Document RpmyxRKv1V8oqvjp58LG8xaqz

^ pc: M. M. O'MARA 8/19/83 - pc: - " " Am. Ind. Hyg. AssiC44<3):I76-I83 (1983) ksb > ^ Evaluation of Emissions from Simulated Commercial Meat Wrapping Operations Using PVC Wrap THOMAS J. SMITH. Ph.D.\ JOHN J. CAFARELLA. M.S.. CHARLES CHELTON. M.S. and STEPHEN CROWLEY department of Environmental Sciences. Harvard School of Public Health, Boston. MA ^luHaraypitr'T <t***inw >5 an elusive health problem with a suspected relationship to exposure to emissions from polyvinyl chloride (PVC) film cut with a hot wire. A study was conducted to determine how the type of wrap cutter (wire or rod) and its temperature afTected the emissions from a simulated occupational wrapping process. The cutting temperatures covered the same range as was measured in Boston retail food stores. A commercial wrapping machine and samples of commercial PVC meat and produce wraps were used. Seventy five percent of the particulate from the hot wire was respirable, and the quantity of emissions was a strong function of the film tension and cutting technique. Particulate emissions did not increase steadily with increasing wire temperature, but plateaued or declined at high temperatures. Particulate emissions from the rod cutter were very low at low temperatures, but exceeded those of the wire at temperatures above 200 0 C. The particulate was 100% dioctyl adipate (DO A, the plasticizer in the wrap) with wire temperatures below 200 C, and was approximately 80% DO A for temperatures above this. Gaseous HCI was not detected in emissions from a hot wire operated below 150 C, but HCI emissions increased rapidly to a plateau for temperatures above 200 0 C. Approximately 20% of the HCI produced at temperatures above 200 0 C was associated with the particulate, which appeared to act as a carrier and transport the HCI through water filled impingers. Field tests are needed to determine if particulate produced in the workplace may also behave as a carrier for HCI. R O c! i V ZI D JUL 1 8 1563 Introduction W. F. CAiNinOLL Polyvinyl chloride (PVC) film is widely used to wrap com mercial products for retail sales, such as meat and produce. The wrap protects the product while allowing customers to inspect its qualities. The simplest and easiest way to cut the film is to draw it across a hot wire, which melts the film so it Because of the uncertain nature of the health problem, a five year prospective epidemiological study of retail food workers has been undertaken by HSPH to try to determine if there is a relationship between respiratory effects and the exposures of retail food workers. can be pulled apart. This process produces emissions when small amounts of PVC stick to the wire and are pyrolyzed. The objectives of the HSPH environmental evaluations were to characterize the nature of the retail food worker Laboratory and field studies have shown that thermal decomposition of PVC film produces particulate from the exposures to air contaminants, and to identify groups of workers with qualitatively and quantitatively different expo film's plasticizer (dioctyl adipate, DOA, or dioctyl phthalate, DOP), and gaseous emissions containing principally sures so that the epidemiological evaluations could be per formed to determine if the exposure groups had significantly hydrogen chloride (HCI).11,2* different risks of respiratory effects. Beginning in the early 1970s, there were a number of case Previous studies have shown that the amount of particu reports claiming occupational respiratory disease associated with emissions from wrap cutting.13'*' Most reported an asthma-like syndrome which has been named "meat-wrappers' late and HCI emissions appeared to be related to cutter temperatures, and that the rod cutter has much lower levels of emissions than the wire.12'6' However, these studies did asthma". However, the precise nature of this disease has remained elusive, and some-investigators have concluded not relate the amount and composition of emissions to the cutter temperature in sufficient detail to be able to pre i that it may not be a single entity.t4'*' As a result of the concern about emissions from the hot dict exposures from the operating temperature. They also did not evaluate the effect of work practices on the amount wire cutter, the industry developed a new type of cutter, the of emissions. "cool rod", which had a larger thermal mass than the wire so it could effectively cut the wrap at lower temperatures. A thermostat was added to control the temperature and keep it low. Preliminary field surveys by the Harvard School of Public Health (HSPH) showed that both types of cutters were widely used in meat and produce departments, and that both had a broad range of operating temperatures. Although our main interest was the workroom exposures of retail food workers, we studied the emissions from the wrap cutting process in the laboratory for several reasons. First, a large number of experiments could be performed under carefully controlled conditions. Second, the emissions could be collected from the source without concern about the disruption that might be caused to work activities in a Thi* work was supported through a contract from the Society of Plastic Industries, and in part by Occupational and Environmental Health Center Grant No. 5 P30 ES00002-20 from the National Institute for Environmental Health Sciences. retail store. Finally, work activities and localized air move ment vary considerably from day to day in retail stores which would add large variability to the observations and substantially increase the numbers of samples that must be Cap9Wq 1M1 Amencan industrial Hvq<n Association tH im Ina. Hyi Assoc J (44) Marcr., 1983 BFG05450 TU02S402 sampling apparatus. collected to characterize a given relationship between emis sions and cutting conditions. Therefore, particulate emis sions from the hot cutter were collected and concentrated within a glovebox, and the HC1 emissions were collected with a glass hood. These procedures may cause artifacts because the emissions stay concentrated for a longer time than they would in a work environment. These possibilities will be discussed in the interpretation of the results. Our results confirmed earlier findings by some investiga tors, and contradicted others. There were four important new findings: (I) particulate emissions are not linearly related to wire or rod temperature, but plateau or fall off above critical cutter temperatures; (2) the rod cutter can be a major source of emissions when operated at temperatures above 165 C (330 F); (3) cutting technique and film tension are important, and poor technique can produce substan tially higher emissions than good technique; and (4) the HC1 emissions followed the particulate levels, except there were no detectable HC1 emissions below a wire temperature of 150 C. Methods A standard meat wrapping machine (Heat Sealing Equip ment Mfg. Co., Model624, Cleveland, OH), with thesealing pad disconnected, was installed in a specially constructed glovebox (see Figure 1). The machine was placed with the cutter at the front of the box so that the operator could simulate normal cutting operations while the box was sealed and all emissions were collected within the box. The cutting machine had interchangeable hot wire or "cool" rod cutters. The temperature of the wire cutter was adjustable by varying the voltage across the wire. Because of the low mass of the hot wire it is likely that the wire temperatures were some what underestimated. The cool rod was a Film Cutter Model AT 1000 (Borden Co., North Andover, MA) the tempera ture of which could be adjusted by an internal potentiometer or an external voltage controller (for low temperatures). Temperatures were measured with an Omegatemp low mass, electronic thermometer(Omega Engineering Inc., Stamford, CT). The relative humidity inside the box was measured with a psychrometer(Psychron Model 566, Bendix Corp., Envi ronmental Science Division, Baltimore, MD). Several types of measurements were made of airborne particulate. Total particulate mass was collected on a 1.0 jun pore size. 37 mm. Teflon membrane filter in closed face cassettes operated at 1.7 Lpm. Respirable particulate was collected with a 1.0 (tm pore size Teflon filter preceded by a 10 mm. nylon cyclone operated at 1.7 Lpm. Particulate number-size distributions were measured with a Laser Opti cal Particle Spectrometer, Model CSASP-100 (Particle Measurement Systems, Boulder, CO). To Collector a. Positions before cut Figure 2 -- Diagram ol the glass hood for collecting the emissions from wire cutting. American Industrial Hygiene Association JOURNAL (44) 3/83 BFG05451 Os) 177 Wir Ttmpwatvre <*C) Figure 3 -- Total particulate emissions as a function of wire temperature and cutting technique. The following experimental procedure was followed: 1. the interior of the glove box was flushed with a large flow rate of filtered air for ten minutes; 2. the wire or rod cutter was set to the desired temperature, 3. the box was sealed and 30-180 cuts were made with the cutting technique selected; 4. the sampling devices were run for 20 minutes; and 5. the relative humidity was measured inside the box. Two types of cutting techniques were used: "good practice", quick movement across the cutter while maintaining a high tension on the film; and "poor practice", laying the film on the cutter with little or no tension on the film so the cut is slow and the separation is caused by the weight of the film pulling away from the cutter. If relative humidities higher then ambient were to be studied, then a pan filled with distilled water was heated and/ or boiled inside the box to saturate the air, then the pan was taken out and equilibrium established. The wire or rod was cleaned of carbonized deposits with steel wool before each experiment, and as soon as cutting was completed. Triplicate samples of the aerosol produced at 150,177,204 and 232 C by "bad practice" hot wire cutting were analyzed for di(2-ethylhexyl) adipate (DOA). DO A was extracted from the samples with methanol and the extract was ana lyzed by gas chromatography usinga3 mm.l.D., 1.6 m., 3% 171 SE-30 column and a flame ionization detector. Separate studies demonstrated 100 percent recovery of DOA from spiked Teflon filters. Air concentrations of HC1 were measured by sampling in 10 mL of water in a midget impinger and analyzing the resultant solution colorimetrically by the mercuric thiocyanate method.<7) Initially, measurements were made on air samples collected in the glove box studies of aerosol emissions. However, the concentrations were unreasonably low, so the experiments were repeated using a glass hood that could directly collect emissions from the hot wire (see Figure 2). A sample of the meat wrap was wound around the wooden block holding the top half of the hood and then pressed down on the hot wire and the bottom half of the hood. Air was drawn in the open ends of the hood, sweeping the emissions into the sampling train. Three impingers, each containing 10 mL of water, were connected in series to the glass hood and operated at 2.0 Lpm. An in-line filter (47 mm glass fiber) was used to remove aerosol emissions before the gases entered the pump. In one set of experiments the filter was placed ahead of the impingers to remove the aerosol before the gases passed through the impingers. Between tests the apparatus was disassembled and cleaned with methanol and distilled water and air dried. Air sample blanks were collected without cutting to verify that the apparatus was clean. The cutting technique in the glass hood experiments was a "poor practice" type because the film only had enough tension to pull the cut ends apart from the wire a few millimeters. Am. Imt. Hyt. Assoc. J <**) Mirch, 1983 0752003 BFG05452 Results ( tice as a function of wire temperature. Neither technique Hot wire and rod cutters produced particulate emissions ( showed a uniform tendency for emissions to increase with during cutting of PVC wrap at all operating temperatures, l wire temperature: good practice emissions declined above a 135-220 C. The particle number-size distribution measured i wire temperature of 150 C. and bad practice emissions with the Laser Particle Spectrometer was independent of j plateaued above 170 C. cutting temperature and had a logarithmic-normal distribu Good practice cutting technique required: 1) quick down tion with geometric mean of 2.1 to 2.3 nm, and a geometric ward contact with the hot wire, drawing the film down to standard deviation of 3.8. Based on this size distribution, 75 form at least a 30 angle with the plane of the film from the percent of the particulate mass should be in particles less roll; and 2) strong film tension to quickly pull the cut ends of than 5 pm in diameter. This estimate was consistent with the the film away from the wire when it softened from the heat. measured fraction of respirable particulate mass (the ratio Under these conditions, the film cut more quickly as the between respirable mass sample and total mass sample), temperature of the wire increased. At temperatures under an average of 86 percent, although the respirable mass 150 0 C, a small amount of clear deposit remained on the wire was usually a larger fraction of the total than estimated from after 60 or more cuts. At temperatures above 170 C, fine the number-size distribution. strands of char accumulated on the wire after about 60 cuts Relative humidity did not affedt the panicle concentration and in some cases reduced the cutting efficiency of the wire from the hot wire or its size distribution over the range hv insulating the film from the wire's heat. of temperatures. J Cutting with "poor" practice occurs if the film separates Particulate emissions from the hot wire were strongly A because of its own weight and loss of mechanical strength related to cutting technique and wire temperature. Figure 3 I from the heat. At temperatures above 150 C, substantial compares emissions resulting from "good" and "poor" prac- 1 visible emissions were apparent with each cut, which. Figure 4 -- Total and respirable particulate emissions as a function of temperature. American Industrial Hygiene Association JOURNAL (44) 3/83 BFG05453 o a oo 179 o UJ o U3 o CL O .o Figure 5 -- Total particulate emissions as a function of temperature and type of wrap film. because of accumulated deposits of film on the wire, become a continuous emission after several cuts. Substantial depos its of char accumulated at temperatures above 170 C. At higher temperatures, the poor practice technique produced I seven to ten times as much particulate as the good pracI tice technique. Emissions from the rod cutter were not sensitive to opera tor technique and had the relationship with temperature shown in Figure4. For rod temperatures 135-165 C, the rod emitted particulate at a rate less than or equal to that of good practice wire cutting. At temperatures above 165 C, emis sions of particulate increased sharply until at 210 C they exceeded those of the "poor" practice hot wire technique. At high temperatures -- above 165 C -- the film parted on edges of the strip in contact with the rod leaving a 2-4 mm wide deposit along the length of the rod. Successive cuts added to the deposit, which became a continuous source of emissions. When the temperature of the rod was increased above 165 C, the amount of non-respirabie particulate (>5 ^m in diameter) stayed approximately constant at 0.01 mg/cut, even though the respirable fraction increased by seven fold. Rod temperatures as high as 195 C have been observed in the stores. The relationship between particulate emissions and tem perature for hot wire cutting with the good practice tech nique was affected by the composition of the film. Figure 5 shows a comparison of particulate emissions from the meat wrap and produce wrap. These films differ in their composi tion. primarily their type of plasticizer: DOA in the meat wrap, dioctyl phthalate (DOP) in the produce wrap. The peak of particulate emissions occurred at a much higher temperature for the produce wrap (190 C) than the meat no wrap (150 C). These temperatures do not correspond to any known physical property of the wrap films, such as soften ing temperature. Analyses of the aerosol from the hot wire cutting showed that the aerosol was 100% DOA (107.6% average with a standard deviation of 10.8) at cutting temperatures below 232 C. At 232 C, the DOA accounted for 79.2% with a standard deviation of 8.6, which was significantly less than 100% (p> 0.05). -- The relationship between the HC1 emissions and wire temperature is shown in Fiure6. The total HC1 emissions per film cut were very small at 150 C but increased rapidly up to 204 C and then apparently plateaued at approximately 45 ^g/cut, similar to the particulate emissions. The relative quantities of HC1 collected among the three impingers showed that an interaction took place between the HC1 and the aerosol. Experiments showed that a single impinger would collect 99% or more of gas phase HC1 if the aerosol were removed with a filter. When the aerosol was present, 4 to 13% of the HC1 penetrated into the second impinger as shown in Figure 6. The amount collected by the second impinger increased with the wire temperature. Thus, there appeared to be an interaction between the HC1 and the aerosol that carried HC1 through the first impinger. Anal yses of the aerosol produced at 202 C and 232 C suggested that there may also have been some adsorption on the aero sol because a significant amount of chloride (15% of the total emitted) was extractable from the particulate (see Figure 6). This change in particulate composition was also apparent in the DOA analyses, which showed that the particulate formed at 232 C was only 79% DOA whereas at lower temperatures it was 100% DOA. Am Ind. Hyi Assoc. J(44) Marcn. 1983 BFGO5454 20752005 Discussion These studies clearly show that cutter type and operating temperature are critical determinants of the quantity and composition of emissions from wrap cutting. They also show that the operator's technique on the hot wire cutters is very important in determining exposure. An unanticipated result was the finding that the rod emissions could exceed those of the wire at the high temperatures that have been observed in the stores. The tension on the film from the wrapping machine was also important for obtaining good practice results. If it was insufficient, the film would not retract from the wire and "poor" practice cuts would result in spite of the operator's attempts to do otherwise. In practice, it was difficult to make a large number of cuts with good technique; approximately one in ten cuts was not sharp and would leave a deposit, unless the operator was very careful. The technician was able to achieve large numbers of good technique cuts by much practice and careful attention during the procedure. Particulate emissions increased with temperature for both the wire and rod cutters as a result of pyrolysis of bits of the film deposited on the cutter. The emissions from the hot wire plateaued at temperatures above 150 C, probably because of the limited capacity of the wire to hold molten PVC. The rod had a larger surface area and could hold larger deposits. At temperatures above 165 0 C, deposits collected on the rod after several cuts formed a source of continual emissions which was observed to smoke for some time after the cutting had stopped. The data presented in the"Results" section do not include this extended emission because the rod was wiped clean once the desired number of cuts had been completed, so the emissions measured represent a lower boundary of the levels that might be seen in the meat wrapping environment. The glass hood experiments demonstrated that hot wire cutting of wrap did not generate significant levels of HC1 until the wire temperature exceeded 160 C. The experi ments also showed that the aerosol formed at hgher temper atures may act as a carrier of HC1 enabling it to pass through impingers that would collect 99% if it were present as a gas only. Passage through the first impinger apparently released some loosely held HC1 because the relative fraction of HC1 collected by the second impinger was substantially higher than the first impinger. Ten to fifteen percent of HC1 was apparently bound to the aerosol because it was carried through three impingers and collected on the glass fiber filter I at the end of the sampling train, even though the third impinger showed little or no HC1. The DOA analyses of the aerosol also showed that the composition of the aerosol changed with increasing temper ature. Above 200 C approximately 20% of the aerosol was something other than DOA. This was consistent with the HC1 experiments which showed that above 200 C, 5 to 7 figl cut of particulate chloride were present but this does not account for all of the non-DOA aerosol. It is likely that the remainder of the aerosol was carbonized PVC fragments, because significant amounts of char were formed on the wire at these temperatures, but not at lower temperatures. This carbonized material would also be a reasonable carrier for.HCl. i I i j i Figure 6 -- Emissions of total chloride (as HCI), aerosol chloride and chloride collected in the second impinger as a function of temperature. American Industrial Hygiene Association JOURNAL (44) 3/83 BFG05455 *0 c NOl O o 181 i tir ' Stone and co-workers observed that the carbonaceous soot from burning torches of mixed PVC and polyethylene plastics had both loosely and tightly bound forms of HCI.'8' They found 19 mg/g of HCI that could be easily stripped from the aerosol with a nitrogen purge, and 27 mg; g of HCI was tightly bound to the aerosol. It was not reported if any of the tightly bound HCI could be removed from the soot by water extraction. Their findings were reasonably consistent with our observation that part of the HCI was carried into the second impinger by the aerosol, while another part was extractable from the aerosol that passes through all the impingers. These findings are potentially very important because they imply that the aerosol may act as a carrier for the HC! to transport it deep into the lungs of the exposed workers, whereas if the same exposure were all gas phase HCI it would be removed by moist upper airways. However, the condi tions of the laboratory experiment do not duplicate the exposures received by workers, because the concentrated emissions from the hot wire were immediately captured by the hood and passed through the collection system. This may have produced an artificial situation that is not present in the work environment because normally the emissions are diluted by mixing with room air as soon as they leave the wire. The increased contact time and high concentration conditions in the laboratory may have enhanced the interac tion between the HCI and the aerosol. Field experiments are planned to determine if aerosol produced under high temperature cutting has a significant HCI content when it reaches the worker's breathing zone. _ Our findings compared quite well with those of Boettner and Ball, if we presume that their smoky and clean cuts were equivalent to our bad and good practice cutting tech niques.19'Their HCI and DO A values were both significantly higher than ours: HCI was 50% higher, and DOA was 100% higher. This was probably the result of better recovery of emissions in their experiments, and wall losses in our exper iments. We both observed that approximately seven times more DOA was emitted by smoky cuts than clean cuts. Our observations that HCI was negligible at 150 C confirms Boettner and Ball's determination that no HCI was pro duced at 135 C. They did not measure the composition of the aerosol so no comparison was permitted with our obser vation of absorbed HCI on the aerosol. Our emission con centrations were more reproducible than theirs, i.e.. had smaller ranges under a'given set of conditions. This may have been a result of our closer control of cutter temperature and film tension during cutting, although there was no direct evidence to support this. Another relevant study was performed by Van Houten and coworkers, who looked at "worse case" emissions from a wire cutter under a variety of temperatures.12' The wire temperature was measured with a thermistor, which we have found may substantially underestimate the temperature if the sensor is too large relative to the wire (they did not report the size of the sensor). They found that total particulate emissions increased with temperature, but they did not observe any tendency for the amounts of emissions to pla teau at higher temperatures. However, they did observe the HCI concentrations plateaued at 171 to 182 C, and were very slight at 132 aC and below. This general behavior of both HCI and. to some extent, total particulate emissions, was consistent with our findings, although the temperatures at which they occurred for Van Houten el al.. were 20 C lower than ours. They also found that the particulate was 100 percent DOA; it ranged from 50 percent at 82 C up to 92 percent at 132 C, but there was no clear trend and no standard deviations were present to show the precision of the measurement. The wrap film appears to decompose at lower tempera tures on the hot wire and rod than expected from bulk PVC studies. Boettner and co-workers presented data on powder samples of PVC polymers and showed that dehydrochlorina tion does not begin to appear on differential thermal anal yses until 240 C.I1(" Apparently, thin deposits of PVC on a wire or rod behave differently than small samples of powder. We observed the formation of char at temperatures above 170 0 C for both rod and wire cutters, which suggest pyrolysis of the flm. The HCI measurements suggest that significant dehydrochlorination of the film deposited on the wire or rod begins at 160 C. However, the PVC does not appear to have been completely decomposed at temperatures within the ( normal range of the cutters. If the dehydrochlorination were i complete, then the ratio of HCI to DOA in the emissions ! would be approximately 1:1 because the film contains approximately 30 percent DOA and 30 percent chlorine by weight. Based on our data at 205 C and 230 C, it appears that approximately 50 percent of the potential HCI yield was released. The data from Boettner and Ball suggest only 10 to 15 percent of the HCI was released at approximately 215 C, but this is not certain because they did not present paired HCI and DOA data from analyses of emissions and their wire temperatures apparently were not as closely controlled as ours.'9' Acknowledgement The authors gratefully acknowledge the laboratory work of Dr. William Hinds and the many helpful comments from Drs. Alfred Cummin, Cyril Gilbert and Warren Cook, and the members of the Medical Committee of the Society for Plastics Industries. References 1. Vandervort. R. and S.M. Brooks: Polyvinyl Chloride Film Thermal Decomposition Products as an Occupational Illness. 1. Environmental Exposure and Toxicology. J. Occup. Med. 73:188-1 91 (1977). 2. Van Houten, R.W., A.L. Cudworth and C.H. Irvine: Evalua tion and Reduction of Air Contaminants Produced by Ther mal Cutting and Sealing of PVC Packaging Film. Am. Ind. Hyg. Assoc. J. 35:281-292 (1974). 3. Sokol, W.N., Y. Aelony and G.N. Beall: Meat Wrapper's Asthma: A New Syndrome. J. Am. Med. Assoc. 226:639642(1973). 4. Brooks. S.M. and R. Vandervort: Polyvinyl Chloride Film Thermal Decomposition Products as an Occupational Illness. 2. Clinical Studies. J. Occup. Med. 79:192-196(1977). 1*2 Am. Ind. Hyi. Assoc. J (A4) March. 19S3 1 BFG05456 0752007 i iir  .5-. Krumpe, P.E., T.N. Finley end N. Martinez: The Search (or Expiratory Obstruction in Meat Wrappers Studied on the Job. Am. Rev. Resp. Dis. IIS:611-618 (1979). 6. James. D.G.: Cool Rod Film Cutting Device Ends Packing Room Fume. Packaging Engineering 52:2b-11 (1975). 7. Zell, D.M., D. Fischer and M.Q. Garner: Photometric Determination of Chlorides in Water.Anai. Chem. 28:11651168(1956). 8. Stone, J.P., R.N. Hazfett, J.E. Johnson and H.W. Csrhart: The Transport of Hydrogen Chloride by Soot from Burning Polyvinyl Chloride. J. Fire Flam. 4:42-51 (1973). 9. Boettner. E.A. and G.L. Ball: Thermal Degradation Products from PVC Film in Food Wrapping Operations. Am. Inti. Hyg. Assoc. J. 4r:513-522 (1980). 10. Boettner, E.A., G.L. Ball and B. Weiss: Analysis of the Volatile Combustion Products of Vinyl Plastics. J. Appl. Polymer Sci. 13:377-391 (1969). 12 August. 1982: Revised 12 October, 1982 20752008 American Industrial Hygiene Association JOURNAL (AC,i 3/83 BFG05457 183