Document bOb4xQdja6yaBg6mbXqg06ZyZ

a+ict C&&meca& ESCAMBIA. PLANT INC. INTEROFFICE MEMORANDUM To: Mr. J. T. Barr Dote: 14 June 1974 From: Mr. D. H. Francke r#; E. P. A. Request for VC1 and PVC Emission Data Copies: R. E. Jones A. K. McMillan E. A. Primeau Attached please find a summary of the information requested by Mr. D. R. Goodwin of the E. P. A. regarding the vinyl chloride and polyvinyl chloride emissions from the Escambia PVC Plant in Pace/ Florida. This plant has a capacity of about 75 million pounds of PVC resin per year. Calculations were made using data collected for monthly material balances start ing in February, 1974. The data runs through May, 1974. Below is a summary of the products and losses as percentages of the total vinyl chloride charged for . the February-May period. Vinyl Chloride Charged PVC Resin Sold Known PVC Losses to the Air Known PVC Losses to the Water Known PVC Scrap to Dump Known VC1 Losses to the Air Known VC1 Losses to the Water Total VC1 Accounted For Total VC1 Unaccounted For 100% 94.79 1.24 .29 .64 ' 1. 18 .04 98.18 1. 82 Total VC1 Out 100.00% DHF:aw Attachments D. H. Francke AP00027793 PROCESS AND EMISSION DATA A) General Process Description Purchased vinyl chloride monomer is suspension polymerized to produce various types and grades o PVC resin. In the polymerization step, only 83 to 87% of the vinyl chloride is converted'to PVC resin. Therefore, it is necessary to vacuum strip at the end of the polymerization cycle to remove the unreacted monomer which is then condensed and recycled to the polymerization step. The recovery and recycle of the monomer is done in a stripping vessel using vacuum recovery equipment, compressors and condensers.. Following the stripping operation, the slurry containing PVC resin and water is stored in blend tanks and then pumped continuously to a centrifuge for dewatering. The water effluent from the centrifuge is treated through a pre-treatment system and flows via an industrial sewer system to the biological system for final polishing prior to discharge to Escambia Bay after approximately 100 days of residence time in the treatment system. The dewatered wet cake from the centrifuge solids discharge is fed to a dryer where the moisture content of the resin is reduced to 0. 2%. From the dryer the resin passes through a screening operation to remove, resin particles larger than 40 mesh and is then sent to storage. From storage the resin is bagged or bulk loaded to hopper cars for shipment to customers. .f AP00027794 S6ZZ3OOOdV re a c to r VC1 le ve l d u rin g cleaning is about 25 ppm- (fro m B endix A n a lysis). y > o 4* J3 O' Pd N N rd m vd C4 00 *h*-. oncOo m po O V o* o > 3 Z 3 34& J0B < 2 < z < 2 i < 3 u S 'O> o M *-4 NO nJ m *rsa Ho mm 0O0 CoM O o 4) o 9 2 3 3 i i o <o i i < Z g Vi a. o V u oEl a "8 B o hP< 0o u >* " I.o ro pn O o Om o o mH Hm oo oo 00 oo o No o >o . io O' o m * io >c fl oo -< >o s(TMM" B4* cs>D -4 4c ^O4' i fk rO' mk sl 0 2 CO CO d rd *0o *- >* ri Vrtl 1 3 u> 0u ft it 'iZ m ^ 2 5 V v# 4) C *a lVi +0* d u 3 O o tS 3 < << Z 22 4< 3 4* <<< 222 >S s8 ** ro* rt W f--oI h ft o U Vi > ft B U 'q & 2S50g.S3f|*a *9H s a f-4>4> > 4>-> V44) <4 50 U 4o- u (44) K O 3 V-H 4J a > M4) a 4) 1 V 4(OJ0 u u 4ou- tad> >4 4) a aH <4 s4 8) > 0 U4) (0 o h 4> 1 ^4 m S 0 > a 01 H *0 4) A oo 045) e 01 4) V &0 Qc o i u 1H0 4aor4l Q u <D a C*-3r >L*s U Vi >* .0 ioi 400)i > 50 O tn > h k a H<t 4) fk 0 to 'O 3 rt B 9 50 o 14 Q JS 00 k 3 w k & >4 r(-4 \UJ .&5 4o-> o r4i) o 4ur4t 4) tf mol *(H4 *s g O ft 444) Zr> <d E $. fc g 4CoH4 <d > Vi 0 o ft rt O > ft r~ N vO PO B 0 X>I. B T) 0 0 T*43 > 0 fc sO I ac - 0 > (4 c4 Q 44)) u H (4 S4) o co 50 u * 96ZZ3000dV d 0) =1 0 .3 4< c O O J3 U > > -< CO >o 3m O o d Ota CD O cn <0 .o s.drl o >0 o TQ N bo d 0< CO + iNO o H ** ^ Ofc CO U 4o-* o d . %P ta * no jC U u O> j: te <m M oo vO o O Ota tHa< bo o Ota s 0crt0 o 8n U o c d .5 u d41 h h 4O-* a O %o d a o M cn A 45 3o o I a ni s 3 Z: O > M A .O * m oo *-> o 00 o* I o ! o in >S d s *o A tal NO o A tat JS Ui O 4o> o d oo !* TdO tai M 00 o ta CfMO o <sf emg o> to II u O > * *o4 00 vO o II 00 H 4d- E 0 4- H 1-4 Ota tal K rsj eg in M h nlU Q|n ji emo ; V.?r. C) Control Devices Source 1 A (Reactor Safety Valve) - No control device, no device should be installed since this device is used to prevent over pressuring of a reaction vessel. Source 1 B (Reactor Jet)- No control device, investigation is currently being made into the use of a Nash Vacuum Pump and Compressor to replace this jet. Reduction ould amount to"$5% of current losses ( 2 lbs/hr reduction). Source 2 (Stripper Jet)-No control device, use of a Nash Compressor and/or vacuum pump to replace the jets. Reduction of ^85% of current losses or 12 lbs/hr Source 3 (Recovery Vent Condenser) - A 3 stage condenser, involving a cooling tower water condenser, refrigerated water condenser and a freon condenser. The flow is 90/10 VCI/H2O at a rate of about 2010 lb/hour. Inlet vapor enters at 70F and 80 psig and exits at 40F and 70 psig. Exit flow is about 8CFM average. Liquid VC1 and water are removed prior to vent. lbs, recovered Efficiency ( lbs. entered ) =* 99%. Source 4 (Blend Tank Vent) - No control device.- This is a saturated air stream at 120F and a 1500 CFM flow and <. 5% VC1 concentration. Source 5 (Dryer Discharges) - Current control devices are Dustex Bag Collectors. Inlet flows to the bag collectors average 9, 122 lbs/hr. of PVC and 18.2 # VCl/hr. with air flows of 34 MCFM @ 120F - 150F. Discharge flows average 34 MCFM @ 120F - 150F with 117# PVC/hour and 18.2 # VCl/hr Particulate efficiency averages 98. 7% with major losses due to bag breakages. New collectors have been ordered. These are dust cyclones and will reduce PVC losses to about 10 lb/hr. AP00027797 C) Control Devices Page 2 Source 6 (Purification Condenser) - This control device consists of an atmospheric condenser which is mounted on top of the purification column. This condenseris cooled with well water. Flows into the condenser are 1, 720 lbs/hour of a 99. 7/. 1/. 2 mixtmre of VCl/H20/inerts. Vent losses are estimated at 19. 75 lbs/hour of a 85/15 mixture of VCl/inerts. Current plans include a revision of the recovery system to eliminate the need for the purification systems and thereby elimination of this emission. AP00027798 D) Fugitive Emissions Total unaccounted losses average 170 pounds/hour. These unaccounted losses include valve stem leaks, pump seal leaks, etc. AP00027799 E) Emission. Control Procedures 1) Process Design Features The reactors are tied into the strippers so that the stripper can be used to drop "hot" reactors into to prevent venting to the atmosphere in case of an uncontrolled reaction. In addition to this feature, an overhead recoveryheader is available to permit venting to a recovery- compressor of any hot reactor instead of venting to the atmosphere. A VC1 Recovery System is used on each reactor batch for the recovery of VC1 and to lower the VC1 released to the atmosphere.......... . 2) Plant Operating Procedures Current operating procedures for the control and prevention of emissions are as follows: a. In the case of an uncontrolled reaction which cannot be dropped for recovery, addition of a short stop agent which stops the reaction, is made. This eliminates the pressure generating reaction and stops any further pressure buildup. b. During each reactor cleaning, the rupture disc is checked for buildup, deteriation, etc. and if replacment is indicated the disc is replaced. c. Headings are taken throughout the plant on a 3 day/week basis and high readings are traced to the source. Upon source detection, T. corrective action is taken. The instrument used is a hydrogen flame analyzer which detects combustable vapors. This instrument is manufactured by the Century Instrument Company. AP00027800 E) Emission Control Pr ;dures Page 2 3) Maintenance Procedures In addition to regular maintenance to prevent equipment failure and resulting losses, a yearly inspection and testing program of the safety relief valves on the plant equipment is carried out. This program is designed to assure that the relief valves will not operate prematurely or fail to operate, thereby causing unnecessary vapor losses. AP00027801 Z08ZZ000dV JO JT O V S 2 2 2 4X jss I* ft" 2 2 si1 +* * oft*; F4 *o< ^- :|8 a S| 2 2 *i! 2 2 3 alj H 0.5 o o JS ? 9 S_ l* 0 1 ft>t iUfat oj ft K 8~ S3 u a ?2 A52 o- ,? 4^''.. > i \ NNN --A r ^ r < ai r* * t iOnk pOk s 0< m^Ok iOo*k sO 0 HO >oo c ga *3 ff u t Q t! 6 r. *fcoet *fsfool>tt 88 uu M M0 w3 w3 **8 - N i ft (m r ft a & g * E o H uU ft kz St 25 K ? a5 kJ 2*S-S35 Qw oo fi <- o 3u. C t o P* o O II < 2eo 23g u* * ft gs .s V 0-4 VM c **o 'Co ft *V0 O & p. V It o ? Bj Ju tftt* ff 2 :e o 3 t u c -eo o* 0 O eak atft>* 3*ft o frtt fitt H 08lZ000dV tt 2 -* 35 ^ * 2 *n 2 2 s 5 H * O B III- w g < oJ OS 8 si > o, 2 1- &o 3 e HO 2 o 2 o- si cv a0 :000. e0<0J o >^ Cu U h *3 0 >u GO *1 m 2 0e o5-* tfSb*at || Um zi**>r. Kre > Ambient Air Da ta. Attached is a plot plan of the Escambia PVC Plant showing ambient concen trations of VC1 as measured with a Century Combustable Vapor Analyzer. These concentrations are "area" type analysis and are not a specific points. The numbers are averages of about 57 readings taken from 11 Feburary 1974 through 5 June 1974. The Century Instrument used analyzes the air for com- bustables by way of a hydrogen flame ionisation. The readout is in ppm range and is a total combustable analysis. Background levels normally run 3 to 7 ppm with this instrument. Additional ambient air sampling tests run during the week ending 15 June. 1974 are tabulated below. These data were obtained by gas chromatographic analysis . of the b ample s taken. Date Wind Point Location 10 June 1974 0-3West Millroom-Lab PVC Office f l 11 June 1974. > 0-3 West PVC Office PVC Lunchroom 2 blocks south of PVC Building 1 block south of PVC Building Z 8 4 5 12 June 1974 0<*atWNkst PYC Lunchroom jjSSfc- 1 block north of ''-Sks- " pvc Building 1 block east of ,, PVC Building Pilot Plant Maintenance Warehouse . Millroom Lab 3 & 7 1 1 1 VC1 Cone. Type .2 Carbon Tube .4 Carbon Tube .4 Carbon Tube . 7 Carbon. Tube . 1 Carbon Tube .5 ' Carbon Tube .3 Carbon Tube . 1 Carbon Tube 1. 3 1.4 . 13 .5 Carbon Tube Carbon Tube Carbon Tube Carbon Tube AP00027804 -2- 13 June 1974 0-3 East Millroom Lab / Lab 10 PVC Office Maintenance Warehouse 2. 9 Fence North Amines Cooling Tower H IZ North side of PVC bldg. South side of PVC bldg. 13 (4 1 block north of PVC bldg. (, North Fence // 1 block downwind of PVC bldg. IS 2.4 2.9 1.5 .5 .5 .5 4.7 .6 .3 .6 Grab Grab Grab Carbon Tube Carbon Tube Carbon Tube Grab Grab Grab Grab 4.1 Carbon Tube Analysis of the water streams from the PVC Plant on June 13 are listed below: Stream Exit from. Reactor Building Exit from the Finishing Building Inlet to Foxtrot Pond Exit from Foxtrot Exit from Echo Exit from Charlie Exit from Hotel Exit from Bravo Exit from Alpha ddixj VC1 21.5 .9 15 1 1.2 J {69. 0 ppm is a 3 month average ...from the Material Balance Analy: First PVC Treatment Pond * 0 0V \0 Intermediate Ponds 0 0 Discharge to Bay , AP00027805 Emission Point Data Emission point data are as listed below and discharge points are a8 marked on the plot plan. Emission Point* Location Height (A) Configuration Flow Velocity Temperature Vent #1A VentJlB Along walls 45 * 55 8" dia. 4" dia. 1 NA ~ 140F 2170CFM** 6200 ft/min ^ 200F Vent Z Vent #3 Vent #4 Vent #5 jf (2 Jets) * * + 55 4" dia. 2170CFM** 6200 ft/min ^200F 55 1" dia. 7. 78CFM 1430 ft/min 40F 55 11 3/4'* dia. 933 CFM 1250 ft/min 120F 60 North 20 1/2*' dia. 11MCFM 4900 ft/min 150F South 25x28 23 MCFM 4900 ft/min 15 0F Vent #6 * 45 1/2" dia. . 57 CFM 42 ft/min 70F Vent #7 * 45 10" dia. 2000 CFM 3670 ft/min 70F Vent #8 * 45 6"x6" 2000 CFM 8000 ft/min 90F * See paragraph B under Process and Emission Data ^ I = Intermittant **Based on usage of 870#/hr/jet steam usage (200#steam) plus 140CFM jet capacity AP00027806 AP00027807 10-13 Jl/ME I<\74 VCL STORAGE "Date z!1,h4 6/^ WC. PlawT P<-ot Pt-AfV S 7 AA/ AP00027808 POLYVINYL CHLORIDE PLANT <ii/>i > oQ < AP00027809 EPA broadens its definition of carcinogen A new and apparently precedent-set ting criterion for determining the car cinogenicity of a substance was used in Environmental Protection Agency ad ministrator Russell E. Train's Oct. 1 decision to suspend all major uses of Shell Chemical's aldrin and dieldrin pesticides. Mr. Train is of the opinion that, for purposes of "carcinogenicity testing," tumorogenic substances and carcinogenic substances are "synony mous." This interpretation could have a far-reaching impact on EPA regula tion of chemical substances. However, it is in sharp contrast to policies of other agencies, such as the Food and Drug Administration, which define carcinogenic substances as those that actually cause cancerous tumors. Since laboratory evidence has linked aldrin and dieldrin to liver tumors-- benign and malignant--in mice and rats, Mr. Train says that he has con cluded that the chemicals also can pose a "high risk of causing cancer in man." According to an EPA enforce ment official, this criterion will guide future EPA rulings concerning cancercausing substances. Indeed, an official in EPA's pesticide divieion tells C&EN, the words "tumorogenic" and "carcinogenic" won't be in the pro posed pesticide regulations that EPA intends to issue next week. In their place, "oncogenic" is used to designate any substance that will produce tu mors--cancerous or otherwise--in test objects. In Mr. Train's suspension order, he explains his criterion and why aldrin and dieldrin are carcinogenic to man. He defines a carcinogenic substance as "one that increases the incidence of benign or malignant tumors in exposed animals, decreases the latency period between exposure and onset of tumor, or results in unusual tumors." To Mr. Train, the distinction between benign and malignant tumors is not a reliable indicator of carcinogenicity, because, "in the thinking of most experimental ists, the induction of a benign tumor is merely a stage in the subsequent oc currence of a malignancy." Hi* opinion is backed by a Health, Education, and Welfare advisory panel that has recom mended that a "finding of carcinoge nicity" be made when a substance is "judged positive for tumor induction in one or more species." Although Shell protested during EPA suspension hearings on the pesti cides that mice were an inappropriate test species and that there is no corre lation between mice and human be havior, Mr. Train points out that not only does Shell use mice in their own cancer experiments, but that mice re sponse is similar "pathologically" to that of man. In addition, Mr. Train re futes arguments that carcinogens are species specific, in that, for instance, aldrin and dieldrin might affect mice but not any other species, including man. He cites research data from orga nizations such as the International As sociation for Research on Cancer, which have found species specificity "unsubstantiated." Further, Mr. Train says it is impos sible to establish a "safe level" of expo sure of aldrin and dieldrin to man be cause no one has been able to deter mine the relationship between the dosage of these two chemicals fed to test animals and the subsequent can cerous response. Nor has anyone come up with a threshold level of carcinogenic effect for the chemicals, Mr. Train points out. He does think that there is a theoretical "no-effect" level although it has not been demonstrated conclusively in laboratory studies. The lowest tested dose of 0.1 p.p.m. aldrin Cancerous cells show up In this mouse tumor. SPA may now Interpret resulte from such animal studies differently and dieldrin still produced "significant tumors" in mice. Anyway, Mr. Train concludes, any substance that will induce cancer in animals "at any dose level, no matter how high or low," should be treated with caution. He admits that high doses of the chemicals were given to the test animals, but points out that, with a limited number of animals, this is necessary to determine gross effects. For instance, Mr. Train says, a chemi cal that causes one cancer in every thousand men is considered disastrous, but there are very few laboratories that can accommodate 1000 animals and even then, he says, one incidence of cancer "might be disregarded as chance." Suspension of the major uses and production of an estimated 10 million pounds of aldrin and dieldrin in 1975 will remain in effect during cancella tion proceedings, which, when conclud ed, would ban permanently production of the pesticides. Cancellation hearings on the risk vs. benefits of the chemi cals have been in progress for a year, and no early end to the hearings seems in sight. j j ij j Ij 11 j /i j / Oct. 14. 1974 CiEN 13 U<lLlol islcL^ receive-- 0GT1 8 1974 APCl CHEM. GROUP LAW DEPT. AP00027810