Document z50kJNLN0xZD2Zay8qB2qMNg

Ayer et aL A COMPARISON OF IMPINGER AND MEMBRANE FILTER TECHNIQUES FOR EVALUATING AIR SAMPLES IN ASBESTOS PLANTS A study of the present being carried out by the I Health Service,* in coopera Howard E. Ayer, Jeremiah R. Lynch, Julius H. Fanney Departments. This study i Division of Occupational Health, Public Health Service Department of Health, Education and Welfare Cincimtati, Ohio membrane filters and other size and other characteristli only one aspect of the stud; C:.\ Introduction Although data had been presented by a number of other authors,*iX,m! the more definitive studies of health hazards in the United States asbestofLL j industry are probably those of the Pennsylvania Department of Labo^ ~ \ and Industry in 1935s and the U. S. Public Health Service and the North,;-- Carolina State Board of Health in 1937.* These studies of asbestos plants, 1? n xj sampling to fiber counts fix microscope objective for the It should be emphasized t pinger counts are develope being proposed as a substi presented, rather, (1) as : pinger counts to fiber coun dication of the relationship correlated exposure to asbestos dust with the incidence of ;isbestosis among^''*' workers. The exposure was evaluated by the same instruments and methods^- o membrane filters, and (3; : of better information, to e.- used in studies of other dusty trades (e.g,, grauite finishing, cement manu--, dust exposure. factoring).1 The sampling instrument used was the Greeuburg-Smith^ , impinger, and samples obtained were counted in the sampling liquid using.,-.;, a 1 mm. deep counting cell, allowing the particles to settle for at least 20D," minutes and counting the settled particles with a microscope equippedI with a 16 mm. lOx objective. In the studies which correlated dust exposure^j with asbestosis it was found that the incidence and severity of asbestosis -- increased with increasing dust exposure, and that those workers exposed CD to dust concentrations of less than 5 million particles per cubic foot (mppcf) did not develop asbestosis even after long-continued exposure. Because epidemiologic studies m the U. S. are based on the impinger method, this method has been adopted as a standard by those concerned with prevention of asbestosis. Using the impinger to evaluate effectiveness of environmental controls, it has'been possible to notably reduce the inci dence and severity of asbestosis. In Great Britain a customary procedure in monitoring air in asbestos processing plants has been to count only the fibers collected in air samples. The thermal precipitator, Owens jet sampler and other instruments have been used for sampling, but the roost common air sampling method at the La. CD Impinger Method foi The impinger method \va in silicosis-producing occup. of quartz and particles of larger. The associated cour. lOx objective to examine pa or 1 mm. depth), is capat minimum size. The impinp essentially the same as me sequently by the American gienists,* the American F Textile Institute.'1 In gene sampling medium, and eiti may be used. In the study r as the sampling medium, . pinger. When this method present time is probably membrane filtration. After sampling, the filter is treated to render it transparent, and fibers are counted using a microscope equipped with a 2 ram. or 4 nun. objective. Fiber concentrations determined 1 by these methods have not been compared to criteria developed in the 3 U.S.A, because these fiber-counting methods do not yield results directly comparable to those obtained by the impinger sampling and counting method. plants, few asbestos fibers doubtedly been observed b when all particles with a le: as fibers, the fiber count i.s pinger dust count. Elector methods reveal that most diameter -- hundredths to t PP<S?qWT WA N0T A RC0RD UfOF IT^ in rcS7R ES* 'NC- 0,0 N0T COME FROM 3271 BB 0020779~7 ER AND MEMBRANE kLUATING AIR SAMPLES PLANTS vnch, Julius IL Fanney . Public Health Service ration and Welfare Ihio ON a number of other authors,**1 rds in the United States asbestos nsylvania Department of Labor lie Health Service and the North These studies of asbestos plants the incidence of asbestosis amoug he same instruments and methods ., g- *e finishing, cement manu- s the Greenburg-Smith the sampling liquid using particles to settle for at least 20 :les with a microscope equipped is which correlated dust exposure idence and severity of asbestosis and that those workers exposed million particles per cubic foot i after long-continued exposure. . S- are based on the impinger s a standard by those concerned mpinger to evaluate effectiveness ssible to notably reduce the inci- r* in monitoring air in asbestos die fibers collected in air samples, tpler and other instruments have muon air sampling method at the tion. After sampling, the filter is "3 are counted using a microscope Fiber concentrations determined red to criteria developed in the hods do not yield results directly apinger sampling and counting Ayer et ctLil PrG FftSS A study of the present environment in asbestos processing plants is being carried out by the Division of Occupational Health, U. S. Public Health Service,* in cooperation with the various concerned State Health Departments. This study uses the impinger technique, air samples on membrane filters and other special samples for evaluating respirability, size and other characteristics of the asbestos. This paper is concerned with only one aspect of the study, the relationship of dust counts by impinger sampling to fiber counts from membrane filters, using a higher resolution microscope objective for the latter. It should be emphasized that, although ratios of membrane filter to im pinger counts are developed herein, the membrane filter method is not being proposed as a substitute for the impinger method. The ratios arc presented, rather, (1) as a rough indication of the relationship of im pinger counts to fiber counts reported from abroad, (2) to give some in dication of the relationship for fiber counts from personal samplers using membrane filters, and (3) as an index which may be used, in the absence of better information, to estimate the proportion of asbestos in a mixed dust exposure. Methods Impinger Method for Sampling and Counting Asbestos Dust The impinger method was developed for the evaluation of atmospheres in silicosis-producing occupations. The impinger is a very efficient collector of quartz and particles of a similar density, M micron diameter and larger. The associated counting method, which uses a microscope with a lOx objective to examine particles which have settled in a shallow cell (0.1 or 1 mm. depth), is capable of detecting quartz particles to this same minimum size. The impinger method as used in the present study was essentially the same as methods described by PHS Bulletin 217 and sub sequently by the American Conference of Governmental Industrial Hy gienists,* the American Public Health Aasociation,*" and the Asbestos Textile Institute.11 In genera], either water or alcohol may be used as a sampling medium, and either the Greenburg-Smith or midget impinger may be used. In the study reported here 95 per cent ethyl alcohol was used as the sampling medium, and samples were taken with the midget im. pinger. When this method is applied to air sampling -in asbestos textile ;-* plants, few asbestos fibers are seen in the samples, a fact which has un doubtedly been observed by all those who have used the method. Even when all particles with a length-to-width ratio of 3 or greater are counted as fibers, the fiber count is usually less than 10 per cent of the total im pinger dust count. Electron micrographs of samples collected by other methods reveal tliat most airborne asbestos fibers are very small in diameter -- hundredths to tenths of a micron -- even when many microns THIS DOCUMENT WAS NOT A RECORD OF _____________________ PPG INDUSTRIES,.INC. DID NOT COME FROM TPS FILES AND CANNOT BE AUTHENTICAI ED BY PPG INDUSTRIES. INC. | BB 002Q780 I  &* '-L-v^vV'^ -- * --- Iktev;-. TZp:-^ ~".,S|.. ^ --'------ 276 Annals New York Academy of Sciences Ayer et al.: long. Fibers of such small diameter may not be collected by the impinger; if collected many of them would not settle to the bottom of the countine cell in the allotted settling time, and those that did settle would usually not be visible by the microscopic technique employed. Thus it is not sur prising that fiber counts with a 1G mm. objective are low. Correlations between dust concentrations from impinger counts And concentrations of asbestos fibers as determined by other methods are not now available. Membrane Filter Sampling and Counting S-fo total fibers and fibers longer of fibers longer than 5 p v. As implied in the discuss* on the filter may be readily better resolution so that srr fiber which b routinely cou yet been determined, but it ir without difficulty. For the environmental study of the asbestos textile industry now under o Lj way, it was desired to have another routine sampling and counting method for asbestos fibers that could also be used with the personal samplers that have recently become available. A medium for dust sampling that has come into wide use in recent years is the membrane filter. Membrane filters will collect particulates as small as 0.1 p. diameter with practically 100 per cent efficiency.11 The particles are collected on the surface" and the filter may be examined directly with incident light, rendered transparent and examined by high resolution transmitted light microscopy, or the c;- oo cze particles may be transferred to an electron microscope grid and examined by electron microscopy." c/2 uj Membrane fdter sampling was thus selected as one of our air sampling methods for this study. The sampling medium used was Miliipore Type AA, a membrane with a stated pore size of 0.8 ft, for which surface collec o * r tion of a very high proportion of the finest particulates had been reported. The filter holder used was a`plastic, factory-loaded holder containing a 37-mm. diameter filter with an open area of approximately 850 mm.1 The oo zz sampling rate employed was 1*5 or 2 l./mim, giving a face velocity on the order of 3 cim/sec. In counting, it was desired to see the smallest diameter fibers, but for routine counting the electron microscope was obviously unsuitable and it was felt that an oil immersion objective would be impractical. A 4 mm. f43x; phase contrast objective was therefore selected as that which would Ficuss 1. Fiber concentrat centrations by midget impinge give the highest resolution compatible with routine examinations of the -Sa filter. With the 4 mm. objective the limit of resolution is % micron; the limit of visibility may be somewhat smaller. For counting, a segment was cut from the filter, placed face up on a microscope slide, rendered trans The samples reported up ing primarily asbestos tex: impinger and a membrane parent by a mixture of 50 per cent dimethyl phthalate and 50 per cent a few centimeters of one S diethyl oxylate, and a eoverslip placed on tup. A Porton eyepiece graticule i was used to define the counting field and to assist in making length judg ments of fibers. Phase contrast illumination was used for all counts. i Three modes of counting were used; (1) counting all fibers with length \ to width ratios of 3 or greater, (2) counting only those fibers longer than 5 microns, and f3) counting oniy those fibers longer than 10 p.- Counts of .5 taken with both samplers and keeping the sampler as these samples were of 20were taken in fixed locatior Because optical systems h be reemphasized that cour. iMIb DOCUMENT WAS NOT A RECORD OF IT^O INC. DID NOT COME FROM j BB 0020781 ... - rH,wniiu WUMINUI bfc. AUTHENTICATED------------ nr of Sciences it be collected by the impinger; to the bottom of the counting i that did settle would usually s employed. Thus it is not suribjeetive are low: Correlations r counts and concentrations of tods are not nmv available. and Counting tos textile industry now under sampling and counting method nth the personal samplers that a for dust sampling that has mbrane filter. Membrane filters diameter with practically 100 ect*d on the surface'3 and the uit light, rendered transparent ittm' '^ht microscopy, or the m^^ ope grid and examined Ayer et aL: I PPG HIES total fibers and fibers longer than 10 /a were made on all samples: counts of fibers longer than 5 /a were made on only a fraction of the samples. As implied in the discussion on impinger sampling, the count of fibers on the filter may be readily increased by using a microscopic method with better resolution so that smaller diameter fibers are seen. The minimum fiber which is routinely counted by the method described above has not yet been determined, but it is believed that J/ /a diameter fibers are counted without difficulty. tenu one of our air sampling ium used was Millipore* Type OJSfor which surface collecparticulates had been reported, iry-loaded holder containing a if approximately 850 mm.1 The u, giving a face velocity on the naffest diameter fibers, but for was obviously unsuitable and would be impractical. A 4 mm. re selected as that which would b routin examinations of the. if resolution is *4 micron; the *- For counting, a segment was moscope slide, rendered transhyi phthaiate and 50 per cent ijx A Porton eyepiece graticule > assist in making length judgn was used for aH counts, counting all fibers with length goniy those fibers longer than rs longer than 10 p. Counts of MILLIONS OF PARTICLES PER CUBIC FOOT (IMPINGER AT 100*) Ficukb 2. Fiber concentrations by membrane filter (430 x ) vs. particle con centrations by midget impinger (100 x ) fiber preparation. re Sanrpits for Comparison The samples reported upon in this paper were taken in plants produc ing primarily asbestos textiles. Each pair of samples included a midget impinger and a membrane filter sample taken simultaneously and within a few centimeters of one another. The "breathing zone" samples were taken with both samplers held in the same hund, following the worker' and keeping the sampler as close to the worker's head as possible. Most of these samples were of 20-minute duration- "General air" sample pairs were taken in fixed locations, usually for a duration of 30 minutes. Because optical systems having different resolutions were used, itshould bo reemphasized that counts by the two methods are not directly com- THIS pJOuCGUMENT WAS NOT A RECORD OF I PPG T E5USTRIES, INC. DID NOT COME FROM U?iL----------------- ITS FtL.S AND CANNOT BE AUTHENTICATED BY PFS^INDUSTRIES, INC. 3274 F - t 7~~ ' 4 i 278 Annals New York Academy of Sciences " ISO oX 100 INDIVIDUAL plants a---------- * Ayer et a- undoubtedly would have : particulates, but the degn counts made on membran mental study thus cannot studies. Uez<c! 80 * U1J so 3 /./ K3 It was presumed that t membrane fiber counts m between plants. To corap selected: fiber preparatio Operations such as windin: cw 40 thus excluded. From thes available, with at least fivt .O to 25 pairs for an operatl Regression lines and co O' plant and operation. Corrt varied from small negativt rather low. Least squares 0 2 4 6 S 10 12 14 IS MILLIONS OF PARTICLES PER CUBIC FOOT (IMPINGER AT I00>) O ; CD zero to very large numbe FiCVRE 2. Fiber concentrations by membrane filter (450 x ) vs. particle con centrations by midget impinger (100 x ) carding1. parable. To emphasize the fact that fiber counts with the 4 mm. objective 1-- !-- CD O AjVMD ' r n tzo are not equivalent to impinger dust counts with the 1G mm. objective, the former are reported in the units of fibers per ml., while impinger particle counts are presented in the usual U. S. unit of millions of particles per cubic foot (1 mppcf -- 35.3 particles per ml.). Where the same optical system is used one may expect more or less comparable counts from the two sampling systems. For example, direct comparisons were made on a relatively few short duration membrane filters with light dust deposits by counting with the same 16 mm. (lOx) objective used for the impinger samples. For these sample pairs the fiber counts on the membrane filters averaged about the same as the impinger fiber counts, and particle counts averaged about one-half the impinger particle counts. A point which should be mentioned is that any comparisons between impinger dust counts and fiber concentrations as determined by some r. other method can only reflect present conditions. Although the manufac turing processes and equipment used now are essentially the same as those used in the 1930's, the quality'of the fiber is said to have improved greatly in the past 10 years -- the proportion of the asbestos received at 0 24 MILLIONS OF PART the textile plant which is "dust" is much smaller than previously. This Ficure 3. Fiber concentre, eentrationi by midget imping THIS DOCUMENT WAS NOT A RECORD -OF ' PPG INDUSTRIES, INC. DID NOT COME FROM IT'S FILES AND CANNOT BE AUTHENTICAi ED BY PPG INDUSTRIES, INC. 1 BB 0020783 J 3275 of Sciences -40IVI0UAL 0--------- 0 PLANTS K-----------------`< 12 ia is MPINCER AT 100*} t* j x) vs. particle coa Ayer et al.: Comparison of Techniques 279 undoubtedly would have some effect on the ratio of fibers to nonfibrous particulates, but the degree of effect cannot readily be determined. Fiber Cm.E,ON- 1 counts made on membnme filter samples taken in the ongoing environ- 'j mental study thus cannot be related directly to the past epidemiologic ,__ -- studies. TT'. ' -'l ~ i" Analysis of Data ;! C... It was presumed that the ratio between impinger particle counts and membrane fiber counts might differ between operations and also differ between plants. To compare these ratios, five main operations were selected: fiber preimration, carding,spinning, twisting, and weaving. Operations such as winding, braiding, fiber treatment, and inspection were thus excluded. From these five major operations 230 sample pairs were available, with at least five pairs for each operation in each plant, and up to 25 pairs for an operation in a larger plant. Regression lines and correlation coefficients were determined for each plant and operation. Correlation coefficients, corrected for size of sample, varied from small negative values to 0.95, with the majority of them being rather low. Least squares liues of best fit had slopes varying from almost zero to very large numbers, and often passed through the axes at con- -. *" J ~ "" ' ( 'i f/E! V' /> L _J '.0 V,.' ts with the 4 mm. objective rh the 16 m objective, the mL, while impinger particle of millions of particles per Where the same optical runparable counts from the mparisons were made on a with light dust deposits by tve used for the impinger ts on the membrane filters counts, and particle-counts ousts. any comparisons l*etween a as determined by some ns. Although the manuface essentially the same as r is said to have improved >{ the asbestos received at Her than previously. This Ficuke 3. Fiber concentrations by membrane filter (430 x ) vs. particle con centrations by midget impinger (100 x) spinning. this Document was not a record of PPG IftoEtSTRIES, INC. DID NOT COME FROM IT'S Fli^-AND CANNOT BE AUTHENTICATED BY PR*,INDUSTRIES* INC. i B8 00 2Q7Q4 J ' &27S 280 Annals New York Academy of Sciences CO til UJ Ayer et the latter. These results presented in FIGURES 6 a particles and the count; concentrations. Percent; samples with lower con sented as cumulative per CD Q_ CL- cj cO a 3: tu * :r When results of the tt against each other, the rz dust counts may be deter in the same plant, and dif in different plants. Fjgu mL by membrane filter ; preparation in four plan* from each plant Fiber < of best fit vary from 4 \ for carding in four plan ml. Figure 3 shows pair Figure 4. Fiber concentrations by membrane filter (4G0 x ) v*. particle con centrations by mullet impinger (100 x ) twisting. siderable distances from the origin. If one pictures, however, an operation where a concentration which is fairly uniform is sampled by two methods, ii each of which has considerable variability, one encounters a situation in which the coefficient of correlation may be very small and a regression line with almost any slope might be found. In many of the operations this situ ation was believed to exist. To make our analysis of the data relevant to the known situation therefore, the additional condition wits imposed that when the dust count was zero the fiber count would also be zero. A "least squares" line through the origin was then determined graphically by trial and error. Such plots were made for total fibers by membrane vs. particles i by impinger and fibers longer than 10 by membrane vs. impinger par ticles, with results as shown in the following section. A second method of analyzing the data was undertaken because in our environmental surveys of asbestos textile plants a number of membrane filter samples were taken without companion impinger samples. To obtain meaningful data from the 230 sample pairs available, all plants were pooled, and the first two operations (fiber preparation and carding) were i. i considered together as were the latter three operations (spinning, twisting, and weaving). This gave 98 sample pairs in the former group and 132 in Figure 5. Fiber concent: centration* by midget iropii THIS DOCUMENT WAS NOT A RECORD PPG INDUSTRIES, INC. DID NOT COME FR< IT'S FILES AND CANNOT BE AUTHENTlCAi BY PPG INDUSTRIES, INC. | BB 0020735 1 3277 i Tty of Sciences Ayer et aL: Comparison of Techniques 2S1 the latter. These results are discussed in the following section and are presented in figures 6 and 7. From .these two sets of data the counts for particles and the counts for fibers "were separated according to specific concentrations. Percentages were computed representing the number of "samples with lower concentrations than those specified. These are prep GO UJ sented as cumulative percentage curves. o _,,l Results - 'n* LjZ Total Fibers `/J When results of the two methods of sampling and counting are plotted f ' -> cr; CTM against each other, the ratios between membrane filter counts and impinger -- dust counts may be determined. These ratios differ for different operations - 'X CL- in the same plant, and differences are also noted between similar operations :0 o in different plants. Figure 1 presents the comparison of total fibers per ml. by membrane filter sample vs. the impinger count in mppcf for fiber u.t. preparation in four plants. The line of best fit is indicated for the samples from each plant. Fiber concentration ratios to 1 mppcf from these lines of best fit vary from 4 to 11 fibers per ml. FIGURE 2 shows similar data for carding in four plants, 1 mppcf ratios range from C to 27 filers per rot FIGURE 3 shows pairs from spinning, with ratios from 3 to 13 fibers ne filter (450 x) vs. particle con ns- pictures, however, an operation arm is sampled by two methods, <me encounters a situation in very small and a regression line '.any of the operations this situmalysis of the data relevant to nal condition was imposed that nt would also be zero. A "least determined graphically by trial ibers by membrane vs. particles jy membrane vs. impinger par- Si; ag section. was undertaken because in our plants a number of membrane on impinger samples. To obtain airs available, all plants were preparation and carding) were operations (spinning, twisting, in the former group and 132 in THlS DOCU Ai MILLIONS OF PARTICLES PER CUBIC FOOT (IMPINGER AT 100*1 Figure 5. Fiber concentrations by membrane filter (430 r.) vs. particle concentrutions by midget impinger (100 x) weaving. ,T BY sPFpINIQL0EUSINSaD7uJRf,%g, eWCA,ENmSN,c0ISTNiCmBp aALUJtrrH^5E0NMcToEICrfAdSTOoMDf 11111 327S . -------- -.k a'5trrt *' -y ,f ri'"'^' -- 2S3 Annals New York Academy of Sciences Table 1 Ratio or Fjbkrs/cc (430x MF) to MPPCF (loox iMnNcta) by Plant a no Operation ^----- - Plant Operation'--^. Fiber preparation Carding AB C D U 8 4 11 9 14 6 27 Ratio ol average concentration by operation 8.6 12.7 Spinning 6 6 3 13 BJJ Twisting 6 2 7 14 6.5 tw1 Weaving Ratio oX average concentrations by plant 6 20 4 15 12.5 8.0 8.6 5.2 15.5 Ratio ol grand means 9.4 Ayer et aL: FIBER PF SPINNINC per mL for each mppcf. Figuhe 4 gives twisting, where ratios vary from 2 to 14 fibers per ml. for each mppcf. Figure 5 shows weaving; ratios range from 2 to 20 fibers per ml. for each mppcf for the four plants shown. These data are presented in tabular form in TABLE 1. The ratio of the averages of membrane fiber to impinger dust count for each operation was obtained from the averages of all samples at the particular operation in the four plants. Similarly, the ratio for each plant was obtained from the averages of all samples taken at the five operations in that plant The difference in ratio of averages between plants probably reflects the fact that each plant was surveyed for a brief period. It may be accounted for by different fibers being processed at the time of the survey, by seasonal or other atmospheric differences, by the type of product being produced, or by any number of other factors. It is obviously impossible, considering these data, to give any single ratio that would accurately represent all processes at all times in each plant. The ratios as presented,, however, do not demonstrate the fact that at many of the ..operations both fiber and particle counts were consistently low. This is shown graphically in fioure 6, where cumulative percentage is plotted vs. concentration by the two methods. The data are divided rather arbitrarily into the operations of carding and fiber preparation, and the subsequent operations of spinning, twisting and weaving. For Figure G. Proportion of sati counts lower than specified cone* carding and fiber preparatior parallel, with 90 per cent of 90 per cent of the membrant the subsequent operations i generally lower, with almost per mL In this set also, how5 mppcf and 50 fibers per r broken down to the five indh range from G fibers per ml. i criteria of an equal proportu tion is to be used, 5 mppcf i. ml. The ratio of means of fit that a value of 50 fibers per r the other method of data an;t THIS DOCUMENT WAS NOT A RECORD OF PPG INDUSTRIES, INC. DID NOT COME FROM IT'S FILES AND CANNOT BE AUTHENTICATED .BY PPG INDUSTRIES, INC. i BB 002078? | 3279 , of Sciences '? (100X IMFIXCER) BY IN Ratio of average concentration by operation II 27 12.7 13 8.3 14 12.5 latio of grand means 9.4 t' oj. where ratios vary from Jiows weaving; ratios range ie four plants shown. These l The ratio of the averages ach operation was obtained alar operation in the four btained from the averages at plant. The difference in ts the fact that each plant ited for by different libers ional or other atmospheric ;ced, or by any number of tiering these data, to give j all processes at all times honstrate the fact that at * counts were consistently ire cumulative percentage ds. The data are divided ?g and fther preparation. ?isti;ng and weaving. For Ayer et al.: Comparison of Techniques FIBER PREPARATION AND CARDING 2S3 SPINNING .TWISTING ANO WEAVING TOTAL FIBERS PER ML-430*MF 1 Ficukk 6. Proportion of samples of asbestos dust and asbestos fibers with [ counts lower than specified concentrations. | | |. J ;* j carding and fiber preparation it may be seen that the two lines are almost parallel, with 90 per cent of the impinger samples less than 5 mppef and 90 per cent of the membrane filter samples less than SO fibers per ml. In the subsequent operations it may be seen that the fiber counts were generally lower, with almost 60 per cent of the samples less than 10 fibers per mL In this set also, however, 90 per cent of the samples were below 5 mppef and 50 fibers per ml. respectively. When the data are further broken down to the five individual operations, the fiber counts per mppef range from 6 fibers per ml. to 14 fibers per ml. Over-all, however, if the criteria of an equal proportion of samples of less than a given concentra tion is to be used, 5 mppef is exceeded no more often than 50 fibers per ml. The ratio of means of fibers per ml. to mppef from table 1 is 9.4, so that a value of 50 fibers per ml. for 5 mppef is reasonably consistent with the other method of data analysis. THIS DOCUMENT WAS NOT A RECORD OF PPG INDUSTRIES, INC. DID NOT COME FROM It'S FILES. AND CANNOT BE AUTHENTICAUCD y. PPG INDUSTRIES, INC. ' -J--~---q--b-- 20788^ 32QQ NOTE: THIS DOCUMENT DID -L ... U-, . f - t i 1 ---*i-- - 1- * .n * -A M .k a < li* .< , J t J i fc -- !-- 4r * * 4 i . 4 > * * * K h * JAj ^ 4 -- < ^*1 . ef f u n f- e ^ W , j , , 4 * ^ .^^ 1 2S4 Annals New York Academy of Sciences Tabu: 2 Ratio or Fibers/cc Loncer than 10* (430* 1IF) to MPPCF (100x IMFINCER) BY PLANT AND OPERATION Operation plant A B C D Fiber preparation 5413 Carding 2526 Spinning 24l4 Twisting Z135 Weaving 3844 Ratio of average concentration by plant 2Jt 2.7 1.2 4.5 Ratio of average concentration by operation 1.7 3.5 2J> 2.5 3.8 Ratio of grand means 2.7 Fibers longer titan 10 ft The concentrations of fibers longer than 10 ft determined from the same membrane samples as above was also compared with the particle count by impinger. Table 2 shows the ratios of 10+ ft fibers per ml. to impinger counts in mppcf for 5 major 0|ierations in four plants. The ratios of aver age concentrations range from 1.7 to 3.8 by operation and from 1JZ to 4.5 by plant; individual operations are even more variable (one or two samples with extreme ratios can cause a considerable change in the least squares line which best fits the data;. The data exhibit about the same degree of scatter indicated in FIGURES 1-5. Figure 7 shows the cumulative percentage of 10+ ft fiber concentrations and impinger concentrations. To compare the 10+ ft fiber concentrations with the concentrations by impinger. the pairs were divided into two groups as before; fiber preparation and carding in one, and spinning, twisting and weaving in the other- In both sets of data, 90 per cent of the counts from impinger samples were less than 5 mppcf and in both sets. 00 per cent of the concentrations, of fibers longer than 10 ft from the membrane filter samples were below 15 per ml. (Individual operations had cumulative percentages equal to the percentage below 5 mppcf at 9 to 21 fibers longer than 10 ft per ml.) Both methods of analyzing the comparative data suggest that the ratio a lAUL/UMthli <T'S FILESS/Mnn (vui a RFffipn nc D! NT C0,V1E FR0X1 . BY PPG INDUSTRIESJNa AUTHENT,CATED Figure 7. Proportion of i than 10a with counts lower tl of impinger (lOOx) count than 10 ft is approximate! Not enough of the mem counted for fibers longer impinger samples. Howevt brane filters counted in tl the 10 ft fiber counts on tl than 5 ft appears to be eq a 10 ^ fiber count of 3 per * *< Samples were also tab' triator of the design sugt were compared with comp sampling and counting \ "['bT 0020 789__l 3281 NOTCOME FROM PPGFILES Sciences ixMF) to MPPCF )re*ATION Ayer ct al.: Comparison of Techniques FIBER PREPARATION ANO CARDING 285 c-o i t determined from the same with the partide count by fibers per nil. to impinger plants. The ratios of avereration and from 1J2 to triable (one or two samples hange in the least squares about the same degree of 10-r ft fiber concentrations 0+ ft. fiber concentrations rs were divided into two ing in one, and spinning, . of data, 90 per cent of the 5 mppcf and in both sets, >nger than 10 ft from the (Individual operations had e below 5 mppcf at 9 to 21 data suggest that the ratio I . Figure 7. Proportion of samples of asbestos dust and asbestos fibers longer than 10a with counts lower than specified concentration. of impinger (lOOx) count to membrane filter (430x) counts of fibers longer than 10 ft is approximately 1 mppcf to 3 fibers per ml. Fibers longer than S ft Not enough of the membrane filter samples taken with impingers were counted for fibers longer than 5 ft to make direct comparisons with the impinger samples. However, in comparing the 5 ft fiber counts on all mem t brane filters counted in this manner, with either the total fiber counts or I the 10 ft fiber counts on the same filters, a count of 6 fibers per ml. longer ! than 5 ft appears to be equivalent to both a total count of 10 p.r ml. and i I ' a 10 ft fiber count of 3 per ml. Size~telective Sampling ! Samples were also taken on membrane filters using a horizontal elutriator of the design suggested by Hamilton and Walton.'1 These samples were compared with companion unelutriated samples- Although the usual sampling and counting variances cause difficulty in drawing firm con- | THIS DOCUMENT WAS NOT A RECORD OF j DB 0G20 79 0_j| PPG INDUSTRIES, INC. DID NOT COME FROM ----------------------------- -- IT'S FILES AND CANNOT CE AUTHENTICATED 3 BY PPG INDUSTRIES, INC. i 2S6 Annals N**\v York Academy of Sciences Ayer et aL elusions from the relatively few sets of samples taken so far, these sample pairs give average ratios (unelutriated to elutriated) of 1.3 for total fibers and l.G for fibers longer than 10 fu In 58 per cent of the cases the total 6. Drkessen, W. CL, J. M. Sayers, H. F. Easom bestos textile industry. 7. Bloomfield, J. J. & J. M fiber count was higher in the unelutriated sample, and in 59 per cent of the sets the 10+ /l fiber count was higher in the unelutriated sample. of industrial dust. Pub! 8. Csalley, L J, H. E. A Lainhart. 1964. Ow These results suggest that the major portion of those fibers counted on the dustry in the United E membrane filter are "respirable" as the respirable fraction was defined 9. Transactions of the Fift Governmental Industr by the Johannesburg Pneumoconiosis Conference.1* II t[ Summary .`-rr methods, Washington. .10 Subcommittee on Dust P, in air analysis: The s Two hundred and thirty pairs of samples taken by two widely-used Hlth. Assn. Yearbook. sampling and counting methods were compared. These methods were count ing of total dust in impinger samples, and counting fibers only on a mem brane filter. 1L Air Hygiene and Manufat Method for detennini . asbestos textile institut 12 Fitzgerald, J. J. <fc C. G At the present time, based on samples collected in four asbestos textile plants, the over-all ratio appears to be one mppcf by impinger (lOx ob tration through the M 13. Li.vokke.v, C. F. K. P face collection efficien jective) to 10 fibers per ml, 6 fibers per ml, longer than 5 fu or 3 fibers 10: 7. per mL longer than 10 p when the membrane filter samples are counted with phase contrast illumination on the transparentized filter using a 4 14. Fraser, D. A. 1953, Ah* airborne particulates and electron microscor mm. (43x) objective. Individual operations may have, ratios differing 15. Hamilton, R. J, & W. H t1 from these values by a factor of two. Size selective sampling using hori I zontal elutriatnrs ahead of the sampling device indicated that the major able Dusts. Presented . held in Oxford, March 16 Orenstein, A. D., Ed. l: portion of the fibers counted would reach the lung if inspired. Acknowledgments Data in this paper were the result of sampling and impinger counting o OO UJ o held at the Universit: Little Brown. Boston, 1 I by C.JU. Brown, H. L. Byrd, G. H. Edwards, R. G. Edwards, G. W. Fisher, J. C. Lumsden, S. B. McKee and J. R. Stanton. Membrane filter samples LCTM C5 1 were counted by T. E. Beaver and K. M. Kelly. Electron micrographic siz ing was done by G. II. Edwards. a_ a.B~- References Lanza, A. J., W. J. McConnell & J. W. Fehnel. 1935. Effects of the in halation of asbestos dust on the lungs of asbestos workers. Public Health Rep. 50:1-12, Reprint No. 1665. Donnelly', J. 1936. Pulmonary asbestosis: Incidence and prognosis. J. In dust. Hyg. 18: 222-228. McPiiekters, S. B. 1936. A survey of a group of employees exposed to as bestos dust. J. Indust. Hyg. 18:229-230. Shull, J. R. 1936. Asbestosis: A roentgenologic review of 71 eases. Radiology 27: 279-292. 5. Fulton, W. B.. A. Dooley, Jl L. Matthews & R. L. Hoirrz. 1935. Asbes tosis. Part IL The nature and.'amount of dust encountered in asbestos fabrication plants. Part III. The'effects of exposure to dust encountered in asbestos fabricating plants on the health of a group of workers. Pa. Dept, of Labor & Ind. Bull, 42. oX oa: LLUJ * ^3* 'o C' o ,4 - ,A this document WAS not a KECora of ul. nsSro CANNOT BE AUTHENrtCAi ED by PPG INDUSTRIES* INC. .- | BB 0020791_| 3283 r\^jy~rVr. 0^''T"T7.'V- '>. V : * ^ > * t -V-.V, ^ "1# w * *4 ,, V** *. 'V Sciences Liken ao far, these sample ted) of 1.3 for total fibers ent of the cases the total ile, and in 59 per cent of the unelutriated sample, hose fibers counted on the ible fraction was defined . >* iken by two widely-used hese methods were countng fibers only on a mera- d in four asbestos textile icf by impinger (lOx obger than 5 ft. or 3 fibers iter samples are counted irentized filter using a 4 ay )--e ratios differing pling using horiin^Bjd that the major C inspired. " g and impinger counting . Edwards, G. W. Fisher, Membrane filter samples lectran micrographic siz- Ayer et al,: Comparison of Techniques 2S7 6. Dkeesskn, tV. CM J. M. Dallavalle, T. I. Edwards, J. W. Miller, R. R. Savers, H. F. F.asom & M. F. Trace. 1938. A study of asbestosis in the as bestos textile industry. Public Health Bull. 241. 7. Bloom field, J. J. Si J. M. Dallavalle. 1935. The determination and control of industrial dust. Public Health Bull. 217.,, 8. Cralley, L. J-, H. E. Ayer. P. EL Entemjne, A. F. Henschel & W. S. Lainhakt. 1964. Occupational health study of the asbestos products in dustry in the United States. Proc. of Caen: intern. Congr. on Asbestosis. 9. Transactions of the Fifth Annual Meeting of the National Conference of Governmental Industrial Hygienists. 1942. Subcommittee on' standard methods, Washington. 10. Subcommittee on Dust Procedures in Air Analysis 1935-3G. Dust procedures in air analysis: The sampling and analysis of industrial dusts. Am. Pub. Hlth. Assn. Yearbook. 11. Air Hygiene and Manufacturing Committee, Asbestos Textile Institute. 1963. Method for determining asbestos dust concentration approved by the asbestos textile institute. 12. Fitzgerald, J. J, & C. G. Detwilul 1957. Optimum particle size for pene tration through the Millipore Filter. AMA Arch, of Ind. Hlth. 15: 3-8. 13. Lixdkken, C, I~, F. K. Pelrock, W. A. Phillips & It. D. Taylor. 1964. Sur face collection efficiency of large-pore membrane filters. Hlth. Physics 10: 7. 14. Fraser, D. A. 1953. Absolute method of sampling and measurement of solid airborne particulates -- combined use of the molecular filter membrane and electron microscopy. Arch. Ind. Hyg. &. Occ. Med. 6: 412. 15. Hamilton, R. J. &. W. H. Walton. I960. The Selective Sampling of Respir able Dusts. Presented at Symposium on Iniwled Particles and Vapors held in Oxford, March 29-Aprii 1. Pergamon Press. Oxford, England. IS. OrensteU'T, A. D-, Ed. 1959. Proceedings of the Pneumoconiosis Conference held at the University-of the Witwatersrand, Johannesburg, Feb. 9-24. Little Brown. Boston, Mass. CD CO -- LU Q __j ZD oo o o=: u_ ..j. uj oo EL. 1335. Effects of the in to* workers. Public Health !<nee and prognosis. J. In- f employ es exposed to Re view of 71 cases. Radiology i*