Document gaa7Oo89o02p3NQKmJjaV5j0G

EOR A GEON COMPANY TECHNICAL DOCUMENT INTERNAL GEON USE ONLY AMBIENT VINYL CHLORIDE EXPOSURES FROM RESIDUAL VINYL CHLORIDE IN BVC RESINS AND COMBOUNDS by Sam Butler Chuck Daniels Walt Edwards Jim Griffin Mike Mele Alan Olson Kirsten Reading Jim Summers June 3, 1993 DISTRIBUTION Lou Maresca Drucilla Knutsen Ashok Shah Mike Marshall woody Ban Authors ALTC CTF-2 C/1 w OD oon o / / / r TABLE OF CONTENTS AMBIENT VINYL CHLORIDE EXPOSURES PVC RESINS AND COMPOUNDS ABSTRACT FROM RESIDUAL VINYL CHLORIDE IN Rage 1 INTRODUCTION 9 EXPERIMENTAL 9 POWDER MIXING AND EXTRUSION IN THEALTC LAB. 9 PLASTISOL MIXING AND FUSION IN THEALTC LAB, 10 COMMERCIAL COMPOUNDING ON ALGC Z-LINE. 10 BULK RAIL MONITORING 11 COMMERCIAL WAREHOUSE MONITORING ; 11 ANALYTICAL METHODS 11 DISCUSSION 12 CONTROL OF AVCM EXPOSURE 12 MODELS FOR AVCM EXPOSURE 12 POWDER MIXING PROCESS 13 EXTRUSION PROCESS 14 PLASTISOL FUSION IN THE LAB. 14 BULK CONTAINER OPENING AND SHORTTERM EXPOSURE 15 COMMERCIAL COMPOUNDING 16 COMMERCIAL WAREHOUSE 19 MATERIAL BALANCE 19 MANUFACTURING CONCERNS ANALYTICAL RELIABILITY FALSE READINGS LABELING PRACTICES IN THE U.S. REFERENCES APPENDIX 20 >0 cn 26 10 O) 27 00 O 28 o to 29 30 ABSTRACT Previously, 8.5 ppm (weight basis) residual vinyl chloride (RVCM) in resins and compounds was set as the upper limit to assure that workers not contact more than 0.5 ppm 8 hour time weighted average (volume basis) ambient vinyl chloride (AVCM) in breathing air. The 8.5 ppm RVCM was based on models which assumed instant dispersion of VCM from the source into the whole room. One model assumed a material balance on measured RVCM would account for the VCM in the air. It is the conclusion of this study that the assumptions were in error. AVCM levels are not assured by RVCM alone, but are the joint responsibility of the supplier to control RVCM and the user to adequately ventilate. The previous practice of controlling RVCM to <8.5 ppm with a ventilation rate of 5 air turn-overs per hour, does, result in an AVCM level <0.5 ppm. AVCM concentrations are higher in plumes than in the surrounding air. Thus the VCM is not uniformly distributed in the room. Material balances showed that AVCM levels should have been significantly higher than were detected in the sealed lab with controlled ventilation. More than 90% of the RVCM that was lost from the resin during powder compounding had disappeared. One explanation is that the VCM reacted with residual catalyst in the resin, or stabilizers, or air, or another ingredients in the compound, to form a chemical not detected as VCM in the compound or the air. Our new data indicate <0.5 AVCM is adequately met with existing commercial resins and compounds for processes including lab processing, commercial compounding, commercial processing, and warehousing. One exception was a warehouse situation where resins are normally stored at >8.5 ppm RVCM and action is required to reduce AVCM. Laboratory powder mixing and extrusion in a room where the air turn-over was turned down to 1/hr, showed AVCM at 0.01 ppm. Monitoring the oven lab during the fusion of plastisols gave widely varying AVCM results. Most Reiszner badge results were at or below the detectable limit. However, duplicate badge samples on the lab operator were 0.24 ppm AVCM, higher than the material balance predicted values, indicating the operator had been in a plume. The AVCM level above a rail-car manway was tested just after opening and found to be non-detectable. The AVCM level inside the (i) C/1 *0 3 C3 car did exceed the 15 minute exposure ceiling level of 5 ppm. This indicates a need to warn about entering closed shipping containers before adequate ventilation. Commercial compounding process monitoring showed 38 samples <0.5 AVCM and 1 sample >0,5 AVCM- The high reading, 1.08 ppm AVCM, was on a packager operating a fork-lift and is suspected to be high because of hydrocarbon interference. Commercial warehouses in Henry, and ALGC showed <0.5 ppm AVCM, however Pedricktown, which ages dispersion resins containing above 8.5 ppm RVCM, showed readings >0.5 ppm AVCM, situation being corrected. a Air monitored by pumped charcoal tubes (OSHA acceptable) and by Reiszner badges was analyzed by headspace-gas chromatography. Residual vinyl chloride in resin, compound, and product was . measured by direct injection- or headspace-gas chromatography. These methods have been shown to be accurate and precise under laboratory conditions. False negative readings for both air and PVC measurements would be unlikely. However, false positives could occur since other chemicals could elute with vinyl chloride in the gas chromatography analyses. Most competitors label their resins with a VCM warning. A summary of the experimental results is listed in the following tables. (2) r <.v 0D & LABORATORY OPERATIONS 'SAMPLING METHOD NIOSH & Reiszner badge TYPE MONITOR area DURATION hours 2.5 Reiszner badge NIOSH & Reiszner badge Reiszner badge personal area U area 2.5 0.25 0.25 2.5 personal 2.5 TEST DATE 1/93 1/93 1/93 1/93 1/93 1/93 ENVIRONMENTAL V CONDITIONS 1.0 air turn overs/hour, no point source ventilation, 55 lbs/hr tt In the sealed resin drum tt 1.0 air turnovers/hr, no point Bource ventilation, 2.9 lbs/hr RVCM OPERATION ppm ' 103c1) powder mixing H storage 1.15 .. 45(1) storage sheet extrusion HH - area 0.25 1/93 In sealed powder mix " drum ` area 3 1/93 1.0 air turn overs/hour, no point source ventilation 102(1> personal 3 1/93 fl (i) rvcm was purposefully higher than normal in this study. storage plastisol fusion AVCM, ppm 8 samples <0.01 to 0.01 2 samples <0.01 0.55 1.5 8 samples <0.01 2 samples <0.01 to 0.01 <0.01 8 samples <0.01 to 0.04 0.24 i)Q3QZSZ COMMERCIAL WAREHOUSES SAMPLING METHOD Reiszner badge " TYPE MONITOR 2 area & 2 personal area DURATION hours ; 8 8 TEST DATE 2/93 2/93 ENVIRONMENTAL CONDITIONS; - - M8 2/93 - 13 area & 2 personal 8 8 2/93 - 2/93 RVCM-!; ppm some >10<2> OPERATION Henry Dispersion resin aging Avon Lake Bldg. 438 Warehouse Avon Lake Bldg. 452 - Avon Lake Bldg. 421 some. >iot2) Pedricktown Warehouse and dispersion resin aging (2) Aging to remove RVCM so later users `do not need to monitor.. (3) Rectification is underway to reduce AVCM levels. AVCM, ppm 4 samples 0.11 to 0.14 6 samples 0.02 to 0.11 4 samples 0.02 to 0.11 15 samples <0.25 15 samples <0.036 to 1.0 with 10 samples <0.5 & 5 samples >0.5<3> COMMERCIAL BULK SHIPPING SAMPLING METHOD hv meter TYPE MONITOR; area DURATION hours instant - t TEST ENVIRONMENTAL DATE;: CONDITIONS 92 Inside an unvented bulk rail car with lid just opened; operator exposure only if he sticks his head in the car(^. 92 " RVCM ppm 1.4; 1.8 " 92 1.3 . " < 92 3.1 it 92 - 92 " 4) AVCM outside t ie railcar manhole-was not detectable. 0.03 3.0 OPERATION unloading i M ft " AVCM, ppm 2 readings 1.5 to 2.5 2 readings 1.5 to 2.0 2 readings 1.0 to 3.5 2 readings 3.5 to 4.0 0.0 2 readings 9 to 15 U1 0038222 COMMERCIAL COMPOUNDING OPERATION SAMPLING METHOD NIOSH & Reiszner badge TYPE MONITOR area DURATION hours : 24 ; test DATE 1/93 Reiszner badge " personal 24 24 1/93 1/93 ENVIRONMENTAL CONDITIONS ALGC Bldg. 441, Buss and FCM operation 1.3 air turn-over/hr, point source ventilation. H RVCM ppm : 8 ppm note*1* " . " H 24 1/93 N " n 24 1/93 ft tl n (Tj RVCM was purposefully higher than normal for this study. (5) Suspected hydrocarbon interference to the VCM measurement'. OPERATION ALGC Buss Rigid PVC compounding " - weighman " - powder mixer operator " - operator w - packager, fork lift operator AVCM, PPm 9 samples <0.01 to 0.07 0.04 0.05 0.04 1.08 note^ 0\ QQOSQZZZ COMMERCIAL COMPOUNDING OPERATION SAMPLING METHOD Reiezner badge TYPE MONITOR personal DURATION. TEST hours DATE a 1/93 8 1/93 8 1/93 "8 8 1/93 1/93 ENVIRONMENTAL CONDITIONS Louisville compounding Henry compounding on FCM mill Henry compounding on FCM & mill Henry compounding on FCM & mill Pedricktown compounding on FCM & mill RVCM PPm OPERATION 0.03 ' PVC compounding t 'stock change .cleaner, control rgom, supervisor, .'rover, packager 2.9 ;rigid PVC compounding, powder mixer, packager, operator 0.38 3.6 . ^rigid PVC compou nding, operator tt 2.5 PVC compounding, operator, weighman, powder mixer AVCM, ppm 10 samples <0.05 to 0.28 3 samples 0.02 to 0.05 2 samples <0.01 to 0.14 0.16 10 samples <0.01 to 0.02 600382SZ COMMERCIAL VINYL PROCESSING SAMPLING METHOD OSH TYPE MONITOR personal DURATION hours o.s TEST DATE 1975 ENVIRONMENTAL CONDITIONS trailer just opened. ' 2 1975 1 to 4 1975 1 to 3 1975 " * * 1.5 197S - m - 1.2 197S jut opening trailer. m " , I to 2 1975 - - - 0.25 1975 just opened trailer. 3 1975 - 0.25 1975 just opened railcar. - "1 1975 - " 1&5 1975 " "3 1975 - " 1 1975 ** 0.5 to 4 1975 RVCM ppm <1.0 <1.0 <1.0 1.0 <1.0 35 <1.0 <1.0 <1.0 0.55 0.55 0.05 <1.0 <1.0 <1.0 OPERATION unloading flexible pellet compound from trailer flexible profile extrusion, operator, packager rigid cellular powder extrusion, operator, material handler, mixer flexible pellet extrusion, operator, lift driver flexible powder extrusion, mixer operator opening C saroplinq truck . flexible powderextrusion, powder mixer, operator. rigid pellets, unloading rigid pellet extrusion, lift operator, operator opening resin railcar rigid powder extrusion, mixer operator rigid pellet extrusion, lift operator, operator flexible pellet extrusion, operator flexible pellet extrusion, operator, material handler flexible pellet extrusion, lift operator, operator AVCM, ppm 2 samples <0.01 4 samples .<0.01 3 samples <0.01 4 samples <0.01 2 samples <0.01 0.02 4 samples <0.02 to 0.40 <0.01 | I | 3 samples <0.01 0.46 2 samples <0.01 to 0.23 2 samples <0.01 N H 1 | 2 samples <0.03 to <0.04 4 samples 0.06 to 0.31 3 samples 0.04 to 0.07 (8) r CA & ct c INTRODUCTION A statistical model for ambient vinyl chloride monomer (AVCM) in warehouse situations correlates this AVCM to residual vinyl chloride monomer (RVCM) content of the stored suspension PVC resins This model indicates that an RVCM less than 8.5 ppm is sufficient to assure an AVCM of less than 0.5 ppm (the OSHA action level for vinyl chloride). Critical assumptions are that the AVCM is uniformly and instantly distributed throughout the whole building and that the air turn-over is 0.6 times per hour. A physical model for AVCM in PVC processing buildings calculates that at an RVCM less than 15 ppm, the AVCM would be less than 0.5 ppm<3>. This model assumes that the AVCM is uniformly and instantly distributed throughout the whole building and calculates that the air turn-over is at least five times per hour. It also assumes all VCM lost from the resin goes to the room air. Neither of the above papers address dispersion resin storage nor processing. Investigation of the variations in AVCM within a building or room shows that VCM is highly variable and is concentrated in plumes coming from the point source <4>. Measurements of AVCM in various commercial processes were made in the mid-1970s, however, few measurements were made where both RVCM and AVCM were known. When RVCM was below 1 ppm, personal monitored AVCM was always below 0.5 ppm(s>. EXPERIMENTAL POWDER MIXING AND EXTRUSION IN THE ALTC LAB. Since the initial experiments run with plant resins yielded no detectable AVCM, a second trial was run with pilot plant produced resin with very high levels of RVCM. Suspension - Pilot plant resins typically have high porosities, which means that they strip easily and end up with low RVCM levels. Geon 110X334 is a low I.V., low porosity resin. The pilot plant recipe and stripping conditions were adjusted to produce a resin with high RVCM. Results are listed below. 3 LQX 193--637--088 193-637-090 193-637-092 193-637-128 Average RVCM rppffil 163 30 7 -208 103 weight 30 28 27 27 (lbl o (9) T A 6021 ft3 room was ventilated at 1.0 air turn-over/hr. The exhaust was in the center of the room and the powder mixing in the corner. Geon 110X334 resin, specially made at 103 ppm RVCM and well above the normal production level, was Henshel mixed in a generic rigid PVC compound that contained 85% resin. The compound was mixed to 160F, the lid opened and filler added, then continued mixing to 180F. This compound was dropped to a cooler, cooled to 130F, then dropped to a drum. The rate was 55 lbs/hr. Monitoring was done with both Reiszner Badges and carbon tube and pumps. Placement of monitoring was in several positions in order to try to find the plumes. An 8928 ft3 room was ventilated at 1.0 air turn-over/hr. The exhaust was in the center of the room and the Brabender extrusion in the corner. The above generic compound at 45 ppm RVCM, was extruded at 2.9 lbs/hr at a melt temperature of 400^- Monitoring was done with both Reiszner Badges and carbon tube and pumps. Placement of monitoring was in several.positions in order to try to find the plumes. PLASTISOL MIXING. AND FUSION IN THE Plastisol #1 contained 67 PPM VCM. Plastisol #2 contained 137 PPM VCM. Plastisol Recipe - PHR Geon* 136, Lot P29N012 Geon* 217, Lot H2B0708 DOP ESO Synpron 940 11 57 3 2 IZ 9998 12 57 3 2 Fusion Conditions: Forced draft Blue-M oven - 375 F for 5 minutes lOOg plastisol in 9ttX13" aluminum cake pans - 3 pans in oven at one time Fused samples cut from slab and placed hot in a 4 oz. then sealed with electricians tape. Testing was done later. The fusion lab was exhausted at 1 air turn-over/hour. jar 3 days COMMERCIAL COMPOUNDING OH ALGC 2-LINE. The Geon Company commercial resin with the highest typical RVCM content is 110x377. This resin is currently produced at ALGCP. The polymerization is run at 82c, which is close to the glass transition temperature. The resin has very low porosity and is very difficult to strip. In addition, this resin is used frequently on Z-line. CA <X 8 to (10) For the plant monitoring trial conducted in ALGCP, the resin plant made two lots of 110x377 with RVCM in the range of 7-10 ppm. The resin sat in a silo for several days before the monitoring test was run. RVCM data for the resin as used is listed below. Date Time RVCM (ppm) 1/26/93 18:00 4.75 24:00 3.87 1/27/93 08:00 2.36 BULK RAIL MONITORING Several rail-cars of resin were allowed to sit in the sun at Deer Park for several days. They were opened and immediately monitored with an h-nu meter for instantaneous levels of AVCM. The AVCM reading was taken in the center of the manway opening, just above the resin, and two feet above the manway. Weather conditions the day of the test were surishirte, 70F, slight- . breeze. COMMERCIAL WAREHOUSE MONITORING Five Warehouses were selected for this study. Forty area and four personal monitorings were conducted, which was based on facility size and resin storage quantities. Warehouse storage included both suspensions and dispersion resins. ANALYTICAL 1 s. Charcoal from badges and tubes was analyzed at Brecksville Research Center using a headspace-gas chromatography method <ft). This is the same method used for routine determination of AVCM in plants and warehouses. For residual vinyl chloride analysis, resin was taken at the start of the Henschel mixing or plastisol preparation. Compound was taken as it came from the Henschel and, again, before extrusion; plastisol was sampled before curing. Sheet samples were taken at the exit of the extruder, and plastisol product was taken as it came from the oven. All samples were stored in completely filled and sealed glass jars. Most were analyzed within eight hours of being taken; if not, they were stored at 36C until analyzed. Analysis was by gas chromatography using either a direct injection method m at ALTC for RVCM levels above 1 mg/kg (ppm) or a headspace method (Standard Test Procedure 1005H) at ALGC for levels below 1 mg/kg. Results for samples analyzed by both methods were comparable. vw* a <2 O (ID -r DISCUSSION CONTROL OF AVCM EXPOSURE There are many factors affecting AVCM exposure. 1) The amount of VCM supplied to the process or the RVCM of the resin or compound. 2) The general ventilation or air turn-overs per hour, which removes AVCM. 3) Local point source ventilation, which removes the VCM before it can come into the room. 4) The rate of resin or compound processed. 5) The percentage of VCM lost from the resin or compound during processing (Many previous measurements indicate that about 60% of the RVCM is lost during processing (3), however lost VCM could be lost to the air or could be chemically reacted to non-VCM products.) 6) Worker nearness to the VCM source or plume. 7) Air volume in a room for the dilution scenario where the room air is well mixed. Of the above factors, the RVCM is controlled by the Geon Company on resins and compound, but several other factors are. controlled by the resin or compound processor. Thus it is a joint responsibility of The Geon Company and the user to control AVCM exposure MODELS FOR AVCM EXPOSURE There are models for calculating AVCM exposure. One model uses statistically fitted data for warehouse exposure (1>, however this model is only for suspension resins in a well mixed warehouse and is not applicable to melt processing. Uniformly mixed room models <3'4> were considered, however, most processes are not well mixed rooms and concentrated VCM in plumes often exist<4>. It was therefore decided to measure AVCM exposures under well controlled conditions to find the relationship of AVCM to RVCM in the resin or compound. Some of the assumptions in the new model are: AVCM ______L&YQt) 1. (ventilation turn-over/hour) These above assumptions are the same for the old t3) and new models. Other assumptions which differ from the old model are: C,-7 GD CO O (12) OLD vs HEW MODEL COMPARISONS VENTILATION NEW MODEL 1 air turn-over/hour AVCM DILUTION EXPOSURE as measured find AVCM concentrations in a plume highest measurement in a plume OLD MODEL 5 air turn-overs/hour from a material balance instant distribution of AVCM into the whole room average calculated for 1 the room | Scale-up of AVCM to manufacturing scale involves RVCM in the resin or compound and the rate in lbs/hr. Worst case ventilation is assumed to be l air turn-over per hour. We find that air in a powder mixing operation or extrusion operation is quite stuffy at l/hoiir, a piastisol fusion operation is quite Smoggy, therefore 1/hour is expected to be the worst case acceptable. It is important to make several AVCM measurements so as to be able to find concentrations in plumes. POWDER MIXING PROCESS AVCM results are listed in Table 1 for a powder mixing operation. The highest readings are 0.01 ppm above the mixer and 3 feet from the mixer. Since we started with an uncontaminated room, these measurements may be 60% of equilibrium values and were corrected for scale-up (see Figure 1.) A summary of results and scale-up follows: | AIR turn-over/hour RATE, lbs/hour RVCM, ppm MEASURED AVCM, ppm CALCULATED AVCM, ppm 1 55 103 0.01 1 18,000 fl | for a 1500 lb Henshel on a 5 minute cycle 1, 3, 8.5, 9 0.05, 0.2, 0.45, 0.5 This indicates that RVCM in resin could be as high as 9 ppm before the AVCM would be expected to reach 0.5 ppm in a commercial powder mixing process. EXTRUSION PROCESS AVCM results are listed in Table 1 for a sheet extrusion process. The highest readings are 0.01 ppm on the personal monitor of Juan Rodriguez. Since we started with an uncontaminated room an equilibrium consideration was made as was made for powder mixing. A summary of results and scale-up follows: | EXTRUSION AIR turn-over/hour RATE, lbs/hour | Compound RVCM, ppm MEASURED AVCM, ppm CALCULATED AVCM, ppm LAS VALUES 1 2.9 45 Q. 01 :;ScS^D* 1 1500 for a siding extrusion 1, 2.6, 3, 8.5 0.2, 0.5, 0.6, 1.6 1 | ( I 1 B This indicates that RVCM in compound could be as high as about 3 ppm before the AVCM would be expected to reach 0.5 ppm in a commercial extrusion process. This also corresponds to a starting resin RVCM of 6 ppm. PLASTISOL FUSIONUEN THE LAB. Production samples of Geon 217 blending resin and Geon 136 dispersion were chosen to make the plastisol. In order to have sufficient quantities of VCM present in the plastisol to produce measurable AVCM levels, an artificially high RVCM Geon 217 sample was prepared. The 217 had 12,450 PPM of vinyl chloride monomer. Two plastisols were made, the first contained 67 PPM RVCM and the second 137 PPM RVCM. The plastisols were fused in aluminum cake pans at 375 F for 5 minutes in a standard Blue-M oven. Each pan contained lOOg plastisol. 1900g of plastisol was fused each hour. One oven was used with about 3 pans being fused at a time. Plastisol #1 67 PPM RVCM was fused during the first hour and plastisol #2 137 PPM RVCM during the second hour. Duplicate Reiszner static badges were placed in locations near the oven door, in the oven exhaust hood, on the operator, on the floor and in the room corners. The exhaust hood was dampened such that there was only 1 air change per hour in the room. The lab had a volume of approximately 6300 cubic feet. (14) The fused slabs from plastisol #1 were tested at 30 PPM rvcm and the #2 slabs at 79 PPM- The area was monitored during the 2 hours of actual fusing of plastisol, plus 30 minutes after fusion was completed. Based on air flow rates and plastisol RVCM vs slab RVCM, and assuming that all the VCM went into the room, then the average AVCM reading for the 2.5 hour period should have been 0.084 PPM on a volume basis by material balance. The duplicate lab operator samples measured 0.20 PPM. The monitor on the floor measured 0.04 PPM. All the other monitors tested at 0.01 PPM or less. The H-NU meter registered 0 to 5 PPM at various locations in the room. High readings were noted near the stack of fused slabs. This lab data was scaled to a rotocaster at 130 lbs/hr and is summarized in the following table. | AIR turn-over/hour RATE, lbs/hour Compound RVCM, ppm MEASURED AVCM, ppm CALCULATED AVCM, ppm LAB VALUES 1 2.6 102 0.24 i 130 for a rotocaster 1, 2.6, 3, 0.2, 0.5, % o | 8.5 j 1.6 I BULK--CONTAINER OPENING AND SHORT TERM EXPOSURE One area of concern for AVCM exposure was the opening of PVC rail-car manways. RVCM in the resin could be released as the rail-car sits in the sun, resulting in a "blast" of air high in AVCM reaching an operator as he opens a manway on the car. Several rail-cars of resin were allowed to sit in the sun at Deer Park for several days. They were opened and immediately monitored with an h-nu meter for instantaneous levels of AVCM. The AVCM reading was taken in the center of the manway opening, just above the resin, and two feet above the manway. Weather conditions the day of the test were sunshine, 70F, slight breeze. Readings at the manway level and just above the resin were the same. No RVCM was detected two feet above the manway. Results are listed on the next page. C/T 03 03 O (15) Certificate of Analysis Lot# RVCM (ppm) BFGX1339 1.42 ACFX55986 1.81 BFGX1582 1.27 BFGX1385 3.13 BFGX1228 (xsoo) BFGX1018 0.03 3.01 RVCM reading,ppm Compartment at manway B 2.5 BC 1.5 B 1.5 BC 2.0 A 1.0 AC 3.5 A 4.0 AC 3-5 B0 BC 0 B BC 9 15 2' above 0 0 0 0 0 0 0 0 0 0 0 0 AVCM levels at any point above the manway opening on a railcar were consistently zero. As long as operators do not place their heads inside the manway, they should not be exposed to any measurable level of AVCM. If someone puts his head inside the rail-car, he could be exposed to an AVCM level above the 5ppm ceiling for 15 minute exposure. This would be a violation of the OSHA confined space entry rule, so this is not something that should ever occur. POMMJSRCIAL.-gQltPQUlfDING Building 441, ALGC Compound East was selected as the test location. This PVC compounding facility has two independent production lines consisting of a Farrel Continuous Mixing (FCM) operation and a Buss Extrusion line. The exposure area around the fcm was chosen for this study. The air exchange rate in B/441 was approximately 1.3/hr. Major ventilation equipment included: mill hood exhaust, RS to GS screw, GS outlet, pelletizer and strip tank exhaust. The building size is approximately 38,000 FT2 on floor level one. The sample duration selected for this study was 24 hours. During this time, 87241 Black and 87371 was produced using 110x377 resins stored in silo 202. Certificate of analysis RVCM was 8ppm for the PVC resin. An RVCM was also run on a silo sample prior to compounding. That result was 4ppm. C/i (16) i r A test area audit was conducted prior to the study. Based on projected emissions source points, (16) area locations and (4) personnel monitorings were selected. The two sampling methods were as follows: 1) Area monitoring using a Dupont model 2500 pump with SKC 6x70 or 10x110 charcoal tubes. Established flow rates of 45cc/rainutes for eight hour and 500 cc/minute for twenty- four hour monitorings. 2) Area and personnel monitoring using a Reiszner "MiniMonitor" passive docimeter badge. Device weight of 35 plus 1.35 g of unexposed activated charcoal granules. grams algo Plant Monitoring Buss Line LOCATION # OF SAMPLES RESULT (PPM) 1. Buss Line 1st Floor 9 <.01 - 0.006 2. Weighman 1 .04 ( 3. Henschel Operator 1 .05 4. Line Operator 1 -04 5. Packager 1 1.08 6. Mill Area 3 .05 7. Z Line Henschel 1 <.01 8. Henschel Feed Hopper 1 .03 9. Pellet Cooling 1 .01 10. 4th Floor Resin Station 1 <.06 Also monitored was Louisville, Pedricktown mill line, and Henry mill line. These results are summarized in the following table. U7) CV7 V>.vs' "V r r COMMERCIAL COMPOUNDING OPERATION SAMPLING METHOD Reiezner badge TYPE MONITOR personal DURATION . . TEST hours ; DATE 8 1/93 ENVIRONMENTAL" CONDITIONS Louisville compounding RVCM ppm 0.03 8 1/93 Henry compounding on 2.9 FCM & mill " "8 1/93 Henry compounding on 0.38 FCM & mill - -8 1/93 Henry compounding on 3.6 FCM 6 mill 8 1/93 Pedricktown 2.5 compounding on FCM & mill .OPERATION .PVC compounding: stock change cleaner, control room, supervisor, rover, packager rigid PVC compounding, powder mixer, packager, operator rigid PVC compounding, operator M PVC compounding, operator, weighman, powder mixer AVCM, ppm- 10 samples <0.05 to 0.28 3 samples 0.02 to 0.05 2 samples <0.01 to 0.14 0.16 10 samples <0.01 to 0.02 ( 18) sz COMMERCIAL WAREHOUSE Five Warehouses were selected for this study. Forty area and four personal monitorings were conducted, which was based on facility size and resin storage quantities. Warehouse storage included both suspensions and dispersion resins. Please note that 5 results at Pedricktown were above the section limit for Vinyl Chloride. Appropriate plant follow-up is in progress as of this writing. Jse LOCATION 1. Henry B/735 2. Avon Lake B/438 3. Avon Lake B/452 # OF SAMPLES 4 6 4 RESULTS (PPM. 0.11 - 0.14. <.02 - .11 <.02 - .11 (of> f 1 V. 4. Avon Lake B/421 15 <'2S 5. Pedricktown Main Warehouse 15 <.036 - 1.0 5 results >.5 ppm* * Appropriate plant follow-up is in progress as of this writing. MATERIAL. .BALANCE A material balance was done for the low air exchange rate experiments run in ALTC using a CSTR (continuous stirred tank reactor) model. The model assumes instantaneous and complete mixing of any VCM released from the compounding process with the air in the room. The mathematical equations for powder compounding of the suspension resin and for curing of the plastisol are listed in the Appendix. For powder compounding, the material balance shows that there should have been a significant level of AVCM in the room. The average level throughout the room for the two hour test would have been about 1.7 ppm AVCM. Since we know that the well-mixed model does not accurately model compounding processes, we can apply the results of the calculations in a qualitative manner to the plume theory. Any badge near the Henschel mixer should have read an AVCM level significantly higher than 1.7 ppm AVCM. However, no reading near (19) <77 05 CO o tv 1 ppm was recorded* The RVCM that disappeared from the resin in the compounding process had to go somewhere. An explanation, based on the data we have, is that the RVCM reacts with residual resin catalyst, stabilizers or other ingredients to form something not detected as RVCM in the compound or AVCM in the air. The data seems to indicate that at least 90% of the RVCM simply reacts in this manner. The analysis of the material balance for dispersion Plastisol curing follows the same pattern. According to the well-mixed model, there should have been an average of about ppm AVCM throughout the room over the course of the 2.5 hour experiment. The actual results included one reading of 0.24 AVCM, which must have been in a plume. 0.16 ppm MANUFACTURING CONCERNS Tables 1 and 2 are a summary of the resin records used in the RVCM study to determine the effect that changes in RVCM specifications would have on current BFG production. The focus was on the poundage that would be effected by any specification changes. In some instances the actual poundage estimates for the given period analyzed were not available, however, the size of the sample for which poundage figures were available was large enough that is doubtful these missing poundage numbers would greatly influence the final estimates given below. Indeed, during the early phases of this analysis, with one quarter of the data presented in the final analysis available, the percentages that were being computed were quite close to the numbers shown below. The following tables can be summarized as follows: I. The current data analysis is based oh a TOTAL production of 1316 MM mass and suspension resin pounds. A. If the RVCM threshold limit is reduced to 5 ppm and manufacturing practices remain the same, the amount of product having levels above this value would increase by 3.5%. 1. Since bag-by-bag selection of material is not economical, any resin with production pounds in excess of 5 ppm would be affected. Thus the TOTAL poundage affected would represent 55.6% of the current poundage. In a similar fashion B. If the RVCM threshold limit is reduced to 3 ppm and manufacturing practices remain the same, the amount of product having levels above this value would increase by 7%. (20) 1. Since bag-by-bag selection of material is not economical, any resin with production pounds in excess of 3 ppm would be affected. Thus the TOTAL poundage affected would represent 64.5% of the current poundage. and C. If the RVCM threshold limit is reduced to 1 ppm and manufacturing practices remain the same, the amount of product having levels above this value would increase by 29%. 1. Since bag-by-bag selection of materials is not economical, any resin with production pounds in excess of 1 ppm would be affected. Thus the TOTAL poundage affected would represent 78% of the current poundage. (21) ce 01 w PLANT A LG C ALGC ALGC ALGC ALGC ALGC ALGC ALGC ALGC ALGC D PAR K D P ARK DPARK DPARK DPARK DPARK DPARK DPARK HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY LVILLE LVILLE LVILLE LVILLE LVILLE LVILLE LVILLE NIAGARA NIAGARA NIAGARA NIAGARA NIAGARA NIAGARA PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN RESIN 101X9 102EPF5 110X334 110X377 110X368 110X394 110X412 110X432 110X450 G-3400 102EPZ X426LD X4 300 X4 50 FY X 4 7 0W X 4 71W xsoora X500FG 12 3A 170X100 171 1 7 1B 173 178 1 78B 179 190X100 193 198 199 213 217 250X100 110X440 F-76TRD G- 2 6 D G-27D G-30D G-37D G 2 7 LIVD 1 0 0C13 110X334 110X380 110X429 GEON30N GEON3 0W 1 1 0X4 2 6G 1 1 0X4 3 3G 1 1 0X4 50 F 121 1 21A 1 2 1AR 121X10 121X10B RVCH SUMMARY FOR 5 AND 3 r'PM RVCM Records MEAN MEDIAN MIN Specs MAX {ppa) 12 385 460 588 20 8 43 59 121 44 131 126 639 175 157 4 206 88 63 1 171 19 199 344 50 101 1 1 70 63 177 63 16 19 21 13 73 35 71 27 82 149 852 49 145 325 623 76 447 295 13 0 164 358 23 155.08 0.03 1.36 2 .54 0.04 0.01 0.01 0.03 0.05 0.36 0.03 1.07 1 . 16 0.06 0.07 0.03 0. 11 0.05 5.14 0.48 1 .41 3.38 0.99 1. 55 4 .19 0.74 0.95 3.93 1.98 1.94 3.41 30.13 0.11 4.82 0.09 0.16 0 .68 0 .03 1 .61 0.18 1.52 3.53 0.68 0 .04 0.14 0 .09 0.36 0.16 0.06 3.19 0.86 0.39 0.58 0.46 0 . 20 0.02 0 . 59 1 .46 0 .01 0.01 0 .01 0 .01 0.04 0.05 0.02 0.70 0.56 0.03 0.05 0.02 0.05 0.02 4.60 0.48 0.66 2.21 0.58 0.66 2.38 0.46 0.35 3.93 1.78 1.58 1.95 8.0 0; 0.12 3 .75 0.07 0.02 0.20 0.03 1.16 0.15 1.19 1.28 0.68 0.00 0.12 0.00 0.17 0.09 0.05 1.21 0.43 0 . 30 0 . 26 0 .30 0.01 0.00 0.01 0.06 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.01 0.00 0.30 0.48 0.00 0.33 0.00 0.00 0.08 0.00 0.95 3.93 0.05 0 .05 0 .00 0 .00 0.07 0 .40 0.00 0.00 0 .00 0 .00 0.00 0.00 0.16 0.00 0.00 0 .00 0.00 0 .00 0 .02 0 .02 0.01 0.00 0 .01 0 .00 0 .01 0 .01 1857.00 2.73 47.41 172.55 0.62 0.01 0.10 1.19 0.52 4.64 0.20 6.47 35.70 1.79 0.51 0.06 10.62 0.54 11.40 0.48 8.50 11.55 7.58 6.50 6 0.79 6.76 0.95 3.93 6.13 6.50 21.23 480.46 0.17 14.14 0.40 1.62 4.31 0.16 7.73 0.50 9.76 46.74 13.29 0.45 1.21 2 . 70 7.81 1.23 0.93 3 4.77 7.26 1.94 5.90 1 . 78 0-2 0-2 0-2 0-8.5 0-2 0-8 . 5 0-8.5 OA -- 4l 0-8.5 0- . 99 0-6.5 0-6.5 0-8.5 0-8.5 0-8.5 0-10 0-2 0-6.5 0-8.5 0-8 . S 0-6.5 0-6.5 0-8 . 5 0-1 Total Lbs (millions) 1.42 32.69 43.50 46.16 2.08 0.79 3 .94 5.45 11.27 2.14 19.45 12.59 276.53 22.76 22.72 0.52 30.76 11.42 2.58 6.68 0.18 15.07 1 .60 4.41 2.94 2.72 6 .53 1.92 2.80 116.00 10.00 120.00 38.00 23 .00 48.00 19.02 23.69 117.00 19.70 76 .23 14 .17 4 .64 6.11 10.28 . Est.Total Lbs (ai11ions } > 5ppa 0.12 0.00 1 .70 2.43 0 . 00 0.00 0.00 0.00 0.00 0.00 0.00 0.20 6.92 0.00 0.00 0.00 0.15 0.00 1.19 0.00 0.39 0.00 0.96 0.40 0.13 0.00 0.00 0.17 0.17 1.33 1.40 0.00 1.16 0.00 0.00 0.00 0.00 1.62 0.00 0.46 3.02 0.96 0.00 0.00 0.00 0.00 0.00 2.69 0.07 0.00 0.03 0.00 Est.Total Lbs < ai11ion ) > 3ppa 0.12 0.00 3.69 5.97 0.00 0.00 0.00 0.00 0.00 0.05 0.00 0.90 15.58 0.00 0.00 0.00 0.15 0.00 1.84 0.00 0.90 0.01 2.45 0.72 0.17 0.00 0.59 0.43 2.56 1.74 0.00 1.62 0.00 0.00 9.86 0.00 2.92 0.00 1.39 5.09 2.06 0.00 0.00 0.00 o'. 00 0.00 4.23 0.26 0.00 0.1? 0.00 U z) i z\ymzsz CONT'RVCH SUMMARY FOR 5 AND 3 PPM RVCM P LANT PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN SCOT SCOT SCOT RES IN 12 3A 12 4 1 24A 1 2 5A 136 137 138 180X10 180X5 180X7 184 18 6A GEON3 0 GEON31 1267 1467 ga 6 Records 452 102 129 262 90 152 478 285 202 38 21 48 405 140 1073 141 72 MEAN MEDIAN 6.86 6.86 3.25 6.19 5.03 0.35 6 .17 3.75 2.51 2.42 8.64 4.85 0.08 0.43 15.19 3.96 1 .63 2.84 4.36 2 .46 3.07 4.36 0.33 2 .86 2 .22 1 .00 2 .09 3.41 2.49 0.05 0 .05 1.00 2.80 1 .60 MIN Spies MAX ( ppa ) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.07 0.25 0.20 458 . 00 52 .06 1 4 .49 283 .70 17 .88 1 .35 373 .00 92 .00 20 .68 14 .02 3 3 .30 4 2 .97 0 .98 50 .17 1254 .37 72 . a 3 5 . 44 0-8.5 0-8.5 Total Lbs (nillions) 6.64 2.40 7.22 3.41 3.23 15.91 14.91 7.95 1.12 0.73 1..36 Est.Total Lbs (nillions) > 5pp 2.31 0.63 2.73 1.63 0.00 4 .76 4.03 1.34 0.15 0.31 0.43 0.00 Est .Total (Billions > 3pp 3 . 28 0.99 3.67 2.08 0.00 7.72 6.43 2.09 0 . 27 0.42 0.62 0.00 *. zzssmzzz (*2 ) 1 KVCM SUMMARY FOK 1 PPM-------ZS PLANT RESIN ALG C ALGO ALGC ALGC ALGC ALGC ALGC ALGC ALGC ALGC D PARK DPARK DPARK PARK DPARK DPARK DPARK DPARK HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY HENRY LVILLE LVILLE LV I LLK LVILLE LVILLE LVILLE l.VI LLE NIAGARA NIAGARA NIAGARA NIAGARA NIAGARA NIAGARA PTOWN I'TOWN PTOWN PTOWN PTOWN PTOWN PTOWN I'TOWN 101X9 102EPF5 110X334 110X377 110X388 110X394 110X412 110X432 110X450 G-3 4 0 0 102EP2 X426LD X4 300 X4 50FY X470W X471W X 500 F B X500FG 1 23A 170X100 171 1 7 1B 173 178 1788 179 190X100 193 198 199 213 217 250X100 110X440 F-76TRD G-26D G-27D G-- 3 0 D G-37D G27LIVD 1O0C13 110X334 110X380 110X429 G EON 30 N G EON 3 0W 1 1 OX 4 2 6G 110X433G 1 1 OX 4 50 F 12 1 1 21A 1 2 1AR 121X10 i 2 ixi on J38Z2Z R eco r 12 365 460 588 20 8 43 59 121 44 131 126 639 175 157 4 206 88 63 1 171 19 199 34 4 50 101 1 1 70 63 177 63 16 19 21 13 73 35 71 27 02 14 9 852 49 145 325 623 76 447 295 130 164 3 58 23 MEAN 155.08 0 .03 1.36 2 .54 0.04 0 .01 0.01 0.03 0.05 0.36 0.03 1 .07 1.16 0.06 0.07 0 .03 0.11 0.05 5.14 0.48 1.41 3 .38 0.99 1 .55 4.19 0.74 0.95 3.93 1 .98 1 .94 3.41 30.13 0.11 4 .82 0.09 0.16 0.68 0.03 1.61 0.18 1.52 3.53 0.88 0.04 0.14 0.09 0.36 0.18 0.08 3.19 0.86 0.39 0.58 0.46 MEDIAN 0 . 20 0.02 0 .59 1.46 0 .01 0.01 0 .01 0.01 0.04 0 . 05 0.02 0.70 0.56 0.03 0.05 0.02 0.05 0.02 4 .60 0.48 0 .66 2.21 0.58 0.66 2.38 0.46 0.95 3.93 1 .78 1 .58 1.3 5 8 .'0 0 0.12 3.75 0.07 0 .02 0.20 0.03 1.16 0.15 1.19 1.28 0.66 0.00 0.12. 0 .00 0.17 0.09 0.05 1.21 0.43 0.30 0.26 0.30 MIN 0.01 0,00 0.01 0.06 0.00 0 .00 0 .00 0 .00 0.01 0.00 0 .00 0.01 0.00 0.00 0.00 0 . 00 0.01 0.00 0 . 30 0.48 0.00 0.33 0 .00 0.00 0 .08 0.00 0.95 3 .93 0 .05 0 .05 0 .00 0 .00 0.07 0 .40 0.00 0.00 0.00 0.00 0.00 0.00 0.16 0 . 00 0 . 00 0.00 0.00 0.00 0.02 0.02 0.01 0.00 0.01 0.00 0.01 0.01 MAX 1857.00 2.73 47.41 172.55 0.62 0 .01 0.10 1.19 0.52 4 .64 0 .20 6.47 35.70 1.79 0.51 0.06 10.62 0 .54 11.40 0 .48 8.50 11.55 7 .58 6.50 60.79 6.76 0.95 3 .93 6 .13 6 . 50 21.23 480.46 0.17 14.14 0.40 1.62 4 .31 0.16 7.73 0 . SO 9.76 48.74 13.29 0.45 1.21 2 .70 7,61 1.23 0.93 34.77 7.26 1 .94 5.90 1 .78 Spacs (ppm) 0-2 0-2 0-2 0-6.5 0-2 0-6.5 0-6 . 5 0-1 0-8 . 5 0-8 . 5 0-8 . 5 0-8 . 5 0-8.5 0-8 . 5 0-10 0-2 0-8.5 0-8 . 5 0-6 . 5 0-6 . 5 0-6 . 5 0-8 . 5 0-1 % Total RftCO rdJ > lppn 25.00 0 . 27 35.87 68.71 0.00 0.00 0.00 1.69 0.00 11.36 0.00 34.13 26.76 1.14 0.00 0.00 0.49 0.00 95.24 0.00 39.77 63.16 32.16 4 5.06 76.00 19.80 0.00 100.0 68.57 69.84 67,80 96.41 0.00 73.68 0.00 7.69 16.44 0.00 53 . 52 0.00 60.98 59.73 26.64 0.00 0.69 1.54 6.10 1.32 0.00 54 .92 29.23 6.71 19.83 13.04 Total Lbs (Billions) 1 .42 32 . 69 43 . 50 46.18 2.08 0 . 79 3.94 5.45 11.27 2.14 19.45 12.59 276.53 22.76 .2 2.72 0.52 30 . 76 11.42 2.56 6.68 0.18 15.07 1.60 4 .41 2.94 2.72 6.53 ' 1.92 2.80 116.00 10.00 120.00 36 .00 23.00 48 .00 19.02 23.69 117.00 . 19 .70 78.23 14.17 4.84 6.11 10.28 Est .Total (million > 1 ppm 0.36 0.09 15.60 3 1.73 Q . 00 0.00 0.00 0.09 0.00 0.24 0.00 4.30 74.00 0.26 0.00 0.00 0 . 1S 0.00 2.46 0.00 2.66 0.06 6.79 1.37 0.87 0.00 2.02 1.90 4.43 1.89 0.00 2.06 0.00 0 .77 19.73 0.00 12.31 0.00 11.60 14 . 1S 31.17 0.00 0.14 1.20 0.00 7.78 1.41 0.41 2.04 CONT'RVCM SUMMARY FOR 1 PPM RVCM PLANT PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN PTOWN SCOT SCOT SCOT RESIN 1 23A 124 1 2 4A 1 2 5A 126 13 7 13 8 180X10 1 8 OX S 180X7 184 1 86A GEON30 G EON 31 12 6 7 1467 G86 Ra co c ds 452 10 2 129 262 90 15 2 478 285 202 38 21 48 405 140 1073 141 72 KEAN 6.86 6.66 3 . 25 6.19 5.03 0.35 6 .17 3.75 2 .51 2.42 8.64 4.85 0.08 0.43 15.19 3 .96 1.63 KEDIAN 2.84 4.36 2.46 3 .07 4.36 0.33 2.86 2.22 1.00 2.09 3.41 2.49 O.OS 0.05 1.00 2.80 1.60 MIN 0.00 0.00 0 .00 0.00 0 .00 0.00 0.00 0.00 0.00 0 . 00 0 .00 0 . 00 0.01 0.01 0.07 0.25 0.20 MAX 458 .00 52.06 14.49 283.70 17.88 1.35 373 .00 92.00 20.68 14.02 33.30 42.97 0.98 50.17 1254.37 72 .83 5.44 Spacs (ppa) 0-8,5 0-8.5 % Total Racocds > lppn 73.23 65.69 64 .34 70.61 75.56 1.97 74.90 69.12 SO . 00 60.53 61.90 60.42 0.00 0.71 48 .37 90.78 84 .72 Total Lbs {Billions) 6 .. 64 2.40 7.22 3 .41 3.23 15.91 14 .91 7.95 1.12 0 .73 1.. 36 Est .Total Lb & [Billions ) > 1 p po 4.86 1 . 54 5 . 10 2 . 58 0.06 11.92 10.31 3.98 0.68 0.45 0.82 0.00 (25) S3 21 ANALYTICAL RELIABILITY Vinyl chloride in the air is absorbed by charcoal in pumped tubes or in Reiszner badges; pumped tubes are approved by OSHA for air sampling. The charcoal is then removed and analyzed by headspace-gas chromatography. The precision and accuracy of the combination of air sampling and GC analysis has been studied earlier "Recoveries of vinyl chloride by the Reiszner Badge compare favorably with results via carbon tube sampling. (Recovery=badge result/tube result expressed as per cent.) In addition, the precision of testing expressed by the coefficient of variation (standard deviation/average) looks good, relative to OSHA requirements." "OSHA requirements for combined precision and accuracy are easily met at all concentrations, despite the tendency for Reiszner results to be slightly higher at very low monomer levels. The following 95% confidence limits were determined for the Badge." less than 0.5 0.5 to 10 greater than 1.0 ppm ....... Allowed 7.8% less than to 36.4% greater than the "true" value 1.9% less than to 17.5% greater than "true" value 7.7% less than to 10.3% greater than "true" value Plus or minus 50% of | "true" value I Plus or minus 35% of | "true" value E Plus or minus 25% of "true" value -- These results were based on tubes and badges prepared in the laboratory under controlled conditions. Extensive testing under actual plant conditions showed much poorer precision for both badges and tubes (see reference 9 for complete details). However, this probably indicated an actual variability of analyte concentration in the air. Similar variability was seen in a study done by the University of Utah Research Institute <ll): "Comparative measurements of organic vapors with active (pumped charcoal adsorption tubes) and passive (diffusion) monitors showed good agreement in steady-state laboratory chamber tests but considerable discrepancies in field trails on painters, in which the monitors were worn on opposite labels under fluctuating conditions of vapor concentrations and air movement." "Continuous recordings by an I.R. gas analyzer of air samples drawn from the left and right lapels of a painter working under simulated field conditions showed large differences in (26) <7T OD % concentration corresponding to the differences found between monitors." it is important to note that this variability occurred with the pumped tubes, which are accepted by OSHA for air monitoring, as well as with the Reiszner badges. Area monitoring provided additional confirmation that variability of the personnel monitoring results was due to variability of the air and not to problems with the monitors. In these tests the monitors and air were stationary; for the personnel monitoring tests, they were in motion. Sixteen badges were attached to a rack and exposed in building 461 of the Avon Lake General Chemical Plant. This was done eleven times. The mean concentration ranged from 0.34 to 4.94 ppm and the coefficient of variation (relative standard deviation) ranged from 5.2 to 10.0%3 , which are similar to those for the laboratory tests. The relative standard deviation of Standard Test Procedure 1-005H for. residual vinyl..chloride monomer in PVG is 2:63% at 2.09 ppm and 4.16% at 62.66 ppm <ia>. The precision of the direct injection analysis for Vinyl chloride has not been determined but the relative standard deviation is expected to be less than 5% for levels above 10 ppm. FALSE .READINGS There has been a concern about false positive and false negative results for both air and PVC testing. False positives are possible for both analyses since gas chromatography can not identify components. Therefore, if one or more other components elute with, or instead of, vinyl chloride, the peak area and concentration will be artificially high. ' False negatives are less likely for analysis of PVC and the charcoal in badges and tubes. However, the tubes and badges may not accurately sample the air to which they are exposed. Results, discussed above, show this to be unlikely. (27) r LABELING PRACTICES IN THE U.S. The following table summarizes the practices of Geon competitors- Many label all resins. c smpany B jrden (E CP) C arbide GG mosa C xy \ rista .. PVC LABELING IN U.S. Labeling Policy All Blending & Susp. Resins are labeled. Signal label not used in Canada. No warnings in MSDS or labels because RVC dOpprri. Emulsion & Disp. resins labeled per "OSHA Std" ianguage. Susp. resin labeled on hopper cars on tag, so hatch is ventilated first. No labels on compounds. Bottle resin not labeled ("mother's milk"). All other resins labeled. Compounds not labeled. "Dust & Fines & VCM can be hazardous at processing temperatures" Not haz. mtl. under Sara III. Comments Lawyers overruled rest of company. Even bottle resin is labeled. Policy partially based on BFG's 8.5ppm limit Revising policy and. labels with help from industrial hygienist. May move to labeling disp. and susp. resin ( and maybe compound) in a very conservative fashion due to USEPA citations for VCM releases. Warning label language is not OSHA language. Policy driven by lawyers. Monitoring 2 years ago showed 5 ppm RVCM in resin gives 99% confidence of being below OSHA limits. Can talk to Oxy's people & inspect their data. Oxy monitors customer plants on demand and always finds AVCM ND. Label-all resin and compound. (28) Driven by responsible care and . liability issues. VCM Warning fa i OSH-A- Std. ........... ...... I^*.! 00 o REFERENCES 1. M.M. O'Mara, L.B. Crider, R.L. Bowles, C.J. Toraakek, " A Physical Model for the Diffusion of Vinyl Chloride Monomer from PVC under various Conditions of Storage", BFG Technical Document, August 25, 1975. 2. L.B. Crider, M.M. 0'Mara, R.L. Bowles, Society of Plastics Engineers, "Safety and Health with Plastics", National Technical Conference, page 196, Denver, November 8-10, 1977. 3. M.J. Ahmed, H.S. Haller, "Modeling of VCM Concentrations During Thermal Processing of PVC", BFG Technical Document, March 31, 1983. 4. A. G. Andreopoulos, A. N. Karayannis, N. C. Markatos, Trans IChemE, 7P Part B, 75 (May 1992). 5. R. Kriiszynski, "OSHA - NIOSH - APA, 2nd Quarter 1975 Update.", Inter-Organization Correspondence, July 8, 1975. (Appendix,). 6. Born, J.W.? "Method of Test for Vinyl, Chloride Monomer in Air Using Activated Charcoal Monitors and Headspace-Gas Chromatographic Analysis"; BFGTD; January 30, 1986. 7. Samples were dissolved in tetrahydrofuran (% gram in lOmL); luL of the solution was injected into a Hewlett Packard 5890A GC equipped with a flame ionization detector and a 10 foot by \ inch (2mm ID) glass column packed with 1% SP1000 on Carbopack B. A response factor was determined by standard addition of VCM to the PVC solutions. 8. F.M. Galloway, "Diffusion Modeling of VCM Concentrations in Air",BFG Inter-Organization Correspondence, Jan. 15, 1993. (appendix). 9. Born, J.W.; "Evaluation of the Reiszner MiniMonitor Charcoal Badge as a VCM Monitor - Part II"; Status Report? August 12, 1985. 10. Fairlie, A.M.; "Evaluation of the Reiszner MiniMonitor Charcoal Badge as a VCM Monitor - Validation"; BFGTR; November 18, 1985. 11. Division, University of Utah Research Institute, December, 1981. 12. Environmental Reporter, PP 120:1088 through 120:1091 October 7, 1988, EPA Method 107. i\j (29) r APPENDIX 1 iFGoodrich INTER-ORGANIZATION CORRESPONDENCE i J.Summers____________ FROM P . M . C,a 1 1 nvay_______ SUBJECT FIELD POINT OR OEPT 4 BLDG NO FIELD POINT OR 06PT 4 BLDG NO AT.TP PSr-N DATE YOUR LETTER OATE THIS LETTER ----------------------------- 1/15/9.1____ Diffusion Modeling off VCM Concentrations in Air This is to summarize my ideas concerning the possibilities of modeling the concentration of VCM in the air within compartments once it has been released from PVC during processing. SUMMARY The well-mixed model is a reasonable approach toward establishing general guidelines for VCM levels. Modeling of warm plumes containing higher VCM concentrations from some points of a process could.be done, but would be rather specific from one example to the next. ` Handling these with local forced convection (e.g., suck up the plume into an exhaust) is an obvious answer anyway. Computational fluid dynamics would be appropriate only in an extreme case. If, in the process of data collectidn, some anomalies turn up, I would be happy to help try to interpret them in terms of some of these ideas, but I do not see any approach toward a general model, other than the well-mixed one, that would be of any value at this time. DISCUSSION The simplest approach is to assume that within a given compartment the concentration is uniform; i.e., the air is well mixed. I understand that your current approach is based on this assumption. There is a lot of justification for this assumption in most practical situations. I have reviewed the work of Gary Huvard, for example, (G.S.Huvard, "Tremco's Mono Acrylic Sealant: Modeling the Release of Controlled Exposure Monomers and Solvent in Indoor Environments", Corporate Research Report, February 9, 1983), in which he modeled the air concentrations in rooms of solvents and monomers released from caulking material. He used the well-mixed assumption, citing several HVAC sources for support. He was primarily concerned with domestic environments where typically every effort is made to minimize air exchanges with the surroundings. In most manufacturing environments, effort is made to promote air exchange with the surrounding, thus leading to more forced convection currents and making the well-mixed assumption even more justifiable. The next level of complexity would be to define subcompartments, or zones, within a given total compartment, which are regions that are partially isolated from each other but still have some opportunity for air exchange. Examples of this are the socalled "zone models'* used in modeling transport of smoke from fires in multi-compartment buildings, where typically each compartment, or room, has two zones; an upper hot layer and a lower cool layer. The>` irO^WNv LlTHO IN U S A well-mixed assumption is made for each layer, but mixing between the layers is assumed to be negligible (except for some well defined conditions) . The volume of gas in each zone in a given compartment is computed from models for fire-driven (buoyant) flows. There is a slight analogy between the smoke transport models and the release of VCM from PVC during processing in that most of the VCM is probably released in areas of the process where the PVC is hot. Thus, a hot buoyant plume, transporting released VCM, would tend to form above the process. If this was relatively undisturbed, it would form a hot layer at the top of the compartment which would then contain most of the released VCM. However, it is likely that the temperature is not high enough to give enough energy to the plume to overcome the forced convection currents that exist in most manufacturing environments; thus, the plume would be dispersed before it could form a distinct "hot zone" at the top of the compartment. However, there could be local areas where this might be a consideration. For example, as we were touring the PVC..compounding building at ALGC.on,1/12/93, at one point we were on a mezzanine above a hot extruder machine. . Molten. PVC was exposed to the air as it went into the machine. There was also a partial ceiling above the mezzanine. It was distinctly warmer on the mezzanine than on the floor level where the machine was; you could feel a rather sharp thermal gradient as you came down the stairs. In a situation like that, if VCM is being released from the molten PVC it is likely that it would be concentrated in the hot air at the mezzanine level. Anyone in that area might be exposed to a concentration much higher than the average concentration in the whole room. In fact, warm zones like that above processing equipment are probably a pretty good tip-off to likely places to look for higher concentration levels as well, since the mechanisms for dispersion of temperature and concentration differences are similar. Local zones like this could probably be modeled to give at least a ball-park estimate of what the VCM concentration levels might be, but the correction for a situation like that is obvious; i.e., local forced convection near the extruder to break up that hot plume. The last level of modeling complexity would be to use computational fluid dynamics to compute in detail the velocity, temperature and concentration profiles existing in the room, or some part of the room. This would only be appropriate for looking at one specific case in which the results are deemed to be very important (e.g., a litigation case). It would not produce general results which could be used to establish guidelines. Distribution: RVCM Label Team F.M.Galloway 7-,- distribution f- RICHARD J. KRUSZYNSKI GUI ugct t'Ull-LJ neco point on ocpt. s. ei_oc aIo. t l CLEVELAND D/5401 (RdK-061-75) OSHA - NIOSH - EPA 2nd Quarter, 1975 Update J LtTlGH OATE THIS EETTCR 7/ 8/75 This report is presented as a quarterly effort to keep the plastics group informed of current activities in product-related environmental affairs. OSHA-NIOSH: Well, April 1st, 1975 came, April 1st 1975 saw and April 1st, 1975 finally conquered. The new OSHA guidelines have been set and it looks like they're here to stay. Within the past few weeks, the Supreme Court refused to hear an appeal from plastics industry, representatives who attempted to. have the Occupational Safety and Health Administration standards for employee exposure to vinyl chloride monomer eliminated. Many new inroads have taken place by the B.F.Goodrich Chemical Co. in regards to lowering residual vinyl chloride monomer levels of our resins. Although the levels have been reduced drastically in lost of the resins, these changes are being related to compound problems that are occurring in our customers' plants. These problems are being studied and, hopefully, a quick solution can be worked out. In my last quarterly report, I discussed our Bag Labeling Program and our objective to obtain "Heal World Data.." The objective of the Bag Labeling Committee was to collect enough data from customers' plants and laboratory testing so as to obtain an exclusion from OSHA in regards to elimination of the PVC warning stickers from bags and boxes of compound and resin. Our program has just been completed. The results of our monitoring programs and lab testing are nOw being compiled. Jerry Cohan will write up the formal report and this will then be forwarded to OSHA for review. I have included a copy of a rough summary of sampling results. The data that has been taken from our customers' operations shows that 49 out of 52 static or area samples showed levels that were either non-detectable or within the OSHA "action level" of 0.50 ppm. Our lower limit of detection at our Avon Lake Labs is 0.01 ppm. The remaining three area samples that showed higher levels of AVCM were taken in areas where direct workmen's breathing zones were not involved. Another plus factor is that all 42 personal samples taken fell<^withifl the OSHA "action level" and in most cases, were less than our lower limit of detection of 0.01 ppm. '(( on c tll.o IJ1I r ISTRIBUTION GE 2 _ JLY 8, 1975 Snvoronmental Protection Agency is was stated In my first quarter report, the primary concern of EPA Ls to develop an air emissions standard for vinyl chloride. It was originally reported that EPA intended to publish the proposed standard by June 20, 1975 in the Federal Register. However, as of bhis writing (July 8, 1975), the proposed standard has not been published. Perhaps the final standard will be pushed up a bit also. This is the only plastics related EPA information that I have to report at the present time. ][y next report in Fall of 1975, should contain the results qf OSHA's actions regarding our Bag Labeling Program. Also, more definite state nents regarding EPA1s actions.should be available; * JK: cj s JJS/GRT/DAP/WTR RJJ/DLA/RJA JDT/BFT/TAG File f . .1.. : TITLE MIL IL L. OOv .>! )\I J l.U/Vil'A[IY SK ETCMES_ANp. COMPUTATlONS.j. iI 4- --A- : <V ",u7'/ ^- "jm" ;-v! ,4' Ay IY yi _ V . -`1 ^iN\f / 'xs> <<vr.jev 1^ <*n 1 ( ..Y < ft: 0AT[7 Ti V 1 O; .[1 vii V -7 1 vL i! : <T\ ' 7 i i V)! !... _! . >K ^\i ! --! -i..... .! i ; L_ k! | >>! ! I ISJS.-- r I 1 I i! * -- \I I i \ I Til LI' II <I IIi : TML H. r. GOODRICH COMPANY SK: ETCHEf -Sr-AiNDi CO- MiPUTj A TrIOi NrS'j- H = -x 0 -v ?<$:; -i* \' i O' o! q'i t q! r ! II : vf .!'.vi.* V| vi vi v; v v I _! _ I ! JJOonB NOip. l o `'n'r1'"'; q\ !i <vs!i o'. Ii x!i ^ v! Vil Vi! V Vl < (> $ ..Jo.0:J.j4.M$ * $.5kI"JL' i 1 _ iUPt > i'--'l HbN C\ Civ Qk| n t -- l-.-aj-'-i i: i-49\ T Ti ". ; ; fi i\"v\\T'hi! -p t -q--j--- j--|---* .hi Nt"---- i s. ; i | R; S: i ! | f. j i ! K-` > - -V $ s ^ - o, *v >. <S 1 -r-4-- _i----- S--1.J....: *T tIVv- TO- ^v 'tj ^ i al 11 | !! C^ S5 5 \ h ~t V' rf X"tf(trXt-r,xfr^A'&Ttik'T<-lnO;i-6xs|x0: f1 Nr f ~N*V A- T 10 -- iA|4 t "T.... }n ti$xH']--r Hi i A; GO V. iipzz rfffiSI Hitt \\ 1 l\*v> V: Vl L__ . ______ ""Vr -I S| A XK (5 Q b. W~N\J .<o vo ' : U' t Qi i I _j tv I : ^ rx X :;ii 'Iv :o. -r^- A y-. vO, N 4Gv V $i til11 j t i < M -ALUr i- ^-0 --!'.)! i - s! H' ! S! Cnt Cii fi vV X l fifln )(S i4 -?k>*.>s* I -O' r !I k 1I ;i 11 1 j' i : i 1 Ml: U. r.' GOOUUICIl COMPANY 1 l i i -| j S !<: GTC N E S _A N D. COM P U T AT 10 N S L 1 11 ' f 1 1 ! % 1 1 1 ' N 1 * ' JO(7 NO. : 1 ------------------------- -----------;------------------------------------- ^---- N,--------------- --------------------- -------- - I-------A-X : 3 V: ^ V! Vi V 1.....\.....i Q' | \> O' Ql Jn f I V vl "1 - : DATU' !i ii .11 .... ! _J r- *!1 j1 1i ! !* !i ji - ! I..--... . I I f MAT BALANCE APPENDIX-- SUSPENSION RESIN - POWDER COMOUNDING AIR VCM V ROOM ^ AIRx VCM^ ASSUMPTIONS 1. Complete mixing of VCM-Air in room. 2. 1# mol VCM = 359 cu. at room T,P. ft V dc Q (cln - cout) ct = Cla + (ct_0 - cln) exp (zQt) ct-0 = 0 (V) couC=cin (1 - exp (-Q.t)) (V) B416/Lab 103 V = 6021 cu ft Q -- 100 cfm, 1 volume change air per hour. cin" SS# cmpd / 0.85# resin (103-45) VCM / #Mol hr 7 #crapd 106#resin / 62.5IVCM 359scf/______ Min Hr #Mol / lOOcf air 60 Min - 2.60 ppm (1 - exp (-loot) ) (6021 ) t (min) 0 30 60 90 120 gout (ppm) 0 1.02 1.64 2.01 2.24 c room) ft Vv O? o? o rk CS r DISPERSION PLASTISOL FUSION AIR VCM N ROOM AIR\ VCM' ASSUMPTIONS 1. 1# mol VCM CU ft @ room T,P. 2. Complete mixing of air VCM in room. V_dc_ * QiCi - QoCo =* Q(ct - Cc) dt Gout = cln + (ct_0 - cu)exp(-Qt ) <=t-o ~ 0 (V ) Gout = cu(l - exp (zQt . ) (V ) B413/Lab 115 V = 4743 cu ft Q = 80 cfm, 1 volume change air per hour. RVCM = 38PPM, VCM released by plastisol during compounding 1ST HOUR cln " 38# VCM 960*2 106# Plast-1 hr Plastisol # #Mol 359scf/ Min Hr 453.6g 62.5# #Mol /80cf air 60 min = 0.192ppm VCM 0.192 ( 1 - exp r-80tl) [4743] t fmin) 0 30 SoutXREm) o 0.076 60 0.122 2ND HOUR [cc.0 = .122] RVCM = 67 cout = cin + [.122 - cin] exp (~80t) (4743) 67#VCM / 960*2g plast/ # / #Mol / 359scf / Min Hr 106# plast 453.6g / 62.5# / #Mol / 80 cfm / 60 Min \ = 0.339 ppm VCM w(V 7 CO o rt CO r 0.339 + [.122 .339]exp(^80t) (4743) = 0.339 - 0.217exp(-80t) (4743) t (min) 0 30 a...- (ppm) .260 .157 3RD HOUR (1/2 hour) <=out = <=ln + (ct-0 ' cin)exp(^t) cin - 0 (V) cout = 0.2 60exp(^80t) (4743) _t (min) 0 30 c,,,,. (ppm) .260 .157 r <v: O) or.