Document By15QdVNvqvE87o2mB1p67xVL

*d J. 0. Fannin P. M. Zakriski SUBJECT POINT UB OEPT. C. QlDQ. hj0, Louisvi11e FieLQ POlNiT OH OEPT. & BLOG, no!------------ Brecksville R & D Center, D/8506 Louisville's Vinyl Chloride Personnel Monitoring Program OH LtiTTEa OATE THIS LETTER July 28, 1976 Thank you for the opportunity to visit the Louisville plant and examine in depth the personnel monitoring program for vinyl chloride- Your dual tube program to determine the incidence of vinyl chloride breakthrough has shown that vinyl chloride is breaking through the primary tube about 30% of the time. However, after close examination of your measure-* ments, it is clear that only 13% of the time is this breakthrough greater than 35% of the amount retained on the primary tube. I judge this situation as serious and in need of explanation and correction but not critical. The cause of the breakthrough is not clear to me. It can be related to poor sampling practice, dirty PMCC tubes (inadequately cleaned) and excessive amounts of water on the PMCC tubes. It is likely that all of these contribute to the problem. It is my judgement that thermal cleaning is a small factor. It is more likely that unusually large quantities of water are responsible for the breakthrough. Unusually large quantities of water are those which arise;from collecting suspended water (fog and mists) not just the water from humidity conditions. Your PMCC dual tube program from ^February 1976 through July 1976 generated 215 sets of measurements (2 tube combinations). This represents 9% of the 2524 personnel monitorings conducted during that period. The program shows that in 63 cases (29%) of the 215 dual tut tests breakthrough ; greater-than . 10% occurred. These breakthroughs are not related to tl amount of vinyl chloride retained. My calculations show the following distribution among the 63 breakthrough cases; PPM Range <.ll .11 to .43 .44 to .86 .87 to 1.63 1.64 to 3.26 >3-27 Percent of the 63 Breakthrough Incidents II 14 24 22 19 10 Two qualifications need to be imposed upon your results to bring them into proper per spective. First of all, the number of breakthroughs (63) should be reduced to 55 based upon a probable artifact of dual tube programs. Eight dual tube sets showed virtually nothing in the primary tube and all of the measured vinyl chloride in the secondary or back up tube. It seems that the tubes were reversed and the primary was sampled, labeled or analyzed as the secondary and vice versa. Then, we must not use 10% breakthrough to the secondary as significant but rather choose 35% . When we use this new criteria we see that 13-5% of the 215 tests showed breakthrougf This says that 13.5? of the time single tube personnel monitors will be out of the 35? accuracy interval identified by OSHA for measurements In the range 0.5 to 1.0 ppm. In othe words one out of every 10 personnel monitorings is a bad test but which one is it? It is safe to assume that if we can find the reason for this breakthrough we can reduce this uncertainty. 1'C-IHI-C ),'m UNO US BFG44728 1-000 9 .1 *\ -2- I examined your personnel monitor program in further depth, looking at the laboratory's work,, the calibration of the pumps by your people, the selection of the test subjects the circumstances of actual field sampling (the incidents of the six hour sampling period' and the storage of the PMCC's before and after sampling. Based upon my examination l make the following comments: a) Proper care is made to calibrate PK pumps prior to and following the six hour sampling period. b) Good attention is given to the selection of employees to be tested. c) inadequate attention is paid to observing the employee, his pump and tube during the six hour sampling period. d) Insufficient attention is paid to the water "puffing" during the analysis of PMCC's. When water "puffing" is observed the measurement should be voided. e) The laboratory needs to analyze freshly cleaned PHCC tubes ('20) to determine the average signal level from cleaned tubes. Then subtract out this value as a "blank". 1 took the time to observe personnel samplngs which were in progress- 1 found.a samplin which was subjected to massive water sprays >md mists. This water can be retained on th PHCC tube. I found a case when the pump and tube were resting on the window sil with no one around. Finally, ?n a spot check for r nten and pumps of 10 samplings 1 could find only 2 tests- Not knowing where the others what they may be picking up or if they ar in use is a serious question. It is a good idea to pay closer attention and record the conditions during the six hour sampling. The incidence of water "puffing" was observed in the lab and l feel it tells us that a great deal of water (too much) is on the l tube. When the tube is removed from the Bendix Flasher after analysis, a puff of jets up from the point where the tube seats in the downstream solenoid. In udditi hi to that, occasionally water is so abundant that it collects in the connecting tube !,. the flasher and chromatograph. This is concluded from observations of water put'i >. i blowing the second, third and fourth triggering of the flasher inject switch .* r first "puff" is observed and before any PHCC tube is installed in the flasher. ' .i. ,-ru this puffing observation is not even recorded. It should be recorded and <: the basis of voiding a measurement. It is a good idea to continue the present .; * .?>e runs employing new PMCC tubes and 300C cleaning. Fifty or so tubes seem ; j t -- to establish if 300C cleaning will eliminate the breakthrough observed from :,v. x.fr to July. It does not seem likely that this will eliminate the breakthrough because breakthrough is not primarily associated with low level snmpl*s. It is clear from humidity studies at Brecksville that humidity up to $2% (relative) does not alter the performance of PMCC's or produce significant breakthrough. There is, howev the likelyhood that water in droplet form (fog, mist) can be filtered by the PMCC thus depositing quantities of water much greater than that resulting from S2% relative humidit BFG44729 1-000 10 T f -3- air. This very large quantity of water could easily account for breakthrough. It should also be associated with the water puffing observed in the laboratory. In the light of suspected high water charges to the PMCC your proposal to try a desicant study is wise. 1 have two recommendations for desicants, silica gel and calcium hydride. The silica gel recommendation comes from work Chuck Titus completed. In testing a vinyl chloride point source for composition, a silica gel trap was placed upstream of a carbon trap. Vinyl chloride and other monomers were transmitted by the silica gel but water was retained. We do not know if this case was 100% transmittion of vinyl chloride through the silica gel. However, analysts of the silica gel should clear up this point. And l would analyze the silica gel with the carbon disulfide procedure just as if it was carbon. The basts of this silica gel recommendation is a mass spectroscopic analysis of the carbon bed which was performed for Chuck by Bob Lattimer of the Brecksville lab* The calcium hydride recommendation comes from work being carried out by John Nikora at the Avon Lake Development Center. John is using a calcium hydride tube ahead of a carbon tube in a gas stream. The stream purges a water sample (and thus becomes virtually saturated with water) to strip vinyl chloride from the water. The calcium hydride removes the water but does not appear to retain the vinyl chloride which is retained on the carbon bed. This is a preliminary conclusion because the calcium hydride bed has not been analyzed for retained vinyl chloride. John has only prepared various mixtures of VC1 and water (at the part per billion level) and obtained good straight line relationships between analyses and VC1 content. You can get more details and keep up to date on this work by discussing the work with John at Ext. 288, Avon Lake Development Center. In summary, I believe you have a breakthrough problem which is serious but not critical. I do not know if this same problem exists at the other PVC production facilities but I will find out. I recommend you continue your breakthrough program with 50 sets using new and clean flasher tubes and 50 tests employing desicant tubes. You and the other plants will be receiving the PMCC/humidity report from Brecksville. Finally, 1 plan to examine all BFG data and experiments and prepare the best statement of accuracy and uncertainty relating to personnel measurements employing the/PMCC/flasher-method. & j'b cc: E. B. Katzenmeyer, Jr. - Akron R. W. Strassburg _ ii. ilv ;Ni kora - .Avon Lake DC J. M. Whitney - Avon Lake GC C. H. Lufter - Brecksvi1le C. E. Titus J* W. Born O. S. Kurtz J. A- Klupar Cleveland P. H. Lawrence Louisvi1le S. S. Michels R. R. Taylor R & D Files Paul M. Zakriski bfG44130 1-000 11 T -10- Corporate Environmental Service 8504-76, SR, 8-24-76 TABLE 1 I I Comparison of VCM Analyses of PMCC's Charged Under Pressure vs Vacuum Man I fold Port No. Exposure Time (min.) VCM Weight ____ Lya) Air Volume Sampled (ml) VCM (ppm) V.CMav,9 (ppm) A. Analyses of PMCC'sCharged at 1.09 ppm Under Pressure*: 1 360 19*0 15,200 0.49 2 11 5-20 3,730 0.55 0.54 3 " 6.78 5,110 0.52 0.025 4 11 12.8 8.920 0.56 5 11 12.4 8,640 0.56 B. Analyses of PMCC's Charged at 1.00 ppm Under Vacuum*: 1 360 3*75 2,930 2 11 7*95 4,430 3 " 7-30 4,210 4 n 4.31 3,Mo 5 " 3-96 3,710 0.50 0.70 0.68 0.49 0.42 0.56 0.11 * The A sample tubes were charged under pressure directly from a small gas cylinder containing a standard 1.09 ppm VCM-air mixture. The B sample tubes were charged by drawing a 1.0 ppm VCM-atr mixture through them with a vacuum pump. The B samples were charged at 30$ relative humidity (see Table 4). Analysis of the VCM from the Monitor Tubes Instrumentation The analytical equipment included a Bendix HS-10 Flasher (Reference 2) connected to a Hewlett-Packard Model 571IA gas chromatograph (G.C.). The G.C. was equipped for flame ionization detection (FID). We substituted an external manual bypass line for the internal solenoid bypass line of the Bendix Flasher. We had been unable to eliminate solenoid valve leakage. We used a 3/16 x 36 inch Porapak QS column. The injection port and detector temperatures were 200 and 250C, respectively. The oven temperature program included 2 minutes at 80C, a 32C per minute rise, and 2 minutes at 160C. The resulting VCM retention time was 3*9 to 4.1 minutes. The Bendix Flasher heater block temperature was 250#C. The carrier gas was helium We allowed two minutes per sample for thermal desorption. BFG44731 1-000 B.F.GOODRICI 22 Rcicerch end Development Cen -II- Corporate Environmental Service 850^-76, SR, 8-2A-76 Calibration of the Gas Chromatograph We calibrated the gas chromatograph for each series of samples with fresh standard VCM solutions. The solutions consisted of weighed amounts of VCH dissolved in carbon disulfide (CS2). Shaking the VCM standard samples for 30 minutes insured complete solution. Dividing the weight of injected VCM by the resulting average G.C. peak area gave the calibration factor (Fcaj) in nanograms per area unit (ng/a.u.) Figure k presents two VCM calibration curves. One curve emphasizes the higher concentrations and shows an of 2.61 ng/a.u. The other curve gives an Fcaj of 2.35 ng/a.u. by emphasizing the lower VCM concentrations. We used the average FCal value of 2.^8 ng/a.u. Standardization of the 1.00 ppm VCM-Air Stream We injected 5~ml samples of the 1.00 ppm VCM-air stream from the supply manifold of the vapor generator into the G.C. column. The samples came from all five ports. Table II confirms that the VCM concentration was 1.00 ppm within 5.3%. Percent Recovery of VCM from Monitor Tubes By "percent recovery" we mean what percent of the 1.00 ppm of VCM drawn into the monitor tube is found (i.e., recovered) by G.C. analysis. Recovery of VCM from PMCC's Table 4 shows the effects of charging time and relative humidity upon the percent recovery of VCM from PMCC's. Figure 5 is a graph which summarizes the results. The relative humidity of the VCM-air stream from 30 through 92% relative humidity (R.H.) did not significantly affect the recovery of VCM from the PMCC's. The amount of VCM which was recovered from the PMCC's at 30, 50, and 92% R.H. was the same within 8$ of the average. This was true over the exposure range from about 60 minutes to 360 minutes. The average recovery for all three relative humidities was 0.72 ppm after 90 minutes and 0.59 ppm after 360 minutes exposure. The amount of VCM recovered during the 1.00 ppm exposures from 100 to 380 minutes duration decreased at 5 x 10" ppm. per minute at 30% and 50% R.H. and at 8 x lO"** ppm per minute at 92% R.H. Even at 92% R.H. a 0.10 ppm decrease required more than 100 minutes exposure. Therefore, corrections for humidity effects in the 1.0 ppm range are unnecessary for exposures of less than 100 minutes at 10 ml per minute. Table 5 summarizes the results from the latter study. It also lists the results from two earlier duplicate studies. The results in Sections A and B agree well with each other. The agreement between the results in Sections A and C is not as close. The PMCC's of Section C had not been cleaned for VCM as well prior to use. However, all three studies support the same conclusion. The PMCC's yield only about a 60% recovery when charged at the 1.00 ppm level at 10 ml per minute for six hours. BFG44732 1-000 B.F.GOODRIO 23 Research end Development Cent -12- Corporate Environmental Service 8504-76, SR, 8-24-76 FIGURE 4 (si!un DEUV) Bsuodsey g-g BFG44733 -13- Corporate Environmental Service 8504-76, SR, 8-24-76 TABLE IV Effects of Charging Time and Humidity on VCM Recovery from PMCC's*. \. A. Measurements at 30fc Relative Humidity Manifold Port No. 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5 Exposure Time (min) 15 it ti ii 30 ii n it ii 60 n ii ii 11 90 ii it ii ii 180 n it ii 360 ii ii it VCM Weight (Uq) 0.417 0.310 0.391 0.663 0.416 0.930 0.396 0.725 0.699 0.691 0.856 1.41 0.5)2 2.18 0.422 2. 1 1 1.74 1.86 1-29 1.15 3.45 2.90 2.83 2.56 3.1 1 3.75 7.95 7.30 4.31 3.96 Air Vo1ume Sampled (1) 0.129 0.098 0.165 0.201 0.186 0.372 0.176 0.339 0.351 0.303 0.534 0.744 0,368 1.28 0.281 1.02 0.927 0.972 0.900 0.778 1.78 1-74 1.87 1.69 1.94 2.93 4.43 4.21 3.41 3-71 VCM (ppm) 1.26 1.24 0.93 1.29 0.87 0.98 0.88 0.84 O.78 0.89 0.63 0.74 0.54 0.67 0.59 0.809 0.734 0.749 0.561 0.578 0.758 0.652 0.592 0.593 0.627 0.500 0.702 O.678 0.494 0.418 * PMCC's are Bendix Personnel Monitoring Collection Columns. VCMavo (ppm) 1.12 0.17 0.87 0.05 0.63 0.06 0.69 0.09 0.64 0-05 0.56 0.11 BFG44734 B.F.GOODRICH 1-000 25 RctcarcH and Development Cent Manifold Port No. J 2 3 4 5 1 2 3 4 .5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 , -)4- Corporate Environmental Service 8504-76, SR, 8-24-76 TABLE IV (continue) B. Measurements at 50% Relative Humidity Exposure Time (min) VCM Weight (yg) Air Volume Sampled (1) VCM (ppm) 15 30 11 11 60 n 11 90 11 11 11 180 11 11 11 360 11 n 11 n 0.257 0.302 0-57^ 0.362 0.324 0.796 0.672 0.965 0.260 0.837 1-19 0.803 b 1.59 1.66 1.63 1.59 1.74 1.52 2.02 3.23 3.34 2.75 3-32 2.95 6.20 4.57 3.66 4.66 b 0. M0 0.156 0.154 0.158 0.154 0.309 0.318 0.378 0.193 0.318 0.568 0.412 0.520 0.678 0.792 0.756 0.894 0.927 0.796 1.18 2.02 1.67 1.68 1.67 1.98 2.87 2.64 2.33 3.23 2.33 0-914 0.757 (I.46)a 0.896 0.823 1.01 0.827 1.00 (0.527)a 1.03 0.820 0.762 - 0.917 0.820 0.843 0.696 0.734 0.747 0.670 0.626 0.782 0.640 0.778 0.583 0.845 0.677 0.614 0.564 (porn? 0.85 0.06 0.97 0.07 0.83 0.04 O.74 0.05 0.68 0.08 0.68 0.09 BpG44l35 1-000 B.F.GOODRICH ,,, Research 26 and Development Cent Manifold Port No. 1 2 3 4 5. 1 2 3 4 .5 r 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 -15- *f Corporate Environmental Service 8504-76, SR, 8-24-76 TABLE iV (continued) C. Measurements at 32% Relative Humidity Exposure Time (min) VCM Weight (yg) Air Volume Sampled (1) VCM (ppm) VCM,W,, (pp$ 15 11 11 11 11 30 n ii n n 60 11 11 ti 90 11 n 11 180 11 11 ti 360 11 it n 11 0.397 1.29 0.826 0.543 0.330 0.905 0.971 0.491 1.02 0.479 0.330 0.944 Kl-3 0.893 1.60 2.22 1.83 1 .90 1.60 1.92 2.06 2.25 3.52 2.39 4.22 7.32 4.88 5.38 6.00 3.66 0.136 0.201 0.168 0.148 0.136 0-339 0.408 0.268 0.306 0.222 0.145 0.672 0.600 0.479 0.798 0.954 0.945 1.02 0.945 0.888 1.16 1.48 2.21 1.43 2.75 4.36 3.52 3.52 3.89 4.21 1.14 2.51 1.92 1.44 0.949 1.04 0.931 0.717 1.30 0.844 0.890 (0.550)a 0.737 0.729 0.784 0.910 0.758 0.729 0.662 0.846 0.695 0.595 0.623 0.654 0.599 0.657 0.542 0.598 0.603 (0.340)a 1-59 0.44 0.97 0.16 O.78 0.05 O.78 0.08 0.63 0.03 0.60 0.03 1 rejected this value in calculating the average ppm value. This sample was lost during the G.C. analysis. BFG44736 1-000 B.F.GOODRICH 27 Reieerch rwf Development Cent* Effect of Humidify on the Retention of VCM by Bendi'x Flasher lubeg -16FIGURE 5 Corporate Environmental Service 8504-76, SR, 8-24-76 s? c 0 CQ *-s o o >+ O\**o* o V) 0 <v d c 0 tv <0 E F-~ 0 0 s QJ S_ d </> 0 0 Q__ 1 XT 1 l_L_l o o <0 o BFG44737 o B.F.GOODRIO 1-000 28 Re*crch nd Devtlopment Cent -17- Corporate Environmental Service 8504-76, SR, 8-24-76 TABLE V Effects of Charging Time and Humidity on VCM Recovery from PMCC's. II. A. Measurements3 Hade on 5/18 19/76: Frai ~ 2.48 ng/a.u. Charge Time (min) Air Volume Samp 1ed (1iters) Concentration of VCM in Air (ppm) Concentration of VCM in ,Air (ppm) Measured by Flasher Tubes 30% R.H. 50% R.H. 92* R.H. 15 0.150 30 0.300 60 0.600 90 0.900 180 1.80 360 3.60 1.00 |1 ii 11 ii 11 1.12 1.17 0.87 0.05 0.63 0.06 0.69 0.09 0.64 0.05 0.56 0.11 0.85 0.06 0.97 0.07 0.83 0.04 0.74 0.05 0.68 0.08 0.68 + 0.09 1-59 0.44 0.97 0.16 0.78 0.05 0.78 0.08 0.63 0.03 0.60 0.03 B. Measurements3 Made on 4/28 29/76: F cal = 5.31 ng/ a.u. 180 1.80 1.00 0.59 0.13 0.60 0.08b 0.59 0.07 C. Measurements3 Made on 3/17 - 4/14/76: Fmi = 2.53 ng/a.u. 16 0.160 30 0.300 60 0.600 100 1.00 180 1.80 395 3.95 1.00 II 11 II II II 0.53 0.11c 0.43 0.05c 0.43 0.04c 0.36 0.09c 0.40 0.04c (0.20 0.07)c 0.92 0.58b 0.87 0.12b 0.75 0.05b 0.58 0.11b 0.55 0.03b 0.50 0.07b 1.20 0.16 1.05 0.10 1.05 0.10 0.77 0.08 0.79 0.10 0.79 0.06 a Each flasher tube (PMCC) value is the average of five measurements involving five separate flasher tubes. 13 These measurements were made at 70% relative humidity. C Each flasher tube (PMCC) value is the average of two sets of five measurements each, made at two separate times. BFG44738 1-000 29 B.F.GOODRIO Research end Development Cent VC M~ Collecting Efficiency of MSA Tubes vs. Exposure and HumicM -18- Corporate Environmental Service $504-76, SR, 8-24-76 FIGURE 6 -X^K- 1 ~T3 TS 3 ' q> 5 > mo *5 too o cr } C5 < CQ Iw O3 .E 51 <v W |__ X UJ CO 0O* <0 ro6 W3A BFG44739 1-000 B.F.GOODRICH 30 Reieerch end Development Ccnte -19- Corporate Environmental Service 850*4-76, SR, 8-24-76 The rapid initial decrease in ppm of VCM in the air is partly due to residual con tamination in the PMCC's which was not removed during the thermal cleaning process. This residual VCM of about 0.12 to 0.22 micrograms (yg) represents about 16 to 32 percent of the I5"niinute average VCM concentration at 92% R.H. Decreasing that 1.59 ppm value by 32 percent would give 1.08 ppm. Thus, the recovery values above 1.00 ppm could and probably do result from Initial residual amounts of VCM. This small weight of residual VCM in the PMCC at the start is proportionately signifi cant at the lowest exposure times. As the exposure time increases, the residual VCM becomes proportionately less and less significant. Recovery of VCM from MSA Tubes The partial recovery of VCM from PMCC's raised the question of what recovery the MSA tubes give. Table VI shows the effects of charging time, charging rate, and relative humidity on the percent recovery of VCM. Figure 6 summarizes and compares the results graphically. The charging conditions were different in the MSA tube case from the PMCC case. The same dynamic vapor generator and 1.00 ppm VCM-air stream were used. However, in this study each MSA tube received a nominal 5"liter charge. The exposure periods were 30, 180, and 360 minutes at 30% and 92% R.H. and 30, 60, 180, and 360 minutes at 50% R.H. Thus, the flow rates through the tubes were nominally 166, 83 28, and 14 ml per minute for charging periods of 30, 60, 180, and 360 minutes, respectively. Figure 6 shows two sets of results. The average total recovery of VCM by MSA tubes remained constant throughout 6 hours exposure at each relative humidity. The total includes the VCM recovered from the front section plus the back section of each tube. However, the level of recovered VCM varied with the relative humidity. The average total recovery was O.76 + O.OI ppm at 30% and 50% R.H. but was 0.56 ppm at 92% R.H. The scatter of data points was much wider at 92% R.H. than at 30% and 50% R.H. Table VII shows the effect of relative humidity and flow rate through the MSA tubes on VCM recovery, increasing the flow rate decreased the total VCM recovery at each relative humidity. The average total VCM decreased from 0-75 ppm at 10 ml per minute to 0.70 ppm at 166 ml per minute during the 30% R.H. exposures. In creasing the flow rates from 10 to 166 ml per minute decreases the average total VCM recovery from 0.79 to 0.7*4 ppm during the 50% R.H. exposures. We recovered an average total of 0.61 ppm VCM at 10 ml per minute and 0.48 ppm at 166 ml per minute during the 90% R.H. exposures. Figure 7 shows the rate of decrease of total recovered VCM with increasing flow rate at each relative humidity. The rate of decrease is the same at 30 and 50% R.H. The rate of decrease at 92% R.H. is 2.6 times that at 30 and 50% R.H. This graph shows the marked adverse effect of high humidity on the recovery of VCM from MSA tubes. Again, a much wider scatter of data points occurred at 92% than at 30% and 50% R.H. Figure 7 also includes the results of charging MSA tubes for 30 and 60 minutes at the same nominal flow rate (28 ml per minute) as for 180 minutes' VCM charge. Those results show the same kind of exponential decrease in VCM recovery which we observed during the PMCC study. BFG44740 1-000 B.F.GOODRICH 31 Re*erch end Development Cente Sample No. 31 32 33 28 29 30 34 35 36 16 17 ?8 7 8 9 13 14 15 4 5 6 1 2 3 10 11 12 19 20 21 22 23 24 25 26 27 -20- Corporate EnvironmentaI Service 8504-76, SR, 8-24-76 Exposure (min) 30 180 360 30 30 60 60 180 360 30 180 360 TABLE VI Summary of MSA Tube - VCM Study. 1. Flow Rate (ml/min) 189. 137. 168. Air Volume Sampled (ml) 5,670 4,110 5,040 R.H. <S) 30 32.4 27.6 15-5 5,832 4,968 2,790 30 10.5 12.4 20.7 150. 189. 169. 3.780 4,464 7,452 4,500 5,670 5,070 30 50 33-5 27-7 23-9 1,005 831 717 50 85-7 92.6 81.3 5,142 5,556 4,878 50 21.9 31-3 32.1 1,314 1,878 1,926 50 27-8 21.3 22.6 5,004 3,834 4,068 50 11.1 13.2 13.1 152. 195212. 3,996 '<,752 4,716 4,560 5,850 6,360 50 90 25.7 27.5 23.4 4,626 4,950 4,212 90 12.7 13.0 11.2 4,572 4,680 4,032 90 BFG44741 Total VCM Amount (yq) (DDm) 10.2 7-36 9-15 11.1 9.35 5.54 0.70 0.70 0-71 0.74 0.74 0.78 7-49 8.80 14.6 8.26 10.6 9.88 0.78 0.77 0.71 0.72 0.73 0.76 2.31 1-75 1.62 0.90 0.82 0.88 10.4 11.1 9-33 0.79 0.78 0.75 2.59 3.53 3-99 10.5 7.47 8.04 7.50 0.77 0.73 0.81 0.82 0.76 0.77 0.73 0.75 0.76 7.05 6.49 5-93 0.60 0.43 0.36 7M 6.26 7.48 0.63 0.49 0.70 6.60 6.43 7.08 0.56 0.54 0.69 B.F.GOODRIC1- 1-000 32 Restweh md Development Cnt< Sample No. 31 32 33 28 29 30 34 35 36 16 17 18 7 8 9 13 1*4 15 4 5 6 1 2 3 10 M 12 19 20 21 22 23 2*4 25 26 27 -21- Corporate Environmental Service 8504-76, SR, 8-2*4-76 TABLE VI I Investigation of VCM Breakthrough in MSA Tubes Exposure (min) 30 Flow Rate (ml/min) 189. 137. 168. Air Volume Sampled (ml) 5,670 4,110 5,040 R.H. (*> 30 VCM Amount (uo) Front Back % 8.542 6.477 8.013 1,618 0.8840 1.141 18.9 13.6 14.2 180 32.*4 5,832 30 9-788 1.271 13-0 27.6 4,968 8.678 0.6764 7.8 15.5 2,790 4.916 0.6250 12.7 360 10.5 3,780 30 6.518 0.9688 14.9 12.*4 4,464 8.116 0.6867 8.5 20-7 7,452 13.14 1.456 11.1 30 150. 189. 169. 4,500 5,670 5,070 50 6.521 1.739 26.7 8.641 1.966 22.8 8.402 1.476 17-6 30 33-5 1,005 50 1.894 0.417 22.0 27.7 831 1.463 0.289 19.8 23-9 717 I.309 0.3H 23.8 60 85.7 5,142 50 8.619 1.804 20.9 92.6 5,556 8.818 2.321 26.3 81.3 4,878 8.031 1.301 16.2 60 21.9 1.314 50 2.588 - - 31.3 1 ,878 3.011 0.515 17-1 32.1 1 ,926 3-163 0.824 26. 1 180 27*8 5,004 50 9.920 0.603 6.1 21.3 3,834 7.473 - - 22.6 4,068 8.042 - - 360 11.1 3,996 50 6.322 1.173 18.6 13.2 *<,752 7.616 1.454 19.1 13.1 *1,716 7-590 1.563 20.6 30 152. 195. 212. 4,560 5,850 6,360 90 4.119 2.931 71.2 3.883 2.608 67-2 3.803 2.127 55.9 180 25.7 4,626 90 5.015 2.477 49-4 27.5 *1,950 3.994 2.261 56.6 23.*1 4,212 4.390 3*090 70.4 360 12.7 *4.572 90 4.093 2.508 61.3 13.0 4,680 3.927 2.503 63.7 11.2 *4,032 4.159 2.924 70.3 BFG44742 B.F.GOODRICH 1-000 33 Research and Devdortmeni Cent* -22- FIGURE 7 Corporate Environmental Service 850W6. SR, 8-2*1-76 VCM- Collecting Efficiency of MSA lutes vs. Row Rale & Relative Humidif - U40 BFG44743 1-000 B.F.GOODRIO 34 Rocarch nd Development Cent- -23- Corporate Environmental Service 8504-76, SR, 8-24-76 Search for the Missing VCM Our results up to this point led to a search for the missing VCM. By the missing VCM we mean the difference between the amount of VCM charged and the amount recovered as found by analysis. First, we examined the possibility that the G.C. calibration factor had changed with out our knowledge. This thought arose in part from considering the results in Table VC. The VCM values for each charging period seemed to increase by factors of two to three as the relative humidity increased from 30 to 32% R.H. However, we had analyzed one humidity set of samples at a time. Considerable time elapsed be tween the measurements of each two sets. During these interim periods we used the same G.C. column to analyze other pollutants. Contrary to our usual custom we neglected to recalibrate before each humidity set of samples in Table VC. Table VII! permits a check of past VCM-Porapak Q.S column calibration factors with time and hence with column history. The calibration factor on 4/20/76 was 47$ larger than on 3/1/76. And the factor on 4/26/76 was 2.2 times as large as on 3/1/76. Thus, the differences among the humidity set values in Table VC could have been due to unsuspected changes in calibration factor. However, we calibrated immediately be fore all sets of measurements in Tables VA and VB. The agreement within and between the latter two tables of values is good. So, we had to look further for an explan ation of the missing VCM. TABLE VIII Variations in Calibration Factors for G.C. Analysis of VCM 1ibration Date 1/29/76 3/01/76 4/20/76 4/26/76 4/29/76 5/11/76 Calibration Factor (nq/area unit) 5 33 : *>3 1 '1 > * r ; *; . *+ " Notebook 447 Paqe 4891 4923 Notebook 469 Paqe 36 40 42 52 We next studied the possibility of loss >' VC* by transmission completely through the PMCC. We charged two PMCC's in ser.ps at each of the five manifold ports. Table IX shows the results and compares with the results from single PMCC's. These values all resulted from charging na1ly for 180 minutes at 10 ml per minute and 30$ R.H. with a 1.00 ppm vc^-air mixture. Except at port no. 4 the second PMCC in series collected less than 2 percent as much VCM as the first tube did. That Is, we recovered as much VCM (in ppm) from the first tube in each series as from the single PMCC at each port in the comparison study. These results seemed to rule out transmission as the cause of VCM loss. 1-000 B.F.GOODRICH 35 Research and Development Centc -24- Corporate Environmental Service 850*1-76, SR, 8-2*4-76 ManIfold Port No. 1 2 3 4 5 1 2 3 A 5 TABLE IX Search for VCM Breakthrough During 180 Minutes at 30% R.H. A. Measurements Involving Two Flasher Tubes in Series VCM Weight (ua)- Front Back Total Air Volume Sampled (1) VCM (ppm) 2.34 0*361 2.70 1.56 0.68 . 2.21 0.0311 2.30 1.82 0.49 A.55 5*00 0.0776 0.102 4.63 5-10 2.68 2.30 0.68 0.85 A.55 0.0746 A.62 2.68 0.67 B. Measurements Involving Single Bendix Flasher Tubes - - 3*27 2.09 0.61 - - 2.36 1*91 0.48 - - 4.63 2.50 0.72 - - 1.20 1*33 0.35 - - 6.71 3.74 0.70 VCMa (ppm) 0.68 0.07 0.57 0.11 BFG44745 B.F.eOODRIO 1-000 36 Research end Development Cent -25- Corporate Environmental Service 8504-76, SR, 8-2*1-76 Our third thought was that the BFG Chemical Company plant laboratories calibrate their G.C.'s with standard PMCC's. They prepare the standard PMCC's by charging them directly with VCM of certified concentration in air from a cylinder under pressure. We calibrate our G.C. by injecting aliquots (usually 5 microliters) of successive standard solutions of VCM in carbon disulfide. So, we charged PMCC's as the plant laboratories do by flowing the certified VCM-air mixture directly through each PMCC, one at a time. We flowed both a 53 ppm certified mixture for about 6 minutes and a 1.09 ppm mixture for 6 hours at 10 ml per minute. Table X shows the results for comparison with the results from charging PMCC's by vacuum pump suction. The percent recoveries of VCM averaged 59*3%. 633t ar*d 67*10% respectively. The three sets of measurements gave the same percent recovery within experimental error. Therefore, the method of charging does -not account for the loss of VCM. PMCC No. 12 31 32 68 T-17 TABLE X Recovery of VCM from PMCC's Charged Directly vs. by Suction A. PMCC' s Charged from a 53 ppm Cylinder for 6 minutes Flow Rate (ml/minute) Amount of VCM per Tube Charged (up) Found (up) Found (ppm) 19.0 18.5 20.8 l*i. *1 *i0.5 15-4 15.0 16.9 11.7 33-0 9.00 9.79 8.99 7.07 19-0 31 35 28 32 30 Recovery {%) 58 65 53 60 58 B. PMCC's Charped from a 1.09 ppm Cylinder for 360 minutes 2 10.1 3 1.78 22 8.80 10.3 1.81 8.97 6.62 1.20 5.28 0.70 0.72 0.64 64 66 59 C. PMCC'*> Charped by Suction : 1.00 ppm for 360 minutes 57 7.98 61 7.33 63 6.48 6*t 8.96 7.30 6.70 5.93 8.96 6.20 4.57 3*66 4.66 0.85 0.68 0.62 0.52 85 68 62 52 Our fourth attempt to explain that loss i nvolved charging 12 PMCC's du ring about 6 minutes' time. Then, we flowed room air through 8 of the tubes for 360 minutes. The results of G.C. analyses of the two sets of PMCC's appear in Table XI. The percent recoveries of VCM averaged 53-3% before and 563% after the six-hour flow of VCM-free air. Again, the two sets of measurements agree within experimental error. Therefore, six hours flow of VCM-free air through the VCM-charged PMCC's did not remove any measurable amount of VCM. BFG44746 1-000 37 B.F. GOODRICH Research nd Development Cenlc -26- Corporate Environmental Service 8504-76, SR, 8-24-76 PMCC No. 12 31 32 68 T-17 TABLE XI Effect of 6-Hour Airflow through VCM-Charged PMCC's A. Without Airflow after the 6-Minute VCM-Air Charging Flow Rate (ml/minute) Amount of VCM per Tube Charged (pq) Found (pq) Found (ppm) 19.0 18.5 20.8 14.4 40.5 15.4 15.0 16.9 11.7 33.0 9.00 9.79 8.99 7.07 19.0 31.0 34.6 28.2 32.0 30.5 Recovery (%) 58 65 53 60 58 B. With 6-Hours' Airflow after the 6-Minute VCM--Air Charqinq 10 19 77 81 3-11 T-5 T-18 10.1 16.3 16.4 11.0 3 a.9 37.0 31.2 8.18 13.2 13.3 8.90 28.4 30.1 25-4 4.72 3.62 7.88 5.20 9.16 14.9 13-8 31 15 31 31 17 26 29 a .These values are omitted on a statistical basis. 58a 27 59 32a 50 54 Our fifth attempt to discover the cause of the missing VCM involved desorbing the VCM from the PMCC carbon with carbon disulfide. We modified the NIOSH-approved procedure for MSA tubes only by using 3 ml per sample instead of i ml. We removed all the carbon from each exposed PMCC and submerged it in 3 mi of carbon disulfide. After half an hour of shaking each sample, we injected three onemicroliter aliquots into the G.C. The purpose of this approach was to see whether carbon disulfide desorption would recover VCM from the carbon more efficiently than thermal flashing does. Table XII provides the answer. The VCM recovery values averaged 646&. The close agreement of this average recovery value to the various average recovery values from thermal desorption indicates that thermal flashing and the Bendix flasher are not the cause of the missing VCM. TABLE XI I Desorption of VCM from PMCC carbon with Carbon Disulfide PMCC No. Flow Rate (ml/minute) Amount of VCM per Tube Charqed (pq) Found (pg) Found (ppm) Recovery <*) 13 11.4 22 23.8 54 5.66 55 9.93 58 25.4 10.4 21.8 5.18 9-07 23.2 5.64 12.8 3-36 6.24 16.9 BFG44747 0.54 0.59 0.65 0.69 0.73 1-000 38 54 59 65 69 73 B.F.GOODRici Rerch nd Development Cen -27- Corporate Environmental Service 8504-76, SR, 8-24-76 At this point we could only conclude that the missing VCM must remain on the PMCC carbon in a form or manner which resists removal by thermal flashing or carbon disulfide desorption. Determining that form or manner is beyond the practical scope of Project 8504 (Method Development). A practical question did remain though: does the MSA tube also show this VCM loss? We charged MSA tubes just as we had charged the PMCC's with VCM. The results of the analyses appear In Table XIII. The three groups of samples in order represent exposures at 30%, 50%, and 92% R.H. The nominal volume of 1.00 ppm VCM-air mixture which we sampled was 5.0 liters in every case. The variations in flow rate resulted from Inherent differences in the activated carbon packing in each tube and In the diameter of the openings in the broken tips. The MSA tubes gave average recoveries of 70 to 79% at 30% and 50% R.H. compared with 56 to 97% for the PMCC tubes. As Figures-5 and 6 show at exposure times of 90 to 360 minutes at 30% and 50% R.H., the MSA tubes gave larger percent recoveries than the PMCC's did. In general, both the MSA tubes and the PMCC's showed large deficiencies of VCM recovery. TABLE XIII VCM-Collecting Efficiency of MSA Tubes MSA Tubes Flow Rate (ml/minute) Amount of VCM per Tube Charqed ( g) Found ( q) Found (ppm) Recovery (%)a 31~33b 28-30" 34-36e 137-189 16- 32 10-21 10.4 -14.4 7.09-14.8 9.6O-I8.9 7.36-10.2 7.49-14.6 0.70-0.71 0.74-0.78 0.77-0.78 70 75 77 16-I8b 13-15 1- 3d 10-12 150-189 81- 93 21- 28 11- 13 11.4 -14.4 12.4 -14.1 9.74-12.1 10.1 -12.1 8.26-10.6 9.33-U.l 7.47-10.5 7-50- 9-15 0.72-0.76 0.75-0.79 0.76-0.82 0.73-0.76 74 78 79 75 13--21b 22-24d 25-27e 152-212 23- 28 11- 13 11.6 -16.2 10.7 -12.6 10.2 -11.9 5.93- 7-05 6.26- 7.49 6.43- 7.08 0.36-0.60 0.49-0.70 0.54-0.69 47 61 60 a average value b 30* exposure C 60' exposure ^ 180* exposure e 360' exposure ,f J yr\ Paul M. Zakriski IZUJl&.IUjL Richard D. Hardest-^ BFG44748 jt> B.F. GOODRICH J-000 39 Research and Development Cent -28- Corporate Environmental Service 8504-76, SR, 8-2*1-76 REFERENCES 1. U.S. Department of Health, Education, and Welfare, National Institute for Occupational Safety and Health, Physical and Chemical Analysis Branch, Physical and Chemical Analytical Method #178, Vinyl Chloride In Air, September 3, 197*1. 2. B.F.Goodrich Chemical Company Standard Test Procedure, Bendix Flasher Procedure. 3- R. D. Hardesty, P. M. Zakriskl, and J. M. Whitney, VC1 Personnel Monitoring Cross Check Program, February 9, 1976. 4. (Letter of 2/17/76 from S. C. Tenhaeff to D. T. Wright), Maintenance Problem with Hewlett-Packard 5711 Gas Chromatograph/Flasher System. 5. (Letter from J. A. Baclawski to J. M. Whitney dated 2/25/76), VC1 Personnel Monitoring Cross Check - Rationalization of Low Recovery. BFG44749 1-000 4V0 B.F.GOODRIC R**rch end Development Cen -29- Corporate Environmental Service 8504-76, SR, 8-24-76 APPENDIX I Description of the Dynamic Vapor Generator The dynamic vapor generator has four integrated functions by design. It regulates the flow of one or more vapors or gases at part per thousand levels into a mixing chamber. It also regulates the flow of purified dried air into the mixing chamber. By humidifying the latter diluting air it can provide controlled relative humidity. The vapor generator mixes the vapor or gas of interest with the dilution air to give part-per-mi11 ion concentrations of the vapor or gas in dry or humidified air. The vapor generator design permits the use of diluting gases, other than air. it also permits the simultaneous mixing and dilution of several vapors or gases with the diluting gas. Figure 3 (in this Appendix) illustrates the dynamic vapor generator. The Century Lab-Crest Mark 111 flowmeter (A) regulates the flow of VCM in this study at 9.0 ml per minute. The regulated flow of VCM at a standard concentration (508 ppm in air in this case) passes through the thermometer well (B) into the mixing chamber (C). The mixing chamber is a modified Greenburg-Smith 500 ml impinger. The diluting air (or other gas) passes through the oi1-and-metal particle-filter (D), the silica gel drying chamber (E), and the activated carbon purifying chamber (F). The purified dried air (at 30t R.H.) then flows through rotameters (G) and/or (1) into reservoirs (H) and/or (J). A resistance heater tape surrounds the bottom fourth of reservoir (H). Heating water in reservoir (H) is necessary in order to achieve 32% R.H. In this VCM study we adjusted rotameters (G) and (l) to control the humidity and give a combined diluting air flow rate of 4.6 liters per minute. Mixing the diluting air at 4.6 1pm with the 508 ppm VCM-air mixture gave a 511-to1 dilution and a 1.00 ppm VCM-air mixture. The diluting air and the 508 ppm VCMair mixture mix turbulently in chamber (C). The 1.00 ppm mixture then flows through tube (K) into supply manifold (L) and out through the five ports. Five personal sampling tubes are attached to the ports of the vacuum manifold (M) with short Tygon sleeves. Tygon tube (N) connects the vacuum manifold through a regulator valve to a vacuum pump. We insert the tubes to be charged inside the five ports of the supply manifold (L). The inlets of the tubes are positioned along the vertical axis of that manifold. The regulator valve on the vacuum pump controls the flow rate of the air through the monitor tubes connected to the vacuum manifold ports. We pre-measure the airflow rates through the monitor tubes connected to the ports. Then without disturbing the connections we proceed to charge the tubes with the VCM-air mixture. Five duplicate sample tubes are the result in each series of exposures. The samples are analyzed by gas chromatography in the usual manner. BFG44T50 1-000 41 B.F.GOODRICH Rscch and Development Cent* -30- Corporate Environmental Service 850W6, SR, 8-24-76 APPENDIX I Dynamic Vapor Generator BFG44751 1-000 42