Document JrpBke2RVv9R95mr9vqDwd3rv

CONFIDENTIAL EMISSIONS TEST REPORT NEW REACTOR D-700 CONOCO CHEMICALS COMPANY P. 0. BOX 91 ABERDEEN, MISSISSIPPI 39730 JUNE 21, 1982 SAL 0001101.17 EMISSIONS TEST NEW REACTOR D-7QO CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI I. INTRODUCTION A. Source Information B. Testing Personnel C. Process Description II. SAMPLING RESULTS SUMMARY A. Emissions Test Results B. PVC Production During Tests III. SAMPLING PROCEDURES A. Sampling Apparatus B. Sampling Procedures IV. ANALYTICAL PROCEDURES A. Analytical Apparatus 1. Slurry Samples 2. Reactor Opening Loss Samples 3. Reagents B. Analytical Procedures 1. Slurry Samples 2. Reactor Opening Loss Samples V. CHAIN OF CUSTODY VI. APPENDIX - Operating Data Sheets Sampling Data Sheets Analytical Calculations Analytical Data Standard Gas Certifications Reactor Opening Loss Calculations CHFIDCHTUL SAL 000110118 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI I. INTRODUCTION In accordance with 40 CFR 61, Section 61.67 an emissions test was conducted on June 8 and 10, 1982 for the newly constructed PVC polymerization reactor D-700. The purpose of the test was tQ demonstrate compliance with 40 CFR 61, Sections 61.64 (a)(2) and 61.64 (e) (1)Cii) . A. Source Information The emissions test was conducted at: Conoco Chemicals Highway 25 Aberdeen, Mississippi 39730 Questions concerning the testing or this report should be directed to: R. A. Frohreich, Chief Process Engineer Telephone Number (601) 369-8111, extension 2239 Dave Mahler, Environmental Engineer Telephone Number (601) 369-8111, extension 2285 B. Testing Personnel The testing and analytical work was conducted by Conoco Chemicals Company personnel at the Aberdeen facility. Personnel involved in sampling and sample analysis were: Jimmy Grace - Quality Control Technician Dave Mahler - Environmental Engineer Wayne Myers - Engineering Co-Op Student C. Process Description Vinyl chloride monomer is received at the plant site in railroad tank cars and is transferred by compressors to the VCM storage sphere. The VCM is then pumped to fresh VCM batch tanks which are located in each of the two reactor areas. &&L. 00U0119 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI INTRODUCTION (CONTINUED) C. Process Description (Continued) PVC is produced batchwise from the vinyl chloride monomer. The plant produces only one type of resin; that by the suspension polymerization process. The reactor is first . evacuated to remove air and inert gases and is then charged with carefully measured quantities of VCM, water and a suspension agent. The reaction is initiated with a peroxidetype catalyst. Following completion of polymerization, the reaction is killed and the unreacted VCM is removed from the reactor in the recovery sequence. The recovery system is designed to recover unreacted VCM for recycle to subsequent batches. During the pressure recovery phase, VCM vapors are pulled from the reactor and compressed, condensed and liquified. The liquified VCM is routed to the recovered monomer tanks. As a part of the recovery operation, steam is injected into the slurry to aid in the removal of residual VCM in the slurry. When the reactor pressure approaches atmospheric pressure, vacuum pumps are started and the reactor pressure is reduced to a vacuum. These VCM vapors are also compressed and liquified and the liquified VCM is routed to the recovered monomer tanks. By this time the slurry has been brought to a boiling condition by action of the steam injection and the vacuum. The steam blows through the slurry and purges the vapor space above the slurry before being pulled out of the reactor by the recovery system. Cooling water is then put on the reactor in order to cool the slurry down prior to dumping the reactor. When the recovery step is complete, the PVC/water slurry is dumped through a slurry screen which removes oversize material.. The slurry is "tdeh^pumped to Blend "tank's.- After product^pumpdutV tlie reacfdr is rinsed to the blend tanks, which completes the reaction batch cycle. The PVC/water slurry is pumped from the blend tanks to a centrifuge in the drying area, and the extracted water is discharged to the process sewer. Wet cake is fed to a dryer in which- the resin is dried. The dried resin flows through a product sifter where oversize material is again separated and collected. Resin is then pneumatically conveyed to storage silos and eventually shipped from the plant in trucks or hopper cars. SAL 000110120 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI CfclDEHT II. SAMPLING RESULTS SUMMARY A. Emissions Test Results The following is a summary of results from sampling and analyzing the reactor D-700 vapor space and dumped slurry for each of three batches. D-700 Batch Date Resin Grade Vinyl Chloride Emissions (Lbs. VCM/MM Lbs. PVC) Vapor Space Slurry Total D-700-223 6/8/82 5385 32 82 114 D-700-224 6/8/82 5385 31 57 88 D-700-229 6/10/82 5305 96 107 203 The process employeed at this plant utilizes the reactor for polymerization and stripping. Because VCM flashes out of the slurry and into the vapor space while the reactor is being dumped, it is impossible to accurately segregate the slurry residual and the reactor opening loss. The standard specifies maximum emissions of 400 lbs. VCM/MM lbs. PVC from the slurry and 20 lbs. VCM/MM lbs. PVC from the reactor vapor space. Thus, a maximum emission of 420 lbs. VCM/MM lbs. PVC from the entire reactor is applicable to reactor D-700. The results of the emission tests summarized above demonstrate that reactor D-700 was in compliance with Sections 61.64 (a) (2) and 61.64 (e) (1) (ii) for each of the three reactor batches tested. The total emissions from reactor D-700 did not exceed 420 lbs. VCM/MM lbs. PVC produced. B. PVC Production During Tests -"The Aherdeeh^planC had"~a"ri'omi'nal PVC produerton ^capaclTjr of - - ' 450 MM lbs. PVC/year at the time of the D-700 emission test. This corresponds to a daily average of 1,233 M lbs. PVC/day. During the two days of emission testing the plant produced 1,007 M lbs./day and 712 M lbs./day respectively, which is less than the nominal plant capacity. The maximum production rate can only be achieved in the winter time when colder cooling water temperatures are available which allow for higher catalyst loadings. Furthermore, the maximum production rate requires perfect reactor scheduling without any reactor or dryer downtime due to maintenance or other delays. ------ SAL 00011012 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI SAMPLING RESULTS SUMMARY B. PVC Production During Tests (Continued) Since the plant utilizes a batch polymerization process, the plant's production rate is increased or decreased by processing more or less reactor batches per unit time. The size of the reactor batches varies only slightly over long periods oftime for any specific grade of resin produced. During the emission testing, reactor D-700 was charged with quantities of VCM which are representative of the current batch size for reactor D-700. SAL 000110122 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI III. SAMPLING PROCEDURES A. Sampling Apparatus Reactor opening loss samples were taken with the apparatus shown on Figure 1. The components of the apparatus are described as follows: 1. New, unused teflon tubing; 6.4 mm OD. 2. Stainless steel plug valves with teflon plugs, stainless steel hose connectors and interconnecting tubing. 3. 15 liter Calibrated Instruments multilayer gas bags. 4. Rigid container, sealed airtight with RTV gel sealant. 5. Gast vacuum pump. 6. Calibrated MSA sampling pump rotameter. 7. Needle valve. 8. h" Tygon tubing. B. Sampling Procedures The reactor slurry samples were obtained from a sample tap on the reactor dump line. The sample line was kept open throughout the dump period, which purged the line between samples. The three slurry samples were grabbed 5, 10 and 15 minutes into the dump period. 2-ounce samples bottles with screwed on covers were used to collect the samples. ^Immediately following sampling, the sample bottles were taped_ shut with electrical tape-and labeled with The date and" ' reactor batch number. The samples were then stored under refrigeration and analyzed within 24 hours of collection. Reactor vapor space samples were taken immediately after the slurry dump and before the rinse. The sampling train was tightly assembled, using a sample bag which had been leak checked overnight. The sample bag was evacuated until the rotameter indicated zero flow, and the valve was shut. Reactor D-700's top manway was then opened, and the sampling line was lowered into the top of the reactor. The sampling line was purged at a rate of 1 to 2 liters per minute for approximately 1 minute. Immediately following this purge, the sample was taken. The top of the reactor, the middle of the reactor and the bottom of the reactor (within 6 inches of the bottom) were each sampled for 5 minutes, in that order. Flow through the SAL 0001103.23 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI III. SAMPLING PROCEDURES A. Sampling Apparatus (Continued) sampling train was maintained at 0.5 liters per minute throughout the sampling period. At the end of sampling, the valve above the sample bag was shut, the pump turned off and the sampling apparatus disassembled. The sample bag was transferred to the plant's laboratory and analyzed within 24 hours of sample collection. SAL 000110124 h FIGURE 1 REACTOR OPENING LOSS SAMPLING APPARATUS Tygon SAL 003J01, VJ EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI IV. ANALYTICAL PROCEDURES A. Analytical Apparatus 1. Slurry Samples The following apparatus was used to -analyze slurry samples for residual VCM: a) Vials, unused with Septa and aluminum seals. b) Analytical balance, with minimum sensitivity of 0.001 gram. c) Syringe, 100 ul, Hamilton #710. d) Vial Sealer, P-E No. 105-0106 e) Filter Flask, 250 ml capacity. f) Buchner funnel, 7 cm. g) Filter paper, Whatman No. 4 h) Pans, disposable aluminum i) Cotfk borer j) Oven, Blue-M natural draft k) Gas chromatograph, P-E F-42 Headspace Analyzer l) Chromatographic column, stainless steel 2 m X 3.2 mm, containing 0.4% Carbowax 1500 on Carbopak C. m) Thermometer, minimum range 0-100 C, minimum sensitivity 0.1 C. n) Sample Tray Thermostat System, Haake Model N3B. o) Integrating Computer, Spectra Physics, SP 4000. p) Barometer, sensitivity 0.1 mm Hg. q) Thermometer, accurate to 1C, near GC. r) Recorder, Houston Instruments potentiometric strip chart recorder set on 1 mm full-scale response and 1 cm per minute chart speed. 2. Reactor Opening Loss Samples . The following apparatus was used to analyze gas bags for VCM: a) Gas chromatograph, HP model 5711 A with a 1.0 ml heated sampling loop in a manual sample valve and flame ionization detector. b) Recorder, Perkin Elmer potentiometric strip- chart recorder set on 1 mv full-scale response and 1 cm per minute chart speed. SAL 000:110126 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN. MISSISSIPPI IV. ANALYTICAL PROCEDURES (CONTINUED) A. Analytical Apparatus (Continued) 2. Reactor Opening Loss Samples (Continued) c) Chromatographic column, stainless steel, 2.0 m X 3.2 mm, containing 80/100 mesh chromosorb 102. d) Flow meter, rotameter type with 0-200 ml/minute capacity, with control valve. e) 6 gas regulators, dual stage brass. f) Thermometer, mercury type, accurate to 1 C, mounted near the sample loop. g) Barometer, accurate to 0.1 mm Hg. h) Pump, ITT diaphram. i) Tubing, teflon, 6.4 mm O.D. j) Sample bags, Calibrated Instruments multilayer bag of aluminized Mylar. k) Data Processor, Spectra Physics SP4000. 3. Reagents The following reagents were used to analyze slurry samples and gas bags for VCM: a) b) c) d) e) f) Nitrogen Gas, zero grade. Hydrogen Gas, zero grade. Compressed Air, zero grade. Calibration Gas, 496 ppm VCM! in nitrogen, certified. Calibration Gas, 2188 ppm VCM in nitrogen, certified. Calibration Gas, 4015 ppm VCM in nitrogen, certified. B. Analytical Procedures ~1. Sinrry Samples ~~~~~ - "-- The following procedure was used to analyze slurry residual VCM: a) Sample Storage Slurry samples must be stored under refrigeration. Analyses will be performed after sample has cooled to at least 30 C and within 24 hours of collection. SAl- 000110127 X EMISSIONS TEST NEW REACTOR D- 700 CONOCO CHEMICALS COMPANY ABERDEEN. MISSISSIPPI IV. ANALYTICAL PROCEDURES (CONTINUED) B. Analytical Procedures (Continued) 1. Slurry Samples (Continued) b) Slurry Samples Tareavial with septum and seal and record weight. Tare and record the weight of an aluminum pan (label pan). Set up vacuum aspirator and filter flask assembly. Pour slurry into funnel and start vacuum until steady stream of water stops (15 sec.). Remove vacuum. Use the cork borer transfer about 1.5 g. to the vial and seal it. Take another plug for solids and transfer to an aluminum pan. Quickly weigh and record the pan and save this sample for determination of total solids. Weigh and record the vial weight. c) Preparation of Gas Chromatograph Flame should be lit at least, on^ hour prior to analyses. Nitrogen pressure of 1.4 kg/cm gives 32 cc/minute. Set hydrogen at 2.1 bar and air at 2.8 bar. The oven should be set at 65 C, the neddle at 150C, and the injector and detector at 150C. d) Programming the Chromatogrpah Set "1" at 3 seconds, "A" at 07, "BF" at 09 and "S" at 2 minutes. Set "Analysis + Backflush Times" toggle switch at "X0.1" position. Both range and attenuation should be at a value of "1". e) Preparation of Sample Turntable The bath should have been on at least 4 hours prior to use. Check the temperature in the turntable to insure it is at 90 + 0.5C. Insert vials in turntable as follows: Positions 1 and 2 are old standards in the vials from a previous run; positions 3-6 are the calibration standards; positions 7-30 may be samples of any type.- Condition all samples including standards at least one hour (not to exceed five hours) before analyzing. After conditioning, start the chromatograph program and recorder. SAL 000110128 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN. MISSISSIPPI IV. ANALYTICAL PROCEDURES (CONTINUED) B. Analytical Procedures (Continued) I. Slurry Samples (Continued) f) Determination of Total Solids Place the label aluminum pans in the oven. The oven should be at 110+5 C. Allow the samples to dry for 45 minutes. Remove and cool to room temperatures. Record the weight and calculate the fraction of total solids (dry weight divided by wet weight). Record this value on the data sheet. g) Calibration Standards Calibration standards are prepared by filling the vials (at least two sets per series of samples) with the vinyl chloride/nitrogen standards, rapidly seating the septum and sealing with the aluminum seal. Use only stainless steel lines from the cylinder to the vial. The sample line from the cylinder must be purged into the hood prior to filling vials. After line purging, place the end of the tubing into the vial near the bottom and allow purging of the vial for one minute at 1000 cc/minute. (Hold the vial with finger tips only to minimize temperature affect of hand). Slowly remove tubing and quickly cap the vial with the septum and seal. Inject 100 ul of deionized water into vial. 2. Reactor Opening Loss Samples , The following procedure was used to anlayze reactor opening loss samples for VCM: a) Sample Storage Samples must be kept out of direct sunlight during storage. Analysis will be performed after the bag has equilibrated to laboratory temperature and within 24 hours of sample collection. b) Sample Recovery Connect the sample bag to the GC sample valve inlet direct with Teflon ferrules. Prepare the equipment such that the sample gas will pass through the bag, through the sample loop, through the pump, rotameter, and flow control valve at a rate of 100 ml/minute. SAL 00011012? EMISSIONS TEST NEW REACTOR D-7Q0 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI ANALYTICAL PROCEDURES (CONTINUED) B. Analytical Procedures (Continued) 2. Reactor Opening Loss Samples (Continued) c) Analysis The GC should be previously prepared such that the calumn oven and sample loop are equilibrated at 130C and 100 respectively. The hydrogen flame should be ignited at least one hour prior to analysis using constant and previously optimized flow rates for both hydrogen and air. Carrier flow at 30 ml/minute should also be established at least one hour prior to analysis. (Leave instrument on overnight during routine use.) Observe the baseline to determine both noise level and drift at the 1 x 10^ sen sitivity are acceptable. 2 With the recorder on and an attenuator setting of 1 x 10 , purge the sample loop with the sample gas at least 30 seconds at 100 ml/minute. Stop the pump, and when the rotameter just indicates no flow, activate the sample valve to inject the sample gas. Mark the injection point on the chart and read the loop temperature and laboratory pressure and temperature. Record all conditions on the chart. Start the SP4000. With the laboratory pressure and temperature, determine the water content of the sample at analysis from a water saturation vapor pressure table (assume 100% relative humidity in the sample). d) Instrument Calibration % Fill evacuated Calibrated Instruments bags with the standard gases, always using the sample bag for the same concentration, daily. Use only stainless steel or the appropriate Teflon tubing connectors. Allow bags to equilibrate to laboratory temperature. Measure each standard gas in duplicate each day using the same GC conditions and procedures described in the above analysis section. Record all conditions as described earlier for each analysis. sftL. ooono1 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN, MISSISSIPPI CONFIDENTIAL CHAIN OF CUSTODY Slurry samples and gas bags were handled as follows: Slurry Samples Sample Run No. Taken By Transported By Analyzed By 1 Wayne Myers Wayne Myers Jimmy Grace 2 Wayne Myers Wayne Myers Jimmy Grace 3 Wayne Myers Wayne Myers Jimmy Grace Reactor Opening Loss Gas Bags Sample Run No. Taken By Transported By Analyzed By 1 Wayne Myers Wayne Myers Jimmy Grace 2 Wayne Myers Wayne Myers Jimmy Grace 3 Waybe Myers Wayne Myers Jimmy Grace SAL 000110131 EMISSIONS TEST NEW REACTOR D-700 CONOCO CHEMICALS COMPANY ABERDEEN. MISSISSIPPI CONFJDfNTMt VI. APPENDIX The Appendix of this report contains the raw data and calculations used to obtain the sampling results. This information is presented in the following order: Operating data sheets for reactor D-700 during the three batch sampling period. Sampling data she'ets. Reactor opening loss and slurry residual analytical calculations Integrator printout data. Standard gas certifications. Reactor opening loss calculations. SAL 000110132