,and EGB’sarebecomingincreasinglypopular. Nodataareavailable ontheperformanceofthesecontrolsystemsforcadmium emissions.

Emissions

Emissionsresultingfromallfourprocessingstepsinclude particulatematter. Additionally,emissionsfromthe pyroprocessingstepincludeotherproductsoffuelcombustion

suchassulfurdioxide(SO2),nitrogenoxides(NOx),andcarbon monoxide(CO). Cadmiumemissionscanbeexpectedfromtherotary kilnorpreheater/precalciner,ifusedintheprocess. Cadmium thatmaybepresentintherawmaterialandthefuelcan potentiallybeemittedfromeitherofthesesources. Astheuse ofmunicipalwasteandhazardouswastebecomesmoreprevalentas anauxiliaryfuel,thelikelihoodforcadmiumemissionsmay increasesignificantly. Table8-5providesemissionfactors basedonemissiontestsconductedduringanewsourceperformance standard(NSPS)developmentforPortlandcementplants

PHOSPHATEROCKPROCESSING

Phosphaterock,acalciumphosphatemineralknownas 4 apatite(Ca10(PO )6F2)isminedandprocessedbybeneficiation, dryingorcalcining,andgrinding.12 Asanaturalimpurityin thephosphaterock,cadmiumemissionsmayoccurduringany thermalprocessingoftherock. Floridaorescancontain2to 15partspermillion(ppm)ofcadmium;NorthCarolinaores,10to 25ppm;Tennesseeores,0.1to2ppm;andwesternores(mainly Idaho),2to980ppmwithamedianconcentrationof200ppm.12

8-20 TABLE8-5.CADMIUMEMISSIONFACTORSFROMPORTLANDCEMENTFACILITIES

EmissionFactor

Process kg/Mgoffeed lb/tonoffeed Ref.a

RawMill-AirSeparator 4.43x10-7 8.87x10-7 1

FinishingMill-WeighHopper 7.56x10-7 1.51x10-6 1

FinishingMill-AirSeparator 1.30x10-6 2.59x10-6 1

ClinkerCooler 8.7x10-6 1.7x10-5 2

WetRotaryKiln 1.1x10-4 2.2x10-4 2

Source:Reference10. a1.Sourcetest:EmissionsfromDryProcessRawMillandFinishMillSystemsatIdealCement Company,Tijeras,NewMexico.ETBTestnumber71-MM-02.1972.

2.Sourcetest:EmissionsfromWetProcessCementKilnandClinkerCooleratIdealCement Company,Seattle,Washington.ETBTestnumber71-MM-03.1972.

8-21 Whilefertilizerproductionisthemajoruseforphosphate rock,therockcanalsobeusedtoproducephosphoricacidand elementalphosphorus. Theseindustriesarenotconsideredpart ofthephosphaterockprocessingindustry;however,becausethey usephosphaterockasarawmaterialtheyarebrieflysummarized. Phosphoricacidisproducedbydigestingthephosphaterockwith sulfuricacid,andpurephosphorusismanufacturedinanelectric arcorblastfurnacefromphosphaterockandsilica.13,14

Anumberoffertilizersareproduceddependingonthe phosphatecontentoftherock. Thesefertilizersare: normal superphosphate,triplesuperphosphate;andammoniumphosphate. Nothermalprocessingisusedinfertilizerproduction; therefore,itisnotconsideredasignificantsourceofcadmium emissions. Additionally,nodataareavailablethatquantify cadmiumairemissionsduringthechemicalreactionsthatproduce . Sincethereisalsonothermalprocessingduring phosphoricacidproduction,thisprocessisnotconsideredtobe acadmiumemissionsource.

UsinganEAFinelementalphosphorusproductionisa thermaltreatmentprocessandtheEAFmaybeacadmiumemission source. Therockisheatedinalargefurnacewithcokeand silicatotemperaturesof1300°to1500°C(2372°to2732°F)to producephosphorusvapor. Elementalphosphorusiscondensedtoa liquidincoolingtowerswithwaterspraysat45°to55°C(113° to131°F).14

Dryingorcalciningphosphaterock,arealsothermal treatmentprocesses,andareconsideredcadmiumemissionsources forphosphaterockprocessing. Thesepotentialemissionsources arediscussedinmoredetailintheprocessdescription.

8-22 TableC-3inAppendixCliststhosecompanies,asof January1,1991,whichminephosphaterockandtheirannual capacities. In1991,154.5millionMg(340milliontons)of phosphaterockwereminedand48.1millionMg(106milliontons) weremarketedafterprocessing.15 Ofthatamount,Floridaand NorthCarolinaproduced149.8millionMg(330millionpounds). Theremaining4.7millionMg(10.3milliontons)wereminedin Idaho,Montana,Tennessee,andUtah.15 Annualplantproduction capacitywasreportedas59.1millionMg(130milliontons)in 1991.16

KnownEAFoperationsthatproduceelementalphosphorus fromphosphaterockareinPocatello,Idaho;SilverBow,Montana; andSodaSprings,Idaho(Monsanto). Theseproducersandtheir capacitiesarelistedinTableC-4inAppendixC.

ProcessDescription3

Figure8-3providesaflowdiagramfortheoverall processingoftherockformanufacturing. Beforemanufacturing fertilizerorelementalphosphorus,theminedphosphaterockmust bebeneficiated,dried(orcalcined),andground.

Beneficiationrequiresremovingclay,sand,ororganic materials. Dependingonwheretherockwasmined,itmayneedto besegregatedbycertainsizesforwashingandfurthergrinding. Onceaparticularsizeisreached,andallforeignmaterialshave beenremoved,awetrockmillgrindsthefinalslurrytothe consistencyoffinebeachsand. Hydrocyclonesseparatethe slurryintorockandclay;thentherockisfilteredoutand allowedtodryinpiles. Sincethisisawetgrindingprocess withnothermaltreatment,nocadmiumemissionsareanticipated.

8-23

Theorganiccontentoftherockdetermineswhetherthe rockisdriedinrotaryorfluidized-beddriers,orheatedin fluidized-bedcalciners. Rockfreeoforganiccontaminantsis driedindryersattemperaturesaround120°C(248°F). Rotary driersaremostcommonlyusedandoperateonnaturalgasorfuel oil(Nos.2or6). Rockwithorganiccontaminantsisheatedto 760°Cto870°C(1400°Fto1598°F),usuallyinfluidized-bed calciners. Afterheatingordrying,therockisconveyedto storagesilosonprotectedconveyorsforprocessinginthe grindingmill. Sincedryersandcalcinersareheatprocesses,it isbelievedtherewillbecadmiumemissionsfromthesesources.

Aftercalciningordrying,thephosphaterockissentto thegrindingmills. Rollerorballmillsareusedtogrind calcinedphosphaterockintoafinepowder;60percent(by weight)ofthispowdertypicallypassesthrougha200-meshsieve. Arotaryvalvefeedstherockintothegrindingmills,and circulatingairstreamsremovethegroundrock. Anyremaining oversizedparticlesaresentbackintothemillforregrinding. Finalproductrockisseparatedbyacycloneforuseinthenext manufacturingstep,usuallytomakefertilizers.

Whilegrindingisnotaheattreatmentstep,itis believedtobeasourceoffugitiveparticulateemissionsfrom blownfinerock. Thisparticulatemaycontaincadmiumtracesas notedpreviously.12

EmissionControlMeasures3

Controlequipmentusedforphosphaterockdryersusually consistsofscrubbersorelectrostaticprecipitators. Fabric filtersarenotused. Venturiscrubberswithlowpressuredrops (12inchesofwater,or3,000Pascal,(Pa))canremove80to 99percentoftheparticulatesthatare1to10micrometersin

8-25 diameter;andforparticulateslessthan1micrometer,10to 80percentmayberemoved. Scrubberswithhighpressuredrops (30inchesofwater,or7,500Pa)canremove96to99.9percent (1to10micrometerparticulates)or80to86percent(lessthan 1micrometerparticulates). Electrostaticprecipitatorscan remove90to99percentofallparticulates. Ifawetgrinding processisused,adryingstepanditsparticularemissionsare eliminated.

Calcinersalsousescrubbersandsometimesfabricfilters. Oneoperatingcalcinerusesanelectrostaticprecipitator. Grindersusefabricfiltersandscrubberstocontrolemissions. Operatingtheaircirculatingstreamsingrindersatnegative pressureavoidsfugitiveemissionsofrockdust.

Materialhandlingsystemsforgroundrock,suchas elevatorsandconveyors,haveahighpotentialforfugitive emissions. Theseemissionscanbecontrolledbycoveringand enclosingconveyors,whichhavecontrolleddischargepoints. Materialtransferareasarehoodedandthehoodsevacuatedtoa controldevice. Storagesilosorbinsthatareventedtothe atmosphereusuallyhavefabricfilterstocontrolparticulate emissions.

Electricarcfurnaceemissionsfromelementalphosphorous productionaremostoftencontrolledbyfabricfilters,although venturiscrubbersandelectrostaticprecipitatorscanbeused.17

Emissions3

PhosphateRockProcessing-- Themajorcadmiumparticulateemissionsfromphosphate rockprocessing(toproducefertilizers)areassociatedwiththe processesofdrying,calcining,andgrinding. Theseemission

8-26 sourcesareshownbysolidcirclesonFigure8-3. Since beneficiationinvolvesslurriesofrockandwater,thereareno significantcadmiumparticulateemissions.

Nodatawereavailableconcerningmeasuredcadmium emissionsfromdrying,calciningandgrinding. Thereare emissionfactorsinAP-42foruncontrolledPMfromthese particularprocesses,butnoinformationisavailableoncadmium levelsinthisPM. Therefore,thesefactorscannotbeusedto provideinformationonpotentialcadmiumemissions.

Nospecificdataforcadmiumemissionsfromphosphaterock processingwerefoundintheliterature,andnoemissiontest datawereavailabletopermitthecalculationofcadmium emissions.

ElementalPhosphorusProduction-- OnlyoneEAFreportedcadmiumemissionsinthe1990TRI, FMCCorporationinPocatello,Idaho. Totalreleasesof3,966kg (8,725lb)werereported;emissionfactorswereusedtoestimate nonpointreleasesof88kg(194lb)andtheremainingpoint releasesof3,878kg(8,532lb)werebasedonmonitoringdata.9 AccordingtotheBureauofMines,anumberofEAFplantswere closedby1991,andthe1990TRIprobablyreflectsthisdeclining production.9,15

CARBONBLACKPRODUCTION

Carbonblackisanindustrialchemicalusedasa reinforcingagentinrubberproducts,suchastires,andasa blackpigmentforprintinginks,surfacecoatings,andpaperand plastics.17 Cadmiummaybeacontaminantintherawmaterials usedandmaybeemittedduringcarbonblackproduction. Table8-6providesalistingoffacilitiesproducingcarbon

8-27 TABLE8-6.CARBONBLACKPRODUCTIONFACILITIES

Annual capacityb Type of a Company Location process 103 Mg 106 lb

Cabot Corporation Franklin, Louisiana F 141 310 North American Rubber Black Division Pampa, Texas F 32 70

Villa Platte, Louisiana F 127 280

Waverly, West Virginia F 82 180

Chevron Corporation Cedar Bayou, Texas A 9 20 Chevron Chemical Company, subsidiary Olevins and Derivatives Division

Degussa Corporation Aransas Pass, Texas F 57 125

Belpre, Ohio F 59 130

New Iberia, Louisiana F 91 200

Ebonex Corporation Melvindale, Michigan C 4 8

General Carbon Company Los Angeles, California C 0.5 1

Hoover Color Corporation Hiwassee, Virginia C 0.5 1

J.M. Huber Corporation Baytown, Texas F 102 225

Borger, Texas F and T 79 175

Orange, Texas F 61 135

Phelps Dodge Corporation El Dorado, Arkansas F 50 110 Colombian Chemical Company, subsidiary Moundsville, West Virginia F 77 170

North Bend, Louisiana F 109 240

Ulysses, Kansas F 36 80

Sir Richardson Carbon & Gasoline Company Addis, Louisiana F 66 145

Big Spring, Texas F 52 115

Borger, Texas F 98 215

Witco Corporation Phenix City, Alabama F 27 60 Continental Carbon Company, subsidiary Ponca City, Oklahoma F 66 145

Sunray, Texas F 45 100

TOTAL 1,471 3,240

Source: Reference 16. aA = acetylene decomposition C = combustion F = furnace T = thermal bCapacities are variable and based on SRI estimates as of January 1, 1991

8-28 black,theirannualcapacity,andproductionprocesses. Total annualcapacityasof1991was1.47millionMg(1.62million tons).16

ProcessDescription18

Carbonblackisproducedbypartialcombustionof hydrocarbons. Themostpredominantlyusedprocess(which accountsformorethan98percentofcarbonblackproduced)is basedonafeedstockconsistingofahighlyaromatic petrochemicalorcarbochemicalheavyoil. Cadmiumcanbe expectedtobepresentinthefeedstock. Althoughthecadmium contentinthefeedstockusedtomanufacturecarbonblackisnot known,cadmiumcontentinpetroleumcrudehasbeenreportedat 0.03partspermillionbyweight(ppmwt).19 Figure8-4contains aflowdiagramofthisprocess.

Threeprimaryrawmaterialsusedinthisprocessare, preheatedfeedstock(eitherthepetrochemicaloilor carbochemicaloil),whichispreheatedtoatemperaturebetween 150°Cand250°C(302°Fand482°F),preheatedair,andan auxiliaryfuelsuchasnaturalgas. Aturbulent,high- temperaturezoneiscreatedinthereactorbycombustingthe auxiliaryfuel,andthepreheatedoilfeedstockisintroducedin thiszoneasanatomizedspray. Inthiszoneofthereactor, mostoftheoxygenwouldbeusedtoburntheauxiliaryfuel, resultingininsufficientoxygentocombusttheoilfeedstock. Thus,pyrolysis(partialcombustion)ofthefeedstockis achieved,andcarbonblackisproduced. Mostofthecadmium presentinthefeedstockwillbeemittedasPMinthehotexhaust gasfromthereactor.

Theproductstreamfromthereactorisquenchedwith water,andanyresidualheatintheproductstreamisusedto

8-29 preheattheoilfeedstockandcombustionairbeforerecovering thecarboninafabricfilter. Carbonrecoveredinthefabric filterisinafluffyform. Thefluffycarbonblackmaybe groundinagrinder,ifdesired. Dependingontheenduse, carbonblackmaybeshippedinafluffyformorintheformof pellets. Pelletizingisdonebyawetprocessinwhichcarbon blackismixedwithwateralongwithabinderandfedintoa pelletizer. Thepelletsaresubsequentlydriedandbaggedprior toshipping.

EmissionControlMeasures18

Duringthemanufactureofcarbonblack,high-performance fabricfiltersareusedintheoilfurnaceprocesstorecover additionalcarbonblack;however,theyalsocontrolPMemissions frommainprocessstreams. Fabricfiltersreportedlycanreduce PMemissionstolevelsaslowas6mg/m3 (normalm3),andwillbe usedbyfacilitiestooptimizetheirmanufacturingperformance.

Foroilfurnaces,acyclonecanbeusedforparticle agglomerationupstreamofthefabricfilter. Asinglecollection systemoftenservesseveralmanifoldedfurnaces.

Emissions

Thelocationsofcadmiumparticulateemissionsourcesfrom theoilfurnaceprocessareshownbysolidcirclesinFigure8-4. Thegreatestreleaseofcadmiumoccursduringpyrolysisofthe feedstock,makingthereactorthemajoremissionsourceduring production.

Nodataareavailableconcerningcadmiumemissionsfrom thethermalprocess. Itisalsonotknownhowefficientlyfabric filterscapturecadmiumemissions. Theonlyavailabledataare

8-31 foremissionsfromtheoilfurnaceprocess. Thesedatashow cadmiumemissionstobelessthan5.0x10-5 kg/Mg (1.0x10-4 lb/ton)fromthemainprocessvent.19 Thisdata sourcewasacompilationofreporteddataandnottestdata;it shouldbeusedwithcaution.

MOBILESOURCES

Historically,themajoremissionsmeasuredandregulated underTitleIIoftheCleanAirAct(CAA)frommobilesourcesare

CO,NOx,andhydrocarbons(HC). Emissionfactorsforthese specificpollutantsamongthedifferentmotorvehicleclassesare compiledinAP-42,VolumeII.21 Gasoline-poweredmotor,on-, light-dutyvehiclescomprisethemostsignificantmobileemission sourcesbecauseoftheirlargenumbers. Accordingtothe1990 StatisticalAbstract,1988nationwideregistrationswere estimatedtobe183.5millioncars,trucks,andbuses. Ofthat number,140.7millionwerepassengercarsand42.8millionwere trucksandbuses.22 Asof1991,thetotalvehiclemilestraveled (VMT)intheUnitedStateswas3,457,478millionkilometers (2,147,501millionmiles).23

Potentialcadmiumemissionsresultfromtracequantities presentinthepetroleumcrudeoilfeedstockforfuelandmotor oil.19 Uncontrolledvehicleemissionshavedeclinedbecause catalyticconvertersandunleadedgasolinearerequiredalong withState-regulatedinspectionandmaintenanceprograms. Therefore,malfunctioningvehicleswouldbetheemissionsource fromcadmiumcontainingfuelormotoroil. Tirewearmayalsobe asourceofcadmiumemissionsifanycadmiumispresentasan impurityinthefinishedtire.

8-32 Astudyconductedin1979characterizedexhaustemissions fromnoncatalyst-andcatalyst-equippedvehiclesunder malfunctioningconditions.24 Nocadmiumwasdetectedinthe exhausts. Amorerecenttestwasperformedin1989to characterizeexhaustemissionsoflatemodelcarsfortoxic pollutantslistedorundergoingreviewforlistingunder California’sairtoxicsprogram.25 Particulatesampleswere takenandanalyzedfor31tracemetals,includingcadmium. Of the31tracemetals,only18weredetectedintheexhaust; cadmiumwasamongthegroupofmetalsthatwasnotdetected. The studyusedx-rayfluorescenceforthemetalsanalysesbutdidnot stateadetectionlimitforcadmium. Basedonthisstudy,the presenceofcadmiuminautoexhaustcannotbeexcludedbutmaybe presentatalevelbelowtheunspecifieddetectionlimitofthe analyticalmethodusedinthestudy.

REFERENCES 1. Houck,G.W. IronandSteel. AnnualReport:1991. Bureau ofMines,U.S.DepartmentoftheInterior. Washington, DC. December1992. 2. Trenholm,A.R. MidwestResearchInstitute. DRAFT PreliminaryDescriptionoftheIntegratedIronandSteel IndustryandofBlastFurnaceBasicOxygenFurnaceand SinteringOperations. EmissionsStandardsDivision, OfficeofAirQualityPlanningandStandards,U.S. EnvironmentalProtectionAgency,ResearchTrianglePark, NC. October1991. 3. U.S.EnvironmentalProtectionAgency. CompilationofAir PollutantEmissionFactors-4thed.,VolumeI,Stationary PointandAreaSources. AP-42. OfficeofAirand Radiation,OfficeofAirQualityPlanningandStandards, U.S.EnvironmentalProtectionAgency,ResearchTriangle Park,NC. 1985. 4. Easterly,W.E.etal. MetallurgicalCoke. (In)Air PollutionEngineeringManual. A.J.Buonicoreand W.T.Davies,eds. VanNostrandReinhold,NY. 1992.

8-33 5. Huskonen,W.W. AddingtheFinalTouches. 33Metal Producing. 29:28-131. May1991. 6. Coleman,R.etal. SourcesofAtmosphericCadmium. EPA-450/5-79-006. OfficeofAirQualityPlanningand Standards,U.S.EnvironmentalProtectionAgency,Research TrianglePark,NC. August1979. 7. ITCorporation. EngineeringEvaluationoftheCharging MachineHoodSystematJewelCoalandCokeCompany. Knoxville,TN. December1992. 8. U.S.EnvironmentalProtectionAgency. ElectricArc FurnacesandArgon-OxygenDecarburizationVesselsinSteel Industry-BackgroundInformationforProposedRevisionsto Standards. DraftEIS. EPA-450/3-82-020a. OfficeofAir QualityPlanningandStandards,ResearchTrianglePark, N.C. July1983. 9. U.S.EnvironmentalProtectionAgency. 1990Toxic ChemicalReleaseInventory. OfficeofToxicSubstances. Washington,DC. January1993. 10. MidwestResearchInstitute. PortlandCementManufacturing -AP-42Update. DRAFT. TechnicalSupportDivision, OfficeofAirQualityPlanningandStandards,U.S. EnvironmentalProtectionAgency,ResearchTrianglePark, NC. July1992. 11. PortlandCementAssociation. U.S.andCanadianPortland CementIndustry: PlantInformationSummary. Portland CementAssociation. Skokie,IL.1990. 12. Mumma,C.,K.Bohannon,andF.Hopkins. Chemical TechnologyandEconomicsinEnvironmentalPerspectives. TaskVI-CadmiuminPhosphateFertilizerProduction. EPA560/2-77-006. OfficeofToxicSubstances,U.S. EnvironmentalProtectionAgency. Washington,DC. November1977. 13. Dolan,M.J.,andR.B.Hudson. PhosphoricAcidand . (In)Kirk-OthmerConciseEncyclopediaof ChemicalTechnology. M.GraysonandD.Eckroth,eds. A Wiley-IntersciencePublication,JohnWileyandSons,New York,NY. 1985. 14. VanWazer,J.R. PhosphorusandthePhosphides. (In) Kirk-OthmerConciseEncyclopediaofChemicalTechnology. M.GraysonandD.Eckroth,eds. AWiley-Interscience Publication,JohnWileyandSons,NewYork,NY. 1985.

8-34 15. Stowasser,W.F. PhosphateRock. AnnualReport:1991. BureauofMines,U.S.DepartmentoftheInterior. Washington,DC. November1992. 16. SRIInternational. 1991DirectoryofChemicalProducers: UnitedStatesofAmerica. SRIInternational,MenloPark, CA. 1991. 17. GCACorporation. SurveyofCadmiumEmissionSources. EPA450/3-81-013. OfficeofAirQualityPlanningand Standards,ResearchTrianglePark,NC. September1981. 18. Taylor,B.R.,Buonicore,A.J.,andW.T.Davies,eds. CarbonBlack. (In)AirPollutionEngineeringManual. VanNostrandReinhold,NY. 1992. 19. Smith,I.C.,T.L.Ferguson,andB.L.Carson. Metalsin NewandUsedPetroleumProducts. (In)TheRoleofTrace MetalsinPetroleum. T.F.Yeh,ed. AnnArborScience Publishers,AnnArbor,MI. 1975. 20. Serth,R.W.,andT.W.Hughes. PolycyclicOrganicMatter (POM)andTraceElementContentsofCarbonBlackVentGas. EnvironmentScienceTechnology,14(3): 298-301. 1980. 21. U.S.EnvironmentalProtectionAgency. CompilationofAir PollutantEmissionFactors-4thed.,VolumeII,Mobile Sources. AP-42. MotorVehicleEmissionLaboratory, OfficeofMobileSources,U.S.EnvironmentalProtection Agency,AnnArbor,MI. 1985. 22. U.S.BureauoftheCensus. StatisticalAbstractofthe UnitedStates: 1990(110thed.). Washington,DC. 1990. 23. Telecon. N.Dobie,EPA,OfficeofMobileSources,with T.Campbell,MidwestResearchInstitute. Highway StatisticsonTotalVehicleMilesTraveledintheU.S.in 1990. November1992. 24. Urban,C.M,andR.J.Garbe. RegulatedandUnregulated ExhaustEmissionsfromMalfunctioningAutomobiles. SocietyofAutomotiveEngineers,Inc.,TechnicalPaper Series,No.790696. June1979. 25. Warner-Selph,M.,andJ.DeVita. MeasurementsofToxic ExhaustEmissionsfromGasoline-PoweredLight-Duty Vehicles. SocietyofAutomotiveEngineers,Inc., TechnicalPaperSeries,No.892075. September1989.

8-35 SECTION9 SOURCETESTPROCEDURES

INTRODUCTION

Anumberofmethodsexisttodeterminecadmium(Cd) emissionsfromstationarysources. SeveralEPAofficesandsome Stateagencieshavedevelopedsource-specificordedicated samplingmethodsforCd. Otherindustrysamplingmethodsdo exist,butnoneofthesemethodshavebeenvalidatedandwillnot bediscussedinthissection.

SubsequentpartsofthissectionwilldiscussEPA referenceorequivalentsamplingmethodsforCd. Sampling methodsfallintooneoftwocategories: (1)dedicatedCd methodsforspecificsourcesor(2)multiplemetalssampling trainsthatincludeCdformultiplesources. Eachcategoryof methodswillbedescribed,differencesamongthemethodswillbe discussed,andacitationwillbeprovidedformoredetailed informationaboutthemethods.

Samplingmethodsincludedinthissectionwereselected fromEPAreferencemethods,equivalentmethods,draftmethods,or Statemethods. Tobeareferencemethod,asamplingmethodmust undergoavalidationprocessandbepublished. Toqualifyasan equivalentmethod,asamplingmethodmustbedemonstratedtothe EPAAdministrator,underspecificconditions,asanacceptable alternativetothenormallyusedreferencemethods. Also includedinthissectionisadraftmethod,whichisunder development.

9-1 MULTIPLEMETALSSAMPLINGTRAINS

Method0012-MethodologyfortheDeterminationofMetalsEmissions inExhaustGasesfromHazardousWasteIncinerationandSimilar CombustionSources1

Thismethodwasdevelopedforthedeterminationofatotal of16metals,includingCd,fromstackemissionsofhazardous wasteincineratorsandsimilarcombustionprocesses. Method0012 allowsforthedeterminationofparticulateemissionsfromthese sources. Adiagramofasamplingtraintypicalofamultiple metalssamplingtrainispresentedinFigure9-1.

Thestacksampleiswithdrawnisokineticallyfromthe source. Particulateemissionsarecollectedintheprobeandon aheatedfilter;gaseousemissionsarecollectedinaseriesof fourchilledimpingers: twocontainanaqueoussolutionof diluteHNO3 combinedwithdiluteH2O2 andtwocontainacidicKMnO4 solution. Samplingtraincomponentsarerecoveredanddigested inseparatefront-andback-halffractions. Materialscollected inthesamplingtrainaredigestedwithacidsolutionsusing conventionalParr®Bomb,ormicrowavedigestiontechniquesto dissolveorganicsandtoremoveorganicconstituentsthatmay createanalyticalinterferences. ThedetectionlimitforCdby ICAPisapproximately5ngCd/ml.

Thecorrespondingin-stackmethoddetectionlimitcanbe calculatedbyusing(1)theproceduresdescribedinthismethod, (2)theanalyticaldetectionlimitsdescribedintheprevious paragraph,(3)avolumeof300mlforthefront-halfand150ml fortheback-halfsamples,and(4)astackgassamplevolumeof 1.25m3:

9-2

AxB D C where: A=analyticaldetectionlimit,µgCd/ml B=volumeofsamplepriortoaliquotforanalysis,ml C=samplevolume,drystandardcubicmeter(dscm) D=in-stackdetectionlimit,µgCd/m3

MethodologyfortheDeterminationofMetalsEmissionsinExhaust GasesfromHazardousWasteIncinerationandSimilarCombustion Sources2

Themethodwasdevelopedtodeterminetheemissionsofthe samemetalsasMethod0012fromhazardouswasteincineratorsand similarcombustionsources. ThismethodissimilartoSW-846 Method0012insamplingapproachandanalyticalrequirements.

CARBMethod436-DeterminationofMultipleMetalsEmissionsfrom StationarySources3

Thismethodisapplicablefordeterminingtheemissionsof metals,includingCd,fromstationarysources. Thismethodis similartoSW-846Method0012insamplingapproachandanalytical requirements. Method436suggeststhattheconcentrationsof targetmetalsintheanalyticalsolutionsbeatleast10times theanalyticaldetectionlimits. Thismethodmaybeusedinlieu ofAirResourceBoardMethods12,101,104,423,424,and433.

9-4 EPAMethod29-MethodologyfortheDeterminationofMetals EmissionsinExhaustGasesfromIncinerationandSimilar CombustionSources(Draft)4

Thismethodisapplicablefordeterminingtheemissionsof metals,includingCd,fromstationarysources. Thismethodis similartoSW-846Method0012insamplingapproachandanalytical requirements.

ANALYTICALMETHODSFORDETERMINATIONOFCADMIUM

Thissectioncontainsbriefoverviewdescriptionsofthe fiveanalyticaltechniquesgenerallyusedfortracemetal determinations: (1)inductivelycoupledargonplasmaemission spectrometry(ICAP),(2)direct-aspirationorflameatomic absorptionspectrometry(FAA),(3)graphite-furnaceatomic absorptionspectrometry(GFAA),(4)hydride-generationatomic absorptionspectrometry(HGAA),and(5)cold-vaporatomic absorptionspectrometry(CVAA). Eachtechniqueisdiscussed belowintermsofadvantages,disadvantages,andcautionsfor analysis.

TheprimaryadvantageofICAP isthatitallows simultaneousorrapidsequentialdeterminationofmanyelements inashorttime. TheprimarydisadvantageofICAPisbackground radiationfromotherelementsandtheplasmagases. Althoughall ICAPinstrumentsutilizehigh-resolutionopticsandbackground correctiontominimizetheseinterferences,analysisfortraces ofmetalsinthepresenceofalargeexcessofasinglemetalis difficult. Anexamplewouldbetracesofmetalsinanalloyor tracesofametalinalimed(highcalcium)waste. ICAPand FlameAAhavecomparabledetectionlimits(withinafactorof4), exceptthatICAPexhibitsgreatersensitivityforrefractories (Al,Ba,etc.).

9-5 FlameAAS (FAA)determinations,asopposedtoICAP,are normallycompletedassingleelementanalysesandarerelatively freeofinterelementspectralinterferences. Eitheranitrous- oxide/acetyleneorair/acetyleneflameisusedasanenergy sourcefordissociatingtheaspiratedsampleintothefreeatomic state,makinganalyteatomsavailableforabsorptionoflight. Intheanalysisofsomeelements,thetemperatureandtypeof flameusediscritical. Iftheproperflameandanalytical conditionsarenotused,chemicalandionizationinterferences canoccur.

GraphiteFurnaceAAS (GFAA)replacestheflamewithan electricallyheatedgraphitefurnace. Thefurnaceallowsfor gradualheatingofthesamplealiquotinseveralstages. Thus, theprocessesofdesolvation,dryingdecompositionfororganic andinorganicmoleculesandsalts,andformationofatoms(which mustoccurinaflameorICAPinafewmilliseconds)maybe allowedtooccuroveramuchlongertimeperiodandatcontrolled temperaturesinthefurnace. Thisallowstheremovalofunwanted matrixcomponentsbyusingtemperatureprogrammingand/ormatrix modifiers. Themajoradvantageofthistechniqueisthatit affordsextremelylowdetectionlimits. Itistheeasiest techniquetoperformonrelativelycleansamples. Becausethis techniqueissosensitive,interferencescanbeaproblem; findingtheoptimumcombinationofdigestion,heatingtimesand temperatures,andmatrixmodifierscanbedifficultforcomplex matrices. FurnaceAA,ingeneral,willexhibitlowerdetection limitsthaneitherICAPorflameAAS.

HydrideAA (HGAA)utilizesachemicalreductiontoreduce andseparatearsenicorseleniumselectivelyfromasample digestate. Thetechnique,therefore,hastheadvantageofbeing abletoisolatethesetwoelementsfromcomplexsamples,which maycauseinterferencesfortheanalyticalprocedures.

9-6 Significantinterferenceshavebeenreportedwhenanyofthe followingispresent: (1)easilyreducedmetals(Cu,Ag,Hg), (2)highconcentrationsoftransitionmetals(>200mg/L),and (3)oxidizingagents(oxidesofnitrogen)thatremainfollowing sampledigestion.

Cold-VaporAA (CVAA)usesachemicalreductionto selectivelyreduceHg. Theprocedureisextremelysensitivebut issubjecttointerferencesfromsomevolatileorganics, chlorine,andsulfurcompounds.

SUMMARY

Alloftheabovesourcesamplingmethodscollectasample foranalysisofmultiplemetals,includingCd. Significant criteriaandcharacteristicsofeachmethodarepresentedin Table9-1. Thistableisasummaryofinformationpresentedin variousmethods. Themajordifferencesbetweenthemethods involve: (1)thetypeofimpingersolutions,(2)theamountor concentrationofimpingersolutions,(3)thesequenceandtypes ofsampletrainrecoverysolutions,and(4)theuseand/ortype ofparticulatefilter.

InassessingCdemissionsfromtestreports,theageor revisionnumberofthemethodindicatesthelevelofprecision andaccuracyofthemethod. Oldermethodsaresometimesless preciseoraccuratethanthosethathaveundergonemoreextensive validation. Currently,EPAMethod301from40CFRPart63, AppendixAcanbeusedtovalidateorprovetheequivalencyof newmethods.

9-7 TABLE 9-1. CADMIUM SAMPLING METHODS Method Filter Impinger Range Chemical interference Detection limit EPA 29 Quartz or glass 1 X empty (optional) ngCd/ml to µg Cd/ml Excessive chloride, iron 5 ng Cd/ml

(Draft) fiber 2 X HNO3/H2O2 1 X empty

2 X KMnO4/H2SO4 1 X silica gel SW-846 0012 Quartz or glass 1 X empty (optional) ngCd/ml to µg Cd/ml Excessive chloride, iron 4 ng Cd/ml

fiber 2 X HNO3/H2O2 by ICAP 1 X empty

2 X KMnO4/H2SO4 1 X silica gel OSW-BIF Quartz or glass 1 X empty ngCd/ml to µg Cd/ml Excessive chloride 5 ng Cd/ml

fiber 2 X HNO3/H2O2 1 X empty

2 X KMnO4/H2SO4 1 X silica gel

9-8 CARB 436 Quartz or glass 1 X empty ngCd/ml to µg Cd/ml Excessive chloride 5 ng Cd/ml

fiber 2 X HNO3/H2O2 2 X KMnO4/H2SO4 1 X silica gel REFERENCES 1. EPAMethod0012,MethodologyfortheDeterminationof MetalsEmissionsinExhaustGasesfromHazardousWaste IncinerationandSimilarCombustionSources,TestMethods forEvaluatingSolidWaste:Physical/ChemicalMethods. SW-846,ThirdEdition. OfficeofSolidWasteand EmergencyResponse,U.S.EnvironmentalProtectionAgency, Washington,DC. September1988. 2. MethodologyfortheDeterminationofMetalsEmissionsin ExhaustGasesfromHazardousWasteIncinerationand SimilarCombustionSources,MethodsManualforCompliance withtheBIFRegulationsBurningHazardousWastein BoilersandIndustrialFurnaces. EPA/530-SW-91-010. OfficeofSolidWasteandEmergencyResponse,U.S. EnvironmentalProtectionAgency,Washington,DC. December 1990. 3. CARBMethod436,DeterminationofMultipleMetals EmissionsfromStationarySources. StateofCalifornia AirResourcesBoard,Sacramento,CA. 4. EPAMethod29. MethodologyfortheDeterminationof MetalsEmissionsinExhaustGasesfromIncinerationsand SimilarCombustionSources(Draft). 40CodeofFederal Regulations,Part60,AppendixA. Washington,DC. 1992.

9-9 APPENDIXA NATIONWIDEEMISSIONESTIMATES

9-10 EMISSIONSFROMCADMIUMPRODUCTION CadmiumRefining BasisofInputData 1. The1990TRIreportedemissionsforallproducersof cadmium(seeTable4-3)tobe4.2Mg(4.6tons). 2. EmissionsreportedintheTRImaygiveabnormallyhigh valuesbecausetheTRIdatamayincludeunusualand accidentalreleases. However,intheabsenceofother data,thenationwideestimateswillbebasedonthese data. CadmiumPigmentsProduction BasisofInputData 1. Theemissionsreportedinthe1990TRIforproducers ofinorganicpigmentswere1.6Mg(1.8tons). These dataarepresentedinTable4-8. 2. EmissionsreportedintheTRImaygiveabnormallyhigh valuesbecausetheTRIdatamayincludeunusualand accidentalreleases. However,intheabsenceofother data,thenationwideestimateswillbebasedonthese data. CadmiumStabilizersProduction Noemissionfactorsareavailableforcadmiumemissions fromthissource. OtherCadmiumCompoundProduction Noemissionfactorsareavailableforcadmiumemissions fromthissource.

A-1 EMISSIONSFROMMAJORUSESOFCADMIUM SecondaryBatteryManufacture BasisofInputData 1. The1990TRIreportedemissionsforallmanufacturers ofsecondarycadmiumbatteriestobe0.32Mg(0.35 tons). 2. EmissionsreportedintheTRImaygiveabnormallyhigh valuesbecausetheTRIdatamayincludeabnormaland accidentalreleases. However,intheabsenceofother data,thenationwideemissionspresentedinSection3 arebasedonthe1990TRIreport.

A-2 EMISSIONSFROMCOMBUSTIONSOURCES CoalCombustion Coal-FiredUtilityBoilers-- BasisofInputData 1. FromTable6-8,emissionfactorforbituminouscoal combustion=3.0x10-14 kg/Jandforanthracitecoal combustion=7.3x10-15 kg/J. 2. BituminouscoalcombustionsystemscontrolledbyESP’s withanaveragecadmiumcontrolefficiencyof 75percent. 3. Anthracitecoalcombustionsystemsuncontrolled. 4. Energyfromcoalcombustioninutilitysectorfrom Table6-1. Calculations AnnualEmissions=3.0x10-14 kg/J*16.939x1018 J/yr* 0.25 +7.3x10-15 kg/J*0.018x1018 J/yr =128.37Mg/yr=141.51tons/yr Coal-FiredIndustrialBoilers-- BasisofInputData 1. FromTable6-8,emissionfactorforbituminouscoal combustion=3.0x10-14 kg/Jandforanthracitecoal combustion=7.3x10-15 kg/J 2. Nocontrolofemissionsfromindustrialboilerswas assumed. 3. Energyfromcoalcombustioninindustrialsectorfrom Table6-1. Calculations AnnualEmissions=3.0x10-14 kg/J*2.892x1018 J/yr +7.3x10-15 kg/J*0.009x1018 J/yr =87.64Mg/yr=96.61ton/yr Coal-FiredCommercialandResidentialBoilers--

A-3 BasisofInputData 1. FromTable6-8,emissionfactorforbituminouscoal combustion=3.0x10-14 kg/Jandforanthracitecoal combustion=7.3x10-15 kg/J 2. Nocontrolofemissionsfromindustrialboilerswas assumed. 3. Energyfromcoalcombustionincommercial/residential sectorsfromTable6-1. Calculations AnnualEmissions=3.0x10-14kg/J*0.130x1018 J/yr +7.3x10-15kg/J*0.032x1018J/yr =4.17Mg/yr=4.60tons/yr Oil-FiredUtilityBoilers-- BasisofInputData 1. FromTable6-15,emissionfactorfordistillateoil combustion=4.7x10-15 kg/Jandforresidualoil combustion=7.1x10-15 kg/J 2. Airpollutioncontrolmeasuresassumedtoprovideno cadmiumemissionreduction. 3. Energyconsumptionfromfueloilcombustionfrom Table6-1. Calculations AnnualEmissions=4.7x10-15 kg/J*1.201x1018 J/yr +7.1x10-15 kg/J*0.091x1018 J/yr =6.24Mg/yr=6.88tons/yr Oil-FiredIndustrialBoilers-- BasisofInputData 1. FromTable6-15,emissionfactorfordistillateoil combustion=4.7x10-15 kg/Jandforresidualoil combustion=7.1x10-15 kg/J 2. Airpollutioncontrolmeasuresassumedtoprovideno cadmiumemissionreduction.

3. Energyconsumptionfromfueloilcombustionfrom

A-4 Table6-1. Calculations AnnualEmissions=4.7x10-15 kg/J*1.245x1018 J/yr +7.1x10-15 kg/J*0.436x1018 J/yr =8.91Mg/yr=9.82tons/yr Oil-FiredCommercial/ResidentialBoilers-- BasisofInputData 1. FromTable6-15,emissionfactorfordistillateoil combustion=4.7x10-15 kg/Jandforresidualoil combustion=7.1x10-15 kg/J 2. Airpollutioncontrolmeasuresassumedtoprovideno cadmiumemissionreduction. 3. Energyconsumptionfromfueloilcombustionfrom Table6-1. Calculations AnnualEmissions=4.7x10-15 kg/J*1.395*1018 J/yr +7.1x10-15 kg/J*0.255x1018 J/yr =8.31Mg=9.16tons/yr WoodCombustioninBoilers-- BasisofInputData 1. Woodcombustionrateinboilersis1.0x1011 Btu/hr, whichisthesamerateas1980givenonp.6-35. Boilersassumedtooperateatcapacity,8,760hr/yr. 2. Heatingvalueofwoodis4,500Btu/lbbasedon midpointofrangepresentedonp.6-35. 3. Emissionfactorof8.5x10-6 lb/tonofwoodburned. 4. Nodataavailableoncontrolofcadmiumemissionsfrom woodwasteboilers. Calculations AnnualEmissions

A-5 = 1.0x1011 Btu/hr*8,760hr/yr*8.5x10-6 lb/tonswood 4,500Btu/lb*2,000lbwood/tonwood*2,000lbHg/tonHg =0.41ton/yr=0.38Mg/yr MunicipalWasteCombustors-- BasisofInputData 1. Theemissionfactorsforuncontrolledsystemscontained inTable6-19wereaveragedtoobtainthefollowing "typical"emissionfactors: MassBurn-5.3g/Mg Modular -1.2g/Mg RDF -4.4g/Mg 2. Electrostaticprecipitatorsachieve95percentremoval. Spraydryersystemscombinedwithfabricfiltersor ESP’sachieve99and98percentremoval,respectively. Ductsorbentinjectionsystemscombinedwithfabric filtersorESP’sachieve99and95percentremoval, respectively. Basedonresultsforothercombustion sources,wetscrubbersystemsachieveabout75percent removal. 3. The1990MWCprocessingratesareassumedtobeequal tothosepresentedinWasteAge,November1991,and tabulatedinthecalculationtablebelow.1 Calculations

ControlledEmissions AnnualEmissions = ProcessRate*EmissionFactor * (100-Efficiency) 100 Thecalculatedemissionsaretabulatedbelow:

A-6 Process Control AnnualEmissions Combustor Control rate, Emission efficiency, type statusa 106 Mg/yr factor,g/Mg % Mg/yr ton/yr

MassBurn SD/FF 7.190 5.26 99 0.38 0.42

MassBurn SD/ESP 7.190 5.26 98 0.76 0.83

MassBurn DSI/FF 1.077 5.26 99 0.06 0.06

MassBurn DSI/ESP 1.077 5.26 95 0.28 0.31

MassBurn ESP 13.806 5.26 95 3.63 4.00

MassBurn U 0.517 5.26 0 2.72 3.00

RDF SD/FF 2.809 4.37 99 0.12 0.14

RDF SD/ESP 2.809 4.37 98 0.25 0.27

Modular WS 0.630 1.21 75 0.19 0.21

Total 8.39 9.24 a SD=Spraydryer FF=Fabricfilter ESP=Electrostaticprecipitator DSI=Ductsorbentinjection WS=Wetscrubber U=Uncontrolled SewageSludgeIncinerators-- BasisforInputData 1. Totalsludgeprocessedannuallyis1.5x106 Mg (seep.6-51) 2. FromTable6-20,thebest"typical"uncontrolled emissionfactoris26g/Mg. 3. Mostunitsarecontrolledbyeitheraventurior impingementscrubberandaveragenationwideefficiency is75percent. Calculations AnnualEmissions=26g/Mg*1.5x106 Mg/yr(1-0.75) =9.89Mg/yr=10.91tons/yr MedicalWasteIncinerators-- BasisofInputData 1. Fromunpublisheddatacontainedinthemedicalwaste incineratorNSPSbackgroundfiles,annualprocessrates are0.204x106 Mg/yrforpathologicalwasteand 1.431x106 Mg/yrformixedmedicalwaste.

A-7 2. FromTable6-23"typical"uncontrolledemissionfactors are2.5g/Mgformixedwasteand0.18g/Mgfor pathologicalwaste. 3. Fromunpublisheddatainthemedicalwasteincinerator NSPSbackgroundfiles,approximately3percentofthe MWI’sfiringmixedwastesareequippedwithpollution controldevices. Approximately45percentofthe controlledMWI’sareequippedwithdryscrubbersand approximately55percentwithwetscrubbers. Dry systemscanachieveatleast97percentcontrolwhile wetsystemscanachieveabout38percentcontrol. Calculations AnnualEmissions 2.5g/Mg*1.43x106 Mg/yr*0.97 +2.5g/Mg*1.43x106 Mg/yr*0.03*0.45*(1-0.97) 2.5g/Mg*1.43x106 Mg/yr*(0.03*0.55*(1-0.38)) +0.18g/Mg*0.204x106 Mg/yr =3.55Mg/yr=3.91tons/yr

EMISSIONSFROMNONFERROUSSMELTINGOPERATIONS 1. Primaryleadsmelting BasisofInputData Datain1990ToxicReleaseInventory(TRI)System. There are3primaryleadsmeltingfacilitieswithintheU.S. All 3facilitiesreportedcadmiumemissionsresultingfromall operationsfortheyear1990. Thesedataarepresentedin Table7-2. Accuracyofthesedatacannotbeverified. Therefore,thesumofthecadmiumemissionsforall3 facilitieswillalsobetheestimatednationwidecadmium emissionrate. Calculations Thesumofthecadmiumemissions(forboth,nonpointand pointsources)forall3facilitiesis14.3Mg(15.80tons), whichisalsotheestimatednationwideannualemissionrate ofcadmiumfromprimaryleadsmeltingfacilities.

2. PrimaryCopperSmelting

BasisofInputData

A-8 Datain1990TRI. Fourfacilitiesreportedtotal facilitywidecadmiumemissionratesfor1990. Thesedata arepresentedinTable7-5. Calculations Fourfacilitiesreportedcadmiumemissionratesfor1990as requiredunderSuperfundAmendmentsandReauthorizationAct (SARA)TitleIIIprovisions. Thesumofthecadmium emissionratesreportedbythe4facilitiesis5.6Mg (6.2tons). 3. PrimaryZincSmelting BasisofInputData Datain1990ToxicReleaseInventorySystem(TRIS). There are4primaryzincsmeltingfacilitieswithintheU.S. All 4facilitiesreportedcadmiumemissionsresultingfromall operationsfortheyear1990. Thesedataarepresentedin Table7-8. Accuracyofthesedatacannotbeverified. Therefore,thesumofthecadmiumemissionsforall4 facilitieswillbetheestimatednationwidecadmiumemission rate. Calculations Thesumofthecadmiumemissions(forboth,nonpointand pointsources)forall4facilitiesis5.7Mg(6.3tons), whichisalsotheestimatednationwideannualemissionrate ofcadmiumfromprimaryzincsmeltingfacilities. 4. SecondaryCopperSmelting BasisofInputData Datain1989and1990TRI. Threefacilitiesreportedtotal facilitywidecadmiumemissionratesfor1990. Onefacility reportedcadmiumemissiondatafor1989. Thesedataare presentedinTable7-11. Calculations TheTRIdatabasecontainsemissionsreportedfor1990by3 facilities. Anotherfacility(SouthwireCo.)reporteddata for1989. Forthisstudy,itisassumedthatthecadmium emissionsatSouthwireCo.for1990wasthesameasthat reportedfor1989. Thereareatotalof6secondarycoppermanufacturing facilitiesintheU.S. Rawmaterialsusedatthese

A-9 facilitiescanvarysignificantly. Therefore,itisnot validtoassumethatthecadmiumemissionrateisdirectly proportionaltotheplantcapacity. Thisisconfirmedby thefactthatthecadmiumemissionfactors(seeTablebelow) forthe4facilitiesrangebetween4.1x10-5 and0.028 kg/Mgofproduct. ______Facility Cadmiumemissionfactor(kg/Mg) ______CerroCopperProducts 0.0097 FranklinSmelting&Refining 0.028 GastonRecyclingInds. 4.1x10-5 SouthwireCo. 5.43x10-4 ______Thecadmiumemissionfactorsreportedabovehavebeen estimatedusingtheTRIdatainTable7-11andplant capacitydatacontainedinTable7-10. Cadmiumemissionfactorsarenotavailabletoseparately estimatetheemissionratesatthe2facilitiesforwhich therearenoTRIdata. Therefore,itisconservatively assumedthattheemissionfactorof0.028kg/Mgestimated forFranklinSmelting&RefiningCo.maybeapplicablefor estimatingcadmiumemissionsattheremaining2facilities. Thisapproachisconservativebecausetheemissionfactor chosenisthehighestofthe4factorsestimatedabove. Basedonthisemissionfactorandthecapacitiesgivenin Table7-10,thetotalcadmiumemissionrateforthe2 facilitiesisestimatedtobe5.0Mg(5.5tons). Thus,the nationwideannualcadmiumemissionrateforall6facilities isestimatedtobe10.8Mg(11.9tons).

5. SecondaryZincRecoveryfromMetallicScrap BasisofInputData 1. CadmiumemissionfactorsfromSPECIATEdatabasegiven inTable7-13. 2. Figure7-6containingtheprocessflowdiagram.

Calculations

A-10 Majorprocessstepsarepretreatmentofscrap,meltingthe sweatedscrapandrefining/alloying. Eachofthesesteps canbeperformedinvariousways. Becausethespecific processdescriptionforeachofthe13plantslistedin Table7-12isnotknown,weneedtocreateaconservative modelprocessflowscheme. RefertoFigure7-6. Forthisstudy,assumethatallplants producesecondaryzincfromdiecastproducts,residue skimmings,andothermixedscraps. Pretreatmentofthese materialsiscarriedoutinoneof4waysasshowninFigure 7-6. Ofthe4cadmiumemissionfactorscorrespondingto these4pretreatmentmethods,theemissionsfactorfor reverberatorysweatfurnace(generalmetallicscrap)isthe highest. Thisemissionfactoris0.01232kg/Mgofzinc produced. Therefore,itisassumedthatall13plantsuse thereverberatorysweatfurnaceforpretreatment. Next,themeltingiscarriedoutinoneof4ways. However, Table7-13doesnotcontaincadmiumemissionfactorsfor meltingoperations. Therefore,itisassumedthatall13 plantsalsocarryoutthemeltinginreverberatorymelting kettlesandthecadmiumemissionfactorformeltingisthe sameasthatforpretreatment(0.01232kg/Mgofzinc produced). Finally,zincingotisproducedinoneof2ways. The cadmiumemissionfactorforbothmethodsisreportedtobe 0.00091kg/Mg. Therefore,thetotalcadmiumemissionfactorforthethree majorprocessingstepsisestimatedtobe0.02555kg/Mgof zincproduced. Thetotalzincproductioncapacityforall 13plantsisreportedtobe58,000Mg. Therefore,the nationwideannualcadmiumemissionrateforall13plants resultingformthemajorprocessingstepsisestimatedtobe 1.5Mg(1.7tons). 6. SecondaryZincRecoveryfromEAFDust BasisofInputData CadmiumemissionfactordataforrecoveryofzincfromEAF dustarelimited. Cadmiumemissionfactorforonlyone processingstepusingtheflamereactorisavailable. The emissionfactorisreportedtobe2.1x10-4 kg/MgofEAF dustprocessed. Therefore,forthisstudy,itisassumed thatallfacilitiesproducingzincfromEAFdustusethe flamereactor.

A-11 Calculations Table7-14presentstheEAFdustprocessingcapacityforall 9facilitiesintheU.S. ThetotalEAFdustprocessing capacityis533,200Mg/yr. Ifallthisdustisprocessedin aflamereactorequippedwithafabricfilter,the nationwideannualcadmiumemissionrateresultingfromthis processingstepisestimatedtobe112kg(246lb). Itmust benotedthatothermajorprocessingstep,thecalcining kiln,willalsoemitcadmium. However,thecadmiumemission ratecorrespondingtothisstepcannotbeestimateddueto thelackofemissionfactordata. EMISSIONSFROMMISCELLANEOUSSOURCES IronandSteelProduction Theavailabledataareinsufficienttopermitthe calculationofemissionestimates. PortlandCementProduction BasisofInputData 1. The1990totalproductionofcementwas70.6x106 Mg (77.8x106 tons)ofwhich95.7percentwasPortland cement. TotalproductionofPortlandcementwas 67.5x106 Mg(74.5x106 tons). 2. Cadmiumemissionfactorsforselectedprocessesinboth dryprocesskilnsandwetprocesskilnswerepresented inTable8-5. Theseemissionfactorsarebasedon quantityofrawmaterialfeed. 3. Basedondataintestreports,thefollowingaverage conversionswereobtainedforrawmaterialquantities toclinkerquantities: wetprocess:1.7Mg(1.87tons)offeedproduces 1.0Mg(1.1tons)ofclinker dryprocess:1.6Mg(1.76tons)offeedproduces 1.0Mg(1.1tons)ofclinker 4. Basedonthetwoprocesses,cadmiumemissionsfrom a dryprocesskiln,includingpreheater/precalciner,are notanticipatedtobeanylessthanemissionsfromthe wetprocess. Therefore,theemissionfactorsin Table8-5areusedforallPortlandcementproduced.

A-12 Calculations Annualemissionsfromrawmill-airseparator: 6.75x106 Mg*1.6*4.43x10-7 kg/Mg=47.8kg/yr =0.048Mg/yr Annualemissionsfromfinishingmill-weighhopper: 67.5x106 Mg*1.6*7.56x10-7 kg/Mg =81.6kg/yr=0.082Mg/yr Annualemissionsfromfinishingmill-airseparator: 67.5x106 Mg*1.6*1.30x10-6 kg/Mg=140kg/yr =0.14Mg/yr Annualemissionsfromclinkercooler:

67.5x106 Mg*1.6*8.7x10-6 kg/Mg=940kg/yr =0.94Mg/yr Annualemissionsfromwetrotarykiln: 67.5x106 Mg*1.6*1.1x10-4 kg/Mg=11.9Mg/yr Totalannualemissions: 0.048Mg/yr+0.082Mg/yr+0.14Mg/yr+0.94Mg/yr+ 11.9Mg/yr=13.1Mg/yr=14.4tons/yr PhosphateRockProcessing Noemissionfactorsareavailableforcadmiumemissionsfrom thissource. CarbonBlackProduction BasisofInputData 1. The1990totalcapacityforcarbonblackproductionwas 1.47x106 Mg(1.62x106 tons).2 Nodatawere availableforactualproductionofcarbonblackin 1990. 2. Anemissionfactorof5x10-5 kgofCd/Mgofcarbon black(1x10-4 lb/ton)isused.3

A-13 3. Theemissionfactorisbasedonlyontheoil-furnace processwhichaccountsfor99percentofallcarbon blackproduction. 4. Cadmiumemissionsarebasedonproductioncapacityand notactualproduction. Useofactualproductiondata wouldshowalowervalueforcadmiumemissions. Calculations Annualemissions=5x10-5 kg/Mg*1.47x106 Mg= 0.07Mg/yr=0.08ton/yr MobileSources Noemissionfactorsareavailableforcadmiumemissionsfrom thissource.

REFERENCESFORAPPENDIXA 1. Kiser,J.V.L.,andD.B.Sussman,MunicipalWaste CombustionandMercury:TheRealStory. WasteAge,November 1991,Pp.41-44. 2. SRIInternational. 1991DirectoryofChemicalProducers: UnitedStatesofAmerica. SRIInternational,MenloPark, California. 1991. 3. Serth,R.W.,andT.W.Hughes. PolycyclicOrganicMatter (POM)andTraceElementContentsofCarbonBlackVentGas. Environ.Sci.Technol.,14(3):298-301.1980.

A-14 APPENDIXB SUMMARYOFCOMBUSTIONSOURCECADMIUMEMISSIONDATA

A-15 TABLE B-1. SUMMARY OF COAL COMBUSTION EMISSION DATA

Emission factor

kg/1015 J lb/1012 Btu Industry Facility Control Coal sectora typeb statusc typed Mean Range Mean Range

U PC/DB ESP B 1.1 -- 2.6 --

U PC/DB WS B 0.52 -- 1.2 --

U PC/DB MP/ESP B 0.82 -- 1.9 --

U PC/DB MP/ESP B 0.60 -- 1.4 --

U PC/DB ESP B 11 4.9-23 26 11-53

U PC/DB UN B 59 49-72 140 110-170

U PC/DB ESP B 2.8 -- 6.6 --

U PC/DB ESP B 4.2 -- 9.8 --

U PC/DB ESP B 1.6 -- 3.8 --

U PC/DB UN B 18 -- 41 --

U PC/DB UN B 5.2 -- 12 --

U PC/DB UN B 4.7 -- 11 --

U PC/DB ESP B 1.9 -- 4.5 --

U PC/DB ESP B 3.1 -- 7.1 --

U PC/DB UN B 4.3 -- 10 --

U PC/DB UN B 4.0 -- 9.2 --

U PC/DB UN B -- 4.3-6.0 -- 10-14

U PC/DB ESP B <2.0 -- <4.6 --

U PC/DB ESP/WS B <2.0 -- <4.6 --

U PC/DB MP B 130 59-210 290 140-490

U PC/DB MP/ESP B 20 -- 46 --

U PC/DB VS B 0.84 -- 2.0 --

U PC/DB ESP B -- 0.095-0.26 -- 0.22-0.60

U PC/DB MP B 13 6.5-24 31 15-56

U PC/DB UN B 18 10-32 42 24-74

U PC/WB MP/ESP B 0.82 -- 1.9 --

U PC/WB ESP B 0.24 -- 0.56 --

B-1 TABLE B-1. (continued)

Emission factor

kg/1015 J lb/1012 Btu Industry Facility Control Coal sectora typeb statusc typed Mean Range Mean Range

U PC/WB ESP B 0.27 -- 0.63 --

U PC/WB VS B 0.037 -- 0.086 --

U PC/WB ESP B 0.60 -- 1.4 --

U PC/WB ESP B 1.1 -- 2.6 --

U CY WS B 210 -- 490 --

U CY ESP B 1.3 -- 3.0 --

U CY ESP B 0.47 -- 1.1 --

U CY ESP B 0.15 -- 0.35 --

U CY ESP B 0.47 -- 1.1 --

U CY UN B 12 9.5-15 28 22-35

U CY ESP B 0.34 0.30-0.39 0.80 0.70-0.90

U S FF B 0.14 -- 0.33 --

U S MP B 1.8 -- 4.2 --

U S MC B 9.5 -- 22 --

U CY UN SB 1,900 -- 4,400 --

U CY WS SB 210 -- 490 --

U PC VS SB 1.7 -- 4.0 --

U PC ESP SB <0.17 -- <0.40 --

U NA ESP SB 0.17 -- 0.39 --

U NA ESP SB 0.73 -- 1.7 --

U PC/DB MC L 11 -- 26 --

U PC/DB MC L 2.2 -- 5.1 --

U PC/DB ESP L <1.5 -- <3.5 --

U CY ESP L 0.52 -- 1.2 --

U CY CY L 6.9 -- 16 --

U CY ESP/WS L 13 0.77-25 31 1.8-59

B-2 TABLE B-1. (continued)

Emission factor

kg/1015 J lb/1012 Btu Industry Facility Control Coal sectora typeb statusc typed Mean Range Mean Range

U SS MC L 2.3 -- 5.3 --

U SS ESP L 0.82 -- 1.9 --

I PC/DB ESP B 8.6 -- 20 --

I PC/DB ESP B 0.21 -- 0.49 --

I PC/DB MC B 200 -- 460 --

I PC/DB MC/WS B 0.42 -- 0.98 --

I PC/DB UN B 124 -- 290 --

I PC/DB ESP B 17 -- 39 --

I PC/WB MC B 0.65 -- 1.5 --

I SS MC/ESP B 0.0039 -- 0.0090 --

I SS MC B 0.082 -- 0.19 --

I SS MC B 0.40 -- 0.93 --

I SS UN B 2.1 1.8-2.4 4.8 4.1-5.6

I SS UN B 8.6 6.9-9.4 20 16-23

I SS UN B 15 -- 35 --

I SS UN B 3.7 -- 8.7 7.4-10

I SS UN B 9.5 -- 22 20-25

I SS UN B 19 -- 45 25-65

I OS UN B 16 -- 37 --

I OS UN B 5.2 -- 12 --

I OS UN B 77 26-130 180 60-300

I OS UN B 43 -- 100 --

I OS MP B 24 19-29 56 44-67

I SS UN B 5.6 -- 13 --

I SS ESP B 0.56 -- 1.3 --

I SS UN B 4.7 -- 11 --

B-3 TABLE B-1. (continued)

Emission factor

kg/1015 J lb/1012 Btu Industry Facility Control Coal sectora typeb statusc typed Mean Range Mean Range

I SS ESP B 1.8 -- 4.2 --

I SS UN SB 6.0 2.1-9.9 14 4.9-23

I SS UN SB 34 -- 78 --

I SS MP/ESP SB 2.5 -- 5.7 --

I SS UN SB 120 -- 290 --

I SS MP/ESP SB 6.0 -- 14 --

C PC/DB UN B 5.5 -- 13 --

C PC/DB MC/WS B 0.15 -- 0.35 --

C SS MP B 2.4 -- 5.6 --

C OS MP B 0.52 -- 1.2 --

C S UN A 0.99 -- 2.3 --

C S UN A 1.5 -- 3.5 --

C S UN A 0.60 -- 1.4 --

R NA UN B 67 -- 160 --

R NA UN B 13 -- 31 --

R S UN B 3.8 -- 8.9 --

R S UN B <19 -- <44 -- aU = utility, I = industrial, C = commercial, R = residential. b PC = pulverized coal, DB = dry bottom, WB = wet bottom, CY = cyclone, NA = not available, SS = spreader stoker, OS = overfeed stoker, S = stoker. cESP= electrostatic precipitator, WS = wet scrubber, MP = mechanical precipitation device, UN = uncontrolled, VS = venturi scrubber, FF = fabric filter, MC = multiclone, CY = cyclone. dB = bituminous, SB = subbituminous, L = lignite, A = anthracite.

B-4 TABLE B-2. SUMMARY OF MUNICIPAL WASTE COMBUSTOR EMISSION DATA

Combustor Control Concentration a b µ c Facility name type technology g/dscm @ 7% O2 Commerce MB/WW UN 960 Commerce MB/WW UN 1,600 Commerce average MB/WW UN 1,280 Quebec City - Pilot MB/WW UN 1,000 Quebec City - Pilot MB/WW UN 1,500 Quebec City - Pilot MB/WW UN 1,200 Quebec City - Pilot MB/WW UN 1,300 Quebec City - Pilot MB/WW UN 1,100 Quebec City - Pilot MB/WW UN 1,200 Quebec City average MB/WW UN 1,220 Vancouver MB/WW UN 1,200

Babylon MB/WW SD/FF 1.00 Babylon MB/WW SD/FF 5.00 Babylon MB/WW SD/FF bd Babylon average MB/WW SD/FF 2.00 Bridgeport MB/WW SD/FF bd Bridgeport MB/WW SD/FF 4.00 Bridgeport MB/WW SD/FF bd Bridgeport average MB/WW SD/FF 1.33 Bristol MB/WW SD/FF 2.00 Bristol MB/WW SD/FF 1.00 Bristol MB/WW SD/FF 2.00 Bristol average MB/WW SD/FF 1.67 Commerce MB/WW SD/FF 0.400 Commerce MB/WW SD/FF 2.00 Commerce average MB/WW SD/FF 1.20 Fairfax MB/WW SD/FF 9.00 Fairfax MB/WW SD/FF 6.00 Fairfax MB/WW SD/FF 6.00 Fairfax MB/WW SD/FF 5.00 Fairfax average MB/WW SD/FF 6.50 Gloucester MB/WW SD/FF bd Gloucester MB/WW SD/FF bd Gloucester MB/WW SD/FF bd Glouster average MB/WW SD/FF 0.00 Hempstead MB/WW SD/FF bd Hempstead MB/WW SD/FF bd Hempstead MB/WW SD/FF bd Hempstead average MB/WW SD/FF 0.00 Kent MB/WW SD/FF 4.00 Kent MB/WW SD/FF 4.00 Kent average MB/WW SD/FF 4.00 Long Beach MB/WW SD/FF 18.0 Marion County MB/WW SD/FF 3.00 Stanislaus County MB/WW SD/FF 2.00 Stanislaus County MB/WW SD/FF 2.00 Stanislaus County MB/WW SD/FF 2.00 Stanislaus County average MB/WW SD/FF 2.00

B-5 TABLE B-2. (continued)

Combustor Control Concentration a b µ c Facility name type technology g/dscm @ 7% O2 Haverhill MB/WW SD/ESP 38.0 Haverhill MB/WW SD/ESP 18.0 Haverhill MB/WW SD/ESP 10.0 Haverhill average MB/WW SD/ESP 22.0 Millbury MB/WW SD/ESP 13.0 Millbury MB/WW SD/ESP 22.0 Millbury MB/WW SD/ESP 32.0 Millbury MB/WW SD/ESP 6.00 Millbury MB/WW SD/ESP 18.0 Millbury MB/WW SD/ESP 7.00 Millbury MB/WW SD/ESP 11.0 Millbury average MB/WW SD/ESP 15.6 Portland MB/WW SD/ESP 4.00 Portland MB/WW SD/ESP 4.00 Portland average MB/WW SD/ESP 4.00

Pinellas County MB/WW ESP 8.00 Tulsa MB/WW ESP 390 Tulsa MB/WW ESP 140 Tulsa average MB/WW ESP 265

Vancouver MB/WW DSI/FF 4.00

Dutchess County MB/RC DSI/FF 3.00 Dutchess County MB/RC DSI/FF 3.00 Dutchess County average MB/RC DSI/FF 3.00

Dayton MB/REF UN 1,200 Dayton MB/REF UN 1,100 Dayton MB/REF UN 1,950 Dayton MB/REF UN 1,300 Dayton MB/REF UN 1,500 Dayton average MB/REF UN 1,410

Dayton MB/REF ESP 30.0 Dayton MB/REF ESP 19.0 Dayton average MB/REF ESP 24.5

Dayton MB/REF DSI/ESP 11.0

Biddeford RDF UN 1,100 Mid-Connecticut RDF UN 500 Mid-Connecticut RDF UN 567 Mid-Connecticut RDF UN 1,100 Mid-Connecticut RDF UN 600 Mid-Connecticut RDF UN 617 Mid-Connecticut average RDF UN 677

B-6 TABLE B-2. (continued)

Combustor Control Concentration a b µ c Facility name type technology g/dscm @ 7% O2 Biddeford RDF SD/FF bd Mid-Connecticut RDF SD/FF bd Mid-Connecticut RDF SD/FF bd Mid-Connecticut RDF SD/FF bd Mid-Connecticut RDF SD/FF bd Mid-Connecticut average RDF SD/FF 0.00

Semass RDF SD/ESP 10.0 Semass RDF SD/ESP 7.00 Semass average RDF SD/ESP 8.50

Detroit RDF ESP bd Detroit RDF ESP bd Detroit RDF ESP bd Detroit average RDF ESP 0.00 Albany RDF ESP 33.7

St. Croix MOD/EA DSI/FF 2.00

Dyersburg MOD/SA UN 238 N. Little Rock MOD/SA UN 360

Barron County MOD/SA ESP 220 Oneida County MOD/SA ESP 920 aMB = mass burn, WW = water wall, RC = rotary water wall, = REF = refractory wall, RDF = refuse- derived fuel-fired, MOD = modular, SA = starved air, EA = excess air. bUN = uncontrolled, SD = spray dryer, FF = fabric filter, ESP = electrostatic precipitator, DSI = dry sorbent injection. cbd = below detection limit.

B-7 TABLE B-3. SUMMARY OF SEWAGE SLUDGE INCINERATOR EMISSION DATA

Emission factor

-3 Incinerator typea Control statusb g/Mg dry sludge 10 lb/ton dry sludge

FB IS 0.15 0.30

FB VS/IS 0.55 1.1

FB VS/IS 1.4 2.9

FB VS/IS 0.27 0.55

FB VS/IS 0.0035 0.0070

MH UN 49 98

MH UN 0.0010 0.0020

MH UN 5.3 11

MH UN 51 100

MH VS 0.17 0.35

MH VS 0.65 1.8

MH IS 1.5 3.0

MH IS 1.2 2.4

MH VS/IS 1.9 3.8

MH VS/IS 7.8 16

MH VS/IS 2.7 5.4

MH VS/IS 0.32 0.64

MH VS/IS/AB 2.1 4.2

MH CY 32 65

MH CY 0.86 1.7

MH CY 4.4 8.8

MH CY/VS 25 50

MH CY/VS/IS 8.1 16

MH ESP 0.17 0.35

MH FF 0.014 0.027 aMH = multiple hearth, FB = fluidized-bed. bIS = impingement scrubber, VS = venturi scrubber, UN = uncontrolled, AB = afterburner, CY = cyclone, ESP = electrostatic precipitator, FF = fabric filter.

B-8 TABLE B-4. SUMMARY OF MEDICAL WASTE INCINERATOR EMISSION DATA

Emission factor

g/Mg of waste 10-3 lb/ton of waste Waste Control No. of Facility typea statusb runs Average Range Average Range

Fox Chase M VS/PB 3 9.95 -- 19.9 --

Southland M DSI/ESP 3 0.297 0.264-0.360 0.593 0.528-0.719

Royal Jubileec M UN 2 1.31 0.951-1.66 2.61 1.90-3.32

Mega NA VS/PB 3 2.93 1.16-6.27 5.86 2.33-12.5

St. Bernadines M UN 3 0.619 0.552-0.703 1.24 1.10-1.41

Sutter (1988) M UN 6 1.05 0.420-1.86 2.11 0.840-3.71 Sutter (1987) M UN 3 1.12 0.118-2.44 2.25 0.237-4.88

Stanford M UN 3 1.01 0.475-1.47 2.02 0.949-2.93 M VS 3 0.751 0.558-0.926 1.50 1.12-1.85

St. Agnes M UN 3 1.59 1.24-2.04 3.19 2.48-4.09

Cedars Sinai M UN 3 2.22 1.45-3.40 4.44 2.90-6.80 FF 3 <0.00258 <0.00245- <0.00516 <0.00490- <0.00265 <0.00531

Nazareth M VS/PB 2 1.48 0.714-2.24 2.96 1.43-4.49

Kaiser M WS 3 2.57 1.06-5.31 5.13 2.12-10.6

USC M UN 3 2.60 0.921-3.69 5.20 1.84-7.38

Borgess G500 UN 14 4.39 1.27-21.9 8.77 2.54-43.9 DI/FF 9 0.0101 0.00615-0.0147 0.0203 0.0123- 0.0294 DI/FF+Cd 2 0.0152 0.0150-0.0154 0.0303 0.0299- 0.0307 DI/FF+Ce 3 0.0723 0.00642-0.195 0.145 0.0128- 0.391 RB UN 10 1.63 0.722-2.52 3.26 1.44-5.03 DI/FF 9 0.0131 0.00959-0.0156 0.0261 0.0192- 0.0313 G100 UN 2 1.56 0.792-2.33 3.12 1.58-4.66

University of M UN 3 4.32 3.26-5.86 8.65 6.53-11.7 Michigan VS/PB 3 2.69 2.11-3.81 5.38 4.22-7.63

Lenoir M UN 9 3.04 1.05-6.78 6.07 2.10-13.6

Cape Fear M UN 9 5.68 3.90-6.74 11.4 7.81-13.5

B-9 TABLE B-4. (continued)

Emission factor

g/Mg of waste 10-3 lb/ton of waste Waste Control No. of Facility typea statusb runs Average Range Average Range

AMI Central M UN 3 0.461 0.403-0.575 0.923 <0.807-1.15 Carolina <0.000-4.67 <0.000-9.34 P UN 6 0.902 1.80

Morristown M UN 6 6.86 4.72-8.25 13.7 9.43-16.5 SD/FF 3 0.0213 0.0104-0.0293 0.0427 0.0208- 0.0587 SD/FF+C 3 0.0121 0.00955-0.0145 0.0242 0.0191- 0.0290 aM = mixed medical waste, NA = not available, G500 = mixed waste from 500-bed hospital, RB = red bag waste, G100 = mixed waste from 100-bed hospital, P = pathological waste. bVS = venturi scrubber, PB = packed bed, DSI = duct sorbent injection, ESP = electrostatic precipitator, UN = uncontrolled, FF = fabric filter, WS = wet scrubber, DI = dry injection, C = carbon addition, SD = spray dryer. cSampling method suspect, results biased low. dCarbon injection at 1 lb/hr rate. eCarbon injection at 2.5 lb/hr rate.

B-10