Document 6RRxMJdy5mEyqJv1woabb4ZO6

"ersar mc --/ends (?) OPTS 6203; Pea (Flres) ------------------ File Written Convn. December 14. 1463 Ms. Denise Keehner Chemical Regulations Division Office of Toxic Substances U.S. Environmental Protection Agency Room 341, East Tower Waterside Mall Washington, D.C. 20460 Subject: PCS Capacitor Explosion In Finland Dear Denise: I am enclosing several writeups of the events surrounding an explosion of PCB capacitors that occurred In Finland on August 2, 19S2. I Just received these from Dr. Dingier of Zurich Insurance In Zurich, Switzerland. Items of major Interest Include: Analytical results Indicating that about 10X of the tcdfs were 2,3,7,8-TCDF. Cleanup methods used and protective clothing for workers. Reoccupancy criteria: 100 *g/m* PCSs 100 ng/m* total furans 10 ng/mJ 2.3.7.8-TCDF , Medical surveillance program and results. I do not see anything In these reports that Is surprising In light of what we have seen and experienced In One Market Plaza and Binghamton. Apparently, banks of capacitors can act Just like transformers. Please let me know If you have any questions concerning these reports. Very truly yours. Robert A. westln RAW/lg Senior Chemical Engineer 715C-7S End: Ruohlo, V. *0n the Brink of Catastrophe." Industrial Insurance. 1993. Pp. 8-13. Elo, 0.A. and P. Vuojolahtl. 'Capacitor Fire In a Paper Mill at lmatra, Finland.* undated manuscript. SB50 VERSAR CENTER X O *OX 1S4 SMINCFIEU}, VIRGINIA J2151 TILirHONI 17031 7S0-30OO TELEX M11I3 MONS 213971 BY I'EIKKO RVOHIO Eiuo-Gutitu Oy's KaukepdB Milts a on* of the largest forest industry production facilities tn Finland. tna^rni1 \n%. |-'3* ^993 HONS 213972 The oil containing PCS which leaked onto the floorfrom the dam aged capacitor banks war cleaned up using plastic buckets and ab sorbent. EXPLOSION in the capacitor ROOM A Imle after midnight the night shift supervisor of the pa perboard line heard two succes sive loud explosions from the mill's electrical switchgear room He notified the Cue bri gade and the mill electrician on dun from the line's control room Soon after this, heavy smoke forced the paperboard line supervisor and the person nel on dull m tbe adjoining pulp tank room to vacate the control room The fire brigade made a swift response, but since no flames could be seen, it took almost half an hour to pinpoint the site of the explosion, a capacitor bank behind a 10 kV switchgear facility. The pressure wa\e created by the explosion escaped through the door from ihe capacitorfacility seen tn the background into the 10 k V jwrtchgear room seen In theforeground The areas from which sam ples mere taken can be seen in the soot covering the floor initial uncertainty ABOUT THE TOXIC EFFECTS OF PCBS Oil containing PCBs is used in capacitor banks as a dielec tric and coolant In the accident at Kaukopaa. the gasified Liquid surged with explosive force into the open A couple of hours af ter the explosion ten people were taken to hospital They were all suffering from the same svmptoms nausea, heavy per spiration and headache By this time it was clear that thegasi- fed liquid contained PCBs, but lo-j '983 Airborne PCB concentrations mere measured almost consiamh dur ing the first fem dart The picture thoms a mraiunng ream /coking for a suitable location for one of the suction pumpt used in the sam plinf The PCB andfuran samples from turfacet mere collected from areas measuring 10 cm x 10 cm using cotton wadding dipped in acetone or alcohol. 4. MQNS 213973 no one on the site was familiar wuh the effects or exposure to PCBs or with ns treatment Ini tial mformanon about the potemtal effects of exposure to the (as was found in the library of the mill's doctor As ihe ni|ht when the explo sion occurred was warm, those exposed to the reaction products of PCSs wilted for treatment on the hospital (rounds Tbeir s>mptoms disappeared in the fresh air. The first symptoms may have been caused by exposure to chtonne and hydrochloric acid, which had also been produced in the decomposition of PCBs. As soon as the first symptoms were reported, the fire brigade dosed ofT the contaminated area with wanurt( si(ns Because of the initial uncertainty about the effects of exposure to PCBs, several mil! employees moved about in the contaminated area without proper personal protec tion The total number of people uho bad been in the contaminated area and who therefore had at least a theoret ical chance of slight exposure was estimated at 150, among them industrial hygienists and cleaning personnel. Ail these people were later interviewed. Their movements and lengths of suy in the contaminated area were noted and samples of their blood were taken. Follow-up studies have indicated that the exposure to PCBs and ns reac tion products was at a low level, very likely thanks to the speedy action takeo immediately after the incident IKFORMATION STARTS TJUCKUKG IK The fust outside expert lo reach the site of the explosion was an industrial hygienist from the Lappeenrinu Regional Inrutute of Occupational Health. He miualed gas concentration sampling dunng the earl) morn ing hours The results, which indicated that the level of PCB concentrations in the form of gas in the contaminated area was sufTieieot lo require person al protection, were received on the afternoon of Ihe day after the explosion. Two chemical safn> experts from Industrial 10 The importance ofskin protection was not fully understood immedi ate!) after the accident Scaffolding and plastic sheet was used to create a makeshift shelter foe donning protective wear and equipment Insurance armed on the morn ing following Ihe explosion. They brought wuh them litera ture oo PCBs acquired from abroad and on the hazards of its decomposed elements. On the afternoon of the time day samples of soot were flown to the Chemical Instrument Center of The Uotvertity of Helsinki for analysis The prelimmsry results indicating the presence of dibenzofurans were received on the evening of the tame day . The capacitor manufacturer ar rived at the mill tt xboul the same time, bringing with him information oo the constructioo of the capacitor banks and on the amounts of PCBs used in them. CLE-tK-i'PASD MEASLREME^JS HASP IK HASP A large number of samples were inalyzed to establish Ihe size of the contaminated area nd ihe concentrations of hat- HONS 213974 Change ofclean-up shifts. The two "spacemen" on the right'hand side ha\e lust returned from a one-hour dean-up shift white the two men on the left are getting ready for their turn on duty. protective clothing and equipment were used In areas which had been cleaned up until sufficiently low concentrations were registered. be done wearing protective Clothing and using protective devices The pulp unk depinment, including the control room, un derwent thorough cleaning, which consisted of alternating vacuum cleaning and washing In less than a week, the conctO' muons in that area had fslleo below hazardous levels, and the depanmem was out of danger Measurements indicated that vacuuming, water blasting and wiping with alcohol were suffi cient for areas with low concen trations and for painted metal surfaces. The PCBs adhered so ttrongl) to concrete floors us areas which contained electncal installations and to plastered and bnck walls that the only solution was to resurface them with epoty pauL During the first few days there was uncer tainty about the level of PCS concentration* that should be reached a the dean-up, as no official mmmm allowable concentration of PCBs in air or oo surfaces had yet been estab lished in Finland or elsewbere. The maximum allowable con centration of airborne PCBs without protection, issued three days after the incident by the fool labor protection district, was 10 pa'a1. The maximum allowable concentrations of PCBs on surfaces, which were obtained from the focal labor Cmtecooe district one month ter. wen 100 pg/m' for PCBs, 100 na'm' for hirwzu as t whole and 10 na'm1 for 2, 3, 7, t tetraehlorodibcnzoftiran. No instruc tions have bees issued as to how samples from surfaces art to be taken, an omission which has prompted some debate. ardous substances. The Institute of Occupational Health under took to measure the PCB con centrations, and the Chemical instrument Center of the Uni term) of Helsinki the furan concentrations The extent of the meisurements performed is shown b) the fact that the num ber of PCB concentrations anaIvzed amounted to twentv and the number of individual sam ples taken went far be>ond 200. The clean-up operanons. which were started without de* I). c-e ;\rfarmcd t?) In Out side com pin). The immediate problem was to train the clean up personnel. The importance of slue protection and the use of self-contained breathing ap paratus were emphasized The clean-up work focused first on the area where the pulp tanks were located because they are of crucial importance to the operation of the board ma chines The operation of this department requires the contin ual presence of personnel, whereas occasional work on electncal installations can well WHAT REALLY HAPPENED* Capacitor bank explosions do eoi occur evetyda) Fuse bums in elements uside capacitors are not rare, but an explosion which destroy ao enure capac itor bank, requires the presence of several simultaneous factors At KaukopU these factors may have included increased use or thyristors with the acquisition of a new paper machine, excep tions! distribution in the electn cal network and a long penod HONS 213975 of warm weiihcr. The com bined e(Tecl of these factors could have set off a number of element fuse bum The unbal ance current protection deuce was defeaive. and ai a result, one capacitor unit blew up Be fore the overcunent relay was able io disconnect the supply of electricity to the defective ca pacitor bank, voltage had in creased considerably as a result of harmonics This caused a chain reaction that destroyed yet other capacitor units. The capacitor bank consists of 96 capacitor units Two of them were desiroved by the electnc arc, another two units ere destroyed mechanically, and a further 11 units were da mated by pressure. The over current relays cut off the supply of electricity to the capacitor bank two seconds after the elec tnc arc i Tbe incident immediately stopped the operation of three board machines, one paper ma chine. an external pulper station and some machinery used ut the conversion of paperboard. Because of the contamination in the area housini the pulp tanks, pulp production had to be halted the day after the acci dent. This led to considerable losses m production. T wo board machines were back tn operation within three days thanks to a by-pass ar rangement. Other affected pro duction machinery was operat ing at full capacity within seven days of the explosion. KATIOSUTDE COSCEFS As soon as investigations had indicated that a risk of similar accidents exists wherever capac itors and transformers contain ing PCBs are used, the matter became a topic of national con cern A quick survey showed that there are nearly 100.000 PCB-filled capacitors and 250 PCB-fillcd transformers in Fin land At the moment, the situation is under control An ample supply of guidance material has been issued by both insurance companies and manufacturers. Public discussion in the afierci3.-, of ;hc ccc dent bas been A Guide to Loss Prevention Tbe scope of industrial safety in Finland bas expanded at a fast rate Consequently, the number of laws, statutes, regu lations and instructions, and the volume of research have grown steadily As a result, even safety professiooals have found it bard io keep up with tbe ever-in creasing wealth of reference ma terial. The situatioo has been partic ularly difficult for plant design ers who art expected to design safe plants, machinery, and equipment but who seldom know when to find the relevant information. Years ago, the participants at a safety meeting sponsored by Industrial Insurance expressed the wish that Industrial Insur ance compile a list of laws, stat utes, rtfulauons, and instruc tions pertaining to industrial safety. The decision was taken to pool efforts and resources ana, under tbe leadership of In dustrial Insurance, to compile extensive, and the participants have included numbers of pri vate organizations and public bodies Tbe media have spared no effort tn covering issues invol' uig PCBs If there are any people still w orlung in electrical engineering, Gre protection, or industrial safety today who are unaware of the discussion on PCBs, they must be deaf and blind. Regardless of the intensity of the discussion, the approach io the issue has been basically sound Various committees have investigated nsks involved in the use of equipment con taining PCBs and 'he reduction of such nsks The Labor Protec tion Fund has received grant applications which contain pro posals for dev eloping ways to measure PCB concentrations, the concentrations of ns reac tion products, and investigation of iheir effects In summary, one can say ihai the warm Au gust mg-hi shook ih: nanon a guidebook containing tbe principal sources of information on industrial safety and fire pro tection. Two worlong groups were set up for tbe task: one for industrial safety, tbe other for Gre protection. PROTECT!OS CASSOTBE REDUCED* Perhaps the most difficult task to be faced in accidents of this type is how to motivate staff io use protective wear and equipment This problem is en hanced when ihe substance against which protection is re quired cannot be seen, smelled, or fell. At Kaukopkl, familiarity wuh protection required in the use of chlonne bleaching turned out to be helpful Because of the nik of accidental chlonne emis sions. the mill has to maintain sufficient stock of protective wear, self-contained breathing apparatus and gas masks Em ployees must also be trained io wear protective clothing on the job. The outside clean-up per sonnel had to be given this training and they required fre quent forceful reminding sboui ihc imporance of pc'sa-;l fro- HONS 213976 Since the firsi edition of ihe Guide in 1977, iwo supplement* have been issued, in 1973 and I960 It was initially agreed tbat a rented edition should be pub lished five yean later, tn 1932. The bulk of the practical revi sion as in the bands of loti prevention experts from Indus trial Insurance with the help of a supporting organization cootitling of active safety and fire protection professionals repre senting Industrial Insurance's policy holders Information on each subject is subdivided into three sec uons. section A Lists laws, sa lutes, government decisions, regulations, and instructions: secuon B lists Finnish Stan dards, dau sheets issued by in surance companies, and other similar items, secuon C lists lec tures, articles tn journals, short reports, and studies published outside Finland (e g NSC Data Sheets) Although the compilation aims at practicality and simplic ity, as is reflected ui the choice of entries, the formulation of each entry, and the disposition of the Guide, Industrial Insur ance recommends a short intro ductory meeting with the users of the Guide The charge to pol icyholders for the Guide covers only the pnnung costs The Guide is available to non-poh- cyholder enterprises, orgamza- uons, and individuals at a - slightly higher pnee. O A Plant Designer 's Guide to Loss Pretention is a source book, a bibliographic work of reference listing relevant items oflegjslation. statutes, govern ment decisions, regulations and instructions Loss prevention it divided into four subficlds -- in dust nil safety -- fire protection -- environmental protection -- civil defense, The same major headings art used in the tndustnal safety and fire protection sections. Letter symbols (P* Fire Protection, T Industrial Safely) keep infor mation in the two subfields dis tinct from each other A sample page from the Guide. lection The main point in pro tection against PCBs and us reaction products is the protec tion of skin and the respiratory organs Great care must also be taken not to spread the toxic substances to non-coniaminaied areas Disposable overalls and hoods, rubber boots, plastic gloves and disposable footwear covers turned out to be handy banc protective wear. Disposa ble overalls were sufficient alone in low concentrations, while higher concentrations re quired additional protective overalls developed for civil de fense use The disposable hood |ave good protection to hair and nrck and the filer-equipped facial mask could also be worn under it Immediately after use the disposable overalls hoods, gtov es and footwear covers were pia:ed m special trash bags for hazardous w asie The light pro* icc ve overalls developed for -r.s -1 "s, i~ct `933 civil defense use, the self-conutned breathing apparatus and the gas masks were washed with the utmost care before reuse. The extent of the Kaukopha in cident is also demonstrated by the following statistics on the materials required in the clean up, measuring, inspection and related work Item used respiratory filter* disposable overalls protective hoods footwear covers 3,300 4,900 1.200 4,100 disposable gloves 2,800 glass jars (for samples) 600 foam sealant 129kg absol absorbent 4,000 civil defense overalls in use at one ume (in laundry about 330 93 pain) compressed air cylin- 2,000 ders (filled) paint for resurfacing 2,3001 indicated above, the number of people with possible exposure to PCBs was about 150, As no medical treatment was Lnow-n for such exposure, medicinal carbon was prescribed. If uken in large amounts, it speeds up the removal of toxic substances from the body. The amount of medicinal carbon taken was considerable, as the daily dose for each exposed person was 50 g for one week In addition to the ten people exposed to PCBs and taken to the hospital, five other exposed employees showed some similar symptoms of exposure soon af ter the accident This group of 15 has been under closer super vision So far analyses of blood, fai tissue and other samples have not revealed anything alarming Follow-up studies of this group are still in progress O Folfow-up examinations have shown no cause for alarm As HONS 213977 13 CAPACITOR FIR IN A PAPER MILL AT IMATRA, FINLAND O.A.Elo and P. Vuojolahti I, INTRODUCTION 1. Early stages After two consecutive and strong explosions in the electricity room, the control room of the pulp null storage department sounded an alarm at 12:10 AM on 3 August 1982. The control room became filled with smoke, which caused the shift supervisor, the storage department attendant, and their assistants to leave the room. When the fire brigade arrived, the formation of smoke was abundant. After some thrity minutes of investigation, the explosion was found to have taken place in the capacitor bank situated behind the 10 kV contactor station. The damaged capacitor caused all three board machines and paper machine No. 8 to stop almost immediately. 2, Persons immediately exposed The capacitors contain a liquid known as polychlorinated biphenyls (PCS). PBC is used both as an insulating and a cooling agent. An electric arc caused decompose. One substance given off was hydrochloric acid, which causes acute irritation. Immediately after the explosion, it was not known that the damaged parts contained PCB. Thus respirators were not used. Only after some 15 people had symptoms of irritation did it became obvious that toxic gases were spreading as a consequence of the explosion. The most common symptoms reported by this group were headache, smarting and redness of the eyes, nausea, irritation of the throat and respiratory tract, and itching and smarting of the MGNS 213978 2 skin. Due to these symptoms, ten people were taken to the out patient department of the Imatra District Hospital some two hours after the explosion. The symptoms resolved slowly in fresh air. About five hours after the explosion, the ten most heavily exposed were interviewed at the company health service center in Vuoksenniska Blood and urine samples were taken m the laboratory the next morning Within 24 to 48 hours all ten reported that the symptoms had slowly disappeared. As investigations proved that not only PCB but also dibenrofurans had been given off by the explosion, it was decided that the exposed persons should be carefully examined and followed. The test program arranged for this purpose is based on the findings of both foreign and Finnish specialists; also considered were the views of the physicians who participated in the work. 2. Persons exposed later Since at the initial stages there was no knowledge of the presence or the effects of PCB or its disintegration products, many members of the mill staff moved about in the polluted areas without the proper protective equipment or clothing. Some of the people who thus became exposed had come to restart the machines and operations. Also, despite the clear smell of gas in the area, people moved about and inspected the affected area without protective clothing. Some of the exposed were merely people who out of curiosity wanted to get as close to the explosion area as possible even though posters prohibiting entry had been erected at an early stage. The exposed also included one person who was there to take samples for indus trial hygienic measurements and others who were cleaning up. Al together some ISO people were thus slighly exposed. All of them have been interviewed and records made of the places and times at which they were in the areas. In addition, blood samples have been collated from all of them for later determination of the levels of Similar tests will be carried out simultaneously with a control group of about 200 people m order to determine the present level of PCB in' the blood of a normal Finnish population. HONS 213979 4. Examination and treatment of the exposed. in this study those immediately and more heavily exposed form group At and those later and less heavily exposed comprise group B. Group A was initially examined at intervals of a few days and later at about monthly intervals. Group B was interviewed and examined during the early stages. No clinical follow-up tests have been carried out for group B. In future, tests to determine the levels of PCB m the blood will be carried out; decisions about further measures depend on their results. There is no treatment recommended for this sort of exposure. Animal tests have indicated that large doses of carbon tablets during the acute stages could help clear the substances from the body. Thus every exposed person was given 50 g of carbon daily for an average of one week. 5. what is thought to have happened Harmonics, the exceptional connection positions of the network, and possibly also the hot weather - which had lasted for a long period blew a number of fuses in the capacitors. The imbalance relay was not operational, and the over-current relay, due to its high setting, does not yet operate in a case like this; therefore one of the units "exploded". The explosion caused the impedances of the network to move closer to resonance condition, and the harmonics caused the voltage to increase. This situation led to the destruction of other units when the capacitors went into resonance condition. A total of eleven units were destroyed by pressure, two units were destroyed by external electric arc, and two more units were destroyed mechanically. It is estimated that altogether about 100 litres of PCB leaked out from the destroyed units. The over-current relays .tripped the capacitors from the network two seconds after the electric arc. MONS 213980 4 . Initiation of investigations Some three to four hours after the explosion, the Labor Protection Manager was asked to come to the mill. He called an industrial hygienic expert from the Lappeenranta District Occupational Health Center to test the concentrations of gas. MD P. Vuojolahti was alerted at about 4:00 AM to organize the required medical measures. At 7:00 AM a conference was held at the KaukopiiS mill in which he explained the toxicity of PCS and the possibility of the presence of the even more dangerous pyrosynthetic derivatives. We decided to call in specialists to carry out investigations, first from the Industrial Insurance Company and a representative of Oy Nokia Ab. The first samples of soot dust were flown to the Department of Chemistry of the University of Helsinki m the after noon of 3 August 1982. That same evening we were informed that the samples analyzed contained furan. Thus,' that day the Industrial Insurance Company immediately contacted specialist institutions in Europe to collate the information available from reports on similar accidents or investigations on PCB derivatives, products. On 4 August 1982 Prof. Lindberg from the University of Helsinki arrived at the scene of the accident. At his instructions, the systematic collection of floor samples was started in the mill departments thought to have been polluted. On 5 August Dr. Reggiani from Hoffman La Roche in Basel, Switzerland arrived at the mill. He was known to have extensive experience in the Investigation of PCB derivatives and the clearing up of polluted areas* The Institute of Occupational Health participated in the investiga tions during the initial stages by measuring the concentrations of PCB in the air and, later, in dust samples. The University of Helsinki concentrated on the collection of samples of PCB dismtegraderivatives and the analysis of the results of soot dust tests. HONS 213981 5 7. Effects of the explosion The explosion took place in the board null capacitor bank behind the 10 kv contactor station. The pressure wave, accompanied by clouds of soot and gas, went mostly through an open doorway into the 10 kv contactor station and further into the adjacent 3 kv contactor station. The filter and control rooms of the storage department are located directly above capacitor room and the neighboring room. The cloud of soot and gas entered these rooms through openings in the floor; the openings include three uncovered cable holes, one of which led underneath the floor level of the instrument cross-connection room situated under the control room. The pollued area at the floor level of the storage department was limited to the surroundings of the control room and the end of the floor. Some traces of PCB were also discovered on the story below the floor level of the storage department; these were, however, considerably lower than the concentrations detected at the floor level. The gases and soot were not able to penetrate into the board and paper machine halls. Once the ele,ctric power had been cut off, it was no longer possible to run the three board machines and two paper machines, the outer pulper station, and the Kamyr pick-up machine. After the pulp storage tanks had been filled, the pulp production line also had to be stopped. HONS 213962 6 8. Cleaning and other follow-up measures 8.1. General Depending on the concentration of PCB, the cleaners used different kinds of protective clothing: V5S suits and compressed air respira tors; disposable protective clothing with filter mask; or disposable clothing that included gloves and head protection. The vacuum cleaners used were all fitted with microfilters. Once collected all the waste was packed in metal containers or specially made plywood cases. 8.2. Control room and cross-connection rooms The control room and cross-connection rooms were vacuum cleaned several times with microfilter-equipped vacuum cleaners and wiped several times with methylene chloride and methyl alcohol. 8.3. Switching station and adjacent rooms The switching station and the adjacent rooms were vacuum cleaned as above, wiped with methyl alcohol, and high-pressure washed with warm water. The walls and the ceiling were painted twice over with epoxide pitch,and the floors received a coating of epoxide, 8.4. Room of the explosion The floor, an area of approx. 30 m*, was pick dressed and the plaster was removed from the walls. It was vacuum cleaned several times (as in 6.1.) and wiped several times with methyl alcohol. The floor and walls were ground. The floor underwent the following treatment: moisture insulation, new concrete, and epoxide paint. The walls and celling received double coatings of epoxide pitch. HONS 213983 7 8.5. Mill area The storage department was vacuum cleaned and washed with water two or three times. The Ka 4 beater department was vacuum cleaned, washed with water, and partly wiped with methyl alcohol. The room adjacent to the control room was vacuum cleaned and washed with water several times. The surroundings of the bobbin machine were vacuum cleaned and partly wet-wiped. 8.5. Electrical and other appliances in the various rooms These were vacuum cleaned with microfilter-equipped vacuum cleaners several times and wet-wiped with methylene chloride and methyl alcohol several times* 8.7. Special equipment and supplies for these operations The following list of items was used: 3,300 respirators 4,900 overalls of non-woven material 1,200 protective hoods 4,100 shoe covers 2,800 disposable gloves 40 ordinary overalls 45 pairs of rubber boots 35 raincoats and hats 600 glass jars 120 kg of seaming foam 4,000 kg of ABSOL absorption agent 93 VSS suits, in circulation (350 washes) compressed air i 4 1 x 200 BAR: 1,087 units at 3.8 - 12.8 1,195 units at 13.8 - 18,1 ' c. 2,000 litres of paint HONS 213984 8 8.8. Cleaning waste All cleaning waste, a total of about 100 tons, was first packed in closed casks or boxes which were stored m the vicinity of the mill, either in bundles or in containers. The waste materials were later sent to be destroyed in England. 9. Results of the measurements at KaukopHI Mills 9.1. Concentrations of PCB measured in air samples The first measurements for the levels of PCB in the air were carried out about five and a half hours after the explosion,at a place close to pillar B 2 in the storage department. An open cable channel leads to this site from the condenser room. The concentration of PCB in the air at this site was 16,000 pg/m* . A similar test was done in the capacitor room itself; the result was 8,000 pg/ra'. (This latter figure can hardly be a correct reading.) A sample was taken in the control room of the storage department only at 4:00 PM on the day of the explosion; the finding was 7? pg/rn'. By this time the level of PCB in the air close to pillar B 2 had gone down to 80 pg/m1. Further tests were carried out in the capacitor room only on 12 August_1982, and the reading was still high, 1,600 pg/m1. The electricity department was later fitted with special ventilation, for which suction fans equipped with activated carbon filters were used. By 6 August *1982 the level of PCB was below the critical value of 10 pg/m* in areas other than the electric powfer areas (i.e., the capacitor room, the 10 kV and 3 kV contactor stations, and at point B 4 on the second floor). By 18 August 1982 the concentration in the 3 kV contactor station had dropped below the critical level. On 29 September 1982 the concentrations recorded in the capacitor room and the 10 kV contactor station still exceeded the critical level. Only on 12 November 1982 was the concentration in the capacitor room below the critical level; at that time PCB levels of 4.9-6 pg/m' were measured. MONS 213965 9 Mechanical ventilation had been arranged for the electric power area* by the use of temporary pipe constructions. The ventilation system was fitted with activated carbon filters; these reduced the level of PCS' by 80 - 90 t. 9.2. Determination of PCB and pyrosynthetic products on surface samples At the time of the accident it was not possibleto measure the con centrations of PCB on dust samples in Finland. The Department of Chemistry of the University of Helsinki carried out tests to deter mine the contents of furan from 3 August 1982 onwards. Samples taken from the 10 kV contactor station on the day of the accident con tained 10,000 ng/m' of furan. On 4 August 1982 samples were collected from the storage department; the lowest findings were about 10 ng/m*, the highest 1,500 - 3,600 ng/m'. In the electric power area, in the 3 kV contactor station, the levels detectedwere 508 - 630 ng/m'. Tests were carried out m the capacitor room on 11 August, when the concentration on the floor varied from 11,600 to 25,900 ng/m' and on the walls from 209 to 576 ng/m1 . Dust samples were also sent for analysis to Sweden, to Prof. Rabbe of the Department of Organic Chemistry of the University of UmeA. These analyses showed concentrations of furans at varying degrees of chlorination, biphenyls and concentrations of tetra- and pentachlordibenzodioxins about tenfold less than the levels of furans. . ' After the explosion at KaukopaS, the Institute of Occupational Health in Helsinki began to develop methods to determine the levels of PCB and its pyrosynthetics in dust and soot samples. However, MONS 213966 10 such analyses were done on dust samples taken on 24 August only, by which time a fair amount of cleaning had already taken place. Analysis of soot samples taken in the capacitor room after the explosion showed that some dibenzofurans at various degrees of chlorination had been created m the explosion. Sample No. 1 con* tamed about 20 |ig/g of tetrachloride benzofurans, and isomers 2, 3, 7, 8 2 pg/g. -Soot sample No. 2 con tained about 90 ug/g tetrachloride benzofurans, and 6,5 pg/g isomers 2, 3, 7, 8. . Traces of polychlorinated pyrenes and polychlorinated biphenyls were detected in the soot samples. No dioxin was found in the samples; in this respect the results of the analyses carried out by the Institute of Occupational Health differed from those reported by Prof. Rabbe. 9.3. Concentrations in mill products The concentrations of PCB in mill products were also measured. The tests proved that the content of PCB in paperboards did not exceed the critical value. The levels of PCB in clean washing water and used washing water were also recorded. II. EXAMINATIONS OF PERSONS EXPOSED TO PCB AND ITS PYROSYNTHETICS 1. Composition of the group and data on their symptoms This report'deals with the examinations of the people who comprise group A, i.e., those with the heaviest exposure. The group contains the storage department attendants and supervisors on duty at the time of the accident, members of the fire brigade, and labor pro tection personnel,_a total of 16 persons thought on the basis of HONS 213987 11 of the interviews to have been the most heavily group contain* only men. exposed. The Of the acute symptoms, 63 * had headache, 63 * experienced smarting of the eyes, and 50 % reported irritation of the throat and res piratory tract. Nausea, dizziness, increased sweating, and irrita tion of the skin were markedly less common. The acute symptoms of all the exposed persons disappeared within 24 to 48 hours. Table 4 presents the results of seven month* of follow-up with regard to diagnosed diseases or changes in health status among the exposed group. The more unusual conditions observed include ridges in the nails of two persons and reduced immunity to infections among two persons. The acne of an exposed person worsened visibly and required a dermatologist's treatment. Nine had lesser skin symptoms such as flaking of the face skin, redness, itching, and pimples. One of them also had increased pigmentation of the neck skin. Immediately after the explosion one person had an allergic-type, itchy rash on the upper part of his body that lasted for five days and disappeared by itself; this does not fit the picture of chlorine- induced acne. There was not one case of chlorine-induced acne. The association of these mild skin symptoms with PCB is questionable. The exposed were requested to check for any possible skin changes, and the reporting of milder changes can merely be the result of close, active observation. ' HONS 213988 12 2. Measurements of PCB levels 2.1. Blood and its subfractions Some of the blood samples taken on the day of the axplosion were sent for analysis to Sweden, to Prof. Rabbe of the Department of Organic Chemistry of the University of Umei, On IS August 1982 we were informed by telephone that no PCDP could be detected or that the concentrations were below the detection limit of 0.IS ppt 0.15 pg/g. No dioxins were found, either. 2.2. Adipose tissue Biopsies of adipose tissue have been taken from some of the most heavily exposed. These samples also await analysis by the Institute of Occupational Health. 2.3. Urine Urine samples have been collected from the group with the heaviest exposure. The samples are now being stored at the Institute of Occupational Health in the event of future analyse*. 3. Measurements to reveal effects 3.1. Hematologic measurements Samples hav* been taken to analyze the levels of enzyme activities in the red blood cells. These determinations have not yet been done. HONS 213989 n Typical blood tests (for La, Hb, Hkr, MCHC, and leuk ) and differential counts of the leukocytes were followed for a period of six months. Patient No. 3 has continually had elevated La levels that range between 26 and 37s the reason ia rheumatic arthritis. Patient No. 6 has also had moderately elevated La levels (between 12 and 22), with no clear reason for the elevation. Other have had mild fluctuations above the La level considered normal (i.e., 10), but these increases do not appear to be a result of exposure. For the differential oamts of the leukocytes, an unusual finding was that the proportion of lymphocytes exceeded 40 % in 43 of the 96 analyzed samples, i.e., in half the samples. Fifteen of the sixteen studied had at least one elevated level of lymphocytes. A similar finding of elevated lymphocyte levels has been reported in Japan for patients with Yusho disease. No clear anemic trend was observed. A slightly increased percentual distribution of eosinophiles was found in 18 of the 86 samples; these increases are probably not connected with the exposure. The blood samples were analyzed at the company health service center in Vuoksenniska. 3.2. Biochemical analyses Liver functions were followed by analyses of enzymes (Gamma-Gt, AFOS. ALAT, and ASAT) over a period of six months. The determina tions were made by the Institute of occupational Health in Helsinki. The reference values were: 40 U/l for Gamma-Gt; 35 U/l for S-ASAT and S-ALAT; and 60 - 270 U/l for AFOS. The following values in excess of the reference values were found: Gamma-GT, 35 of 95 values ASAT, 7 of 95 values; ALAT, 22 of 95 values; and AFOS, 10 of 95 values, pathological enzyme levels were detected at least once for twelve of the exposed, and four persons had completely normal enzyme levels at each examination. Figures 1-4 show the changes in enzymes activities for the individuals as a function of time. HONS 213990 As seen in the figure, about two weeks after the explosion, there was a slighlty rising trend in the activities of Gamma-Gt, ASAT, and ALAT enzymes at group level. N6 similar trend was detected for the entire group at later stages; rather, variations probably had other causes. In some cases alcohol may have been a factor con founding the initial group trend of increased liver enzyme activities. An ASAT/ALAT ratio greaterthan 1 is highly indicative of increased liver enzyme activity'caused by alcohol. Of the eleven in the group for whom samples were analyzed for enzyme activities after two weeks of follow-up, seven had an ASAT/ALAT ratio under L and four had a ratio over 1. The latter had consumed slight amounts of alcohol a couple of days before the samples were collected; the liver enzyme activities of two of them did not exceed the normal limits. Thus the slight increase in liver enzyme activities detected at group level dan at least in part be interpreted as an effect of exposure to PCB and its pyrosynthetics. No essential change was noted in the levels of AFOS. Serum proteins and their electrophoreisis were also followed. The reference values for proteins were 65 - 80 g/1. The determina tions were done by United Ofhtyneet) Laboratories. Protein levels above the reference values were found in 20 of 99 samples. The increases were only slight and most likely due to slight dehydra tion caused by fasting. Ten of 98 electrophoreisic values deviated from the reference values. No specific changes were dis covered. Hypoalbuminenia was not detected at all. Thus the levels of exposure in question did not bring about noticeable effects in serum proteins. HONS 213991 15 TO follow the fatty "meYabolisn the levels of cholesterol, trigyicerides, and the electrophoraisis of lipoproteins were measured frcm the serum. A clear increasing tendency was noted for triglicerides in the serum. For many of the exposed, the peak was reached a week after exposure: the peak was reached two weeks after exposure for the remainder. Four weeks after exposure the baseline values were attained again. The baseline value for the ten exposed persons from whom all four samples were collected was 1.39 mmol/1. For this group the mean peak level of triglycerides was 2.14 mmol/1; their mean level at four weeks was 1,13 mmol/1. At group level a 54 t increase was found (Figure 5). A clear decreasing tendency in the level of cholesterol was observed for the same group. The lowest levels were attained, again, at either one or two weeks after exposure. The mean baseline value was 6.25 mmol/1, the mean of the lowest values was 4.62 mmol/1, and four weeks after exposure it was 5,69 mmol/1. Compared with its baseline value, a mean drop of 26 % occutred in the level of cholesterol (Figure 6). The electrophoraisis of the lipoproteins was followed. Electrophoreisic measurements of the samples collected the first few days after the explosion could not be done because the samples had been stored in deep freeze. The first electrophoreisic analyses were carried out for the samples taken two weeks after the explosion, at a time when the peak values for trigylceridee had already pasted. In electrophoraisis triglicerides ere transported in prev?-fractions. HONS 213992 16 Increased levels of pre-j$-fractions were detected as follows: 2 weeks 4-5 weeks 12 weeks 3/12 1/15 3/16 18-19 weeks 26 weeks 4/16 5/16 Otherwise the electrophorelsis of the lipoproteins was normal. The results for these samples do not allow any conclusions to be drawn. lectrophoreisic analyses at earlier stages after exposure would probably have provided more information. At least sobie of the changes detected in the fatty metabolism were an effect of the exposure. Alcohol may be a confounding factor that accounts for some of the changes. 3.3 Cytogenic determinations It is known that chlorinated dibenzodioxins and dibenzofurans are mutagenic. Therefore chromosomal aberrations and sister chromatid exchanges were studied in the lymphocytes of ten persons in the group with the- heaviest exposure. A reference group was composed that contained a group of workers with similar smoking habits from the Tainionkoski factories. Chromosomal aberrations reflect damage that can lead to the rise of somatic chromosomal mutations. The analysis of sister chromatid exchanges reveals changes in the organization of the chromosomes within the DNA strands. Many substances known to be mutagenic cause increased frequencies of sister chromatid exchanges in cell cultures. The results should be examined at the group level only, as the meaning of intraindi vidual variations is not yet clear. Compared with the reference group, the exposed did not have statis tically significantly more chromosomal aberrations. Smokers in HONS 213993 n these groups did not have more chromosomal aberrations than non smokers. The frequencies of sister chromatid exchanges among the exposed were not higher than the frequencies among the referents. The frequencies were somewhat higher among the smokers, but the difference from the nonsmokers was not statistically significant. The findings indicate that the exposure to mutagenic agents was not heavy enough to induce an increase in chromosomal damage. Appendix 8 presents the data in detail. 3.4. Immunologic tests 3.4.1. Tmmunoglobins and complements The level* of immunoalectrop'horaisis and complements |_C^, C3> and CI-INH CHane method)*! were followed in the exposed. The determinations were made at the Clinical Inrunological Laboratories. The samples were taken one month (ten persons) and six months (sixteen persons) after exposure. The immunoeleCtrophorelsic values were norrral except for one sarple taken six months after exposure* in this sample a polyclonal increase was detected in IGM. Five of the samples contained amounts of complements that deviated from the normal values. The deviations were usually values above the normal range. No sign of deficiencies was detected in the complements. It can thus be concluded that the exposure did not alter either the complement system or the immunoglobins. 3.4.2. Cell immunologic tests These were done at the Transplantation Laboratory of the University of Helsinki, by Drs. E. Taskinen (D. Med. Sc.) and R. Renkonen (D. Med. Sc.). The mononuclear cells were isolated by the method of Ficoll-Xsopaqua (FIP) . In addition to overall and differential cell counts, the staphylococcus technique and monoclonal antibodies were used to study the following subclasses of T-lymphocytes: T cells (OXT-11); T-helper cells (0KT-4)i and T-suppressor cells (OKT-8). Surface and/or intracytoplasmic immunoglobinpositive tests were used to study B cells. Furthermore, the follow ing mitogens were used for stimulation tests of the lymphocyt st HONS 213994* 18 phytohemagglutin CPHA); concanavall A (ConA); and pokeweed mitogen (PWM). Samples were taken from ten persons five weeks after the explosion. Since the results contained many pathologic findings, control samples were taken from six persons twelve weeks after the explosion and, again, from five persons six months after the explosion. Results The FI? isolation method yielded a range in the number of cells from 9 x 10 to 58 x 10; the means were 33.8 x 10 five weeks after the explosion and 24 x 10 twelve weeks after the explosion. The differential cell count revealed that three of the ten samples had more than 10 4 large granular lymphocytes (LGL cells). In all of the samples studied (8/8), the level of T cells was low (< 35 I) five weeks after the explosion. Four of the samples taken seven weeks later had normal T cell levels. The ratio of T-helper cells to T-suppressor cells was low (<1,4) in six of seven samples taken five weeks after the explosion. The ratio was still low in two of the samples taken six months after the explosion. No decisive conclusions could be made with respect to the percentages of B cells. The results of the stimulation tests varied according to the mitogen and the time' of the test (see the table below) . HONS 213995 19 TABLE. Frequencies of reduced mitogenic responses of the mononuclear cells in the blood (< 66 % of normal control values) among persons exposed to PCB, Duration after the explosion PHA ConA PWM five weeks twelve weeks six months 4/10 1/6 0/5 3/10 5/6 1/5 5/9 2/i 0/5 Thus the responses that reflect the functions of T cells were reduced for nearly half of the exposed, end, except for PHA, the reduction was still observed twelve weeks after the exposure. Only one of the exposed had reduced ConA responses six months after the explosion. Summary Reduced percentages of T cells, a reduced ratio of T-helper/T-suppressor cells, and reduced responses of T cell functions in stimulation tests were found in the blood samples taken from persons exposed to PCB five weeks after exposure. Seven weeks later these changes had levelled off to some extent, and the percentages of T cells had become normal. Six months after exposure only two of the ten had a reduced ratio of T-helper/T-suppressor cells,The results of the stimulation tests indicated that T cell functions had normalized. The findings suggest that some temporary damage occurred in the number and functions of T cells. 3.5. Neurophysiological examinations The nerve conduction velocities were measured for the first time on 1 October 1982'and for the second time on 9 February 1903. The motor conduction velocities (MCV) were measured from the median, the ulnar, and the peroneal nerves. The distal sensory conduction velocities (dSCV) were measured from the median and the ulnar nerves, and the sensory conduction MONS 213996 20 velocities {SC/) vrere measured frcro the sural nerve in the area of the calf. The table below presents the group means and standard deviations (SO). Nerve Under 40 years old Mean SD Over 40 years old Mean SD Median, MCV, m/s Median, dSCV, m/s Ulnar, MCV, m/s Ulnar, dSCV, m/s Peroneal, MCV, m/s Sural, SCV, m/s 57.0 47.5 59.6 43.8 47.2 36.0 0.8 3.1 4.0 4.8 5.2 3.3 58.2 45.8 57.7 41.0 50.9 38.2 4.8 3.2 5.5 4.8 6.1 4.0 t The means were somewhat lower than the mean conduction velocities of referents. At group level significantly slower sensory conduction velocities were measured for the ulnar and sural nerves, and significantly slower motor conduction velocities were measured for the peroneal nerve. These findings are indicative of damage to nerve axons. The measurments were repeated about six months after the explosion for the 16 exposed. The following table shows the results of these latter measurements. Nerve Median, MCV, m/s Median, dSCV, m/s Ulnar, MCV, m/s Ulnar, dSCV, m/s peroneal, MCV, m/s Sural, SCV, m/s Under 40 years old Mean SD 60.1 54.2 58.9 45.9 -51.0 39.3 5.3 9.4 3.4 3.0 6.1 2.6 Over 40i years old Mean SD 57. 50.0 59.7 44.3 54.5 40.1 4.9 4.9 4.8 4.2 5.8 3.3 HONS 213997 21 The means for the latter measurements were higher than those for the previous measurements and were now on the same Level with the means for referents. The mean amplitudes of the sensory nerve potentials were also higher than for the previous measurments. Furthermore, all of the conduction velocity measurements fell m the normal range of variation. The measurements were taken by Docent SeppalSinen from the Institute of Occupational Health, m in whose opinion the findings indicated temporary, nild neuropathy of the sansory distal nerves in particular. HONS 213998 t CONTROL ROOM CROSS-CONNECT ICNROOfi 135 x 15 . HONS 214Q00 TABLE 1. Diagnosed diseases and Tindings during the seven-month follow-up (3 August 1982-3 March 19831 Patient s number Diseases, signs, symptoms 1 Barber's itch. Some pimples on the face. Aching Joints. 2 Acne worsened. Repeated sinus infections (treated four tines with antibiotics). One respiratory infection. Ridges in the nails. 3 One respiratory infection. Gastroenteritis once. A Redness in the nasal area two weeks after exposure. One respiratory infection. Ridges in the nails. Erysipelas. 5 One respiratory infection. 6 One respiratory infection. Flaking, itching rash on the face. T Itching rash on the arms two months after exposure. 8 One respiratory infection. Pimples on the' face, temporarily. 9 Nonspecific urethritis. 10 11 One respiratory infection. Slightly increased pigmentation of the skin on the neck. 12 Pharyngitis once. One respiratory infection. One attack of gout. 13 Four or five aphthae in the mouth. Dry skin. 1A 15 Itching, allergic-type rash on the upper body between A August and 9 August 1982. 16 One respiratory infection. Cold sores twice. ,, HONS 21*001 i Figures 1-6 mo 1 S-gamma CT rt , . M nto 5 S-trigIy*erldo8 nso 6 S-Cholesterols HONS 214002 TABLE 3 THE RELATION BETWEEN PCB-LEVELS AND OTHER FINDINGS ?CB 3.8. 6.8. gahhaGTI' AFOSt 1 3.5 2 4.1 30 4 11.2 5 5.2 6 19,9 7 48,3 8 33,4 90 10 10.0 11 0 12 0 13 36,9 14 6,9 IS 4,5 16 19,8 0 20,3 20,6 25,5 19,0 31,9 13.4 11,2 0 7.7 0 0 7.1 0 5,0 6,3 - - + + + +0 0 0 + 0 - - + + + 0 +0 0 + 0 - ALATt + + + + + + 0 0 0 + 0 + +H ASATt ,+ - + +(-) + +0 0 0 +H 0 - KolestI - + - + + + + 0 + 0 0 + ++ TriglyT Cell. Nerve Imhuh.Tesi Conduct lot X VelocityI -0+ ++ _- + - +- -+" ++ ++ -+ +? 0 +0 00 00 ++ -- + + + + +0--- ++ HONS 21*003