Document v1mKDmpkrVMGDEddwKyxQ5NR

t ICCeoOuhMtcUi L ROBERT P* MASTINOS lionamq S janO^Jky CHARLES M WAUHtA LOS ANQcLCS OFFICE S5S SOUTH fk-OWtW STREC1 LOS ANOCLC&, CALIFORNIA 00071 TEkCFHONc t2iat 003-0000 oranoc county office eos TOWN CENTER DRIVE COSTA MESA. CALIFORNIA 92626 TELEPHONE (71*1 64HIOO west los anGel.es OFFICE (200 OCEAN AVENUE SANTA MONICA, CALIFORNIA 00*01 TELEPHONE (213) *51-1200 WRITER'S DIRECT DiAl NUMBER 202-457-9423 LAW OFFICES OF Paul. Hastings. Janofsky & Walker *. PARTNCAjMl* INCkUOlNfi MOftiHONAL CQKPOAAtlONt twelfth FLOOR IOSO CONNECTICUT AVENUE. N. W, WASHINGTON, D. C. 20036 TELEPHONE (2021 223-9000 TWX, 7)0 - 22 - 9062 TELECOPIER (202) *52-l*0 February 5, 1991 ATLANTA OFflCC georgia`P4Cific center 133 PEACHTREE STREET, n e ATLANTA, GEORGIA 30303 TELEPHONE ( + 0*1 $60-9900 CONNECTICUT OFFICE loss WASHINGTON BOULEVARO STAMFORD, CONNECTICUT 06901 TELEPHONE (203) 96i-7*QO NCw YORK OFFICE 9 WEST 37- STREET NCW YORK. NEW YORK (Oo9 TELEPHONE (212) 032-6100 TOKYO OFFICE TORANOMON OHtORl BUILDING *-3. TORANOMON I-ChOME MINATO-KU TQKYO lOS TELEPHONE (03) S07-0730 OUR file NO Jonathan Busch Chemical Manufacturers Association 2501 M Street, N.W. Suite 330 Washington, D.C. 20037 i 111* Dear Jon: Enclosed is the EPA document from which excerpts were taken by EPA for placement in the Hazardous Organic NESHAP docket. Sincerely, RBD:cs R. Bruce Dickson For Distribution fcv CMA CHEMSTAR DIVISION Ftntn, _------ ------------ nd. No_JLEt_anlii^i Din ?/*'/(I------------------------------- SL 063620 Locating and Estimating Air Emission? from Sources of Vinylidene Chloride (U.S.) Environmental Protection Agency Research Triangle Park, NC Sep 85 PB86-117611 [ 0.1 e1 Commerce ferioari Tedwcal bvformadon Service 1 SL 063621 r SEB\ United States Env.ranrr.entai ?*otac!ion Agent-/ 0**Kt Of Air Quality Planning and Standards Pesaaren Triangle sark NC 2771 ' EPA-4S0 4-8A.C07k Seoiamoer 1985 Locating And Estimating Air Emissions From Sources Of Vinyiidenr Chloride NATIONAL TECHNICAL INFORMATION SERVICE SL 063622 --------------------------------------------------------1---------- ------------------------------------ -------------- -- TECHNICAL REPORT 3ATA raa j/i :nt rtatnt ot/ort Lomruj i ^6corso EPA-450/--84~007k 2. 3 CiH.sT $ ACS5CSION SO PB8 6 11 7 61 1 ; * TlT.e AN0 S'-BTlT,.* ,5. REPORT OAT Locating And Estimating Air Emissions From Sources Of Vinyiidene Chloride j September 19So 'i 9a*0*MI*l 3 ORC AN.CA* CN tOOE 1 ' i.'-OP s: 'i, afl?C^v:SO :h3*;,'1(;a: C` 1 ,z j *o Ea*C = M.`:C .]*<; A" o Acr**ss: o \: | i * ^ ^ iijcSw''' S6 4 0 ^ ^ 53 ! Office Of Air "ualitv banning And Standards ii V. 3. Environmental Fro; eccion Aaenc" Research Triar.ale 3arh, VC 2"71' '3 Swf L:Vlt\Tiav NOTES ?A Proiecc Officer: Thomas F, Lahre i 1 1 13. T /PE OS 03' A lC JaC0 CO , E A 0 ID -4 s*cnsc*<\o AO = sc ' i To assise groups interested in inventorying air emissions of various potentially toxic substances, ?A is preparing a series of documents such as this to compile available information on sources and emissions of these substances. This document deals specifically with vinyiidene chloride. Its intended audience includes Federal, State and local air pollution personnel and others interested in locating potential emitters of vinyiidene chloride in making gross estimates of air emissions therefrom. This document presents information on 1) the types of sources chat may emit vynilidene chloride, 2) process variations and release points that may be expected within these sources, and 3) available emissions information indicating the potential for vinyiidene chloride release into the air from each operation. vr ____ 4. OtSCWl^TOffS v wo*os Ai*o nocuMtNT iNAi fjis b.iOENTiPfERS.CPEN ENOEO terms Vinyiidene Chloride Emissions Sources Locating Air Emission Sources Toxic Substances CCSaTi 1 teld, Croup is 3'5^'3VT OSri'EV;*r IB SECUR.T- CLASS This Reporti 20 SECURITV class , True papet PAPe,A222C-.:.3*. *.,'71 *< ul 13. ""3 "i I Jtio.ct 2\ NO. OF G5 74 22 PRICE SL 063623 EPA-450/4-84-007k September 1985 Locating And Estimating Air Emissions From Sources Of Vinylidene Chloride U S ENVIRONMENTAL PROTECTION agency Office Of Air And Radiation Office Of Air Quality Planning And Standards Research Triangle Park, North Carolina 2771 1 September 1985 SL 063624 This report has been reviewed by the Office Of Air Quality Planning And Standards. U S Environmental Protection Agency, and has been approved for publication as received from the contractor Aooroval does not signify that the contents necessarily reflect the views and policies of the Agency, neither does mention of trade names or commercial products constitute endorsement or recommendation for use EPA-450. 4-84-QO7k ii SL 63625 CONTBiTS Figures................................................................................................................................... Tables. .... ......................................... ...... ................... ..... iv v 1. Purpose of Document...................................................................................... 2. Overview of Doemenc Contents................................................................ 3. Background........................................................................................................ Nature of Pollutant ...... .............................................. Overview of Production and Uses.............................................. 4. Vinylidcne Chloride emission Sources .............................................. Vinylidene Chloride Production.................................................. Perchloroethylene and Trichloroethylene Production...................... 1.1.1-Trichloroethane Production.............................................. Polymerization of Vinylidene Chloride ................................ Use of Vinylidene Chloride in Specialty Chemical Production................................................................ Vinylidene Chloride Copolymer Fabrication ...... Volatilization From Waste Treatment, Storage and Disposal...................................................................................... 5. Source Test Procedures............................................................................. 1 3 5 5 7 10 10 19 28 33 44 45 33 55 References............................................................................................................................... 58 Appendix - Fugitive Emission Calculations for Vinylidene Chloride Production From1,1,2-Trichloroethane ................................... A-l References for Appendix................................................................................................. A-8 SL 063626 FIGURES 1 Chemical use tree for vinylidene chloride. ......... 8 2 Basic operations chat may be used for vinylidene chloride production from 1,1,2-trichloroethane.................. 12 3 Basic operations that may be used in perchlcro- ethylene and trichloroethylene production by chlorination of ethylene dichloride ......................................... 20 4 Basic operations that may be used in perchloroethylene and trichloroethylene production by oxychloTination of ethylene dichioride........................................................................ 23 5 Basic operations that may be used in the production of 1,1,1-trichloroethane from ethane......................................... 30 6 Basic reactions involved in the polymerizatioo of vinylidene chloride with a comonomer. . ....................... .... . 34 7 Basic operations that may be used for the production of vinylidene chloride copolymers . ......................................... 37 8 Basic operations that may be used in the coating of cellophane with bigb-VDC copolymer. .......... 46 9 Method 23 sampling train............................................. ............................... 56 SL 063627 t } TABLES ( I Nu?Sr l P*ge 1 Physical and -henical Properties of 7inylidene I Chloride............................................................... .................................... 6 2 Estimated Controlled and Uncontrolled Vinylidene Chloride Emission Factors for a Hypothetical Vinylidene Chloride Production Facility ....................................................... 14 3 Producers of Vinylidene Chloride. .- .................................................. 18 4 Estimated Controlled and Uncontrolled Vinylidene Chloride Emission Factors for Hypothetical Percblcroethylene/ Trichloroethylene Production Processes ................................ 23 5 Facilities Producing Perchloroethylene and/or Trichloroethylene............................................. 27 6 Facilities Producing 1,1,1-Tricfcloroethane.......................................... 32 7 Vinylidene Chloride Copolymers, Production Methods, and Applications....................................................................................... 36 8 Estimated Controlled Vinylidene Chloride Emission Factors for a Hypothetical Vinylidene Chloride Polymerization Plant................................................. 40 9 Potential Emission Controls for Polymer Plants....... 42 10 Facilities Producing Polyvinylidene Chloride......................................43 11 Estimates of Uncontrolled Emission Factors from High-VDC Copolymer Fabrication Processes ............................ 49 12 Facilities Fabricating High Vinylidene Chloride Copolymers.....................................................................................................51 A-l Estimated Process Lint Composition in VDC Production. . . . A-3 A-2 Estimated VDC Emissions from Valves and Pumps................................A-S A-3 Estimated VDC Emissions from Relief Valves..................................... A-6 A-4 Fugitive Emission Controls end Estimated Controlled Emission Races............................................................... A-7 SL 063628 SECTION 1 PURPOSE OF DOCUMENT EPA, States and local aii pollution contiol agencies are becoming increasingly avare of the presence of substances in the ambient air that may be toxic at certain concentrations. This awareness, in turn, has led to attempts to identify source/receptor relationships for these substances and to develop control programs to regulate emissions. Unfortunately, very little information is available on the ambient air concentrations of these substances or or. the sources that may be discharging them to the atmosphere. To assist groups interested in inventorying air emissions of various potentially toxic substances, EPA is preparing a series of documents such as this that compiles available information on sources and emissions of these substances. This document specifically deals with vinylidene chloride. Its intended audience includes Federal, State and local air pollution personnel and others who are interested in locating potential emitters of vinylidene chloride and making preliminary estimates of air emissions therefrom. Because of the limited amounts ef data available on vinylidene chloride missions, and since the configuration of many sources will not be the same as those described herein, this document is best used as a primer to inform air pollution personnel about 1) the types of sources that may emit vinylidene chloride, 2) process variations and release points that may be expected within these sources, and 3) available emissions information indicating the potential for vinylidene chloride to be released into tnc air from each operation. The reader is strongly cautioned against using the emissions information contained in this cocument to try to develop .in exact assessment of emissions from any particular facility. Since insufficient data are available to develop statistical estimates of the SL 063629 Accuracy of these emission factors, no estimate can be made of the error that could result when these factors are used to calculate emissions from any given facility. It is possible, in some extreme cases, that orders-of-magnitude differences could result between actual and calculated emissions, depending on differences in source configurations, control equipment and operating practices. Thus, in situations wh?re an accurate assessment of vinylidene chloride emissions is necessary, source-specific information should be obtained to confirm the existence of particular emitting operations, the types and effectiveness of control measures, and the impact of operating practices. A source test and/or material balance should be considered as the best means to determine air emissions directly from an operation. SV> SECTION 2 OVERVIEW OF DOCUMENT CONTENTS As noted in Section 1, the purpose of this document is to assist Federal, State and local air pollution control agencies and others who are interested in locating potential air emitters of vinylidene chloride and making preliminary estimates of air scissions therefrom. Because of the limited background data available, the information summarized in this document does not and should not be assumed to represent the source configuration or emissions associated vith any particular facility. This section provides an overview of the contents of this document. It briefly outlines the nature, extent and format of the material presented in the remaining sections of this report. Section 3 provides a brief summary of the physical and chemical characteristics of vinylidene chloride, its commonly occurring forms and an overview of its production and uses. A chemical use tree summarizes the quantities of vinylidene chloride consumed in various end use cate gories in the United States. This background section may be useful to someone who needs to develop a general perspective on the nature of the substance and where it is manufactured and consumed. Section 4 focuses on major industrial souree categories that say emit vinylidene chloride to the air. This section discusses the pro duction of vinylidene chloride, its use as an industrial feedstock, and processes which produce vinylidene chloride as a byproduct. For each major industrial source category described in Section 4, example process descriptions and flow diagrams are given, potential emission points are identified, and available emission factor estimates are presented that show the potential for vinylidene chloride emissions before and after controls employed by industry. Using trade publications and other sources, individual companies are identified that are reported to be involved with either the production or use of vinylidene chloride. SL 063631 1 The final section of this document summarizes available procedures J or source sampling and analysis of vinylidene chloride. Details are not prescribed nor is any EPA endorsement given or implied to any cf these sampling and analysis procedures. Ac this time, E?A generally has not evaluated these methods. Consequently, this docment merely provides an overview of applicable source sampling procedures, citing references for those interested in conducting source tests. This -jcument does not contain any discussion of health or ocher environme :al effects of vinylidene chloride, nor does it include any discussion of ambient air levels or ambient air monitoring techniques. Comments on the contents or usefulness of this docmnent eze welcomed, as ij ary information on process descriptions, operating practices, control measures and emissions information thac would &r.ib a c?A to improve its concents. All comments should be sent to: Chief, Source Analysis Section (MD-1A) Air Management Technology Branch U.S. Environmental Protection Agency Research Triangle Park, N.C. 27711 SL 063632 SECTION 3 BACKGROUND NATURE OF POLLUTANT Vinylidene chloride (VDC) ia a clear liquid with low viscosity at room temperature, It haa a aweet odor similar to other chlorinated hydrocarbons, which can be detected at concentrations of about 300 ppm-* The chemical name for vinylidene chloride is 1,1dichloroethylene; it ia also refered to technically as vinylidene dichloride. The structure of VDC is illustrated below: Vinylidene chloride ia practically insoluble in water, but it soluble in most ocher polar and nonpolar solvents. It is very soluble is chloroform and ether, nd is soluble in benzene, acetone, and ethanol. Chemical and physical properties of VDC are summarized in Table l.*^ In the presence of air or oxygen, vinylidene chloride can fora a peroxiae compound that is violently explosive. The peroxide also initiates polymerization of the bulk VDC. Commercial grades of VDC typically contain about 200 ppm of hydroquinonc monoethyl ether (MEHQ) inhibitor, which prevents the formation of peroxide and aponeancous polymerization. Other impurities in coamereial grade vinylidene chloride include trans-l,2-diehloroethylene (900 ppa), vinyl chloride (830 ppm), 1,1,1-erichloroethane (130 ppm), cis-1,2-dichloroothylene (10 ppa), 1,1-diehloroethane (<10 ppa), ethylene chloride (<10 ppa), and trichloroethylene (<10 ppm).*'^ Vinylidene chloride liquid it very volatile, with a vapor pressure of 660 mm Hg at room teeperature. Its vapor burns readily when ignited. The flash point of the liquid is about -15'C, the lower explosive lisic of the vspor in air is 7 percent, and the upper explosive limit is 16 5 SL 063633 TABLE 1. P\:3ICAL AND CHEMICAL PROPERTIES OF VINYLIDENE CHLORIDE2 Synonym*: VDC, 1,1,-diehloroethylene, 1,I-dichloroethere, vinylidene dichloride Chemical formula CI2C - ce2 CAS registry number 75-35-4 Molecular weight, g/mole 96.9 Density C20'C liquid), g/cm 1.2137 Selling point, *C 31.56 Melting point, 'C -122.56 Flash point, *C open cup closed cup -16 -26 Autoigniticn temperature in air, *C 513* Flammable limits in air, volume percent 5.6-16.0 Latent heats, kJ/mole vaporization (at boiling point) fusion (at freezing point) 26.46 6.51 Heat of combustion (25*C liquid), kJ/mole 1095.9 Heat of polymerization (25'C), kJ/mole -75.3 Heat of formation, kJ/mole liquid vapor Heat capacity, J/mole-K liquid vapor (25*0 Critical properties Temperature, *C Pressure, MPa Volume, cs^/aol* Vapor pressure, 0*C 10*C 20*C :o*c k?a Water solubilities at 20*C, g/lOOg Vinylinene chloride in water Water in vinylidene chloride Dielectric constant (16*C liquid) Viscosity (20*0, centipoise -25.1 1.26 111.27 67.03 280.8 5.21 218 28.92 44.54 66.34 95.91 0.25 0.035 4.67 0.33 *VDC stabilized by MEHQ. SL 063634 6 percent. The decomposition products of VDC exposed to oxygen include formaldehyde, phosgene, and hydrogen chloride. 1 ' 2 The residence time of vinylidene chloride in the atmosphere is about 23 hours, where residence time is defined as the time required for the concentration to decay to 1/e (372) of its original value. The major mechanism frr destruction of VDC in the atmosphere is reaction with hydroxyl radicals. The principal reaction products an formic acid, carbon monoxide, chloroacetyl chloride, hydrogen chloride, phosgene, and formaldehyde.^'^ Vinylidene chloride can also be polymerized to produce polyviny1idene chloride (PVDC) polymer chains made up of monomer units joined head to tail; H Cl H Cl E Cl > ti tii -- c-c-c-c-c-c -- til'll H Cl E Cl a Cl Vinylidene chloride can also be polymerized with other Donomers to produce polyvinylidene chloride copolymers, OVERVIEW OF PRODUCTION AND USES Vinylidene chloride was first used in the late 1930'* by Dow Chemical Company. VDC is produced commercially by the dehydrochloriaation of 1,1,2-trichloroethane with lime or caustic.^ It may also be recovered as a byproduct of chlorination and oxychlorination reactions to produce other compound*.^ Today, production of VDC in the United States exceeds 90,700 megagrams (Kg) per year.^ Exact production figures are not available because the producers of VDC consider these to be proprietary. Figure 1 gives a chemical use tree summarizing the production and use of VDC. 2 ' 3 5-7 The main use of VDC is in the production of VDC copolymers. About 68,000 Mg of VDC are consumed annually in the pro- 2 duct ion of polymers containing VDC. In the United States, the generic term "Satan" is used to refer to high VDC-content polymers. Saran formerly was a trademark of the Dow Chemical Company and is still a Dow trademark in other countries. PVDC homopolymer (Saran A) is difficult to fabricate and for this reason is not used. However, copolymers of 7 SL 063635 c: \/ AiGH C -- Cl ------- K 0 * HaCl /\ 2 5H K Cl S/ c*c /\ H Cl VDC I VDC ?wouct:q I v;c 513 M Cl \/ cc /\ H Cl VOC / H othar vinyl lonoaar haat miiiator Kisft-WOC eapelvur* or Lo-VDC capelyanr* - Sarin 'ilia Ja.'Ul eaatinya Him rtuciMi ru; Saeain? Synthatie ;iSara 0 Cl 2/ c Cl 20 cftloraaeotyl CAlonda launM:c.a m laa product.:, an of '.air in and pAaraacautieala Figure 1 Chemical use tree for vinylidene chloride.2.3,5-7 06363& 9 si* vinylidene chloride with viayl chloride (Saran B), alkyl acrylates (Saran C), and acrylonitrile (Saran F) are widely used. Polymers con taining VDC are resistant to photodegradation and chemical attack, and hecause of their high density and crystallinity, they are impermeable to a wide range of gases. The low permeability of VDC copolymers to water and oxygen is the main reason for the commercial importance of vinylidene chloride.^' In addition to the production of Saran polymers, VDC is also used as a chemical intermediate in the production of chloroacetyl chloride.^ Chloroacetyl chloride is a chemical intermediate in the production of pharmaceutical products and tear gas.^ Formerly, a major use of VDC (about 60,000 Mg/yr) was as an intermediate in the production of 1,1,1trichloroethane. However, the VDC-based process for 1,1,1trichloroethane was only used at one plane, and was replaced in the lac* 197Q's by a vinyl chloride based process.* 9 SL 063637 3CTIC:; 4 VINYLIDEKE CHLORIDE EMISSION SOURCES This section discusses vinylidene chloride emissions tree processes where the chemical is produced, processes where it is used as a chemical intermediate, and processes where it is produced as a byproduct. Process and emissions data are presented for each source category. The following industrial processes have bees identified as potential sources of VDC missions; vinylidene chloride production, e perchloroethylene and trichloroethylene production, 1,1,1-trichloroethane production, e VDC polymerisation, e use of VDC in specialty chemical production, VDC copolymer fabrication, and volatilization from waste treatment, storage, and disposal. VDC is a byproduct of perchloroethylene and trichloroethylene production and of 1,1,1-trichloroethane production from ethane. 7INYLIDHJJE CHLORIDE PRODUCTION Vinylidene chloride is produced domestically by the dehydrochlorination of 1,1,2-trichloroethane with sodium hydroxide.^'** Three plants in the U.S. produce VDC; each of these produces a number of other chlorinated hydrocarbons by a variety of processes,^ The raw material 1,1,2-trichloroethane is produced as coproduct in the chlorination and oxychlorination of ethane, ethylene, and ethylene bichloride (1,2,-dichloroethane) to produce chlorinated Cj species.^ At the plants using the 1,1,2- cricnloroethane dehydrochlorination 063&38 10 process, additional VDC nay alto be recovered at a byproduct of various chlorination and oxychiorination processes.^ These processes ar discussed in later sections. Process Description The reaction for the dehydroehlorination of 1,1,2* trichloroetkane to produce VDC is as follows: H Cl 'f H Cl \/ 1,1.2-cri chloroethane sodium hydroxide VDC The reaction is carried out with 2 to 10 percent excess caustic and product yields ranging from 85 to 90 percent.** Basic operations that cay be used in the production of VDC froo 1,1,2-trichloroethane are chovn in Figure 2. Concentrated sodiua hydroxide (Stress 1) is diluted with water (Stress 2) to about 5 to 10 weight percent and it mixed with the 1,1,2-crichloroechane feed (Stream 3) and fed (Stream 4) to the dehydtochlorination reactor. The reaction is carried out in the liquid phase at about 1Q0*C vlchout catalyst. Because the aqueous and organic reactants are not miscible, the reaction is carried out in e liquid dispersion. The dehydroehlorination reactor is continuously purged with nitrogen (Stream 5) to prevent the accumulation of monochloroacetylene impurity in th* product VDC. The nitrogen is discharged from Vent A.^ The VDC-contaiaing product from the dehydroehlorination reactor (Stream 6) is separated in a decanter into an aqueous phase (Stream 7} and an organic phase (Stream 8). The aqueous phase, comprising a sodium hydvoxide/sodiua chloride solution, is divided. One fraction (Stream 9) is recycled (Stream 4) to the hydrochlorination reactor, and the other fraction (Stream 10) is steam stripped to remove organics and discharged to a wastewater treatment system (Discharge F).11 The organics from the aqueous phase (Stream 11) are combined with the organic phase from the decanter (Stream 8). The combined organics (Stream 12) are fed to a drying column, where residual water is removed ss a bottoms stream (Scream 13). The water removed from the drying SL 063639 1 SJ fAOM finishing COLUMN FUGITIVE EMISSIONS OVERALL PLANT Figure 2 NOTE: The mirteri in Lhl* figure refer lo prote-ji rlrciHir ajk! the letter* devignate process venti. it diliuMed n the text, Basic oreratloits that May be used for vlnyljdcnc chloride production from 1,1,2-t richloroethane. SL 063640 column is fed to the steam stripper with the aqueous stream from the product decanter (Stream 10).** The organic stream from the drying column (Stream 14) is fed to a distillation column, which removes unreacted 1,1,2-trichloroethane as overheads (Stream 15). The unreacted trichloroethane is recycled to the dehydrochlorination reactor. Purified VDC product, removed as bottCES from the finishing column (Scream 16), is used onsite or stored in pressurized tanks before being shipped to users,** Emissions Uncontrolled VDC emission factors for VDC production are given in Table 2. The table also lists potentially applicable control techniques and associated controlled emission factots. The emission factors were developed based on published data for individual plants and for general processing techniques. Because of variation in process design, age of equipment, and other process parameters, actual emissions vary for each plant. Process Vent Emissions-- Process vents which are sources of VDC emissions include the reactor nitrogen purge vent (Vent A, Figure 2) and the distillation column vents (Vents B, Figure 2). Uncontrolled VDC emission factors are estimated at 6.2 kilograms VDC per megagram VDC produced (kg/Hg) for the reactor vent, ar.d 0.7 kg/Hg for the distillation vents.*^ Emissions from the reactor vent can be controlled by incineration with an efficiency of about 98 percent or higher.*4,*5.39 ^h# Mjop products of VDC incineration are CO2 and BC1. However, under poor incinerator operating conditions, other products may be formed, in cluding formic acid, carbon monoxide, chloroacetyl chloride, phosgene, and formaldehyde. Incineration destruction efficiency varies with emis sion stream properties and incinerator operating parameters. The 98 percent efficiency level is a conservative estimate of the control that may be expected at a temperature of at least 870*C anc a residence time of at least 0.75 seconds.*5,39 emission reduction may be greater than 98 percent for incineration of VDC with these operating parameters, SL 063641 13 TABLE 2. ESTIHATED CONTROLLED AND UNCONTROLLED V1NYLIDENE CHLORIDE EMISSION FACTORS FOP. A HYPOTHETICAL VINYLIDENE CHLORIDE PRODUCTION KAC1L1173 tat*aIon oeurco Source dealtnation^ VBC tala*ion factorc MftaitUI ly applicable control* Vricint raductloo Control lad VUC lalrrjiNi (actor Dlatlilotlon vent* tornp* and beadliny* Fugitive peace**! * A B B *.l kp/Hp 0.1 kp/Hp HA 0.H kp/hrS HA* Refrigerated ceirdeueer HA frtBiuviicd took efilpcrated condrnaai HA Quarterly l/H** Monthly I/m Monthly t/H; double tU on putapo; and rupture dioka oh it* <0.11 kp/Hp BA 0.0*1-0.010 kp/Hp 0 o.oto kg/Hs 0 0.010 kp/Hp HA 0.IP-0.IS kp/Hp HA m* HA HA 0.0}* kp/Hp 0.1* kp/kr it O.fo kp/hr 0.11 kp/hr *Aay plvaa tUflOtM chDrUa plant ) a| tat caafl|atattoa and level at control Iraa Mala KpaiUtllul facility. T>* itiiai ia tacoaia|ti ta caatact plant paraanaal to cant i rat tlac ealatenre af anilIlap operation* and total rut teclrno I opy at a particular facility priar ta aotlnotlnp caalaataina ttacrclirata. ^Letter* refer ta aanta iial|utf in Npuro I. *blaaian factara ta tataa af l|/N| (afar ta iilt|rMcf Inylidaaia chloride anil tad par aepaprea af vlnyltdcae clt I artda produced, la caaea uhaae a particular aourre deal(sat ion appltaa to aultiple opera! lone, thaaa lartora rapraaent coahinad aaiaatoaa lar all, not each, o( tltaaa operation* within Ilia hypothetical (acttlty. *Hot available. 'Jtandliag refer# ta leodlap af harp**, truck*, ralleara, etc. lOarivatioae af aalaalaa rataa aad potential etalaalon raductloaa are pivea In the AppendIa. rupitiva ralaelu* rata ia lodepeadant of ploat capacity. SThie rata rapraaente a relatively unceetrol led facility where na alpail leant leak dataclloo and repair pragma* are ia place ta Halt fuplliva eniaaiona. 1*1/11 refer# to loapectioa and aaiatenaoro. Indue!ry reporto Dial nore all infant I/It prupenaa are practiced than ere reflected la Ike reduction* In thie lob I*. (See tent lot dio~ua*ion.l three note air intro* araauir* reeull in fupitiv* eat**ion control efficiantito oa biph oa 10 to 15 percent.t*lF sl> 0636^2 *nd would also increase at higher temperatures and longer residence times.^ A 98 percent reduction of reactor venr emissions corresponds to a controlled VDC factor of about 0.12 kg/ Mg VDC produced. Con trolled reactor vent emissions reported for specific plants range from 0.063 kg/Mg to 0.090 kg/Mg.16 VDC emissions from the distillation column vents can be controlled either by aqueous scrubbing or by refrigerated vent condensers with an efficiency of about 90 percent.14,1^ The distillation column vents can also be combined with reactor missions and controlled by incineration with a 98 percent or greater control efficiency.^ The control efficiency attainable using these techniques is dependent on operating parameters and emission stream characteristics. A 90 percent reduction of distillation column vent missions corresponds to a controlled scis sion factor of about 0.07 kg/Mg VDC produced. Controlled VDC emissions reported for distillation vents at specific plants range f.rom 0.18 kg/Mg to 0.38 kg/Mg. Storage and Handling Emissions-- Vinylidene chloride emissions result from the storage of VDC product and intermediates containing VDC (Source D, Figure 2). VDC emissions also occur where VDC product is loaded to tank cars of trucks. Insufficient data were available to estimate uncontrolled emissions of VDC from srorage and handling. Controlled storage and handling emissions of VDC reported for two specific facilities were about 0.036 kg/Mg VDC produced.16 The types of controls used to attain this emission rate are not known. Controls typically used to reduce storage and handling emissions include pressurized storage tanks and refrigerated vapor condensers. Process Fugitive Emissions-- Fugitive emissions of VDC, 1,1,2-trichloroethane, and Cher volatile organic compounds result from leaks in process valves, pumps, compressors, and pressure relief valves (Source D, Figure 2). Fugitive (missions from a typical VDC plant were estimated based on process flow diagrams, process operation data, and emission factors developed by E?A SL 063643 15 for typical process emission sources. The techniques used to estimate the numbers of various fugitive mission sources and the total fugitive emission rate are described in detail in the Appendix. The estimated uncontrolled fugitive emission rate for VDC production from 1,1,2-trichloroethane is about 0.96 kg VDC/hr- Fugitive mission rates differ from plant to plant depending on the number of valves, pumps, flanges, etc., the age of the equipment, and the level of emission control used. Chmical process streams in the production of chlorinated hydrocarbons such as VDC generally contain chlorine and HC1, as veil as hydrocarbons. These compounds are extremely corrocive and irritating when exposed to the moisture in ambient air. Thus, it is general practice to control such fugitive missions in order to prevent corrosion of outside equipment and generation of unpleasant odors.* Table 2 gives control efficiencies for preventative maintenance programs, the use of double mechanical seals of pumps, and the use of rupture disks with relief valves. Other controls which may be used include the use of welded pipe in place of flanges, special construction materials for piping and valves, enclosure of pumps, and intensive preventative maintenance during plant shutdown. In addition, inspection and maintenance programs practiced at tome plants may be much more intensive than those shown on Table 2, such chat most leaks are repaired vithi*. as little as 1 day. With Chase additional controls, industry reports fugitive emission control efficiencies as high as 90 to 95 percent.*'*^ Secondary Emissions-- Secondary emissions result from the handling and disposal of process waste streams. In VDC production from 1,1,2-trichloroethane, wastewater from the VDC production reaction (Source F, Figure 2) is a potential source of secondary VDC missions. Specific data were not available to estimate emissions from the treatment and oisposal of this stream. Figure 2 shows that a wastewater stripper typically is incorporated as part of the VDC production process to reduce VDC missions and for product recovery.** Emirsions from treatment of 16 SL 063644 contaminated vastevater are discussed in further detail in the sectic ectir led VOLATILIZATION FROM WASTE TREATMENT, STORAGE, AND DISFOSAL. Source Locations ..ajor vinyiidene chloride producers and production locations are listed in Table 3.-3 SL 063645 TABLE 3. PRODUCERS OP VINYLIDENE CHLORIDE20 Manufacturer Dov Chttical O.S.A. PPG Industries, Inc. Chemicals Croup Chemical Division Location Freeport, TX Plaquecine, LA Lake Charles, LA Note: This listing is subject to change as market conditions change, facility ownership changes, plants are closed, etc. The reader should verify the existence of particular facilities by consult ing current listings and/or the plants themselves. The level of VDC emissions from any given facility is a function of variables such as capacity, throughput and control measures, tad should be determined through direct contacts with plant personnel. SL 063646 Id PEaCHLCROETHTLESE AfJD ."RICHLOSOiliiYLEKE PRODl'CTIC:; PerchloroethyLeoe (PCE) anti trichloroethylene (ICE) are produces separately or is coproduct* by either chlorination or oxycblorination or ethylene dichloride (EDC) or other Ci chlorinated hydrocarbon*. A number of byproducts are produced in each of these reactions, ins'.-ding viaylidene chloride (VDC). These byproducts cay be isolated and refined, recycled to the process, or discfcaged in various waste streacs. For instance, in a case where VDC is produced froo 1,1,2-tri.hlaroethar.e at a facility which also produces TCE and PCE, the VDC byproduct cay be recovered and purified by distillation in the VDC finishing section of the 1,1,2-trichleroethane process.*2 g4u caterial ratios and reactor conditions determine the relative proportions of PCE, TCE, and any byproducts produced. s Rescript icr.s Ethylene Dichloride Chlorination Process-- The overall reactions for the chlorination of EDC to produce TCE and PCE, are as follows; Cl Cl Cl Cl H - C - C - H 3C1, **!*- -- C C i\ * /i ah ci ci EEC PCI * 4HC1 byproducts Cl Cl '/ h- c - c -a a/ * h EDC 2C12 Cl Cl --* C C 3BC1 byproducts Cl H TCE VDC is among the byproducts produced in these reactions. Basic operations that may be used in the EDC chlorination process are shown in Figure 3.21 F.thylene bichloride (Scream l) and chlorine (Scream 2) are vaporized and fed to the reactor. Other chlorinated C2 hydrocarbons or recycled chlorinated hydrocarbon byproducts may also be fed to the IS SL 063647 nfcmt OrtCANlC $10HM*t Pit UMIUtt i Cj Cltt OHM* Itil HlWlIK S MtOM OK** 1-MOUSSfS HIAVI IMIS UX.UUM l*iS to INt.ltM It*) ION 1*t f SIU'UU. MHilllvi I M'iMi *_> vi um i PI Ml NQU: Ih* iim4icu in thll (i^nfc iUi In inJ the Icllcn dnitjiNic inuuu vmli. tUtMt ilnotiyli |Ik tir*nni, *% tliicimirtl In tins tenlt Iha tiaavy Hun lc|MCicitl I Ina I |troiluU Figure 1. Basic operations that May In* used III peri li lurucl ae rind L i'it Ti lorot'tliy 1 cno production by clilor tnat inn oi cthylisii* dltli <or Ido.' * SL 063648 reactor. In the reactor, a chlorination reaction is carried out at 400* to 450*C, slightly above atmospheric pressure. The reaction produces the desired products ICE 'and PCE, as well as byproducts, including vinylidene chloride and hydrogen chloride. Hydrogen chloride byproduct (Stream 3) is removed as an overhead stream from the chlorinated hydrocarbon mixture (Stream 4) produced in the reactor. The chlorinate! hydrocarbon mixture (Stream 4) is neutralized with sodium hydroxide solution (Stream 5) and dried,^ The dried crude product (Stream 7) is separated by a distillation column into crude TCE (Scream 8} and crude PCE (Stream 9). The crude TCE (Scream 3) is fed to two columns in series which remove light ends (Stream 10) and heavy ends (Stream 13). TCE (Stream 12) is taken overhead from che heavy ends column and sect to TCE storage; Che heavy ends (Stream 13) and the light ends (Stream 10) are combined, stored, and recycled-^* The crude (Stream 9) from the PCE/TCE separation column is sent to the PCE column, where PCE (Stream 14) is removed as an overhead scream to PCE storage. Bottoms from this column (Stream 15) are sent to a heavy ends column and separated into heavy ends and tars. Heavy ends (Stream 16) are scored and recycled, and tars are incinerated.^ Ethylene Dichloride Oxychloriaation Process-- The overall reactions for the production of perehloroethylene and trichloroethylene by EDC oxychlorination are as follows: Cl / C - H Cl, ^* H Cl CuCl. Cl Cl 2H2Q byproducts Cl Cl Cl Cl Cl \1 CuCl, N * H-C-C-H+l/2 /\ Cl1, 1/4 01,---------- C -C / t- 3/2 H,0 byproducts \* HH Cl H The crude product contains 85 to 90 weight percent PCE plus TCE, and 10 to 15 weight percent byproduct organics, including VDC. Essen tially all byproduct organics are recovered during purification and are 21 SL 063649 recycled to the reactor. The procese is very flexible, so that the reaction can be directed toward the production of either PCE or TCE in varying proportions. Side reactions produce carbon dioxide, hydrogen chloride, and several chlorinated hydrocarbons. Figure 4 shows basic opera lions that may be used in oxychlorination.* Ethylene dichloride (Stream 1), chlorine or hydrogen chloride (Stress 2), and oxygen (Stream 3) are fed in the gas phase to a fluid bed reactor. The reactor contains a vertical bundle of tubes with boiling liquid outside the tubes vhich maintains the reaction tempera ture at about 425*C. The reactor is operated at a pressure slightly above atmospheric, and the catalyst, which contains copper chloride, is continuously added to the tube bundle with the crude product. The reactor product stream (Stream 4) contains the desired products TCE and ?C, as well as byproducts including VDC. This stream is fed serially to a water cooled condenser, a refrigerated condenser, and a decante*'. The noncondensed inert gases (Scream 5), consisting of carb n dioxide, hydrogen chleride, nitrogen, and a small amount of uncondenscd chlorinated hydrocarbons, are fed to an absorber, where hydrogen chloride is recovered by absorption in process water to make byproduct hydrochloric acid (Stream 6). The remaining inert gases arc purged (Vent A). ^ In the decanter, Che crude product (Stream 7) is separated from the aqueous phase and catalyst fines (Stream 8) and sent to the drying column for r&oval of dissolved water by azeotropic distillation. The dried product (Scream 10) is separated into crude TCE (Stream 11) and crude FCE (Stream 12) in a PCE/TCE column. The aqueoua phase from the decanter (Stream 8) and the water from the drying colvsi (Stream 9) are sent to waste treatment. 21 The crude TCE (Stream 11) is sent to the TCE column, where light cuds (Stream 13) are removed to be stored and recycled. The bottoms (Stream 14), containing mainly TCE, are neutralized with ammonia and then dried to produce finished TCE (Stream 13) which is sent to the TCE storage. The crude PCE (Stream 12) from the PCE/TCE separation column is fed to a heavy ends removal column where PCE and light ends (Stream 16) go 22 SL 063650 OXYUh CHtOA.Hf OR HfOHOCfN--- 7" CNLOH1DC \J> fOC STORAOC C CltiOfllHJtrCO MrlANlCS rnuM 01 ut pHoctssts lHS 10 MClKtHAOOH Mil: Tt.* nu*h*M I,, tfcl. flour* ref*r t,, Jl"* **`e 1*1 tei iIciIjimu |iiotutt vcul > . LOADING ,, dilLu.leJ ,, I. .I ' Figure 4 Basic operations Hint may In* used in pen 11 it'i oel liy I ene and t 1 Ulil ut net Ivy 1 one production t>y oxychlorlunl ion of elliylem dIrhluiide. SL 063651 overhead to a PCE finishing column, and the heavy ends (Stieam 17) remaining as the bottoms are sent to the organic recycle system. Here, the organics that can be recycled (Scream 18) are separated from tars and sent to the recycle organic storage. The tars are incinerated. The PCI and light ends (Stream 16) from the heavy ends column are fed r.o a light ends removal column. Light ends (Stream 20) are removed overhead and are stored and recycled. The PCE bottoms (Stream 21) are neutralized with ammonia and then dried to obtain finished PCE (Stream 22) which is sent to PCE storage. Emissions Process Vent Emissions-- Table 4 gives emissions data for process vents in PCE and ICE production processes which say contain VDC. The table lists; uncontrolled VDC emission factors, potentially applicable control techniques, and associated controlled VDC emission factors.These emissions data were developed based on published information for general operations used in the production of PCE and TCE. Variations in process design, feed materials, and reaction conditions have a substantial effect on amounts of VDC and other byproducts produrcd by PCE and TCE production processes. As a result, VDC emissions vaty for each plant. Chlorination ttoceaa- Vents containing VDC in the ESC chlorination process include the neutralization and drying vent (Vent A, Figure 3), and distillation column vents (Vents B, Figure 3). Uncontrolled VDC emission factors are estimated at 2.3 kilog-.ams per megagram PCE and TCE produced (kg/Mg), for the neutralization and drying vent, and 0.106 kg/Mg PCE and TCE, for the distillation vcnt.^ Emissions from both of these sources can be controlled by refrigerated vapor condensers with an efficiency of about 80 percent.*4 The control efficiency attainable using refrigeration if 'ependent on emission stream characteristics and condenser operating temperatures. Oxvchlorinatioa process - Vent streams containing VDC in the EDC oxychlorination process include the reactor vent (Vent A, Figure 4), the drying column vent (Vent B, Figure 4) and the distillation column vents 24 SL 063652 TA1II.E . ESTIMATED CONTKOLEEI) AND UNCONTKOI.I.IT) V I NYU DENI. Clli.HU I l)E EMISSION I ACTOIIS I OK IlYroTItEI lCAE I'EHClILOKOKTIiy LENE/TK ICHI.UHOETIIYEENE I'HoDIICTI ON ITUICI SSI S'' Eauiion aource Source deaifnation^ Uucontfo tled VDC* vaiaiiun factor (k|/H|)c I'ott iu iil tp i||it i cable COU*lull Pricent ridiict too CuntioIled VIK' uaiaiion tact nr KeuuiiuAtion and dry inf vent OiitilUtioo vanta A 1.5 0.10b el Igctaled tottdcitaer Kuf I ig**i ated cuiiJcnact HO HO l>. 5(1 0.011 UiTthlorimt >9n K) teictor vent Oijrini coIum vent DtililUtimi venta A 1.* llitiail oxidation 98 <0.0)b 1 0.4 llierail oaidalion 981 <0.008 C 0.098 Ai)uct>ui tciukliiiif 91) 0.010 Any tiven parchloruethy laue/trirhloroelhy 1M plant MU vary in ronf Igui al ion and level ul runitul tin* t > hypothetic* I factjltiea. Tit reader ia encouraged to runtacl plaul |in iihiil- 1 tv evut Ha Ilia e* latrine uf rail! ini 0|>erat tune and control technology at a particular lacilily prior to ettivuttui enikaiona Ihertlina. bLel tera refer to vent* designated in Figure J fur chlorination nuJ Figure k for gN^cliluitmtiuii. 'Eaiaaion factort in taraa of kg/K| refer to kilogtawof vinyl idane cMui i>la milted per autaprau ul perrl>inr>il.y lena and tr i cli loroethy I ana produced, la caeea where a particular aoiirre Jec iinal inn nppl iea to nlti|-i< upeial kiu, theac factora repreerut coabiacd miaaiona (or alt, nut each, ul theae opera! nine villi in thu hypothetical laiilily. SL 063653 (Test C, Figure 4). Dncontrolled VDC amission factors have been estimated at 0.4 kg/Mg PCS and TCE produced, for the drying column vent, and 0.098 kg/Mg PCE and TCE, for the distillation column vent.22 The uncontrolled emissiou rate given is Table 4 for the reactor vent va* calculated based on a total chlorinated hydrocarbon emission factor for the reactor vent of 21.3 kg chlorinated hydrocarbon emitted per segagram ?CE and TCE produced.22 It was assumed that the fraction of the total chlorinated hydrocarbons comprised by VDC in the reactor vent is the same as the fraction of chlorinated hydrocarbons comprised by VDC in the drying column vent. In the drying column vent, VDC comprises about 18 percent of total chlorinated hydrocabon emissions. Thus, VDC emissions from the reactor vent were estimated as follows: <21.3 kg Cl species/Mg) x (0.18 kg VDC/kg Cl species) 3.8 kg VDC/Mg Emissions from the reactor and drying column vents can be controlled by incineration with an efficiency of about 98 percent or higher.Emissions from the distillation column vent can be controlled by aqueous scrubbing with an efficiency of about 90 percent.^ Other Emission Sources-- In addition to process vents, poce tial VDC emission sources from PCE and TCE production include storage and handling operations, process fugitive sources, and treatment and disposal of process wastes. VDC is expected ro be emitted from recycle storage tanks in both the chlorination and oxychlorination processes for PCE and TCE production (Vent C in Figure 3, and Vent D in Figure 4). VDC is also emitted from any pumps, compressors, flanges, and valves which are exposed to streams containing VDC (Source E in Figure 3 and Source I in Figure 4). Insufficient information is available to estimate VDC emissions from these sources. Source Locations Major producers of perchloroethylent and/or trichloroethylene are listed in Table 5.^ 26 SL 063654 TABLE 5. FACILITIES PRODUCING PERCHLOROETHYLENE AND/OR TRICHLOROETHYLENE20 Company Location Chemical produced PCEa TCb Diamond Shamrock Corp. Dow Chemical L'.S.A. E.I. duPont de Nemours and Co. PPG Industries, Inc. Stauffer Chemical Co. Vulcan Materials Co. Deer Park, TX Freeport, TX Pittsburg, CA Plaquemine, LA Corpus ChTisti, TX Lake Charles, LA Louisville, KYC Geiscar, LA Wichita, KB X XX X X X XX X X X aPCE perchlotoethylene bTCE * trichloroethylene c?lan: has been on standby since 1981. Note: This is a liat of major facilities producing perchloroethy lene and/or trichloroethylene by any production process. Current information on which of these facilities produce these chemicals by ethylene dichloride chlorination or oxychlorination is not available. This list is subject to change as market conditions change, facility ownership changes, or plants are closed down. The reader should verify the existence of particular facilities by consulting current listings or the plants themselves. The level of emissions from any given facility is a function of variables, such as operating conditions, product slate, throughput and control measures, and should be determined through direct contacts with plant personnel. Dnder some operating conditions, byproduct VDC production may be negligible, resulting in negligible or tero VDC emissions. For instance, Dow chemical has indicated that VDC Byproduct is only produced in trace quantities at all of its three plants.1* 27 SL 063655 1,1,1-TRICHLOROETHANE PRODUCTION Host 1,1,1-trichloroethane produced in the U.S. is Bade fron vinyl chloride. An ethane-based process currently is used to a lesser extent. Also, a process starting fron VDC formerly vas used to produce 1.1,1trichloroethane.^*^ This section discusses VDC emissions free the ethane process, which produces VDC as a byproduct. VDC is not produced as a byproduct of the vinyl chloride process. Process Description In the ethane process for 1,1,1-trichloroethane production, ethane is chlorinated to produce the product 1,1,1-trichloroethane and a nunbar of byproducts. The sain reactions in the chlorination process are as follows: CH3-CH3 > c;. CH3-CH2CI eci ethane ethyl chloride hyrodgen chloride CH3-CH2C1 Clj-A^. CH3-CHCI- HC1 1,1-dichloroethane ch3-ch2ci --*-ch2-ch2 HC1 ethylene ch3-chci2 ci2-^-ch3-cci3 HC1 1,1,1-triehloroethane cr3-chci2-^-w*ch2-chci HC1 vinyl chloride CH3-CCl3--- CH2-CC12 HC1 vinylidene chloride Minor quantities of 1,2-dichloroethane and 1,1,2-trichloroethane are also produced. Raw material ratios and reactor conditions determine the relative prop "Cions of 1,1,1-trichloroethane and byproducts produced. If 1,1,1-tric roethane is the only product desired, ethyl chloride and 1.1- dichloroe:, ;ne can be recycled to the chlorination reactor, and vinyl chloride and VDC can be catalytically hydrochlorinated to yield 1.1- dichloroethane and 1,1,1-trichloroethane, respectively:^ 23 SL 063656 FeCl. ch2cuci * aci ---i ch3-chci2 FeCl. ch2cci2 + aci ~---L CH3-CC13 Sasic operations which nay be used in the ethane process for 1.1.1-- trichloroethane are shovn in Figure 5.^ In this process, byproduct chlorinated species, including VDC, are recycled and converted to 1,1,1trichloroethane.^ Chlorine (Stream 1) and ethane (Strean 2) are fed to the chlorination reactor along with recycle streams of 1,1* dichloroethaue (Strean 13) and ethyl chloride (Stream 17). The reactor is adiabatic, with a residence tine of about 15 seconds, and is maintained at a pressure of about 600 kiloPascals (5.9 atmospheres) and an average temperature of about 40Q*C. No catalyse is used. The reactor exit stream (Stream 3) is a gas containing ethane, ethylene, vinyl chloride, ethyl chloride, VDC, 1,1-dichloroethaae, 1,2-dichloroechane, 1,1,2-trichior -ethane, 1,1,1-trichloroethane, hydro gen chloride, and minor amounts of other chlorinated hydrocarbons. This stream enters a quench column, where it is cooled, and a residue com prising mainly cetrachlcroethanes and hexachloroethane is removed (Stream 4). The overhead stream from the quench column (Stream 5) is fed to a hydrogen chloride column, in which ethane, ethylene, and HC1 are removed from chlorinated hydrocarbons. A portion of the overheads (Stream 5) containing HC1 is used to provide the HC1 requirements for VDC and vinyl chloride hydrochlorination in a later step. The remainder (Stream 7) is purified for use in ocher processes. The bottoms from the HC1 column (Stream 8), containing chlorinated hydrocarbons, are fed to a heavy ends column, where a bottoms stream (Stream 9), comprising mainly 1,2-dichloroechane and 1,1,2trichloroethane, is removed for use in other processes. Overheads from the heavy ends column (Stream 10), containing 1,1,1-triehloroethane, vinyl chloride, VDC, ethyl chloride and 1,1-dichloroethane, are fed to the 1,1,1-trichloroetnene column, which removes the product as a bottoms stream (Scream 11). SL 063657 29  t FUGITIVE E MISSIONS OVERALL ANT <>> TO HCI <$>! PURIFICATION <> <> <8> TO Chlorination reactor CHLORINATION REACTOR P FCI product OuEnch COLUMN HYDROCLORl NATION recovery COLUMN COLUMN --) REACTOR |~V HEAVY ENOS I.M-TRI- 1.1 -OlCMLORO- ChlOROETHANE , ethane y COLUMN COLUMN I COLUMN <j> T > \k> y 9suO <> Ct, ethane <5> <$> STAfllLlIER SPENT I V* CATALYST CTHOLORINATION"j,_ filter HI AC TOM <$> 1FMMi A* FfrRom T rPROOUCT DlCHLORO- RECOVERY PRODUC cthane COLUMN COLUMN STORAGE J>L NOTE* INTERMEDIATE <$> --jSTORAGE THE NUMBERS IN TWS FIGURE REFER , O PROCESS STREAMS. AS DISCUSSEO IN THE TEXT. AND LETTERS DESIGNATE PROCESS VENTS. Figure 5. Basic operations llmt May De iist-il in tlie production of 11 >l-trlchloroelluine fro* ellinm-. SL 06 Overheads from the product recovery column (Stream 12) are fed to another column, where 1,1-dichloroethane is removed as boctoms (Stream 13) and recycled to the chlorination reactor. Overheads frca this column (Stream 14), containing mainly vinyl chloride, VDC and ethyl chloride, are fed along with the HCt byproduct stream (Scream 6) to a hydrochlorinatioc reactor. This reactor operates at a temperature of about 65*F, a pressure of about 430 lo?*, and with ferric chloride catalyst. Alternatively, these byproducts may be used in other processes at che plant. The hydrochloriaation reactor converts vinyl chloride and VDC to 1,1-dichloroethane, and 1,1,1-trichloroethane, respectively. Thus, the reactor product stream (Stream 15) consists of unreacced ethyl chloride and 1,1-dichloroethane and 1,1,1-trichloroethane. This product stream is mixed with ammonia to neutralize residual HC1 and catalyst. Spent neutralized catalyst is removed in a filter and the product is then fee to a product recovery column. The bottoms from this column (Stream 16), mostly 1,1,1-trichloroethane, are recycled to the 1,1,1-tricbloroethane column. Overheads (Stream 17), composei of ethyl chloride and 1,1dichloroethane, are recycled to the chlorination reactor.^ Emissions Potential VDC emission sources from the ethane process for 1,1,1trichloroechane include (l) the vents for che 1,1,1-trichlorcethane and 1,1-dichlcroethane distillation columns (fource A) and (2) ptocess fugi tive sources, such as valves, flanges, pumps, relief valves, and drains, located between the chlorination reactor and the hydrochlorination reactor. Data from one plant indicate that the concentration of VDC in che distillation column vents is negligible.^4 Data are not available to estimate fugitive emissions of VDC from the ethane process. Source Locations Major producers of 1,1,1-tricblorocth;.ne are listed in Table 6.^120 information i* not availub'. * on which of these plants use the vinyl chloride process and which use the ethane process. SL 063659 31 TABLE 6. FACILITIES PRODUCING 1,1,1-TRICHLOROETHANE^, 20 Company Location Dow Chemical U.S.A. PPC Industries, Inc. Vulcan Chemicals Freeport, TX Lake Charles, LA Geismar, LA Ncce: This is a list of major facilities producing 1,1,1:richloroethar.e by any production process. Current information on which of these facilities produce this chemical from ethane or vinyl chloride is not available. This list is subject to change as market conditions change, facility ownership changes, or plants are closed down. The reader should verify the existence of particular facilities by consulting current listings or the plants themselves. The level of emissions from any given facility is a function of variables, such as througfcpi: and con.ro' measures, and should be determined through direct contacts with plant personnel. SL 063660 32 POLYMERIZATION OF VINYLIDENE CHLORIDE Vinylideee chloride ncncoer is polymerized with a variery of other oonooer* to produce copolyoers with special properties. VDC copolyoers can be civided into high VDC polyoers, generally containing 70 to 93 percent VDC, and low '.DC pclyr.ers, generally containing 10 to 7C percent VDC. High VDC polymers are unique is their low peroeafci1ities to oxygen, water vapor, and other gates. They also have good clarity and a glossy appearance. High VDC polyoers typically are used as vapor barrier coatings on various filo substrates, such as paper, polyester, po 1 ypropy 1 ene, or polyethylene. High VDC polyoers are also used to oake Saran films wr.ich can be used alone or laminated to other plastic files. The cooonocers cost cocooniy used to produce high VDC polyoers are vinyl chloride, acrylic eeid, acrylic esters and acrylonitrile." In low VDC polyoers, the VDC occccer generally is added to ioprcve tne Place retardant properties of the finished pclyoer. Low VDC polysers are used as flaae resistant coatings, saturants, dipping coepoundt, and adhesive!. They can also be sprayed onto fiber* and textiles. Typical coconoeers used in these applications include acrylie esters, vinyl acetate, end vinyl chloride. VDC is used with styrene and butadiene oonooers to produce a flaae retardant styrene-butadiene latex for carpet becking. VDC is also used with acrylonitrile to produce oodacrylic synthetic fibers. Eaissions froa VDC polymerization are discussed in this section. Eoissions of residual VDC oonooer from the subsequent processing and fabrication of VDC in polyoers are discussed in a later section. Copolvoer production and fabrication generally are carried out at separate facilities. Process Description The resctions involved in the production of vinylidene chloride copolyoers sre illustrated in Figure 6, The major process used to produce VDC copolyoers is eaulsion polymerixetion. This polymerization technique can be used to produce both latex resin and dried resins.^ 33 SL 063661 'Mat U:;u;ir tMilit ii Ii \I \ > v i I l Cim art ! - CiiMi / I- 9 - dvyfia >1 *' - lUitttMst cast it JTtu;t <'i B- fttt Figure 6 Basic reactions involved in the polymerization of vinylidene chloride with a comonomer. 34 SL 063662 Suspension polymerization is also used to produce dried resins, and solution polymerization is used to produce copolymers for synthetic fiber production.2^ 2^ Table 7 summarizes the types of VDC copolymers produced ar.d identifies production processes, typical cononomers used, VDC contents, and major applications for each resin type.2^'2^ emulsion Polymerization- basic operations which may be used in emulsion and suspension 1 ^A polymerization of VDC copolymers are shown in Figure 7. ** VDC monomer (Stream 1), comonomer (Stream 2), and water (Stream 3) are metered and charged to a batch reactor along with surfactants and initiator (Stream The batch is agitated, resulting in the formation of ar. emulsion of aqueous and organic phases. After charging, the batch is heated, causing the activation of the initiator. Initiator* generally usaa in emulsion polymerization are water soluble peroxides which dissociate to form free radicals when heated. The initiator radicals, on contacting the organic monomer phase, initiate the polymerization of VDC and the comonomer. The po1ymerization reaction is exothermic; therefore, after the reaction has commenced, the reactor must be coded. In emulsion polymerization, the reactor temperature typically is held at about 30*C and the reaction duration is about 7 to 8 hours. Additional monomer and initiator may be fed to the reactor during the course of the reaction. The degree of completion is 95 to 98 percent.^'^ Generally, the polymerized batch (Scream 5) is stripped of unreacted monomers (Stream 6) using steam and vacuum. This can be done either in the reactor itself or in a separate stripping vessel. The unreacted monomer is recycled if possible (Stream 7). For some types of copolymers, contamination of VDC with other comonomers precludes the recycle of VDC. Stripped polymer (Stream 8) is transferred to a holding tank, where it may be mixed with other polymer batches to ensure product um* foimity. 26 Emulsion polymerization can be used to produce either a latex (Stream 9) or a dry product (Stream 10). A latex is an emulsion of polymer particles in water, which is sold or used uadried. If a dried product is desired, water is removed by coagulation and dewatering, 35 SL 063663 TABLE 7. VINYLIDENE CHLORIDE COPOLYMERS, PRODUCTION METHODS, AND APPLICATIONS24 Polymer description Production processes Typical VDC content (percent) Comonomer a used Applications High VDC latex Earn la ion Thermoplastic resin Emulsion, Suspension Solvent soluble resin Emu1sion Low VDC latex Emulsion Hodacrylic resin Solution suspension 90 70-85 90-05 40-70 30 Acrylic esters Acrylic acid Aery lonitrile Vinyl chloride Vinyl chloride Vinyl chloride Acrylic esters Acrylic acid Acrylic esters Vinyl acetate Vinyl chloride Styrene-butadiene Acrylonitrile Barrier coatings for packaging films and paper packages Packaging film, composite (multilayered) packaging films, atonof i lament s Barrier coating for cellophane Flame retardant coalings, dips, adhesives, etc., carpet bucking Synthetic fibers SL INITIATOR. SURFACTANT MONOMER RECYCLE MONOMER RECOVERY <$> ? HOIIOHtH kllyile 1 fugitive T EMISSIONS I OVERALL I PLANT NOTE: The nuafcers in this figure refer to process lire***, as discussed In (lie text, ami the letters designate process vents. Figure 7. B.slc operetta,,, that ,, be used fur the ..r.ulueriu,, ,,f ,,,,y,,.|,,.lc. SL 063665 dry rlsin (EMULSION OR _ SUSPENSION) LATER RESIN (EMULSION) >.=` followed by drying with hot air. The dried product comprises polymer particles. The particle diameter produced by emulsion polymerization is 100 to 150 nanometers.^ Suspension Po lymer ization-- The same basic steos arc used in suspension polymerisation as in emulsion polymerization (Figure 7).^ VDC (Stream 1), comonomer (Stream 2), and water (Stream 3) are metered and charged to a batch reactor along with surfactant and initiator (Stream 4). In suspension polymerization, the initiator used is soluble in the organic phase. The batch is agitated Co produce a suspension of the organic phase in the water phase. After charging, the bateh is heated to activate the initiator, which causes polymerization. The reactor must then be cooled to remove the heat of polymerization.^'^ Suspension polymerization generally is carried out at about 60'C. The duration of the reaction is 30 to 60 hours and the degree of completion is about 85 to 90 percent.* After the polymerization reaction has reached the desired degree of completion, the polymer (Stream 5) generally is stripped of unreacted monomer (Stream 6) by steam and heat, either in the reactor or in a separate vessel. Unreacted monomer is recycled (Stream 7) or vented to a control system. The batch is then transferred to a holding tank (Stream 8) and dried (Stream 11) with hot air to produce dry polymer particles. Typical particle diameters produced by suspension polymerization range from 150 to 590 microns. Solution Polymerization-- As noted above, solution polymerization is used on the production of vinylidene chloride copolymer synthetic fibers. These fibers typically have a low VDC content. 2' 2 5* 26 Information is not available on the specific steps used in the solution polymerization of VDC copolymers; however, the general reaction steps are expected to be similar to those used in the solution polymerization of other resins. In a solution polymerization process, the polymerization reaction is carried out in a solvent which dissolves both the monomers and the 38 SL 063666 finished polymer. Monomers, initiators, and solvents are charged to a polymerisation reactor where the reaction typically is carried out under elevated temperature and pressure. The process can be carried cut either in a batch process or a continuous process. In either case, the product is a homogeneous mixture of solvent, polymer, and unreacted monomers. 'Jnreacted monomer may be stripped f-om the po lymer/so 1 vent mixture by the use of steam and heat. In a batch process, the stripping step may be carried out either in the reaction vessel or in a separate stripping vessel. In a continuous process, a separate stripping vessel is required. ` The resulting polymer solution is used in a "spinning" process to produce the product synthetic fiber. In this step, polymer fibers are extruded into a zone of hot vapor or water, where the fibers are solidifiec and solvent and residual monomer are removed. This step is discussed in the subsequent section on the fabrication of VDC copolymer products. 25 Emission? Potential VDC emissions sources at VDC polymerization plants include: VDC unloading and storage; opening and cleaning of mixing, weighing, holding, and reaction vesaels; relief valve discharges; stripper and monomer recoverysystem vents; evaporation of residual VDCfrom the finished copolymer; and process fugitive sources, including valves, flanges, pumps, compressors, relief valves, and process drains. Data are not available to estimate uncontrolled VDC emissions from these sources. However, in response to previous EPA surveys, several polymerization plants have reported total controlled emissions of vioylidene chloride and controlled emissions from a number of individual sources. 16 ' 2 5 Controlled VDC emission rates reported for VDC polymerization are given in Table 8. Emission factors given in Table 8 for reactor emission!, monomer recovery emissions, stoi age and transportation emissions, and total 39 SL 063667 TABLE 8. ESTIMATED CONTROLLED VINYLIDENE CHLORIDE EMISSION FACTORS FOR A HYPOTHETICAL VINYLIDENE CHLORIDE POLYMERIZATION PLANT* Emission source Source designation** Controlled VDC emission fsctorc Reactor Monomer recovery Pnloading/storage Process fugitive Total 3 D A G 3.5 kg/Mgd 0.33 kg/Mgd 2.1 kg/Mgd 2.8-11 kg/Mg* 1.4-7.0 kg/Mgd *Any given plant may vary in configuration and level of control from this hypothetical facility. The reader is encouraged to contact plant personnel to confirm the existence of emitting operations and control technology at a particular facility prior to estimating emissions therefrom. ^Letters refer to vents designated in Figure 7. cEmission factors in terms of kg/Mg refer to kilogram of vinylidene chloride emitted par megagram of vinylidene chloride polymerized. In cates where a particular source designation applies to multiple operations, these factors represent combined emissions for all, not each, of these operations within the hypothetical facility. The types of controls employed were not reported. dBased on industry estimates. Based on EPA projections of monomer emissions from small and medium sized generic polymerization plants with quarterly inspection and maintenances of valves, pumps, compressors, flanges, relief valves, and process drains.29 industry reports chat more stringent inspection end maintenance programs are practiced than are reflected in the factors cited in this table. (See text for discussion.) These more stringent measures result in fugitive emission control efficiencies as high as 90 to 95 percent from an uncontrolled situation where no significant measures are taken for leak detection and r 'sair.^ 40 SL 063668 plane emissiona are based on industry estimates in response to EPA surveys.1'2^ Data are not available on the specific controls used to attain the reported emission rates. Techniques which can be used to control process and storage emissions of monomers from polymerization plants are identified in Table 9. The emission rates given in Table 3 for process fugitive sources are based on EPA estimates of emissions from typical small to medium sized generic polymerization facilities using monthly inspection and maintenance of fugitive sources22 The control efficiency of monthly inspection and maintenance for fugitive emissions is about 30 percent. Additional controls which might be used include doufc1e mechanics 1 seals on pumps, enclosure of pumps, rupture disks cn relief valves, use of welded pipe instead of flanges, special construction materials for piping and valves, and intensive preventative maintenance during plant shutdown. In addition, intensive inspection and maintenance programs practiced at some plants can insure repair of most leaks within 1 day instead of 1 month. With these additional controls, many plants have achieved fugitive emission control efficiencies as high as 90 to 95 percent,1'19 Source Locations Table 10 lists producers of VDC copolymers. The table also gives plant location and identifies the types of VDC copolymer produced at each facility.2^'2^*1 1 SL 063669 TABLE 9. POTENTIAL EHISSION CONTROLS FUN I'OI.YMLH PLANTS30 VBC ulailiai *! atore|a ral)H[lt>lUl SlIUI C (ItBiinttivu1 Cowlt o 1 Icebn?lofty Laadiat linea, VC alara|t lank, k Furg4 to noiioMr recovery airing, wi|hia| and balding ijfia, luciniralloa, itlv l laah aata l>iir|>i luu or ciiboa dao(|>i ioa laaa i olvaiil cleaning, at era plains, baler pieton, rraclor purla air blower, rices purge, ale., uaad Stripping folyMi ha 1 dia| taaba, aUU| Dewatering, ii)ia| valva diachargao Stripping vaaaal want Itcattii ayataai eabauat veuln a.id haoch-out pat c It Tank veala Cantrifa|a venta, dryer vaat atache t E r Vnted in ilMigkait or Monomer recovery cycle* B1ivitU|, conUiuoeiil, ioet i u*eul et ion, loft roved oyeretor litiiU|i elc. Vented 10 aataaaier recovery eyetea lul lowed bp 1acinarat ton, aolveut abeurptloa or carbon adsorption tualanraa to cul led al 1 recovery venta and/or refrlgerat ion In rondanaa VltC lul lowed by incineraliou, aolveut abaorplioa or cailion adsorption. Ellippiag Stripping rl* * eoarcaa <!<, pwf1, flanraa, ale. in N|ri ?. Not* lhl U wtmu da nut covti fu|iiiva eaiaaiona lioa 06367 f TABLE 10. FACILITIES PRODUCING POLYVINYL IDKNK CIIMIHIDK1 8. 20. 25 2<1. 31 Kanuf acturer Location Produc t an American Cyanamid Co. Borden Chemical Div. Dow Chemical Co. E.l. DuPont de Nemoura and Co. W.R. Crace and Co. Dewey ant) Atmy Chemical Div. Honaanto Co. Horton Chemical Co. National Starch and Cheat. Co. Olin Corp. Reichhold Polymera, Inc. Rohm and Itaaa, Inc. A.E. Staley Hanufncturing Co. Pace, FL Campton, CA llliopolitt, IL Dalton, CA Midland, MI Circlevillc, Oil ayntliet ic ( ibera low VDC latex high VDC latex, high VDC latex, i Owenaltoro, KY Decatur, AL Ringwood, IL Heredoaia, IL Pisgah Foreat, MU ChetwoId, DE Briatol, PA Knoxville, TN Lemont, IL high VDC latex ayiiLhelic libera high VDC latex low VDC latex high VDC latex low VDC lute) low VDC latex low VDC latex high VDC latex `Tolywinylidwna chlorilt topalywar fornutations sad khi ara )<aurju4 iw Table U. total Tlii a listing ia subject la change as Mihsl roaditioaa chance, facility ownership t hanger. pi ant a ara cloaad. ale. Tba reader should warily the aaisteaca of paiticolar lacililiea by con aw 11 inn current listings aad/or the piaota tlieaaalvaa. Tlia Icwcl ui VDC eaiaaiuua Iriaa any |in facility ia a function of war tab lea auch aa capacity, throughput and conical aeetutea, and aliuuld ba dataiaiuad through direct contacts with plant peraonncl. SL 063671 OSE 3F VINYL ID ENE CHLORIDE IN SPECIALTY CHEMICAL PRODUCTION Vinylidene chloride i* used as a chemical intermediate in the production of ehloroacetyl chloride. This use of VDC is minor in comparison with its use in production of VDC copolymers. The structure cf chlcrsaeetyl chloride is as follows; Cl 0 a- cI - cIf ii H Cl Its main use is in the manufacture of chloroacetophencne, the principal ingredient is tear gas. It is also used in the manufacture of pharmaceuticals.^*^ Process Description and Emissions Information on the process used to produce ehloroacetyl chloride from VDC is net available, nor are data available to estii-ite VDC emissions from the process. Dow Chemical reports n- losses of VDC to the environment from the process used in ehloroacetyl chloride production. Source Locations The VDC process for ehloroacetyl chloride is used by the Dow Chemical U.S.A., of Midland, Michigan.^ Production facilities are located at Dow Chemical's Michigan Division plant at Midland. 18 44 VINYL IDEA'S CHLORIDE COPOLYMER FABRICATION When VDC is polymerized to produce VDC copolymers, some residual VDC monomer ranaics in the polymer. During subsequent fabrication, when the polymer is heated, dissolved, or otherwise treated, some of this residual monomer evaporates, resulting in VDC emissions. The ieve. cf VDC residue in the polymer, and hence the magnitude of VDC emissions, is dependent cn the type of polymer, the comonomer used, and whether stripping technology is used at the polymerization plant. As noted in the previous section entitled POLYMERIZATION OF VINYLIDENE CHLORIDE, VDC copolymers can be divided into two groups: htgh-VDC copolymers (79 to 9C percent) used to form moisture and vapor barrier coatings and films; and Icv-VDC copolymers (1C to 70 percent), where VDC is added mainly to improve the flame retardant properties of the finished polymer. VDC copolymers are also produced in a number cf forms: dried suspension resin, dried emulsion resin, latex, and polymer solution. The types of VDC copolymers produced, production methods used, and applications cf the different types were summarized in Table 7 in the section entitled POLYMERIZATION CF VINYLIDOJE CHLORIDE. Generally, copolymer fabrication and copolymer production arc carried out at separate facilities. Process Descriptions Process descriptions are presented below for three methods of fabricating high-VDC resins: coating of cellophane with copolymer solution, coating of paper or placcic film with latex, and extrusion of dried emulsion or suspension resin. Information was not available on the specific methods used to fabricate Isw-VDC resins. Cellophane Coating with Copolymer Solutions-- Basic operations that may be uied in the coating of cellophane with VDC copolynt- are shown in Figure 8.^2 The inputs to this process are a dried, high-VDC copolymer and pretreated cellophane film. The VDC polymer Is first dissolved in a solvent mixture in a closed tank. The solvent mixture includes methyl ethyl ketone and tetrahydrofuran as the 5 SL 063673 VENT vcur COPOLYMER Figure 8. Basic oper.it Ions that M.iy |>c u-ie.1 In the coal In,, of cellophane with high-VIX: co|Ht| I- SL 063674 primary solvents and toluene as a diluent. Following the disso1uticn step, additives such as wax, talc and silica are added in a closed blender. The polyoer solution is then fed to a dip tank, through which a cellophane fils strip is drawn. After passing through the dip tank, the cellophane, now coated with 7TC ccpclyrer solution, is run through a dryer consisting of two chanters. In the first, dry air at 90 to 1^0'C is passed over tne fils, resulting in removal of the solvent.^* In the second chanter, the fils is conditioned in worn hue is air. Solvent laden vapor is collected fros the drying chaster and ducted to carbon absorption beds. Solvent stripped free cue beds is purified by distillation and recycled to the process. Heating air fros the second drying chaster is vented wo the atcosphere. Ccating with Latex Ccpclyr.er As noted if. the earlier sect wen of FCLYMSF.IIATTCN OF V Iim CHLCK1LE, a latex is a polyoer esulsio.. in water. Materials typically coated vwth high-VDC latex include paper products and plastic files. The latex say first be blended with additives, such as wax or pigoents, or diluted with additional water in a vented nixing tack. The latex is pusped from the sixing task to a holding tank and then to the dip tank. The holding tank allows reduction of any foas that say fore during sixing. The saterial to be coated is rolled through the dip tank and then to a drying oven. In the drying oven, water is rtsoved f/oo the latex and the latex fores a barrier fils. Beth the dip tank and the drying over, are vented to the atcosphere. 1 1 Extrusion of Therscplastic Copolyeer-- The raw saterial for extrusion it dried esultioo or suspension resin in the fore of a powder or taall granules. The pclyeer it cixed with additives such as plasticizer in a high-intensity blender. Mechanical energy dissipated in the blender heats the resin to about 170*C.^ The blender is vented through a hood, usually to a roof stack. From the bigh-ictentity blender, the 'esin is fed to e ribbon bleeder, where it is hocoger.ized further sec cooled. The blended resin nay be SL 063675 stored pr.or Co extrusion, or may be extruded immediately after blending.^ The extrusion process used on high-VDC copolymers is a blowr-fila process. The resin compound is fed to an extruder, and is extruded through a die in che fora of a cube, becoming colter, in the extrusion process. Toe end of the tube is then pinched off, and air is blcvc. into the tube, expanding it into a bubble. The bubble is then coded by another blast of air and flattened before undergoing further processing to form a f*lm.^- Emissior.s Emission sources from VDC copolymer fa,:tication induce: polymer storage vents, polymer mixing and blending vents, and finished polymer drying. Emissions from individual sources have not been quantified, however, total VDC emissions from copolymer fabrication can be estimated by mass balance from the concentrations of residual VDC conomar in che polymer entering and leaving the process. " ci - Co where IT the overall uncontrolled emission ftetor for the fabrication process, g VDC/Kg copolymer processed, C^ the concentration of residual VDC monomer in the copolymer entering the dissolver (Figure 8) ppmw, end C0 che coecentretion of residual VDC monomer in the copolymer leaving the dryer (Figure 8),ppmw. Table 11 summarize* data on residual VDC levels in raw and fabricated high-VDC copolymers, and presents estimates of uncontrolled VDC emission factors for high-VDC copolymer fabrication procesees. Data are not available on the residuzl VDC levels in low-VDC copolymers. It should be noted tbet the mass balance technique of estimating emissions involves the assumption that the only removal mechanism for VDC free the copolymer is by emissions to che atmosphere. Thus, emissions estimates developed by the mass balance technique would be worst case estimates. In high temperature fabrication processes and drying processes, some of the residual VDC may be polymerized. -a SL 063676 TABLE 11. ESTIMATES OF UNCONTROLLED EMISSION FACTORS FROM KIC-H-VDC COPOLYMER FABRICATION PROCESSES Process Celophaze coating Latex coating Extrusion Copolymer VDC concentration s (otnv)a Raw Processed resin^ resir.c (c,) (eg) Estimated uncontrolled emission factor (g/'Mgl^ 10-120 50-2000 2-25 neg* 0-500 neg* 10-120 50-1500 2-25 ^Reference 32. ^P.asin entering the dissolver (Figure 8). cResir. leaving the dryer (Figure 8). -Calculated from residual VIC levels. Emissions are expressed ir. terms of grans of VCC per Mg of copolymer processed. er.eg negligible. 19 SL 063677 VDC emissions from copolymer fabrication generally are uncontrolled. In the first stage of tbe drying operation for cellophane coating, emissions containing solvent are captured and ducted to a carbon adsorption system. The solvent is then desorbed and recycled to the polymer dissolving operation. Some of the VDC vaporizing in the drying process would be captured on the adsorbers. However, unless some process is used to separate the captured VDC from the solvent, the VDC would be recycled to the process and eventually emitted from the second stage of the drying process, from the adsorber vent, or from another vent. Source Locations Table 12 gives a list of plants fabricating high-VDC copolymers, along with plant locations.31,32 jit iBCiudt# plants that produce PVOC-coated cellophane, plants that apply ?VDC barrier-coatad to paper and plastics, and plants that extrude VOC copolymer. Information is not available on the locations of fabrications of low-VDC polymers. Such plants would be classified under Standard Industrial Classification (SIC) code 282. 50 SL 063678 TABLE 12. FACILITIES FABRICATING HIGH VINTLIDENE CHLORITE COPOLYMERS21.32 Company Plant Location Processes Cellophane Barrier coating coating Resin extrusion Allied Chemical Ccrp. Pittsvilie, ?A American Bag and Faper Philadelphia, PA American Can Co. .''eenah, WI Amtech, Inc. Odenton, MD Consolidated Paper Wisconsin Falls, VI Crown Zellerbach Portland, OR Cryovac Div. of W. R. Groce Icwa Park, TX Simpscnville, SC Curwood Div. of 3emis New London, WI Daniels Rhinelander, WI Diversa-?ak St. Petersburg, FL Dow Chemical Midland, MI .1. dupont do Kenour* Richmond, VA Circleville, OH Clinton, LA Tecuaseh, KS FMC Corp. Fredericksburg, VA Marcus Hook, PA Gordon Cartons Baltimore, MD Green Bay Packaging Green Bay, WI Hercules Covington, VA Interstate Folding Box Middletown, OH X X X X X X X X X - X X X X X X X X X X X X X Continued 51 SL 636?9 TABLE 12- (Conc'd.) FACILITIES FABRICATING HIGH VINYLIDENE CHLORITE COPOLYMERS31.32 Company Plant Location Processes Cellophane Barrier coatitg coating Resin extrusion Michigan Carton Co. Mil print Minnesota Mining and Manufacturing Olin Corp. Qlinkxaf t Oscar Mayer Hexham Rhinelander Oiv. of Sc. Regis Sealed Air Corp. Standard Packaging Ihilmany Union Carbide Corp. Z usbri1 Battle Creek, MI Milwaukee, VI Decatur, AL Pisgah Forest, NC Covington, IN West Monroe, LA Chicago, IL Davinport, IA Los Angeles, CA Madison, VI Nashville, IN Philadelphia, PA Memphis, IN Rhinelander, VI Fairlavn, NJ Clifton, NJ Kaukauna, VI Centerville, LA Cincinnati, OH X X X X X X X X X X X X X X X X X X X X Note: This listing is subject to change as market conditions change, facility ownership changes, plants are closed, ete. The reader should verify the existence of particular facilities by consulting current listings and/or the plants themselves. The level of VDC anislions from any given facility is a function of variables such as capacity, throughput and control measures, and should be determined through direct contacts with plant personnel. 52 SL 063680 V0LATIU2ATIQN FROM VAST! TREATMENT, STORAGE, AND DISPOSAL Considerable potential exists for emissions of volatile substances, including VDC, from waste treatment, storage, and disposal facilities. VDC is expected to be present in the following wastes: stiil bottoms and wastewater from VDC production, perchloroethylene and trichloroethylene production, and trichloroethane production; wastewater and off-specificatioa polymer from VDC polymerization; and still bottoms and wastewater from specialty chemical production processes where VDC is used as a feedstock. (See separate sections on emissions from these processes.) In addition, VDC cay be present in wastes from other processes. Pot entia 1 Sources VDC nay be emitted when waste containing VDC is present in surface impoundments for treatment and storage of wastewater, open treatment and storage tanks, and land-treatment areas for solid wastes and sludges. The above treatment and storage facilities may be located at the site of generation of the waste, or at a separate commercial waste treatment plant. In addition, publicly owned treatment works (POTVs) may emit VDC if they receive wastewater from plants producing VDC either as a main product or as a byproduct, or from plants using VDC as an intermediate. Volatile compounds also cay be emitted from solid wastes during and even after disposal in a covered landfill. Reference 33 summarizes general theoretical models for estimating volatile substance emissions from generic waste treatment, '*ge, and disposal operations, including surface impoundments, landfills, land treatment (landfarming), wastewater treatment, and drum handling and storage operations. If facilities of the above types are known to bandle wastes containing VDC, the potential for air emissions should be considered. -Em. isstons 34 A pilot-scale study was conducted by EPA to evaluate the partioning of severe, volatile organic pollutants, including VDC, in conventional wastewater treatment processes. The tested wastewater treatment system 53 SI 063681 consisted of a sequence of primar- clanfifr, aeration basin, and secondary aeration basin. Vast er influent contained an average of 10.7 parts per billion (ppb) VDC. Over 98 percent of the VDC entering the pilot treatment system was found to evaporate, with about 65 percent evaporating froc the primary clarifier and 33 percent iron the aeration basin. The 96 percent evaporation rate correaponds to an emission factor of 0.98 grans VC per gran VDC in the wastewater feed. It should be noted that these tests were conducted at low VDC concentrations (about 10 ppb); the emission ractor aay change at higher concentrations. 54 SL 063682 SECTION 5 SOURCE TEST PROCEDURES Vinylideae chloride emissions can be measured using E?A Reference Method 23, which was proposed in the Federa 1 Register on June 11, 1980.^ The method has not been validated by EPA for vinylider.e chloride, Dut a similar analytical procedure has been used to measure occupational exposures to VDC.^^ In Method 23, a sample of the exhaust gas to be analyzed is drawn into a Tedlar or aluminized Mylar bag as shown in Figure 9. Tedlar is considered a more reliable bag material than Mylar for VEC.^ The bag is placed inside a rigio leak proof container and evacuated. The bag is then connected by a Teflon sampling lice to a sampling probe (stainless steel, Pyrex glass, or Teflon) at the center of the stack. The sample is drawn into the bag by pumping air out of the rigid container. The sample is then analyzed by gas chromatography (CO coupled with flame ionization detection (FIE). Analysis should be conducted within one day of sample collection. The recommended GC column is 3.05 a by 3.2 mm stainless steel, filled with 20 percent SP--2100/0.1 percent Carbowax 1500 on 100/120 Supelcoport. This column normally provides an adequate resolution of halogenated organics. (Where resolution interferences are coccuntered, the CC operator should select the column best suited to the analysis.) The column temperature ahould be set at 100*C, Zero helium or nitrogen should be used as the carrier gas at a flow rate of approximately 20 ml/ain. The peak area corresponding to the retention time of vinylidsne chlor.de is measured and compared to peak areas for a set of standard gas mixtures to determine the VOC concentration. The range of the method is 0.1 to 200 ppm; however the upper limit can be extended by extending the calibration range or diluting the sample. To avoid absorption of VDC by the Tedlar bag, the sample should be analyzed as 55 SL 63683 FLOW METER CHARCOAL TUBE figure 9. Method 23 sampling train.^ 56 SL 063684 000 as posible after collection, preferably on the caae day. The nethod does not apply when vinylidene chloride is contained in particulate matter. 57 SL 063685 1 REFERENCES 1. Hushon, J., gad M. Koinreich. Air Pollution Assessment of Vinylidene Chloride. EPA-450/3-78-015. U.S. Environmental Protection Agency, Washington, DC. February 1978. pp. 7-20. 2. Crayton, M., ed. Kirk-Othmer Encyclopedia of Chemical Technology. Third Edition, Volume 23. John Wiley and Sona, New York, NY, 1983. pp. 764-798. 3. Edney, E., S. Mitchell, and J. Bufalini. Atmospheric Chemistry of Several Toxic Chemicals. EPA-60C/3-82-092. U.S. Environmental Protection Agency, Research Triangle Park, NC. November 1982. pp. 31-35. 4. Cuppitc, Larry. Fate of Toxic and Hazardous Materials in the Air Environment. EPA-600/3-80-084. U.S. Environmental Protection Agency, Research Triangle Park, NC. August 1980. pp. 3-6. 5. Standifer, R.L., and J.A. Key. Report 4: 1,1,1-Trichloroethane, Perehlorotthylene, Trichloroethylene, and Vinylidene Chloride. In: Organic Chemical Manufacturing-Volume 8; Selected Processes. EPA-450/3-80-028c. U.S. Environmental Protection Agency, Research Triangle Park, NC. December 1980. pp. XII-1 to 111-17. 6. Neufeld, M.L., M. Sittenfield, M.J. Plotkin, I.F. Volk, and R.E. Boyd. Market Input/Output Studies--Task I: Vinylidene Chloride. EPA-560/6-77-033. U.S. Environmental Protection Agency, Washington, DC. October 1977. pp. 67-70, 156-163. 7. Hawley, Cessner G. The Condensed Chemical Dictionary. Tenth Edition. Van Nostrand Rcinhold Co., New York, NY. 1981. pg. 232. 8. Modern Plastics Encyclopedia. Modern Plastics. October 1982. 9. Reference 6. pg. 156. 10. Reference 5. pp. IX-1 tc II-6. 11. Reference 5. pp. III-15 to III-17. 12. Reference 6. pp. 67-70. 59<10A):113-114. 13. Reference 5. pp. IV-15, IV-1S, and 14. Reference 5. pp. V-5 to V-7. SL 063686 58 15- Mascone, D., O.S. Environmental Protection Agency. Memo and Addendum to J.R. Farmer, EPA, entitled 'Thermal Incinerator Performance for NSPS", June 11, 1980. 16. Eisen, ?., M. Sanders, and M. Samuels. Human Exposure to Atmospheric Concentrations of Vinylidene Chloride. Prepared by Uapora, Inc. (Project 507-13) for the U.S. Environmental Protection Agency, Research Triangle Park., NC. February 4, 1982. pp. 2-1 and D-l to 0-3. 17. Reference 6. pp. 84-86. IS. Letter from Alice R. Moyer, Chemical Manufacturers' Association, to Thomas F. Lahre, O.S. Environmental Protection Agency. March 19, 1985. 19. Letter from Thomas E. Lingafe1 ter, Dow Chemical, to Thomas F. Lahre, U.S. Environmental Protection Agency. February 13, 1985 20. 1983 Directory of Chemical Producers, United States of America. SRI International. Menlo Park, CA. 1933. 21. Reference 5. pp. 111-8 to 111-14. 22. Reference 5. pp. iv-i: to iv-is. 23. Reference 5. pp. I1I-4 to I1Z-8. 24. Reference 5. pp. IV-6 to IV-10. 25. Vinylidene Cnloride Monomer Emissions from the Monomer, Polymer, and Polymer Processing Industries. Prepared by Arthur D. Little, Inc. (ADL Reference 76086-31) for the O.S. Environmental Protection Agency, Research Triangle Park, NC. April 1976. pp. 16-42. 26. Reference 6. pp. 9-20. 27. Wilkins, Clynd* E. Industrial Process Profiles for Environmental Use: Chapter 10. Plastics and Resina Industry. EP/.-600/2-77-023 j. U.S. Environmental Protection Agency, Research Triangle Park, NC. February 1977. pp. 55-58. 28. Synthetic Fiber Production Facilities--Background Information for Proposed Standards. EPA-450/3-82-0lia. U.S. Environmental Protection Agency, Research Triangle Park, NC. -October 1982. pp. 3-1 to 3-77. 29. Control of Volatile Organic Compound Leaks from Synthetic Organic Chemical and Polymer Manufacturing Equipment. EPA 450/3-83-006. U.S. Environmental Protection Agency, Research Triangle Park, NC. March 1984. pg. 4-3. 59 SL 063687 30. Vinyl Chloride - A Review of National Emission Standards. EPA-450/3-82-003, U.S. Environment*1 Protection Agency, Research Triangle Park., NC. February 1982. pg. 4-2. 31. Reference 16. pp. 1-3 to 1-4. 32. Reference 25. pp. 43-61. 33. Evaluation and Selection of Models for Estisating Air Emissions from Hazardous Waste Treataent, Storage and Disposal facilities. EPa-450/3-84-020. Prepared for the D.S. Environmental Protection Agency by GCA, Corp., Bedford, MA. Dec. 1934. 34. Petrasek, A.C., Jr., Barry M. Austern, and Timothy V. Nsiheisel. Renoval and Partitioning of Volatile Organic Priority Pollutants in Wastewater Treataent. Paper presented at the Ninth U.S.-Japan Conference on Sewage Treataent Technology, Tokyo, Japan, September 13-29, 1983. 20 pp. 35. Method 23. Deteraination of Halogenated Organics free Stationary Sources. Federal Register. 45<114):39776-39784. June 11, 1980. 25. Forest, Denis. A Seep ling and Analytical Method for Vinylider.e Chloride in Air. American Industrial Hygiene Association Journal. October 1979. pp. 888-893. 37. Severs, L.W. and L.K. Skory. Monitoring Personnel Exposure to Vinyl Chloride, Vinylidene Chloride and Methyl Chloride in an Industrial Work Environment. American Industrial Hygiene Association Jouraal. September 1975. pp. 669-676. 38. Telecoo. Joseph Knoll, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina, with William Battye, CCA/Technology Division. EPA Reference Method 23. January 26, 1984. 39. Memorandiaj from R. Rosensteel, U.S. Enviromaentti Protection Agency to T. Uhrc, U.S. Environmental Protection Agency, entitled "Review of Draft Reports on Emission Factors for Potentially Toxic Substances," November 10, 1984. 60 SL 063688 APPENDIX A PROCESS FVCITIYI EX 13 SIC.'.' CALCVUTICNS FGR VINYLIDEKE CHLORIDE PRODUCTION rac.y i,i,:-Taichl:sc~llne Fugitive emissions of v iay 1 ider.e/ch leride (VDC) and other volatile organics r * su 1: from leais in process valves, pump scale, compressors, sample connections, oper.-er.ded 1 ices and pressure relief valves. Fugic:ve VDC emission rates frc? these sources were based os a process tlcw d;agra= (Figure I), process operation data, a fugitive source inventory for a hypothetical plant,* and EPA emission factors for process fugitive source**.* The first step m estimating fugitive emissions ef VDC was to list, the process streams in the hypothetical plant. Iheir phase* (i.*., gaseous or liquid) were then identified free the process flew diagram snd their compositions estimated. For the reactor product stream, the composition was estimated based on reaction completion data. For a stresr froc a distillation column or other separator, the composition was estimated based on the composition of the input stream to the unit, the unit description, and the general description of the stream of interest (i.e. overheads, bottoms, or aidedrtw). After the process streams were characterised, the number of valve* per stream were estimated by dividing the total number of valves at the plant equally among the process streams. Similarly, puspa were apportioned equally among liquid process stream*, and relief valves were epportioned equally among all reactors, columns, and ocher separators. The locations of any compressors wers determined from the process flow diagram. Although the above sources probably are not apportioned equally among the process lines st an actual plant, the equal apportionment algorithm provides the best estimate of the number of sources per line given the available data. Emissions were then calculated for pumpe, compressors, valve* in liquid snd f,a line service, and relief vslvee. Emissions from flange* A-l sL 063689 tod drain are minor in comparison with these sources and were therefore neglected. Fugitive missions from a particular component were assunec to have the same coeposition as the process fluid flowing through each component. For valves in liquid service, for instance, VDC missions were determined by taking the product of: (1) the total caber cf liquid valves in VDC service; (21 the average VDC content of tne streams passing through these valves; and (3) the average fugitive emission rate per valve per unit time at measured by EPA. Emissions from valves in gas service, pusps and compressors were calculated ir. the same manner. For relief valves, fugitive emissions were assumed to have the ccmpoiiticn cf the overhead stream from the reactor or column served by the relief valve. Emissions from the various fugitive source types were summed to obtain total process fugitive emissions of VDC. Because emissions from process fugitive sources do not depend on tr-eir sire, but only on their number, total process fugitive emusicns are not dependent on plant capacity. Thus, the overall emissions are expressed in terms ct kilograms per hour of operation* HYPOTHETICAL FUST FUCITIVE SOURCE INVENTORY 725 process* valves e 13 pusps (not including spares! 2 compressors 23 safety relief valves PROCESS LISE COMPOSITION AND R'CITIVE SOURCE INVENTORY Of the approximately 16 major procass lints in the production process, 9 contain at least soma fraction of volatile organics compounds (VOC) and 6 contain vinylidene chloride (VDC). Compositions of the major process streams (identified in Figure 2 in the section or VINYLI DENE CHLORIDE PRODUCTION) are estimated in Table A-l. A fugitive emission equipment count was cot avsilable for a VDC production plant. However, studies of other synthetic organic chemical manufacturing plants indicate that a typical fugitive equipment count is as follows:1 063690 SL* TABLE A-l. ESTIMATED PROCESS LIME COMPOSITION IN VDC PRODUCTION* Stream nittber4 Phece --Stream. composition (weight fraction)------- We ter NaOH NaCl C12HC-CH2C1 VDC Other 1 Liquid 0 1.000 0 0 00 Liquid i.co 00 0 0C 3 Liquid 0 0 4 Liquid 0.75 0.061 0 1.000 0 0.191 00 c0 5 Liquid 0 00 0 0 1.00 6 Liquid 0.77 0.011 0.073 0.024 0.121 0 7 Liquid O.SO o o 0.086 0 0 0 S Liquid 0 0 0 0.164 0.836 9 Liquid 0.90 o.oi: 0.086 0 0 0 0 10 Liquid 0.90 o.oi: C.086 0 0 1 A Liquid 0 0 0 0.164 0.836 0 0 12 Liquid 0 0 0 0.164 0.836 0 13 Liquid 0.90 0.012 0.086 0 00 14 Vapor 0 0 0 0.164 0.836 0 15 Liquid 0 0 0 1.000 00 16 Liquid 0 00 0 1.000 0 *Stre*m nusberi correepond to thoae howa in Figure 2. A-3 SL 063691 15 process velvet per major process line, 1 pump (not including spares) per major liquid process line, 1 compressor for each gas line requiring pressurization, and 2 relief valves per pressure vessel or column. EMISSION CALCULATIONS VCC emissions from valves in liquid ar.d gas service, and for pumps were calculated as follows: (Total VDC emission race for ucream type) (d of streams) X (Average vrC content for stream type) X {4 of pumps or valves per stream) X (Emission rate for individual rate for individual pumps or valves) These calculations are sv=marized ir. Table A-2. Similarly, emissions from relief valves were calculated fer each vessel or column processing VDC: (VDC emissions rate) (2 relief valves per vessel) X (VDC fraction is vessel overheads) X (0.104 kg esissions/hr/rlief valve) These calculations are summarized in Table A-3. No compressors art expected to be in VDC service in the VDC-frnm-l,l,2-trichloroethane process. Total uncontrolled process fugitive emission rates for VDC production are given in Table A-4, along with controlled emission races for various combinations of emission reduction techniques. The emission reduction techniques studied were quarterly and monthly inspection of valves and pumps, the use of double mechanical staled pumps, and the use of rupture disk1- in tandem with or in place of relief valves. A-4 SL 063692 TABLE A-2. ESTIMATED VDC EMISSIONS FROM VALVES AND PUMPS Source type Number of streams Average VDC weight fraction* Number of sources Source VOC emission rate* (kg/hr-source) Emissions (kg/hr) Valves vapor liquid Pumps TOTAL as 3 3 0.54 0.60 0.60 30 45 3 o o *0 0.0056 0.0071 0.140 0.191 0.089 0.42 fraction ol VOC streae in each source type comprised of VDC. A-5 SL 063693 TA3LE A-3. ESTIMATED VDC EMISSIONS FROM RELIEF VALVES Vessel Reactor Phase sep-ratcr Drying eoluzn Finisnizg column Stripping column Number of relief elves 2 4 2 VDC weight fraction in overheads 0.56 0.56 0.50 0.50 0.50 TOTAL Emissions* (kg/hr) 0.12 0.12 0.10 0.10 0.10 0.54 Relief valve emission rate of 0.1C4 kg 'hr-valve was used to calculate mi* SICOS.- A-j SL 063694 TABLE A-4. EUC1TIVE EMISSION CONTROLS AND ESTIMATED CON THOM,ED EMISSION RATES Source dtii Uncontrolled VPC emission rate (kg/lir) Quarterly 1/Ha Percent control Emits ion* (kg/lir) Control led emission rates Konllily I/H HontLly 1/M double scat a; rupture disks Percent control Emitsion# (kg/tir) Percent control Emissions (kg/hr) Vapor valvea Liquid valvea Pumps Relief valvea 0.140 0.191 0.089 C.fj 64.0 44.0 32.5 0 0.051 0.107 0.060 0.545 73.0 59.0 60.8 0 0.038 0.078 0.035 0.545 73.0 59.0 100.0 100.0 0.038 0.078 0 0 OVERALL 0.965 21.0 0.762 29.0 0.696 88.0 0.116 *I/H refera to inapection and arinttniu.'f of valvea and pump*. Control pcrccntogcu aie from Reference 2 SL 063695 REFERENCES FOR APPENDIX 1. Standifer, R.U, and JJL Key. Report 4: 1,1,1-Trichloroethane, Perchloroethy lene, Trichloroethylene, and Vinyl idene Chlo.ide. In: Organic Chemical Kanufactux ing--Vo ' vase 8*. Selected Processes, EPA-450/3-30-02Sc, U.S. Environmental Protection Agency, Research Triangle Park, NC. December 1980. pp. IV-1 to 17-22. 2. Fugitive Emission Source* of Organic Compounds--Additioaa1 Information oc Emissions, Emission Reductions, and Costs. EPA-430/3-82-C10, l\S. Environmental Protection Agency, Research Triangle Park, NC. April 1982. A-8 SL 063696