Document GmERv0Z0x9y40e055qY1EbE6v

Process and Emission Data - Avon Lake Plant A. Brief Process Description and Flowsheet Vinyl chloride (VCK) is unloaded from railroad tank cars or tank trucks into pressurized storage spheres. VCM, emulsifiers, and catalysts are metered into polymerization vessels wherein PVC is produced through a chemical- reaction in an aqueous medium under controlled conditions of temperature and pressure. After the reaction reaches a predetermined completion, the contents are transferred to a secondary vessel wherein steam in injected and the VCM containing vapors are pumped to a recovery system. The VCM containing vapors are compressed, cooled, condensed, decanted, and recycled to the process for reuse. The stripped PVC resin water slurry is then pumped to blending tanks where the batches from multiple reaction vessels are blended for product uniformity. From the plant tanks the PVC resin water slurry is pumped to a dewatering centrifuge where approximately 90/1 of the water is removed and subsequently discharged to the industrial sewer system. The PVC resin wet cake is conveyed from the centrifuge to.a flash dryer where essentially all the remaining water is removed.- At this point, the dry. res in-is' buoyant in- ah air stream.and enters a two-stage collection system for separation of conveying, air. The. PVC'resin is then screened and air conveyed to storage for bulk shipment, for compounding, or bagging. It should be emphasized that there are numerous transfer and storage points as PVC manufacture via the suspension process consists of a batch process. Major equipment items installed at this plant for suspension resin comprise 70 polymerizer vessels of varying sizes, 24 stripper vessels, 19 blending tanks, 10 entrainment separators, four VCM receiver tanks, 10 compressors, six VCM condensers, one vent compressor and condensor, eight centrifuges, eight flash dryers, eight primary and secondary collection systems, 16 product storage bins, and 16 product storage silos. A simplified process flowsheet showing all major equipment processing steps is attached. The suspension resin production rate in 1973 for this plant was approximately 25,500 pounds/hour. B. F.mlssions Continuous and intermittent sources of emission to the atmo sphere are shown on the accompanying flowsheet as designated by letter. Emission data are correspondingly tabulated below. t* .i..i. ft o o 1,. Emission Source (A) - Safety Relief Devices This emission source is intermittent in nature and consists of rupture discs and pressure relief devices for numerous and varied equipment items. Also included are discharges which occur as a result of manual venting to reduce the pressure thereby preventing greater losses. Operational history has shown that atmospheric releases occur essentially only from the polymerizer vessels. In such case the emission would consist of VCM, water, PVC, or some combination thereof depending upon the stage of the VCM conversion when venting occurs. There are 70 separate relief valve equipment dis charge pipes and 70 separate manual pressure relief devices associated with the polymerizer vessels. These discharge pipes are four inches in diameter and vary in height from 50 to 65 feet above grade. 2. Emission Source (B) - Poly Evacuation This emission source is generally of a continuous nature inasmuch as the poly evacuation blower services multiple polymerizer vessels on a slightly greater frequency than two per day per. vessel. Use of this system occurs during polymerizer cleaning opera tions after the stripped PVC resin water slurry has been transferred to the blending stage. The purpose of this system is to insure low VCM exposure levels to workers upon entry to the vessels. The following tabulation better characterizes this source. 3 blowers,each exhausting 12 polymerizers at 1,500 scfm through a 0.83 ft. diameter vent at a height of 50 feet above grade. 1 blower exhausting 8 polymerizers at 2,500 scfm through a 1 ft. diameter vent at a height of 50 feet above grade. 1 blower exhausting 10 polymerizers at 3,500 scfm through a 1.2 ft. diameter vent at a height of 50 feet above grade.2 2 blowers, each exhausting 8 polymerizers at 2,500 scfm through a 0.83 ft. diameter vent at a height of 65 feet above grade. 3- Emission Source (C) - Blending The atmospheric emission from the blending operation results from the continuous purging of the vapor space in the atmo spheric pressure blending tanks with fresli air. This practice is both a safety and health-related one aimed at maintaining VCM concentration well below the lower explosive limit and at VI oCl o Co ^ 2L- acceptable levels for worker exposure. The following tabulation characterizes this source. Number 1 1 4 3 4 3 "1 1 1 Emission Rate 1,200 1,200 3,500 4,000 6,000 4,300 1,660 1,660 1,660 Vent Size 0.67 ft-\ 0.67 ft.2 1.0 ft.2 1.0 ft-2 1.0 ft.2 1.5 ft.2 1.0 ft.2 1.0 ft.2 1.0 ft.2 Vent Height 16 16 55 55 55 45 40 40 40 4. Emission Source (D) - Vent Condenser This source consists of the discharge from the VCM recovery system and contains inert gas, water vapor, arid VCM. Prior tp. discharge,. the .gases are' treated in a two.-stage cbmpression condensation:system. The emission-is intermittent in nature as required to discharge the inert gases contained in the system. There are two separate recovery systems installed. Vent pipe size for each system is .33 square feet in area discharging at an approximate height of 63 feet above grade.5 6 5. Emission Source (E) - Steam Jets After vessel cleaning and prior to putting such vessel into service, a vacuum is pulled on the vessel to remove any oxygen contained in the system. This discharge contains no VCM. A discharge is periodic and exhausts through a two inch in diameter vent at a height above grade of 74 feet. 6. Emission Source (M) - Secondary Dust Collection System This emission source contains water vapor, air, VCM, and small amounts of PVC resin. Two of the secondary collectors consist of aerodynes, four are Aerotecs, and two are Dustecs. The following tabulation characterizes this source. C/T v~. * 'm 6 - I? Emission Rate 6,800 scfm 8,800 scfm 9,000 scfm 9,000 scfm 24,000 scfm 25,000 scfm 16,000 scfm 16,000 scfm Vent Size 1.67 ft. diameter 1.67 ft. diameter 1.67 ft. diameter 1.67 ft. diameter 1.4 ft. x 1.6 ft. 1.9 ft. x 3.75 ft. 1.4 ft. x 3.75 ft. 1.4 ft. x 3.75 ft. Vent Height 50 ft. above grade 50 ft. above grade 50 ft. above grade 50 ft. above grade 64 ft. above grade 58 ft. above grade 58 ft. above grade 58 ft. above grade 7. Emission Source (Q) - Resin Transfer Discharge The tailings from the screener operation are air conveyed to a cyclone separator where the conveying air is separated and the resin subsequently rescreened and collected for sale as off-grade product. The conveying air discharges from the cyclone at a rate of 4,000 scfm through a duct eight inches in diameter at a height above grade of 56 feet. 8. Emission Source (N) - Storage Bins ; Sixteen product :stOTage bins afe utilized. Each bin is equipped with a bag type collector for separating the product conveying air from the product. The tabulation below characterizes the venting conditions from this source. Number of Bins 2 5 19 Emission Rate 1,600 scfm 1,725 scfm 3,750 scfm Vent Size 1.15 ft. dia. 1.25 ft. dia. 1 ft. dia. Vent Height 50 ft. 64 ft. 57 ft. 9. Emission Source (P) - Storage Silos Sixteen fluidized storage silos are used, each of which are all equipped with bag type dust collectors for separation of the product from the conveying air. No single discharge pipes exist for these collectors. Rather, the conveying air is exhausted via multiple discharge ports. The height of these silos vary considerably from 40 to 70 feet above grade. C/l w con E- V Avon Lake Plant - Suspension Process Emission Summary Emission Source^) Lbs/Hour A A B C D M, N, P, Q M, N, P, Q Solid & Liquid Losses Fugitive 60.8 VCM 30.4 PVC 109.8 VCM 136.1 VCM 104.0 VCM 68.1 VCM 119.1 PVC 356.2 PVC & VCM 682.8 PVC & VCM Lbs/I00 lb PVC 0..24 0..12 0..43 0..54 0..41 0,.27 0..47 /\ 1..41 2..68 Number of Sources (3) 140 140 7 19 2 41 41 (1) Average hourly for 1973 (2) Reference to Flowsheet (3) Emissions stated are per.total number of vents. fc % Cv* %N*t. CS o o cn E- S CONTROL DEVICES (General Discussion Relating to B.F.Goodrich Plants) 1. Monomer Emission Control Monomer which is not converted to polymer is flashed from the polymer slurry and condensed after compression to suitable pressure levels. Inerts present, in the system, however, will not be condensed and will be saturated with monomer. In B.F.Goodrich plants it is common practice to reduce the monomer released to the atmosphere with the inerts by passing the vent stream through a refrigerated condenser or by compressing the vent stream to higher pressures and recondensing before releasing the vents (F.mission D) . The main monomer condensers are cooled either with cooling tower water (85 F. summer temperature) or with refrigerated water (40 F.) in the newer plants. The amount of monomer leaving these condensers will'de pend upon the inert load in the system which is quite variable. However, the concentration of monomer will' be about 807, in the summer with a tower water cooled unit or about 72% with a refrigerated water cooled unit. Passing the vent stream from these condensers through a refrigerated vent condenser or raising.the pressure and recondensing will reduce the monomer concentration in the vented stream by removing 80-90% of the .monomer content. - Monomer concentrations in this vent depend upon the relative tempera tures and pressures used, a combination of high pressure and low temper ature being theoretically the best combination. There are, however, practical limitations. Since the gases emitted from suspension polymer ization processes contain water vapor, operation of a refrigerated vent below 32 F. can cause ice formation on the condenser tubes, thus limit ing the condensing capacity of the unit. A refrigerated vent operating at 40 F. and 40 psig will contain about 50%, vinyl chloride, while at 80 psig and 40 F. the monomer concentration will be about 30%. Lower temperature can reduce the monomer concentration further if the vinyl chloride is dry enough to avoid icing. It should be possible to decrease monomer emissions at this point even further (perhaps by 80%) by absorption in a solvent, then regenerating the monomer for reuse. The presence of water, however, introduces problems with emulsion formation. There are also problems of contamina tion of the solvent due to polymer formation, and control of solvent losses. Another possible alternative is adsorption on activated carbon or another adsorbent. Polymer formation in the absorbent with subsequent reduction in adsorbent capacity is a problem, however, Theoretically reductions of concentration of the order of 90% or better should be possible if polymer formation can be prevented. 2. Particulate Emission Control Particulate control in B.F.Goodrich plants (Emissions M, N, P and Q) is achieved with high efficiency particle collectors. Collectors on pro duct bins and silos are bag collectors with efficiencies of 99.9+% for C/T cn o o M r- & 2.-' particulate I'm,.:-:?, ion Control (continued) particles above 44 microns. Final collectors on driers are either bag collectors or high efficiency cyclone collectors with efficiencies in excess of 997,, for particles above 44 microns. Bag collectors represent one of the most efficient devices for removing particulates from gases. Where cyclones are used for particulate separ ation, wet scrubbers offer an alternate possibility for supplemental treatment of the gas streams if additional particulate removal is nec essary. An effective device of this type probably would remove 907- or better of the particles escaping a high efficiency dry cyclone collector. ). FUGITIVE EMISSIONS (General Discussion Relating to B.F.Goodrich Plants) Fugitive emissions listed for B.F.Goodrich plants represent the difference between monomer in and polymer out of a production operation which cannot be otherwise identified. In the tabulations for B.F.Goodrich plants, fugitive emissions may be either monomer loss or polymer loss. Some plants establish "loss factors" by adjusting measured or estimated losses to reflect fugitive emissions and in these cases the fugitive emissions may appear low or to be non-existant. Fugitive emission estimates are given in tabular form for our .various plants. E. EMISSION CONTROL PROCEDURES (General Discussion Relating to B.F.Goodrich Plants)' ^ A sincere effort is made in plant design, operation and- maintenance to minimize emissions of monomer to the atmosphere. Some design features and operating practices are listed below. Knock-Out Tanks Knock-out tanks are provided in vapor recovery lines leading to monomer compressors. These permit separation of liquid entrainment which would otherwise be transferred to the compressor's, causing damage and shutdown. In this way these vital pieces of equipment can be kept on stream a greater proportion of the total time. 2. Double Rupture Disc Assemblies Polymerizers and other pressure vessels handling polymer slurry are usually equipped with rupture discs to protect the vessels against rup ture due to excessive pressure caused by unusual operating conditions or exposure to fire. Occasional disc failures occur due to fatigue, incor rect installation or defective discs. In order to minimize the likelihood of these failures, the use of a double disc assembly is established practice in critical service, since the failure of two discs in series due to these causes is unlikely. Also, with the double disc assembly the discs can be mounted in the shop under close supervision, lessening the chance of a faulty field installation. 3. Vacuum Recovery Before opening pressure vessels containing monomer vapors, a vacuum is pulled on these tanks to remove as much monomer as possible through the recovery system. By this method as much as two-thirds of the vapor re maining in the tank at atmospheric pressure can be removed before the vessel is opened. E-7 e/r JO cn o o <2> EMISSION CONTROL r'i 'CEDURES (continued) 4. Reduction of Residual Monomer Before Drying Since most of the monomer remaining in the polymer will be vaporized and emitted to the atmosphere during drying, it is important to reduce the monomer content in the polymer as low as possible before this operation. Steam sparging is employed after completion of the polymerization to improve monomer removal. 5. Emergency Shortstopping Reactions can go out of control due to loss of agitation, loss of cooling water and other unusual circumstances resulting in vented charges unless preventive measures are taken. Emergency shortstopping of the reaction can often prevent this. Some of our plants are equipped with such a sys tern. 6. Organic Vapor Detection Within the last year our plants have adopted a procedure of frequent checking of possible or .suspected sources of leaks with an organic vapor .analyzer. With .this sensitive device leaks too. small to be detected by odor'can be located and the leaks eliminated. This is an effective pro cedure for reducing fugitive emissions. < Cvl vi O o o c- s CONTROL DEVICES AND ECONOMIC DATA AVON LAKE PLANT (Partial List) Control Devices for Emission N (6 Units) 'ine PVC dust is removed from a 2,000 CFM air stream that is exhausted to the itmosphere from PVC storage bins. These units are Flex-Kleen Model 84-BV-25 inits containing 250 sq. ft. filter area. The vendor states that these units vill remove 99.9+7. of particles over 44 microns. The size of the unit is 3' -4" < 3'-4" x 8'-7". The total installed cost of these units totalled $90,438 in 1970, when the ?quipment was retrofitted. Major equipment cost for this installation was 542,600. [t is estimated that the unit will last at least 15 to 20 years. Annual opera ring costs consist of: Air'for Blowback Maintenance Bags . :$ 400 400 800 $1,600 Control Device for Emission D fen as from the main recovery compressors is compressed from 45 psig to 85 psig _nto a vent condenser cooled with refrigerated water. The quantity of monomer ondensed in this unit has not been measured. However, calculations show that lbout 757. of the monomer carried by inerts from the main condensers should be conlensed in this unit. 'he total installed cost of compressor and vent condenser is estimated at $17,000 n 1971. Installation was made as part of the original plant. laintenance costs on this system are high due to tramp polymer formation in the system. In order to reduce maintenance downtime in this area, study of alterna:ive systems is now under way. Operating and maintenance costs are estimated as 'ollows: Electricity Refrigerated Water Maintenance.Supplies and Labor $4,000 400 5,000 $9,400 e- ? oooo a: a. WT si C3 o HK C- /D AMBIENT SAMPLING DATA B.F.Goodrich has conducted a limited amount of vinyl chloride grab sampling with subsequent gas chromatography analysis. This data long with the description of the sampling point locations relative to the plant and the wind vector occurring during sampling are shown on the attachments. The gas samples were taken using metal vacuum containers, 250 milli liter glass containers, tedlar bags, or syringes. The gas was then transferred to the gas chromatography via the appropriate size injection syringe. A typical gas chromatography used for the analysis was a Hewlett Packard Model 5710A or equivalent equipped with a 8'x 1/8" Parapak QS 50/80 mesh column. For this work* the gas chromatography was calibrated using two vinyl chloride standards of 10 and 100 parts per million. The appropriate concentration to establish a calibration curve (area under the VCM peak) down to 0.05 parts per million was obtained by dilution of the standard volumes. In obtaining the preliminary ambient data,'the sample locations A - L were placed bn the wind vector occurring during sampling period rather than the prevailing wind vector. Therefore, the sample points were not stationary and no conclusions could be drawn concerairg the long term effect of vinyl chloride on a specific residential area, etc. surrounding the plants. The purpose of the measurements were to determine orders of magnitudes of concentration up and down wind from which to generate future sampling requirements. Cor n tra tio n s E xpressed In Farts Per BM illioCn By VDolu Sample L o ca tio n A (ppm) co m 6m cm O o O) -, O O Ci % M* M* NT" --1 co CO CO co t--i >--< co o co >< < W kJ CQ S<H< X<I P< CD O O (M X 'TON CU < 2Q O o 0 o^ CO <N ^ CO C\ H r-H tO >o O Q 2; CU 2 O O -=r rr' >-< CU m* m* q< q) o)o> O 2; m* ^ t-* w ooo CN V? * o *- cm co -<n uo ooo O O O r-i CD CD CD c"U* C*NNN Vi in \d cm o CM r-t o oo O O O *-l CM CM CM W CQ a S < < Q >< cu o2 o o (N * ^ 'I in CO CD o o LO r-t O CM O O O O r-t CM CM CM in o H n ONN * ooo CU 0) ro yA c o * x: o <ao) c o5 *-> CQ o o> U c o T<D) yA CU * or >, u C Q) ffi > cn **"f A o yA >< w % <C t--~i CQ O CU H < < y~i Oa O CD Cn CO CM O CM < o oo W yA CQ t-H < >< < Q < CU 2 O O O 2; m1 in m* cm o oo CD CD CD CO CO CO O OO <& --j --i o or > r( (0 o E 0) H X) ro co > < ta 4~t co Q tD c r~t a--I fD CO < CwU * L_ CV1 - SAMPLING LOCATION RELATIVE TO PLAN' MILES v? S'-Y V 0.4 0.5 G 0.6 0.8 J` t.O -.5 3.0 F 2F*--A-- I L K Y WIND DIRECTION c>j C/[ c; o ** ^ - 13 S' ;<r. Vn <t fvH u 2- i- <o t *.. ifMVJ:--'- - _s.. _=v=_! -^.o v> -* __-__'___n_T__v__ *__'_ . v9 ^ ' iL uvoh. i1":iNfiooflyooAiDP7M0VmWI- -_v_5___!(_0_._.>._. y^no .hh/" j ___________ IV. n^0My& * ^rirFN^'^ Ills' - : {.// ' <L IT l;P >? && g>;::1if*: >:v* -i > r (--<,^ 7V- " avo :s AiMnODr ;,:/!MnOD *f / /#'* ;/ i*. I, a* /tell)r?i v //--7/-- VmvnvNOI "* 1^.04 MirtOf h ifc1M2 iv ... __ Hint----lOVOM-- if . ________ J. r. ';l\\ c-: i/TTZnA Tv.;-', \TfrdliiNOtirp^piQ/vnQttns j j.nvy 3>iV7 Noiw/ A>*4^n^li&wi4ra%a^.aT';% jj* - - . , aW sc: '.' . "' Jy ULgvoh-----------------'L i*. .*'.- o ,'v ** ' *;<- <>jVJ U' K*vi C~< O ^ *41 Cl