Document 82EDOV4eM26Qo6b5E7NNboJNK

ASBESTOS EMISSIONS FROM RAG1IOUSE CONTROLLED SOURCES by in F. Harwood, Ph.D.*, David K. Oostreich'-"-', Paul SicbertV: and John D. Stockham* Prepared for Presentation at the American Industrial Hygiene Conference Minneapolis, Minnesota June 1-6, 1975 IT Research Institute, 10 U'cst 35th Street, Ckie-ve, llinois' 60u 16 EPA, Control Systems Laboratory, Research Triangle Park, N.C. 27711 GAF17173 ASBESTOS EMISSIONS FROM E/CliOUSE CONTROLLED SOURCES tv' Colin F. Harwood, Ph.D.*, Daviil K. Oestrcich**, Paul Siebert`v, and John D. Stockham'- ABSTRACT There is virtually no in formation published on the absolute efficiency of baghouses in reducing the emissions of fine particles of asbestos. This lack of information is unfortunate because serious occupational health problems may result from the common practice of recirculating air to con serve energy. Emission testing has been conducted at five asbestos processing plants where the emissions are controlled by baghouses. The results showed that the mass removal ef ficiency frequently exceeded 99.99%. Membrane filter samples of the effluent were examined by optical and electron micro scope. It was observed that despite the high mass efficiency, the number of fibers emitted, which were greater than 1.5 ;.m in length, was about 10 `-lCr fibcrs/mJ, while the number cf fibers less than 1.5 ym was 10~-10^ fibors/n^. The signifi cance of the sice of the fibers in terms of probable health impact is briefly discussed. Prepared for inclusion in the 1975 American Induscrial Hygier.c Conference (June 1-6, 1975) in Minneapolis, Maun c-r oca. TIT Research Institute, .10 West 35th Street, Chicane, Illinois 60616 ** EPA, Control Systems Laboratory, Research Triangle Parle, N.C. 27711 i i GAF 17174 ASBESTOS EMISSIONS FROM IWCllOUSE CONTROLLED SOURCES INTRODUCTION' A recent survey of asbestos processing plants revealed that baghouses are by far the most common device for control ling the emissions from these sources'. Asbestos has been declared a hazardous pollutant by the El'A and is known to cause asbestosis, and is strongly associated with the forma tion of cancers, particularly mesothelioma. It is surprising, therefore, that a review of the published literature has revealed only one reference to the efficiency of baghouse devices to limit the emission of asbestos fibers. In that paper7, the efficiency was quoted simply as 99.9927. on a mass basis and no details were given of the mcasurment technique used. This paper details the efficiency testing of baghouses located at five asbestos processing plants. Information was obtained, not only on the mass efficiency of the dust col lector, but also on chc efficiency as a function of the site of the fiber challenging the filter. Site efficiency was obtained by counting the number of fibers from a given volume of air by means of optical and electron m: c r o sc eg v The use of an electron microscope enabled the smallest fibers, as small as 0.02 urn in diameter and 0.06 i;m long, to be counted. The procedure used by the authors is quite different from the OSHA procedure3 which requires that onlv those fibers longer than 5 urn be measured. At >5 ;jn unambiguous measurements by mu'tiple observers may be made using the relatively simple optical microscope with phase contrast at 430X. The occupational health standard is based on information gained on the incidence of asbestosis among those workers occupationally exposed to asbestos at known concentration levels of fibers greater than 5 iim. The importance of this study in terms of industrial hygiene is chat it gives information on the possible conse quences of Che common practice of recirculating air from the baghouse bade into the plant. The relationship between asbestos particle sine and the induction of cancer has not yet been established; until such time, it is prudent to quantify the concentration of fibers of all sizes present in plant air. EXPERIMENTAL Plant Location Five different plant locations were selected for stuck.-. The basis for the selection was the nature of the asbestos processing category which they represented. Thev included: two asbestos ore refining mills, two asbestos cement product plants (where the asbestos fibers can be considered tc be bound into the product), and an asbestos textile plant (where the asbestos fibers car. be considered as loosolv bound) . In all the locations, ehvyr.otilc asbestos was used exclusively. Baghouse emission control devices were used at all the locations, and details of the kytrhouse construction sr.d GAF 17176 T operation are summarized in TaMc I. These installations are typical of good, accepted modern practice in the industry. Recirculation was practiced ac three of the five locations in order to conserve heat during the winter months. Sampling Procedure Sampling was conducted using, where possible, the EPA pro cedure detailed in the December 23, 1971, Federal Register, 36, 247. Two standard, EPA Method 5, isokinetic sampling systems as shown in Figure 1 were utilized to allow simultaneous sampling both upstream and downstream of the baghouse. The sampling probes and nozzles were fabricated from stainless steel. The probes had an I.D. of 1.27 cm (0.5 in.) and a length of 107 cm (42 in.), and the nozzles were 0.63 cm (0.25 in.) I.D. Sampling points in the duct work were selected in re gions where the most stable flow patterns existed. The ports were located, where possible, eight to ten diameters upstream from any bends, elbows, junctions, or other constrictions in the stack or duct. In some instances, it was not possible to collect samples by the isokinetic sampling method. This was because of physical limitations in the plant design which made such points inaccessible, or because of the elese proximity of accessible sections to fans or bends. here, hirh volume samplers fitted with 20 cm x 25 cm (8 x 10 ; n.) membrane filters were used within the baghouse close to the exit. On the upstream side of the baghouse, samples were drawn through a cyclone followed by a 10 cm (4 in.) men-rone filter of 0.8 vim pore size. Or the downstream side, no GAF17177 Table I baciiouse specifications Plant Locacion Product llaghouse Make I Waukegan, Illinois Asbestos Cement Parsons 2i 3 4 5 Denison, Mnrshvillc, Asbestos, Eden Mills, Texas N.C. Quebec Vermont Asbestos Asbuslos Mi,l led Milled Cement Textiles Asbestos Asbestos Uhcclabrator Wheelnbrntor Wheelabvator Whcelnbrator :nd Model bughouse Special IV:; i j'.n Model 26 Model .11? Spec; nl. Du:: i;;.n Model 8-607 Scries VI11 Capacity min- (cfm) Air to Cloth Ratio eftuft^ 1,132 (AO.000) 2.0:1 736 (26,000) 2.6:1 473 :i6,000) 3.0:1 127,000 (4.500,000) 3.0:1 R, 490 (300,COO) 3.2:1 vjmbcr of Compartments 4 1 1 11 8 l'otal Humber of Bats SOO 734 304 79,200 2,638 liar Size L':.a:ncter, cm 12.7 (5") 12.7 (5") 20.3 (3") Lcnpth, m 3.05 (10") 3.18 (10'5") 2.82 (9'3") bag 12.7 (5") 4.27 (14') 20.3 (8") 5.33 (17'6") Permeability cfm:ft^ at 15 + 5 15 + 5 15 + 5 25 16-20 11 0.5" H20 Bag Construction Material Weave Threads/in. Pressure Drop in. I!20 (Max.) Cleaning Method Cotton Sateen Cotton Sateen 96 x 60 6 5.7 j Mechanical Mechanical !Snake Shake Cotton Sateen ! ^ 2.2 : Mechanical ; Shake Cotton Sateen -3 [ W. c c h a n i c a 1 Shake Cotton Sateen 96 x 60 4 Mechanical Shake GAF 17178 Filter Holder Fdg'irc 1. Sampling nr ran gamer, t GAF17179 cyclone was used: the air stream was lead directly to the membrane filter. Afeer sampling, the filter was placed in a marked plastic folder. 1-iaterinl adhering to the inside of the probes and tubes was washed into a sample collector using acetone. The collector was then marked and sealed. Sample Analysis The mass collection efficiency was determined by accurately weighing the samples collected before and after the baghouse. The filters were weighed on an electronic balance having a sensitivity of + 0.1 mg. On the upsgream side, the dust deposited in the probe, the sampling train, and cyclone was washed with water into a weighing bottle. It was then dried in a vacuum oven at 110C, cooled and revcighcd. From these l 1 weights, the mass efficiency was calculated. Fiber Counting Optical microscope analysis was performed using the method described in the K10SH criteria document on asbestos' . A portion of the membrane filter, approximately 1 centimeter square, was removed from the central part of the filter arc mounted on a slide. Using a 1-1 solution of dimethyl phehaiate and diethyl oxalate, the filter was allowed to clear clean cover slip was placed on top of the sample and the fiber concentration determiner). The light microscope ueed was equipped with phase-contrast and polarized light. The objective lens of 4 mm resulted in a total magnification of 500X. From, randomly chosen fields, the number of fields for a total count of 100 fibers rats noted (with a minimum of 10 fields observed), or 100 fi> iis were observed when the i OAF 17180 distribution was sparse. For the electron microscope analysis, a circle of the sample filter 3.5 mm in diameter was cut. This piece of filter was placed on top of a carbon-coated 100 mesh electron microscope grid. The grid with the filter on it was placed in a condensation washer usinj; acetone as a solvent. The filter medium was dissolved away by the acetone,depositing the fibers of the sample undisturbed on the carbon substrate of the grid. The specimen was then counted on a Hitachi HU-11 transmission electron microscope at a magnification of 16.364X. The optical microscope analysis enumerated the fibers greater than 1.5 urn in length and a minimum diameter cf 0.5 um. The electron microscope analysis counted fibers dovm to 0.06 um in length and 0.020 um in diameter. Calculation of Fiber Numbers To relate the number of fibers to the asbestos concentra tion in the air, the following equation was used: no. of fibers raJ of air fno. of fibers counted! no. of lie Ids ] " effective filter area, cm 2 [area of microscope's field of vie..-, where: Effective filter a r c a Effective filter area Effectivc filter area Area cf f ield of view [velum,c of air sampled, m. j 61.7 c for 4 in. filter in EiY. t r a : n 63.2 cm^ for 4 in. lli-Vol filter 425.^ era 2 for S" x 10" lli-Vol fi ter 6 . 5 If. 10 (500>:. cm for optical rate cscope GAF17181 Area of field of view 1.34A 10 - 7 cm 2 for electron micro scope '16 , 364X) Efficiency Calculation To calculate the efficiencies, both by mass and by number, the following relationships were used: Mass Efficiency (7,,) = 10011 - Mass in a given Mass in the same vol umc volume of ol air air aftor the baghousc before the baghousej Number Efficiency (7.) = 100 fl - No. of fibers per cubic meter after the baghousc 1 [ No. of fibers per cubic meter before Lhe ETTphouse, RESULT? Hass Removal Efficiency The mass efficiencies of die five beghouscs arc given in Table II. It is observed that the efficiency is extremely high and, in all instances, exceeded 99.997',. At two loca tions, the mass collected on the upstream, side of the filter was too small to be weighed, although a faint coloration of the otherwise pure white filter could be observed. Fiber Removal Efficiency In Table III, die number of fibers up stream and down stream of die five baghouses arc presented. In general terms, the number of fibers greater than 1.5 urn in length exiting from the baghou.se is of the order of If/1 fibers per cubic meter, while the number of fibers greater than 0.06 am in length is of the order of 10 s fibers per cubic meter. A similar ratio is found in the number of fibers approaching the bag'nouse, where 109 and 10J `` fibers per cubic meter of GAF1712 Tenb 1 c III UPSTREAM AND DOWNSTREAM EIDER CONCENTRATIONS AND REMOVAL El'T'7 C 1. EMC 1. ES FOR I'.ACIIOUSE H'NTROEEEn AS I'.ESTOS EMISSJON SOURCES AT FIVE PLANT LOCATIONS Plane Locacion 1 i-.'nukcgnn, Illinois Sampling Location ------------------------------------------- 1 Fibers Counted by Fibers Counted by | Optical Microscope at Electron Microscope a Magnification of at a Magnification of 50 OX 16,36AX Fibers per Removal Fibers per Removal Cubic Efficiency Cubic F.f f j.ciencv Meter 7. Meter Upstream > 101U Dov/n scream 6. A x 10J > 99.99 > 101* 1.1 x 107 > 9^.99 Earshville, Upstream S.l x 10" N.C. Downs cream 1. A x 10* > 99.99 2.5 x 10iL 3.3 x 10y 93.69 j 1 Denison, Texas Asbestos, Quebec Ecler. Mills, VernonC Upstream jl.O x 10 Dotmstrean 2.9 x 10* Upstream 2.2 x lo`J Dov:ns cream S.2 x 103 TT Uostrearn l.A x 10' Dovnstrean A . 5 x 10' i 97.IS 99.96 3.2 x 107 1.A x 107 ! 1.2 x 101" 1 1 l.A x 10y j 1. A 10' > 99.9 i- 3 x lO1- 57.90 n n dq 9o 99 GAF17183 I0 greater chan 1.5 um anti greater than 0.06 tin arc fount!, respectively. Very high removal efficiencies arc found in all instances, based on the available data, it docs not appear chat the re moval efficiencies are sir.c dependent. Similar e ff i c i ene i os arc found with the greater chan 1.5 tim fibers and the greater than 0.06 nm fibers. One exception to die very high efficiency values occurred at Denison. At this locacion, high grain loadings were observed in the upstream duct, and it is possible that partial blockage of the probe prevented fibers from reaching the filter. The number of fibers on the dov.-nstrenm side was comparable to the other locations. Despite very high collection efficiencies, very large numbers of fibers were found to exit from the baghouse. This i.s shown in Table IV, where the total number of fibers exiting per minute and their concentration in the air stream has been tabulated for each plant location. DT5CUSSTON The Recirculation of Cleaned Air The result of a survey taken as part of this scuriv has revealed that die recirculation of air exhausted from baghouses used to clean asbestos dust from air is common prac tice in die industry. The American Conference of Covevr-cntr.i Industrial Hygienists has the responsibility through : Ventilation Committee of publishing a Manual of Recommended Practice, which establishes the governmental vccommcnd.it ions as regards the recirculation of air within a plant. Tiie.sc I GAF17184 TabIs IV THE NUMBER OF FIBERS EXITING FROM FIVE UACHOUSE CONTROLLED ASBESTOS EMISSION SOURCES AS COUNTED BY OPTICAL AND ELECTRON MICROSCOPES Plane Location Number of Fibers Exiting Niinibcv of Fibers Exiting Croat or than Greater than 1.5 pm length 0.06 pm length --------------------- [ Concentration Number per Concentration Number per. Fibers in'-* Minute Fibers m'-* Minute | Naukegan, Illinois 6.A x 103 1.2 x 106 1.1 x 107 1.2 x 1010 'larshvilie, N.C. 1.4 x 10/k 1.0 106 3.3 x 109 2.4 x 1011 Denison, Texas 2.9 x 104 1.4 x 10G 1.4 x 107 6.6 x 109 Asbestos, Quebec Eden Mills, j Vernon: S.3 x 105 , r n n4 A`5 x 10 1.1 >: 1011 3.fi x 107 1.4 x 109 1.S x lO1^ C^ 11 1.3 x 10 | 1. 1 x 1C'1 GAF17185 i2 recommendations are considered and cited in the standards of the Occupational Safety and Health Administration (0S1LA) of the U.S. Department of Labor. The recent energy crisis has occasioned a need to conserve heat as a national policy. Prior to this, regulations pro hibited the recirculation of plant air containing hazardous materials even though recirculation was economically attractive. This regulation was predicated upon the premise that no con trol device is foolproof. A temporary loss in control effi ciency would create considerable risk of exposure of workers to dangerous levels of toxic materials. The impact of the energy crisis has led the ventilation committee to propose circumstances which would allow the recir culation of cleaned air, provided certain criteria were met by the system. A formula has been proposed which would estab lish the limits of contaminant concentration in the exiting air; CR . -i (TLV - where = maximum permitted concentration of contaminant in exit air from the collection after cleaning, in any consistent units TLV = threshold limit value of contaminant CQ = concentration of contaminant in workers breathrng zone with local exhaust discharged, outside Qj = total ventilation flow through affected space, cfn, Qp = recirculated air flow, cfm K = an "effectiveness of mixing" factor, usually varying from 3 to 101 1. i GAF 17186 Ic should be remembered Chat this equation has only been pro posed, not accepted, at this stage. For asbestos, the TLV value would be that presently accepted as the OSHA standard, that is,' 2 fibers per cubic centimeter (where a fiber is greater than 5 pm and with an aspect ratio of greater than 3:1 length to breadth). As has been Shown in this paper, only a small portion of the true asbestos concentration in the recirculated air is accounted for by this standard. Two methods are proposed that would increase the sa fety of such a system and guard against a temporary breakdown in the collection efficiency. One is that a back-up filter should be fitted to remove the contaminant in the event of a failure of the primary control. The application of such a system would obviously add considerably to the costs. The second method would be to use a monitor which would stop the recirculation, or activate a rapid acting by-pass to the out side air, in the event of a failure of the collector. The problem would be to produce a monitor for asbestos which would be fast acting and modest in cost. At the present time, no such monitoring device exists for asbestos, despite con siderable research effort to achieve this objective''. Exposure Levels for Asbestos The present accepted standard for asbestos exposure has been determined by OSHA to be 2 fibers per cubic centimeter, for fibers longer than 5 um, The measurement method, and: the rationale behind this exposure level is well presented in the N10SH document "Occupational Exposure to Asbestos"5. The selection of the site of 5 um as the shortest site of the fibers which are counted, stems from British practice' . ] '! GAF 17187 The British use this criteria Tor two reasons. Firstly, it represents that size which is readily detectable using a relatively simple optical microscope, while detection for smaller fibers and particulary submicron fibers requires the use of .'in elect'von microscope, which is both t i mo-con sumi tv, and costly ro use. Secondly, a wide body of information on the exposure levels and incidence of asbestonis has been generated over many years using the 5 pm size limit in estab lishing the exposure levels. Titus, the use of the optical microscope gives an index of exposure which has had wide practical applicability. The problem with using this index is that the ratio of the numbers of fibers greater than 5 pm in length to chose less than 5 urn in length is very much a function of the in dustry. Thus, textile-producing industries have a greater proportion of longer fibers than, for example, those industires using asbestos as a filler in plastics or.paints. The fiber size considered to be most dangerous is not established at this time. Evidence in the literature on this subject is confusing and reveals the uncertainty. Stanton7 has performed experiments that nave indicated that fibers, including fibers of materials other chan asbestos but having a similar morphology (c.g., alumina), in the size range of 10-100 pm arc most harmful, and that their ability to produce tumors decreases with decreasing length. however, his find ings have been criticized because of the method of application. The method required chat AO g of the powdered asbestos be placed in contact with the pleura and held in place by a fiber glass GAF 17188 patch attached surgically. Thus, the normal free movement of the particles was prevented. This work might also be criticized from the standpoint that,in all probability, the larger fibers do not find their way through the human body defense mechanisin';. rout0, on the other hand, has performed experiments using asbestos dry milled to a fine size. Two size groups were used, one with 957. of the fibers less than 5 uni, and a second with 997, of the fibers less than 3 um (with 937. less than 1 urn). The size was checked by electron microscopy, which also showed that the fibrous structure had been retained through the grinding process. It was found that with both groups, tumors developed in 407. of the rats after intraperitoneal injeccion of the asbestos. However, the small sized fibers required a longer incubation period than the larger fibers (about 20 months versus 12 months) to achieve the A07. tumor level. A further important point when discussing the exposure index for asbestos is that it is based on the probability of a worker contracting asbestesis, not cancer. Hith the present exposure levels, the probability of a worker gerev.r asbestosis is 17. if a 30 year working life in the industry is pursued. There is no definitive information on the chances of developing a cancer. Rccognic:on of the carcinogenic aspect of asbestos is comparatively recent, and it is esti mated that, in humans, a 30 year incubation period is re quired before the cancer develops. Therefore, there arc no adequate dose records upon which to relate the delayed rosnon.se. GAF17189 !t> Reasonable scientific predictions are made based on ani mal experiments but, because of the massive doses used to accelerate the incubation period, there is justifiable doubt as to the conclusions reached. Again, a degree of suscepti bility may be required within an individual such that a small dose might be fatel to him, while a much larger dose would leave, a second person unaffected. It is apparent from the literature that there are many unanswered questions about the danger associated with asbestos. Until the answers to these questions are more clearly under stood, asbestos should be treated with caution. CONCLUSIONS Baghouses used for the control of asbestos emissions have been shown to operate at very high efficiencies on both a mass basis and on a number of fibers basis. Even with these high efficiencies, there are substantial numbers of fibers which penetrate the fabric filter. Recirculated air may contain enough fibers to be of concern from the stand point of cancer induction but the dose/response relationship between asbestos exposure and cancer is not well understood. There is certainly, a possibility that a human health penalty could off-set the economic advantages of recirculation. If bag failure results while recirculation is being practiced, the probability of undesirable health effects could be expected to increase substantially. It is recom mended that a rapid-acting detection system be developed which would activate a by-pass system in the event of a bag failure. 17 GAF17190 It has been shown that recirculated air contains well below the current Threshold Limiting Value of 2 fibers per cubic centimeter (fibers equal to or greater than 5 pm in length). This would imply that workers would be protected from asbestosis unless the bags were to fail. ACKNOWLEDGMENTS The help and cooperation of Johns-Manville, Raybestos, and GAF in performing these studies is gratefully acknowledged. The help of Erdmann Lubecke, Tom Blaszak, Dr. Ranade, Anant Somudra, and David Becker of the IITRI staff is also acknowledged. 18 GAF 17191 REFERENCES Harwood, C. F., Sicbcrt, F., and Blaszak, T., "Assessment of Particle Control Technology for Enclosed Asbestos Emissions", Report No. EPA 650/2-74-088, October 1974. Coldfield, J., and Brandt, F. E., ''Dust Control Techniques in the Asbestos Industry", Presented at the AIHA Conference, May 1974. Edward, G. II., and Lynch, J. R. , "The Method Used by the Public Health Service for Enumeration of Asbestos Dust on Membrane Filters", Ann. Occup. Hyc-. 11, pp. 1-6, 196S. Rossi, R. C., Gaulin, C. A., Gerber, R. M., and Van Passen, 11. L., "Evaluation and Development of Instrumentation for Process Control of Air-Borne Asbestos", Interim Report, EPA Grant No. R802394-01, April 1974. (Private communication, Bruce Harris, Project Officer). "Occupational Exposure to Asbestos", U.S. Dept, of Health, Education, and Welfare, HSM72-10267, 1972. "Standard for Asbestos Dust Concentration for Use with the Asbestos Regulations, 1969", Dept, cf Employment and Productivity, Her Majesty's Factory 7nspector atc, Technical Note 13, 1970. Stanton, K. F., and Wrench, C., "Mechanisms cf M;srthc1ioma Induction with Asbestos and Fibrous Class", J. of the National Cancer Institute, 43, p. 797-32.7, 1973. Pott, F., lluth, F., and Friedrichs, "Tumors of P.ats After IP Injection of Pou'dered Chrysotile and Bento (a) pyrene, Zbi. Balct. Hyc., I. Abe. Oric- B 133, 463-469, 19"7. 19 GAF 17192