Document RvwMEXnVQ1NRGmB9rxD8R67v

I' ~ aJ. i,n h s f r . Med., 1965, 22, 13. h, 9 en! reli A STUDY OF THE CONCENTRATIONS OF 949 Ion' POLYCYCLIC AROMATIC HYDROCARBONS IN 3ad `me GAS WORKS RETORT HOUSES urn ich BY ind mb P. J. LAWTHER. B 7 COMMINS, and R E WALLER tnd md From the .Wedim1 ResemcA Councrl Air PdluflonResearch Unu, SI. Bartidunnu. .r F..#pdral Medical C d w , L&R, E.C.1 sal :pc (RECEIVED =on cwai imnm SEPTEHRER 30, 1%) Measuremmts of ?heconcentration of 3:4-benzpyrcne and other polycyclic arornatlc hydro- carbons have been made io gas works retoTt houses ofseveral types. The tarry fumes whichescaped from retnm contained extremely high concentrations of polycyclic hydrocarbons, but in general men w m only exposed to these very briefly, The mean concenrration of 3 :4benzpyrene determimd horn long-period samples at sites representative of normal working conditions in works was 3 WgJrn.', over 1 0 times the normal level in the Wty of London. Above the retorts in an old horizontal retort h o w the concentrationwas over 200 ,ug.:in.a, about f0,000 times that in the City, and the `top-man' working there could & exposed to this in tbe nomal MWSC of his duty. We found no working arcas in the vertical retort houses where men could be exposed to such massive ~ n t r a t ~ of npo~lycycljc h y d m r b o o s . Apart from defining these special conditions abovc horizontal retorts w r resul~sdid not reveal any gross differences in poIlution or the general air in horizontal and continuous vertical retort houses. The tarry material produced when coal is distilled lished beyond doubt, it was more difficult lo deter- during the manufacture of coal ga5 contains a large mine whether t h a t was any risk of carcinogmcsis proportion of polycyclic aromatic hydrocarbons associated with :he inhalation of Iarry fumes in gas some of which are carcinogenjc Although prccau- works. In an aocmpaoying paper DOLI, Asha, tions are taken to avoid the discharge of tarry fume Gammon, G u m , Hughes,Tyrer, and Wilson (1%5) inlo the working environment, some does escape, have now shown from an eight-year study of the and an investigation was carried out to find whether mortality of gasworkers that here is an increased men were exposed during their hours of cmpioyment mortality from lung cmcer among men cmpioyed m to conccnimtions of polycycIic aromatic hydrocarbons greatly in scess of those measured in the gcuaal atmosphcrc. The carcinogenic properties of coal tar have been retort houses. T h e work reponed here was done.to determine whether these men were exposed toconcentrations of palrcyclic aromatic hydrocarbons weate in excess o l those found in the general studied for many years, and indad much of our atmosphere. There was a suggestion from the present-day knowledge of chemical carcinogens is mortality study that the risk was greater among men based OR the classjml work carried out more than 30 years ago at ;he Royal Cancer Hospital using an extract of tar prepared by the Gas Light md Coke Company. The way in which this led to the identifica- who had workcd with horizontal retorts than among those who had worked with continuous vertical retorts. When we started our investigation in 1959, horizontal retons were in use mainly at small works, tion o the polycyclic aromatic hydrocarbon 1/3:4-bnnpyrcnc as one of the most potent cadno- 1genic components of the tar has bem rccalied by Kennaway (1955). Although the connexion bdweto exposure to tar !and the development of skin cancer had k e n estab- and thae wem gradually k i n g c l d down so that production could be concentrated at k g e works having continuousvertical retortsoc coke ovens. W e first studied conditions in a hcuizontd retort house at Rocitester, Kent More the opp&ty was lost ;o us. We thm went to a modern vertical retort 13 house ne one at 1 extended The F ;:-dUSU!il 1::e Rivi 90 hori CaPWilY ceased a' house h where IT good wi doorwa) geatly 1 was qua c!ouds 11 fzr five men wer fims no area arc escaped employe were no: retorts ( tion fa retorts t material !::ses wt retorts 1 directly partimi; unblock (Fig. 3). belt (ab{ expod The F on the : dentiai i ws a r I3j3. I 9 total g The bui large w These n fortable Weather =era1 those in and the men wa. for the Rtor? h FXrn5UT house near Brighton and ID an OM,but re-equipped one at Birmingham. The investigation was later extmded to works in a fcw other areas. Descriiion of Works Studid The Rochester works was situated in an open industria1 area on the outskirrs af the town close to the River Medway. The retort house contained 90 h m h m t a l nforts with a total gasmaking capacity of 24 million CU. ft./day. Gas production at the Rochester works in 1960, and the retort house ha.? s i n e bttn pulled down. At the level where most of the men worked the ventilation was good with strong draughts blowing thmugh open doorways and unglaaed windows. Visibility was geatiy reduced when the hot coke from the retort was quenched in a stream of water. giving rise to clouds of steam (Fig. I). These conditions prevaiiad for five to f0 minutes in every two hours, and the mm were brietly enveloped in the steam. At other times not much pollution was visible in the working area around the retorts. Any tarry fumes which escaped into the air rose vertically, and the men cmployai in charging and discbarginy the mom were not heavily e x m e d to rhc fumes. Above Ihc retorts (Fig 2) condiuons were far worse; in addi- tion IO the fumes rising up rrom thc face of tbe 1-5 there werc also substantial amounts d tarry material emjttcd intmittcntly from 'asccnoion' pipes which carry the distillation products from the retons to the gas main. Only o m man worked dimtly above the :ctorts, tbe 'topman', and he was particularly cxposai to tarry fumes when he was unblocking open ascmsion pipes with an auger {Cis.3). The man who operated the coal conveyor btli (abovc and to OM ude of the retorts) w a s also exposed to the tarry fumcs to ;Ilimited extent. The Panslade works, which we next visited, wds on the sea-shore just to the west of the l a w midential area of Brighton and Hove. The retort house was a modern one which came into operation :n 1953. It contained continuous v e n i a l reforts with a total gasmakinycapxity of 15 million cu. k i d a y . Tbe building was totaIly enclosed, and there were large windows at *klevels where men worked. These were normally left open to maintain a COIIIfortable tempentun.? in the building, bur in stormy weather they sometimes had to be closed. The general working conditions were much cieaber thar. those in the old horizontal Etait housc at Rochester, and the mount of physical work required of the men was less. There was also much k%.sopwrtuniry far the scape of stearn. dust or tarry fumes into the retort housc The only operation involving regular exposure to t m y fumes was the 'rodding' which was I I6 RRlTISH J0L'RNA.L OF I N D U S T R I A L ,MEDICINE totalgas-making capacity of30miilioacu.ft.,'day.T k buildiap,was *wellventilated by means of optningsin the walls and roof, and extractor fans ;had also been installed. Although the returts were of the s a m e &sign BS those at Portslridc, there were soox minor differenas in the mode ofo ~ t i o o .The men there charged the retorts at hourly intervals and roddcd them only when required; interruptions in the flow of coal were indicated cemotety on a p a m L The= ing ?t in. Y 10 in. (20-32 A -75-4 cm.; overail, using a %@ZX high-volurreesunplerrunningat about 1.5rn?fmin. At R a h e s l e r we tried to assess the very highconcenwatiom nf ?oliurion nhoue rhe rernrn ismg a mask technique. One of us wore an dtic~ontindustrkk mask' for A pariod ofurn hour do= to wiare one man was working. W EC Q U I ~only &hate tho volume of air inspired Juring thar pmud, but anntha sample was taken simultaneously with a i n m u n d d u flw usmg the $ortabit sampler ?ham in Figure 5 . in this the tar was cotlectsd on a filter was again no w d u g area comparable with that tnade of tne same marerial 3s that in the mask. backed abow the retorts at RoEhester, and there did not with a glass fibre tilta. This enabled us to detcrminc the 3.p- to be anywhere in the works where men would be exposed to massive quantities of ? a r e fumes. We also visited the nearby Saitlcy works in Birmingham, where some of the men bad ken amsromed 10 Wearing masks in !he COIJTSC sf tbdr dut-iitsin the retort house. This gave us rhe 0pp01-t~- nity to explore the use of masks as a sampling tech- nique, but we did not install fixed samplers there. The works was smaller than the one at Windsor St., hut it had the same type of continuous verticd mtorts. Later in our investigation we also arranged for mask sampiiog to be done at the Stapleton Rd. works, Bristol, where there were continuous v e n i a l retorts, and at New WortIey, keds. one of the few works using intermittent vertical retorts. Sampling and Aoalyticnl Methods We wished to estimate the long-tern expasure to tarry Fumes of men in retort houses, and for this purpose we installed a continuous sampler at Rochester. Brighron [Portshdet, and Birmingham (Windsor St) s u ~ i v c l y . The sampler was in cash case placed close to where men worked. with its air intake about 1 R. above bmthing levet. The tar and other smgwdecl matter were collected on weighed $lass fibre filters wntained m a +in. { 1015 cm.) diameter brass filter holder. A i r was drawn tnrough the filtm aL a rate of 1-5 to 3 litrslmin, and the total volume was :ccot-ded with a dry gas meter. At Brtghton md Birmingham the filter holder, pump. and gas meter were contained in a metat cabinet. At Rochater it was not possible to place all the apparatus close co IIM Ffo.5 --Mark and porhbk ympler in M C a t Rochbwr. sampling point, and to avoid using a long inlet tube we kft tha filter holder outside rhe cabmet and proacted ' t :ffirency of the filter of The mask, and u rhe resulrs from radiant heat with a bright metai shield. In all CWR showed It to he close to IO?:,w used rhc same type of rhc temperatureof the Uter holder was mainrained below Inask in uuc "urthcr s t u k in other works. At 81rmtng- S O T . !t was shown in laboratory experiments (Comrnins ham [Saltley), Ortsrol, and L d s thFsc m k s wcrc worn end Lawher, 1958; CNLmniW, 1961) that samples of hy reron .iniiw workers for one, and in some case^ two, hydrcxarbonsof speciai intuert. notably 3:4-bcnzpyrerne, shrfts, and the filters were returnrd to us for analysas. once cdlened on a Mter paper did not suffcr apprtciable nlc filters were cxtractcd with cyclnl~xancfor three loss at tbh tempuaturc M e s were collened for hours m 3 Soxhlct appantus, and the darkcolawed periods of two to four weeks, and at least :wosuccessive extract was reducbd to small volume and transferred to a sam~icswere taken at each works. We JWO took il d u m n ut' alumina. The hydrocrarhnm wre %eparared number of ~ r m g l eov~er short periods (a few minutes tu chromatugxapnidly by ,-fution wirh cyclohexane and two hours) in dense fuma swing from retorts. T k wee mlfsbedon w@ed glasr fibre filtersueto m m r - &tmnii: The folk aotnauh horantl l:2-krU thcm;f rhcne.$ wrorrni datile during c from th: stable in with the 1:Z-krv caronen, reliably WiYSiS The zc:tlysi: Table I nitude i much h M&AN < POLY RElOR1 .... '. horn thr filters;some hydrocarbons arc not cornplctciy j:4 .. a b l e m %irand some could not be auxlratekydetermined with the i I I I U l f l d methods used. Only t h t arnouns of I :2-benzpyrenc. 3 :4-benzpyrenc, 1 :12.henzperyfcne, and YoDli*'ng ai''r tnhc iigurcl d e r to the maximumfor a - h pollaism in Zl sampks VRrYIPS Q)mlMs,lmn ' mronene are quoted here since these hydrocarbons can be i reltahly determind under the conditions oiaarnpline and analysis used in this study. TALfLE 3 MEAN SHIFT ZXXrOSURE TO pc)f.YCYCLlC AROMATIC HYDROCARBONS IN Kk77DRT HOUSES (MASK SAMPLES) I RHltftS 1 Exgoairs per Eight-hr. Shift Wd.1 , The mean concentrations obtained from the analysis of the long-period samples art given in Table I . The results are of the same order of mag- (Saltlcy) nitude in each orthe three retort houses and t h y are . much higher than those found in urban atmospheres. I:l-Bcnxmrmc ) , ' 1 . 5 'f :n '' e d a d d TA~ILE1 M E A N CONCENTRATION OF SUSPENDED MATTER A N D POLYCYCLIC AKOMATIC HYDROCARBONS IN THE RETORT HOUSE A I R OF CAS WORKS (LONG.PEItIOD SAMPLE51 Thefipurss in varmthesra r d a tu the range. plete shifts, and the resuits are expttzsmzd in this form i Concentralion +z./rn.' air) in Table 3. Since there was some variation in the time for which each mask was worn. the results have -- I Suspended Mum:. ! I.OM I -I__ -.~ [--jj'--iI:Z-Ecmpyrc::r cnromtcne 1:IZ-Benzpcryknc >-L&nrpy?mc. II 2.9 A.0 i ) 925 0.9 n.7 24 1.4 all been calculated as for an eight-hour shift. One of the masks usedat Brut01 and severat ot'tbsc used I_1_,10_0 - at Lacds gave very low results, hardly higher than II would br expected fmm urban air polfution; it is [ 1.n 1.7 possible that these results may be due to bad fit but They have been included when Ealculatiry the mean *Canskimi ol iar wrh some caul dusr. values. .To enable us to compare the results with those from the long-period samplcs we have a b The samples aiso mntaiwd a far higher proponjorl estimated m a n concentrations in the air from tbe ofpolycyclic hydrocarbonsthan urban air pollution: mask samples b'assuming that a breathes at Rocttestrr m a r l y O.S% of the s ~ ~ p r mnatjte~r 10 m.' of air during an eight-hour shift (I.C.R.P., was 3:4-benzpyrcne. The results from 20 hen- 1959). The resuit from tbe mask worn bv one of US Penad samples which we took, mainly at poHsla&, for o m hour at Rochester has been multiplied by extended over a very wide range, a d in Table 2 we eight for inclusion in Table 3. The portable SfImpltr have quoted only the maximum mmmtion of which we used at the same time gave a amcentration each pollutant. These then indicate the highest con- of 3 :4-benzpyre?k of 216 ps-/m?;c m v k n W i t h centrarions thnr could mnc&&jy be bmtm for the rerult from the mask .mpk would indicate that Very brief periods by a man in a ~ o fh-i. of these sampies contained over 1 % of some the breathing rate w 5.25 rn?jeiphthour shift, 3 1 4 - b ~ - about half of h a t assumed fOT the m k m g man. . Pyrene. The mask samples should give a direct measure of exposun to the polycyclic hydrocarbons for con-,- tlndicatcs paunt m x m o m . ' Particle Size Dlstribdon.-At Rochester samples were taken in the retort house with a cascade impactor and a thermal precipitator for d m s c c examination of the particles and droplcts present io the air. The sampks consisted mainiy of very small 1 'i 'i',, J t .i i $ I 18 tarry droplets: their size could not be catimated Satisfactorilywith the optical microscope, andfurther samplos ware taken at Rochester and Portdads for exammation with the electronmicra~~pe. The droplets were then s n to vary between 0.1 and 1 micron in diameter (Waller. Brooks. and Cartwnght, 1963) and all could therefore be consi&red to be respirable. We did Dot find `smokea-tes' of the type 3ecn io domestic pollution, but some coaldust particles were present with diameters from about one micron upwards. In Table 4 tbe concentrations of the carcinogenic hydrocarbon 3;4-hzpymme Found in this study have been compared with those which we have measured elsewhere. The long-pcriod and mask samples give similar results aod show that men working in retort houses are exposed to concentrations of 3 :4hzpyrene that are mom than 100 times those normally found in the City of London. Tlie CONCENTRATION OF 1:bBENZPYRENE IN RETORT very high urban figure of 1 p.g.jm? quoted was found dunng a temperature inversion and !asted For a few hours only. Apart from this instance we haw not found any wstaincd concentration of 3:4-lxnrpyrene, even in the more polIuted industrial towns, or `hbusy streets or poorly ventilated tunads which approached those found in retort h o w . The sampje taken above the retorts at Rochester repm m t s mom than 10,LXXI times the average level of 3:rl-bcnrpyt-enein the City of Lnndon, and one m a n waa e x p o d lo this hthc normal mume of hi5 duty. This is probably the most jmponant point to note when any comparison is sought between the condi- tions in horizontal and vertical retort houses. Although wc had found even h g k concentratjom in the phima isruing straight from retorts in both hwizaniat and vertical mtort hwsc$ rnm could only be exposed to these very briefly. Apart from defining t f s~paially hi& concentratioos a b v c thc hori- mntal retorts at Rocnestcr the tabniqlns wc used did not reveal any gross differences in polluuon in the general atrnosphen of the retort house3 studied. T h e results from the long-pcriod samples uabie I) showed that the mean concentrations of polycyclic hydrocarbons in the hbdzoural reton house (Rochester) grater than those in t k two vertical retort houses (Portslada and Birmingham). but since there wen? variations in the results from successive samples taken in a s h works we could not estabIish without il more extensive study whether there was a consistent difference between the poIlution in the two types ol retorthouse. Our pilot experiments using masks were carried out to try to asess esposures directly without having to consider variations rn cunccotr;lrion from one part o f the retort house toanother. There are obvious difficulties in the use ol masks as samplers; it is sometimes difficult to ensure that they fit pcrktly, and a leakwit1 lead to an underestimaleof pollution; mnvcrscly, the usual cxposummay be overestimated Saause the inan. protected by a mask. may venture into a highly polluted atmosphere which he wouid normally avoid. Thc`personal` samplers which have been developed in recent years (Sherwood and Gretnhalgh, 1960)would not be subjecf lo the same disadvantagcs. but we considered that we had alrudy obtained sufficient ioformation for our present purpose by the use of masks and fixed smplws. Wecan summarize our prescnt findings as tdlows: (a) Retort house workers are liable io be exposed to mu& higher concentrations of polycyclic am- matic hydrocarbons than is the gene& population. (6) A wide range or exposures IS to bt expected in any retort house but there is one worlung aiviron- ment, above honzontrtl retorts. whrre exposurr~m y be far geater than elsewhem. (c) Apart from the s p e c k 1 conditions above horizo~~rraeltorts, for wbicfi there is no countupan in a vertical retort house, we have not in uur plrsent study demonstrakd any gross differences between the pollution in the two t y p a of Mort house. It follows that the total amount of 3 :4-bempyrene and other poolycyclic hydrocarbons inhaied by a gas- worker during his working life will be strongly in- fluenced by whether he has been employed as a top- man in a horizomal retort house for any substantial period. Many works have changed From horhmtd to continuous vw,ical retorts during the past few decades and it is. thenfarc d%cult to BSSCSS rht i EXPOSURE TO POLYCYC'C AROMATIC NYDROCARBQNS 19 I overaU exposure of long-servioe employees. O m chmnic bronchitis among retort house worken 0th further point that was brought IO our notice was *k (Doll et af., 1965) and the uccurrence Of ChlS disease fan that all of the works had (0 be 'bladxd-out' in the general population is more d m t to estab- I during the Second World War. This was Eslid to 1kh. It has nOt previoudy beui s u 5 t o d that tarry mi- 1 mtrict the ventitation, and the whole atmosphere of fumes are concerned in the devdopment of thig i thc retort house was then far more pollulcd than at disease, but ths is the only common Urban pohltanl 1 in 1 prcsenr. Some of the men incItxdcd in the nudy of pmenl io retort houses in substantidy hjgha con- i d 1 Doll es 01. (1965) could have k e n employed in rctort 1 houses during this perid. I Our findings are of interest in relation to generai '~KLSE aimospkrjc pollution, but it is important to dis- centrations tbao are found ourside. Many Studies have indict& air pollution ;LS a possible factor in the development of chronic bronchitis. Pembcrton and Goldberg (1954) showed a relationship between 1two . i''.on ' 1~: the ! tinguish beiwecn the type of pollution which we nave foundin the retart houses and that in urban air. In a retort coal is distilled in the absence of air, and tbe produ~tsinclude tar, carbon monoxide, and other combustible gases together with some hydrogen sulphide. but with very little sulphur dioxide or black smoke. When coat is; burnt on a domestic fire, mortality from chronic bronchitis and pollution by sulphur dioxide, although this gas should perhaps be regarded jn Britain as but an index of gene& pollution. In one af the retort houses (Portslade)the possibiiity of m p w m to sulphur dioxide was considtaed. The gcneral atmosphere of the retort house was found to be alkaline rather than acid: the con- r i d , distillation products, includins tar, are emitted cach centration of ammonia was more than sufficjent to >'ing ' dme the fire is refuelled bur then js m b u s t i o n with mutralin the sulphur dioxide when both we= ioonue5 it is ' ',:tb. ton'. ctIec* lure Jdd lave and anle had iked ' JWS : Wd arc- tion ran- ma>' bow emission of smoke and sukphur dioxide. ln mare absorbed in dilute hydrogen peroxide. It is not efficlcnrheating appliances little tar or smoke is p m known whether the# gases cfectiveiy n e u t d i z one duced, and the main products are CO,and sulphur another when Inhaled together, but the conccntra- dioxide. Other sonrces. suc!i as m t o r vehicles. emit tion of sulphur dioxide (0.35 p p m . ) was within the Merent pollutants: the composition of the mixture normal range found in urban air. One operation led presenr in urban air has been dizusscd hy Comrnins to the rdcasc of sulphur dioxide into the retan house (1963. %"le tar, containing 3:4-benzpwene and atmosphere: when il hot retort was filled with coke other pdycyctic hydrocarbons. is prcsent in the ~ J I after having been scurfed, fumes containiag sulphur throughout Great Britain, and :he increased in- dioxide were emitted lor about 15 rninutcs. This cidence o f lung cancer among retort house workers happened three times a day, hut the men tended to exposed tn tar is therefore relevant to the stuay of avoid this local pollution by gas and dust. By lung cancer in the general population The results sampling in the worst of'the fumes we Found a mn- of this study mnid that or DoU PI a/.(t965) should be centration of 6 p.p.m. of sulphur dioxide. This is cansidered in rciation to two alternative hypotheses. more than is found in the general atmosphere hut it It has been shown (Stocks,1960)that the urbanjrural is only just outside the inaximum albmb1e con- gradient in lung cancer mortality i s associatfxl ~ i ~mihration for eight-hour cxposu~es in mdutty exposure to 3A-benzpmne m the air. From a Con- (5 p.p.m.)- Since the men could only haw been sideration of the lung cmca mortaiity among retort exposed to these conditions momentarjly and s~nce house workm and in the gmeral population it is we could Tmd no ncher %mrcc of sulphur dioxide in apparent that the relationshipbemeen mortality and the retort house we concluded that there could be no exposure to 3'4benzpyrene is far from linear. If a special hazard from th~spollutant in gas works. non-linear relationship hetwtxn dose arid O a t 1 1 were Other occupationai stdies have s u g g d that ex- %sumad then the increased mortality among gas- posure to dug may be a fmor in the development *workerscould be compatible with Stocks' hypo- of.chronic bronclutis (Higgiins, -me, Gihn, thesis. On the other hand, there may be other andWood, 1 9 5 9 ) a J although wedidnot examine -%of .ween rem LW- Y intop ntial mtal few ; the explanations of Stocks' findings (Buck and Brown, 1964; Waller, 1965). The failure to find a more massive discrepancy between garworkers and the general public despite :he gross difTcrcnEs in expo- Sure misht well bc interpreted as an indication that 3:Ckmpyrene is of doubtful relevance in thesa populations. l7me questions may be answered if the continued study sbould reveal any great excess of mrtallty among the top-men cornpard with mort house workers. Any link Setween the in& morraliry from this aspea in detail, it was obvious that the men were exposed to more dust [ban tfie general population. In the horizontal retort house at Rochester much coal and coke 3ust was &s.& into tht air but most of it was too coarse to inhale. In cundusioo we rhould stress that the pnscnt study was undertaken in an attempt to assess the probable exposure to polycyclic aromatic hydrocarbons of men who had worked Io retort buSeS at any time pior to 1%1. If the mortality study continues, the effect of progressive changes in g s pro- i, i ? .I ! i/ /.I ,'I I '1 I 't '1 .I {! I. :i '! .. .' I. .:, I Brit. J UR I ~i.am*k'i. 2.C o c h r a . A. L.. C(lmn I . C.. and Wcod. C. H. Pf of or tO ve is Ben2 mater]; and o varnish of pain L*' >et bermn matlc 1 quanti! some 9 bcnzenc &nz when r prefere: cimm: In t t ;emrat Cai onn. SDccial arise Frl sample! and rar The catecho nic,acid William in the u In <lfliR 13 ",, is benzent and I O , 3'