Document KRkErDJMEJXQmwyDre2GaM4gQ

Comments- Concerning- the Atmospheric Chemistry of Vinvl ChloriJttRoger Atkinson A. Vinyl Chloride. OH Radical Reaction. In addition to the flash photolysis-resonance-fluorescence data, of Perry et al., Liu and coworkers (A. Liu, W. A. Hulac and C. D. Jonah, J. Phys. Chem., 4092-4094, 1989) have used a pulsedi radiolysis-resonance- absorption method to determine absolute rate constants* for- the gas-phase: reaction of the OH radical with vinyl chloride over-the temperature range: 313-1173 K in the presence of 1 atmospnere of argon-diluent. The rate constants obtained by Liu et al. over the temperature range common to the Liu et al. and Perry et al. studies (313-423 K) are in good agreement with those of Perry and coworkers. A product study of the gas-phase reaction of the OK radical, with vinyl chloride, in the presence of N0X, has recently been carried out by Tuazon et al. (. C. Tuazon, R. Atkinson, S. H. Aschmana, H. A. Goodman and A. M. Winer, Int. J. Chem. Kinet., 20, 241-265, 198S) using long pathlength Fourier transform infrared (FT-IR) absorption spectroscopy to monitor the reactants and products in irradiated ethyls nitrite - MO vinyl chloride - air mixtures in the presence and absence of ethane (used. to scavenge any chlorine atoms produced from the OH radical reaction). The major products observed were formaldehyde (HCHO) and. formyl chloride (HC(O)Cl), with the measured yields (corrected for secondary reactions of these products with the OH radical) being 0.96 and 0.83, respectively, in the presence of ethane and 0.89 and 0.80, respectively, in the absence of ethane. These product yield data show that HCHO plus-HC(0)Cl account, foressentially all of the vinyl chloride reacted, and that: Cl atom production in this OH radical reaction with vinyl chloride is minor, at most. These. data then agree with the reaction sequence shown on page A-46 of the vinyl chloride document. oooo^s SPl-05636 t July 20, 1989 Dp. Richard Corey Toxic Air Contaminant Identification Branch California Air Resources Board 1102 Q Street P.0. Box 2815 Sacramento, CA 95812 Dear Dr. Corey: As promised in our telephone conversation of June 10, 1989, 1 enclose: comments concerning the atmospheric chemistry of vinyl ctoloride, trichloroethene and tetrachloroethene (perchloroetnene). I hope that these comments are of use to you. Roger Atkinson Research Chemist OOGOZ7- SPI-05635  M2-- I. F.'H. Pt,imurib Mtatstryof AfTCTUuienFitlikiaaonmrnoiMWIIlrSutYrr- ofVi^OiondtCeMRro< w*i"**Oilonir(drF(i?d Conus and of Foods. H.MSO. London Memoir R. R,, Pettrr J. M 4 Johnson-M. N (19741. Proportional monaniy amonf-vinyl chloride* workers. Lancet i. 397. SAS (Nauooih-Academy of Sornenr I!9*0) DrMkmp- Ham mi Htaitk. Vot. 3. p. 31. National Academy-Press. Wathnfton.: Nieholsoo.w. J,, Hamaood. E. C. Minie H. k. Sdikoff 1. i. (1973). Monabty rtperunei o( a cohort-of vmyi chlonde.-polyviByl chiendc-workers. tanret h.4197. Nicboboo W. J_ HimialMi|(i Pr K. k Tan-D. (1994). Trad* m oncer moruhry usoaf ewtev ia the lyatheac potymtTT industry. la Jmutinal Hasarit af Hama and Svwnrttr ffarunurr. p. fcJ. Alan R. Liai. New York. ORC (19741. Mortality data coUacudby ORC canoermat th* eftets af vinyl chlendt exposure a PVC fabrication. Otad by Bair<)9*2l On M. G. Ttnrnrr R. R. 4 Holder B. H. (19751. Vinyl chloride exposureia a comrolled adtmnal enttroaaem. Areki tmrm. HUM 30. 333. Raamif V. Gotbe R. k Wacbtmenter C. A. (1974). Th* matafeaieny of chlorocthytent- oxide. ehloroacetaldehydes 2-chloroethanoi aad ehloroaceoc add. conceivable mctabobia of vinyl chloride. Ckmaca-Bial. hurractmi 21. 221. Real W. 4 Wtber-H. (19741. Stand der epidemiolopachtg Forsehust uber die Vraylchlond-knakhat. Zjnubi. Art>Mad. AriSdaus 24. 9*. Real W_ Wtoe- H. k Grtacr E. (197*1. Epidcmioiofy suidy of the manahry of workers exposed to vuiyi chloride ta FRG. Paper presented at the 19th International Conference for Occupational Health. Duorovsic. Sepusher. Schseidesnan M. A.. Mantel N. A Brewr C. C (1975). rrwa room io win how to fet froa the Uhonton* tc Part Avenue and 39th Street .4in. .V.Y. Acad. Set. 244. Z57. Sebkoff 1. J. f19751. Toxisrv of vinyl ehioride.-potvvmvl ehionde. Am. S.Y. Acad Sci. 244. SptnaJUB.riaaaabiRiHf7fl`Anfioaamiimaf.i in vmvi chloride poiwmvt chlonde-worker*. J. nvy tfed 20.427. Swan J- Orermsirt 4 Valesfco M. (19431 Conintwmom to the study df dneasr by vinyl chloride. Stri. /memo IS. 947 (in Italians Szsdkowski 0. k Uhnerr-G. (19121. VmvleHlond alt Knakhenstusache: Eine Btbfeofraploc: VKE. Frankfurt. Tabertbaw J. R. k Gaflrr W R. (1974t Mortality study of workers at the- aieaafacmee-of vmyi chlondc aad iu pollinen: J. wruy. Mot 14. 309. Thenault G. (1912). Cancer mortality of Canadian workers exposed to VCM: a ihrwycar followup. J. eccup. Hat. 24. 730. Thenault G. 4 Allard P. (19(11. Cancer mortality of a pond of CaBadiamworker*. exposed to vmyi chlondc monomer: J. empt Med. 22.671. TorkdsocT. R_ OfeirF.fcRow-V. R. (19611 The toeioty of vmyi chlondrsdeuniuiuil by repeated exposure- of laboraiorranissals--Saw ndr/fiy. Att. J. 22.254. US PHC*'--Deparuaear af Health. Education aad Wel fare (1910). Sebeted abnnets oa the araaofematy of VCM. OHEW. WeihiaftOB. DC Verborft F. G. 4 Vo**l E. (1977). Vmyi chlondc awa reness m DrotepiutaatHpaapanir. Maiaiica Act. 4L 327. Viola P. I_ (1949). PaihotOfT of vinyl ehloride. Proceoainp of the 14th International Coopts oa Oeeupauonal Health. Tokvo. Wataaabe-P. G. 4 Gchriar P. J. (1974). Oose-depeadent fate of vinyl chlondc- aad tu posxfme relauoaship to OBCDfemdry ia m. Laa. Htik Parpen. 17. 142. Wtawcucr R_J_ Stnaprr W. Wapoaer J. IL Jobs J- Falk. H. k Carter C 119741 Neoplastic risk amonr workers exposed to vmyi chionoe. Aim. S.Y. Acad. Sci. 271. 40. Weber H_ Retai W. k Grieser t (19*11 Gersas vesapaneas on somtdjty and mortality of workers eapesed to vmyi chionce. Litrw. Hhh prrjptct. 41. ?<. Wilson R. H. McCosck W E_ Tatum C. F. & Ctessh J. L. M94T). Ooeupatioaal acro-osteolyss. J. Am. mtd Ajt. 201. *77. P Ei Rtr 00002S SPI-05634 tz Vinyiehlnnds wtn n nd 4rknbdrwmn Weihawk Or M TTiomas lor hit hetp- Food-Safety'Counai (1960k PrrpoM*- Svsiem-for Foods <n >ih me-calculations im Dr D M Conning for hit hetp- SafetyAssessment: Final Rtpon F5C. W ashington. DC. on nh the manutcnpt |Alio.m- Cosmrf; Toxtcdl. 1910. IL 7)|) of Fonnan' D- Bcbdcb B-- Stafford J. it Doll R. (19161. Vinyl ,,`ly CfTRCNCXT chlonde and.angtosarcom* of the bver--e repon-of ihe- to otv regmerof caacsu Jr. J. ind. Med. 42-750. ACCIH--Aipencan- Cwfrmoprf OwinuBomt Indus-- For A. J. k Collier P. F. (197T). Mortality esperwnee-of trial Hyptanu f 1974). Threshold Limit Valuot for Chem* workers-expoted to nnyl chlonde monomer me nanu-- '.at icaL Substances-and Physical-Agra*,m the Workroom- faeiurrof polvvmvl chionde m Great Bntam. Jr- aid.; .he Eawroiuwrtmh lntroaad QM|far 1974. ACCIH. Med. 14. 1. 'X- Cinrmnan. OH. Frmet-Beyme- R.. Sdumtz T. It Thicss. A-M. (1971). AndtnomM. W.. Hod D. C. A Kaplan N. U (1910V. A Mor^htitastudie ber VCFN'C Arbenerne- der BASF -ate ptrnl irheme for the incarporatw oi pharmacokiacm^ Akumgeielhcaifu Ludwigthafcn am Rhem- Araruxmrdt: re* a low-dose ntk csussaooe for rhnmcal taraaogcDcssr Secialmrd, Prtrrmrmtd. IX 21*. the exempli *myl ehlondor Texte. gppL Pharmec. SS. 134. Gauvaua X (1976). Vinvf eklando Free; JL Sec. Med. 69,- *. of "be ou- Banc* A. W. (1976V Viov) chlonde and the production of PVC Aror A. Sot; Med. 69. 27T. Birr J. T. f 1912 V. Risk usanmcnr far vinyl rhlondem puipecuvt. Pressnuxt n the75th Annas! Maetsng of the 275. Gaytor D. WV t Kodcll R- l_ (19WV > inmpolauog-. aigomhsr for low-doirr-nsk anessnestrof toxic sub-- sianoecJ. eunrr Oatlc Toxicol. A 305. cal .Air Pollution Control Association. Sew Orleans. USA. Gchnnf P. J- W'aunabc-P. G. X Park C N. (197*). on June >9*2. Retoluuoo of dote-respoase touaty dau for chenurmls- -;ge Baumoet J. J. k Bresiew N. & (1911V Power eonsid- requmarmetabobe aaivauoa: example--vmyi chlonde- on enoeat in eprOcnialofie (tndia of vinyl chloride work* Tnes mwewd <|] J in .nd eri* .ain .ear. css ra AmrJ. Epidem. 114, 722. BcrtuffP. A_ Villa A_ Fna V_ Sail B_ Febri 1_ Mapp C. MircrrC_ Manno M_ March! M. & Botuaso F. M. (1979). An cpidcsmolopcnl study of vinyl chloride ex posed workerr m Italy. Arate kirrada toxtcoL 20. 379. Bufier P. A. Wood S_ EiSsr C-- Suarez L. it Kalian D. J. 11979L.Monthly expcncBcn of wortcnmivmyl chloride GebmirP- 5- W'atanabe-P: G. it Park C N. (1979). Risk, of aagmtucoma in workers exposed to vinyl chlonde as predicted from studies in rats. Toxic. appL Pharmec. 49, 15. Green-T. it Hathwiy D. E. (1975). The biological fate in rats of vinyl chlonde- in relation to its oncogonsiy. GttodeoSioL Imcrocnont 1L 545. monomer preouenon plant. J. oeatp. Med. 2X. 195. Gicul T. It Hathway D. E. (1977V The and sts Byres D. Fiighoi C. Fnfiund A. it W'tmcrfaoim P. (1976V btegcncss of the S-contaisiag mrobolitas ef vttyi cdio- '.bis Mortality and cancer morbidity in a group of Swedish ridc is ms. Chemice .Biol. Interaction! 17. 137. csis VCM and PVC proposes,workers. mr. Hhh Perspea. Hams D. K.. k Adams W'. G. r. (1967). Acro-orhoiyxii * i. IT. 147. oenmag is stcc csgagee.m me polymenssnee of yi :S Csribcrj F. W. (19*! V Dose-response functions m caremo- chiorice. Sr. mtd. J. 3. 712 |SB# and die Weifaull model. Fa Conner. Tczicol. 19. lARC Working Group (1979V Monofrapit! on nr Esolu* -.'SC T.) 255. Cuass A Femur L. D. (19*11. Mertainy among PVC-fabncaiinj employees. Enrtr. Hhh Perspea. 41. 13*. ttJtctt of the Coranotemc Aitk of Chemical! to Hienent. Vol. 19. Some Menomrrr. Pissnss aid Syr.thcnc ZaJlomerr wtd Acrotem p. 377. iDiemsttanal Agency for Chiaase L_ W'arj O- Nieholt W. E. t Fcrrnce L. D. (!90). Research on Csncce. Lyon. :n Breast cancer tnortalttv among PVC fabneatets. J. ocaip. Infanta P. F. (19*IV Observations of the she specific .-.e- SUe. 2X 677. armnogcRiOty of vinyl chionoe to humans. Znctr. Hhh -.ot Conference to Reevaluate the Tcxioty of Vinyl Chloride Penpect. 41. 19. ::r Monomer. Polyivmyl Chionoet and Structural .Analogs Kusnaeh A. M. & McGaughy (1975). Quasuuuvt Risk (19611. Conference sponsored by N1EH5 XIOSH OSH.A .Astetsmenr for Community Exposure to Vinyi Cfiionde. at N7H. Bcthesca. March I960. Srstr. Hhh Perspea. US EPA Report. 5 Decernper. EP.A. W'ashinpos. DC. on -e 1961. 41. 1-231. Cook W A- Grtver P M.. Disnaan B. D. k Magnnson H. J. 11971). Occupational acro-osteoiysn II. An i.-.sus- Lee C. C.. Bhandan J. C- W'ieston J. M.. House w'. B.. Dixon R. C. k Wooes J. S. (197J). Carenogesutty of vtnvl chlonde and virviietne chlcnoe. J. Toxic, emir. tnai hvpene study. Areht mar. Hhh 22. 74 Hhh 4. 15. :d Cooper w C. (1981). Eptdetniolopeai study of vtnvl chlo Lester D.. Greenoert L. A. k Adams W. R. (19631. E5ecu on ric* workers: mortality through December jf. 1972. of single and repealed exposures ef humans and reu io .be tnrsr. Hlth Perspea. 41, 101. vinyl chionoe. Am. me. Hyf. Ass. J. 24. 265. er Creech J. L it Johnson M. N. (19*4). Aopesareoma of the Lopncno N.. Ba\-alc R.. BareneeUi S.. Baruch H- Brouxetu liver ta the manufacture of PVC. J. oecup. Med. 16. 150. G- Gammelhni A.. Com C_ Frexa D- S'ien R.. Lcporai Cnimp K. S. 4: Guess H. A. fl9*0). Drtnkmg Water and C_ Roseliini D. k Rots A. M. (1S77). Inducuon of gese- : of .iih ;a oiic mg i Cancer. Report no. PBIM2S167. NTIS. Washington. Duck B. W.. Caner J. T. L Combes E. J. (1975). Mortality study of workers m a polwmv! chloride producuon plant. Lancet ii. 1197. EPA (1910). Ambient Water- Quality Criteria for Vinvl Chlonds Environmental Protection Agcncv Report. mtltafens sod gene eonvemons by vinyl chionoe metab olites m yeast. Cancer Set. 36. 253. Maltoni C- Lefemine G- dibem A- Coni G. k Carretu D. (1980V Epidemioiogw animale et epidetmolope hu mane 1c eaa de chlorurt de vinyl menomere. In XXe JUtmwn- de Club de Cenetrofentse Chimiaue. ISBN EPA 440'$.((M)7g (October). 2*6315.0075. p. 15. Publications Essenueiles. Pans. : of Ftron V. J_ Henonksea C F. M.. Speek A. J.. Til H. P. i Maltoni C- Lefemme G- CHibeni A.. Com G. k Ca.-Tttti .he Spit fi. 1. (19SI). Lifespan or*) toxic,ty uuOy of vinyl D. 09*1). Carasogeiuary biossssys of ttnyi chlonde mg chloride m ms. fd Coanei. Toxicol. 19, 317. monomer i model of nsk assessment on an experimental the FDaiov* V. S,, Antonyuzhcnko V. A.. Srauievich V. B- basis. Eitctr. Hhh Pmpect. 41. 3. .nd ou- Fedoiova 1. V.. Kryzhanovjkav* N. A.. Bochkareva Maltoni C- A RondineUa R- (19*0). Hepatic anposarcotna T. v.. Goryicoeva LA.i Bul-buJysn M. A. (19*2). y^e is workers exposed to vinyl chlonde in Italy. Acta '.U blastomogenic tuzire of vinyl chlonde la elimco-hyrene- Oncolofice 1. 35 fin Italian). 14^1 ^Ica>e**lc*^ study). Gif. Trudo. prof. ZadaL Meysrsotr 1_ B- k Meer G. C. (1972V Cutaneous lenons m' acco-osieol>-ns. Arens Derm- 100. 224. r cr 0Q0O2S SFI-05633 :oo I. F H. FV*0*! *4- (el The-UK it nor typicai-of the-orldwide- beiweeisrVCM exposurwand ASL in man: ASL anO growth m (he- exposcebpepuiauon. neoptasmu'of a nuraPer-of otner organa have Pretr (d) No account. hat-beartaken of piani im induced:in laboratory rodents bv VCM. Estimation provements occurrmf-priorto I9M and hencr- of the exposure-levels likely to causae lifetime nsk of fewtt-cases: may- occur-in. for example.- the- ASL of 10'* on the basts of these-dan give extremety 1910-3000 penod than arr estimated. from-the- low-levtirldown to 3.9 * 10'ppb) which appear to 19*0-19(0 experience: be-unrealistic, estimates for man. Parr of (he reason for this is that, laboratory studies have-shown that An assumpuonihat the-mk of ASL erased m 1964-- VCM is metabolized in the hvet-tend elsewhemn the- ratherthan m 1974 results in a considerable reduction in the estimateof futurrcaset. For mhertiiumpnom body) to the reacuve metabodies chlorocthylene ox ide and. chloroaceuidebyda The rate of conversion is the number-of nee cater observed-annually thouidsoon beprr to decline-and the-rate of decline will limited at high levels of exposure gmng inaccurate estimates, of the slope:- ot the dose-response- re indicate,which assumption: is nearer-to the truth. lationship: It has not been possible to estimate' the - Tberc-have* bear two other-predicbonr of the- rate of conversoirm man. and hence extrapolation of numberof casesof ASLlikeiy to result fronrpmious. these-lo-nsk_ dose estimates a conjenuraL The exposurmto VCM. Nicholson et el. (19(4) suggese second:pareof the. problcm-of extrapolation at low that-therewiU be-a further-1500 cases of ASL. whiter risk is the selection of the most suitable mathematical Forman aL (1986) conclude, that a further 150-200 model for extrapolation. Using Maltoni's data from deaths might be expected over the next 30 years. Our rats (Maltoni et al. 1981). them a substantial range estimates rely on a more sophisticated model than the - (up to 10*) of low-risk dose estimates: depending on latter estimate and on a larferdata set than the* the mathematical modeland the assumptions used in former-Nevertheless, the conclusions of Formas-er applying-the models- Using the same-`(probn and al. (1986) are similar to ours. Only the experience of log-dose) model and different sub-sets of experi the next few yean will show- which is the best mental data, a large-range of estimates is apin estimate. obtained, even- after correction, for the non-ltnear t 0f metabolism at high dose (which reeuers this range to about 10-). Larger differences are Summary and coodusioa* obtained with olcnladoas using the WeibuU analysis as a haais oflow-dose ashmanon. suggesting that this There is unit doubt that exposure to high levels of is a presietr with the use of mathematical models VCM as a consequence of oeeupauon can resuh in as rather than one associated with the iog-probu anal increased incidence of ASL. A review of 20 epi ysis. Although, them was csssidscable variability tti demiological studies involving about 4;.000 workers the dose-response relationship in the different experi ocaipanonaliy exposed to VCM showed that neo ments reported, in all cases a total meubohzed dose plasms of the liver showed an increase in incidence m of J x 10s ug (equivalent to inhalation of 200 ppm) the majority of studies. For brain cancer me associ was required to produce an elevation in ASL inci ation between exposure to VCM and an increased dence. This dose represents a practical threshold in incidence was less dear because of the lower relative rodents. At this stage a ibar development, mathe risk. Neoplasms of the respiratory tract, digestive matical models for low-nsk dose estimates are not system lymphatic tad haemopoieue system buccal sufficiently reliable or reproducible to en gender cavity and pharynx, cardiovascular* system and confidence m their use. colon stomach were reported to show an increased Using negative epicemiolepeil studies of popu incidence m cne or more- studies, hut to show no lations living in the vicinity of VCM production increase, or m some eases a decrease, m madenee in other studies. In view of the increased incidence of breast neoplasms in rodents exposed to VCM. the studies of Chaise.et el. (1980). who did not unarm these findings in humans, are of importance. The register of ASL cases now contains records of facilities, an estimate of the cose for a 10~* lifeume risk in man may be made (Barr. 1982). Tns value (lOOppb) is similar to the highest estimates derived from* data, and taking biotransfonr.auon data, into account, is suanaauxily larger than the lowest esumates. which are up to I O'* lower 99 persons with unarmed .ASL and occupational exposure to VCM. The average latent period between first exposure to VCM and death from ASL is 21.9 yearn. The matority of cases occurred in autoclave workers, who ere recognized as hating been exposed to extremely nigh levels. Although precise timeter of exposure are not available for"the periods of most (3.9 x 10*' ppb using a multi-hit model). The higher esumates are compatible with oceupauonal experi ence and suggest that the current hygiene-standard of around. 1 ppm is sufficiently low to protea the health of VCM .-'PVC workers. The esumates also pve a considerable safety fanor for the general public consuming PVC-packed food and drink or living interest, the patient of cases roughly suggests that extremely high exposures were necessary for the inducuon of ASL. For example. ASL a tended to occur in larger numoers in some plants than in others, a finding that can be explained most easily by differences m exposure patterns. There a an extensive senes of animal studies on the carcinogenicity of VCM. Some of these precede the epiceaiolopcai. studies confinmag the association near VCM PVC facilities. It has been possible to provide a crude estimate of the number of cases of ASL that may occur in the future from exposure to VCM pnor to 1974. Using the age structure: of employees m one company, the total number of cases of ASL reported to date and the mortality pattern expected frem a normal populauots: iht-possible: future numberof ASL cases has bear stzsiaied.as.tn the repos of 150-300. 000024 SPI-05632 t Vmvt ehlenoe--iter u T*tr-> mm--iiiin< A$LcaM*idaw So < ASL catn oworoc Y*t )t!3 7 IH) l 4 7 i ir* 2 j 4 7 i i t*so 1 > Total... SlMT Eimom I 1 l \ 4 1 3 6 4 7 s 4 4 c Son* J 1 l 1 3 2 A 2 4 j3 i 4 0 31* to* of worti \ X *1 i 0 1 Aooool auu 1 i 1 ] 3 1 2 3 I 3 11 U 9 19 s 10 4 0 ASL - AfW at Unr k*rDom am MM usil (dill akw a IH2). |iuim-->*' Cy^iiw- toy soul 1 *3 4 4 7 12 15 j: VM* 2 33 Mthem 31 Onndnrk- 49 40 9 n 44 9 91 9* 33 nth-- 003 IwMtaa Ilia) ba oora taOl (bai------ ' Vi Use association between aSL and VCM exposure. A hypoiheucal exposed population of 100.000 has been used, but this is unimportant (tee (a) below). An esszcaie of the age dtssnaution within the hypothet ical 'total' exposed population of 100.000 has best based on ITK dau (Fax it Cottier. 1977). For persons already exposed duals; the whole of the various latent periods, the numbers with a latency of 30 yean or more forts only a small proportion of the total. Toe numbers of persons at hsk in the future are calculated by advancing ume in 5-year penods talcing account of the age-dependent death rates in the population at large. Death rates for an intermediate year for the male population of England and Wales have bees used in this calculation and the future cases (column 10) have been obtained by muluplieauon. The incidence figures for long latent penocs (> 25 years) are unreliable or aon-existent but those for latencies of 15-25 yean are fairly constant and values of 0.5 and 0.8 cases 1000 persons have been used for all latency penods over 15 yean to calculate the expected number of eases for the 196* and. 1974 assumptions. The calculation is unrealistic m many respects but the simplifications are unlikely to affect the estimate of future cases by more than a small factor, For example: (a) The population size used for the calculation is probably larger than Use exposed population, but the calculation depends on the rauo of "person-years to come" and "person-vein ex perienced" and this rauo the ume for aoy population size. (b i Exposure level has beempiored. The calcu lations are based on the overall risk to the cohort and although the inodenes figures for sub-cohorts could be higher; the estimate of future eases will change vert" little. Similarly duration of exposure has been ignored. Table 13. Hypo-- eUeuiiutt of fviun ASL cun win{ iw &5cvm Msuaipami toout tbe dttt u vtuefe the Imk baarlm at Cakuteuoes utumnt ft* mk t/wt !944 Ctkttituom uttomac ao mk a/tcr I9"4 Lamer (vrf Casa lo diu ft.HIM >1 5-vr na w date mettcaet fsturr pmont tt rttk Ftnvrt aw Fcnoat at ntk m tiu Syr iso*CTCt Future pnon at mk Future CUB 1-3 4-10 11-13 lfr-20 :i-:* 24-30 31-33 34-0 41-43 30 51If- 0 100.000 0.00 00 1 eidso 0.01 00 11 95J0O 0-12 00 'J W.730 oji k7J0 31 41.400 0.46 34.330 u II 34.730 049 41.730 20 4 21ZX o.a 41.100 13 4 0 ** 600 0.19 31.730 * 45.730 n 00 34.300 1 0 0 * 47.400 9 e so >00.?j0 130 IOQjOOO HMD 71.000 HJOO 1S.750 11.750 10.130 3300 310 0 0 0.00 0.0) 0.14 0.60 0.97 0.9* 0J3 1.71 * 9 7 0.10 0 37 1TJ00 43.000 rrjioa 9.730 3L500 55.000 44J30 :-(.3oo 47.400 40*340 0 0 j 2" 37 57 31 94 i 9 9 23 For oraiA ff t^r ucusssaai tal arjaai w ist i9? t, 1961. 000023 SPI-05631 !*-- ipV 1*1 t I F H Pvaenue *t tt -umT `h TiMr-l) All i r*r*r tt tow npwwT mo mnrrlrn Hrrn-- mcliiWr 1701 YwrT ASL OHir * II II tar nmot Wntm hwr Norm AHwna IUh of ooMO rwnu 1939 l'S24; 40- 1 Ml ; a. usr LSI 3. L*S2* ) Frl4 4 UKI 3 S2 ft Frl. FtS. S-> CL LSI* ci. cs. css. us7. usa C4. US3. US* Cl. c*. USX. US11. US21. US31I 7 S-l Ci. UST. US24 1 Fr* Fall Fr4 US1 US12 l* FrT. Kl. UKl US1 Vi Cli 1 S l. uxs * 03 USld US32 US4 3 CIS. to 4 07. Gd XX*. 01* CIO US1S Jul. Y1 V3 s Oil. CIS. Oil USL US17. US20 FrlO. 01 7 Frt, 04, 1C. 02 Cxi s Ft*, it * Frl. 1)4 US3 J*p2. Y4 IMO OS. 013 1 0*. Old Oil 017; C2d or a : Gd UK4. C21 USi s Frl 1. UK7 4 Swt US30 s Frf UK3 7 i * ino tielx i th: h>* MU er ica. j Tout... Muon J* 1 ASL AapoMRom at iht tavcr ttOl it oot cessment with ether ASL gum: the prunin' tumour stay haw tan of the pensrhnea. The otio mntad FVC udu. For fTTtiMinry ivy. mi Tibii iZ b&.< air is: or Th fL*S3! it mil taw. LAereioi an ftfer. icc poyea Toe data required ere; (1) Annual populations of employees classified by age; (2) Annual exposure estisites for each person in 0); (3) An exposure-response latency model for ASL induced by YCM. avoid using the exposure data directly by relying on the similarity in exposure levels in differing locauons. The exporure.response-latency data indicated under item (3) can be derived from established cases. The key data for these procedures are the set of cases worldwide, together with the descriptive data (Tables 12-1 a). It has been possible to calculate an incidence rate for each latency period for each ex h2' ICC Tin vea late of Tat.- Toe data under iiem (1) ire available in the UK is posure level for each age group (on the basis of the a result of the diti extraced from the relevant UK data and assuming that tt u representative of the occupational records (Fox & Collier. 19*77). Exposure worldwide population) and to use these rates to data for item (2) are more dificuit to obtain, but eaa derive a simple model of dose-response latency that be gleaned from the records that are used to define- can be applied to the population data. The broad the occupational population. The problem of oc conclusions are that, most cases have a latency of cupation changing, which occurred frequently, has about 20 yean and cases will continue to occur for been dealt with by using the principal employment the next 10 yean. category or the highest exposed employment cate In the calculation used to estimate the future gory. The esumauon of ume-weighied avenge ex number of ASL cases (Table 15) an assumption has posures for the least exposed employees is straight been made that when exposures were reduced to low forward, as the exposures were essentially continuous levels, the future risk of ASL became negligible. Two and eonstast. but for autoclave cleaners; mainte dates at which the negligible risk levels were attained nance worsen and laboratory workers, exposurer have bear selected: 1964, when levels were reduced could vary frosrzoo to near narcotic. levels. In the. to hundreds of ppnraad 1974 when the levels were* calculations described below, n bis besmpossible to reduced, to Below 10 ppm following the discovery of 6- I I16- > 313t- 46- *116- Far - 000022 SPI-05630 t Vmvi cniondr--mt ait 1*7 TaMr-t: Allow' lot- r duik- rn* <wr at atailrma-mnHWoIkran---twciirara ITCi*i rr rr 5' 4SL - ka Li I l i t Nan* Rni ti wan* Fubkcaram 1*55 6 7 t * l**0 I v j 4 s 4 7 FI I Ct 1*70 S I Id : Nl. S.! LX1. 1C 3 G3J * C*. Cl 1X3 5 F! n. 04. Ol. GL to 4 II. F4. FJ, F4. F7. SJ 7 FI. F*. CIO. CI1. Cl!. S4 t F10. Fit. <5*. C1J. G13. GU. G17 FI! FI3. 1X4. LXJ. GU 1*10 1X4.1X7. Gil. $*. CM. Git 1 1U. FU. tXL Gr 2 TeiL.. C Cl C7 (.'SI O L'SJ C4. C5. L*S4. LT7. US 10 US!! US16 usn c*. us: C7 CL US1. US1 L'S23 C*. US13 US4. VIS*. US1L USM US!*. US30. US!3 CIO. US7! US)4| us:7. usa usn. uss. um us:: Jl> cu ri. n qi Japl Jap! V3 Y4 I Vial* Mahan Crack a Jouxa All MpgiUCDBi Of |pc bw *nahu oa 01 nt aof a typical AS! tau pmaary tumour wu prokaely of the seteareaia. Tkii ana t cp|F s ostatas of FVC uskk `J hipan. G W- Gimasy; S " Smock C " **. It m Italy; LX Lance KicfCcs: Ct Cocaauevkju*: Jao Japan; Y Yuyeilavia: F Fncce N - Sarwiv; US - USA. Taut G* - an a. * IB Wt Germany. Cara LX! GU. US14. US13 aM L'Sa wo* skews set ta hr iiaaniia* with VCM ooi=i car kcaa wsaenws ftps tat an. lArraol can lUlcr. fOteiupounami. -Do* oat include US31 (mli alive). North .American PV'C plants have not recorded an .ASL case so far. The average latent penod between starting work in an occupation involving VCM exposure and death from .ASL for the 99 cases is 21.9 yean cm Francs. Sweden and the USA between 24 and 25 years, in Germany about IS yean), li is still too early to preact whether the annual number of ASL cases amongst VCM workers has reached a peak. ASL cases appeared earlier in North America than in Western Europe and while the occurrence is tending io decrease in North America (Table 12). it is still high m Western Europe. On the basis of (be data in this case register, it is possible to draw certain conclusions about nsk facton associated with AS! The large number of ASL eases in some factories and the absence of ASL in othen of similar age indicates that variations in manufacturing practices between factories may be the cause. These variations may reflect both differences m the types ofjob earned out by individual workers and differences in engineering practices. The bulk of the cases have occurred, however: in highly exposed autoclave cleaners, with relatively few in other PVC or VCM producuon jobs. So far no wellauthenticated-eases, have occurred in PYC com pounding or fabrication where many more people have been exposed but to a much lower dose. Prediction of future aSL cases as o consequence of pre-1974 exposure The causal relationship between VCM and ASL is proved beyond doubt by the specificity of the tu mour. the high relative incidence of that tumour in highly exposed workers, the consistency of the excess in different pans of the world, the ume relationship between exposure and diegno&s and the doseresponse relationship. An intensive analysis of the pre-1974 cohorts should establish the dose-response curve for ASL after VCM exposure and predict the likely outcome for the future. It will be impossible to collect a complete data set on which to calculate risks of ASL for the whole world, but within a single company there may be closer definition of the cohon. the number of cases and the pattern of exposure. Using these dsts and averaging across the worldwide population exposed to VCM. 'it is possible to calculate the future inci dence of AS! using relatively crude assumptions which ean. only be tested in. ume when the predicuon can be judged against the final outcome. 000021 cp).05629 i. r. h. tiaily hightrthasrthoie-ealcuiaiefcfor-the-ritand--mouKonduhemi stills ranpr-of overlOO-fotd in the etumstes-denved- from-the-differenr rodent expen* memc WHetrthts amount of variability occurs m ihr rctrxpoUtion of the nsk of low-dose exposure to VCM based-solely on differenr experiments- m the same speocsr the-reliability and hence-thc utility of these procedures is open to questions The general rclatiotuhip between the dosradmmntered and the- meidenee-of angiosarcomas derived from 55-week exposure-does not apply to exposures of shorter duration (Table 8). in all expenmems a total metabolized, dose in excess of 5 x 10* pg was required to produce an modcnccof angiosarcoma in excess of 1-2%. This relationship-was seen m both rats and mice and in experiments m which VCM was administered by gavafe- or by inhalation. In longterm inhalation studies, a total-metabolized dose-of S x 10s eg is equivalent to about 200ppm adminis tered over 52 weeks and represents a practical thresh old for this series of expenmems. In conclusion there is a wide variation in the estimates of dose /or a 10"* lifetime nsk. This vanauon is due to the type of mathematical model that is applied, to the assumptions that are made and to the particular experiment thaits used to provide data for the extrapolation. A high level of confidence anaoi be placed on low-dose otrepolations whea variables that would not be eepeaed to alter the expression of risk have a profound effect on the estimated risk. In addition, the mterspedes enrzpelation from experimental ammais to tr.za is largely inrunrve.!; is clear that estimates of risk should take :au account all available data, including epi demiology. to provide a degree of reliability. Risk assessment from human studies ?.tfitter of ASL cases Since 197J. lists of reported ASL cases attributable to VCM exposure in the VCM-TVC industry nave been kept by NTOSH (Spinas it Kaminski. 1978). by LARC and by the VCM Committee of the Associauon of Plasties Manufacturers in Europe (A?ME). Details of 99 cases is the .APME register at T>M*-II CImmoho; of ASL no Couiurv U mini Lwoo-- i * Oinwn a w'n Cmn* 1 G*rmmw 4 Ww Cwnwww 1 Frmacr 2 Franco3 Fftac* 1 UK. UK 1 SWn- SmIammk 1 Co0i I USA 0 USA 3 USA SUaorWofM1 Jopta l Vosmloiii-- 1 Cwrtiniliw* No 0i Tout.. Tout. - 10 4 2 2 < j 2 5 2 42 10 n 9 4 14 2 4 A "For lAr purpem of Urn i tht i Toni-. I > moy. n a aw wamy to wiuly i Inranao of ibat pttau- the end of 19S2 have bees analysed by country and by manufacturing company and plant. The czsas hive been recorded from all major VCM ?VC manu facturing countries (Table 10). but the incidence has not necessarily been m proportion to the PVC procucnon capacity now or prior to 1962. In the absence of data on the-number of workers employed. pro duction capacity is the only available indieauon of the numbers of people potentially exposed. i ne maiomy of the ASL cases are PVC autoclave cleaners or men who have worked in or around autoclaves. There are ASL cases among men wno mxnufxcnred VCM and a few cases were involved both with monomer and with polymer procucuon. Only one ease suffered from both acrc-osteolysis and ASL. The ASL cases tended to occur in larger numbers in some plants than tn others (Table 11). Of the total of 59 ASL cases recorded in North Amenex. 5-t have oecun-ed at four PVC plants, while ever *0 Tibk 10. Diicibatim of ASL am eauatry Countrv- >0. Of ASL aui FYC producuas stmtpuu wstory fkiloiooaa yr) 1*62 1972 USA ** CtfBUV Fnaa Cmm UK Snin Yufouamluw Cacbotloraku loots Bclfsua Sonny TouL.. Europ* Nona Aaeia km of World Torn... : ; :i 14 n 10 3 7 27 33 43 39 21 : 12 i3 1* w JI 12 ! in 1 31 w 31 ASL- NopottT^rau 6. * 1 7W 260 |7 177 20 S 111 23 3S* 23 20 31 -26 709 2316 2090 1133 027 * 302 103 to 771 41 1699 193 ft* 3930 2171 3334 MI 000020 SPI-05628 2' tc r,; p< lc ra in rr. a ih C: at or EL* ZZ9S0'ldS STOOOO iffji *j y! -*7 s r i iJi is llii :ihi srl `i*? Lf *t :* n I? = *S * r i : VI - -* i ffi'**5r 3 * i i. I i I * Is** s=! ix , 1 S' 5 . i L? iizl M- r* | I; l fI ?! * it i?i I"* mi r* H *? J, f * * i j : /! * |1 1 5 5aYi.r*S. 5 ??? s5 =iii <i Hi 4 I Jt * If!-;SrlII -Sr; J f . Hii': 5f! S *3 Hii i 1 1. -f rm a< XXi S4 < * mo "ni 1 S 51 3 t3 |:-s! 1 *+ 5 m z* .5 in r B l ** 2 -1 :3 ** a ; st sin--mmip 1 l |4m |M MNHHNIIMM M INMHIM H(| | {UtUMllX |W f f | l f 4|I1|MjA<II| IW.I MM| yM A >M| I t 104 I. F II rt */ Tahir-* VmU iM<Mr<nnM iwmiw o> wmi niiinm--i m Cow* iprnir IWIIWMW v *1 a br- p nuii AWpOMWI riS* Malt Fma* Mcao w i 0.000 torn :soo 1000 500 230 230 so \ 0 11-145 11.007 10.144 MSI a*} a** ISM ISO* 0 i 7 - ur 3 l 30 17 1 TJ- 1.7 * 10* 4 7 34.7 :i.7 T4 I S a 10* :o.7 *3.3 r>.i T4 J a to* >4 SCO 411 La n a t J 0 10* mo 247 aj IT 4 74 ter- 30.0 son 30.0 IT 4 7 4 . 10* :* o 47 0 34.3 Laa r att Ha 10* s.* to* 3.S )0.3 0 0 1.7 T4 o Lat-ad* 0 0 0 0 IT a 4 Laa # af. 'S we*. kl-k opcntnmcr dim* (or 30 k. hrCD. am. 32-k nwn--. * day-aipoiai (Laa n *L imy TknrnraM km M Mr tldte tbr Haliu--I ter TtMt * time* iia mmaaan iteiif mkmI mv i\ Thus: tr , ,,, 0.011m1 V.(mou,-V.(r.,)x5;5-- -5706,1*4 hr*--- - 1395 m-4 hr The values of 0.045 nr and 0.011 nr are the body surface area of a rat and a mouse, respeenvely. Since toxicity u a function of the concentration of the toxic metabolite in the tissue, the amount transformed must be normalized for mass to estimate an equiv. slent response. Thus Ve must be adjusted on the basis of the bocv weights of a rat <0.25 kg) and a mouse (0.03 kg; by dividing by 0.03'0.25 * 0.12. The V. tor the mouse on a mass-equivalent basts is therefore: i;o< yjr -- 11625 g 4 hr This value of Ve has been used in calculating the total amount of YCVl metabolized (Table 7). From the variety of models (or mathematical extrapoiauon techniques) used for low-dose nsk extra polation (Tabie 41. an arbitrary ehoiee of models has been mace to lest the robustness of the extrapolation from the different animal studies. A log-prooit analysis of the dose that would be expected to produce a lifetime nsk of ASL of 10*` is presented in Table 9. This calculation can be earned out on the basis of the concentration mhaled, the daily dose metabolized or the total quanuty metabo lized dtihng the whole experiment. There is a wide variation in the rmmairri dose depending on the database used for the calculation. The largest san ation between doses derived from the rat experiments is 360-fold (0.025 ppb r. 9.1 ppb) when exposure m ppo is considered, but decreases to 100-fold for other of dose. The results from mice are substantially lower when expressed in ppb (2 x 10*''ppb) but the difference is less for other expressions of dose. Similar calculations of the dose expected to give z 10"* lifeume nsk of .ASL have been eased on i Wesfeul! analysis (Table 9). This is a more 'conserv ative mathematical model and the estimates of dose are aecoreagiy lower. The variation m estimates of dose is. if anything, larger than that observed with tbs log-prosit analysis (for example a 10*' difference between the S values derived from Wjstar and Sprague-Dawley rats;. Tne doses for mice are so much lower than those calculated for tats or man that the issumpuons used in thetr calculation must be suspect. A further calculation to derive the human dose likely to produce a nsk of 10** is given in Table 9 (S calculated for maai. These calculations are based on a V. for man of 1675 yg 8 hr based on correcuons for body surface area and mass. The values are substzn- Tbi t. Vinyl cdkendt (VCM1 do** in* bceaae upsunon* modra* a Sp:ifm Daaky nu npuirt 10 VCM by ___________________________ inhalation Cooes 10.000 10.000 10.000 10.000 10.000 4000 6000 6000 6000 6000 Schadnk* 1 II III IV V l 1! in IV V No. af low 260 IS M 100 23 :to 15 IS 100 23 ' ASMOI malabedad: **-4 br *321 ss:i ;::t 1370 522! 3403 5403 3403 1330 5403 nt (total) 1.4 10" 4.: a to* 1.4 101 1.4 10* 1.4 1C l 4 10* 4 6 * 10* 1.4 a Ifrl.la 10* 14 v 10* Aopouiroai wadiess (*/) MaW 10 0 1.7 1.7 0 10.3 0 0 J.4 0 Fault I3J 0 0 0 1.7 33.3 2J 0 1.7 '.7 Mu 11.7 0 0.1 0.1 o.t 22.0 1.7 0 L3 0.1 no. BTl BT3 BT 10 BT 10 BT 10 BTl BT 3 ETIO BT 10 BT 10 -SehMultc 'T'ttky. 3 davi-v* for 12 wk: II--4 br.djv. J tin.i f,, ; k; 111--* hr'dav, S tin-n for J wk: A--I hr dav. * ayri for a k: V--4 hr day. | day.*, for 2J . CAaeaai scuBodaad tv> m a hour dcnvrd from tht fa--.u. v i* i-i v . 5 X_- S har V u 4 S of the d hr Vfcjyt ^ r 000018 SPl-05626 8 Vinyl i tiloraw i inhalation (Tabled) for rats cxpoiedrorally (Tabic-6) and for mice ezposcd-by inhalauan (Table71 are also presented.. Data from-cxpcnmcRtr with venous ex* posurt penods of short duration are-given m Table 8. For calculating the-amounts of the dose metabo lized in rau in the inhalation experiments, the con stants calculated (Gehring-er al. 1978) have beenapplied. For Witiar rats, the 1C. and V, values derived for Sprigue-Diwley- rau have-been-used. These estimates of metabolized dose, have been in cluded in the tables. For the experiment in which VCM was pveir by gat-age. the data from Fig. 2 were used to estimate the amount of VCM exhaled unchanged. As the for exhalation of VCM was M minutes, these data based, on a 72-hour period give a good estimate of the fraction of VCM exhaled in the; 24 hours between doses. It has been-assumed that the VCM not exhaled was metabolized, an assumption similar to the one used for estimating metabolized dose in the in halation experiments. Greesr & Haihwiy (1975 &. 1977) showed that VCM administered by gavage to Wlstar rau was exhaled and metabolized in a similar manner to that in the Sprague-Dawley rau. and the V. and K, values derived for Sprsguie-Dawley rau have been used. In the experimenu by Feron et aL (1981). wbo used War rata, the same assumptions about V, and K* have been-made. The quantity of VCM administered has been deah with as if it had ben tcminmered by gavage. vcm eoM-ime/xei ' Fig. 2. Summary of don-dependent urinary ud pulmonary excmion of vinyl ehlende monomer(VCMV Ch&afy exesw lion < > upineau meubobta of VCM. while pulmonary rliminanoa () it unchanged VCM. (After Watanabe A. Gehnei (1976)]. For mice, the data have been combined in Table 7. The estimation of the dose metabolized in mice-has been calculated using values for VB that have beenadjusted on the basis that, for a chemical requiring metabolism to tu active form, the quantity metabo lised will be proportional to ibe body surface area and must be expressed m terms of metabolized dose-lcg body mass. This technique has also bees used by Gehnng et al. (1978) for estimating the dose metabolized by man. Tabk S. Vinyl ctiandr dew sue saadtner ai htpauc aagictarcoaa m malt Wnur nu Bimr et i djvi t icr Coaca lepmi Amount ntunnhna al a hr >1 natal) AApoureoms Expmt no. 10.000 6000 2300 500 230 30 l 0 3321 S03 SOM Ml] :<;< :?9 it 0 14 a 10* 1.4 m I01 U 10* SI 10* 6J 10* 1.9 * I0: as . io> 0 29.6 11.3 1-0 101 3.1 0 0 0 IT 7 BT7 BT7 ST 7 ST 7 ST 7 ST !7 ST7. |7 Tbl 6. Vmy* chionde /VCM) bow tod modrasr of hgpauc anpoiarcasu m rtti pvtn VCM bv fiMtt or mrwoon Dew im at) 302 16.63 3-33 1.0 0.3 0.03 0 001 I4.IT 3.0 1.7 0 Amount iV of bowl 30 33 10 2 1.7 1.4 -- SO 32 16.3 % Amount mnooocxea wi now* 6230 2703 730 ;:as 14 1.4 0 13.000 2290 tow m 0 I i totalI :.6 10* * 0 10* 20.10* * 26 10* U6 10* 216.10" 0 (2 a 10* 1.63. 10* * 23 . 10* 29 10* 0 Anposafcom* iiwvuiBW (tel Malt 20 10 0 1.3 0 0 0 49 49 10 0 0 Foma* * < ii.i 0 27 1.4 0 0 *3 16 4 0 0 Moan :i2 12! 0 1C 0.7 0 0 31' 32 1. 0 0, Exrmt no. 8711 ft? (1 ST 11 BTn BTr BT27 BT 11. 27 Ftrso et et. (19111 Catena** from fiaia atm* from WaiaaaBt* Gehrmi (1976) pra*niad.ia Fig. 2 . 'Amnmra nt 2SpnfUt*Dtw*tr rau rtoatd by pvifr vnk VCM m tore ofl S tiaorwa lor 32 WL. (BTT? dot* tor 39 *. ran u>* at sacob oy Firs* rt tL (19111 aae dm* for 13 ^`sur mu rwtrmr a Cae encases ' CM daaav* m PVC. 000017" SPI-05625 1 1 t" Caacr-' ............. ' -- ' --' ' - ' *" (ppnr_______m 4 iif______ mail. Maw- Fawr- Mw Liwmno 30.000 10.000 *000' 23 300 ao mo 130 100 30 23 10 3 1 0 3047 3321 M0) 30M Ml) :d. :i2 1701 1)0* 7)* yn iw u 17 0 1 47 . JO* 1 **->-10* Ml 10* 1.) 10* I I 1C 0.3 > 10' !.! 10* 40 I0> 3 4 10* I f to* 1.0 10* 44|Q*-" ill vr~ 44e 10* 0 144 100 10.) 20.0 0 J.4 U.7 1.7 0 l.l 1.7 0 0 0 0 4).) in 3J.3 m au 0.7 u u 1? 92 0.7 1.7 0 0 0 MO II .7 210 21.7 100 J.l 10ft *0 0.1 4.3 42 at 0 0 0 IT** art IT I TI IT 1 an t: t: at: an i.* IT 15 IT 15 IT 15 IT 13 Tt.l *. 13 Atm Makamei at ntlli. 'haaam ITtaM after ealrUet eWrittiarMirit ip rliOek: Otwtee itMwatrtfiMmin IT 0 it proPaht* tow wmwrieuima b---- o( the aim Plenty multistage model by the Food Safety Council (1980) and. by Gavlor & Kodell (1980) thawed that for the tune 10"* lifetime nth. the Food Safety Council estimated the dote as 2 x 10'-ppm whereas Gavlor i KodeD esthaatad the dote as 5 x 10"*ppm-The difference between these two was due to alternative assumptions on the value of the expansion of the exponential term used. In general. calculations based on the amount of material metabolized or on human data have pro* duced exposure vaiues of about 1 ppm for a 10** lifetime rule. All the other studies have produced exposure vaiues in the ppb note. A large variable appears to be the selection of the mathematical model applied to the experimental data. In the following section two models are used to calculate the exposure for a 10~* risk from a variety of expenmental animal data applying the correction for metabolism used by Gebrmg ti at. (I9?9). Calculation of txpesurt for I0~4 rjk A summary of the crude ASL incidence rates for inhalation studies is Ssr*gue-Dawley ms is gives in Table 3. data for Wistar rats exposed by Table a Summary of euaautataw mfc lueumma for vnvt ehlonPt aioaemtr* Ri(mw Seeem ipww lor 10- fifnsc mi tpptn Comnnu SrOBeiitrunu tt at. (1*751 Dai Kiuaact k McGaiithy (1*731 Gvfens| n mL (1979) Food Safer* Couaal (1*10) at. nap Rat. maa hat hat AaOencw n 1. (1*101 Gavtor a KodtU (1910) Catiborg (1*111 iarr (1*12) Thu paper (Table *) ETA (1*101 nas (istoi Crop k Gees (1*10) 'Kfict Urr (t9tZl *ugpt %b7t KUUd OtdcaiM. hat. aun Kai hat Maa hu Mont* Mta hat Man hat hat Maa hai 1* iHiiimr 73 119 14 140-14Q0 >1000 <ID->I000 20 20 nio-* i.*to-' >1000 0.7 0J 1* * io-* >100 O.fla-*.IO\ :io-" / 0.02-00 : * io-'-:io-* 10*" 0.007-4.14 i h*anaa 4 (-day 3 * I0*1 m|.^k|/dar 0.* m-dav OJ ufdav Frobit uiow K Mantai) Lspi lUopt -- MSI Lapt inept 1). eae-bitl Linear taroufP am Layamon Bxnr&atiomuuoe Oau lactudad t, intar or tef^prebn Depend oft mammauaal mod*) vu*d Onthn Amupt-OeO Wcibuil DNA bredmj uaad for dcnmiu Upper *7.5to metatpcr asm ai bacar neetl AMtay*-Do0 Wcibuti Dcmad frota tair'i Offline cpiOcmelon Leg-proon Lo|-probn mcfcidutt teotrsTuformauon daia for man WtitalJ WcibuU tadudwt bournaaformauofi for man Food or water W*ur Appfrvtef sorter dau io mm*wr Upper 4;*; contoence hrmu 000016 SPI-0S624 I Vinyl thlondf fWte-, *f s g Mi -ill 7*1 5 Hi i 4 5 js M < i ;]! fill PHl! i z a* i sIt 5* s - r- *T i I X 7.1 <* i Xl > s yi r 14 1 i of VCM meiabulistirm n is given-m Fif.-1. On the- basis of this schemer the htfhlv reactive intermediates m the metabolic process tpamcularlv chioroethvleneoxide) react with cellular macromolecules, including DNA to produce the cnucxl lesions leading to mu tation or the induction of cancer: Studies on the quantitative aspect of VCM metab olism have shown that thercts a dose dependency in the rate of metabolismr After-administration of "C-labelled VCM by gavage at doses between 0.2 and 100 mg kg to Wistar rats, the amount of "C excreted, in the urine and faeces and retained in the carcass was estimated over 72 hours (Watanabe-i. Gehnng. 1976). As the dose of VCM was increased, ihc proportion exhaled increased and that excreted in theunnrand faeces decreased.(Fig. 2). The proportion retained in the carcass, also decreased The tame, general trend occurred: after administration by in halation. although the magnitude, of the differences in retention and excretion was less (Watanabe &. Cehnng. 1976). Studies of the amount of non-volatile.material retained in the carcasses of rats exposed to venous levels of "C-labelled VCM for 6 hours demonstrated that the metabolism of VCM appeared- to be is accordance with Michaelis-Mentes tannin (Cehrisg ei al. 1978). The constants for maximum velocity of metabolism (V. is u; metabolized. 6 hr) and the Michaeiis constant fKa in fig VCM litre air) accord ing to the formula: V,S V K.-S (where V velocity of metabolism s u g-6hr and S - cascsctrauon of VCM bemg inhaled) were V. m 8558 u | metabolized.-6 hr and K* S60 u g VCM .litre air. Thus there was a considerable change in the rauo of administered dose to metabolized dose as the exposure concentration increased (Table 5). At the higher doses a smaller proportion of VCM was metabolized than at low doses. Review of earlier calculations of risk There have been a number of attempts to calculate the nsk of ASL development on the basis of extrap olation from experimental data. These have been reviewed by Barr (1982) and as adaptauon of his data is presented in Table a. The introduction of biotnnsformauon data into the estimation of nsk increased the lev-el of exposure calculated to cause a 10"* lifetime nsk. from pans per billion to in excess of one pan per million. A further refinement of the technique, using DNA binding as the measure of dosimetry (Anderson ei al. 1980) provided a similar esumaie of the exposure. A vanety of mathematical models can be used for extrapolating below the experimental dose range, and it is not possible to select from amongst these math ematical modeis on the basis of goodness of fit to expenraenial data. Attempts to do so have shown that most of the models fit the data equally well (Cehnng tt al. 1979). It is equally difficult to select amongst the models on the- basis of the assumed mechanism of action of VCM. Thus a comparison of the lifetime risks calculated using the Armttage-Doll 000015 SPI-05623 SPI-05622 000014 I 5 a* ; lt t % j 5 2 2 ( ttia//c t t at (1910) Ikaummil A Niolow (IVNI) lrt i i IIic m I HcN|tlmliny iia tl Iik im k in mh iim lirm d 9ir ca*c-MiiHttlr4 lwtf Rcit o f m ik dw liei I F. H P\-rnj rr-mi. j*r 190" * t VinyH rftlnnrW wn whietehanwigraudiima- the- food- orbeverage. Since 197* thrtmounr of VCM in P'"C hat. beerr reduced io leu than i mg kg min the melt that the maximum human daily intake of VCM in food and drink it O.lwgdav (Ministry of Agriculture;- Fisheries it Food. 19781. The fount) group with potential exposure to VCM are thote who live in the vicinity of VCM or PVC manufacturing or fabneaung factories. The levels in amPtenr atr around a factory are-very-low (in the parts per 10* range) but much larger population groups, which include all age groups, are Involved. For the workers in VCM manufacture and PVC polymerization and fabrication, the route of exposure is by inhalation. Much of the animal earcmoeenicny data arc based on inhalauon exposure and the human epidemiology is predominantly of populations ex posed occupationally by inhalation. Thus an assess ment of the risk factors and the quantitative nsk of inhalation exposure is the main objective. For the consumer exposed to VCM via food and beverages the route is by ingestion. Relatively few experimental studies have used oral administration and only one study used a comparable exposure pattern-(Feron ri al. 1981). Similarly there are no specific epi demiological dau on oral ingestion. Risk assessment for exposure via the oral route must rely on the existing animal data and on extrapolation from epi demiological and experimental studies of inhalation exposure. After xdwmmration by garage or inhalation; pan of the dose-is exhaled unchanged-and the renumcerti excreted or retained in the carcass. A general schemr ct CH------OCSIH.CHO VH it) f =0 ! IHCH.STH.CH. IV VH OH c =o I CHCH.SCH.CO.H VH Risk an i inunit from experimental animal dau ciu Ci. Assumptions !n earning out a nsk assessment on the basis of animal data, a number of assumptions have to be made. The srst of these relates to the overall dosi metry. Ezpenmenul animals are exposed to concen tration! of vinyl chloride or dosed with amounts of vinyl chlonde that allow an estimate of the amount to which they have been exposed. It is possible to calculate a correction factor for these quantities so that they are applicable to man. However, rau and mice live for relatively short periods of time (up to 2 years) curing which they develop cancers of a type similar to those seen m man. The latent period forthe same tumours m man may be between 10 and 0 years. It is therefore assumed that the lifetime of man is equivalent to the lifetime of an experimental animal species even though the chronological time is substatically different. Strictly speaking, mathematical extrapolation of nsk on the basis of experimental ammai dau pro vides an estimate of the risk at low doses to the experimental animal under consideration. A variety of factors, particularly inherent biological suscepubilitv and differences m raeuboilsm. render the extrapolation of the dau from animals directly to man subject to numerous errors. It is at this point that scientific judgement is required to deride whether these dau are applicable to the human situation. Metabolism Is rats. VCM has been shown to be aeuboliztd extensively, producing a range of excreuon products. t ro.H I` CHCH.SCH.CH, I 'V VHUit OH <1 I T co.h i* I CHCH.SCH.CO.H I NH. fn CO.H I` C -- CM.SCH.CO.H II 0 S(CH.CO.H), r :` (.) Fig. 1. Scbemr showing the mettbobsm of vinyl chionde monomer O'CM) m nu to 5-cosuimn metabolite*. VCM (a) is convened to cfaloroetbytes* oxide fb) which is trans formed spontaneously to chioroactialdebyde (c). These t*o metabolites art mutagenic and besee are considered to be the proximate carcsofcsa. Thr urinary excretion products A'*aryl*5^*hydroxyeUiy))evftaBc (t) S^cirtoaymciiiyi)- cyttemc (f) and thiodiglycoliic aad (g) art denved from these rntnagcsuc' metabolites via id). Gly and Glu are the gjycinc p^ giuusisa residues of flusaUucoc. JAw/vcr Gieu L Halbway (1977)]. 0C0O13 SPI-05621 I F H PntfUO* ei of. TM* f Uww awmwa pr 4ow *! tifw*aiiM>* < imww frwi *ii iwiwii tn f*i earcmofcnmn t*Mv Tuivietar Conew %ppmt Domcmf ZrmatHWM Pitman SowoHoMtmmti- LlVr Mimon-Mie Mmamaan JO.OOO 10.000 10.000 ZSC (female* * 100 mulri- 90 5 (femalei 50 tmatr* IM* tCcWMtl Pan IwMUlat m. (INI). HFTftFT dotes (7-10%) in both animals and man. The doses responsible for acute loxxstv art about I000>foid higher-than the mininnardose for aremogematy and there is frequently no sign of overt organ toxicity pnor to the development of the carcinogenic re sponse. VCM is mutagenic in a variety of test systems including Salmonella typhmurtum (Rannug tt al. 1976). Saccharomvces (Lopneno tt al. 1977) and Drosophila (Verburgt &. Vogel. 1977). usually with some forte of mammalian microsomal metabolizing system to convert-VCM into its active mcabobtes. ehloreethylene oxide and chloroacctaldebydc. The data on the mutagenicity of VCM provide useful qualitative information on its mode of action and metabolism, but are not suitable for the quantitative estimation of r.ik to man. The most useful expersccatai data are derived from long-term animal carcmogeniesty studies. An extensive senes of 17 studies (Malwai ct al. 1981) gives a useful database for risk assessment. Other studies (Feroa er al. 1981; Lee ei at. 1978) tend to confirm the findings of Maitoni. Carcinogenic effects were observed in mice. rats, and hamsters. A complication m the selection of these data for nsk assessment is the variety of tumour types observed (Table 1). Some of these occurred at very high exposure levels, but mammary acenocarciboma in females and ASL in both sexes of both rats and mice occurred at 60 ppm or less, exposures similar to those beiieved to have occurred on manufactunng plants (Barnes. 1976). Epidemiological studies Several maior epidemiological studies on workers exposed to VCM have been reported (Table The main organs that have been associates with higher incidences of cancer in workers exposed to VCM are the liver, lung and brain. Increases in the standard ized mortality ratios of cancers in the buccal cavity and pharynx, of lymphomas and of cancers of the lymphatic and cardiovascular systems have been, re ported in one or tw o studies. The analysis of cancer of the respiratory system is often confounded by smoking, making quantitative analysis of the con tribution of VCM difficult. The excess of liver cancers is due to an excess of ASL in many of the studies. An analysis of the statistical power of various studies for association between VCM exposure and cancer of the lung, liver and brain (Beaumont i. Bresiow. 198!) concluded-that the results for liver were consistent with an aeuolopcxl role for VCM. For brain canerr. where three out of five studies had statistically significant findings. the mulls were morevariable.' positive ftndmp occurring in the studies with the greatest starancal power. The most reason able interpretation was that the data were consistent with a causal association between VCM exposure and an excess of brain cancer. Infante (1981). in reaching the same conclusion, points out that the relative risk for brain cancer u much lower than that for liver cancer. Only two out of eight studies on lung cancer (Beaumont 4t Brestow. 1981) yielded suusiically signincant results and. because studies with a high power were-oegauve. a causal assoaaoon was consid ered unlikely. ASL is the most suitable endpoint for analysis of the risk of exposure to VCM for a number of reasons, it is a rare cancer ra unexposed populations, making attribution to VCM exposure on the basis of work history a reasonable approaen. ASL occurs m both animals and humans exposed to VCM and it is usiikeiy that any other earanogesic effect of VCM will be found to occur at lower exposures than the lowest exposures that induce .ASL For these reasons, most work on the quantitative risk assessment of chronic exposure to VCM has used ASL as the eadpotni to study. Case register The availability of data from a comprehensive case register of ASL cases with a history of occupauonal exposure to VCM provides an opportunity to identify risk factors for the induction of ASL. Persons potentially exposed to rtnvl chloride Current manufacture and use of VCM and PVC results in the potential exposure of four groups of the population. The highest exposure category covers the workers involved in the manufacture of VCM. its polymerization to PVC and certain other industrial uses of VCM. Within this group, certain occupations, particularly autoclave baaing, involve higher poten tial exposure than others, although all groups would now he expected to hate exposures complying with hygiene standards of 1-5 ppm. The next category covers those exposed as a result of using the PVC. Workers in the compounding and fabrication of PVC products are exposed to residual VCM released front PVC on hating (but PVC does not decompose to VCM when hated). In general the exposure levels for these workers are very low m companion to those forPVC polymerization workers (from' 10 to 100 luces lower). ConnnE--rw-ho ai food and drink beverages that have been packed in PVC may ingest unreacted VCM 000012 SPI-05620 F4 Ckmm Timer vi li. So 2. op 11"-202. 191' Eiip-- Orppi InupiAH npm wnp< Review Section I ^JUL 'JlNU I * MV4 fc.V;vC o:i4*o-' nottw> Copw;0r f T9ft EfrnMp^l KC ii w <i VINYL CHLORIDE: AN ASSESSMENT OF THE RISK OF OCCUPATIONAL EXPOSURE* 1. F. H. Pvuchast Central Toxicology Laboratory J. Staftord PWsucs ud Petrochemicals Division and G. M. Paddle Central Medical Group, lmpenal Chemical ladustnes pie. Aldertay Park. Macclesfield. Cheshire. Fngland (Aecnred |4 Oeeember 19(3; rmsioni rrcrrrd 13 January 19(6) Istrodonion riods. since m some plants opernon became faint Vinyl chloride monomer (VCM). more property Mined moaochlorcihanr. is a colourless gas normally handled under pressure as a liquid which boils ai - 14:C ai normal pressure. Discovered around 1S35. VCM*s commeraahzation did not begin until the 1930s and did not reach high volume until after 1945. Presest manufacture is around 12 * 10* tonnes per annum, nearly all of which is used to make the polymer poiyvisyl chloride (PVQ. Until the 19H. VCM was regarded as a material of low human toxicity and the main concerns were related to the compound's narcotic effect. Indeed mere are many reports of employees exposed to VCM monomer in polymer plants becoming cizv and unconscious. Because VCM was considered to be relatively innocuous, it had a threshold limit value (TLV) of 500 ppm. S-hr ume-weighted average [TWA) for many years (ACGIH. 1974; Lester er al. 1963: Torkelson ti ai. 1961). .Measurements of em ployee exposure were infrequent, since most mea surement and warning systems were designed to ensure ihai plant atmospheres were beyond the ex plosive limits, tire and explosion being the main hazards of VCM. Retrospective esumites (Barnes. !9"6) of typical TW'a personal exposures (in ppm) for polymerization workers have been cited as: 1000 in 1945-1955. 400-500 in 1955-1960. 30CM00 in 1960-1970. 150 in mtd-1973 and 5 in 1975. However in some jobs, particularly in the cleaning of the autoclaves m which VCM is polymerized to PVC very much higher exposures, in thousands of ppm. were undoubtedly experienced for short.medium pe- and unconscious from time to ume. The first clear indication of chronic health prob lems assoaa&d with VCM arose to the 1960s in men who entered VCM polymenauon autoclaves to remow build-up of polymer from the walls. Some of these men developed aero-osteolysis (.AOL: Cook et aL 1971: Hams it Adams. 1967; Subu et ai. 1963). Modification of working practices led to a reduction ia the inscer.ce of AOL cases tr. autoclave cleaners. Although AOL is occasionally seen in people not exposed to VCM (Meverson &. Meier. 1971: Wilson et el. 1967) it is a rare disease, in the iate 1960s. studies in rats involving exposure to high conesntrzuons of VCM fer lcr.g periods (Viola. 1969) failed to produce AOL but showed an increase in the incidence of tumours at various ntes. Further studies (MaJtomer ai. 19806: !981:Maltom it Ronaineila. 1980) showed the rare tumour angio sarcoma of the liver (ASL) in exposed rau. and confirmed VCM as an animal carcinogen. Three ASL cases in employees at a PVC polymenzauon plant (Creech it Johnson. 1974) confirmed VCM as a human carcinogen. Other known aeuologseal agenu for ASL in mas were thorium dioxide, arsenic and. possibly, anabolic steroids (Maltom tt ai. 1980). Since 1974. the health hazards of VCM have been the subject of tr.znv investigations, scientific papers, seminars and other presentauons (Conference to Reevaluate the Toxicity of Vinyl Chloride Monomer. Poly(vtnyi Chloride) and Structural Analogs, 1981: Gauvam. 1976: IARC Working Croup. 1979; Selikoff. 1975; Szadkowski it Lehaert. 19EL US DHEW. 19S0). The plethora of information (and misinformation) now available suggests that an ob jective historical case study of VCM w-ould be of A Ionter version of this paper hu been published m Teoeeietieal Aiii Auttmtm. edited by D. B. CUyson. D. Krewski and 1. Munro and published bv CRC Press, lac- Boca Raton. FL (19(3). value. Experimental and human data Abbrmeuons: AOL acro-osteolysit: ASL tnpo- ureoms of Uie liver PVC polyvinyl ehlonde: TLV threshold licit value: TA'A m urae-eiemed average: VCM -- vuni ebionoe monomer: Experimental studies The principal effect seen in the acute and subacute studies is anaesthesia, which occurs ai reuuveiy high 0000221 SPI-05619 vinyt~ Institute: A OMakmet Ifta Societyat The Plastics Industry, Inc. September- 12, 1989 Mr. Robert- Barham, Chief Toxic Air Contaminant- Identification Branch Air Resources Board P.o. Box 2815 Sacramento, California 95812 Re: Draft Report on vinvl Chloride Dear Mr. Barham: The enclosed article was referenced in comments submitted on September 8th by The Vinyl Institute on the Air Resources Board's Draft Report on Vinyl Chloride. I would appreciate it if this report is appended to those comments. Sincerely yours, MNS/pmb Meredith N. Scheck Assistant Director enc.: I.F.H. Purchase, J. Stafford and G. M. Paddle, "Vinyl An Assessment of the Risk of Occupational. Exposure", Fundamentals of Chemical Toxicology Journal. Vol. 25, pp. 87-202 (1987). Chloride: No. 2, 000010 gp\-056A8 Wayne Interchange Plaza II 155 Route 46 West Wayne. NJ 07470 (201) 890-9299 t 4 Purchase; atal. A very- extensive evaluation of the availableinformation atr that: time: is- included: in this- article, and a. vary comprehensive 'examination: of risk, assessment: approaches: to vinyl chloride*is: examined-- Me-believe^ that: this document demonstrates:a much: more studied, and. scientifically defensibleapproach: to~ assessing- risk of exposure to vinyl chloride. In summary; there are atrleas'ktwenty epidemiological. studies which involve over-45,000 workers who: have occupationally been exposed to vinyl chloride.. To dismiss this, body of epidemiological, study in favor-of basing-risk, assessment: on animal data, is questionable at best. In the paper by Purchase, et al., information that is precisely the issue being addressed by DHS is. present. In addition, an epidemiological study of populations living- in the- vicinity of VCK production facilities had. been conducted previous ly. This, study, Barr; et al. 1982, suggests- that 100 ppb re presented the estimated dose representing a 1x10-6 lifetime risk in man. That value is similar to the highest estimates derived from the animal data when taking biotransformation data into account. The studies discussed in the paragraphs above, will be forwarded under separate cover. Finally, the Vinyl Institute is extremely interested in reviewing the revised draft document before it is forwarded to the Scientific Review Panel. Please add our organization to your distribution list. Materials should be forwarded to: Meredith N. Scheck Assistant Director The Vinyl Institute 155 Route 46 West Wayne, Hew Jersey 07470 Thank you for your attention to this matter. Sincerely yours, MNS/pmb cc: Mr. Richard Forey Substance Evaluation Section Air Resources Board P.0. Box 2815 Sacramento, California 95812 Meredith N. Scheck Assistant Director 0C0GC9 SPI-05617 Many of the- epidemiology- studies- that have- been- in the published. 1 iterature. haves been* updated: ire the- paste yeacr or- two. - One example is the study Update of Vinvl Chloride Mortality authored- by Dahair, et. al~ which:vaeupdated^am- recently as-1988 and: further-demonstr-ated a decreasing- cancer- incidence rats* in workers: as the latency period, has: beenr expanded: substantially. The- person- years' in this one particular- study has: been- expanded- from only approximately 4,000 persomyears~-to-over 17,000:persorryearsv thus:a substantial- increase* ixr sensitivity of. the*- study, as only one example. The Chemical- Manufacturers- Association: (CMA) Vinyl Chloride Panelsponsored-epidemiology study-was-updated' as recently as 19B6. It is a very comprehensive epidemiology study consisting of a cohort of over-10,000 workers employed at 37 different plants belonging to 17 different companies. That study identified, at. that time, over 1,536 deaths- These are only several examples- of many epidemiology studies published:on vinyl chloride and DHS's approach to dismiss human epidemiology evidence in their risk assessment is inadequate. Many of the human epidemiological studies point out a statistical ly*^ ignificant association between an increase in lung, liver and brain cancer and exposure to vinyl chloride. For brain cancer, three out of five studies demonstrate statistically-significant findings, although the results were somewhat variable. Positive findings occurred in studies with the greatest statistical power. Most reasonable interpretation of the data is consistent with the causal association of vinyl chloride exposure and an excess of brain cancer, however, the relative risk calculation for brain cancer is much lower than that for liver cancer. Only two out of eight studies on lung cancer yield statistically-significant results, and because studies with the higher power were negative, a causal association is unlikely. It is for these reasons, therefore, that the incidence rate on the angiosarcoma is the most- suitable end-point for analysis of risk of exposure to vinyl chloride for a number of reasons: 1. Vinyl chloride angiosarcoma, is a rare cancer in unexposed populations, thereby making the utilization of angiosarcoma as a demonstration of vinyl chloride exposure on the basis of work history truly a reasonable approach. 2. Angiosarcoma has been demonstrated to occur both in animals and humans when exposed to vinyl chloride. 3. It is therefore demonstrated unlikely that any other car cinogenic result from vinyl chloride would incur lower exposures than those lowest exposures that would induce angiosarcoma. Recent publications entitled vinvl Chloride. AD__Assessment of the Risk of occupational Exposure,, was published in 1987 in the Fundamentals of Chemical Toxicology, laumal, Volume 25, pages 187 to 202, 1987, authored by 0000G8 SPI-05616 t 2 Point One: Pharmacokinetic- Information- There- ara several publications'in the literature not cited. in the; DH_ document, that address the incorporation of pharmacokinetics: in low dose- risk estimation for* chemical, carcinogenesis- One. suet article was. published. as. far-back as 1980 in Toxicology-and- Applied Pharmacol ogy. authored^ by Andersorrr Hoel. and.' Kaplair-- That document demonstrates- how to incorporate the pharmacokinetic, information on vinyl chloride into a risk- assessment approach^ forr low- dose; risk, estimation. There-are numerous otherr publications: on the phar macokinetics of vinyl chloride as well. Another- such, document, published in 1981 in the- Archives of Toxicology authored by Bolt, Filser and Buchterv demonstrates significant information that is relevant when extrapolating low level carcinogenic- risk estimates, from the existing-data base. The DHS document fails, to incorporate any of the established pharmacokinetic information in its treatment of theoretical risk for vinyl chloride.. A number of studies indicate that probably a reactive metabolite, not vinyl chloride per- se is responsible for its toxicity. Although some inhaled vinyl chloride is excreted unchanged, depending on dose, a varying amount is metabolized- The metabolism of vinyl chloride has been the subject of numerous, studies and it is currently thought that vinyl chloride is metabolized by epoxidation with subsequent production of chloroacetaldehyde. The further oxidation and conjugation with glutathione are responsible for the metabolites found in the urine. Cehring, et al. analyzed, the metabolic and carcinogenic data from man and laboratory animals, and used several models to predict the incidence in man from the animal data. They found that all models over-predicted the risk to man unless corrections were made for the varying rates of metabolism and for the surface area differences of the different species. Point Two: Epidemiology. There have been many published, epidemio logical investigations of occupational workers exposed to vinyl chloride at a variety of occupational exposure levels. Vinyl chloride may, in fact, be one of the most epidemiologically-studied industrial chemicals in the literature. To dismiss that data and relegate it only for comparative purposes to animal data is unacceptable. OHS demonstrates a bias towards the utilization of animal experiments as a priority over human evidence in their approach to risk assessment. This results in a dramatic over estimate of likely human risk at the low environmental levels being addressed by the document. The DHS goes on to state that risk extrapolations based on the human data yield results they judge to be comparable. The practical aspect of responding to an order of magnitude or two in risk assessment can often be dramatic, therefore risk estimates that yield order of magnitude different estimates of risk are extremely important. When adequate or substantial human evidence exists, that data should be given preferential treatment in the risk assessment process. 000007 SPI-05615 Wiyt Institute A DMvonat The Socmtyet The Ptasocrtndustry: me Sptiobr8r 1989 Hr. Robert-Barham-, Chief. Toxic Air Contaminant: Identification: Branch Air Resources Board Attn: Vinyl Chloride 1102 Q Street: Sacramento, California 95812 Re: Draft Report- on vinvl Chloride Dear Mr. Barham On August 29th, the Vinyl Institute* received the preliminary draft report on vinyl chloride dated July 1989 being prepared by the California Air Resources Board (CARB). There has been, therefore, a limited amount of time for our membership to thoroughly review the documents prior to the comment, deadline. Nevertheless, after reviewing the document, there are at least two areas of discussion that are inadequately treated in the California Air Resources Board (CARB) document. Therefore, most of the comments will be spent on those two areas. They are the phar macokinetic knowledge of vinyl chloride in the risk assessment approach and a total inadequate treatment of the large number of epidemiology studies in the published literature. These are very concisely dismissed by the Department of Health Services (DHS) as being unacceptable to be used in the risk assessment process for regulatory purposes. If thinyo\ aln*titute 13 an operating division of the Society .Incustry, Inc. its members include Air Teed ??eaicals' Borden Chemicals & Plastics, Certain- GeorcHlS3fa_tlon' D.ow Chemical USA, BFGoodrich Company, InduCc>f?0Ja^1On' ccidental Chemical Corporation, PPG tSoeXei U thlnteCa Inc*' and vista Chemical Company. thf <r0K?arues account for more than 80% of the. chloride. productlon of Both vinyl chloride and polyvinyl 000GG6 SPI-05614 Wayne Interchange Plaza II 755 Route 46 West Wayne, NJ 07470 (207) 890-9299 Pages 3- In the past; residual. vinyl chloride concentrations- in PVC' resin*: at tha_ time of shipment; verraa. high- as- 2000 ppsv Curently, PVC- resins contain: about: 10- ppm: residual, vinyLchloride. at:the time, of: shipment and- ma>r lose-vinyl, chloride- at- a rate: of-20 to -50 percent: pet month, during* storage-. In. addition; most-of the. vinyl chloride, will, vaporize- and: escape- during- the high: temperature processesin which PVC resins are melted and made into final products.. Pace A-27 Landfill Emissions- Emissions- of- vinyl chloride, fromlandfills mainly occur* by two mechanisms: 1) direct vinyl chloride emission from disposed wastes which contain vinyl chloride (i.e.. chlorinated organic compounds) ? and and the formation of vinyl chloride, from: the biodegradation of chlorinated, hydrocarbons. It is hoped that by making the previously described suggested changes, the readers of Report A will more readily understand that the major source of VCM emissions in California in landfills is from chlorinated organic waste disposal, -not from the disposal of PVC fabricated consumer and construction industry products. Thank you for the opportunity to comment on the Part A and B Reports. Please feel free to call me at (216) 374-2962 should you have any questions on our proposed additions to these documents. Sincerely, 6661W Kathleen E. Stimler Manager, Government Relations ocooos SPI-05613 Pages 2- 2) Wu, W.; Steerrland*,. IT.; Brown, D.; Wells,- V.; Jonas, J.; SchulteP. anctHalperin,-W. "Cohort- and. Case.-contraLAnaLysec of- Workers: Exposed- to- Vinyl Chloride. -- An. Update.*-. HXOSH- Report: Draft; October-, 1988. K~3) Wong; 0.; Whorton, M.D.; Ragland, 0.; Klassanr, C.; Samuels, D. andrChaxtoxr,. "FinaL Reportr -- An-Update; of: an Epidemiology Study of Vinyl Chloride Workers', 1942-1982.". Prepared- for- Chemical- Manufacturer's Association, October; 17, 1986. The second area of concern- with the CARB document: is more an issue of semantics ? nevertheless, we offer- it- for your consideration The PART" A Report at pages A--1, A-17 and A-27 accurately states the following facts, but we would like to see clarifying phrases added or sentences reordered as described below. Pace A--1 to A--2 * PVC is fabricated for use in several products of which many are used by the construction industry. In California, the identified sources of vinyl chloride emissions are landfills, PVC production and fabrication facilities, and sewage treatment plants, not PVC fabricated products for consumer or construction industry use. Page A-17 to A-18 Plastic Materials and Consumer Products. Plastic products made of PVC and other-vinyl chloride polymers are ubiquitous in. most homes. Because vinyl chloride monomer can remain in the PVC resin for an extended period of time, an indirect source of indoor vinyl chloride emissions may come from the release of unre&cted vinyl chloride monomer* from these plastic products. However-, emissions of unreacted vinyl chloride monomerhave been substantially reduced due to improvements in monomer-stripping technology (Wheeler, 1987). Thus, Consumer products made of PVC resins no longer contain elevated residual levels of vinvl chloride monomer and, therefore. are not expected to be an important contributor of indoor levels of vinvl chloride. 000004 SPI-05612 BESumluU*r Th iWooaneft-Comoarnr3925 Emeo**v Parkway Amen, one *UI3 * September- 6, 1989 Mr: Robert- Barham,. Chief Toxic. Air Contaminant Identification Branch Air Resources: Board Attention: Vinyl Chloride P.0. Box 2815 Sacramento, CA 95812 Comments on Technical Support- Document: Proposed Identification of Vinyl Chloride as a Toxic Air Contaminant: Part A and Part- B Reports------------------------------ Dear Mr. Barham: The BFGoodrich Company welcomes this opportunity to comment on the above-captioned documents and ve would like to commend CARB for accurately assembling and summarizing the extensive data describing vinyl chloride's uses, emissions, physical properties and exposure in California. We have only two comments for your consideration. First, the primary deficiency of the CARB document on identifying VCM as an air toxic from landfills is that it fails to note these important epidemiology studies: 1) Doll, Sir R., (1988) "Effects of Exposure to vinyl Chloride: An Assessment of the Evidence", Scandinavian Journal of Work, Environment, and Health, 14(2) :61-78. 000003 SPI-05611 -2- > Septambecl, l rti An adequate- review- of the. medical, studies of the effect of exposure to vinyl chloride can: not: be satisfactorily completed before the end of the first comment periods Therefore, a request is. being made, forr an extension of the. initial comment period. If you have questions', please call the writer at 216-796-2698. Sincerely, CAS:cas C A See Environmental Engineer Corp Environmental Engineering 000Q02 SPI-05610 I rn 4kwurOhio* 44U6>0001' CORPORATE" ENGINEERING September- 1, 1989 Air Resources- Board Toxic Air Contaminant Identification Branch P.0. Box 2815 Sacramento, California 95812 ATTN: Vinyl Chloride Mr. Robert: Barham, Chief Dear Mr Barham: The following comments are offered in response to the "Report to the Air Resources Board on Vinyl Chloride - Proposed Identification of Vinyl Chloride as a Toxic Air contaminant". Clarification is requested concerning the relationship between the California ambient air quality standard for vinyl chloride - 10 ppb, as it was discussed in the report, the level of concentration of vinyl chloride which poses "no significant risk" to the population 0.3 micrograms/day and the interaction of these two values in the regulation of toxic air contaminants. In the sampling and determination of the concentration of vinyl chloride, the use of analytical techniques comparable to and as re liable as the method outlined in the report should be permitted. 000001 SPI-05609 * I. COMMENTS RECEIVED SPI-05608 B. Comments., froarWaste Management-of North'America. Inc. 46 C. Commits 72 II. AIR RESOURCES B ,S TO COMMENTS ON PART A A. Response*to the Comment from the United States Environmental 79 Protection Agency B. Responses to the Comments from Haste Management of North America, Incorporated 79 III. DEPARTMENTOF HEALTH SERVICES'RESPONSES TO COMMENTS*ON PART B A. Responses to Comments from Waste Management of North America, Incorporated 82 B. Responses 88 SPI-05607 PART'C TABLE. OF CONTENTS PABL-C I. COMMENTS RECEIVED A. Comments fronrth. B. Comments- fra^the e & Rubber-Company Goodrich-Gob C. Comments froarx&e D. Comments from Dr. Roger Atkinson of the^Statfwide Air Pollution Research Center at the University of California, Riverside II. AIR RESOURCES' BOARD STAFF RESPONSES' TO COMMENTS ON PART A A. Responses to Comments from the Goodyear Tire & Rubber Company B. Responses to Comments from the BF Goodrich Company C. Responses to Comments from Dr. Roger Atkinson III. DEPARTMENT OF HEALTH SERVICES RESPONSES TO COMMENTS ON PART B A. Responses to Comments from theQMnyl Chloride Institute B. Responses to Comments from the Goodyear Tire & Rubber Company C. Responses to Comments from thd^BFGoodrich^Company IV. LANDFILL GAS TESTING PROGRAM UPDATE V. AIR RESOURCES BOARD STAFF LETTER TO GOODYEAR TIRE & RUBBER COMPANY REGARDING THE REQUEST FOR AN EXTENSION OF THE FIRST COMMENT PERIOD EAR! C ADDENDUM I. COMMENTS RECEIVED A. Comment from the United States Environmental Protection Agency Pane 1 3 6 27 31 31 32 33 38 39 40 41 44 SPI-05606 PART C STAFF RESPONSES TO PUBLIC COMMENTS ON THE VINYL CHLORIDE REPORT Prepared by the Staffs of the Air Resources Board and the Department of Health Services October 1990 SPI-05605 t row Comments: - page. 2 NOn Radical Reaction. A rate constant-for the gas-phase reaction of the MO- radical with vinyl chloride, has. recently been obtainedusing- a relative rate- techniquer (R. Atkinson, S. M. Aschoann and M. A. Goodman;. Int. J. Chea- Kiner., 299-307, 1987). Combining the measured rate constant ratio at 298 t 2 t of ktNO^ vinyl ehloridel/klNO^ ethene) * 2.08 i C.v9 with the room temperature rate constant for the reaction of the rati cal. ith ethene of 2.1 x 10-1^ em^ molecule"' s"' (R. Atkinson,. S. M. Ascasann and J. M. Pitts, Jr., J. Phys. Chem., 2, 3*154-3457, 1988) leads, to a rate constant of kdtfO^ vinyl chloride) s A.4 x 10"* cm^ molecule"' s"* at 298 i 2K. Lifetime. As noted, the lifetime of vinyl chloride in the troposphere is calculated by combining the measured rate constants for the gas-phase reactions with OH and NO^ radicals and 0^ (and other gas-phase loss processes, if applicable) with measured or estimated acaientconcentrations of OH and NO^ radicals and O^. Few, if any, reliable real time measurements of ambient tropospheric OH radical concentrations exist to date. The most reliable global tropospheric OH radical concentration value is that derived from- the ambient tropospheric concentrations and emission inventory of methylchlorofomr, leading to an annually and diurnally averaged global tropospheric concentration of 7.7 x 10^ moleculecm"3 (Prinn et al., 1987). For the NO^ radical,, the measured lower- tropospheric concentrations over continental areas range from <1 part-pertrillion (ppt) up to 430 ppc (see R. Atkinson, A. M. Winer and J. N. Pitts, Jr., Atmos. Environ., 20, 331-339, 1986). An average value of 10 ppt (2.4 x 10 molecule cm"^) seems-reasonable, with cr.e recognition that this concentration is uncertain at any given- tine oy a factor cf t 10. 000029 gp\.05687 *# ** l Comments: - page 3 V \ \ \ With these ambient OH and NO^ radical concentrations', the ra.1 euU. lifetimes of vinyl chloride with, respect: to. reaction wittr OH andoNO-jradicals are then 2.3- days and 220 days, respectively^ Since the-llfetlmem of vinyl chloride with respect to reaction with Oj is. (Table-IV-2) -50 oavs (using- the rate data cf Zhang et al. anc Cay et: al.) , the OH radtea.3. reaction appears to be the dominant tropospneric loss process for vinyl, chloride. \ r ooooco SPI-05638 II. AIR RESOURCES BOARD STAFF RESPONSES' TO COMMENTS ON PART A SPI-05639 Conieonts? from-the- SoodyeacrT frr& Rubber Company- 1. Comment? Clarification is. requested., concerning' therelationshlpbetweerrthe: California IQ.ppb;ambient.air*quality standard for vinyl chlortde-andrthe0.3-ug/day concentration-of-vinyl chloride: which: poses, no: signifleant, risk. to. the-popislatlorn- Responser This, comment 1r. addressed?1n Part-C,. I III- Department of Health Services?Responses;to- Comments- on-Part B- 2. Comment? In; the^sampllngrand;determination- of the?concentretiorr of-vinyl chloride?, the?usnof analytical techniques?comparable toand: as re 11 ab 1 er asw the: method- out 11 ned-1 rr the: report shou 1 d be?permitted. Response; The ARB did not intend to imply that the sampling and analysis techniques?described-in the preliminary-draft report on vinyl chloride:should be?the?only method?used.by facilities testing for vinyl chlortde.- Comments fronr the B.F. Goodrich Company 1. Comment; On page A-l and A-2. the report should clarify that polyvinyl chloride (PVC) products used by consumers- and the construction industry are not sources of vinyl chloride. Response: Page A-2 of the second draft report states that finished commercial PVC"products are not expected to be significant sources of vinyl chloride due to current processing and shipping procedures. ARB staff can not conclude that these products have absolutely no-vinyl chloride associated with thenr. 2. Comment: On page A-17 and A-18, the report should emphasize that consumer-products of PVC no longer contain elevated-residual levels of vinyl chloride.monomer-and are not expected to be important contributors to indoor levels.of vinyl chloride. Response: The last sentence on page A-17 of the preliminary draft report states: "Thus-, consumer products made of PVC resins no longer contain elevated..levels of vinyl chloride monomer and, therefore, are not: expected-to be an important contributor of indoor levels of vinyl chloride." 3. Commontr On page-A-27'the ninth line from the top, the report should insert "l.e., chlorinated organic, compounds* after, "which contain vinyl chloride*. Response: The preliminary draft report states: "Emissions of vinyl chloride from landfills mainly occur by two mechanisms: 1) direct vinyl chloride emissions from disposed wastes which 1- OGOOCi SPI-05640 tr \.Il %* conteiir vinyl- chloride;: andr2) ther format! on? of vinyl chloride? fromrthe:biodegradation- of' clrlorinatedshydrocacbons./ The "chlorinated: organic: compounds^ referred to im the-comment-are addressed:by the- second- mechanisirc- C. Comments from*. Dr: Roger- Atkinson- of therStatewide^ Air PollutionResearch* Center-atthe.University of California, Riverside 1. Comments The report* (page*A-44-) should- Indicate-, that the results of thfe study of Liu and coworkerr (A. Liu,. W.A- Muloc,- and C.D~ Jonah-.. Journal of Physical Chemistry. 93.. on4092=4094. 1989) which-determined-absolute-rate constants: for the gas-phases reactionof the hydroxyl radical with vinyl chlortde:over*the: temperature range-of 313 to 423"K' agree-with- those- of Perry-and coworkers. Responser The second draft of the report ref 1 acts: this additional information on page A-41. 2. Comment: The report (page A-44-) should include the most-rellable estimated average hydroxyl radica.1 concentration-of 7.7 X 10 molecules cra~J derived by Prinn and coworkerr (Prinn et al., 1987) through the use of the ambient tropospheric concentration and emission inventory of methyl chloroform. Response: This additional Information is Included on page A-41 in the second draft of the report. 3. Comment: The report (page A-46) should Indicate that a study by Tuazon and coworkers (E.C. Tuazon, R. Atkinson, S.M. Aschmann, M.A. Goodman, and A.M. Winer, International Journal of Chemical Kinetics. 20, pp. 241-265, 1988) confirmed the-study by Pitts and coworkers (Pitts et al., 1984) which demonstrated that the reaction of one molecule of vinyl chloride with hydroxyl radicals yields one molecule of formyl chloride. Response: This additional Information is Included on pages A-42 and A-43 in the second draft of the report. 4. Consent: The report (page A-47) should Include new data (R. Atkinson, S.M. Aschmann and M.A. Goodman. International Journal of Chemical Kinetics. 19, pp_ 299-307, 1987 and R. Atkinson, S.M. Aschmann-and J.N. Pitts, Jr. . Journal of Physical Chemistry. 92. pp. 3454-3457, 1988) concerning the rate constant of the gasphase reaction of vinyl chloride, and the nitrate radical. Response: This new data Is Included on pages A-43 and A-44 in the second draft of the report. 2- - 000032 spi-05641 * III. DEPARTMENT"OF HEALTH SERVICES RESPONSES. TO COMMENTS ON PART B SPI-05642 t Response to Coi tnts: I. General. comwiir Co--nt: "There ara at least- two* araaa. of dlseosaion chat are inadaquaealy treated......... They ara the pharmacokinetic, knowledge of vinyl chloride, in the risk assessment: approach- and= a_ total- inadequate* treatment: of the large number* of studies in the published, literature." (sic) Response: DHS staff note the usefulness of the commenter's general, suggestions advocating more explicit consideration of the pharmacokinetic model and. of the epidemiological data in the- quantitative'* risk, assessment-. Therefore, in the revised document; DHS staff have described-quantitatively the Michaelia-Henten kinetic modal, as developed by Gehring-et al. (1978), which the. comsenters specifically mention- The model has been included in the risk analysis of the major epidemiological study and- in the quantitative analysis of the animal studies. II. Specific comments A. Concerning the assertion that the risk assessment does not adequately treat pharmacokinetic.knowledge of vinyl chloride: 1. Comment: The DHS risk assessment did- not cite several pbamacokinatically oriented studies. One such study was. Anderson et al. (1980). Another was Bolt et al. (1981). Response: DHS considered both the references that the commentar mentioned. The original DHS risk assessment cited one of these two references, as well as many ocher references on pharmacokinetics. See pages 2-1 through 2-17, and especially page 2-4, where Bolt et al. (1981) is cited. The original. publle announcement listed the Anderson et al. (1980) paper, but the DHS risk assessment, did not cite that reference because. Che original. DHS risk assessment, did not use the pharmacokinetic approach in the quantitative modelling of risk predictions. That reference obtained- a multistage risk estimate in the lower end of the range of risks, consistent with the DHS calculations for the early Haltoni data- that Anderson- et al. used. The revised risk assessment now cites Anderson et al. (1980). 2- Comment: "The DHS document: fails to Incorporate any of the established pharmacokinetic information in its treatment of theoretical risk for vinyl chloride." Response: The revised document now Includes a pharmacokinetic model in the quantitative prediction of risk. The original version of the document included on pages 8-1 and 8-6 a summary of the implications of the pharmacokinetic, information and concluded-that the pharmacokinetic analysis is not- quantitatively necessary (for: laboratory* rodents) because of sufflcint bioaaaay data, at exposures below- the saturation concentration for** rats. This view-is consistent with an independent analysis, of Krevski et al -- 000033 SPI-05643 (1987). They- reported. that when? baaing- eha quantitative risk, analysis forr racs on dose*: balow. 200*500 ppmp. which is within cha linear. range- of dose- response, there* is-virtually no difference? between-unit risks obealnad. using- administered-dose-and-delivered-dose, aa- obtained* In. a pharmacokinetic nodal-- The revised analysis did- find, a greeter* difference, and. Cha revised version of Che doetinane.parfozBK.cha pKaraaoaklttaelc nodal. 3. Conaancr A reactive* metabolite-is probably responsible for- VC" coxielcyr Response: DHS: agrees-. The original-vinyl chloride risk assassmant document seated at pager 8-1, "the oncogenicity of vinyl chloride appears to ba due-to one or more: reactive- natabolites,. rechar- thaic. cha* parent molecule". Alsor the first santanea in Chapter 2, Metabolism.- and- ?haznacokinadcs, stated-. "Experimental, evidence, has., suggested chat, vinyl chloride must undergo cransfomadon to a reaedve natabollte(s) by the liver to ba toxic." 4. Congener "It is currently thought: Chat- VC is nacabolized by epoxidadon with subsequent production of chloroacetaldahyda. The further oxidation and conjugation with glutathione are responsible for the natabolites found in cha urine." Response: The risk assassaanc mentioned both cha epoxidadon process andthe conjugation vieh glucachiona -- on pages 2-1 and 2-13 respectively. Both also appeared in eha XARC diagram, which is Fig. 2.1. 5. Consent: Cehring found chat savaral models overpredicted the risk co man unless corrected for varying races of metabolism end for surface area differences of Che different species. Response: In 1978 Cehring et el. used pharaaeokinecies in fitting a probit model to observed cancer races in Che rat bioassay. Those authors Chen went on to use surface area scaling on the assumed- rate of metabolism co extrapolaee the results from racs co conpare co a human risk measurement, derived from an occupational study (Fox and Collier, 1977) . In 1979 Cehring c al. used the same, pharmacokinetics in fitting four models co observed, cancer races in Che rat: bioassay. Those authors, then went on to extrapolate all four results from racs to compare to an occupational risk study that was then recently completed-by Equitable Environmental. Health (EEH, 1978). The comparison by Cehring et al. considered the probit prediction to be in satisfactory agreement with the new human- measurement without any scaling of risk by surface area. Of Che remaining chree models, Che auchors reported one as being coo low and Cha ocher two as being coo high. A follow up study of the occupational, group (Vong et al., 1986; see comment: B-2 below) subsequently indicated much higher rates of human, liver angiosarcoma, chan hadtha earlier study. These last two occupational studies (EEH, 1978 and Vong at al., 1986) remain unpublished. B. Concerning Cha assertion that Che risk assessment does not adequately treat the large number of epidemiology studies in the published literature: 000034 SPI-05644 2 r* 1. Co--ncr To dismiss^ tha^ largo- number- of epidemiology studies* and- co relegate- thamr- only* to comparisons: with animal isunaecapeable. "DHST demonstrates-- a bias- cowards- Cha utilization: of animal, experiments: as- a priority- over* humam avidanca- in chaic approach, to risk- assessmentr. Thia rasulca- in a dramacirovaraarlaaca of- lilcaly human- riak- at cha low* environmental.. lavala. being addressed- by cha document." Tha DHS: judges chat, riskextrapolations based: oir cha human- data- ara- comparable co those of: cha animal predictions, yetr differences of- an. ordar- of magnitudes or two in risk: assessment can. ofcan- have* a-- dramacic. practical-- effect "Whan, adequate or substantial-- human- avidanca existsv that data should, ba given prafarandal. treatment in cha rislc. assessment process.* Responsa: Tha original documanr poincad ouc ac pagas 1*4, 7-55, 8-7, B-2 and B-3 that cha apidamiologieal data ara important: Co- consider- in cha rislc. assessment: but mostly ara not sufficient to construct reliable, dose-responses functions. One of cha main, reasons- for this, limitadotr- is the inability-- of Che occupational, studies co account for cha effaces of sex, tuaorr sice endage of exposure, alL of which are found co ba important in cha animalcarcinogenicity results. Also, chare ara large uncertainties of exposure in Che occupational studies. Tha original document did. make Che comparadve statement ehae, taking all the limicadons of Che occupational studies into account, aehe human risk ascimacas ara consistent with chose obcainad for laboratory- animals." (page 1-4). See also page 8-10. Using suggestions of the commancar about pharmacokinetics, DHS has revised. Che estimate of lifetime unit risk for all cancers to ba 4.5 x 10*^ ppb , based on an. occupational study by tfsxweller ac al. (1976). Thia. asdnaca is only a factor of four less chan cha base animal pradicdona. Such a result represents reasonable consistency, considering that Che occupational results may not taka proper account of cha greater sensitivity found in females, tha greater risk to children, and tha inability of cha human studies to detect any, except relatively large increases, in any specific, type of tumor. The DHS has revised tha document to include tha two most reliable human results, both from cha Waxveiler at al. (1976) study, which do now overlap the narrowed range of risk for animals. 2. Comment: Two updated epidemiology studies, one of over 10,000 workers, ara cited in support of the commancar's position that "DHS' a approach, to dismiss human, epidemiology- evidence in their, riak assessment- is inadequate." Response: Tha original document reviewed epidemiology studies on pages 7-31 through 7-55 and developed quantitative analyses in Appendices B and C. The. documanr cited both tha studies mentioned by tha commancar. Tha first is the. paper of Daher ar il. (1988), which is cited at page 7-46. This paper, which is leas than two pages in length, continues to follow- the same- 593 Dow employees as did tha study of Ott et al. (1975). Tha number-of parsons in the study is still too small to expacr to detect any- affect. Tha second study mentioned by the commancar is the epidemiological-- follow up for tha Chemical Manufacturers Association (CMA), which was summarized in the Tables B-l and B-2 of cha original documanr. This, study- recorded-359 cancer deathsThe SMR for liver and-biliary cancer vaa very large, 641, and the SMR. for ^ra^-n cancer , 180, was statistically significant-. On- page B-10 that study- 000035 SPI-05645 3 I l\.h: fi- vu-iUo; cicadaas-providing sonar evidence against a relationship between* lung: caneat and-vinyl chloride. exposure-. This, work: for-the CMA wan liJtad in the original- bibliography- by the- corporate^ author*, Environmental Haaleh Aaaoeiacaa (1986). Tha risk. assessment has-been raviied. Co uiaa consistent means- of- rafaraneing- chia unpubliabad-worker aa Wong- et al. (1986). DHS a caff haa not put nuchr weight on chia. work- because it doea not appaat Co beprocaading- Co Cha- peer-reviavad-literature and- ic la problaaadc co relate iaoC~ of: tha-smritas-Co axpoaura- 3 Co--anet Liver* angiosarcoma. "Is Cha most suitable end-point for-analysis of risk- of- azpoaura co vinyl chloride." (a) Tha "aoat reasonable inCarpraceclon of Cha data, la eonsiaCant wide. thm causal, association of vinyl chlorldaand an axcess- of brain cancarr howavar, cha raladve risk calculation for brain, cancar is much lower chan chat for liver cancer." (b) "Only two out of eight seudies on lung cancar- yieldacadadeally-algnlficant rasulca, and because- scudias with cha higher power* ware negative, a causal, associadon is unlikely." (sic) (c) "Vinyl chloride- angiosarcoma. is a rare cancer in unexposed populations, thereby waking cha utilization of angiosarcoma aa a demonscracion of vinyl chlorida exposure on cha basis of work hiscory truly s reasonable epproach." (d) "Angiosarcoma has bean demonstrated co occur both in animals and humans whan exposed to vinyl chlorida." Response: Liver angiosarcoma plays s major role in cha current risk assessment, for tha reasons given by tha comnear. Nevertheless, other sensitive indicators of carcinogenesis, such as breast cancer observed in rodents and several cancers in humans- are also considered. A. Co--end a recant paper- by Purchase ec al. (1987) "demonstrates a much more studied and scientifically defensible approach to assessing risk-of exposure co vinyl chloride." Response: The approach of Purchase ec al. is noc defensible by current standards of risk assessment in. Che U.S. The models chat they use in their risk assessment to Interpret data have become of marginal importance compared co Che muldscage (or single scage) --dal, which has more biological plausibility and also provides more scable escimaces of confidence limics on risk. Expressing Chelc results as dose producing one-in-e-million risk, they use ehe marginal models to produce an excessively large range of dose, vich Che highest, does being 10** cha lowest dose.. The higher* doses ere said Co be cons is cent vich Cha occupational experience, hue chare is no support for thac. stateasnt in spite of a lengthy*' analysis- of. dace on liver angiosarcoma- in vinyl chlorida vorkars in several countries prior co 1982. Tho only dose chac cha paper* derives from the human scudias is from e sketchy environmental effaces analysis of Barr (1982). See Che next item. 5. Comment: Bart (1982) conducted. an. analysis of liver angiosarcoma cases chat: could- be- located aaong populations inferred, to be living, in- the. vicinity of VCM production, facilities. The results- suggest Chat "100 ppb represented Che 000036 4 SPI-05646 I AFT estimated dose representing a 1 x 10'6 lifetime risk in man. That value is similar to the.highest estimate derived from the animal data when taking biotransformation into account." Response: The risk assessment did not cite the study of Barr (1982) with its brief analysis of liver angiosarcoma because chat analysis is so unsubstantial epidemiclogically and the work remains unpublished in the peerreviewed literature. As a counter to Barr's brief analysis, a well considered recent assessment by the Committee on the Evaluation of Carcinogenic Substances, National Health Council of the Netherlands (1987), published in the scientific literature, has found carcinogenic risk based on published occupational studies to be one in a million per ppb, which was about the same as found in the original DHS risk assessment, 2.1 x 10'6/ppb, before revising the model to take account of the pharmacokinetics of vinyl chloride. 000037 5 SPI-05647 z Response to Comments: . The Goodyear Tire and Rubber Company Comment: "Clarification is requested concerning the relationship between the California ambient air quality standard for vinyl chloride - 10 ppb, as it was discussed in the report, the level of concentration of vinyl chloride which poses "no significant risk" to the population - 0.3 micrograms/day and the interaction of these two values in the regulation of toxic air contaminants." (sic) Response: As pointed out at page A-l of the risk assessment document, the Air Resources Board in 1978 adopted 10 ppb as the ambient air quality standard for vinyl chloride in California. That standard is not to be exceeded in air within the jurisdiction of the Air Resources Board. The rate of intake of vinyl chloride which poses "no significant risk" under Health and Safety Code 25249.10 is 0.3 j*g/day. OHS determined that intake rate to ensure that the estimated lifetime risk of cancer from intake of vinyl chloride by all routes is less than 10*^ or one chance in a hundred thousand, taking the carcinogenic potency of vinyl chloride to be 2.3/(mg/kg- day) in accordance with the U.S. EPA (1984) assessment based on a diet study. For exposure by inhalation alone that EPA potency is equivalent to a unit risk of 7 x 10* ppb*3- vinyl chloride for a 70kg human breathing 20 m3/day with 40% absorption. Thus, the potency used to calculate the current intake rate for no significant risk corresponds to a unit risk that is above the range of unit risks for inhalation in the revised risk assessment document. See Figure 8.1 of the revised document for more information. The quantitative relationship of the 0.3 Mg/day intake rate to the 10 ppb air quality standard is obtained by converting the 10 ppb (26 ug/m ) to its equivalent intake rata of 210 Mg/day for a human breathing 20 nr/day with 40% absorption. Thus, the air quality standard, which was set at the detection limit at the time of adoption (1978), is 690 times greater than the existing DHS determination of intake rate posing "no significant risk." oooors 6 SPI-05648 I Response to Comments: The B.F. Goodrich Company Comment: The "primary deficiency of the CARB document on identifying VCM as an air toxic from landfills is chat it fails to note these important epidemiology studies: 1. Doll, Sir R. , (1988) "Effects of Exposure to Vinyl Chloride:an Assessment of the Evidence", Scandinavian Journal of Work, Environment, and Health, 14(2):61-78. 2. Wu, W. ; Steenland, K.; Brown, D.; Wells, V.; Jones, J.; Schulte, P. and Halperin, U. "Cohort and Case-Control Analyses of Workers Exposed to Vinyl Chloride - an Update". NIOSH Report Draft, October, 1988. 3. Wong, 0.; Whorton, M.D.; Ragland, D. ; Klassen. C.; Samuels, D. and Chaxton, K. "Final Report An Update of an Epidemiology Study of Vinyl Chloride Workers, 1942-1982". Prepared for Chemical Manufacturer's Association, October 17, 1986." Response: Reference to Doll's recent review of cancer mortality in occupational studies is a useful addition to the risk assessment, and it has been included in the revised document. ' The Wu et al. study has recently been published in the Journal of Occupational Medicine 31(6) 518-523 (1989). That study provides useful additional Information on following up the worker outcomes for one of the four plants of the Waxweiler (1976.) study of vinyl chloride workers. DHS staff has included a discussion of this recent work in the revision. The Wong et al study, an industry-wide compilation, remains unpublished. Nevertheless, the original version of the risk assessment did cite it by authors in Tables B-l and B-2, and by corporate authorship, Environmental Health Associates, in the list of references. The revision uses a consistent method to site this work (Uong et al., 1986). SPI-05649 1 IV. LANDFILL GAS TESTING PROGRAM UPDATE SPI-05650 I IV. LANDFILL GAS TESTING PROGRAM UPDATE Landfill Gas Testing Program data on page A-31 of the preliminary draft report were amended on page A-29 of the second draft to include test results through December 1989. 000040 SPI-05651 V. AIR RESOURCES BOARD STAFF LETTER TO THE GOODYEAR TIRE AND RUBBER COMPANY REGARDING THE REQUEST FOR AN EXTENSION OF THE FIRST COMMENT PERIOD SPI-05652 AJR RESOURCES :- * 7jjj* ; tc\ zv: .A.:a*^tsTo Ca mt: ECaRD G#orQt DtufcmfMin, rs r> * s t rv H 5 L*' `if*1 September 29, 1389 C.A. See Corporate Environmental Engineering Department 1100 Goodyear Tire & Rubber Company 1144 East Market Street Akron, Ohio 44316-0001 Dear Ms . See: Thank you for your response to the craft report p-peosed ! d en t i f i e a t i on of Vinyl "Chloride as a Tcric Air Cant ami rant. Your comments will be considered and addressed Part C of the- second draft of the report. in The second draft of the report will be mailed to you and other members of the public fcr final review. It will include Parts A, B, and C of the report as well as an executive summary which summarizes Parts A end B. A 2C-day comment perioc will be given for your review. During this comment period, only comments on the executive summary anc any revisions made to the report will be accepted. All of the comments received and our responses will then be incorporated as an addendum to Part C. The final draft report, including Part C, will then be submitted to the Scientific Review Panel for its review. The Scientific Review Fanel has requested that all public comments be directed to the Air Resources Board within the time spans allotted for the two comment periods. In accordance with this process, we are unable to extend the first comment period as you requested. If you have any questions, please call me at (916) 322-7072. Sincerely, Robert Barham, Chief Toxic Air Contaminant Identification Branch 000061 SPI-05653 DRAF uniiThe Goodyear Hre&R erComyaixy Akron, Ohio outaie -0001 CORPORATE ENGINEERING September 1, 1969 Air Resources Board Toxic Air Contaminant Identification Branch P.0. Box 2815 Sacramento, California 95812 ATTN: Vinyl Chloride Mr. Robert Barham, Chief Dear Mr Barham: The following comments are offered in response to the "Report to the Air Resources Board on Vinyl Chloride - Proposed identification of Vinyl Chloride as a Toxic Air Contaminant". Clarification is requested concerning the relationship between the California ambient air quality standard for vinyl chloride - 10 ppb, as it was discussed in the report, the level of concentration of vinyl chloride which poses "no significant risk" to the population 0.3 micrograms/day and the interaction of these two values in the regulation of toxic air contaminants. In the sampling and determination of the concentration of vinyl chloride, the use of analytical techniques comparable to and as re liable as the method outlined in the report should be permitted. 0000- SPI-05654 t -2- September 1, 19S& SH) f\ FT "T U f\ n t An adequate review of the medical studies of the effect of exposure to vinyl chloride can not be satisfactorily completed before the end of the first comment period. Therefore, a request is being made for an extension of the initial comment period. If you have questions, please call the writer at 216-796-2698. CAS:cas Sincerely, C. * C A See Environmental Engineer Corp Environmental Engineering OOGO^o SPI-05655 I. PART C ADDENDUM COMMENTS RECEIVED SPI-05656 t UNITED STATES ENVIRONMENTAL PROTECTION AGENCY REGION IX 215 Fremont Street San Francisco. CA 94105 June 8,1990 Genevieve Shiroma, Chief Toxic Air Contaminant Identification Branch Air Resources Board Atm: Vinvi Chloride P.O. Box 2815 Sacramento, CA 95812 Dear Ms. Shiroma, Thank you for the opportunity to comment on the Air Resources Board's technical support document entitled "Proposed Identification of Vinyl Chloride as a Toxic Air Contaminant" dated May 1990. Please incorporate the comments listed below into the the final report. Also, the Environmental Protection Agency's risk assessment group is conducting a detailed review of the report. Any additional comments resulting from this review will be delivered by June 22 of this month. Ms. Barbara Cook of your office assured me that these additional comments will be addressed by the Scientific Review Panel. Please note that the Operating Industries, Incorporated (Oil) landfill is- currently a federally listed Superfund site. As part of the Remedial Investigation at the site, EPA is conducting a 12-month ambient air quality study at the Oil landfill. Twenty-four hour air samples are being collected every eighth day at nine permanently located stations (including 2 background stations) near the landfill. The detection limit for vinyl chloride for this study is 0.30 parts per billion. Meteorological data is also being collected for this study. The results of this study will be used to support EPA's risk assessment for the Oil landfill. Please include the following paragraph in the Executive Summary: The Operating Industries. Incorporated (Oil) landfill is currently a federally listed Superfund site. Subsequent to the Air Resources Board's vinyl chloride sampling during 1987, the Environmental Protection Agency (EPA) has implemented more stringent landfill gas control measures. EPA has also selected a remedy for landfill gas control that is expected to substantially reduce landfill gas emissions from the Oil landfill. It is fully anticipated that these control measures will substantially lower the levels of vinvi chloride in the ambient air in the vicinitv of the Oil landfill. Thank you for the opportunity to comment. SPI-05657 I /asie :vianacem>r- .Norm Amenca. inc. Government aita:-;2S L Sireet. Suite srO '.acramento. s.--..o-"in sSSU e'iO.t--ST* */*oa-4^r June 11, 1990 Genevieve Shiroma, Chief Toxic Air Contaminant Identification Branch AIR RESOURCES BOARD P.O. BOX 2815 Sacramento, CA 95812 ATTENTION: V^nyl Chloride SUBJECT: PROPOSED IDENTIFICATION OF VINYL CHLORIDE AS A TOXIC AIR CONTAMINANT BY THE CALIFORNIA AIR RESOURCES BOARD (ARB) Thank you for the opportunity to provide comment on the ARB' s proposal to identify vinyl chloride as a toxic air contaminant, waste management of North America (WMNA) is a comprehensive waste management services company owning and operating, among other things, landfills and waste hauling companies in the State of California. In addition, Chemical Waste Management, Inc. (CWM) provides comprehensive hazardous waste management services including hazardous waste collection, transportation, treatment, and disposal in California. Both WMNA and CWM are supportive of your efforts to identify vinyl chloride as a toxic air contaminant. Indeed, identification of this compound as a toxic air contaminant is mandated by state law by virtue of the fact that it is identified as a hazardous air pollutant pursuant'to federal law. However, we are concerned about the bases for identification that are contained in your staff report in two primary areas: 1. Presence of vinyl chloride in the atmosphere and the inference that landfills in California are the principle source of this proposed toxic air contaminant, and 2. The degree of public health risk that is posed by vinyl chloride. GO GO SPI-05658 ARB/Vinyl Chloride June 11, 1990 Page 2 t LANDFILLS AS A SOURCE OF VINYL CHLORIDE On page A-23, the second paragraph states, "Based on the emission estimates for two landfills in California (BKK and Oil), landfills are the largest identified source category of vinyl chloride emissions in the state. The information necessary to estimate vinyl chloride emissions for the hundreds of other landfills in California is not available." Other references to landfills being the largest source of vinyl chloride emissions are made elsewhere throughout the report. It is erroneous to assume that these two landfills are representative of all landfills. Both BKK and Oil are landfills that are currently included on the state superfund list of hazardous substance release sites. Both of these sites are reported to have accepted significant quantities of waste vinyl chloride during their operating life. In fact, contrary to the statement made above, significant information DOES exist that landfills are NOT a significant source. The Air SWAT programs mandated by Health and Safety Code Section 41805.5 show that waste management units operated by WMNA are not a significant source of vinyl chloride emissions. Unfortunately the ARB's report makes only passing reference to the Air SWAT data. Even this passing reference indicates that, while the presence of vinyl chloride has been detected in some landfills, the concentrations and amounts are vastly lower that those represented by BKK and Oil. Rather than attribute vinyl chloride emissions to landfills, the report would be more accurate in attributing such emissions to superfund sites that once received vinyl chloride waste for disposal. Attached to this letter I have included summary tables of the Air SWAT results for six of the landfills owned and operated by WMNA. This data shows that, while vinyl chloride is detectable at low to very low levels within the landfills themselves it is, with only minor exception, virtually undetectable in surface samples and in downwind ambient air samples. Finalization of the rulemaking for vinyl chloride as a toxic air contaminant should be delayed until this recent and very critical information can be properly incorporated into the report. In fact, section 39660(f) of the Health and Safety Code mandates that DHS and the ARB give priority to the evaluation and regulation of substances as air toxic contaminants based on a variety of factors including amount or potential amount of emissions and ambient concentrations in the community. To proceed with identification of vinyl chloride as a toxic air contaminant while identifying landfills as the largest source of emissions based on two unrepresentative sites would be 0300 -G SPI-05659 ARB/Vinyl Chloride June 11, 1990 Page 3 a disservice to the waste management industry and contrary to state law. This is made even more true by not using readily available Air SWAT data which provides a much more accurate indication of the true contribution of waste management units to emissions of vinyl chloride. PUBLIC HEALTH RISK OF VINYL CHLORIDE While we do believe that it is ultimately appropriate to regulate vinyl chloride as an air toxic contaminant, we are concerned that the unit risk factor that you have attributed to this compound is overly conservative. I have also attached to this letter a copy of a brief paper on Carcinogenic Risks from Landfill Emissions dated June 6, 1988. This paper was submitted in comment on a preliminary draft document circulated by EPA in March, 1988, "Air Emissions from Municipal Solid Waste Landfills--Background Information for Proposed Standards and Guidelines". This information provides a much more realistic assessment of the health risks posed by municipal landfills not only from the standpoint of vinyl chloride but a number of other compounds as well. In summary this brief paper, based on an assessment of the cumulative impact of all landfill emissions, concludes, "Using a dispersion model for area emissions, we find that for persons spending their whole lives 100 m from the edge of such a landfill the lifetime risk is about 20 x 10'6, while even for persons staying permanently at the edge of the landfill the lifetime risk is only 50 x 10`6." In addition, I have attached some specific comments prepared by Dave Dolan, Waste Management Inc. toxicologist, listing specific concerns we have pertaining to the risk assessment information contained in the ARB's Technical Support Document for Vinyl Chloride. The report Mr. Dolan cites in his second item (U.s. EPA, 1985) is entitled, "Techniques for the Assessment of the Carcinogenic Risk to the U.S. Population due to Exposure from Selected Volatile Organic Compounds from Drinking Water via the Ingestion, Inhalation, and Dermal Routes". OOOCi SPI-05660 ARB/Vinyl Chloride June 11, 1990 Page 4 t RECOMMENDATION Due to the fact that the ARB knows that the Air SWAT data is now available to assess the impact of vinyl chloride, identification of vinyl chloride as an air toxic contaminant should more properly be delayed until this information can be included in the report to provide a realistic assessment of landfills as a very limited source of risk to adjacent communities. Thank you for the opportunity to comment on your draft Technical Support Document. If you have any questions or concerns pertaining to these comments, please do not hesitate to contact me. se ctaries A. White, Manager Regulatory Affairs CAW:fal Attachments cc: Dave Dolan Sara Broadbent Sue Briggum 0G0C>S SPI-05661 ATTACHMENT 1 TAILS 2.1 SUMMARY Of ALTAMONT ASWAT RESULTS (ppbv) Primary Target Momtonno Compound Vinyl Chloride Suoolemertta! Taroet Monitorinaiomoounds Benzene Ethylene Oibromide Ethylene Oichloride Methylene Chloride Perchioroethylent Carbon Tetrachloride Methyl Chloroform Tri ehloroethylene Chloroform Methane Gas Characterization Landfill Gas Integrated Surface Sample 3.000 3.0 <500 <50 32 21,0001*) 5.AOO <5 210 8,600 430 480.000,000 <2.0 <0.5 <0.2 <1.0 0.8(b) <0.2 0.5(0 0.9(d) <0.8 3.000 Ambient Air (Net Oownwind Upwind Concentration) 24-Hour Continuous 0.0 - - - Note: ppbv * parts per billion by volume. (a) This result is potentially due to limitations of the analytical methods specified by the ARB in the Testing Guidelines (i.e., a non-Calderon constituent may coelute with methylene chloride). (b) Altamont integrated surface sample value of 0.8 ppbv for perchloroethylene is similar to the ARB background value of 0.6 ppbv for the Bay Area Region (1985 data) in wnich Altamont is located. (c) Altamont integrated surface sample value of 0.5 ppbv for methyl chloroform is nearly identical to the laboratory detection limit (<0.5 ppo), and is well below the ARB background value of approximately 2.1 ppbv for the Bay Area Region (1985 data) in which Altamont is located. (d) Altamont integrated surface sample value of 0.9 ppbv for trichforoethylene is similar to the ARB background value of approximately 0.5 ppov forth# Bay Area Region (1985 data) in which Altamont it located. oooo- 2-5 SPI-05662 I TABU 2*1 SUMMARY OF LANCASTER ASWAT RESULTS (ppbv) Primary Target Momtonno Comoound Vinyl Chloride Suoolemental Taroet Monitoring Compounds Benzene Ethylene Oibromide Ethylene Dichloride Methylene Chloride Perehlorocthylene Carbon Tetrachloride Methyl Chloroform Trichloroethylene Chloroform Methane Cat Charactanxation Landfill Gasfd Integrated Surface Sample 4,629 <2.0 571 <1 <20 2,094 578 <5 824 442 40 62,000,000 <2.0 <0.5 <0.2 <1.0 0.4(a) 0.8 <0.5 <0.6 <0.8 10,000 Ambient Air (Nat Oownwind Increase Comoared with Upwind Concentrations) 24-Hour Continuously Directionally Control led<) 0.2(b) 0.0 0.0 0.0 0.0 4.0(b) 0.2 <M 4.8 <b) 0.9(b) 0.0 0.0 . 0.0 0.0 0.0 2JW 0.0 0.7(b) 3.7(b) 0.0 0.0 Not*: ppbv a parts par billion by volume. () ^neaiter integrated surface sample value of 0.4 opbv for perehloroathylene is similar to the ARB background value of 0.2 ppbv for the Southwest Oesert Region in which Lancaster is located (see Table 2-2). (b) These downwind ambient increments are greater than expected considering the low concentrations for the integrated surface samples and landfill gas samples. However, these downwind increments are less than 5 ppbv, which corresponds to inherent data uncertainties 2? AI4l56,#nt *,rdata associated with limitations of the analytical methods specified by the ARB in the Testing Guidelines. (0 Based on composite data which indudes ail samples. 0w r> U r^r- 2-5 SPI-05663 TAILS 2*1 SUMMARY OF DAVIS STRUT ASWAT RESULTS (ppbv) Virtyi Chloride Benzene Ethylene Oibromide Ethylene Oichloride Methylene Chloride Percnloroethylene Carbon Tetraehloride Methyl Chloroform Trichloroethylene Chloroform Methane Gas Characterization Landfill Gas <500 <S00 <1 <20 <60 <10 <5 <10 <10 <2 530,000,000 Sample <2.0 <2.0 <0.5 <0.2 8(b) <0.2 <0.2 1.1(0 <0.6 <0.8 <2000 Ambient 24>Hour Ambient Air (Net Downwind Increase Compared with Upwind Concentration)!*) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.4W) 0.0 0.0 Note ppbv parts per billion by volume. (a) Based on composite data for all samoling days. (b> Oavis Street irrtagrataP surfact samoic value is higher than exoected considering the noneetecuon of this constituent in the landfill gas samote. The reoorted value may nave been affected by amoient baclcground levels, which may exceed 10 pebv in the Bay Area (see Table 2*2), ano/or limitations of the ASWAT analytical methods soeeified by the ARB In the Testing Guidelines, which may result in data uncertainties of approximately S PObv. (c) Davis Street integrated surface sample value of 1.1 ppbv is higher than exoecteo considering the nonaetection of this constituent in the landfill gas sample, ihe reoorted value is similar to tne ... bacltgrouna value of 1.4 ppbv based on 8AAQMD data tor San teanoro. (Sec Table 2-2.) (d) This downwind concentration increment mav be due to limitations of the ASWAT analytical methods specified by the AR8 in the Testing Guidelines, which may result in data uncertainties of approximately S ppbv. The results presentee above oe not include the pnmary sampler results, which have a contamination bias of approximately 4 ppbv. 000051 2-5 SPI-05664 TAILES-1 SUMMARY OF DURHAM ROAO ASWAT RESULTS (ppbv) Primary Taroet Momtorino Compound Vinyl Chloride Supplemental Target Monitohna Compounds Benzene Ethylene Oibromide Ethylene Dichloride Methylene Chloride Perchloroethyiene Carbon Tatrachlohde Methyl Chloroform Trichloroethylene Chloroform Methane Cat Characterization linQTill yis integrated Surface Sample 3,000 <2.0 1,000 <1 <20 7,500 5,200 <S 300 2,000 260 520,000,000 <2.0 <0.5 0.2 <1.0 0.3<*> <0.2 1.3> <0.6 <0.8 <2,000 Ambient Air (Net Downwind increese Compareo to Upwind Concentration) 24-Hour Continuous 0.0 - -- Note: ppbv * parts per billion by volume. (a) Durham Road integrated surface sample value of 0.3 ppbv for perchloroethyiene is similar to the ARB background value of 0.6 ppbv for the Bay Area Region (1985 data), in wnich Durham Road is located. (b) Ourham Road integrated surface sample value of 1.3 ppbv for methyl chloroform is similar to the ARB background value of approximately 2.1 ppbv for the Bay Area Region (1985 data), in which Durham Road is located. 000G5- 5*2 SPI-05665 I TAILS 2*1 SUMMARY Of SRAOLIY ASWAT RISULTS (ppbv) Momtenno Comoound Vinyl Chloride Supplemental Taroet Momtonnc tomoounds Benxane Ethylene Olbromide Ethylene Diehloride Methylene Chloride Perehloroethyiene Carbon Tetrachloride Methyl Chloroform Trichloroethylene Chloroform Methane Gat Characterization!*) Integrated Landfill Gas Surface Samde Ambient Air (Net Downwind Increase Compared with Uawind Concentrations)!*) 24-Wour Continuous Directionally Controlled 33,8251*1 2.2 0.2 0.0 900 <3 <30 2570 375 <5 <10 1435 <4 500,000,000 1.5 <0.5 <0.2 <1.0 0.4 0.2 7.5(81 <0.6 <0.8 <5,000 0.1 0.0 0.0 0.0 <0.1 <0.1 0.8 0.0 0.3 -- 0.3 0.0 0.0 1.3 0.1 0.0 0.0 0.0 0.6 -- Notes: 1. ppbv parts per billion by velum*. 2. The landfill gas samples were collected in Oeeember 1987/August 1988, integrates surface samples in May 1988/Augutt t988, and ambient samples in May 1988. Based on composite data, which include all downwind samples. <81 This result may have been affected by sample matrix interferences, coetution of constituents with similar GC retention times, and other inherent limitations of the ASWAT analytical methods specified by the ARB in the Testing Guidelines. Ambient air concentration results confirm that Bradley Landfill gas amissions for this constituent do not affect offsite air quality. <4 Value is similar to the ARB range of background values (1.11* 7.07) for the South Coast Region. 0 oortc-' 2-5 SPI-05666 t TAIL! 2*1 SUMMARY OF KIRBY CANYON ASWAT RISULTS (ppbv) Vinyl Chlonde Benxene Ethylene Di bromide Ethylene Dichioride Methylene Chloride Perchloroethyiene Carbon Tetrachloride Methyl Chloroform Trichloroetnyiene Chloroform Methane Gas Charactenzation Landfill Gas 41,00016) 2.500 <1 <20 59.000(b) 2,100 <5 190 2,200 2,000 2,600.000 Emission Screening -- -- -- -- -- -- -- -- -- -- <50.000 Ambient Air 24-Hour Continuous Downwind'*) <2.0 <2.0 <o.s <0.2 <1.0 0.7(0 <0.2 1.0W) <0.6 <0.8 -- Note: ppbv * parts per billion by volume. (a) One sample day with two collocated samplers. (b) These results may have been affected by sample matrix interferences, co elution of constituents with similar GC retention times, and otner inherent limitations of the ASWAT analytical methods specified by the ARB in the Testing Guidelines. Ambient air concentration results confirm that Kirby Canyon Landfill gas emissions for these constituents do not affect offsite air duality (in fact, they were not detected in the amoient samples). (c) This concentration is higher than expecteo considering the low concentration detected in the landfill gas sample. However, this concentration of 0.7 ppov is similar to BAAQMD/ARB results for the San Jose/Bay Area (0.S-0.8 ppbv mean with 1.6 ppbv maximum). Therefore, the Kirby Canyon results are attributable to background conditions. (d) This concentration is higher than expecteo considering the low concentration detected in the landfill gas sample. However, the concentration of 1.0 ppbv is similar to BAAQMD/ARB results for the San Jose/Bay Area (0.6-4.1 ppbv mean with 47.3 ppbv maximum). Therefore, the Kirby Canyon results are attnbutaole to background eonaitions. G n r> 'JU Oc' 1 2-5 SPi-05667 TAILS 2-2 COMPARISON OP KIRBY CANYON AMBIENT AIR RESULTS (ppbv) (BASED ON THE 24440UR CONTINUOUS DOWNWIND STATION) AND AVAILA1U REGIONAL DATA Mean Maximum Number of Observation! Kirby Canyon Vinyl Chlonda Benzene Ethylene Dibromide Ethylene Oichloride Methylene Chloride Ptrchloroetriyiene Carbon Tetrachloride Methyl Chloroform Trichloroethylene Chloroform <2.0 <2.0 <0.5 <0.2 <1.0 0.71*) <0.2 i.oin <0.6 <0.8 S.F. Bay Areal*) (O 1.8-3.2 0-0.01 0.05-0.07 0.7-4.3 0.5-0.8 0.2 0.6-4.1 0.3-0.7 0.03-0.05 San JoselW (0 4.4 (0 (0 2.6 0.S 0.1 1.8 0.3 0.05 Kirby CanyonlO <2.0 <2.0 <0.5 <0.2 <1.0 1.3 0.2 1.2 <0.6 <0.8 S.f. Say Araai*) (0 15.6 0.1 0.3 . 11.9 1.6 0.5 47.3 1.0 0.1 Kirby CanyonW) 2 2 2 2 2 2 2 2 2 2 S.F. Bay Area'** (0 9i 82 84 82 84 83 83 43 84 Nota: ppbv parts par billion by volume. (a) Based on available ARB data for the San Francisco Bay Area {California Toxic Air Quality Oata Summary of 1985 Toxic Air Quality Oata. Preliminary). (b) Basao on available 1986 BAAQMD data for San Jose (Toxic Air Monitoring Summary, 1986-1987, 8oard of Directors Meeting, Seotember 2,1987). (c) Information not available for this report. (d) One sample day with two collocated samplers. (e) This concentration is higher than expected considering the low concentration detected in the landfill gas sample. However, this concentration of 0.7 ppbv is similar to BAAQMO/ARB results for the San Jose/Bay Area (0.S-0.8 ppbv mean with 1.6 ppbv maximum). Therefore, the Kirby Canyon results are attributable to background conditions. (f) This concentration is higher than expecead considtring the low concentration detected in tne lancfill gas sample. However, the concentration of 1.0 ppbv is similar to BAAQMO/ARB results for the San Jose/Bay Area (0.6-4. 1 ppbv mean with 47.3 ppbv maximum). Therefore, the Kirby Canyon results art attributable to background conditions. OOOC^ 2*7 SPJ-05668 ATTACHMENT 31 DATE: June 11, 1990 FROM: TO: RE: Chuck White David Dolan oO >VP Comments on the Identification of Contaminant" Air Resources Board's "Proposed Vinyl Chloride as a Toxic Air I have reviewed Parc B of tha Air Resources Board's "proposed Identification of Vinyl Chloride as a Toxic Air contaminant". The ARB is to complimented for the thoroughness of this report. There are, however, several issues that deserve some attention. Given the short amount of time available for this memo, please excuse its terseness. First, why bother using the linearized multistage model under the pretense that it is a true mechanistic modal (which it is not), when a simple linear regression usually yields nearly identical estimates of q,* (rJ * 0.98)(Personnel conversation with Curtis, Travis, Oak Ridge National Laboratory)? Second, the discussion of uncertainty in the quantitative risk estimates is given short shrift. Although the uncertainties or absence of exposure data in the occupational cohort studies is mentioned, there is no discussion of the conservatism built into the risk estimates by the selection of data for extrapolation, and the extrapolation assumptions, and the effects their underlying assumptions may have on the risk estimates. For instance, the use of the most sensitive sex/strain/specias instead of the average may alter risk estimates by "several orders of magnitude." (U.S. EPA, 1985) Similarly, the issues the extrapolation of rodent potency estimates to humans, particularly on the basis of surface area, and the use of upper 95th percentile estimates of carcinogenic potency instead of the MLE, may alter potency estimates by an order of magnitude, or more. (U.S. EPA, 1985) Third, it is perplexing that the Krewski et al. (1987) chapter is referenced, yet the 36-fold lower carcinogenic potency factor they derive is omitted from the brief discussion. Some discussion on the merits and limitations of the Krewski et al. analysis is necessary. Fourth, the,ARB cites the concordance of the potency estimate derived from the Drew et al. (1983) study and the Maltoni at al. (1984) experiments, it is unclear whether the Maltoni experiments were conducted in his medieval castle/laboratory where the mycobacterium infection is endemic, or in some other facility. (Personnel conversation with E.E. McConnell, National Toxicology Program) In the U.s., mycobacterium infections in test animals would likely violate GLPs, and Berve as grounds for invalidating a study. 0G005G SPI-05669 I Fifth, the recommended use of a potency factor derived from animal instead of the human occupational study of Waxveiler et &1. (1376) is not robust, given that the human data already represents an upper-bound estimate in the target species of concern (i.e., humans). The - additional rationale that the selection of the highest animal estimate is justified by the limited evidence of an effect by age at first exposure (Drew et al., 1983) suggests that perhaps the ARB should consider using a true mechanistic model, perhaps one based upon the MVK model paradigm, as the basis of its potency determinations. cc: Jim McHenry 000057 SPI-05670 ATTACHMENT 2 Carcinogenic Risks from Landfill Emissions. Addendum to Comments on a Preliminary Craft Document circulated by tbs EPA in March,. 1SS8: Air Emissions frsa Municipal Solid Wasre landfills -- Sacksraund Tr.famag.ian for Pronosaci Standards and Guide lines by Edmund A.C. Crouch, ?h. D. and Laura C. Green, Environmental Health and Toxicology Group Meta Systems Inc. Cambridge, MA 02141 ?h. 3. Produced at the request of he National Air Pollution Control Techniques Advisory Committee to the EPA and Waste Management Inc. June 6, 1988 000053 SPI-05671 CONTENTS Introduction............................................................................... 1. Data. Summary .......................................................................... 2. Methodology................................................................ .... 2.1 Emissionrates.................................... ......................... 2.2 Nationwideaverage risk............................................. 2.3 Worst case risk estimates.............................. .... . 1 2 8 8 10 12 0D0C5D SPI-05672 T An estimate of carcinogenic riilea from landfill emissions. Introduction On May 18, 1988, we presented testimony to the National Air Pollution Control Techniques Advisory Committee to the E2A on the risk assessment aspects of a Preliminary Draft Document, "Air Emissions from Municipal Solid Waste Landfills ~~ Background Information for Proposed Standards and Guidelines." The gist cf that testimony was that the carcinogenic risks predicted by the Draft Document were incorrect. The analysis that follows is our attempt to derive such risks more correctly. In particular, we derive estimates of "average" and "worst-case" risks of cancer that could be attributed to volatile organic compounds that may be emitted from municipal solid waste landfills. The estimates are all "standard" in the sense that they are deliberate overestimates, predicated upon "no-thxeshold" models for ail chemical carcinogens off interest. It is our toxicologic opinion that many of the chemicals of interest here are ir. fact likely to contain thresholds in their dose-response curves for carcinogenesis, such that the very low-level exposures involved carry with them no excess risk of cancer to humans. Nonetheless, we have not "taken credit" for this probability, but instead modeled all compounds as if they carry excess risks of cancer at ail non-oero levels of exposure. 020CCO SPI-05673 2 1. Table l.l summarizes measurements of landfill gases collected at 6 .municipal landfills, labelled A to H. All these measurements are given in ppm by volume, and include entrained air (the amount of which can be estimated from the nitrogen and oxygen content of the gas). The landfills labelled B, C, G and H were used in the past for co-disposal of municipal waste and hazardous waste, although this practice has now ceased. This former practice of co-disposal is likely to have led to emissions of larger quantities of chemicals of interest than would have cccured from the disposal of municipal solid waste alone. Table 1.2 shows the average concentrations of components of the emitted gases from each landfill. These averages may he compared with the values given in the EJA Draft Document, Table 3.9. Despite the differing data sources, the average, concentrations are very similar. In this data, carbon tetrachloride was never detected, whereas the ESA data has an average concentration of 0.0115 ppmV. Also, the average concentration of 1,1-dichioroethene (vir.yiidene chloride) is a factor 10 lower here than in the ESA data. In both cases, the concentrations were very low even in the ESA data. Also shown in Table 1.2 are the molecular weights of all the measured components, together with upper bound estimates of "unit risk" for the known carcinogens. These estimates were taken directly from the Carcinogen Assessment Group (CAG) assessments' where they have made such estimates. Otherwise they come from CAG estimates for "potency" of a compound, and assume that a human breathes 20 m3/day of air, and that 100% of a compound is absorbed. At this breathing rate, if a material is present at 1 ug/m3 in air, a person will inhale 20 ug/day or 3.33 * 10 mg/kg-day for a 60 kg person. For a compound with potency ? mg/kg-day, this results in a unit risk (risk frcm 1 ug/m3 of air) of 2.33 x 10 ^ P. The estimates in Table 1.2 generally agree GGGOf1 SPI-05674  3 with, those in the EPA Draft Document, except for carbon tetrachloride, where we take the upper end of a suggested range (EPA uses an average of the range); vinyl chloride, where a more recent estimate by the CAS (which we use) has raised the carcinogenic potency estimate by a large factor; and vinylidene chloride, where our estimate is again substantially higher than that of the EPA Draft Document. Using the molecular weights of the components, together with the unit risks, we can define an average unit risk for the "as measured" average landfill gas. This is obtained by finding the weighted average unit risk for all components, where the weighting factor is the product of molecular weight and volumetric concentration for each component. The result obtained is 1.6 x 10"9 per ug/m3 for the landfill gas including entrained air, and approximately 1.9 x 10"9 per ug/m3 after correction fcr entrained air. We do not use this average, since it is preferable to compute risk estimates on a landfill by landfill case, taking into account the differing concentrations and emission rates at each' landfill. The unit cancer risks estimated by the EPA in Table 2-4 of tne Draft Document suffer from major deficiencies. The first two "scenarios" cannot be justified at all. Averaging together the unit risks of the various carcinogens found in landfill gas could only be justified if there were equal emission ratas (by mass) of those carcinogens, but it is clear that this is incorrect. Furthermore, the Draft Document includes one carcinogen (Table 2- 3, ethylene dichloride) which was apparently never found in their samples cf- landfill gas (Table 3-9, although it is listed twice in Table 3-8) . The "scenario 1" estimate appears to ignore measurements of non-methane VOCs whicn indicate that the major components (certainly more than 75%) are simple alkanes (especially ethane and propane). Furthermore, it is unclear what is meant in this document by non-metbane VOCs. Since chase are OGOCf2 SPI-05675 6 TABLE 1.1 (contd.) Average Concentrations (tcmVl SITE EF G Cartoon tetrachloride nd nd nd Chio rotoenz ene 4.3E-01 nd nd Chloroethane 9.1E-02 7.0E-01 3.8E-01 Chiorefora nd 2.52+00 nd Chioromethane 1.25+00 1.1E+01 3.62+00 0ito rcmochlorcmethane nd nd nd 1/l-Oichloroethana 7.6E-01 2.0E-01 1.6E+00 1,2-Diehloroethane nd nd nd l,1-Dichioroethene nd nd nd t-1,2-Dichioroechene 1.1S-01 nd nd l,2-Dichloropropane 2.2E-01 nd nd c-l,2-Dichlorepropene nd nd nd t-l,3-Dichloropropene nd nd nd Methylene chloride 3.22+00 9.2S+00 1.4E+01 1,1,2,2-Tetrachloroethane 1.1E-01 nd nd Tetrachlorcethene 6.9E+00 3.82+00 1.2S+01 1,1,1-Trichloroethane 2.0E-Q1 nd 1.7E-01 1,1,2-Trichloroethane nd nd nd Trichloroethene 4.1E+00 6.0E-01 2.9E+00 Trichlorofluoroaethane Vinyl Chloride 4.42-01 2.0E-01 7.6E+00 5.32+00 3.12+00 2.6E+00 1,2-Dichloroben2ene nd nd nd 1,3-Dichlorotoenzene 1,4-Dichlorotoenzene nd nd nd nd nd nd H nd 2.15+01 nd nd 6.25-01 nd 2.8E-01 nd nd nd nd nd nd 3.4E-01 nd 1.4E+00 nd nd nd 1.32-01 4.2E+00 nd nd 2.0E+00 Chiorodi1iuoromethane Dichlorodifluorcmethane Dichlorofluoromethane ' 6.9Z-01 6; OE-01 4.OE-01 4.7Z-01 nd nd nd nd 4.6E-01 8.0E-01 2.0E+00 nd Methane Ethane Propane n-Butane n-Pentane n-Hexane Acrylonitrile Benzene Toluene Ethylbenzene Total Xylenes 4.7E+0S 9.5E+02 1.1E+01 nd 4.1E+00 6.12+00 nd 2.7E+00 1-.2E+02 2.4E+01 7.2E+01 5.1E+05 7.6E+02 5.62+01 1.72+01 5.SE+00 6.42+00 nd 2.0E+00 5.8E+01 1.4E+01 3.4E+01 3.7E+05 5.0E+02 1.5E+01 2.8E+00 2.9E+00 S.9E+00 nd 2.0E+00 1.4E+02 2.3E+01 8.4E+01 5.1E+05 1.62+03 2.52+01 nd 1.5E+00 4.0E+00 nd 7.2E+00 5.22+01 2.7E+01 7.12+01 TNMHC (as C6) 9.7E+02 9.9E+02 1.0E+03 1.3E+03 Carbon dioxide Oxygen Nitrogen 3.8E+05 3.7E+QS 3.0E+05 3.12+05 1.7E+04 1.1E+04 6.5E+04 1.8E+04 1.4E+05 1.1E+0S 2.6E+05 1.6E+05 0300' SPI-05676 > 7 TJU3ZX-1.2 At*race concantratlsn wig.iitMi * *** rlafca Avrt?i Mol. cone. weight ppmv unit potency weighted rink unit risfc Carboa tetrachloride Chlorobenzene Chloroethana Chloroform Chlorcmethane oibromochloromethane 1.1-Diehloroethane 1.2-Oiehloroethane 1.1-Dichloroethene t-l,2-0ichloroethane 1.2-Oichiorepropane c-i,2-0ichiorepropene t-l,3-Oiehlorcpropene Methylene chloride 1.1.2.2-Tetrachloreethane Tetrachioroethene 1.1.1-Trichleroethane 1.1.2-?richlrcethane Trichicroetheae Trichicrcfluoromethaae Vinyl Chloride 1.2- Dichlercbenzene 1.3-Dichlorobenzene 1.4-Oichlerobenzer.e Chlerediflucrcmachane Diehlcrcdifluoromethar.e Dichlorcflucrcaethane Methane Ithane Propane n-Butane r.-Pentane n-Hexane Acrylonitrile Pensene Toluene Ethylbenzene Total Xylenes nd 1.55+02 4.35-05 1.35-01 2.75+00 1.15+02 3.35-01 8.55+01 3.15-01 1.55+02 2.35-05 3.95- 4,35+00 5,05+01 nd 2.15+02 2.35+00 2.85-02 9.95+01 9.95+01 3.05-05 9.15-02 3.05- 12 9.85-02 9.75+01 3.95-04 1.25+00 1.35- 10 2.45-01 9.75+01 5.15-02 1.15+02 nd 1.15+02 nd 2.75+01 1.45-02 1.15+02 8,55+01 1.75+02 4.75-07 8.75-05 2.CE-01 3.8E- 11 5.SE- 12 1.45+01 1.75+02 9.55-07 8.15- 11 1.75-01 1,35+02 1.95-05 S.7S-02 1.5E- 11 nd 1.35+02 5.05+00 1.35+02 1.71-06 4.0s -11 1.45+00 5.75+00 1.45+02 6.25+01 9.85-05 2.95-01 1.25 -09 nd 1.55+02 nd 1.-55+0 2 2.55-01 1.55+02 1.35+00 8.65+01 9.45-02 1.25+02 4.45+00 1.0E+02 4.75+05 1.65+01 7.75+02 3.05+01 2.45+01 4.45+01 3.85+00 5.85+01 3.12+00 7.22+01 7.35+00 8.65+01 nd 5.35+01 8.0E-05 2.45-01 2.95+00 7.85+01 8.05-06 6.45-11 7.65+01 9.25+01 l.SE+01 1.15+02 4.45+01 1.15+02 TNMHC (as CS) 8.3E+02 Carbon dioxide Oxygen Nitrogen 3.6E+C5 4,45+01 2.85+04 3.25+01 1.55r05 2.35+01 OOGO'BG SPI-05677 I 8 2. tft&afilOSC_ 2.1 Mt Table l.i gives the concentration* of various gases measured la collected gasea at various landfills. Xiao available for each landfill la the rata at which those gases are released, taking the product of total emission rate for landfill gas with theaa concentrations gives the volumetric emission rate for each gas. This volumetric amission rsta may ba converted to a mass amission rate by using the gas density, which ve approximate by assuming all the gases behave perfectly, ror the 8 landfills considered hart/ the average volumetric emission rate ,is 2.7 x lO6 efd par landfill, from an average amount of refuse in plaee of 5.4 x 10$ teas per landfill. This is about 50% higher than assumed in the E5X Draft Document for wet landfills. from the mass emission rate,, we may use air dispersion modelling to estimate the expected long term average concentrations of each component of the landfill gas at various positions off-sita. The product of these concentrations (in ug/m^) and the upper bound Ufiit risk estimate (aeaaured in units cl 3/ug) gives' an upper bound estimate to lifetime risk. The net effect of all the landfill cis can thus be obtained from the sum over all components of the product of mass emission rate and unit riak for each components. Table 2.1.1 showa this product (in units of a3/s) for ail detected components of the landfill gases whieh have unit risks defined. Xlso shown are the sums of products fer each landfill. OOOO'* SPI-05678 \ 9 7XSZZ2*11 Mi tfllnlan -gateVvai.* j (ral/t) SITS' Chloroform 1,2-0ichidroethane l,l-Dichloroethane Methylene chloride 1,1.2,2-Tetrachlcrcathane Tetrachloroethene 1,1,1-Trichioroethane Trichloroethene Vinyl Chloride Benzene Total AB CD 0000 0 3.SB-03 0 3.82-03 4.32-02 5.22-02 0 2.62-01 3.82-04 4.52-02 1.22-02 2.92-02 00 00 5.42-03 1.82-02 5.82-02 1.42-01 2.0E-03 1.62-03 4.72-03 2.82-02 2.42-03 2.32-02 3.32-02 4.92-02 1.52-01 6.42-01 2.72+00 3.32-01 8.82-03 Si .22-03 5.02-02 4,02-02 2.22-01 7.92-01 2.82+00 8.82-01 SITS T GH Chloroform 1.2-Dichicrcethane 1.1-Dichioroether.e Methylane chloride 1.1.2.2-TatrachIoroethane. Tatrachioroethene 1.1.1-Trichloroethane Trichloroethene Vinyl Chloride Ben:ana 0 1.7S-01 0 0 0000 0000 4.6E-03 7,02-03 3.72-02 1.32-03 4;52-02 0 00 4.02-02 1.12-02 1.22-01 2.22-02 1.92-02 0 2.92-02 0 3.32-02 2.62-03 4.32-02 0 1,22+00 3.62-01 1.12+00 2.62+00 6.12-02 2.42-02 8.22-02 4.52-01 Total 1.42+00 S.82-01 1.42+00 3.12+00 Average total 1.42+00 GOGOLS SPI-05679 I t 10 rid average total for all the landfill! la 1*4 and tie maximum ia 3,1 3/s for landfill 8 <vhlen was used in the past for co-disposal). .In every case# the vinyl chloride present- contributes tne majority of the risk. 2.2 Satlonvlde twragi risk ` To sake an astlmate of the nationwide average risk frss landfill gas now requires a scale-up, together with seme dispersion modelling* The landfills discussed in the previous sections have an average amount of refuse in plaee of 3.4 million tons/ which is 1/300 of the total estimated refuse in plaee in municipal, landfills in the *J.$. (25A Draft Document, page 3-1)', The land area of the contiguous U.S. is about 7.84 x 10$ 'Jca3, so that the D.S. land area per average landfill is about 8.7 x 10$ m3, corresponding to a radius of about 52 Sen. as a first approximation, the effects of landfills on the Q.S *** obtained by finding the average effects of a single -and ll cn a. radius of about SO Jem around it. This approximation *uld -a correct if (l) landfills were uniformly distributed over the ^*S*j {2) ti: population were evenly distributed; and (3) r.o landfill had any effect beyond SO Jem. :t is plausible that the third of t..ese is correct, since the chlorinated vocs which contribute to carcinogenic risk are relatively short-lived (vinyl chloride, for example, has a half-life in air estimated at l.S - 1,8 days). The first two are dearly incorrect, but will be compensated by the overestiaaticn of risks to those dose to landfills (see below) . ;or an average landfill, we have an emission rate x C*rC""^eft~c ua^ risk of 1.4'ir3/3( usi-g the standard . gaussian p*u~e model, the average concentration obtained frea this ever the -adial -ange 0.1 to *0 km corresponds to a lifetime risk of * 10 ' Tiiia 4SSU5Res 4 uniform wind rose, a wind speed of 2 m/$, and a s-aple .averaging ever 7 wind stability classes <x, 3, C, Dday, OOOObO SPI-05680 11 Odight, E, 1\, and- emission heights of 1 a, and a receptor height: pf l.S a. This averaging procedure has been found to give estimates within 20% of those obtained using the ISC model- in particular- cases with observed wind rose and stability class data# provided, the average wind speed is* used. The assumed average wind speed of 3 m/i <6.7 mph) is a reasonable satinets, probably a little low (resulting in an overestirata of risk) for moat of the <7<S. For 67 cities in the SO contiguous states, just 3 report average windspeeds lass than 6.7 mph. The minimum distance used for this averaging, 100 m, corresponds to an estimate of the minimum distance frca the center cf a landfill at which people can be expected to be living. If landfill gas is collected at some landfill, it say be collected together at any point over the landfill. However, if it is collected it will, he flared, so that there is negligible exposure of anybody to it. If if is not collected, then the emissions will take place over the whole landfill, and so the nearest person to the landfill say be eleser than 100 a. in that caae, however, the dispersion modelling performed above is a substantial overestimate for estimating exposures close to the landfill (within distances similar to the dimensions of the landfill', . case estimate. 4 8 *>3 dealt with below for the worst SPI-05681 ____ Ci 4veri* =-. a Ufotin. rule of l.t x 10*8, " 00 _ ZS> 40 4r'acli clr-car ineisene. of a. os in ii.a coit.a Wi.. .. _~r4 iQr--T Cifftrin^ n..noctolc7i.8. oils a?re.a wall onl'vM-. ' :0eS"r-,: of 0.11 (Scanas!0 2, on. only _ *" elr] 6* *lT*a 4ny cr'===> BOU.V.*, oh. differ..... need aao-nlr 4apec*a 7 "* des--aac affaeo of vinyl oilorid. in tneso *_!::*as. sj!!e:*s " #*# i. e,.1Mnbi7 m *:" 13 ^ ia Cra" Doconono oo loose. w" "y lav* ,r4i"'tut --^y pi.c. i. i.-. ..odel. *.* technicue described centers relative to landfills is prone 000*0 if locating to lead :: population sunstantial ..... I 12 overestimates of exposure if a aaa.ll. error- is made in. tie location- of t population close to a landfill. :n average exposure may tiea be dominated by those estimated far nearby populations. 2.3 Tterrtr gi risk Tbe worst- ease risk estiaats will be for those persons living near to a landfill. As mentioned above, however, the dispersion modelling used for the nationwide average (both here and in the ESA Draft Document) will give very misleading results close in. if landfill gas is collected together into a single vent, that gas will be flared. The worst case exposure estimate will correspond to a landfill with no collection system, in which case the emissions will take place from ever the wnole surface ares of the landfill. The concentrations from such area emissions are considerably lower than those from a vent pipe emitting the seme total quantity cf gas. at equal -distances from the edge of the area or the vent pipe. To make an estimate of the worst ease emissions, consider the landfill labelled H above (Table 1.1) . This was previously used for co-disposal of hazardous waste as well as municipal wests. The total emission race x unit risk far this landfill is 3.1 n2/s. and la contains 12.S million tens of refuse. The worst case will occur with maximum emissions per unit area of landfill, so we will assume waste piled to a height of 100 feet and with an average density of 1 ccn/cu. yd. (double .that assumed in the ZSA, Draft Document) . The emission rate x unit risk per unit area for this landfill is then 8.2 x 10 ^ m/s. and the landfill covers an area of about 3.2 x 10" a2, corresponding to a diameter of about 640 m. Using a dispersion model for area emissions, we find that for a person spending their whole lives 100 m from the edge of sueh a landfill the lifetime risk is about 20 x 10'`, while even for a person staying permanently at the edge of the landfill the lifetime risk is only SO x 10'. ooGcr;. SPI-05682 I THE DOW CHEMICAL COMPANY MIOLANO MICHIGAN *8674 1803 BUILDING July 3,1990 Ms. Barbara Cook Project Manager California Air Resources Board 1102 Q Street Sacramento, CA 95814 Dear Ms. Cook: Attached are our comments on the unit risk derivation presented in the May, 1990 Draft Technical Support Document, Proposed Identification of Vinvl Chloride as a Toxic Air Contaminant. We appreciate your accepting our comments, which have been submitted to promote the best possible science in the performance of health risk assessments. Pleased call on us should you require additional information. Sincerelv, Neil C. Hawkins, Sc.D. Senior Research Risk Analyst Health and Environmental Sciences 1803 Building (517) 636-8237 kal Attachments 0?OC72- SP1-05683 COMMENT: DRAFT TECHNICAL SUPPORT DOCUMENT, MAY, 1990, "PROPOSED IDENTIFICATION OF VINYL CHLORIDE AS A TOXIC AIR CONTAMINANT" Derivation of a unit risk for Vinyl Chloride (CAS 75-01-4) VCM is dearly a rat and human carcinogen, causing liver angiosarcoma in both species and zymbal gland tumors in rats. Thus, for regulatory purposes, there is interest in deriving a quantitative estimate of a level of no significant risk. There are two general approaches to this problem. One approach has been the use of safety factors or uncertainty factors applied to no-observed* effect-levels (NOEL's) in animals to derive a safe level in humans. The other general approach, which has been used more recently by regulatory agencies, has been the use of quantitative risk assessment to estimate levels of risk for any given exposure. The risk assessment process involves a number of decision points for which there is no scientific consensus as to the correct approach. These areas of uncertainty, induding the presence or absence of thresholds, the shape of the dose response model, and animal to man conversion factors, have been resolved within the agencies through the use of policy decisions as to a default methodology. The default methodology is conservative in nature, so as to protect public health. However, the State of California Cancer Risk Assessment Guidelines as well as EPA and OSTP guidelines on the use of risk assessment clearly state that the default methodology should not be used when other data are available. In particular, epidemiology data and pharmacokinetic information should be incorporated into risk assessment when the appropriate data are available. The DHS unit risk for vinyl chloride of 20 x 10(-5) per ppb, as cited in the CARB Draft Technical Document (CARB, 1990), does not utilize the available pharmacokinetic or epidemiological information. Pharmacokinetic Information Pharmacokinetic (PK) information can be used in two ways to augment risk assessments for vinyl chloride. PK data have been used to demonstrate and explain nonlinear behavior at both the high dose and low dose portions of the dose-response curve. The bioassay data of Maltoni (1979) clearly indicated a plateau in the dose-response curve at high doses. This phenomenon can be explained by the use of a Michaelis-Menton function to calculate metabolite concentrations, as suggested by Watanabe et al. (1976), and implemented by Gehring et al. (1978), Crump (1982) and USEPA (1987). However, this methodology only explains the high-dose results in the animal bioassav rather than addressing the problem of low-dose extrapolation. Low-dose risk assessments utilizing PK data have been discussed by Gehring et al. (1979) and Anderson et al. (1980). 00007^ SPl-05684 I .1 . Purchase et al. (1980) reviewed risk assessments for VCM and showed that, among the linear models used, risk estimates varied from the current DHS value, (equivalent to a unit risk of 20 x 10(-5) per ppb), upwards (less risk) at least a factor of 100-fold. The different values derived from animal models vary primarily on the basis of whether or not pharmacokinetic information has been utilized in the assessment. In evaluating the use of pharmacokinetic data, Anderson et al. (1980) conclude that: "Based on the present understanding of the mechanism of carcinogenesis, we believe this to be a more rational approach to the low-dose extrapolation problem.' Gehring et al. (1979) fit a number of extrapolation models to the metabolized dose of VCM, and showed that risk estimates derived without consideration of low-dose metabolite formation potentially overestimate risk by at least an order of magnitude. For example, the one-hit model applied to metabolized dose predicts a risk of 189 per million at 1 ppm for an occupational exposure (Gehring, 1979). By comparison, use of nominal dose (air concentration), predicts upper bound "risks" of 37,000 per million using the unit risk of 20 x 10(-5) per ppb. Other viable dose response models predict much lower risk. Not withstanding the fact that the health criteria represent one aspect of many inputs considered in the standard setting process, we submit it is essential to base any proposed regulation on the most complete information possible. For this reason we believe that risk assessments for vinyl chloride should include PK data, or preferably, the use of actual human data. Risk assessments derived from epidemiology data In the early 1970's, vinyl chloride was reported to cause a rare form of cancer, angiosarcoma of the liver, among workers who had been exposed at extremely high levels for many years in polyvinyl chloride (PVC) polymerization plants. Since this discovery, there have been approximately 50 angiosarcoma of the liver deaths reported throughout the United States and Canada which have been associated with previous vinyl chloride exposure. Eighty percent of these deaths occurred in four PVC plants where exposures to vinyl chloride were known to have been over 500 ppm in the 1950's and 1960's. Today, there.are strict emission limitations under the NESHAP regulation, and the OSHA regulated 8-hour time weighted average for vinyl chloride is 1 ppm. It is particularly noteworthy that there has never been a reported death from angiosarcoma of the liver among Louisiana chemical workers who have worked with vinyl chloride. Vinyl chloride has not been shown to cause cancer at any other anatomical site in humans. Epidemiologic studies conducted in the 1970's suggested that there may be an association with brain and lung cancer, however, recent updates of these studies have reported either no association, or associations only at a much lower statistical level of significance. 000074 SPI-05685 I -3 - A world-recognized expert in epidemiology. Sir Richard Doll, recently reviewed the existing vinyl chloride literature as it pertains to cancer in humans. He concluded that vinyl chloride is a known occupational carcinogen (only for angiosarcoma of the liver) which is due to high occupational exposure levels which have not existed since this association was reported in the early 1970's. According to Doll, the risk for cancer in communities surrounding vinyl chloride production plants from environmental emissions in today's tightly controlled and well-regulated environment "must be negligible." (Doll, 1988) Generally, risk assessments utilize animal data as the basis for quantification of risk. Human epidemiology data often do not have sufficiently precise exposure estimates or sufficiently well-defined populations to be of quantitative value. Human results are dearly preferred, when available, however, and should be induded in any risk assessment review. .In the case of VCM there are at least three assessments of suffident predsion which utilize the human database to estimate risk. In one analysis (Barr, 1982), negative epidemiological studies of people living near VCM production facilities have been used to estimate human potency. Barr estimates that 100 ppb is the approximate lifetime dose corresponding to a human risk of 10(-6). Purchase et al. (1987) note that Barr's estimate is similar to the highest estimates of 10(-6) dose levels derived from animal data and are orders of magnitude higher than the conservative dose estimates which do not take into account low dose PK. This result is consistent with other observations that humans may be less sensitive than animals to the carcinogenic effects of VCM. Gehring et al. (1979) compared the results of an epidemiological study of approximately 10,000 occupationally exposed workers to the values predicted by four different mathematical, models derived from animal data. They conclude that the observed human results are inconsistent with the two linear non-threshold models used, and are consistent with both the probit model and a linear threshold model. The latter two models predict 10*& risk levels at occupational exposure levels in excess of 1 ppm. These analyses by no means prove the validity of the two models and undoubtedly numerous other models would fit and give quite different results for predicting' the ambient level corresponding to a 10*6 risk level. However, these analyses dfi. show that human epidemiology data can be used to derive risk estimates for VCM exposures and that the models indicate that the linear non-threshold models are conservative by a substantial margin. This is to be expected in light of the well known conservativeness of the models. U.S. EPA, for instance, when presenting risks estimates describes them as upper bounds and notes that: "the true value of the risk is unknown and may be as low as zero" (Federal Register, 1986). In an analysis of alternative modeling assumptions for animal to human extrapolation, Elizabeth Anderson, (1984) as head of the U.S. EPA Cancer 0000s 5 SPl-05686 1 -4- Assessment Group found that alternative plausible modeling assumptions would lead to risk estimates that were 15-fold to 10,000-fold lower than the standard LMS procedure. Thus it is essential to use the available human data to place some perspective on the results predicted solely from animal data. In an independent review of VCM, the National Health Council of the Netherlands (1987) derived ambient exposure levels corresponding to risk levels of 10(-6) in humans. Their estimates were derived from both animal data and from epidemiological human data. While noting that the estimates did not differ greatly, they expressed a preference for the human data and reported a value of 1 p.g/cubic meter as corresponding to a risk of 10(-6). This value is approximately 80 times higher than the exposure level derived using the DHS unit risk of 20 x 10(-5) per ppb. The over prediction of the models can be further demonstrated for VCM by comparing predictions of risk utilizing the DHS unit risk with human exposure scenarios. To make this comparison. Table 1 shows the "risk" predicted from the DHS model for a number of occupational exposure situations. The relevance of the specific exposure scenarios are also discussed below. The specific exposure scenarios used in Table 1 were based upon a retrospective study (Barnes, 1976), in which past typical VCM exposures in PVC plants were estimated as: 1000 ppm in 1945-1955, 400-500 ppm in 19551960, 300-400 ppm in 1960-1970, 150 ppm in mid-1973 and considerably lower afterwards. Considering the latency of carcinogenesis in general, and for VCM in particular, tumor incidence rates noted in the 1980's reflect exposures from the 1960's. It can be seen from Table 1 that incidence rates predicted from the linear animal model are completely incompatible with that observed in actual human studies. For example, in the study examined by Gehring (1970) there were only 5 observed cases in 9677 workers. This is approximately three orders of magnitude less than that which would be predicted by the DHS model. Thus, there are a number of assessments based upon human epidemiological data which would indicate that linear models utilizing animal data overpredict risk by at least one to two orders of magnitude. In the interests of assuring that any proposed regulation is supported by as comprehensive a review of the available health data as possible, we submit these assessments should be incorporated into any risk assessments which will be used for regulatory control. This is particularly important in view of the fact that they are based upon human data rather than on laboratory animal results. It can be seen from the above analysis that standard risk assessment methodology and the use of reported literature results lead to orders of magnitude over-estimates of the predicted risk from emissions of VCM from UOOGTf; SPI-05687 t -5- existing facilities. We recommend that these inconsistencies in the risk estimates be resolved if they are to be used as the basis for any proposed regulation. TABLE 1. "RISK" PREDICTED FROM LMS MODEL (using unit risk of 20 x 10*5 per ppb) Occupational Exposure Scenario Upper Bound on "Risk" (Cases Per 10.000) 400 ppm 30 years 9999 400 ppm 20 years 9995 300 ppm 30 years 9998 300 ppm 20 years 9964 200 ppm 30 years 9960 200 ppm 20 years 9770 100 ppm 30 years 9400 100 ppm 10 years 6090 100 ppm 5 years 3750 000G7*: SPI-05688 t 6- - REFERENCES 1. California Air Resources Board (GARB), Draft Technical Support Document, Proposed Identification of Vinvl Chloride as a Toxic Air Contaminant. May, 1990. 2. Watanabe, P.G., McGowan, G.R., and Gehring, P.J. Fate of l14Cl Vinyl Chloride After Single Oral Administration in Rats. Toxicology and Applied Pharmacology 36. 339-352 (1976). 3. Gehring, P.J., Watanabe. P.G. and Park, GN. Resolution of Dose-Response Toxicity Data for Chemicals Requiring Metabolic Acdvadon: Example - Vinyl Chloride. Toxicology and Applied Pharmacology 44, 581-591 (1978). 4. Crump, K.S. Quantitative Assessment of Human Risk From Exposure to Carcinogens with Special Reference to Vinyl Chloride. Contract for Occupational Health and Safety Division, Ontario Ministry of Labour (1982). 5. EPA. Incorporation of Biological Information in Cancer Risk Assessment: Example Vinvi Chloride. Office of Environmental Health and Assessment U-S. Environmental Protection Agency. EPA/600/d-87/188 (1987). 6. Gehring, P.J., Watanabe, P.G., and Park, GN. Risk of Angiosarcoma in Workers Exposed to Vinyl Chloride as Predicted from Studies in Rats Toxicology and Applied Pharmacology. Vol. 49, No. 1 (June 15. 1979). 7. Anderson, M.W., Hoel, D.G., and Kaplan, N.L. A General Scheme for the Incorporation of Pharmacokinetics in Low-Dose Risk Estimation for Chemical Carcinogenesis: ExampleVinyl Chloride. Toxicology and Applied Pharmacology 55, 154-161 (1980). 8. Federal Register, Volume 51, Number 185, Wednesday, September 24,1986. 9. Purchase, I.F.H., Stafford, J., and Paddle, G.M. Vinyl Chloride: An Assessment of the Risk of Occupational Exposure. Fd. Chem. Toxic. Vol. 25. No. 2. pp. 187-202.1987. 10. Barr, J.T. Risk Assessment for Vinyl Chloride in Perspective. Presented at the 75th Annual Meeting of the Air Pollution Control Association, New Orleans, USA, (June 1982). 11. National Health Council of The Netherlands. A Scientific Basis for the Risk Assessment of Vinyl Chloride. Regulatory Toxicology and Pharmacology 7, 120-127 (1987). 12. Anderson, E.L., Ph.D. Risk Analysis in Environmental Health with Emphasis on Carcinogenesis". Harvard School of Public Health, September 18-20,1984. 13. Doll, R. Effects of Exposure to Vinvl Chloride: An Assessment of the Evidence, Scand J Work Environment Health 14 (1988) '61-78. 14. Maltoni, G and G. Lefemme. Annals of N.Y. Academy of Sciences. 246: 195-218 (1979). 15. Bames. A.W. Vinyl Chloride and the Production of PVG Proc. R. Soc. Med. 69: 277 (1976). oooc SPI-05689 II. PART C ADDENDUM AIR RESOURCES BOARD STAFF RESPONSES TO COMMENTS ON PART A SPI-05690 A. COMMENT FROM THE UNITED STATES ENVIRONMENTAL PROTECTION AGENCY 1. Comment: Please include the following paragraph in the Executive Summary: "The Operating Industries, Incorporated (Oil) landfill is currently a federally listed Superfund site. Subsequent to the Air Resources Board's sampling during 1987, the Environmental Protection Agency (EPA) has implemented more stringent landfill gas control measures. EPA has also selected a remedy for landfill gas control that is expected to substantially reduce landfill gas emissions from the Oil landfill. It is fully anticipated that these control measures will substantially lower the levels of vinyl chloride in the ambient air in the vicinity of the Oil landfill." Response: The paragraph (corrected to Indicate that Oil sampling was performed by the South Coast Air Quality Management Oistrict during 1986) appears in the revised Executive Summary. B. COMMENTS FROM WASTE MANAGEMENT OF NORTH AMERICA, INC. 1. Comment: It is erroneous to assume that BKK and Oil landfills are representative of all landfills since both of these sites accepted significant quantities of vinyl chloride waste during operation. Response: Landfill records of whether or not vinyl chloride waste was accepted may not be a reliable means of predicting the potential for vinyl chloride emissions. For decades, vinyl chloride waste (as well as other halogenated industrial waste which can form vinyl chloride) was disposed in some Class II as well as Class I landfills. In addition. Class III landfills accept disposed consumer products containing chlorinated compounds which can form vinyl chloride. Also, incomplete recording of vinyl chloride waste disposal and illegal vinyl chloride waste dumping have occurred to an unknown extent. However, a statement has been added to the report indicating that the vinyl chloride emissions measured at BKK and Oil may not be typical of all landfi 11s. 2. Comment: Significant information exists that landfills are not significant sources of vinyl chloride emissions. For example, data from the Air Solid Waste Assessment Testing Program (Landfill Gas Testing Program) mandated by Section 41805.5 of the California Health and Safety Code show that six waste management units operated by Waste Management of North America (WMNA), Inc. are not significant sources of vinyl chloride emissions. Response: After considering the data available on potential sources of vinyl chloride emissions, the staff of the Air 00007D SPI-05691 $ Resources Board (ARB) concluded that landfills are a potential major source. Modeled estimates of vinyl chloride emissions for just BKK and Oil landfills were far greater than emissions estimates for publicly-owned treatment works (POTWs) and polyvinyl chloride (PVC) fabrication and production facilities: Source Emissions (tons/vearl BKK Landfill 44-197 Oil Landfill 4-51 POTWs 1.7 PVC fabricationi 0.75 PVC production <0.5 Inventory Year 1987 1986 1985 1982 1988 Although BKK and Oil landfills may not be typical, one cannot rule out the possibility of elevated vinyl chloride emissions from other California landfills using preliminary Landfill Gas Testing Program data. The preliminary data show that vinyl chloride was detected at or above the detection limit in the internal landfill gas at 160 out of 340 landfills tested. Also, vinyl chloride was detected at or above the detection limit in the ambient air near 24 out of 251 landfills tested. However, because landfills vary in the amount and composition of wastes accepted as well as disposal methods used, estimating total statewide vinyl chloride emissions from landfills is not possible at this time. Therefore, the staff report has been revised to indicate that landfills are a potential major source-category. 3. Comment: Section 39660 (f) of the California Health and Safety Code mandates that the Department of Health Services (DHS) and the ARB give priority to the evaluation of a substance's amount or potential amount of emissions and ambient concentrations in the community. To proceed with identification of vinyl chloride as a toxic air contaminant (TAC) while identifying landfills as the largest source of emissions based on two unrepresentative sites (BKK and Oil landfills) would be a disservice to the waste management industry and contrary to the law. Response: In the revised vinyl chloride report, based on available data, the staff of the ARB conclude that landfills are a potential major identified source-category of vinyl chloride emissions. The staff further conclude that sufficient overall data are available to proceed with the identification of vinyl chloride as a TAC as provided in the statutes. Furthermore, vinyl chloride, as a federally designated hazardous air pollutant, must be identified as a TAC pursuant to Health and Safety Code Section 39655. Also, please see the responses to comments 1 and 2. QOOOSO SP/-05692 I 4. Comment: Waste Management of North America. Inc. recommends that identification of vinyl chloride be delayed until Landfill Gas Testing Program data are included in the report. Response: Preliminary Landfill Gas Testing Program data have been included at appropriate places in the revised vinyl chloride report. In addition, a table of Landfill Gas Testing Program data has been provided in Appendix VI. OGOQS1 SPJ-05693 I III. PART C ADDENDUM DEPARTMENT OF HEALTH SERVICES RESPONSES TO COMMENTS ON PART B SPI-05694 RESPONSE TO COMMENTS: WASTE MANAGEMENT OF NORTH AMERICA, INC. June 11, 1990 PUBLIC HEALTH RISK OF VINYL CHLORIDE COMMENT: "While we do believe that it is ultimately appropriate to regulate vinyl chloride as a toxic air contaminant, we are concerned that the unit risk factor that you have attributed to this compound is overly conservative." RESPONSE: The DHS document of May, 1990, provides estimates of unit risk that use data, assumptions and methods that are highly defensible, based *on standard procedures utilized by DHS and EPA. The analysis uses animal and human data. Dose rates to tissue have been obtained from a pharmacokinetic model. The range of unit risks does not include some of mouse data which is up to 2.5 times above the top of the range in the risk assessment, as indicated in the text at page 8-8. The best estimate of unit risk for regulatory purposes is the top of the tightly clustered less than 10-fold -- range, containing numerous results, including human results. The two top points include liver angiosarcoma in the female rat, this tumor being one of the most distinctively linked to vinyl chloride exposure in both rats and humans. DHS staff conclude that this choice is not overly conservative. COMMENT: "I have also attached to this letter a copy of a brief paper on Carcinogenic Risks from Landfill Emissions dated June 6, 1988. . . . This information provides a much more realistic assessment of the health risks posed by municipal landfills not only from the standpoint of vinyl chloride but a number of OOOGS^ SPI-o0C5C6Q95a ocher compounds as well. In summary this brief paper, based on an assessment of the cumulative impact of all landfill emissions, concludes, 'Using a dispersion model for area emissions, we find chat for persons spending cheir whole lives 100 m from the edge of such a landfill the lifetime risk is about 20 x 10^, while even for persons staying permanently at the edge of the landfill the lifetime risk is only 50 x 10*6.'" RESPONSE: The cited paper, which was an addendum to comments to EPA and not a journal article, was produced at the request of Waste Management Inc. and of The National Air Pollution Control Techniques Advisory Committee to the EPA. This response will focus on the unit risk estimate utilized in the document and not on site-specific factors such as emission rates, source areaa and meteorology. The cited paper uses for vinyl chloride a unit risk of 22 x 10*^ ppb'^, expressed as 9.8 x 10*^ (jtg/m^)** in their Table 1.2. This risk estimate is essentially equivalent to the DHS best estimate of 20 x 10*^ ppbfor unit risk. Therefore, DHS is recommending essentially the same unit risk as is the basis for the calculation of risk for landfills that is advocated by the commenter. The cited paper refers to EPA (1985b) as the source for that unit risk, but DHS staff, after obtaining and reviewing that document, calculate that the unit risk corresponding to the EPA's (1985a) potency of 2.95 x 10'1 (mg/kg-day)`1 is 11 x 10 ^ ppb using EPA's own assumption of 50% absorption, which the cited paper's authors evidently did not use. The reason for that EPA unit risk being 55% of the DHS best estimate stems from the EPA analysis combining male and female rats, giving a lower risk than the DHS use of females, the sex with the higher risk in this case. For comparative purposes the EPA 1985b unit risk has now been included in the document at page 8-13 and in Figure 8-1. OCOCf 2 SP/-05696 SPECIFIC DOLAN COMMENTS COMMENT: "First, why bother using the linearized multistage model under the pretense that it is a true mechanistic model (which it is not), when a simple linear regression usually yields nearly identical estimates of q^ie (r2 - 0.98) (Personnel conversation with Curtis Travis, Oak Ridge National Laboratory)?" RESPONSE: The linearized multistage model affords an efficient unified approach to determining carcinogenic potencies of most substances for which there are enough data to determine the potency. Once the appropriate computer software and a knowledge of its use have been acquired, the model is quite convenient to use. This model accommodates pharmacokinetic conversion of dose rate and can account for a vide range of test results in a way that allows extrapolation that is frequently in accord with available knowledge of mechanisms. In contrast, the simple linear regression becomes inappropriate at high doses in many animal experiments. Even in the case of vinyl chloride, the multistage model indicates an improved fit to the data by including a quadratic term. Thus, a simple linear model would introduce some bias into the low dose extrapolation. COMMENT: "Second, the discussion of uncertainty in the quantitative risk estimates is given short shrift. Although the uncertainties or absence of exposure data in the occupational cohort studies is mentioned, there is no discussion of the conservatism built into the risk estimates by the selection of data for extrapolation, and the extrapolation assumptions, and the effects their underlying assumptions may have on the risk estimates. For instance, the use of the most sensitive sex/strain/species instead of the average may alter risk estimates by 'several orders of magnitude.' (U.S. EPA, 1985). 00008" Similarly, the SPl-05697 issues of che extrapolation of rodent potency estimates to humans, particularly on the basis of surface area, and the use of upper 95th percentile estimates of carcinogenic potency instead of the MLE, may alter potency estimates by an order of magnitude, or more. (U.S. EPA, 1985)." RESPONSE: DHS staff do not agree that there is unjustified conservatism built into Che risk estimates. DHS used procedures chat are standard for EPA and DHS, taking account of the pharmacokinetics of vinyl chloride. The DHS risk assessment did not use 1 the most sensitive species and strain. Some of the mouse data resulted in unit risks up to 2.5 times higher chanactually used in Che risk assessment. The assessment does use data for the more sensitive sex in order to protect women as well as men because female rats had a 3-fold higher risk than male rats. The assessment does use the 95th percentile estimates, clearly identifying them by UCL throughout the document. The analysis uses these estimates in the risk assessment only when the ratio of UCL to KLE is less than 3, specifically avoiding-the possibility of that ratio being "an order of magnitude, or more", as found in other circumstances by EPA (1984a). In order to expand che discussion of uncertainties, DHS staff have added a brief paragraph to the document at page 8-13 as follows: "All these estimates are subject to substantial uncertainties as have been discussed in the scientific literature (DHS, 1986, and EPA, 1984a). The available information does not suggest that there is a threshold for vinyl chloride's carcinogenic effect, though this remains uncertain. The multistage model is the best choice based on the plausible mechanism of vinyl chloride carcinogenicity. Nevertheless, our incomplete understanding of cancer makes this choice subject to uncertainty. Furthermore, the present approach uses other assumptions that are 00006" 4 SPI-05698 designed to be somewhat health protective in the absence of precise knowledge. One of the most important of these is the extrapolation from humans to animals on the basis of surface area in accordance with DHS guidelines (1985). This approach may overpredict or underpredict human risk. COMMENT: "Third, it is perplexing that the Krewski et al. (1987) chapter is referenced, yet the 36-fold lower carcinogenic potency factor they derive is omitted from the brief discussion. Some discussion on the merits and limitations of the Krewski et al. analysis is necessary." RESPONSE: The Krewski et al. (1987) result of 0.0058 ppm'^ is based on virtually all the relevant female liver angiosarcoma data of Maltoni at al. (1984) and is unadjusted for lifetime exposure. When adjusted for lifetime exposure the unit risk is 9.7 x 10'^ ppb'*. On this basis, the unit risk of Krewski et al, rather than being 36-fold lower, is actually 45% higher than the result from the analysis (BT-9, 15) which corresponds most closely in the document, 6.7 x 10-5 ppb'^. This is among the highest rodent risks in the assessment. Because of the rather good agreement despite the differing analyses, adding a discussion of the merits and limitations is inappropriate for this document. COMMENT: "Fourth, the ARB cites the concordance of the potency estimate derived from the Drew et al. (1983) study and the Maltoni et al. (1984) experiments. It is unclear whether the Maltoni experiments were conducted in his medieval castle/laboratory where the mycobacterium infection is endemic, or in some other facility. (Personnel conversation with E.E. McConnell, National Toxicology Program). In the U.S., mycobacterium infections in test animals would likely violate GLPs, and serve as grounds for invalidating a study." OGGCr SPI-05699 RESPONSE: Both EPA and IARC have relied on the Maltoni et al. data for their assessments. DHS knows of no documentation that casts serious doubt on the validity of these data. The concordance of the Maltoni data is not only with Drew et al. but also with Bi et al. COMMENT: "Fifth, Che recommended use of a potency factor derived from animal instead of Che human occupational study of Vaxveiler et al. (1976) is not robust, given that Che human data already represents an upper-bound estimate in the target species of concern (i.e., humans). The additional rationale that the selection of the highest animal estimate is justified by Che limited evidence of an effect by age at first exposure (Drew et al., 1983) suggests that perhaps the ARB should consider using a true mechanistic model, perhaps one based upon the MVK model paradigm,, as the basis of its potency determinations." RESPONSE: The human data in itself does not represent an upper bound in humans because (1) that data does not include a lifetime exposure, and there is evidence of greater sensitivity of the young and (2) the human data that is sufficient for the risk assessment includes almost no females. As to the remainder of the comment, the DHS staff note that the MVK model does have the potential to be more closely linked to the biological observations of cell proliferation than the multistage model. When the necessary data are available and the mathematical analysis is adequately established, then an analysis related to the MVK model is worth consideration. 0000^ 6 SPI-05700 RESPONSE TO COMMENTS: THE DOW CHEMICAL COMPANY July 3. 1990 Derivation of a Unit Risk for Vinyl Chloride COMMENT: "VCM is clearly a rat and human carcinogen In order to address cancer concerns, regulatory agencies have recently used "quantitative risk assessment to estimate levels of risk for any given exposure." In order to resolve uncertainties in this process, agencies have developed default assumptions which are "conservative in nature so as to protect public health." California and federal guidelines "clearly state that the default methodology should not be used when other data are available. ... The DHS unit risk for vinyl chloride of 20 x 10*^ per ppb ... does not utilize the available pharmacokinetic or epidemiological information." RESPONSE:. The DHS staff disagree with the assertion that the current DHS risk coefficient does not utilize the available phamacokinetic or epidemiological information. DHS staff, in response to comments on the first draft document, did specifically incorporate the available pharmacokinetic and epidemiological information into the analysis that produced the estimates of unit risk. All the DHS calculations of unit risk in the document under review directly use the available pharmacokinetic information. In addition, the risk assessment specifically shows how the best estimate of upper confidence limit (UCL) for unit risk, 20 x 10*^ per ppb, cited in che comment above, is consistent with human occupational data when adjusted from males in that workforce to females who would be exposed in the general population. Pharmacokinetic Information COMMENT: "Pharmacokinetic (PK) information can be used in two ways to augment risk assessments for vinyl chloride. PK data have been used to demonstrate and explain nonlinear behavior at both the high dose and low dose portions of the dose'responses curve." The Michaelis-Menton "methodology explains only the high-dose results. Low-dose risk assessments utilizing PK data have been discussed by Gehring et al. (1979) and Anderson et al. (1980)." RESPONSE: Contrary to the implication of the comment, Gehring et al. (1979) and Anderson et al (1980) used the Michaelis-Menton methodology to incorporate PK data at all doses, high and low, in essentially the same way as che DHS document. The low dose extrapolations in those two studies differed from the DHS document in that they explored extrapolation not only by the single stage model, as did the DHS document, but also by the log-probit model. DHS staff consider che log-probit model to be inappropriate based on the data that became available after these articles were published and the apparent mechanism of carcinogenesis. COMMENT: "Purchase et al. (1980) reviewed risk assessment for VCM and showed that, among the linear models used, risk estimates varied from the current DHS value, (equivalent to a unit risk of 20 x 10'^ per ppb), upwards (less risk) at least a factor of 100-fold. The different values derived from animal models vary primarily on the basis of whether or not pharmacokinetic information has been utilized in the assessment. In evaluating the use of OOGO- SPI-05701 i pharmacokinetic data, Anderson et al. (1980) conclude that: "Based on the present understanding of the mechanism of carcinogenesis, we believe this to be a more rational approach to the low-dose extrapolation problem." ` RESPONSE: Contrary to the comment, Purchase et al. (1987) do not make clear comparisons among linear models, nor do they make clear comparisons between chose models that use pharmacokinetic adjustment and those that do not. The work of Anderson et al. (1980), does clearly account for the role of the pharmacokinetic adjustment and the role of the carcinogenesis model. Anderson et al. (1980) found that for the multistage model, which extrapolates linearly to zero exposure, "incorporating the pharmacokinetics has only a moderate effect on the low-dose estimates." Table 1 in that work shows that the effect of incorporating Gehrlng's simple pharmacokinetics into the multistage model is to increase - - by either or 4- or 26-fold, depending on the specific choice of rat data - - the extrapolations of risk estimates to low dose. The more appropriate choice of rat data, eliminating the two highest and therefore most saturated exposures, corresponds to the 4-fold increase. DHS staff agree with the Anderson et al. statement about the rationality of the pharmacokinetic approach, provided appropriate data are available. COMMENT: "Gehring et al. (1979) fit a number of extrapolation models to the metabolized dose of VCM, and showed that risk estimates derived without consideration of low-dose metabolite formation potentially overestimate risk by at least an order of magnitude. For example, the one-hit model applied to metabolized dose predicts a risk of 189 per million at 1 ppm for an occupational exposure (Gehring, 1979). By comparison, use of nominal dose (air concentration), predicts upper bound "risks" of 37,000 per million using Che unit risk of 20 x 10*^ per ppb. Other viable dose response models predict much lower risk." RESPONSE: Gehring et al. in their 1979 article did not derive any estimates of risk without using their model for low-dose metabolite formation. The four models in chat article use the metabolized dose at all dose levels. In that article the authors did characterize the results of their models C (linear forced through the origin) and D (one-hit) as overpredicting the number of liver angiosarcomas reported in the Equitable Environmental Health (1978) study. However, a follow up study (Wong et al., 1986) of those worker populations showed a marked increase in incidence of deaths attributable to liver and biliary cancer. Neither of these papers on workers has appeared in the peer-reviewed literature, making acceptance of either of their epidemiological results problematic. All the DHS estimates of risk use a metabolized exposure chat is essentially equivalent to the metabolized dose of Gehring et al. (1978, 1979) at low exposures. For the 1 ppm occupational example in Gehring et al (1979), their assumptions of 40 hr/wk for 35 years does not give a risk of 37,000 per million for the DHS unit risk of 20 x 10*^ per ppb but gives 22,000 per million. This risk is 126-fold greater than the Gehring et al. result of 189 per million, not because of differences in analysis of the data but because of three different choices in applying the results of the rat analysis to the human. In the one-hit analysis for the rat the result of Gehring et al. (1979) is a unit risk of 1.1 x 10*^ per ppb, when adjusted to lifetime exposure. This result is actually somewhat greater than that of the nearest analysis in the DHS document, for BT-1,2, giving an MLE of ql - 0.8 x 10'^ per OOGCrV' 2 SPI-05702 ppb. The sources of the higher risk estimate for DHS ere (1) the use of later Maltoni et al.(1984) data, BT-9,15 for the best estimate rather than the earlier Maltoni and Lefemine (1975) data used by Gehring et al.(1978), resulting in a 3.2-fold increase, (2) the use of the 95% upper confidence level on risk rather than-the mean regression estimate (similar to maximum likelihood estimate), resulting in a 2.3-fold increase , and (3) the use of DHS standard scaling of humans to animals by body weight to the two-thirds power rather than the Gehring et al. ad hoc scaling that has never been accepted, resulting in a 17.7-fold Increase. The result of multiplying all these increases together is an overall 130-fold Increase. COMMENT: "Not withstanding the fact that the health criteria represent one aspect of many inputs considered in the standard setting process, we submit it is essential to base any proposed regulation on the most complete information possible. For this reason we believe that risk assessments for vinyl chloride should include PK data, or preferably, the use of actual human data." RESPONSE: DHS staff agree and have used both in the risk assessment. Risk Assessments Derived from Epidemiology Data COMMENT: After Introductory remarks concerning vinyl chloride in the workplace, the eommenter asserts, "It is particularly noteworthy that there has never been reported death from angiosarcoma of the liver among Louisiana chemical workers who have worked with vinyl chloride." RESPONSE: It is difficult to respond to the comment about Louisiana chemical workers without specific reference to surveillance programs and exposure estimates. COMMENT: "Vinyl chloride has not been shown to cause cancer at any ocher anatomical site in humans. Epidemiologic studies conducted in the 1970's suggested that there may be an association with brain and lung cancer, however, recent updates of these studies have reported either no association, or associations only at a much lower statistical level of significance." A world-recognized expert in epidemiology, Sir Richard Doll, recently reviewed the existing vinyl chloride literature as it pertains to cancer in humans. He concluded that vinyl chloride is a known occupational carcinogen (only for angiosarcoma of the liver) which is due to high occupational exposure levels which have not existed since this association was reported in the early 1970's. RESPONSE: In his review article Doll (1988), cited in the next comment, discusses this issue at length. He concludes in a manner contrary to chat of the comment. "It is, however, still difficult to decide whether vinyl chloride produces small risks of cancer, compared to those due to nonoccupational causes, at sites other than the liver, and, if so, whether, in total, these risks might cause almost as many deathes as angiosarcoma of the liver." COMMENT: "According to Doll, the risk for cancer in communities surrounding vinyl chloride production plants from environmental emissions in today's tightly controlled and well-regulated environment "must be negligible." (Doll, 1988) ." OPOOCn SPI-05703 3 RESPONSE: Doll (1988) did not estimate unit risks or any other numerical measure of the relationship between exposure and response. He did not specify numerically what he or his reference considers to be a negligible risk outside a vinyl chloride plant. Also, he does not consider other environmental exposures such as landfill sices. So his remark is difficult to apply to the present assessment. COMMENT: "Generally, risk assessments utilize animal data as the basis for qualification of risk. Human epidemiology data often do not have sufficiently precise exposure estimates or sufficiently well-defined populations to be of quantitative value. Human results are clearly preferred, when available, however, and should be included in any risk assessment review. In the case of VCM there are at least three assessments of sufficient precision which utilize the human database to estimate risk. In one analysis (Barr, 1982), negative epidemiological studies of people living near VCM production facilities have been used to estimate human potency." , RESPONSE: The previous response to comments from the Vinyl Institute pointed out that che Barr (1982) analysis is too unsubstantial epidemiologically to be considered in this risk assessment. COMMENT: "Barr estimates chat 100 ppb is the approximate lifetime dose corresponding to a human risk of 10' . Purchase et al. (1987) note that Barr's estimate is similar to the highest estimates of 10* dose levels derived from animal data and are orders of magnitude higher chan che conservative dose estimates which do not cake into account low dose PK. This result is consistent with ocher observations chat humans may be less sensitive than animals to the carcinogenic effects of VCM." RESPONSE: Contrary to the comment, Barr, in his Table 1 and consistent with his text, found that the lifetime exposure for 10* risk was greater than 1 ppm, not 100 ppb, which was a mischaracterizacion appearing in che Table 4 of Purchase et al. (1987). The present DHS document gives 0.S ppb as the lower confidence limit on lifetime exposure for 10* risk. Barr offers no rationale for using the weak data he selected from a 1975 EPA report in order to calculate his epidemiological estimate. These data do not appear to be appropriate for that purpose, and such inappropriate use of data would account for disagreement with the DHS value by orders of magnitude. Also contrary to the comment, Purchase et al. (1987) do not specifically comment on Barr's estimate in their text. As stated in che response above, Purchase et al. do not present clear comparisons of effects of pharmacokinetics or of the results of using different basic forms of models for carcinogenesis. Finally, the commenter has offered no supported observations to show "that humans may be less sensitive than animals to the carcinogenic effects of VCM." COMMENT: "Gehring et al. (1979) compared the results of an epidemiological study of approximately 10,000 occupationally exposed workers to che values predicted by four different mathematical models derived from animal data. They conclude that the observed human results are inconsistent with che two linear non-threshold models used, and are consistent with both the probit model and a linear threshold model. The latter two models predict 10* risk levels at occupational exposure levels in excess of 1 ppm." OQOCP' u SPI-05704 RESPONSE: As pointed out in the response above and in the previous response to comments of the Vinyl Institute, the comparisons that Gehring et al. (1979) made are now out of date because of a follow up study of Wong et al. (1966), which found much higher rates of liver cancer incidence in vinyl chloride workers than in the data used by Gehring et al. COMMENT: "These analyses by no means prove the validity of the two models and undoubtedly numerous other models would fit and give quite different results for predicting the ambient level corresponding to a 10' risk level. However, these analyses & show that human epidemiology data can be used to derive risk estimates for VGM exposures and that the models indicate that the linear nonthreshold models are conservative by a substantial margin. This is to be expected in light of the well known conservativeness of the models. U.S. EPA, for instance, when presenting risks estimates describes them as upper bounds and notes that: "the true value of the risk is unknown and may be as low as zero" (Federal Register, 1986)." RESPONSE: The DHS document does use human epidemiology data in the risk assessment. The DHS staff does not agree that the linear nonthreshold models extrapolate conservatively by a substantial margin, relative to actual incidence of cancer. Certainly, such models extrapolate conservatively compared to the log-probit model (Gehring et al. 1979: Model A), but that model is not in accord with present understanding of mechanisms of carcinogenesis applicable to vinyl chloride, whereas the linearized multistage model is in accord with such tinderstanding and therefore most likely to extrapolate to low exposures accurately rather than being overly conservative. The feature of the unit risks that is health protective and might be characterized as in the conservative direction is the use of the 95% upper confidence limit (UCL) in order to provide adequate protection in the great bulk of cases. Any model with sufficient data can incorporate this feature. EPA does call such estimates "upper bounds," a term that is now commonly used for UCL although that usage is not in accord with the strict mathematical definition. The true risk is very unlikely to be exactly zero; so the quote from EPA, though possible as a point of logic, does not appear to enhance the readers perspective, particularly in cases of the maximally exposed individual. COMMENT: "In an analysis of alternative modeling assumptions for animal to human extrapolation, Elizabeth Anderson, (1984) as head of the U.S. EPA Cancer Assessment Group found that alternative plausible modeling assumptions would lead to risk estimates that were 15-fold to 10,000-fold lower than the standard LMS procedure. Thus it is essential to use the available human data to place some perspective on the results predicted solely from animal data." RESPONSE: The main issue in this comment is the question of what is considered plausible. If an extreme curve-fitting model such as the logprobit is compared against the more mechanism oriented linearized multistage model, then many-fold lower risks will be obtained for the log-probit. COMMENT: "In an independent review of VCM, the National Health Council of the Netherlands C1987) derived ambient exposure levels corresponding to risk levels of 10 in humans. Their estimates were derived from both animal data and from epidemiological human data. While noting that the estimates did not 000C9Z SPI-05705 differ greatly, they expressed a preference for the human dapa and reported a value of 1 Mg/cubic meter as corresponding to a risk of 10* . This value is approximately 80 times higher than the exposure level derived using the DHS unit risk of 20 x 10*5 per ppb." RESPONSE: fiS pointed out in the DHS document at page 8-5, the Netherlands council obtained 1.2 x 10*" per ppb for the unit risk of mortality due to liver cancer and 2.5 x 10* per ppb for all cancer. "Both these results were based on estimated atmospheric exposure. When those results are modified to cake account the pharmacokinetics and to provide 95% upper confidence limits, the results are close to the present results." The DHS adjustment of the council's worker exposure of 500 ppm is 115 ppm, requiring a 4.3-fold adjustment upwards of their unit risks to account for pharmacokinetics. The council's average unit risk needs to be multiplied by about 2.3 to estimate the corresponding UCL value. The unit risks resulting from both multiplications are 10-fold greater, or 1.2 x 10*^ and 2.5 x 10* per ppb. These values are about half the corresponding epidemiology estimates in the document, based on the data of Uaxweiler et al. (1976), which was one of the studies used by the Netherlands council for data on mortality due to cancer in vinyl chloride workers. COMMENT: The over prediction of the models can be further demonstrated for VCM by comparing predictions of risk utilizing the DHS unit risk with human exposure scenarios. To make this comparison, Table 1 shows the "risk" predicted from the DHS model for a number of occupational, exposure situations. "It can be seen from Table 1 that incidence rates predicted from the linear animal model are completely incompatible with that observed in actual human studies. For example, in the study examined by Gehring (1970) there were only 5 observed cases in 9677 workers. This is approximately three orders of magnitude less than that which would be predicted by the DHS model. Thus, there are a number of assessments based upon human epidemiological data which would indicate that linear models utilizing animal data overpredict risk by at least one to two orders of magnitude. In the interests of assuring that any proposed regulation is supported by as comprehensive a review of the available health data as possible, we submit these assessments should be incorporated into any risk assessments which will be used for regulatory control. This is particularly important in view of the fact that they are based upon human data rather than on laboratory animal results. It can be seen from the above analysis that standard risk assessment methodology and the use of reported literature results lead to orders of magnitude over-estimates of the predicted risk from emissions of VCM from existing facilities. We recommend that these inconsistencies in the risk estiamtes be resolved if they are to be used as the basis for any proposed regulation." RESPONSE: The incompatibility of predicted and observed incidence rates, as derived in the comments, arises because of the commenter's errors in making the predictions and the citation of incidence data that are not current and that do not permit adequate estimates of exposure. Thus, the commenter has made no sustainable case for overprediction or inconsistency of risk estimates in the document. 0000?'' 6 SPI-05706