occupational exposure to dioxins by thermal oxygen cutting

Occupational Exposure to Dioxins byThermal Oxygen Cutting, Welding, andSoldering of MetalsHeinrich Michi Menzel," Ulrich Bolm-Audorff,2 Erich Turcer,2Hans Gunter Bienfait,2 Gerd Albracht,2 Dirk Walter,3 ChrisEmmel,3 Udo Knecht,3 and Olaf Papke41Zentralstelle fur Arbeitsschutz HLfU, Wiesbaden, Germany; 2HessischesMinisterium fur Frauen, Arbeit und Sozialordnung, Wiesbaden,Germany; 3lnstitut und Poliklinik fur Arbeits- und Sozialmedizin, Giel3en,Germany; 4ERGo Forschungs-Gesellschaft mbH, Hamburg, Germany

This paper focuses on one aspect of occupational dioxin exposure that is novel and unexpected.Exposures in excess of the German threshold limit value of 50 pg international toxicity equivalent(I-TEQ)/m3 are very frequent, unpredictable, and sometimes very high-up to 6612 pg

l-TEQ/m3-during thermal oxygen cutting at scrap metal and demolition sites. The same

procedure involving virgin steel in steel trade and mass production of steel objects gave no suchevidence, even though no final conclusions can be drawn because of the low number of samplesanalyzed. Low dioxin exposures during inert gas electric arc welding confirm previous literaturefindings, whereas soldering and thermal oxygen cutting in the presence of polyvinyl chloride giverise to concern. The consequences of occupational dioxin exposure were studied by analysis ofthe dioxin-blood concentration, the body burden, of men performing thermal oxygen cutting atscrap metal reclamation and demolition sites, in steel trade and producing plants as well as forindustrial welders and white-collar workers. The results concerning body burdens are in excellentagreement with the dioxin exposure as characterized by dioxin air concentration in the workplace.The significant positive correlation between duration and frequency of performing thermal oxygen

cutting at metal reclamation and demolition sites expressed in job-years and dioxin body burdenspeaks for the occupational origin of the observed overload after long times. The results reportedhere lead to consequences for occupational health, which are discussed and require immediateattention. - Environ Health Perspect 106(Suppl 2):715-722 (1998). http.//ehpnet1.niehs.nih.gov/docs/1998/Suppl-2/715-722menzel/abstract.html

Key words: dioxin, PCDD, PCDF, occupational exposure, thermal oxygen cutting, TRK-Wert,scrap metal reclamation sites, demolition sites, body burden, occupational health.

IntroductionConsidering the extent of exposure and the Therefore, it might be appropriate tonumber of papers dedicated to the impact of consider occupational health aspects ofdioxins, it is impressive to note how much dioxins also in a journal dedicated to envi-the environmental aspect outnumbers the ronmental health. In doing so, we will con-occupational point of view, even though the centrate on one specific topic with verylatter has added important knowledge about recent and rather unexpected findings as athe adverse effects on human health (1-4). result of our investigations.

This paper is based on a presentation at the International Symposium on Dioxins and Furans: EpidemiologicAssessment of Cancer Risks and Other Human Health Effects held 7-8 November 1996 in Heidelberg,Germany. Manuscript received at EHP28 May 1997; accepted 3 November 1997.

Results reported here were made possible by grants of the Hessische Ministerium fOr Frauen, Arbeit undSozialordnung, Wiesbaden, for projects to discover, counteract, and eliminate health hazards at the workplace.Part of the investigation also was supported by the Bundesministerium fur Forschung und Technologie, Bonn,project 01 HK750/5

Address correspondence to Dr. H.M. Menzel, Zentralstelle fOr Arbeitsschutz HLfU, Unter den Eichen 7,65195 Wiesbaden, Germany. Telephone: 0611-581(0)428. Fax: 0611-581-221. E-mail: [email protected]

Abbreviations used: See "Appendix."

In Germany 2,3,7,8-tetrachlorodibenzo-p-dioxin (2,3,7,8-TCDD) was classified in1986 as a chemical known to inducecancer in animals (group IIIA2) (5) and in1993 a legally executable threshold limitvalue at the workplace (TRK-Wert) of 50pg/m3 international toxicity equivalent (I-TEQ) was established (6); at the sametime the toxic potency of other members ofthe polychlorinated dibenzo-p-dioxin/poly-chlorinated dibenzofurans (PCDD/PCDF)family were taken into account by I-TEQfactors(7). The TRK-Wert is time weighedaveraged over an 8-hr shift with a short-term exposure limit (STEL) 4 times higherfor 15-min periods. Such short-term periodsmust not occur more often than 4 times pershift. Consequently, the TRK-Wert of 50pg/m3 I-TEQ must also be met when theintegration time of the measurement is 1 hror longer. With this legal basis given, theoccupational health authorities in Hessenfelt compelled to check existing dioxinexposures by investigating all hazardouswaste incineration (HWI) plants in the fed-eral state of Hessen during routine opera-tion (8,9) in 1993. In contrast to ourexpectations, the exposure levels observedwere low: always less than 8% of theTRK-Wert, which therefore can be loweredfor such workplaces. For one employee, whowas engaged in metal-repair work duringthe annual revisions of the plant, elevatedPCDF blood values were observed. Wetherefore investigated the occupationalexposure to dioxins during the annual revi-sions ofHWI plants as well as thermal oxy-gen cutting at metal reclamation anddemolition sites (10,11).

The reason to investigate dioxin expo-sures at the workplaces where metals arewelded, soldered, or thermally oxygen cutwas the expectation that in all cases wheretraces of chlororganics such as polychlori-nated biphenyls (PCBs) in old paint, oil, orthe like are involved, the three criteria fordioxin formation are fulfilled: chlorinatedcompounds, elevated temperatures in theappropriate range, and heavy metal cataly-sis. To our knowledge we were the first tofollow this line of evidence by checking theup-to-then unsuspected thermal oxygencutting workplaces. In this paper we extendour previous findings (10,11) about highdioxin exposures during thermal oxygencutting at scrap metal reclamation anddemolition sites and elevated dioxin bodyburdens in men performing such work fora long time.

Environmental Health Perspectives * Vol 106, Supplement 2 * April 1998 715

MENZEL ETAL.

Materials and MethodsSmpling and Dioxin Analysis ofAir

Air samples were always taken by means ofpumps and filters carried by the personperforming the work. To characterize theenvironment, especially when the work wasperformed in open air, additional station-ary samples were taken. In both cases parti-cles were retained by glass fiber filters andgaseous dioxins were retained by poly-urethane (PUR) foam (NILU Corp., Kjeller,Norway), since from our own observations(10) as well as from the literature (13) weknow that 10 to 30% of total PCDD/PCDFin air can be found as vapor. For economicreasons PUR plugs and glass-fiber filterswere extracted together. Personal samplerswere operated at 3.3 liters/min by means ofAlpha 1 pumps by Ametek (AmetekTesting Calibration Instruments, Largo,FL) and by Gilian (Gilian InstrumentCorporation., Wayne, NJ), which wererecalibrated before each use. The samplerhead was the standard head for total dust(Gesamtstaub/Gas [GGP]-Probenahme-kopf, product of Gesellschaft fur Schad-stoffmefgtechnik GmbH Gutvellbruggen,Neuss-Norf, Germany) (14). The air vol-ume taken varied according to the durationof the activities sampled and typically wasaround 1 m3. Prior to use, the glass-fiber fil-ter was spiked with 250 pg 13C-half-labeled2,3,7,8-TCDD. The working person wassupervised with times and types of activities,and breaks recorded as well as the atmos-pheric parameters: temperature, air pressure,humidity, and wind velocity. Stationarysamples of the environmental air were takenin accordance to immission measurements(15) by the PCDD/PCDF-immission sam-pler head developed by the Berufsgenossen-schaftliches Institut fur Arbeitssicherheit(BIA) (16), manufactured by Gesellschaftfur Schadstoffme&technik; the sampler headalso accommodates two PUR plugs and wasoperated by use of a Gravicon PM4 pump(Gesellschaft fur Schadstoffme1itechnik).The air flow rate was exactly 4 m3/hr and theoperation time was roughly the same as forthe person who carried sampler.

Total dust was determined gravimetri-cally. Prior to Soxhlet extraction withtoluene, 14 I-TEQ congeners, fully 13C-labeled, were spiked on to the filter. Theextracts were purified by use of a multi-column system including carbopack C oncelite 545 (Supelco Corp., Bellefonte, PA).Identification was done by high resolutiongas chromatography and high resolutionmass spectrometry with an SP 2331

(Supelco) and a DB 5 spectrometer UIWScientific, Folsom, CA) or a VG Autospec(Micromass, Manchester, UK). Quantifi-cation was performed by isotopedilutiontechnique using the spiked standards addedbefore extraction. Recovery rates are typi-cally 65 to 115%, the detection limit for2,3,7,8-TCDD is 0.2 to 1.0 pg/m3 for theperson-carried sampler (depending on theair volume taken) and 0.03 pg/m3 for thestationary sampler.

Sampling and DioxinAnalysis ofBloodBetween 60 and 80 ml whole blood wascollected by free blood flow in a pre-cleaned, 2500 IU heparin-containing glassjar; the jar was closed by a Teflon-coveredscrew cap. Samples were transportedfrozen and stored at -20 to -30°C. Whiledonating blood, the donor was questionedby the same interviewer, after a detailedquestionnaire that covered nutritional andsmoking habits, the dwelling environment,and details of type and duration of alloccupational activities performed so far.For the present publication, this question-naire was used to determine job-years ofthe occupation performed at the time ofthe investigation and to check for pastoccupational activities, as well as nutri-tional habits (e.g., catching and eating fishfrom local, possibly contaminated water-sheds) and dwelling conditions (e.g., livingin a flat with pentachlorophenol [PCP]-treated wood covers) which are indicativeof nonoccupational exposure to PCDD/PCDF. A more detailed investigation con-cerning the correlation between the otherparameters requested and PCDD/PCDFbody burden will be the subject of afuture paper.

For PCDD/PCDF analysis, 40 ml of thethawed blood was mixed with 35 ml waterand spiked with 1000 pg octachloro-dibenzo-p-dioxin (OCDD), respectively,100 pg of all other completely 13C-labeled I-TEQ congeners. The PCDD/PCDF and allother fat-soluble compounds were adsorbedtogether with the blood fat onto modifiedsilica (Chem-Elut, Chemelut Hydromatrix;ICT, Bad Homburg, Germany) and elutedwith hexan/isopropanol (3 + 2). Blood fatwas determined gravimetrically and analysisof PCDD/PCDF together with coplanarPCBs followed the same regime as thatdescribed for the air samples. Recovery ratesare 70 to 110% and the detection limit for2,3,7,8-TCDD is 0.5 to 1.0 pg/g blood fat,which is equivalent to 5 fg/ml whole blood.For more detailed information on the

analytical techniques we refer to previouspublications (17-19).

During the investigations reported here,we also checked for overall reproducibilityafter long times by reanalysis of the samedonors: one after 1 year, another after 2years. The analyzing laboratory was blind tothis fact and was informed only after deliv-ery of the results. Differences for the I-TEQvalues were + 12% and -3%. The meanvalue of variation for the individual con-geners analyzed was 16 ± 1 1% (n = 12) and14 ± 10% (n= 11), when considering onlyresults above the detection limit and with-out possible interferences. When all 17 con-geners are considered, mean values ofvariation of 18 ± 15% and 20 ± 17%, respec-tively, were found (n= 17 in both cases). Forthe group parameters (sum of hexa-chlorodibenzo-p-dioxin [hexa-CDD]), sumof pentachlorodibenzofuran [penta-CDF],etc., including I-TEQ values) mean valuesof 13 ± 9% and 12 ± 9% (n= 8 in both cases)was found for the two donors. We interpretthese differences as analytical variations ofthe whole procedure and not as changes inthe body burden of the person reanalyzed,which was taken to be constant. To us itappears that the long-term reproducibilitywith mean variations of less than 15% forthe sum parameters including I-TEQ valuesis well acceptable, while individual compo-nents at very low concentrations may needmore cautious interpretations.

Quntitative Evaluation ofPCDDand PCDF ToxicityPCDD and PCDF represent two familiesof closely related molecules, consisting of135 and 75 members, respectively. Thesecan be subdivided into different groupsaccording to the number of chlorine atoms(four to eight) attached: tetra-, penta-,hexa-, hepta-, octa- chlorodibenzodioxinsand furans (see "Appendix"). When analyz-ing for PCDDs or PCDFs, a single PCDDor PCDF is never found, but rather a largenumber of different ones are found, espe-cially when they are formed de novo incombustion processes. To evaluate thetoxic potency of such vast mixtures, theconcept of toxicity equivalency factors see"Appendix") has been established in thestudy (7) of the North Atlantic TreatyOrganization-Committee on the Challengeof Modern Society (NATO-CCMS) andhas gained both scientific and legislative (6)acceptance. The quantitative number ofeach factor is derived from experimentalanimals and from in vitro experiments (7).Essentially this concept provides factors to

Environmental Health Perspectives * Vol 106, Supplement 2 * April 1998716

OCCUPATIONAL DIOXIN EXPOSURE BY THERMAL CUTrING OF METALS

7 PCDDs and 10 PCDFs ranging between0.001 and 1.0, which is given only to themost potent member, 2,3,7,8-TCDD. Toobtain the quantitative number of the dioxinconcentration representing the potency ofthe mixture analyzed, each concentration ofthe 17 PCDDs/PCDFs is multiplied with itsappropriate toxicity factor and the results aretotaled; this number is the concentration of2,3,7,8-TCDD believed to have the sametoxic potency as the mixture analyzed. Theenormous advantage of this procedure is thatit enables the investigator to make very com-plex mixtures comparable by use of a singlenumber. For this reason, but also to conformto legal demands (6), PCDD/PCDF con-centrations both in air and in blood aregiven in these I-TEQ units throughoutthis paper.

StaisticsStatistical analysis of dioxin body burdenof the different groups investigated herewas performed by Kolmogorov-Smirnov 2-sample test. The dependence of dioxinbody burden on duration of exposure wasdone by linear regression analysis.

Results and DiscussionDioxin Concentration in Breathing AirIndicating Occupational Exposure

Table 1 summarizes relevant data concern-ing the dioxin exposure at workplaceswhere thermal oxygen cutting is performed.The first high exposure exceeding theTRK-Wert about 40-fold, was encounteredduring repair of the waste chute in an HWIplant. During the first two measurementslisted, the double-walled, water-cooledchute, to which the solid hazardous waste isapplied, was repaired by thermal oxygencutting and fitting of new parts. The nextfour values refer to the same activities dur-ing the next revision of an HWI plant ofidentical design, although under very differ-ent conditions due to the knowledgeobtained from the first measurements. Onthe one hand, ventilation was dramaticallyincreased and on the other hand-andmore importantly the chute was pre-cleaned extensively before any repair workwas started. As can be seen, these mea-sures resulted in a dramatic, 10- to 100-fold, reduction of dioxin exposure, eventhough the TRK-Wert still is surpassedand respiratory protection remains neces-sary To explain the results, it is reasonableto assume that during the operation of theHWI plant some of the dioxins formed inthe off-gases may condense onto the

Table 1. Occupational PCDD/PCDF exposure during thermal oxygen cutting as determined by personal air sampling.

Workplacea

HWI: waste chutenot precleaned

HWI: waste chuteprecleaned

Demolition site,plant no.lalblcld2a2b3

Scrub metalreclamation site1 a (same person 1 b)1 b (same person 1 a)2a2b34567a7b8a8b910

Production plantsand steel trading

12a2b3a3b3c

aNumbers refer to diff

Dioxinconcentration,pg/m3 I-TEQ

1.8302.430

802033140

459358

2.4626.612

6440113

98858348

1.18319171520188

771372393

Total2,3,7,8-PCDD,

pg/m321.67820.0386.911

354794

2.680

2039132126552.8741.097

2271.266

953009625118373046861162625

Ratio PCDF/ PCDD(2,3,7,8 congeners)

1.51.91.21.21.11.1

121.042.090.8

271.171.00.373.9

3.417.224.526.02.15.10.83.43.91.15.89.15.0

27.3

1 62 0.271.7 1.655 0.00080.07 11 0.130.8 9 2.03.0 539 0.340.3 1 1 0.7

ferent plants; letters refer to different determinations.

water-cooled waste chute and be preservedthere at relatively low temperatures.When the chute is heated again duringrepair, the dioxins present will be mobi-lized, unless they were removed before byintensive cleaning. Thus, precleaning ofthe waste chute reduces dioxin exposureduring repair, but will mobilize dioxins aswell, which requires protective measures.

The exposure to dioxins in HWI plants,especially during extensive repair, was to beexpected, whereas the generation of dioxinsduring thermal oxygen cutting performedoutside of HWI or MWI plants, was unre-ported. As can be seen in Table 1, six ofseven person-specific air samples at threedifferent demolition sites clearly exceed the

Multiple ofoccupationalthreshold limitTRK-Wert, 50pg/m3 I-TEQ

36.648.61.60.40.62.8

9.27.2

49.2132.2

1.30.82.3

2.03.87.0

23.70.40.30.30.40.40.21.52.70.51.9

0.020.030.0010.0160.060.006

Duration ofsamplingtime, hr

2.12.33.73.93.53.3

6.35.64.41.96.36.34.9

0.44.52.63.13.91.24.04.14.14.77.45.05.64.2

2.96.06.04.64.64.6

50 pg/m3 I-TEQ level, one of them 132-fold. For practical reasons demolition sitesare excepted from TRK-Wert applications(see "TRK-Wert" in the "Appendix"),which does not affect the obligation to pro-tect those exposed to dioxins here just aseffectively as at every other workplace.Demolition site 1 was a coal-operatedpower plant; site 2 was an old gasometer;and site 3 was a crane. In no case were sus-picious materials, like electric transformersor condensers, subject to thermal oxygencutting, but rather steel constructions, high-pressure pipes, and boilers, mostly paintedbut free of grease and oil. In the case of thegasometer some black plastic material waspresent at a few spots, this material was


MENZEL ET AL.

positive for chlorine components, as deter-mined by a quick test. The most surprisingresult is that of demolition site 1d, whichhad the highest dioxin exposure everencountered in our studies, since in thiscase a neat steel rotor of a steam turbinewas cut. While in the case of the HWI sam-ples the ratio of PCDF/PCDD (taking intoaccount only 2,3,7,8 congeners) is ratherconstant between sites 1 and 2, irrespectiveof the dioxin I-TEQ concentration, thisratio varies very much for the samples takenfrom the demolition sites. In addition, thisratio appears to be higher when the dioxinI-TEQ concentration is high.

Other workplaces where old material issubject to thermal oxygen cutting are scrapmetal reclamation sites. Of the 14 person-related air samples from 10 sites, half ofthem clearly exceed the TRK-Wert, onemore than 20-fold. This is-as in the caseof the demolition sites-the more surpris-ing, since all the operations monitoredwere performed in the open air, often atconsiderable wind speeds. The materialworked on was typical for metal reclama-tion sites: big containers, bulky steel con-structions, heavy-duty machines, and soon. The material was always paint coated,and often greasy and oily. Of the paintsamples collected and analyzed by a quicktest for organohalides, about 25% werepositive and there was no obvious correla-tion between positive results and dioxin I-TEQ concentrations at the respectiveworkplaces. With respect to the ratioPCDF/PCDD (2,3,7,8 congeners only) apicture emerges that is similar to the caseof the demolition sites, although both theratio and the dioxin I-TEQ concentrationsare lower, but they parallel each other.

As can be seen from Table 2, theambient dioxin I-TEQ concentrations in thevicinity of the open-air thermal oxygen cut-ting, as determined by stationary sampling,is much lower than for the person-related airsamples, but much higher than ambientopen air values (20). At metal reclamationsite 2 this immission value reaches two-thirdsof the occupationally defined TRK-Wert,which is impressive, since the cutting opera-tions were performed 5 to 10 m apart in theopen air and indicates that we deal here withan environmental problem as well.

From these results it is obvious thatthermal oxygen cutting outside of incinera-tors is also a highly relevant activity withrespect to occupational dioxin exposure. Atmetal reclamation sites as well as at demoli-tion sites, such exposures are very frequent,sometimes very high, and in no way pre-

Table 2. Environmental PCDD/PCDF air concentrations near thermal oxygen cutting of metals in the open air asdetermined by stationary sampling

Mulfiple ofoccupational

Dioxin Total threshold limit Duration ofconcentration, 2,3,7,8-PCDD, Ratio PCDF/ PCDD TRK-Wert 50, sampling

Workplacea pg/m3 I-TEQ pg/m3 (2,3,7,8 congeners) pg/m3 I-TEQ time, hrDemolition site

1 c,d 1.45 4.8 1.5 0.03 4.8Metal reclamationsite2 39 12.3 23.5 0.78 4.43 4 8.1 3.0 0.08 4.24 0.18 1.7 1.9 0.004 1.55 0.15 56.5 0.02 0.003 2.16 3.5 6.5 3.1 0.07 2.37 2.1 15.2 0.95 0.04 2.68 3.8 27.1 1.1 0.08 1.0

&Numbers refer to different plants; letters refer to different determinations.

dictable. The necessity for prompt protec-tive measures at scrap metal reclamationand demolition sites does not require-andcannot wait for-detailed models explain-ing dioxin generation at such places. Incontrast, as we can in no way assume thatat these working places dioxin exposure iswell below the TRK-Wert at all times, wemust do what can be done immediately:using personal respiratory protection.

The validity of this conclusion stronglydepends on how well our results representthe work shift of a person performing ther-mal cutting of metals, especially withrespect to the percentage of the shift cov-ered. As can be seen from the last columnin Table 1, the duration of personal sam-pling is, with one exception, always longerthan 1.2 hr. This exception, determinationla, represents the attempt to trace thesource of dioxin exposure during thermalcutting by operating the person-carriedpump only when cables attached to thescrap metal were burned, while determina-tion lb averages over the whole shift of thissame person. The results show that burningcables is not an important contributor todioxin exposure during thermal oxygen cut-ting, as exposure during this activity islower than the exposure averaged over thewhole shift (ib). Calculating the median ofthe duration of our exposure assessmentsat HWI plants, scrap metal plants, anddemolition plants yields the value of 4.1 hr(n= 27), that is more than half the 8-hrshift. In addition, all determinations-with the exception of la, which servedother purposes-meet the criterion to rep-resent the TRK-Wert, as they all lastedlonger than 1 hr (see "Introduction" and"Appendix"), which even holds true for the

5th percentile (1.4 hr), while the 95th per-centile (6.3 hr) is in excess to 75% of the8-hr work shift.

Thermal oxygen cutting is alsofrequendy performed in steel trading places(cutting to size) and in industrial plants(mass production of customer designedobjects from thick steel plates). As can beseen in Table 1, none of the six person-related air samplers from three plants indi-cate dioxin concentrations in excess to 1.5pg/m3 I-TEQ. These low concentrationsalso correlate with typically very lowPCDF/PCDD ratios and give rise to thehope that in these settings excessive exposureto dioxins may not occur, even though thenumber ofsamples and ofplants is at presenttoo low to exculpate these workplacesalready now.

The high dioxin exposures at thermaloxygen cutting sites prompted us also tolook into workplaces where steel iswelded, even though Linde et al. (12)reported low exposures. As can be seenfrom Table 3, electric welding within anHWI plant leads to exposures close to theTRK-Wert, whether it is done in theboiler or at the precleaned waste chute.This, however, we associate with the liber-ation of preformed dioxins and not as aconsequence of the welding process itself.In accordance with Linde et al. (12) wefind low dioxin exposures on inert gaselectric arc welding when constructionelements for a bridge were mass producedin an industrial workshop. Although itshould be worthwhile to scan additionalwelding workplaces, we do not see a highpriority for the surveillance of workplacesdealing with virgin steel. We shouldrespond differently to the repair welding of


OCCUPATIONAL DIOXIN EXPOSURE BY THERMAL CUTTING OF METALS

Table 3. Occupational PCDD/PCDF exposure during welding as determined by personal air sampling.

Multiple ofoccupational

Dioxin Total threshold limit Duration ofconcentration. 2,3,7,8-PCDD, Ratio PCDF/ PCDD TRK-Wert 50, sampling

Workplace pg/m3 I-TED pg/m3 (2,3,7,8 congeners) pg/m3 I-TEQ time, hr

Welding (electric arc)in HWI: - -Boiler pipes 56 541 1.5 1.1 3.2Waste chute, cleaned 44 1.058 1.1 0.9 2.3Waste chute, cleaned 11 204 1.2 0.2 2.2Industrial workshop 1.7 34 0.5 0.03 4.8Industrial workshop 1.9 236 0.1 0.03 5.8

old constructions, in which we see a closerelationship to thermal oxygen cutting andwhich should be investigated.

Because we started our search fordioxin exposures outside of HWI plantswith the hypothesis that such exposure isdue to chlororganics present when the met-als are heat stressed, we specifically lookedat two workplaces where the presence ofchlororganics during heating of metals isknown. One is the soldering of polyvinylchloride (PVC)-insulated copper plumbing(WICu-Rohr), which leads-depending onthe extent of removal of the PVC insula-tion from the spot to be soldered-to moreor less pyrolysis ofPVC. As seen in Table 4,the TRK-Wert was not surpassed in ourmeasurements, and accordingly, the PCDF/PCDD ratio is on the low side. The resultsnevertheless indicate that such work is to beperformed under very good ventilationonly, especially in narrow rooms, and thatit is essential to remove the PVC insulationgenerously by mechanical cutting and neverby burning.

Another example of involvement ofPVC in the heat treatment of metals is itsrecent use as a component in anticorrosivepaints. As seen in Table 4, thermal oxygencutting of steel plates treated with suchpaint leads to dioxin exposures well in excessof the TRK-Wert. The discontinuation of

the use of PVC as a component in anti-corrosive paint is the solution to this prob-lem, which is the only example of thedioxin exposure during thermal oxygencutting for which we can identify thesource and which therefore shouldbe eliminated.

Dioxin Body Burden ofOccupationally Exposedand Unexposed MenConsidering both the frequent and highdioxin exposures during thermal oxygencutting at scrap metal reclamation anddemolition sites it is of interest to investi-gate whether the dioxins have reached themen performing such work, enhancingtheir body burden. Although exposure todioxins via respiration is of minor impor-tance for the population at large [1.2% oftotal intake according to Travis (20),0.01%, respectively, for people living nearcontaminated sites (21)], this is expectedto be different for men performing thermaloxygen cutting at scrap metal reclamationand demolition sites. Since the TRK-Wertis 2.5 x 103-fold (20), respectively 106-fold(21) higher than the ambient dioxin airconcentrations assumed in these studies,exceeding the TRK-Wert frequently anddrastically should increase the body burdenof those affected.

Table 4. Occupational PCDD/PCDF exposure during soldering and thermal oxygen cutting of metals involving PVCpyrolysis as determined by personal air sampling.

Multiple ofoccupational

Dioxin Total threshold limit Duration ofconcentration, 2,3,7,8-PCDD, Ratio PCDF/ PCDD TRK-Wert 50, sampling

Workplace pg/m3 I-TEQ pg/m3 (2,3,7,8 congeners) pg/m3 I-TEQ time, hr

PVC pyrolysisThermal cutting 80 118 5.6 1.6 2.1of steel coatedwith PVC paint 77 160 5.2

Soldering of PVC 9 38 2.8 0.2 0.5insulated copperpipes (WiCu) 21 136 3.3 0.4 0.4

w1

40-

CDa0' 30

CDCL4-P

30-

00.

0 -20-

10-

0-

T171

-_n ui_s Widt Whita SarmanScrap virgin sollar populaton

Figure 1. Dioxin body burden of men performing ther-mal oxygen cutting, welding, or administrative work for1 job-year or more, compared to the German populationat large (19). The figure shows median and 95th per-

centile of body burden as well as number of workers.

As seen in Figure 1, the dioxin bodyburden of the German population at largein 1994 (n= 134, median body burden17.3 I-TEQ pg/g blood fat) (19) does not

differ from our local control group of malewhite-collar workers (n= 16, median ofbody burden 18.5 I-TEQ pg/g blood fat),while men currently performing thermaloxygen cutting in production plants or

steel trade are higher (n= 7, median 25.1 I-

TEQ pg/g blood fat) and industrial weldersstill somewhat more increased (n= 9,median 29.9 I-TEQ pg/g bloodfat). At firstsight 21 men currently performing thermaloxygen cutting at scrap metal and demoli-tion sites appear not to differ from the lat-ter two groups, since the median is 26.9I-TEQ pg/g blood fat. This figure distortsthe facts indeed, since 20% of the 21 men

had less than 1 job-year, while all membersof the other groups had job-years above 1.Using the cut-off criterion of equal to or

larger than 1 job-year brings the numberfor men at scrap metal reclamation anddemolition sites to 17 and the median ofdioxin body burden to 44.4 I-TEQ pg/gblood fat (Figure 1), which differs from allother groups and indicates the occurrence

of enhanced body burden. The statisticalanalysis of this group as compared to our

local reference group yields a statisticallysignificant difference at p < 0.05.

Assuming that dioxin body burden is a

consequence of occupational exposure, thenthe former should be a function of the


MENZEL ET AL.

150- A Thermal oxygen cutters of virgin steelcD * White-collar workers

100- * Industrial welders

CD 50-

0 10 20 30 40 50

Figure 2. Dioxin body burden as a function of job-yearsfor white-collar workers, industrial welders, and ther-mal oxygen cutters in the steel trade and industrialworkshops.

extent of exposure. This is expressed best interms of job-years, which is the time in yearsspent in the profession multiplied by thefraction of time spent with performing theactivity in question, that is: welding or ther-mal oxygen cutting, while for white collarworkers this fraction is set to one. Plottingdioxin body burden for each of the fourgroups as a function of job-years shouldmake any causal relationship between occu-pational exposure and dioxin body burdenobvious and at the same time should controlfor artefacts, as for instance the confoundingof age, using the white collar group as areference without occupational exposure.

As shown in Figure 2, no dependence ofdioxin body burden from job-years can befound for the white collar workers, theindustrial welders or the thermal oxygen cut-ters working in steel trade or industrial work-shops. The results shown in Figure 3 are verydifferent; the figure shows the increase ofdioxin body burden with the duration ofthermal oxygen cutting at scrap metal anddemolition sites (r= 0.79 atp< 0.001), corre-lating body burden and duration of dioxinexposure with high statistical significance.After very long times the body burden canrise to more than 10-fold of the typical bodyburden of the population at large.

Figure 3 also indicates that the use ofrespiratory protection by four men cuttinghigh-grade steel brings about a reduction ofdioxin body burden. Furthermore, one per-son in Figure 3 differs from all others (soliddot), as determined by use of the question-naire. While for all other persons no indica-tions of additional dioxin exposure fromnutritional habits, dwelling conditions, orprevious occupational activities were found,

300-

cmcU'

1-

cm4-cc

-0 100-0

* See * belowA With respiratory

protectiono Without respiratory

protection-r= 0.79 at p = 0.001 c

0

A0 AsA- o

10 20 3

Job-years

Figure 3. Dependence of of dioxin body bumal oxygen cutters at scrap metal reclademolition site, by number. of job-years. ilikely previous occupational dioxin exposuin paper mill).

he is the only one whose past occilikely to have contributed to hidioxin body burden more than hone, since he worked 14 years aanic in a paper-producing factory1 year at a scrap metal reclamzwhere he performed oxygen cuttprevious occupational activity is ca confounder with respect to dicburden. Since the calculated line3 is not visibly affected by incexclusion of this data point and thtion coefficient r is unchanged, wexdude this data point.

ConclusionOur first reports on novel occidioxin exposures outside of incand chemical plants, that is, the.gen cutting at scrap metal redamdemolition sites, have been verifiresults reported here to the exrequires immediate response (.exposure in excess to the TRKthese sites is so frequent and consithese workplaces can no longer bered safe from overexposure to ditherefore protective actions mustThe exposures can be high and aridictable at all. The excessive dicburden of people performing suchmore than 1 job-year in additi4clearcut time dependence of the ibody burden with job-years 1doubt about the causal relationsi

workplace, thermal oxygen cutting at scrapmetal reclamation and demolition sites.The mechanism of dioxin formation atthese places remains unresolved andrequires additional efforts to exclude suchexposures for the future.

Thermal oxygen cutting ofvirgin steel inthe steel trade and mass production of steelitems was found inconspicuous in our inves-

40 50 tigations, a result that also correlates wellwith the moderate dioxin body burden, notdiffering from the population at large and

irden of ther- the lack of its increase with job-years for the'Person wih men performing such work. However, bothire (14 years numbers of blood and air samples are lowand reconfirmation of this first, desirable,

result is required before these workplacesupation is can be definitely exculpated.is present Of concern is the use of PVC as alis present component in anticorrosive paints, whichs a mech- leads to exposures well in excess of theand only TRK-Wert when steel thus preserved is cutation site thermally. Since in this case the cause ofting. This dioxin exposure is obvious and known, it is-onsidered necessary to exclude such exposure in the)xin body future by banning PVC as a component ofin Figure anticorrosive paint.lusion or From these results we draw the followingie correla- condusions:;e did not * Thermal oxygen cutting at scrap metal

reclamation and demolition sites mustbe replaced by mechanical cuttingwherever possible.

upational * If unavoidable, such work must be per--inerators formed with respiratory protection,rmal oxy- which covers both particulates andiation and organic vapors (in Germany: P3/A2ied by the filters).,tent that * The causes of dioxin exposure in these22). The workplaces must be identified by more-Wert at detailed investigations to replace per-istent that sonal protective measures with actions)e consid- that prevent the formation of dioxinsioxins and under these circumstances.be taken. * Repair welding might be of concern

*e not pre- and should be investigated, while weld-xin body ing of virgin steel gives no reason fori work for concern.On to the * Thermal oxygen cutting ofvirgin steel inncrease of an industrial setting may not be a causeleaves no for concern but should be investigatedhip to the further.

AppendixAbbreviations and Glossary2,3,7,8-PCDD (polychlorinated dibenzo-p-dioxin): sum of the concentrations of allPCDDs carrying a chlorine atom in the2,3,7, and 8 position. The corresponding

sum for the 2,3,7,8-PCDF can be calculatedfrom this number and the ratio PCDD/PCDF in Tables 1 to 4.

2,3,7,8-TCDD (2,3,7,8-tetrachloro-dibenzo-p-dioxin): the most potent, toxic

member of the 210 individual chemicalsthat form the PCDD/PCDF (dioxin)family. This chemical was released inkilogram quantities at the accident inSeveso.


nl rcU-r-

720

OCCUPATIONAL DIOXIN EXPOSURE BY THERMAL CUTTING OF METALS

HRGC (high resolution gas chromatog-raphy).HRMS (high resolution mass spectro-

metry).HxCDD, HxCDF (hexachlorodiben-

zodioxins, hexachlorodibenzofurans): thesecompounds carry six chlorine atoms in thedibenzodioxin (-furan) molecule.HWI (hazardous waste incineration

plant).I-TEF (international toxicity equiva-

lency factors): PCDD/PCDF congener-specific factors ranging between 1.0 for themost potent 2,3,7,8-TCDD and 0.001 forthe least potent congeners like OCDD orOCDF (6,7,22).

I-TEQ (international toxicity equiva-lents): calculated by multiplying the con-centration of each PCDD/PCDF congenerwith its specific I-TEF (see above) and sum-ming these numbers (6,7,22). The resulthas the dimension of the concentrationsused in the calculations and represents theconcentration of 2,3,7,8-TCDD thatwould have the same toxic potency as thatspecific complex mixture of PCDD/PCDF.This procedure permits the expression ofthe toxic potency of very complex mixtureswith a single number, which is importantin the quantitative comparison of mixturesneeded in developing regulations. Thisconcept has been extended to other chem-icals and complex mixtures such asPCBs (22).

I-TEQ congener: a member of thedioxin family for which an I-TEF hasbeen assigned. It carries a chlorine atomin the 2,3,7, and 8 position of the

dibenzo-p-dioxin-, the dibenzo-p-furanmolecule, respectively.

NATO-CCMS (North Atlantic TreatyOrganization-Committee on the ChallengeofModern Society).OCDD (octachlorodibenzo-p-dioxin):

the dibenzodioxin molecule carrying thehighest possible number-eight-ofchlorine atoms.OCDF (octachlorodibenzofuran): the

dibenzofuran molecule carrying eight chlo-rine atoms, the highest possible number.

PCB (polychlorobiphenyl): two cova-lently bound benzene rings (biphenyl) car-rying varying numbers of chlorine atoms,up to ten.

PCDD/PCDF (polychlorodibenzo-p-dioxins/polychlorodibenzofurans): diben-zo-p-dioxin and dibenzofuran moleculescarrying varying numbers of chlorineatoms at different places within the mole-cule, colloquially abbreviated as dioxins orthe dioxin family.

PCP (pentachlorophenol): used as awood preservative in the past and contami-nated by PCDDs to varying degrees.PUR (polyurethane): a plastic that

forms foams with large surfaces capable ofadsorbing all types of gases, includinggaseous dioxins.

PVC (polyvinyl chloride): chlorine-containing plastic that is mass produced.

STEL (short-term exposure limit): con-centration of a chemical at the workplaceelevated above the time-weighed, averagedexposure limit and tolerated only if it doesnot surpass a given multiple of the expo-sure limit and if such episodes do not occur

more often than a given number per shiftand do not last longer than a given time,typically 15 min.

TRK-Wert German (Technische RichtKonzentration): a threshold limit value atthe workplace for cancer causing chemicals.In contrast to the MAK-Wert (German:Maximale Arbeitsplatz Konzentrationen)the TRK-Wert is defined by technical feasi-bility and not by scientific deduction of alimiting value which, if not surpassed, isassumed not to harm the health of humansat the workplace. As a consequence,TRK-Werte must be lowered whenevertechnically possible, thereby reducing theprobability of causing cancer in people. TheTRK-Wert is time-weight averaged over an8-hr shift; short-term levels over a period ofup to 15 min may surpass the TRK-Wertnot more than 4-fold. The sum over suchpeak times must be less than 1 hr per shiftwith the consequence that determinationsaveraging over 1 hr and more must keepwithin the 50 pg/m3 I-TEQ limit or other-wise the TRK-Wert is surpassed. Withrespect to demolition sites, there is a legalpeculiarity because these are excepted fromthe TRK-Wert application (6). This excep-tion requires legal clarification, but evennow it does not in any way interfere withactions to be taken to protect thermal oxy-gen cutters at demolition sites from overtdioxin exposure as well as at any otherworkplace, which on the contrary isrequired immediately.

WICu-Rohr (German: warme [heat]insulated copper pluming): the insulatingmaterial is typically made from PVC.

REFERENCES AND NOTES

1. Manz A, Berger J, Dwyer JH, Flesch-Janys D, Nagel S,Waldsgott H. Cancer mortality among workers in a chemicalplant contaminated with dioxin. Lancet 338:959-964 (1991).

2. Flesch-Janys D. Preliminary results on the exposure to poly-chlorinated dioxins and furans and mortality in a cohort ofworkers of a herbicide producing plant in Hamburg, Germany.In: Current Views on the Impact of Dioxins and Furans onHuman Health and the Environment. Proceedings of TheToxicology Forum, 9-11 November 1992, Berlin:TheToxicology Forum, eds. The Toxicology Forum:Washington,1992;207-217.

3. Fingerhut MA, Halperin W, Marlow D, Mortality among U.S.workers employed in the production of chemicals contami-nated with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). In:NIOSH Final Rpt PB 91125971. Springfield, VA:NationalTechnical Information Service, 1990.

4. Zober MA, Messerer P, Huber P. Thirty-four-year mortalityfollow-up of BASF employees exposed to 2,3,7,8-TCDD afterthe 1953 accident. Int Arch Occup Environ Health62:139-157 (1990).

5. Deutsche Forschungsgemeinschaft. List ofMAK and BAT values.

Appendix IIIA2: 2,3,7,8-Tetrachlordibenzo-p-dioxin.Mitteilung XXII der Senatskommision zur Prufung gesund-heitsschadlicher Arbeitsstoffe. Weinheim:VCHVerlagsgesellschatf, 1986

6. Technische Regeln fur Gefahrstoffe TRGS 901 Nr. 42.TRK-Wert fur chlorierte Dibenzodioxine und -furane.Bundesarbeitsblatt. (Bundeministerium fur Arbeit undSozialordnung, ed.) Koln, Germany:W Kohlhammer Verlag,1993; 71-73.

7. NATO. Pilot Study on International Information Exchange onDioxins and Related Compounds. Scientific Basis for theDevelopment of the International Equivalency Factors (I-TEF)Method of Risk Assessment for Complex Mixtures of Dioxinsand Related Compounds. Rpt No 178. North Atlantic TreatyOrganisation, December 1988.

8. Bo[m-Audorff U, Menzel HM, Murzen R, Papke 0, Turcer E,Vater U. Dioxin- und Furanbelastungen von Beschaftigten inSondermullverbrennungsanlagen. In: Verhandlungen derDeutschen Gesellschaft fir Arbeitsmedizin und Umweltmedizine.V. (Kessel R, ed). Fulda, Germany:Rindt Druckerei,1993;105-108.


MENZEL ETAL

9. Papke 0, Ball M, Lis ZA, Menzel HM, Turcer E, Bolm-Audorff U. Occupational exposure of chemical waste incinera-tor workers to PCDD/PCDF. Organohalogen Compounds21:105-110 (1994).

10. Menzel HM, Turcer E, Bienfait HG, Albracht G, Papke 0,Bolm-Audorff U. Dioxinexposition beim Brennen, Trennenund Schweigen von Metallen. In: Verhandlungen derDeutschen Gesellschaft fuir Arbeitsmedizin und Umweltmedizine.V. (Munzberger E, ed). Fulda, Germany:Rindt Druckerei,1996;323-327.

11. Menzel HM., Turcer E, Bienfait HG, Albracht G, Papke 0,Ball M, Bolm-Audorif U. Exposure to PCDD's and PCDF'sduring welding, cutting and burning of metals. OrganohalogenCompounds 30:70-75 (1996).

12. Linde H, Striefler B, Stuber H. Polychlorierte Dibenzodioxineund -furane an MAG-Schweif3platzen. Staub, Reinhaltung derLuft 55:179-182 (1995).

13. Stockmann R, Hahn JU, Lichtenstein N, Neumann H-D,Schick W. Polychlorierte Dibenzodioxine und -furane anArbeitsplatzen, In: Messungen von Gefahrstoffen BIAArbeitsmappe Nr 7552, Gesamtstaub. Bielefeld, Germany:ErichSchmitt Verlag, 1993.

14. Berufsgenossenschaftliches Institut fur Arbeitssicherheit. BIAArbeitsmappe Nr 7552, Gesamtstaub. Bielefeld, Germany:ErichSchmitt Verlag, 1989.

15. Verein Deutscher Ingenieure. Handbuch Reinhaltung der Luft,VDI Richtlinie 3499 Blatt 2, Messen von PolychloriertenDibenzo(p)dioxinen und Dibenzofuranen. Verein DeutscherIngenieure, ed. Berlin:Beuth Verlag, 1993.

16. Berufsgenossenschaftliches Institut fur Arbeitssicherheit. BIA

Arbeitsmappe Nr 6880 Gesamtstaub. Bielefeld, Germany:ErichSchmitt Verlag, 1989.

17. Papke 0, Ball M, Lis ZA, Scheunert K. PCDD/PCDF inwhole blood samples of unexposed persons. Chemosphere19(1-6):941-948 (1989).

18. Papke 0, Ball M, Lis A. PCDD/PCDF in humans-an update ofbackground data. Organohalogen Compounds 13:81-84 (1993).

19. Papke 0, Ball M, Lis A. PCDD/PCDF und coplanare PCB inHumanproben; Aktualisierung der Hintergrundsbelastung,Deutschland 1994. Organohalogen Compounds 22:275-279(1995).

20. Travis CC, Hattemayer-Frey HA. Human exposure to dioxin.Sci Total Environ 104:97-127 (1991).

21. U.S. EPA. Estimating Exposure to Dioxinlike Compounds.Chapt 9-34. EPAI600/6-88/005B. Washington:U.S.Environmental Protection Agency, 1992.

22. Menzel HM, Albracht G, Bienfait HG, Turcer E, Walter D,Emmel C, Knecht U, Papke 0, Bolm-Audorff U. Dioxin-expositionen und Biomonitoring an Schneidbrennar-beitsplatzen auf Schrottplatzen., Abbruchbaustellen und beiindustrieller Fertigung. Dokumentationsband uber die 37.Jahrestagung der Deutschen Gesellschaft fuir Arbeitsmedizinund Umweltmedizin e.V. in Weisbaden. (Borsch-Galetke E,ed). Fulda, Germany:Rindt Druck, 1977; 497-499.

23. Safe S. Polychlorinated Biphenyls (PCBs), dibenzo-p-dioxins(PCDDs), Dibenzofurans (PCDFs) and Related Compounds:Environmental and Mechanistic Considerations WhichSupport the Development of Toxic Equivalency Factors(TEFs). Critical Review in Toxicology. Vol 21 (McClellan RO,ed). Boca Raton, FL:CRC Press, 1990;51-88.

722 Environmental Health Perspectives * Vol 106, Supplement 2 * April 1998

occupational exposure to dioxins by thermal oxygen cutting

Documents