Document G5nyxoRNwonVkZV1qp5x38xOv

FILE NAME: Z< DATE: DOC#: Z< DOCUMENT DESCRIPTION :Z, DZ Journal ofOccupatIOnal and Enl'llvnmental Hygzene. 4' 525-537 ISSN: 1545-9624 print /1545-9632 online Copyright 2007 JOEH, LLC DOl' 10.1080/15459620701411109 Heavy Equipment Maintenan.ce Exposure Assessment: Using a Time-Activity Model to Estimate Surrogate Values for Replacement of Missing Data Fred W. Boelter,1 John W. Spencer,2 and Catherine E. Simmons1 1 Boelter Associates, Inc., Park Ridge, Illinois 2Environmental Profiles, Inc., Baltimore, Maryland This study onjcJlIr pieces oj heavy construction equipment was conducted to detern'line the concentratioll <d' airhople ashestos jihers during in-jrame maintenance and repair activities, which included aggressive techniques that resulted in visihle dust ji-om work involving ji'iGlion products and gaskets Despite execution oj a carejully planned sampling strategy, approximately 10% (47) oj the samples collected could not he analyzed due to overloading or filter damage. To include the overloaded samples in the data analysis, s1Jrrogate values were estimated /ollowlflg a time-activity model Twelve long-term personal samples, 2 short-term, 30min personal samples, and 31 long-term area samples were modeled. Personal and area time-weighted average (TWA) data were analyzed hoth with and without the estimated surrogate values and compared A total (~f 444 samples were collected over 9 days. Four experienced heavy equipment mechanics removed and replaced friction products and gaskets. Samples were analyzed using NIOSH Method 7400 Phase Contrast Microscopy f{)lIowed hy NIOSH Method 7402 Transmission Electron Microscopy Sample data inFirmation including the surrogate values/or the/ull-shift, TWA personal sample results ranged/rom 0.002 to 0064 ashestosf/cc. Personal, short-term, 30-mill sample results, including the Mo surrogate values, ranged/mm 0 038 to 0 561 ashestosf/cc Full-shift TWA area samples, induding the 31 surrogate values, ranged/mm 0.005 to 0.039 ashestos flec. Area air sample results at the end of the project were similar to levels measured he/ore the start of the project. No jiher cOllcentration huildup within the work area was indicated over the 9-day study AIljull-shift personal and area TWA sample results were helow 0.1 flcc, and shortterm 30-min personal samples were helow 1.0 /Icc. Statistical resu/tsoj the sample data with and without the surrogate values were consistent Use 0/ the time-activity model reduLed the uncertainty associated with this data analysis and provided a cOllsistent logical process f{!I' estimating surrogate values to replace missing data Keywords asbestos-containing, chrysotile, exposure assessment, friction products, gaskets, time-activity model Addl'CSil CQrrespondcn.(:e~; Cai'h~dne E. Sin:untl.ll$, B<ld~x As wclilteil, !ne" l.300tHgsws :Road, Sl.'utc 30i, hui :Ridge. ItlO68~ 5'771; e,nmH: t~nnmo.f'l>@ bo~lt.e.m~()dat~ .O})1)., INTRODUCTION A sbe~tos has been used in certain ,lpplica!ion~ of engine gaskets and friction products in high- a.nd lowspeed automobiles, trucks, buses, light dllty equipment, ,md heavy construction equipment. Workers involved in repair or replacement of gaskets and friction prodDcts used in heavy con~truction equipment are potentially e'Cposed t~) asbe:-,tos fibers. The authors anticipated and hypotiwsiz.ed that resulting airborne asbestos fiber levels would be below 0.1 flee 8hr time-weighted average (TWA) and l.l'l flee 30-min ~hon term exposure limit (STEL). The hypo!itesis wa~ ba"ed on the previous work experience of the authors and results of published studies discussed herein. Conducting an assessment for asbesto'i expomres on mechanics performing maintenance activities 0n heavy equipment presents a number ofchullenges to sample (ollection and :malysis, The analytical methods used to dctermi ne the concentration of fibers are sen:>itive to the presence of other particles on the filters, Activities such as welding, cutring using torches, and grinduig may render !he fiher overlvaded with panicubtc matter, the result being that the sample l:annot be analyzed. This study experienced a percemage of sample~ that were affected by these activitie~ and as a result, 10% of the <;amplcs could not be analyzed. To overcome problelIls assoClateJ with filter overlouJing fmd damage, the authors chose to apply a time-activity model commonly Ilsed In historical exposure assessment to estimate surrogate values for replacement of the missing data.l ] L 'Ji The "tuJy was conducted during June :~003 and performed in a heavy equipment repilir faci] ity by independent mechanic~ with extensive experience working on heavy equipment. Tilt' \vork spanned a period of 9 days and wo., conducted on four pieces of Caterpillar heavy equipment: A D8 Dozer (D8). a 12E Motor Grader (12E), a 930 Wheel Loader (930), and a 955 Track Loader (955 L Brakes. clutches, and engine" determined hkely to be original equipment witb asbestos-coma ill ing c(lmponents were Journal of Occupationai and Environmental Hygiene July 2007 525 dismantled. cieaned. and reassembled. The gaskets, brake linings, and clutch Iining~ were removed and replaced. METHODS Work Layout and Ventilation The repair faci lit)' ,,\'as constmcted of a steel frame and metal ~kin and was fitted with t\'vo. large, i>ectional steel overhead doors. The interior was unfinished and the dimensions wl?re 17.1 III wide by 18.3 m long with a roof that varied from 3.7 m at the side \vall to 8.5 m at the apex. Figure I shows the location of the equipment during the work and the loc:ltion of the overhead doors. The D8 was in Location A throughout the study. Sequentially, the 12E, 930, and 955 wae placed in Location B and worked on during the study. Figure I sho\"s the layout of the facility and the location (If equipment. The repair facility was not equipped with a mech:micalo ventilation s.ystl?m. The study was conducted under routine conditions with doors being opened only for entry and exit of per<;onnel. The overhead doors were opened only to move' e'-luipment anJ supplies in and out of the repair facility or for a few minutes following work activities, such as operating the diesel engines, welding, and grinding or other activities that generated dense smoke or fume. A pedestal-mounted fan \vas used by the me('hanic~ at various times to prov ide air movement inside the repair facility. Prior to and fonowing work activities, the effective ventilationrate (reported as the numbel' of air changes per hour I of the facility was determined by using CarDO[] dioxide (C02) tracer gas. The AmericJIl Society for Te~ting and Materials tA~TlVll Office , :> N ..~ Jl LocalionA e "I !!! .... Bay1 .. ' ...... ' ......... """""",' Bay 2 ':' ~ [ } - I.ocolicn B If Overhead Doors " , ~ FIGURE 1. equipment 17 .llIItlers Layout of repair faCility and location ot heavy Method E 741-00 Standard Test Method for Determining Air Change in a Single Zone by Mean:- of a Tracer Ga<; Diluti~)I] was followed as a guideline.(4) SAMPLING STRAT.EGY AND ANALYTICAL METHODS T he faCIlIty owner~ removed intenering constructlOn equipment and cardbo:Jrd hoxes from the rerair facility. Ten background air sample~ were collected for :lpproximately 145 min at 2ljmin on the day prior to the start of repair activities. Five sample~ \v'ere collected in"ide the facility, and five were collected out<;ide the facility to establish background fiber concentrations. Personal and area air samples were collected each day with sample pumps started prior to the initiation of work activ ilies and were "topped at the end of the day after completIon of work. Consecutive, long-term sample~ Wetl' colkCkd Clll each mechanic and at each area sampling location. Samples Wefe combined into TWAs over the whole sampling period to achieve full-shift TWA concentrations. The "hort-term, per~onal monitoring period" \verc for 30 min and were coUected during :lctivitie~ when the meclwnics were directly impacting the a~be~tos-('ontaining gasket or friction material. These ~amples INere colkcted to reflect pea..k expo~ure cpncentra'tions from activities conducted over a short pl?riod of time. To determine average room fiber concentration" in air, a procedure ba~l?d on Environmental Protection Agency (EPA) asbestos guidance documents was used.(~-71 The occupicd volume of the repair facility was subdivided into 40.320 modules. The volume of each module equaled 0.03 m' _Only areas bet\vecn the floor and a height of :'.7 m \vere included in the random generation of area sample locations a~ this i" the zone \vhere the mechanics worked. Eight of the module:,> \vere "elected using a random number generator. A ~ample cassette was located in the center of the re'ipective module for the duration of each day of testing. Area samples wcrecolkct,xl and analyzed using the same methods a~ the per~onal samples. All background. personal. and area salllples were collected and analyzed following National In:'>titute for Occupationill Safety and Health (NTOSH) Method 7400 for collection and analysis by PCM.'"' All air sample~ were ,:oHeeled at a rate of approximately 2 Limin. Samples with re~u1ts above the limit of detecti.on (LOD) by phaSe c'ontrast micro~copy (PCM) wae subsequently analyzed by NIOSH t.,/fcthod 7402 Transmission Electron Micr%eopy (TEM) to identify fiber type.('J) TEM analysis determine~ the fraction of asbestos fibers prcsent 0n the PCM filters. This fraction was then multiplied by the reM fibers per cubic centimeter dice) re"ult in accordance with NIOSH Method 7402 to ob1ain a PCN! equivalent (peME) asbestos flee. All sample" above the Peiv1 LOD In the study were subs,equently analyzed ~y TENL For sample results below the LOD, the peM LOD value was used in the TWA calculation and/or statistical analyst:s. 526 Journal of Occupational and Environmental Hygiene July 2007 I Missin~ Data. Determine time period (TtOlal) and tasks or activities performed. , , Break down th e individ ual tasks or activities along with the corresponding time for each,separate one. 1Task 1 Time 1 l Task 2 Time 2 Review available.sample results for same tasks or similar activities. , Task 3 Time 3 Select sample result value(s) representative of each task or activity and docume nt the process used. 1 , Perform nNA calculation to estimate a surrogate valu e for the missitlg data point. Th e sum of time for all tasks or activities should equal TtOlal. Is the surrogate value reasonable when compared to values of similar activities? If not, re-evaluate. Combine surrogate. value with other measurement(s) to obtain the full-shift TWA estimate. fiGURE 2. 'Process frameworK for tile time-activity model usea to determine an estimated surroga1e value for missing da1::t. 'Note. Ti'IIS value may be based on an average, mean. maXimum, or some other value. Although not used in the mode! for thiS paper, commercial softvvare for stochastlc modeling IS available to assist with selection 01 a representative value Journal of Occupational and Environmental Hygiene 'July 2007 527 For example, if a sample was analyzed by peM and resulted in < 0.045 flec, a value of 0.045 f/ce was u~ed as the value for peM and PC.ME. For samples nnalyzed by TEM, and where the result after multiplying the peM result by the TEM fraction was below the PCM LOD. the PCM LOD was used as the value in the TWA calculation and/or statistical analysis. Surrogate values for missing (overloaded and damaged filters) samples were estimated by using a time-activity model. This model is consistent with methods descri1:>ed by Leidel et al.(l()! Md Mulhausen and DamiMo,(iI) where professional judgment is used to make inference,> about exposure concentrations and the u~e of estimated surrogates in plaCe, of missing data. The process used in this study is cOI1f!istent with the worker time-activity model described by Ramachandran and Vincent(3) and Ramachandran,\1.2) \vhich is used fol retrospective exposure assessment'>. Figure 2 describes the step-by-step proceSs that determined the e:,timated surrogate value to replace the missing sample data point. Bulk Samples Bulk samples of friction and gasket materials and debris created during the brake and clutch removal process were collected iUld iUlalyzed for asbestos content and mineral type by polarized light micr0scopy (PLM) in accordance with guidelines set forth in EP/l. Method for Deierminalioll (~r Ashestos in Bulk Building Materials. O?) Statistical Methods Descriptive statistical functions .,vere performed to SUlll- marir.,e personal and area data. Industrial hygiene inferential statistics(W,II,13) of the TWA full-shift. personal sample TABLE I. TWA PCME Personal and Area Samples Daily Statistics Perllonal TWA Samples Area Samples Day Data SetA n Min Max Mean Std. Dev. .n 'Min Max Mean AFS 2 ,0.002 0.()41 0.024 0.028 0 APS 1 0.007 8 0.004 O.lXJ6 0.005 1 BFS 3, 0.002 0.033 0.016 0.Ol7 8 0.006 0.008 0,006 2 AJ'S 3 0.005 0.007 0.006 0.001 8 0.005 0.006 0.006 2 APS 0 0 2 BFS 3 0.005 0.007 0.006 0.001 8 0.005 0.006 0.006 3 AFS 2 0.020 0.028 0.024- 0.006 8 0.005 0.022 0.014 3 APS 2 0.ell5 0.046 0.030 0.022 0 3 BFS 4 0.013 0.036 0.024 0,010 8 0.005 0.022 0.014 4 AFS 4 APS 1 0.005 2 0.008 0.009 0.008 0.001 0 8 0.013 0.018 0.014 4 BFS 3 0.005 0.007 0.006 O.OOJ 8 0.007 0.007 0.007 5 AFS 0 0 5 APS 3 0.046 0.055 0.052 0.005 4 0.Q3 0.04 ('.036 5 BFS 6 AFS 3 0.045 0.064 0.053 :I 0.011 0.012 0.01 1 0.010 O.OOJ 8 0.034 0.039 0.036 5 0.007 0.010 (LOOS 6 APS' 6 BFS 1 0.011 4 0.011 0.012 0.011 0.001 3 0.007 0.010 0,008 8 0.007 0.010 {LOOS 7 AFS 7 APS 3 0.014 0.022 0.017 0 0.004 8 0.009 0.022 0.015 0 7 BFS 3 0.Ol4 0.022 0.017 0.004 8 0.009 0.022 G.OIS 8 AI"S 3 0.010 0.017 0.012 0.004 8 b.009 0.019 v.Oll 8 APS 8 BFS 0 , 3 0.010 0.017 0.012 0.004 0 8 0.009 0.019 0.011 9 AFS 9 APS 4 0.008 0.OJ9 0.014 {j 0.005 8 0.008 0,017 O.oJI 0 9 BFS 4 0.008 O.l)l9 0.014 0.005 8 0.008 0.017 O.0l1 Total AFS 0.002 0.041 0.014 0.009 0.005 0.022 0.011 Total APS 0.008 0.055 0.022 0.003 0.004 0.040 0.016 Total BFS 0.002 0.064 CLOIS 0.005 0.005 0.039 0.013 AAFS = TWA fullshilt analyzed data> 390 min; APS = T'>\TA partial shift analyzed data <390 min; BFS = TWA analyzed plus modeled data, Std. De". 0.001 0.001 0.000 0.000 0.005 0.005 0.002 0.000 0.004 0.001 0.001 0,001 0.001 0.004 0.004 0.004 0.004 0.003 0.003 0.002 0.001 0.002 528 Journal of Occupational and Environmental Hygiene July 2007 data and short-tenn 30-min personal sample data were also determined. RESULTS Ventilation The effecti've ventilation rate with all doors closed was 0.13 air changes per hour (ACH) and averaged 4.7 ACH with both overhead repair facility doors open. Background Air Samples Ten samples \vere collected (5 inside/5 outside) but only the inside facility samples wvre analyzed. Results ranged from 0.012-0.025 total fibers/cc by PCM. These sample5. were fUlther analyzed by TEM and no a<;be<;tm fibers were detected (pCME of <0.009 5.bestos f/cc) in any of the five samples. Bulk Samples The removed and replaced brake materials from all four pieces of equipment contained from [5% to 95% chrysotile asbestos. Bulk ~ample analysis for asbestos in brake debrh coUected from all four machines yielded either nondetectable or les& than 1% chry&otile asbestos. Removed and replaced gasket materials positive for asbestos ranged from 3% to 85% drrysotile. Personal and Area Samples A total of 434 personaJ and area samples were collected during work 'Ictiviiles and analyzed: 192 area samples, 101 long-tcnn personal samples, anJ 138 short-term samples. Forty-seven (approximately 10%) of the 434 samples w('re overloaded with paniculate matter and could not be analyzed. Three samples were excluded from the data compilation and statistical analysis due to pump faults that resulted in sample times less than 30 min or sample time unccltainty. Full-Shift and Short- Term Personal Samples f'-ull-shift monitoring periods ranged from 6.5 hr to 9.5 hr, with over half of the samples in the 8-hr time period. Full-shift sample collection ranged from one sample coliected over a full shiJt to collection of five !'amples averaged to determine the I ~ fun-shift 'fWA. From the 101 personal samples, a total of 30 full-shift averaged personal sample<; were expected. For this study, a minimum sampling time of 390 min defined a fl111- shift personal sample. Twenty-one averaged personal <;amplcs met the full-shift criteria and are descrj~d as Dara Set AFS (full shift). The remaining nine personal TWAs were les5. than 390 min due to one or more overloaded samples. The partial- shift samples are described as Data Set APS (partial shift). Sample filters that overloaded with particulate matter or were damaged resulted in inconclusivE' data, which then resulted in T\VA sampJes failing to meet the selected full-<;hift time criteria. To avoid loss of the missing ~arnple data points, a time-activity model (Figure 2) was used to estimate surrogate asbestos fibcrs/cc values. The estimated surrogate value for each missing data point wa<; then entered il1to the PCME TWA calculation. The results of including the surrogate values into the PClvlE TWAs are described as Data S~t Bl'S (full shift). Personal fun-shift TWAs by reM analysis ranged from 0.005 t/cc to 0.049 flce with an arithmetic :lVerhge value of 0.027 t/cc. Pat1ial-shift TWAs by PCM analysis ranged from 0.10 0.09 0.08 u .!:! 0.07 f! .4a1 0.06 u(:j:): 0.05 ~ 0.04 .4a1 (j) c( 0.03 0.02 0.01 0.00 FIGURE 3. TIIIJA personal sample results--asbestos jibe,slee by day ~md data set Nole: The solio cm:.le on the graph moica!es the average value The line extending through the circle Indicates the range of the data AFS indicates Data Set A-TWA data >390 minutes. APS indicates Data Set A-TWA partlalshift data <390 minutes, BFS indicates Data Set 8-TWA fullshlft data including surrogate modeled data. Sample Analysis by NIOSH 7400 followed by NIOSH 7402 Note: The maximum value for Data Se1 BFS IS different from AFS because one of the AFS samples Included an Ol sample that when replaced with the surrogate value, the TWA, was lower. 'Journal of Occupationa! and Environmental Hygiene July 2007 529 <.n eN o TABLE 11. Daily TWA Personal Sample Results with Activity Descriptions (Data Set BFS) Sample Information Results Date Day 1 Day 1 Day 1 Day:? Day 2 Day 2 Day 3 Day) Day 3 Day 3 Day 4 Day 4 Day 4 Day 5 Day 5 Day 5 Person Sampled Mechanic #2 Me-chanic #3 Mechanic #4 Mechanic #2 Mechalllc #3 Mechanic #4 Mechanic #1 Mechanic #2 Mechanic #3 Me~hanic .jf4 Meo2hanic #2 Mechanic #.3 Mechanic .jf4 Mechanic #1 MechaniC' #.3 Mechanic #4 Activity D8 -- Cable control disassembly. 12E-----removed wheels and brake D8 -=-- Removed and replaced right cable control brake 12 E --- Removed and replaced the air filter, exhaust and intake manifolds. valve cover, and engine head D8 - Removed part~ to acce~s the track brakes. Assi~ted with disconnection of the cable control unit from the body of the dozer. 12 ---removed and repla,~ed rear brakes D8 --- Removed and replaced left cable cor!U'ol brake, disconnected cable control unit from body of D8 12 - Rem.oved and replaced rear brakes. cleaned engine head and block, replaced head gasket and engine head, removed radiator connections and gaskets l2E -- Removed and replaced re,1T brakes, cleaned engine head and block, replaced head gasket and engine head, removed radiator connections and gaskets DS --- Removed the left and right track brake bands, replaced left brake band DS - Removed and replaced clutch and cable puller linkage 12E -- Machine fluids replaced, oil filter, crankcase bJeather and air jilter part;, and ga,kets and engi.ne cover replaced. 930 - Removed and replaced the ~vheels, four brake caliper as;;emblies, and brake pads D8 -- Track band brake replacement and adjuf>tment. Assi~ted with lhe removfl1 of clutch lIsing overhead hoist D8 -- Removed parts to access the masler clutch. Assisted with the removal of clutch using overhead hois1 955 --- Removed engine cowling, exbaust manifold, rocker arm assembly, valve cover, engine studs, air intake manifold and nix filler, radiator hose connection, water thermoslat and the engine head aIling with associated gaskets 955 --- Removed parts to gain acceSf> to the rear brake bands. Disassembled and removed the ~teeri.ng dutch linkages and assisted Mechanic #4 D8 --- Removed and replaced dutch drive and driven plate friction linings 955 - Removed gaskets from the intake mamfold, exhaust manifold, valve covel, radiator hose pipe connection, and the oil cooler/water pump. Replaced ferrules, &eals, gaskets and engine Surrogate Estimated Value Time (min)" 0 131 123 0 0 0 0 162 162 (; 350 223 (; 64 155 279 Day 6 Mechamc #- I D8he---adASSisted with repblraackeem, eDnIlI of Rcleumtcohv.e9d55an--dRreepmlaocveeddetlhurteee11leafntdSeidaehbl"rapkllellhearnldinaksa&geembHes (0) Day 6 Mechamc #2 955 --- ReillovcJ hand Day 6 Mechanic #3 D8 - Reinstalled the clutch a~sembly and clutch linkage 0 Day 6 Mech,UJiC' jq 955 -- Reas>embled with gaskets the valve assembly, air intake maniti)ld, exh:mst manifold, I.) lhermostat, oil cooler surface, radialor pipe conneclion surfaces. ilDd air filler iJo;f>emb1e Total Sample Time (min) 490 536 494 509 409 509 405 511 512 504 517 520 493 422 49?- 4~19 33611 43B 574 576 PCMETWA N (Asbestos Fibers!cc) I CU)O:? 4 0.033 3 0.007 2 0.005 2 O.OO? .., 0.005 L. ) 0.028 3 0.036 3 0.013 3 0.020 2 0.1)07 3 O.OOR 2 (l.OO5 5 0.1)45 5 0.052 5 0.064 :., O.Oll 4 0.1,)12 5 0.012 :; iHlll 'o - c ~~ Day 7 Mechanic # L 95r5ad--iaRteomr Ofa\.n-alanodf fleafnt gsiudaerd[).lake bands, replaced tbe oil cooler, radiator pipe, gaskets, and 0 9, Day 7 Mechanic #-3 D8-- Reassembled the clutch linkage and installed the transmission in1er-lock assembly wi1h 0 o(("")) its original gasket 955 -- Assisted mechanil: #1 s::: "0 Day 7 Mechanic #4 DS - Removed the "pony" engine exhaust pipe, pony engine supply tube, water manifold pipe, 0 ~ exhauH and air intake manifold, and cylinder heads. Removed gaskets and cleaned engine cl" ::J head and fianged surfaces ~. bl Day 8 Mechanic #L 955 -- Installed right- and lett-~ide brake banes and steering linkages 0 5. DHy8 Mechanic #-2 955 -- Installed the brake bands on the righl and left side. Reinstalled brake and s1eering 0 m linkages; D& -- Assisted with installation of engine head ::l :So Day?; Mechanic #4 D8 -- Cleaned the exhaust/intake manifolds, front and rear cylinder heads, removed Hnd 0 o::J replaced gaskets 3 Day 9 Mech;mic #1 955 -- Completed steering clutcb and brake linkage inslallation and adjusted brakes. 0 <D ~::J Day 9 Mechanic #2 9D5g5-- ACsosmistpeldetMedesctheaenriincg#c4lutch and blake linkage installation, adjusred blakes; D8 - Assisted 0 I '< Mechanics itl and 4 wilh engine assembly and replaced_exhaust slack (Q w' Day 9 Mechanic #3 D8 --- Drained and leplaced oil, re-i.nsrlliled seats, levers, and steering linkages in cab 0 ::J CD Day 9 Mechanic #4 D8 -- Cleaned valve cover surfaces, instailed leplacernent valve cover gaskets. Reinstalled fuel 0 lines, exhaust and water manifold gaskets, and air filter/i1l!ake pipe. Cleaned the "pony" 's:-:: engine exhaust pipe flange and replaced gasket. Re~lloved and replaced the fuel filtel and -< gaskets N o o liThe amounlof mIle that rhe timf,-activiry model W,lS u~ed ro estimare a surrogatt' va.lue for indu,jon III the T\VA. -...J /J The total arnount of [m~ this mechamc w0rked on Day 6. 564 4 577 4 554 4 547 4 540 4 482 4 489 :; 494 3 502 :; 496 3 0.016 0.014 0.022 0.011 O.otO CI.Ol7 i}'(1l9 0.013 O.OOl-l 0.018 (J1 (JJ --'- TABLE Ill. Short-Term, 30-Min Personal Sample Daily Statistics Day Data Set n MinA Max 1 2 3 4 5 5 6 7 8 9 Total AST Total BST AST AST AST AST AST BST AST AST AST AST 4 <0.044 0.1.146 6 <0.044 0.045 13 <0.044 0.56l 5 <0.043 0.045 l8 <.0.043 OAl3 20 <0.043 0.413 23 < (W43 0.099 30 <0.043 0.170 29 <OJ)44 lU8(~ 9 <0.045 0.045 <0.043 0.561 <0.038 0.56l NNP<' AST indicate~ analyzed data only; BST mdicate, analyzed plus rnodded daw ,~ Mmimum value, were all less than the lImit of JeteLll0n tor the -amp!mg: ll1t'hud. 0.011 f/cc to 0.176 flee with an aritlmletie average value of 0.053 flee. The minimum .rCME value result for data set~ AFS and BFS \vas 0.002 asbestos f/cc, and 0.008 asbestos fice for Data Set APS. The maximum PCME values were 0.041 for Data Set AFS, 0.055 for Data Set A.PS, and 0.064 asbestos fice for Duta Set BFS. The means were similar with Data Set Af"S at 0.014 asbc:'.tos flee, Data Set APS at 0.022 asbestos flee. and Data Set BFS at 0.018 asbestos flee. Table 1 and Figure 3 present the daily statistics for personal PC"ME TWA :'><lmples and Data Sets AFS, APS, and BFS. A comprehen~jve list of work activities and sample int'offilation for fuB-shift TWAs representing Data Set BFS values are pre~ented in Table n. One hundred and thirty-eight shoi1-rerm samples were collecteJ. Of those, 136 sample" were Emalyzed. Two I.amples on Day .5 were owrload('d with pat1iculatc matter and could not be analyzed. The analyzed samples art' designated as Data Set AST (short term). TIle time-activity model was used to estimate surrogate vulues for the two mil.~ing data points and the combination of these values and the aJlalyzed samples are designated as Data Set BST. All short-term samples were collected tor 30 min except for one 34-min sample. Table HI and Flgure 4 present the sample data grouped and analyzed by day. The range of results of the 30-min shot1-tenu samples for Datl Set AST 1:>y PCM analysis WH5. less than the LOD (< O.lM3 flee) to 0.561 flee. The range of results for Data Sets AST and BST by PCJ\1F. analY'ii~ were less than the detection limits of 0.043 and 0.038 asbestos f/cc to 0.561 asbe~tos flcc for both PCt<.-1E data sets. 1'lean valUes were not calculated due to the high fraction of censored data. Gasket and friction-related excursion limit ~amples were grouped and analyzed to detennine expo-;ure concentrations for the various activities. Data were grouped to evaluate the re~u1ts by a variable meh as perceived aggressive activitie<;, including scraping, hammering, riveting. abrasive cutting. and cleaning surfaces using Hdie grinder or compre5.5.ed air. The peM and PCME result). of the sample data grouped 1:>y activity are presented ill Table IV. 1.00 0.80. j 0.60 ... ..-!~ 0.40 -A~~~ c 0.20 0.00 FIGURI: 4. Short-term personal sample results by day. Note: The solid diamond on the graph Indicates the average value The line ex1endlng through the diamond indlca1es the range of the dala. AST indicales Data Set A, BST Indicales Data Set B. Sampie AnalysIs by N!OSH 7400 foilowed by NIOSH 7402 532 Journal of Occupational and 5nvironmel~tal Hygiene July 2007 TABLE IV. Short-Term Personal Samples by PCM and TEM Sample Informationr\ Activity Analysis by NIOSH 7400 Phase Contrast Microscopy (PCM) Analysis by NJOSH 7400 PCM and NIOSI-I 7402 Transmission Electron Microscopy (TEM)B Ne Min. Max. Avg. Std. Dev. Min. Max. Avg. Std. Dev. Friction-Related Work Band brake removal Removal of rlvels and friction lining from brake band or brake shoe Cutting bllild with abrasive disc Removal of disc brake assembly Brake replacement and adjustment Rivet replacement of friction lining to bands or shoes Drilling of brake lining Clutch removal, replacement and adjustment (08 only) Rivet removal from drive and driven plates (D8 only) Cleaning, scraping compressed air cleaning of driVe and driven plates (D8 only) Rivet replacement of friction Iinings to drive and driven plates (D8 only) Gasket-Related Work Removal of head/exhaust and intake manifold/water manifold gaskets Scraping!pmver cleaning of engine head. engine block, and manifold svrfaces Wiping, cleeming parts Use of compressed air t? clean engine pmts Replacement of engine parts and gaskets 17 0.043 0.077 0.048 4 0.044 0.044 O.Ct44 2 0.ll49 0.561 0.305 5 0.0-14 (lO44 lW44 28 0.043 0.104 0.047 4 0.043 0.044 0.044 1 0.038 28 0.043 0.201 0.079 2 0.190 0.230 O.2HJ 4 0.151 0.211 0.180 0.010 0.000 0.362 0.000 0.012 0.001 0.050 0.028 0.026 6 0.048 0.207 0.156 0.056 13 0.044 0.048 0.045 0.001 13 (1045 0.413 0.143 . 0.125 8 0.045 0199 0.065 8 0.045 0.199 O.OM 0.054 0.054 12 0.045 0.074 0.050 0.010 0.043 0.058 (1.045 O.OM 0.044 0.044- 0.049 0.044 0.043 0.043 0.561 0.044 O.()..I.] 0.044 0.305 0.044 0.044 0.044- O.o:l8 0.043 0.099 0.O~8 0.150 0.184 0.167 0.150 0.211 0.170 0.044 0.207 0.127 O.OM 0.045 0.0-1-4 0.045 0.413 0.116 0.045 0.094 0,051 0.045 lW94 O.O~I 0.045 0.045 0.045 0.004 0.000 0.362 0.000 0.001 0.001 0.013 0.024 0.029 0.069 0.000 0,114 0.017 0.017 0.000 4All sample;s Vlere collected for 30 min at apprmmnafely 2 L/ml1l except for one '14-mm <;,lmple rJunng brake rlnlling. BFor all ,ampk, less than the reM LOD, the PCvI LOD is thc~ value u,ed III statistical calculations. C Seyenteen shon-term samples meluded work under tWO activities and are inc [uded m the anaJyses for bmh. Further inferential statistical analysis of personal full-shift long-term TWA (Data Sets A.r:S and RI'S) and the &h6rttern) 30-min data set was conducted. For the short-term 30min data, 96 of the 138 samples were less than the LaD. Inclusion of data points less than the l.OD (censored data) would bias the slatistical results toward the 10\\! end. To be conservative and avoid the bia<; associated with analyzing censored data, only samples above the LOD were induded in the analysis of the short-term 30-min data. By excluding the censored values from the shorHenn, 30-min data set analysis. only the upper bound value~ of the inferential &tatistics are reported. Each data set was determined to be lognormally distributed. The geometric mean (GM). geometric standard deviation (OSD), percenTilc valuc X (95'70), upper rolcnmce limit (95% UTLl, 95% exceedanee fraction point estimate, and 95% upper confidence limit were calculated. The results of these analyses are presented in Table V. Area Samples Seventy-two TWA area samples were calculated from the results of 192 area samples. Of these, 41 area sample results were full-shift TWA samples, with the remainder composed of 31 panial-shift TWA area samples. Sample sets are designated the same as the full-shift personal samples with Data Set AFS representing full-shift TWA samples. Data Set APS representing partial-shift TWA samples, and Data Set BFS, which represented all of the data from the previous two data sets plus the estimated surrogate values for the missing data points. The range of PCM &ample results for Data Set AI:;S \\il~ 0.003 to O.(l47 total t/cc with an arithmetic average value of 0.020 Uce, Tile rangc of PCM sunple re:'!ul[;s for Dum Sct /\PS was 0.003 to 0.066 flee with an arithmetic average value of 0.022 fie..:.. The overall statisti..:.s for the full-shift PCIvlE data sets AFS and BFS were similar with minimum values of 0.005 Journal of Occupational and Environmental Hygiene July 2007 533 TABLE V. Inferential Statistical Analysis-Personal Samples Full-Shift Personal TWA PCME Samples (Data Set AFS jA Full-Shift Personal TW.\ peME Samples (Data Set BFS)B Occupational exposure limit (OEL) flee Number of samp1e~ ~N) Geometric m~an (flee) Geometric stilndard deviation (f/cc) X(95)% (tJee) UTL (95% UCL of X(95%)) (flee) Exceedance fraction point estimate % Exceedance fraction 95S7" lJCUv" 0.1 N=2l 0.0n- 1.98 0.035 0.051 0.07% J.l4% 0.1 N=30 0.013 2.17 0.047 0.074 0.45% 2.50% .4 AFS = TWA fuli-,il1fl analyzed data >390 'mll. BBFS = TWA 1ull-,hift analyzed plus modeled data. CAnaJyzed data from Data Set AST-- data :tbove the LOD onty. Onty upper l'ound staHStlcs reported. Short-Term 30-Min PCME Personal Samplesc 1 N:::::48 0.261 0.359 0.03% 0.23% and 0.005 asbestos f/cc; maximum values of 0.022 and 0.039 asbestos f/ce; and arithmetic average values of 0.011 and 0.013 a<.bestos flee for data <;ets AFS and BFS. respectively. l.'lbJe I and Figure 5 present the PCME data by day. Figure 5 shows that the arithmetic average asbestos flee air level by the cnd of the ~tudy was essentially the same as the begirming asbestos t/ec background level. These re<;ults indicated no buildup of asbestos 11bers in air from one day to the next. All T\VA.. sampl~s were below im 8-hr TWA of 0.1 flee. All shOlt-term 30-min samples were below J f/ce. DISCUSSION T he majority of the activities conducted during thi~ study were, according to the mechanics, typical for the "ervicing required. However. a few unusual practices were also monitored. One unusual practice. the clItting of the brake band. resulted in the highest value for the "hort-term 30-mi.."1 samples (0.561 asbestos flee). The brake band curting took approximately Z.S min and was necessary because the bolts holding the brake band on could not removed. This activity 0.10 0.09 0.08 0.01 .. 0.06 1... 0.05 i 0.04 ! 0.03 0.02 0.01 0.00 q, tJ',+<~ fr ..:~..:~I....~ ' 'l!~'l!~'l~ ! 'b~ ' 'b~ ' 'b~ ' ~'~~'~~~ ' ~'~~~ ' ~~ ' Ib~ ' Ib~ ' Ib~ ' '\~ ' ,~ \" '~ \' ~'~ b' '~ b' I~ ll' I~ ll' I~ ll' qj FIGURE 5. TWA area sample resuils----asbestos I'Ders/cc oy da)" and data set. Note. The solid square on the graph indicates the average value. The line extending through the square indicates the range of the data AFS Indicates Data Set A-TWA data >390 minutes APS indicates Data Set A-TWA parllal-shlft da1a <390 minutes BFS indicates Data Set B-Till/A fu!i-shlft data including surrogate mOdeled values Sample AnalYSIS by NIOSH 7400 followed by NIOSH 7402 534 Journal of Occupational and Environmental Hygiene July 2007 Results PCM (total f/ce) PCME (asbestos f/ce) Work Acth1ities, Analyzed Values. and Surrogate Values Personal Sample Information for Mechanic #4 on Day 5 Sample Time Duration Analyzed Surrogate Full-shift {min) Value Value Value TWA Mechanic removed radiator fan, co~ducted ~craping. and used a die grinder and abra~ive disc on 7:42 to 8:46 64 OLA manifolds and valve cover surf.1ces. Five shont-elm samples collected on this mechanic on [his day included the ~ame or similar activities described for this time period. An average of the four short-rerm samples ~0.413, 0.270, 0.080.0.046, and 0.077 asbestos f/ce) was used as the surrogate value (0.172 asbestos f/cc). . Mechanic used!i die grinder and scrapers/chisels on oil cooler sUlfaces and air inta_ke pipe flange. 8:46 to 9: 18 33 OLA Four short-term samples collected on this mechanic on this day included the same or similar activities described for this time period. An average ohhe four short-tenn samples (0.413, 0.270, 0.080.0.046, and 0.077 asbestos [icc) was used as the surrogate value (0.172 asbestos f/cc). Mechanic used a chis('l, hammer. and die grinder on water pump/head flange, radiator head flange, 9: t 9 to 10:07 48 0.26 0.16 exhaust manifold on engine head. air intake on engine head, valve mechanism assembly, broom cleaned tile floor. wiped down work table, and u~ed a solvent to clean and wipe engine head surface. Mechan.ic used a die grinder and scrapers on the engine block ~Ulface~. Co~ducted parts inventory. 10:07 to 11:48 101 0.019 0.016 Used a grinder on the equipment frame to allow clearance to install engine head. Mechanic replaced head gasket; assemQled fenvies and seals; installed head to engine block; 11:48. to 14:50 182 OLA tighrened engine head bolts with torque wrench; and helped Mechanic #1 remove brake bands for 5 min. Nl) short-term samples were collected on this mechanic during this time period. To determine the surrogate value we took rhe air sample result on Mechanic #1 (0.058 asbestos flee) for the brake \vork and used this valu~ for die 5 min Mechanic 114 assisted him. Lunch break was included during this sample and \vas for 37 min. There was no exposure during lunch. For the remainder of the time (140 min) a value of 0.0 [6 asbestos fiber~/ce was the surrogatevalue. This surrogate value was based on the results of four samples collected on Mechanic #4 on Days 8 and 9 while ]le performed tIle same activities during reassembly of the D8, which ranged from <0.008 to 0.016 asbestos f(cc. Using the time-weighted average (TWA) calculation for the abo\le value the remIr is 0.0 [4 asbestos f/ce for the [82 min. Mechanic completed tightening engine head bolts \vith a torque wrench, replaced valve mechanisms 14:50 to 16:01 71 (1043 0.040 and engine head gasket. Once the sllrrogate values were detennined for th", three OL samples d",scribed above, they were 499 averaged wirh the remainder of the air samples for this mechanic on this day. The total work time was 499 min and th", TWA result was 0.064 asbestos fice. The activities for this mechanic: OIl this day were similar to his activities on Days 7 and 8. Thl' TWA result" for [hose days Wl're 0.022 and 0.017 i3sbe~to~ f/ce. Ba,;ed on compa.ring the T\VA results on Days 7 and 9 to the TWt\ for thj~ mechanic on Day 4. surrogate values were likely bia~ed on the high side. Because the goal was to compare results wirh O. [ flce as an 8-1lf TWA, the bias was reasonable wirhout being unreasonably conservative. 0.l77 0.177 0.014 0.064 w(Jl A OL mdlcat(S ,ample W:l'; over.loaded with particulare matter and could nor be analyzed (Jl was also conducted a second time on another brake band on the D8 and the sample result was 0.049 ashe<;tos flee. Exposures with the highest numerical values occurred on Days I, 3, and 5 of the study. Activities on Day I included removal of brake bands and removal and replacement of rivets and friction lining on the brake bands. Day 3 activities focused on tbe brakes and clutcb systems on the DS, top-end engine rebuild on the 12E, and brake removal and replacement on the, 930. Day 5 activities included DS-rivet removal, cleaning, and replacement on the drive and driven plates; 955-removal and replacement of top-end engine asbesto'>-containing gasket'>. Days 1, 3, and 5 were also the days when 32 air samples (approximately 70% of all the overloaded samples) were overloaded and cOllld not be analyzed. To overcome the inability to include the overloaded sample,> and to reduce uncertainty associated with evaluation of the data, a timeactivity model to estimate surrogate values for the missing data wa~ applied. There w~s also a large data set with multiple samples for each task or similar tasks. As shown in Figure 2, the author~ relied on the analyzed data to assist with detemlination of the surrogate value. Also, for most of the overloaded sample periods, there were one or more short-term 30-min samples collected on the same mechanic during the applicable time frame that were available for use in the model. For each estimate made by following the model described in Figure 2, the authors chose conservative values in recognition that :>hort-term samples, with their higher limits of detection, are likely biased high and are not nece~~aril1 good predictors of long-term exposures.(lO.1~) Therefore, the estimated surrogate result~ are abo likely biased toward higher, exposure valucs. Because the authors were interested in comparing the results with the maximum likely TWAs, it was determined that conservative but reasonable estimations woulJ be pmdent. To determine whether cach surrogate TWA value was reasonable, the result was compared with the TWA result of other mechanics pClil.lrming similar activities. To illustrate the process, Table VJ presents the results of using the time-activity model to estimate values for three sample~ on Mechanic 4 on Day 5 of the project. The p"ttem of results of IWA personal samples in Figure 3 is similar to the pattern of the results for TWA area samples in Figure 5. The similarity in patterns was expected and indicates that the area samples, though much lower in concentration, are reflective of the work activities conducted on each day of the study. Results also indicate that the average value of the ShOliterm samples over the 9-day study were consistent as shown in Figure 4. The PCJvIE mean and range for each data set Oil each day and demonstrate the consistency of these values throughout th4 study. However, the authors would caution users of the timeactivity model for replacement of missing data that there may not always be this level of cO!1."istency for all data "ets. Two pub1iwed studies proVided extensive PCM totaliiber data regarding airborne exposures during maintenance and brake lining replacement in automobiles and other types of vehicles.(I5,IAl Neither of the published studies included analysis to confirm fib~r type in the air samples. The rCM total fiber results of our &tuJy were \vithin the ~xpo&ure range of these previously published ~tudies. in comparison, a number of published )tudie~ provided a smaller volume of data regarding exposures to asbestos while removing ,md replacing gaskets. Although not all of the studies provided TEM analysis results, some did provide 8-hr TWAs. The results of this study \vere consistent with the previously published studies<-5. 17-22) where 8-hr TWAs by peM or fibcr identificati.on was included a~ part of sample analyses using NIOSH method 7402. CONCLUSIONS. R epair and involving maintenance of heavy constmction equipment asbestos-containing friction product'> and gas- kets resulted in exposure concentrati.ons below 0.11/1:(" as an 8-hr TWA and below 1 flee as a 30-mm short-term sample. Average asbestos flee air concentrations in the repair facility work environment by the end of tilt' 9-day study were essentially the same as background levels measured before the study, indicating no buildup of asbestos fibcrs in air from one day to the next. A time-activity model approach to estimate sunogate vallles for replacement of missing data i., an option to consider for exposure assessments when the ability to resample activities is not possible. ACKNOWLEDGMENTS T he authors wish to thank Caterpillar, Inc., for funding this study. Caterpillar is a defendanr in asbesros-relat('"d perwnal injury claims. Caterpil1:rr did not influence the exposure assessment protocols tbat were developed or conducted. We also want to thank the team of professionals who assisted with the stUdy and the professional and diligent heavy equipment mechanics of Hoerr Machinery for sharing their knowledge, skill, and experience. REFERENCES 1. Ramachandran, G.: Retrosp;dlV" expo;ure asst ssm"m usmg Bayeslan method~, Ann O{-{ul', Hyg, .,t5{l5j651-iJii7 (2001, 2. Rann:hantlran~ (;.: OC(u{J(Ju()nal EX{Jo.Ha-e A'i.\{!\..~ment.f()J Air C{lntafJli/lams. Boca Raton, Fla.. eRe Press, Tayior & Flancis Group. 2005 pp. 30R-.~09. 3, Ramachandran, G., and .J.H. Vincent: A Baye-ian approach to retrospective exposure a.,se,sment. Appl Occup Em'mm, !fyg, !4:5J.7--557 re..! (19991- 11. American Society for Te~ting and Materials ( \.STM): Srandard !v(('tJlOd/or D.!t.!rmininr; A~rCh(utf.e ina Sing/{' Z.)nc hy A4.ean~ ofa TrrJeY; Ga, Dilirti(ln (ENI-O()). [Standard] W'f.( Com'lohOi:ken, Pa,' ASTM lnternationaI, 2006 5, Boeltel, EW.. D.M. I'odraza, llnd G.N. Crawford: Airborne fibcr exposure a;se,sment of dry asbestos.contammg gasket; and p,,,kmgs tound in intact indusrdal and maritime fitrmgs. tllHA .J, (6.3)'732-740 (2002) 536 Journal of Occupational and Environmental Hygiene July 2007 6. Research TI'ian~le institute ,RTl): Aleawnng Airhorne Ashe,If[,,\ 1:1,1iowil'Jg an Ahatcmcnt Auwll (tJ.S. EPA 600/4-1'5-(49). Re,earch Triang!e PMk, N C: RTllntt'lTIational, 19R5. 7 Environmental Protection AgenC)" GUIdance for Cu;;troiii'lg Asl)e,lt,),, COlltJming Materials in BllIldlng~ (D.S EPA 500/5-R5-025) Washington. D.e.: lJSEPA. 191'5. !<. National Institute 01' Occupational Safety and Health: A;bes~os and other tibers by peM, NIOSH Method 7400. In NJOSH M!)nual of A.nalyticallvlethodl, 4th Edition. Cmcmnali, Ohio: NIOSH. 1994. 9. Nationallnstitutc of Occupational Safcty and Health: Asbestos and other fibers by TEM. NJOSH Method 7402. In NIOSH Manllal of AnalytiL'al Methods, 4th Editlon. Cin-:innati, Ohio: NIOSH, 1994. JO. Licdcl. N.A., KA. Busch, and J.R. Lynch: Oc,upanonal E).[losufP Samphng Strategy Manual. DEHW (NIOSH) PUb. No. 77-173 Cincmnall. Ohio' 1977. l1. Mulhauscn, .I.R., and J. Oamiano: A Slnlll'gy If!/' Asse,'SIIIi! and Managmg OCcu{Jafllmal E.~po.mre,\, 2nd Edillon. F3.lrtax. Va: AIHA Pre~<;, 199R. pp. 07 , oS, U l. J2. Perkins, R.L., and H. W. Hal'vey: lv/ethnd 101' {)ctcl"fllinatlOll of !\.lhesl'" m Bulk Buzlding Marerials (EPA!600/R93-116 Test Method), 1993. Resealch Tnang!e Park, N.C.: RTI International. 19QJ. 13. European Committee 1'01' St:mdardization (CEN): \t~)rk(!lace Atmosphere,I'---Gllldance Jur IIU' AIsessmenr of E"p{lslIre by Inhalation to Chemical Agents j;n' Comparil.on waft l..,Iml[ vcLiues and iviea.\urem(>nt Strategy lEN 689'19<:15) LStandard]. Dublin. Irdand: CEN. Fd,ruary 1995. 14. HiNardi. S.R.: The (),('uparional Emir.onmr-nt: Its Evaluation, C{lntml, ,md Managemellt, 2nd Edition. Fairttx., \'3.. ' AIHA Prc"s, 20m. pp. 126. 127. 15. Pauhtenbach, 0 . .1... R.O. Richter. B.L. Finley, and Po.l. Sheeh\.\n~ 'An ",valuatio,n of the hIStOrical exposures ot mechanic~ to as be,to, m brake dmt. Al'l'i. Uccul'. Em'iron. Hyg. (18j'7R6-~04 (2003) 10. Nichol,on, W.J., S.M. Daum, W.V. Lorimi'f, H, Velez, and R, Lili,: Imcsngati{ln of Hcultl! !la:ard~ if! B"uke LIIlIn;! Repml" and Maintenance WOl-(crs Occupationaiiy E'J'(l<ed to AlOe'.",).\', 1982 Cmcmnatl. Oh\O: National Instiwte for Occupational Safety and Health, Divlsion of Surv{'t!!,lIlcr, H:tzard EvulmtJOns. anrJ FIeld StudIes (DSHEFS" 1982 17. Chcng R.T., and l-LJ.. l\tlcDcrmott: Exposure to a,;hestos from asbesfo) 1 ga,ket,;. ApJli OCClljJ. Envil'On. Hyg. 6:58fi-,)91 (1991). 18 Spence, S.K.. and p.s..J. Rocchi: Ex.posure to asbestos fibers dunng gasket removal. :\nn O('(up H:vg ,,0(5) ~R3-5Rb (1990'1. j 0. Liukonen, LR" and E W. Weir: Aslxw); cx.posmc IrOlll ga,kcls dunn;; disa'sembly of a medium duty die,e1 engine. Reg,!l TOJ.!Col. Pharmawl. 4i:ll'J-11J (200')). 20. Paustenbach, D.J., A.K. Madl, E. Donovan, K. Cl:wk, K. hhling, and T.e. Lee: Chry<;otJJe a<;bes[os expo~ure a~sociated wIth remo-val of lmtomobi!c exh:mst systr;ns (ca. 1'145-197.") 1') mechaniCS. Re,;ults ot a ;lmulation ,tuJy. .T. E.1pO final. Enl'imn Ep!demJOi J6(2j-1S6--171 (20001. 21. Blake. C.L.. G.S. Dotson, and R.D. Hal'bi~on: As>cs,mem vi aubornc a,bestos c'lpv,urc during the servicmg and handlmg vi automobi!e :i.sbestos-conraHung gaskf'ts. Rpf.u! 'jiH1C.)l. Ph;Jrmacol. -I:o(2):214-222 (:2006) 22. Mangold, C K. Oarl., A. Madl, and D. Pau,tcnbach: An exposure ,tudy of bystdnd.ols and workas during the lll'llIJlallon and removal 01 a,bestos gaskels :md packmg. .T Occu,n. EI11'11"O'1. Hyg 3.87 -98 (2006) Journal of Occupationai and Environmental Hygiene July 2007 537