the radiological accident at the irradiation facility in nesvizh

8/12/2019 The Radiological Accident at the Irradiation Facility in Nesvizh

1/88


2/88


3/88

THE RADIOLOGICAL ACCIDENTAT THE IRRADIATION FACILITYIN NESVIZH


4/88


5/88

THE RADIOLOGICAL ACCIDENTAT THE IRRADIATION FACILITYIN NESVIZH

INTERNATIONAL ATOMIC ENERGY AGENCYVIENNA, 1996


6/88

THIS PUBLICATION WILL ALSO BE ISSUEDIN FRENCH, RUSSIANA N D SPANISH

VTC Library ataloguing in Publication DataTh e radiological accident at the irradiation facil ity in Nesvizh. Vienna :

InternationalAtomic Energy Agency,1996.p. ; 24 cm.

STI/PUB/1010ISBN 92-0-101396-5Includes bibliographical references.1. Nuclear facil itiesAccidentsNesvizh (Belarus). 2. Radiation

injuries. I. International Atomic Energy Agency.VICL 96-00151


7/88

FOREWORDTechnologies that make use of radiation continue to spread around the world:

millions of people are employed in radiation related occupations, and hundreds ofmillionsofpeople benefit from these applications.Use ofintense radiation sourcesfo r purposes such as sterilization of medical products requires special care in thedesign andoperation ofequipmenttoprevent radiation injury to workers.

More than 40 years ofexperiencein radiation processing has shown that suchtechnology isgenerally used safely, andsteady improvementin thedesignoffacili-ties and careful selection and training of operators have contributed to this goodsafety record. However, some cases of circumvention of safety systems have beenregistered and it is documented that the consequences of radiological accidents atindustrial radiation facilitiescan be extremely serious.

The causes of accidents may have some points in common, but at the same timemay be highly specific. A detailed study of these common and specific featuresseems to be of great importance for further improvements in safety systems.

One such event occurred on 26 October 1991 at an industrial sterilizationfacility in Nesvizh, Belarus, whentheoperatorentered theirradiation chamberandwas severely exposed to a lethal dose of radiation. The significant feature of this casewasrelatedto themedical management.Itshouldbeunderlined that some circum-stances of the accident only came to light during the post-accident review made bythe IAEA. To document the causes and consequences of the accident and to definethe lessons learned are of help to those people with responsibility for the safetyofsuchfacilities and to those medical authorities who might be involved in the manage-ment of aradiation event.


8/88

EDITORIAL NOTEAlthough great care has been taken to maintain the accuracy of information contained

in thispublication, neither the IAEA nor its Member States assume any responsibility forconsequences which may arise from its use.

The use ofparticular designations of countries orterritories does no t imply an yjudge-mentby thepublisher,theIAEA, as to thelegal status of such countrieso rterritories, of theirauthorities an d institutions or of the delimitation of their boundaries.

Th e mention of names of specific companies orproducts (whether or not indicated asregistered) does not imply an y intention to infringe proprietary rights, nor should it beconstrued as an endorsement or recommendation on the pan of the IAEA.


9/88

CONTENTS

1. INTRODUCTION 11.1. Background to the accident 11.2. Background to the IAEA post-accident review 12. IRRADIA TION FACILITY 22.1. History of the facility and regulatory control 22.2. Description of the facility 32.3. Radioactive source 32.4. Produc t transport system 62.5. Safety an d control systems 62.5 .1 . Product transport control panel 6

2.5.2. Gamm a mo nitoring probe system 72.5.3. Source movem ent an d safety mechanisms 82.5.4. Access safety features 9

2.6. Operation an d maintenance 113. RADIATION ACCIDENT 123.1. Initiating events 123.2. Entry an d exposure 123.3. Response to the acciden t 144. DOSE ESTIMATION 154.1. Phy sical reconstruction of the accident an d calculationsof

the exposure dose 154.2. Biological dose assessment 174.3. Electron spin resonance studies 184.4. Exposure factors 185. LESSONS LEAR NED 195.1. Priorities 205.2. Interaction of the transport mechanism and the source rack 205.3. Safety systems 215.4. Instructionand training of staff 235.5. Monitoring 235.6. Inspection and audits 245.7. Co-operation 24


10/88

5.8. Initiatives by the IAEA 455.8.1. Training 455.8.2. Dissemination of in fo rma t io n 45

6. M E D IC A L M A N A G E M E N T 466.1. Introduction 466.2. Clin ica l course of ARS 47

6.2.1. Initial medical managem ent (days 0-7) 476.2.2. Cl inical course (days 8-39) 506.2.3. Clinica l course (days40-70) 546.2.4. Clinica l course (days 71-101) 556.2.5. Clinical course (days 102-113) 56

6.3. M edical treatmen t and a rev iew of i ts effectiveness 566.4. Pathomorphological f indings 586.4.1. Whole body study and skin injury 58

6.4.2. Lungs 596.4.3. Gastroin testinal tract 596.4.4. Liver 606.4.5. Kidne ys 606.4.6. Adrenals 606.4.7. Testes 606.4.8. Bone marrow 606.4.9. Spleen 606.4.10. Lymphat ic nodules 606.4.11. Thyroid gland 60

6.5. Conclusions 616.6. Lessons learned 63

6.6.1. The medical community 636.6.2. Internation al collaboration (medical field) 65

A N N E X I: ESTIMATION O F WHOLE BODY DOSEFROM BLOOD CELL COUNTS 66

A N N E X II : RES ULTS OF CYTOGENETIC AN ALYSE S 68ANNEX II I : LIST O F DRUGS, DOSES A N D

ADMINIS TRATION DATES 71REF ERENCES 74CONTRIBUTORS TO DRAF TING AN D REVIEW 75


11/88

1. INTRODUCTION1.1. BACKGROUNDTO THE ACCIDENT

On26 October 1991, a radiological accident occurred in an irradiation facilityin thetownofNesvizh, about 120 kmfromM insk, Belarus.Avarietyofag riculturaland medical products aresterilizedat the facility by meansof an internal transportsystem that passes them, in a controlled manner, close to an intensely radioactive^Co source in a mov eable rack. Following a jam in the product transport system ,theoperatorenteredthefacilitytoclear thefault; atthis point,thesourceisthoughtto have been in the safe position. However, on entering the facility the operatorbypassed a number of safety features an d left the controls in a position such thatexposure was imminent.

Atsome stage, the source rack became exposed and the operator w as irradiatedfo r about1minute. He suddenly felt unwella ndthen noticed the source rackin theirradiation position. The accident was quickly reported to the authorities and theoperator was taken into medical care, first in Nesvizh an d Minsk, an d then fo rspecialized treatment in M oscow. It was estimated that he had received a who le bodydoseof 11 Gy, with localized areas of up to 20 Gy. Despite intensiv e medical treat-ment, he died 113 days later.

1.2. BA CK GRO UN D TO THE IAEA POST-ACCIDENT REVIEWTechnologies that make use of radiation c ontinue to spread around the world.

Millions of people are employed in radiation related occupations and hundreds ofmillions of people benefit from these applications. Facilities using intense radiationsourcesfor purposes such as radiotherapy, radiation processing of products, preser-vation of foodstuffs and gamma radiography require special care in the designan doperation of equipment to prevent radiation injury to workers or to the public.Experience has shown that such technology is generally used safe ly, but on occasionscontrols have been circumvented and serious radiological accidents have ensued.

To the extent that reports on such accidents are incom plete or are unav ailableto the scientific community, potentially valuable informationislost. Althoughthecauses ofacc identsand thetreatmentofpatientsmay behigh ly casespecific, reviewof the circumstances in which they h ave arisen c an y ield generally applicable lessonsthat are of help in prev enting further accidents or in imp roving the response to thoseaccidents that do occur. This was the motivation for reviews by the IAEA of threefatal accidents: inGo iania,Brazil, in 1987[1];in San Salvador, El Salvador, in 1989[2]; and at the Sor-Van facility, Israel, in 1990 [3]; the last two of these accidentstook place in irradiation fac ilities. Two other fatalacc idents hav e occurred atirradia-tors: at the Stimos plant, Italy, in 1975[4];and inKjeller, Norway , in 1982 [5]. It


12/88

is relevanttonote that there hav e also beenanumbero fac cidents involving gammaand electron beam irradiators that have resul ted inserious, but non-fatal , radiat ioninjuries [6-8].

There are more than60 0in dustria l irradiation fac ilitiesof thegamma ray andelectron beam type in use worldwide. Some of these are to be found in developingcountries, where the radiation protection infrastructure may be less strong.However ,as isevidentfrom th eabove references, accidentsare notconfinedtothoseirradiators that are in use in developing countries. The numberof fatal accidents inirradiatorshas given further impetus to the IAEA pro gram me in this are a; one aspectof this work has been the post-accident review reported on here. Three otherimportant features of the IAEA programme are:(1) Thepublicat ionof a Safety Guide[9] and anassociated set of Practical Radia-

tion Safety M anua ls [10, 11];(2) The organizat iono f a series ofregional t raining coursesand theproductionofa training video tape [12];(3 ) The holding of a Technical Comm ittee Mee ting for comp etent authorities and

the designers and operators of irradiators [13].

2. IRRADIAT ION FAC ILIT Y2.1. HISTORY OF THE FACILITY AND R E G U L A T O R Y C O N T R O L

The gam ma irradiation facility at the Biochemical Plant of the PharmindustryCorporat ion inNesv izh , about120 km from Minsk, Belarus, w as designed in 1981an d commenced operat ion in 1984; subsequent refurbishment w as completed in1989. It was built by the All-Union Scientific Research Institute of RadiationTechnology of the former USSR, and is of the 'Pepel' model type. Several suchfacilities are in use in the former USSR republics. The design and construction ofthe plant were carried outacco rding to specifications laid downby the then regula-tory authorities.

Under international classification [9] it is a category II gamma irradiator of drystorage design with an original capacity of 6.72 PBq (182 kCi), which was sub-sequently increased to 30 PBq (800 kCi) of ^Co. It was originally designed forsterilizing peat to be used infertil izers, but is capable ofdeal ing witha wide rangeof products; at the time of the accident, the irradiator was being used to sterilizemedical syringes and haemostatic sponges.

Since its commissioning, two important factors have altered the system ofregulatory control. First, the Chernobyl accident had asignificant impact on partsof Belarus and, although this focused attention onradiationprotection problems, it


13/88

diverted resources from radiation protection in the industrial sector. Second, thebreakup of the former USSR and the emergence of Belarus as a separate sovereignstate produced some changes inorganizations and their responsibilities.

Atthe time of the accident, two relevant organizations existed. The firstwasthe State Committee on the Supervision of Operational Safety in IndustrialandAtomic Power Engineering, which was responsible for thetechnical aspects of thedesign, construction and modifications of facilities. It licensed the facilities and wasalso responsible for the drawing up of the regulations that control plant operation.The second organization was the Radiation Protection Division of the Public HealthInspectorate within the Ministry of Health (Gossannadzor). It had specific responsi-bilitiesin terms of assessing the radiation doses tostaffand the means of controllingthesedoses. Visits by provincial and district inspectorate staff took place approxi-mately every 3months.

2.2. DESCRIPTION OF THE FACILITYComprehensive documentation on the plant as-built, on its condition at the time

oftheaccidentand onsubsequent alterations was notavailable to thepost-accidentreview team, and the following description is based on the original specifications,together with the observations and data presented at the time of visiting the plant.

Inthe 'Pepel'design the source, whennot in use, is stored in a shielded drystorage pit, from whichit can be raised to the irradiation position. The product tobe irradiated isrotatedat a constantratearound the source in a transport containersystem suspendedfroman overhead rail. The irradiation chamber has massive con-crete shielding, the primary barriers of the walls and ceiling being 1.8 mthick anddesigned to reduce accessible dose rates to below 14^Sv/h, a level that is specifiedin theregulations. The facilityhas amaze entranceand avariety of safetysystemsto prevent entry when the source rack isexposed.The source exposure and producttransport systems are controlled from a control room in the product loading area.Figure 1 shows a plan view of the facili ty, with the location of safety devices.

2.3. RADIOACTIVE SOURCEThebasic sourceelementsare ofRussian originand of the GIK-7-4type. They

are in the formof metallicrods,with a diameter of 11mm and a length of 81 mm,each incorporating a nominal70 TBq(1890Ci) of ^Coencapsulated in stainlesssteel. Up to 12elements, together with spacers, are loaded into64 stainless steeltubes placedin four sectionsof 16source pencils each, witha 30 mm gapbetweeneachpencil.Thesectionsareheldtogetherwith stainlesssteelstripstoformaplanarsource (2.2 m X 1.2 m), known as the source rack. At the time of the accident, thesource activitywas28.1PBq(760 kCi).


14/88

Moveable floorto coverpit Emergency'stop'button

Pressureplate

Counter-weights

Gamma monitoring probe(positionS)Drive motorforfloor Source rack

Leadshieldedgammamonitoringprobe(positionX)

Gammamonitoringprobe(positionT)

Gammamonitoringprobe(positionY)

Overheadtracksfor Emergency 'stop'buttonproduct transportsystemFIG. 1. Plan view of thefacility, with the locationof safety devices (for explanation of thegamma monitoringprobe positions, seeSection 2.5.2).

The source rackisconnectedby stainless steel cablesto tw o counterweights.The source rackan dco unterweightsrun in vertical stainless steel guides, asshownin Fig. 2 andphotographs1 and 2. In addition,the source rack isprotected againstinterference from the product boxes by a metal grid. Th e source is raised an dlowered by an electric motor drivenby gears, ahy draulic damperand twotractionchainsattached to the source rac k. D uring'sourceu p 'a nd'sourcedown'movem entsit ispossible to stopthedrive motorand to hold the source rack in anintermediateposition.In anemergency, thec lutch disengages from themotorand the source rackreturns to the safe position under gravity.

In the source pit, water cooled heat exchangers remove the radiogenic heat.In the irradiation chamber, forced ventilation removes the radiogenic heat, ozone(03) and oxides of nitrogen produced by radiolysis.


15/88

1< N

8


16/88

2 .4 . PRODU CT TRANSP ORT SYSTEMThe product to be irradiated is transported around the facility in carriages

suspended from an overhead track. Each carriage has a cradle supporting two stain-less steel boxes, each 500 mm X 560 mm x 800 mm. The cradle can be rotatedin a vertical plane orthogonal to the direction of travel sothat the top box is movedto thebottomandvice versa (photographs 3 and4). Thereare 24irradiation positionsaround th etransport track, and the product rotates around each position twice. Forth e second pass, th ecradle is turned u psidedown. Thet ime spenti neach irradiationposition can be adjusted from th e control panel to ensure that th e product dose is25 kGy (+40%).

Attached to the top of each carriage are linkage arms (photograph 5(a)) thatcouple th ecarr iages to oneanother for movement aroundth e facility. Problems canoccur with linkage at bends in the overhead track. If loading of the boxes isunbalanced, m ovem ent of the carr iage can result in decoup ling of the l inkage (photo-graph 5(b))and the product transport system jams. Thiscanoccur up to three timesa day and necessitates the operator enter ing the i r radiat ion chamber, f i rs t to leverthe jam m ed c ouplings back into place with a metal rod (photograph6), and then torecouple th e units.

2.5. SAFETY AND CONTROL SYSTEMS

In th econtrol roomthere are three adjacent control panels: th eproduct t rans-port control panel; th e gamma monitor ing probe system; and the source movementan d safety mechanisms. These are described in Sections 2.5.1 to2.5.3, and theirlocation in the facility is identified in Fig. 3.2.5.1. Product transport control panel

The system w as originally designed with a microswi tch at each irradiationposition to sense the presence of a carriage and to g ive a conf i rmatory i lluminatedsignal for each position at the control panel. However, since decouplings have takenplace at bends, only th e sensors at each end of the four irradiation rows (i .e. eightsensors in total) are now used. A transport jam is signalled when one of themicroswitches isno t i l luminatedon the control panel.There is no audible alarm toindicate such an event, the o nly indication being the chan ge in the sound of the drivemotor when i t is unable to m ove the product t ransport system. To protect the motorthereis a therm al cut-out mec hanism, wh ich only operates after some t ime and whichis indicated on the control panel . The noise of the dr ive motor cannot readi ly beheard in the control room, thus vigilance over the visible signals (from a productdose viewpoint ) is important . Figure 4 shows th e product transport control panel.


17/88

Irradiator Ventilation ConveyorUp |[irradiation]|Down Ready Dose Storage On On [Entrance

Up Stop DownOnOf f

Mode |amrv(manual/ LamPautomatic) control

Start Stop

Key

Emergencydropping

Power

FIG. 3. Source control panel.

There is no interlock between the product transport system and the sourcecontrol system, therefore in theevent of a jam 'sourcedown' movementhas to bemanually selected at thecontrol console.2.5.2. Gamma monitoring probe system

Twopairsof fixedgamma monitoring probesareinstalledat thefacili ty; theirlocation is shown in Fig. 1:(1) The first pair of probes focuses on the inside of the facilityand comprises a

probe in a lead shield (photograph 7), position X, inside the irradiationchamber, and an unshielded probe, position Y, on the inner leg of the mazeentrance;

(2) The second pair of probes focuses on the dose rates outside the facility andcomprises a probe, position S, at the beginning of the maze entrance, and aprobe, position T, by the drive motor near to where the source movementsystem enters the exposure chamber.Eachpair of probes has a single ratemeter (i.e. two in total) on a panel in the

control room and the operator can choose which probe signal is to be displayed(photograph 8). However, it is not immediately clear which readings are beingdisplayed, hence misinterpretation of the signals can easily occur.


18/88

Charge Rearrangement Discharge

Accumulator2

Irradiator Platform Conveyor Stopelectromagnets Power

Lampcontrol

FIG. 4, Product transport controlpanel.

An interlock arrangement exists that activates the emergency source returnsequence if the dose rate f rom probes S or T exceeds 14^Sv/h.However, the X andYgamm a m onitoring probesdo nothave systemstow arn persons insidethe facilitythat th e source rack isexposed, nor arethey interlocked with other control systemsto provide an automatic safety feature that prohibits entry when the source rack isexposed.2.5.3. Source movem ent and safety mechanisms

Figure3showsthelayoutof thesource control panelin thecontrol room. Thedesign is such that a single key has to be inserted and turned 90 to energize the con-trols. Removalof the keyautom atically de-energizes thecontrols. The samekey has


19/88

tobeusedin a separate control panelat themaze entrance duringthe entry proce-dure.Themaincontrol panelhasbuttonstoraiseandlowerthesource, togetherwitha 'stop'buttonthat holdsthesourcein anintermediate position, withwarning lightsto indicate itsposition.

The stainless steel shaft running through the hydraulic damper of the sourcedrive mechanismhas alarger diameter ateachendwhich, whenin the 'fully u p 'and'fully down 'positionsof thesource rack, actuates the tworespective microswitchesand sends signals to the control panel. The tolerance of these microswitches isunknown. Also, located at ground level at one end of the source rack guide is amicroswitch that senses whenthe shielding plugon top of the source rack is in the'fullydown'position. Pressing the'sourcedown 'or 'sourceup'button immediatelyresults in the lights 'down' or 'up' being illuminated. When source movement iscompleted, the microswitch signals sense the position of the source rack and thestorageor irradiation lightsare illuminated, as appropriate.

If entry is not required, the source rackcan be repeatedly cycled by raisingand lowering without an y constraints. However, if entry is required then certainsequences haveto befollowed; these bring into playanumberof safety featuresandsafety procedures, asdescribed in Section 2.5.4.2.5.4. Accesssafety features

Attheentranceto themaze thereis amotorized, moveablefloor section that,during irradiation, is driven back to reveal adeep pit that covers the full width ofthe mazeentrance.At the timeof theaccident, narrow ledges were present alongthe sides of the pit which, alone, were notwide enoughtopermit anyonetowalkacross. However, thedrive motorfor themoveable floor sectionwaslocated in thepitnear onesuch ledge and, together, they provided apotential pathway across thepit. Sincetheaccident, themotorhasbeen moved, the ledges removedand the sizeofthe pit increased (photograph 9).For access, a small control panel at the mazeentrance drives backthe floor section coveringthepit. However,asinglekey fromthemain source control panelhas to beinserted, and it cannotbe removed unlessthe floor section coveringthe pit ismoved. Thus,the floor position is interlockedwiththe source controls.

Aroundthefirstbendin themaze entrance there is alarge,pressure sensitiveplate covering th e full widthof themaze entrance. Walking over th e pressure platewiththesource exposed would resultin thesource rack being automatically loweredto the safe position.2.5.4.1. Entrance sequence

Withthe source 'fully up',the 'irradiation light' isilluminated, as is the 'noadmittance' light. To gain entry to the facility, the following sequence has to becarried out:


20/88

(1 ) Push the'down' but ton on the control panel. T his i l lum inates the ' d ow n ' lightuntil the m icroswitches sense the i r radiator is 'fully d ow n ' , at which t ime the'down' light goes out and the 'storage' light is i l lumin ated.

(2 ) The safety procedures require thatth e operator wai t4m inutes before enter ingthe irradiation chamber in order to allow the ozone and oxides of nitrogen todisperse. This t ime delay is indicated by the 'entrance' light on the controlpanel;however , thisi s aprocedural controla ndthereis nointerlock toenforcethe delay.

(3) The key should be turned from th e 'ready' to the 'off position, and thenremoved. Even i f the ' source down ' button has not already been pressed,removal of the key automatically act ivates th e 'emergency dropping ' of thesource rack to the storage position.

(4 ) The gamma monitor ing probe system should be observed toensure that th edose rates are acceptable. However , in the event of high dose rates beingrecorded there are nomechanisms to automatically prevent entry or to triggeran alarm.

(5) The operator should examine th e dose rate moni tor and conf i rm, beforeproceeding, that i t is funct ioning correct ly, using a check source attached toth e moni tor . The facility has three moni tors , which are calibrated annually.The monitors consist of a separate ratemeter and long probe (approximately50 cm), which require tw o handed operat ion. Thedetector is of the sodiumiodide type and is pr im ari ly designed for geological survey wo rk, w i th a maxi-mum dose rate of 30 /xSv/h (photograph 10).

(6) The key has to be taken to the maze entrance panel, where it is inserted andturnedto the'on'position beforethe motorized sectionof thefloora t the mazeentrance can be moved to cover the pi t .

(7) Wh ile wa iting for the floor to be positioned, th e operator should take note ofth e portable monitor reading. If the source rack is shielded it should read0.2-0.5 f tSv/h, and if fully exposed, 3-4 f iSv/h.

(8) Proceeding dow n th e maze , th eoperatorcrosses th e sensitive pressure plateand should observe the state of the warning lights and the readings on hisportable monitor.At the timeof theaccident, there were no warninglights.Should anyth ing unusual be discovered, th e emergency 'stop' but tons arepressed, causing th e irradiator to be lowered.

2.5.4.2. Exit an d exposure sequence(1 ) Having completed any necessary workin thei rradiat ion cham ber, th e operator

has to follow 'search and lock-up' procedures. A but ton has to bepressed atthe far end of the chamber, whereupon he/she has 2 minutes in which tocomplete the sequence.

10


21/88

(2) Theoperatorshould make sure that no one is in the irradiation chamber, andthenexit.

(3) At the maze entrance control pan el, the mo torizedfloor should be moved backto prevent further access.

(4) The key should be removed and taken to the main control panel, w here it isinserted and turned to the'ready' position. The ' ready' lightwill illuminate ifall thesafety systems are in a safe condition. Confusingly, at thispointin thesequence both the 'noadmittance' and 'entrance' lights are illuminated.

(5 ) Press th e 'up' button, th e 'up' light isilluminated and the 'storage', 'ready'and 'entrance' lights go out.

(6) When the source rack reaches the 'fully up ' position, as sensed by amicroswitch, the 'up' lightgoesout and the 'irradiation' light is illuminated.

(7) At this point, the conveyor system can be started.Althoughtherearesome c onnections betweenth ec ircui tryof thethree panelsinthe con trol room , none provides interlocking mec hanisms to prevent entry when

highdoseratesare detected in the irradiation c ham ber or to automatically return thesource to the safe position if a product jam isdetected. These arrangements arerequired by the standards detailed in IAEA Safety Series No. 107 [9] in order toprovide adequate levels of redundancy, diversity an d defence in depth fo r safetysystems. Lack of such arrangements places the onus on procedural controls which,as willbe seen, contributed to theaccident.

2.6. OPERATION ANDMAINTENANCEAtthe time of theaccident,the facility w as operating for 24 h ours a day , using

a 3 x 8hour shift system. Each shift consisted of oneoperatorat thecontrol paneland oneassistant toload theproduct intothecarriers. Th eirradiation facility ispartofalarge factory, and electrical and mechanical maintenancestaff are available. Toallow time off, the factory has six separate shift teams.

The qualifications required for an operator are an engineering degree or a'special technical education standard'. The State Committee on the Supervision ofOperational Safety in Industrial an d AtomicPower Engineeringprovides specifictraining on the regulatory requirements and local rules of the facility, and also setsperiodic examinationson these subjects which operators havetopass toretain theirlicence to operate. Initially, these examinationshad to be taken every 3years, butthe curren t requirement is an annual exam ination. Every 3 m onth s,operatorshaveto attenda safety meeting chaired by theu nit head.Averylowturnoverofstaff ha sbeen experienced and nearly all the operators have been employed since startup ofthe facility.

11


22/88

Regulations and local rules require that the operators and assistants wear adirect reading quartz fibre electrometer (QFE) and either a film badge or a thermo-luminescent dosimeter (TLD). Three portable doseratemonitorsof theSRP68 typeare avai lable, whicharecalibrated annually by alaboratoryat the Ministryo fHealthin Mi nsk .All thesafety systemsar epositively checked by amemberofstaff on amon thlybasis, and an external organization carries out preventive maintenance every3 months. Indeed, th e faci l i ty 's lo g book indicated that, on the day before th eaccident, such maintenanc e wo rk had been carried o ut and the standard found to beacceptable.

3. RADIATION AC C IDENT3.1. INITIATING EVENTS

The operator involve d in the accident was a 34 year old man . He had a degreein engineer ing and was the most experienced operator at the plant , having com-menced employment at the facility dur i ngit sconstruct ion. Thu s, he was completelyfamil iar wi th th e faci l i ty , the hazards and the safety systems. He was regarded asa skilled operator capableo fdeal ingwithoperat ional problemsa nd meet ing produc-tion targets.

He and an assistant were workingon n i gh tshift and the accident occurred atapproximately 03:40 hours on the morning of Saturday, 26 October 1991. A jamoccurred in the transport mechanism because of decoupling of the linkage arms oftwo carr iers , as previously shown in photograph 5. Alth oug h this could have beennoticed by c ontin ual observation o f the position indicators on the control p anel, the rewas no intrusive alarm. The operator was in the control room reading a newspaper .While loading the product into the carriers, the assistant first became aware of thejam onhear ing th e characteristic change in noise of the drive motor. Theassistantshouted to the contro l room to alert the operator, who at the time was not we aringhis TLD personal dosimeter or a QFE, both of wh i ch were in his lunch box.

3.2 . ENTRY A N D EXPOSUR EAs in many accident investigations, ambiguities exist in the description of

exactly what took place during th e accident. The operator stated that his ownrecollection of events was somewhat hazy as he was feeling sleepy at the time.However, both the authorit ies and those who treated the operator reported that healso appeared reluctant to give anyprecise details. In the following description,12


23/88

distinctions are made between the known facts, the probable actions taken and therange ofpossible options needed toproduce the end effect.

There is ahigh probability that,onbeing toldof theproduct jam,theoperatorpressed the'sourcedown' button before attempting torectifythe situation. He admit-tedthathe did nottakethe key from thecontrol paneland after theaccident it wasfound in the control panel in the'ready'position. He may haveleft it in this positionin order to speed up the procedure for restarting irradiation once the transport jamhad been cleared, i.e. circumventingthe search and lock-up procedures.

As the operator did not take the key with him, hewouldhave had to cross thepit at the maze entrance without the motorized floor section being moved to theclosedposition. Afterthe accident, no plank or board for bridging the gap was found.However, at the time of the accident the pit was narrower than shown in photo-graph 9 and the drive motor for the moveable floor was inside it. All evidence indi-cates that theoperatorcrossed the pit by stepping on to the top of the motor and thathe took theportabledoseratemonitor with him.Aftertheaccident, thismonitor wasfound at the point where the maze entrance joins the irradiation chamber, and wassubsequently confirmed to be in proper working order.The next safety feature hewould have encountered was the pressure plate in the floor, and it is difficult toenvisage how he could have avoided stepping on this. Subsequent tests showed thatthe pressure plate was in correct working order (tested 25 times) and that it wasimpossible to jump over it (its position prevented a possible run-up).

Theoretically, it would be possible, but extremely difficult, to cross thepressure plate using the product transport system. This could not have been donewhile holdingadoseratemonitor andtherewould have been noreasontoattemptsuch amove becauseof thedanger involved.

Atthis point, even if in his sleepiness the operator had forgotten to lower thesource rack at the control panel, it should automatically have returned to the safeposition. There remains a possibility that the source movement mechanism hadjammedsothatthesource rackwasexposed tosomedegree.However,theproblemappears to have rectified itself, because after the accident it was observed that thesource rack was in the 'fully shielded' position; after testing it was found to beoperat ingsatisfactorily. Thus, it can beconcluded that in allprobability th e sourcerack was in a safeposition when th e operator entered th e irradiation chamber.

Onentering th eirradiation chamber, th etrainofproduct carriers wouldhaveobscured theoperator'sview of the source rack had it been in the irradiation posi-tion. However, had he looked, he would have seen the position of the counter-weights, indicating the position of the source rack.

He walked around the back of the product transport system to where theblockage had occurred (photograph 11(a)) and tried to release the jammed carriagecouplings (photographll(b)).Afterabout 1 minute, he developed an acute headacheand pain in hisjoints and gonads. He felt generally unwell and had difficulty inbreathing. He turned his head to the leftand saw the source rack in the irradiation

13


24/88

position. He did not press th e nearby emergency 'stop' but ton , but ran out of theirradiation chamber and told the assistant that he had been irradiated.

It is agreed by all concerned that wh en he entered the i r radiat ion cham ber thecon trols were in a position such that the sourc e rack sho uld have been in the storageposi t ion. H owev er , wi th the key in the con trol panel and turne d to the 'ready' posi-t ion, exposure was imminent . This leaves the following quest ions unanswered:(1) W hat caused th e source rack to rise to the 'up' position?(2) Wh y did the operator not notice the source rack bein g raised?

Dealing w ith what caused the source to m ove to the i r radiat ion posi t ion, thismay have resulted f rom:(a) Ac cide ntal and unno ticed depression o f the exp osure button (several button s

are close together);(b) Tem porary fai lure of a component in thecontrolci rcui t , e .g. jammed contacts ;(c ) Thelogicof theelectrical circ uits(a sinstalled), together w iththeunusual com-

bination of control settingsa nd safety devices, such that power wasprovidedto the source dr ive mechanism.As to the question of why the operator did not notice the source rack being

raised, this would not have been accompanied by any audible alarms inside theirradiat ion chamber and there were no warning l ights to indicate i ts posi t ion. Thedrive mechanism f or mov i ngth esource rack wo uld cer tainly h ave madeanoise, butthis may have been masked by the noise from the ventilation system and the producttransport system (if he had failed to stop it). In addit ion, he may have been in themaze entrance at thet ime andtherefore been unabletoclearly hear th emechan i sm.

While i t has not been possible to accurately chronicle th e sequence ofeventsleading to the irradiation, i t is clear that:(i ) The specified safe operating procedures were not follow ed and the safety

features were ci rcumvented;(ii) The safety features could be improved to prevent ci rcumvent ion (see

Section5).

3.3. RESPONSETO THE ACCIDENTImmediately after th eo perator exi ted th eroom, telephone callswere made to

the local hospital and the police.An ambulancew as sentand within20minuteshe was admitted to the hospital

in Nesvizh.It was immediately clear to themedical staff, fromhisdescription of theaccident and the ini t ial symptoms of vomit ing, headache, fat igueand tachycardia,that he had been severely irradiated. The medical staff made th e decision to refer14


25/88

him tospecialists fo rtreatment, an darrangem ents were madefor him to bemoved,via Minsk, to the clinicat the Biophysics Institutein Moscow. He arrived thereat19:00 hours the same day. Details of further medical management are given inSection 6.

The police and the factory reported the accident immediately to the Ministryof Health via their 24 houremergency numberan d relevant staff were contacted athome. The factory wasaskednot to use the facilityand toleaveall thecontrolsa nditemsastheyhadbeenat thetimeof theacc ident.On themorningof thesame day,beforehistransfertoMoscow,theoperatorwasinterviewedan dsome dose measure-ments were taken aspartof reconstruction of the accident (see Section4). Alsoonthesameday,theState Comm itteeon theSupervisionofOperationalSafety inIndus-trial and Atomic Power Engineering set up a com mission to inv estigate the acc ident.The commission included representatives of the regulatory authorities, the plantmanagement and thetradeunions.Operationof theirradiationfacilitywassuspendedand the plant's licence withdrawn pending implementation of the commission'srecommendations fo r improvement, which included:(1) Enlarging the size of the pit and mo vin g the motor;(2) Providinganaudible alarmformovementof thesource rack withinthefac ility;(3) Providing visible wa rning signals insidethe facility to indicatethe positionof

the source rack.It is understood that all these recommendations were implemented and by the

t imeof thepost-accidentreviewinNovember 1992 theplantw asbackin full opera-t ion. In the interim, theoperator ha d died in hospital, having survived 113days.

4. DOSEESTIMATIONForco nvenience, this section bring s togetherthenumberofattemptsthatwere

made to estimatethe dose received. However, these estimates spanned aperiodo ft ime, and it is relevantto Section 6 that the exact t iming of when this inform ationbecame available to the medical team in Moscow be borne inmind.

4.1. PHYSICAL RECONSTRUCTION OF THE ACCIDENTAND CALCULATIONS OF THE EXPOSURE DOSEOnthemorningof theaccident, the operator (inhospital) and the staff (at the

plant) were interviewed in an effort todetermine exactly what had happened. Inparticular,inform ation w as sought on wh ere the operator had been standing and theduration of exposure. Th e initial estimate of exposure time was in the range of

15


26/88


27/88

14.5 0.7

Front18.0 __r _ 12.015.5

Back

FIG. 6. Expo sure doses (Gy) estimated by ESR.

It is relevant tonote that the information givenin Figs 5 and 6 suggests thattheoperator had been very close to the source, about0 .5 m from one edge in freespace,withoutanyshielding. Reconstruction of theaccident duringthe post-accidentreview visit identified that thiswas not atru e representation of the exposure condi-tions. Photographs 1l(a)and (b)depictthelikely positionsof theoperator duringthemajor part of his exposure. This isdiscussed further in Section 4.4.

4.2. BIOLOGICAL DOSE ASSESSMENTWhen the operator was in Moscow, samples were taken fo r analysis of

chromosome aberration in thebloodan dbone marrow lympho cytes. Some problemsarose in getting the cell cultures to grow, but by day 5 a dose equivalent of11 1.3 Gy was estimated. Inaddition, a review of the neutrophil and lympho cyteblood counts between days 1 and 8 wasm ade tha t suggested a whole body doseinthe range of9-11 Gy (see Section 6).

17


28/88

As the patient 's symptoms progressed, further biological indicators of dosewere given, in particular the way in which erythema of the skin and subsequentlesionsprogressed; theseprovided good indications that the left side of the body hadbeen closer to the source than the right side. These symptoms did not develop uni-formly , with an apparent increased reaction on the left elbow, the left buttock andthe lower left leg, the latter two reactions suggesting sharp cut-off points that wereindicative of some shielding, although conflicting factors aroseafter treatment (seeSection 6.5 for fur ther discussion).

4.3. ELECTRON SPIN RESO NAN CE STUDIESThe electron spin resonance (ESR) technique can be used to estimate doses

[14, 15]. Some clothing materials are particularly suitable for ESR measurementsand , since clothes are directly linked to the movement, orientation and t ime ofexposure of a person, spatial assessments ofdosecan be made.

Two days after exposure, the clothing worn by the operator at the t ime of theaccident was received by the Dosimetry Laboratory at the Institute of Biophysics,Moscow. His vest, which was made of cotton mixed with viscose, was chosen asthe most suitable apparel from the point of view of radiosensitivity. Samples weretaken at waist level and analysed by ESR. The results, shown schematically inFig. 6, were telephoned to themedical teamon 30 October (day4 ); these are consis-tent with th e later skin effects discussed in Section 4.2.

Following the death of the patient, a tooth, together wi th nail clippings fromth e fingers and toes, were taken for ESR analysis. The dose estimated from th eenamelof thetooth (left molar)w as 14.5 + 0.7 Gy. For thenail clippings, th elengthof timethathadpassedsinceexposureprecluded estimationof theabsolute dose,butdid giveanindicationof therelative dose. Aswith previous measurements, therewasa 20-30% increase indose on the left side relative to the right side of the body.However , th e estimated hand dose was 50% greater than that for the feet.

4 .4. EXPOSU RE FACTORSThe operator 's exposure falls intotw om ain parts: (1) whi lemoving in and out

of th e irradiation chamber; and (2)w hile t rying torearrange th ecarriagecouplings.For item (1), the exposure time is very uncertain, as i t is not known how long

th e source was in the'up' position. Duringthisphase, th eoperatorwouldhave beenshielded to some extent by the loaded produc tcarriages,possibly equivalent to ahalfvalue layer. However, at around waist level there would have been a gap of about10 cm between the upper and lower boxes. The effective gap may have been largerbecause of air spaces at the top of the lower product boxes. Similar ly,therewould18


29/88

have been higher dose rates to thefeetbecause of the 20 cm clearance of the boxesabove floor level.

The same considerations apply to his exposure while rearranging thecarriagecouplings (item(2)).However, other local shielding would have been present. Thenarrow angle of exposure relative to the plane of the source rack is such that smallchanges in position could have had asignificant effect on theshielding, i.e. fromattenuation provided by the stainless steel structure of the source rack and thecounterweight frame, from th e counterweights themselves, from th e productcarriages and from self-absorption in the source.

Thus, the operator was exposed in a complex geometrical pattern that waschangingwith timeand inwhich small movements could have significantly affectedthedosereceived. Thisis one of thereasons why the ESRmeasurements wereofparticular importance.

The initial physical dose assessment provideduseful information to the medicalteam. However, as mentioned in Section 4.1, their perception of the exposureenvironment differed somewhatfromthe real situation at the plant. Hadtherebeenadditional, more detailed dose reconstruction, involving both medicalandphysicsstaff at an earlystage, some medical queries regarding apparent variations in skindose may have been resolved. During thepost-accidentreview visit to the plant,rele-vantitems were photographed and video taped. The latter, in particular, when shownto the medical team, was clearly useful in helping to match the observed clinicaleffects to the possible exposure conditions.

5. LESSONS LEARNEDThe primary objective of carrying out a post-accident review is to identify the

lessonsthat can be learned froman accident and then to disseminate this informationto all who may benefitfromit. One of the problems encountered in this post-accidentreview was that, while the many individuals in the various organizations concernedwere co-operative inproviding interviews, therewas alackofreadily availableandcoherent data and documentation. These are relevant to safety and are reflected inthe observationsandrecommendations(initalic type) outlinedinSections5.1to5.7;they are presented under topics relevant to irradiation safety. Although it is clearwheretheprime responsibilitieslie forimplementing someof the recommendations,many,if notall,ofthese have implicationsfordesigners, manufacturers, operatingorganizations, competent authorities and international authorities.

19


30/88

5.1. PRIORITIES(1) This fatal irradiator accident is the fifth to be reported, joining those at the

Stimos plant, Italy [4]; in Kjeller, Norway [5]; in San Salvador , El Salvador[2];and at the Sor-Van facility, Israel [3]. In addi t ion,a numberof irradiatoraccidents have g ivenrise to serious, but non-fatal , radiation injuries [6-8].A clear lesson to emerge from the accidents that have occurred over thepastfew years is that radiation safety in irradiation plants m ust be given a muchhigher priority by all concerned, i.e. manu facturers, operating organizations,competent authorities an d international bodies.

(2 ) Thereporton the San Salvador acc ident stated that Manyof the recom menda-tions cover procedures and practices already considered to beessential to safeoperation . This statement was as valid for the Sor-Van accident as i t is forthe accident reported onhere.The IAEA has published guidance inthisareain Safety Series No. 107[9],wh i chis recomm ended to those withthe resp onsi-bility for irradiator safety.It is strongly recommended that: a ) A thorough review of radiation safety be carried out at each existing

irradiator to take accou nt of: ft) the lessons learned from this andpreviousaccident reports, and (ii)thecurrent gu idelines from the IAEA,manufacturers and competent authorities;

(b) All designs of facilities shou ld be in accordance with the standard ofSafety Series No. 107 [9].

(3) Thep r ime responsibil i ty for implement ing recommendat ion (2 )rests with th eoperating organizat ions; however, i t is recognized that theexpertiseavai lablevaries significantly, and that the operating organizations may need to seekexpert advice.In planning their workprogrammes, manufacturers, competent authorities an dinternational organizations should take cognizance of the increased priorityadvocated in recommendation (1) and the need to respond to operating organi-zations'requestsfor expert advice. Priority should begiven to those facilitiesthat have been in use for a number ofyears, and particularly those in develop-ing countries where the radiological infrastructure is probably not as strong.

5.2. INTERACTION OF THE TR ANSPOR T M EC HANISM ANDTHE SOUR C E R AC K

(4) In anumberof the previous irradiatoraccidents, aproduct jam had obstructedthe free movement of the source rack andacted as the initiatingevent. Since

20


31/88


32/88

(c ) Independence: achieved throughfunct ional isolat ionandphysicalsepara-tion of components .

(8 ) Altho ugh gamm a monitors had been installed inside the i r radiat ion chambe r,their signalwas notbeing usedto thebest effec t. First , th e ratemeter displayingthe reading could be switched between the output from the two detectorswithout it being clear to the operator which one was being displayed. Second,and more impor tant ,there was no connect ion between th edose rate controlpanel and the source control panel.The safety system must include a monitoring system with built-in redundancyto detect the radiation level in the irradiation chamber. The monitoring systemshould be integrated withpersonnelaccess safety features and warning systemssuch that: a ) If thesystem detects aradiation d ose rate aboveaspecified level, ormal-

functions, or is turned off, access to the irradiation chamber isprevented; b ) If the specified radiation level is exceeded, the system should generate

visible an d audible alarm signals an d termination of the irradiationshould be shown on the console.

(9) Itappeared th at , in the 'ready' positionof the key sw i t ch ,the source could becycled up and down w i thou t any audible alarms sounding in the i rradiat ionchamber . In theory, other safety systems should prevent anyone enter ing theirradiation chamber when movementof the source rack istaking place; there-fore, at f i rs t s ight the alarms might be considered unnecessary. However , ascan be seen from this acc ident , rel ianceonthis beingth ecase is not acceptableand provisionof an a larm would have advocatedth e redundancy and diversitypr inc ip les . Indeed, had such an alarm been provided itwould probably haveprevented th e accident.An alarm that isaudible both inside the irradiation chamber and at all accessportsmust be provided to indicate w hen the source is neitherfully shielded norin the fully exposedposition.

(10) In this and in the earlier accidents in San Salvador and at Sor -Van, ways ofc i rcumvent ingpartsof thesafety system were found. Thisis yetanother reasonfo r pursuing a policy of redundancy, diversity an d independence in safetysystems. However, attention should also be given to assessing how easilysafety features can be overcome. For example, in retrospect i t isclear that theelectric m otor located in the pit prov ided a way of circ um ven ting the safetyfeature, as indeed a commonly avai lable i tem such as a ladder would havedone.The effectiveness of safety features should be regularly reviewed. This isparticularly important at the design stage, but is also relevant to operatingorganizations and regulators during the l i f e of theplant.

22


33/88

5.4. INSTRUCTION AND TRAINING OF STAFF(11) Although there appears to have been a structured training programme at the

facili ty, any accident where the safety procedures have obviously not beenfollowed must call into question the effectiveness of instruction and training.Moregenerally, there will always be a need to review trainingprogrammes.The effectiveness of training, and indeed of many other safety aspects, candepend onhavingacoherent,up todateinstruction ma nual coveringall aspectsof the plant. a ) Training should be reinforced regularly and updated when necessary.

Particular emphasis should be placed on the junctions of the safetysystems andprocedures, and on thepotential consequences offailure tooperate them correctly.

b ) A comprehensive instruction manual coveringplant layout, safety andwarning systems, operating rules, emergency procedures and trainingshould be kept up to date and be readily available to the workforce.

5.5. MONITORING(12) In theearlieraccidents,lack of aportable monitor was acontributory cause;

this does not appear to have been the case in this accident. However,it wasnoted thatthe dose rate meters available were bulky, whichmay accountfortheir not always being used. An additional safetyprecaution should be personalalarm monitors.(a) Personal alarm monitors shouldberoutinely wornbyoperators through-

ou t their work shifts;(b) Procedures should bedrawn up toensure thatthepersonal alarm m oni-

tors are routinely checkedfor satisfactory operation; c ) Personal alarm monitors should be used in addition to not inplace of

portable monitors on entry into the irradiation chamber.(13) Failure of staff to wear their personal monitoringbadges while at work is a

common feature found in investigations of accidents and incidents. This leadsto loss of dose information, which could be valuable inresponding to an acci-dent. However, it is also an indicator of the general lack of attention that staffpay to radiation safety.The importance of w earing personal monitors by s t a f f should continually bereinforced by managements and regulators.

23


34/88


35/88

PHOTOGRAPHS


36/88

Photograph 1.Source rackframe and protection grid. The orange metal frame is the upperpart of the shielding plugfor the source rack. Spaces atbothsides of th eframe allow thecounterweights toslide down.

26


37/88

Photograph 2. Counterweight in the upper position (source in the shielded/storage position).

27


38/88

Photograph 3. Conveyor system withcradles and transport boxes.

28


39/88

Photograph 4. Distribution of carriages inside the irradiation chamber, indicating theconfined space when fully loaded.

29


40/88

Photograph 5. Linkage arms: (a)normalposition: and (b) abnormal position.

30


41/88

Photograph 6. Operator restoring conditions to normal with a metal rod.

31


42/88

Photograph 7. Lead shielded gamma monitoring probe inside the irradiation chamber.

Photograph 8. Gamma monitoring probe system in the control room.

32


43/88

Photograph 9. Dry pit with moveable floor section at themaze entrance.

Photograph 10. Portable dose rate monitor used during the entry procedure.

33


44/88

Photograph 11. Position of the operator in relation to the source frame and carriages duringreconstruction of theaccident: (a) at theback of theproduct transport system; and

34


45/88

(b) trying to release thejammed carriage couplings.

35


46/88

Photograph 12.Herpes labialis an dmouthmucositis: (a) day 4; and (b) day 6.

36


47/88

Photograph 13.H istology of the bone marrow biopsy: complete cell aplasy, withonly a fewmacrophages to be seen (xlOO) (day 16).

Photograph 14. Skin injury to the left elbow (day 16).

37


48/88

Photograph 15. Skin injury to the left hand (day 18).

Photograph 16.Skin injury to the buttocks (day 18).

38


49/88

Photograph 17.Almost total body erythema an d other injuries (day 32).

Photograph 18. Histology of the bone marrow biopsy: a noticeable increase in cellularity(x60) (day 44).

39


50/88

Photograph 19 .Skin injury to the legs;phlebitisof the l e f t leg (day 53).

40


51/88

Photograph 20. General viewof thepatient: (a)from the left side;and (b)from the right side(day 74).

41


52/88


53/88

Photograph 23. Post-mortem examination: right lung-foci o ffungal infection destroying theorgan tissue (x60).

Photograph 24.Post-mortem examination: testes- oedema of tissues, spermatogenic epitheliumisdead (x60).

43


54/88


55/88

(18) Lackofdocumentation relatingto theplant suggested that there hadbeen onlylimited contact betweenthemanufacturersand theoperating organization; thiswaspossibly due to changesin organizations an d their responsibilities in theformer USSR republics.To ensure that lessonsare learned fromradiation accidents arising from awhole rangeofdesignan d operational problems in many irradiation plants,it is important thatongoing contactismaintained betweenthe manufacturersand the operating organizations.

5.8. INITIATIVES BY THEIAEAFollowingtheirradiatoraccidentin SanSalvador, theIAEA enhanceditswork

on radiological protectioninirradiation facilities. Thiswasgivenfurther impetusbythe accidentsat Sor-Van and inNesvizh,an dsome relevant aspects shouldbenoted.5.8.1. Training

The experience gained from theaccidentsin SanSalvadorand atSor-Vanw asgiven prominencein theprogrammeo fregional training coursesinradiological pro-tection. In addition, a series of 2 week regional training courses specifically onirradiator safety fo r senior staff commenced in 1991.Five such courses have takenplaceinthreedifferentworld regions (Latin America, th eMiddle Easta ndEurope,and Asia and the Pacific).

A feature of all the fatal accidents that have occurred at irradiators hasbeenoperator ignorance ornegligenceas to theconsequencesof notfollowingtheproperprocedures.Thereisclearlyaneedtoprovide simple,buteffective, training supple-mentsforoperators.Atraining video tapei s agoodw ay ofreachingawide audienceand the IAEA, in consultation with manufacturers, operating organizations an dnationalbodies, hasproduced suchatraining tape[12],whichisavailabletoMemberStates. This shouldbe used tosupplement normal training regimes an d shouldnotbe regarded as a substitute.5.8.2. Dissemination ofinformation

The information disseminatedonaccidents an dincidentscantakemanyforms.IAEA post-accident reviewreportsnecessarily take some timetocomplete as infor-mation on all aspects has to be compiled, analysed and reported on in depth. Toensure early disseminationof the keypointsofsuchaccidents,asummary, basedonthe findings of thelocal authorities, wasprepared anddistributedassoonaspossibleafter th eaccident.Distributiono fthis summarywas on aninformal basis, primarily

45


56/88

to those know n to have a specific interest in thisareaor to those who had previouslyasked for advice.

To promote dialogue on radiation safety at irradiators and to disseminateinformation further , the IAEA organized a Technical Comm ittee M eeting to reviewRadiological Ac cidents in Industrial Irradiation Facilities and the ir Implicat ionsforCompetent A uthori t ies, Designers and M anufa cturers (Vienn a, 19-22 Aug ust1991).The participants heard presentations on fatal accidents in Italy [4], Norway [5],El Salvador [2] and Israel [3], which prompted some thought and ideas on theproblems of disseminating information, particularly in developing countries. Forexample, manufacturers often have diff iculty in identi fying com petent authori t ieswith regard to new facilities or better safety systems. It was agreed thatth e IAEAwou ld, insuch cases, beableto use itsgood office stoidentify competentauthorit ies.

Some significant steps have been takenin the disseminationof information. Inth e longer term, more formal arrangements m ay have to be made and the IAEAiskeepingthe matter under review.

The IAE A is also considering a pi lot prog ram me (Internat ional Review ofIrradiator Safety (IRIS)) to provide M emb erStateswithadequate advice on radiat ionsafety at gamma irradiators. IRIS will consist of a Secretariat to obtain and collatedata on irradiator types and locations, an Advisory Group to direct and moni toractivities, and on-site review teams to provide practical assistance to M emb er States.IRIS will form partof an integrated system ofassistance, includin g video tape train-in g packa ges, regional train ing courses and IAEA publicat ions.

6. MEDIC AL MANAGEM ENT6.1. INTRODUCTION

Closeobservation of thepatient suffering from acute radiation sickness (ARS)over a period of time(113days) provided agreatdeal of inform ation on the clinicalmanifestations, and very detailed diagnostic findingsand therapeutic measures. Fora comprehensive studyof thecase, th egeneral co urse ofdisease w as subdivided intoanum ber of timeperiods.Fo r each of theseperiods,the special tasks that the me dicalteam had to consider were th e level of information received, th emain cl inical find-ings and the treatment regime. From th e extensive experiencegained by the clinicat the Institute of Biophysics, where the patient was treated, five periods wereidentified.

During the first period (days0-7), th e main task was theurgent collection ofinformation ondoseassessment and evaluation of the uni form ity ofdosedistribution.This was necessary for decision makingrelated to the treatment of bone marrow andgastrointestinal syndromes, which began almost immediately.46


57/88

Th e second period (days 8-39) saw severe clinical manifestations of all theinjured systems an d demanded very intense therapeutic interventions under carefulclinico-laboratory monitoring.

During the third period (days40-70), some results of the treatment wereevident, withthe patient surviving the bone marrow and gastrointestinal sy ndromes.However, at that time new complications arose, possibly related to the severity ofthe treatment.

The fourth period (days71-101) was characteristic only for the patient, whodeveloped new lung infections and progressive metabolic changes. This requiredadditional changes in the therapeutic tactics.

The fifth period (days 102-113) was terminal, characterized by progressivepulmonary insufficiency. Intensive re-animating measures were necessary, buthopeless.

Descriptionsof medical management have been madeaccordingtothesetimeperiods.

6.2. CLINICAL COU RSE OF ARS6.2.1. Initial med ical management days 0-7)

At th e moment when th e operator saw the source in the upper position, henoted a metallic taste in the mouth and a feeling of heat on the face. He left theirradiation room and 5-6 minutes later experienced nausea and vomiting, whichrecurred frequently over th e next fe w hours. He realized immediately that he hadbeen exposed and, in accordance wi th instructions, telephone calls were made to thelocal hospital and thepolice. Twenty minutes later he wasadmitted to the hospitalin Nesvizh.

At the time of adm ission, he was in an excited state but nevertheless com -plained of fatigue, a headache, an ache in the abdomen, and pains in the hands andfeet. He waspale and continued to vomit. Fifty minutes after exposure he had asingle attack of diarrhoea. Th e first medical examination showed tachycardia, anormal blood pressure, which dropped after2 hours, and an elevated body tempera-ture (38.5C). Hispsychological state lowered rap idly , althoughhe stopped vomit-in g 6hours after exposure.

The first blood analysis (2 hours after exposure) showed early neutrophilicleukocytosis (13 X 109/L), with absolute lymphopenia (0.9 X 109/L). The bio-chemical test results werenormal.W hen physicians in thelocal hospital realizedthatthiswas aseverecaseofradiationexposure,emergency treatment wasgiven. Aveinwas catheterized for continuing infusions. The patient was transferred to Minsk4 hours after exposure and then to Moscow, where he was directly admitted(16hours after exposure) to the clinic at the Institute of Biophysics, a specialized

47


58/88

FIG. 7. Schematic representation of thebone marrow punctures an d biopsies performed:(1)sternum; (2) third thoracic vertebra; (3)crest iliaca anterior sinistra; (4)crest iliacaposterior sinistra; and (5)crest iliaca posterior dextra.

hospital forradiation injuries. He did notrequire anyspecial supportfor thejourneyby ambulancefromNesvizhtoMinsk (200km) or byaircraftfromMinsktoMoscow(800 km).

At thetimeofadmission to theclinicinMoscow,thepatienthadlight erythemaof the face, neck andhands, andswelling of theparoticglands (greater on the leftside). Amylasuriaw as up to five times higher than normaland hisbody temperaturewas37.6C, buttherewas nonausea or vomiting.

From th e local hospital s information and an interview with th e patient, th ephysicians in Moscow realized thattheywere faced withasevere case ofirradiation.Therefore, thepatient wasimmediately placedinisolation (22hours after exposure)and provided withaprophylactic; enteral selective decontaminationand hepariniza-tion were started simultaneously. Blood samples were taken for peripheral cellcounts, human leukocyte antigen (HLA)(A and B)typingan d cytogenetic analysis.Bone marrow from five different sites in thebodyw astaken fo rdoseestimation andevaluationofdose distribution (Fig. 7). Thefirst sternal aspirationwasdone2hoursafter admission (18 hours after exposure), and the third thoracic vertebra, theanterior andposterior left iliac crestsand theposterior right iliaccrestwere punc-48


59/88

tured 8hourslater.]Anelectroencephalogram examination w asalso performed andfound to be normal. Thepa tient'ssister wascalled to theclinic for HLA typingasa potential donor.

Severalcallsfor international assistance were made.R.P.Gale (Bone Marrowand Stem Cell Transplantation, Salick Health Care, Inc. , UnitedStates of America)was askedtoprovide haematopoietic grow th factors. Through the Service central deprotection contre les rayonnements ionisants (P . Pellerin) and the Centre de radio-pathologie(H.P.Ja mm et) (both in France), co mm unicatio nwiththe French Registryof Bone Marrow Unrelated Donors was established, and blood samples of the patientand his sister were sent for typing to J.J. van Rood at the Leiden Laboratory,Netherlands.

As a result of these measures, treatment with growth factors (granulocytemacrophage colony stim ulating factor (GMCSF) and interleukin-3 (IL-3)) wasstarted on days 1 and 6, respectively:(1) The results of domestic HLA typ ing were confirm ed by the Leiden Laboratory

an d showed haploidentity between th e patient and his sister;(2) A potential unrelated donor for a bone m arrow transplant (BMT) was found

in France.The gene ral condi tion of the patient was satisfactory between days 1 and5.The

erythema evident on day 1 started to diminish on day 2 and had completely dis-appeared by day 6. Sw elling of the paroti c glands also decreased.

Herpetic lesions (herpes simplex) appeared on the upper lip, the left side ofthe nose and the left cheek (photograph 12(a)). These lesions showed favourableevolution, with a crust on the lip and a tendency to healing on the cheek (photo-graph 12(b)). This evolution was mainly attributed to the antiviral treatmentadministered to the patient on admission.

The rapidly decreasing granulocyte and thrombocyte counts were the majorhaematological signs during this period, indicating a high radiation dose. Thelymphocyte counts were below 10 /L from day 2. The granulocytes fell from4 x 10 I2 /L on day 2 to 0.5 x 1012 /L on day 6. The thrombocytes also fell, from14 0 x 10I2 /L on day 2 to 80 x 10I2 /L on day 6. These haematological data con-firmed that the case was very severe. From the reference curves of the blood cellcounts elaborated previously [16], the mean whole body dose was estimated to beabout 12 Gy. Annex I shows how this was done.

Th eresultsof cytogenetic analysesaregiven inAnnexII. It wasimpossible toevaluate the dose from cytogenetic analyses of the peripheral blood lymphocytes

Such early bonemarrowsampling for cy togeneticanalysis wasnecessary fo rpossible doseassessmentby the socalleddirectmethod, avoiding cultivation. Thismethod isonly informa -tiveduringthe first24-30hours after exposure.

49


60/88

because of: (1) the very low level of circulating lymphocytes; (2 ) their inability toundergo mitosis; and (3) failure of the attempt to produce cultures. The frequencyofaberrant bone m arrow cellsw as 100%.From previous experience, th ewhole bodydose wasestimated to be higher than 6 Gy, andprobably between 8 and 10 G y . Onda y 5, this estimate was refined and the dose w as evaluated at about 11 Gy. Theresults of ESR were received on day7, again confirming the high level of exposure,i.e. between 11 (right side) and 18 Gy (left front) at waist level (Fig. 6).

Day 6marked the end of the so called latent period, which was, however, notas pronounced, since several clinical signs had appeared:(a ) M ucositis in the mo uth (photographs 12(a) and (b)), with oesophagic pain;(b) Diarrhoea with wa tery stools (0.6-0.9 L/d);(c) Some loss of body weight ;(d ) Progressive cy topenia;(e ) M oderate fever (about38C) .

Duri ng this period i t wasdecidednot toperform aBMT,b utrather tocontinuethe t reatment with haematopoiet ic growth factors (GMCSF and IL-3).6.2.2. Clinical course (days 8-39)

This period corresponds to the crit ical phase of ARS, resulting from a combi-nation of haematological syndrome, gastrointestinal syndrome and skin injur ies .

Fever w as constant, between 38 and39C.The haematological syndrome w ascharacterized by deep pancytopenia, confi rmed by complete bone marrow depres-sion (photograph 13). However, the bone marrow investigations performed ondays 23 and 27 showed young forms of myelopoietic cells. At this t ime, theperipheral blood count showed the appearance of a few granulocytes and monocytes.From day 33 reticulocytes were increasingly found.These changes, which indicatedthe reversibili ty of the pathologicalprocess, provided some hope to the physicians.However , th e lymphocyte counts decreased continuouslyup to day 18. The level oferythrocytesw asconstantly low, demanding repeated erythromasstransfusion. Like-wise, the continuous decrease in thrombocytes required almost daily thrombomasstransfusions. The patient received antibiotherapy of broad spectrum, includingantiviral drugs.Mucositis in the mouth w as still moderate, withsmall erosions, and recoverybegan on days 14-15, exhibiting epithelization on days 28-30.

Manifesta t ions of the gastrointestinal syndrome were moderate (three to ninestools per day, with avolume vary ingfrom 0.5 to 1.5L/d) ow ingto total parenteralfeeding. Any attempt to begin enteral feeding led to a deterioration in the syndrome.Between days36 and 39thereweresignsofintestinal bleeding, w hic h required trans-fusions of large quantit ies of red cells an d platelets, as well as large volumes ofliquids.The gastroduodenal endoscopy showed alargenumberofsmall erosionsand50


61/88

haemorrhages. The whole mucosa bled. Treatment with heparin was stopped. Themo st significant resulto fthis syndromewas a rapid deteriorationin thegeneral stateof thepatient, whose w eight dropped from 70 kg on the day ofadmission toaround60 kg on day 39.

Skin injuries appeared on day 11,with epilation thatwasmore evident on theleft side of the body, beginning on the head an d extending to the chest and pubis.Secondary erythema appeared between days 11 and 15 on the left side of the

face,ear and neck. Drydesquamationoccurred at these places on day s24-25. Fromthis time, the patient demonstrated a gradual extension of skin injuries, whichappeared successively onvarious parts of the body.On day11, erythema with skinoedema appeared onboth elbowsan dboth iliaccrests.However, some time laterallthe signswere more pronounced on the left sideof the body. Photograph 14 showsskin injury to the left elbow an dphotograph 15 the dorsal surface of the left hand,with marked localizationson the fingerjoints. After a fewdays, blisters an dsmallerosions appeared onthese areas.Also on day 11,erythema appeared on theleftbut-tock, and extended to the rightsidewithin a few days (photograph 16), At the endof the first month, the gradual process of skin lesion extensions resulted in almosttotal body erythema (photograph 17). Wet desquamation was especially marked onthe left handan d left leg. Some signs ofskin recovery were visibleat thebeginningof the second month, with skin injuries not as painful as expected. Severe painappeared inboth feet aroundday 20 andpersisted for 1week.Thecharacter ofthesepainswas nottypicalof classical radiation injury. Aquestionw asraisedo n thepossi-ble combination of three causes, i.e. radiation, the sideeffects of the growth factorsand some local circulatory disorders. After areduction in the growth factorsdose,the pain diminished and was only exacerbated by contact and pressure. Painfulbalanoposthitis developed on day 28, but this was cured within 1 week.

Skin biopsies of the left iliac region were carried out on days 3 and 16. Onday3,no significant signs of radiation injury wereshown ; dystrophic changes werefound in one of five to six normal cells, while skin appendices were normal,indicating a skin dose of lower th an 20 Gy .On day 16, more specific changes wereobserved, such as a decrease in epidermis cellularity, dyske ratosis, wh ich consistedof alarge increase in the numberofm elanocytes, andmarkedinfiltrationsurround-ing the dermal vessels. These clinical findings confirmed the possibility of spon-taneous skin recovery. At this time, the pathologist estimated the skindose to theleft iliac region to be around 8-10 Gy.

Regular chest X ray examinations showed almost normal lung pictures ondays 9 and23,w ith the exception of a slight vascu lar net. O n day 38, a solid shadowcouldbe seen in the middle lobe of the righ t lung and som e infiltration s in both lun gs.However, there were no clinical symptoms of any lung disease at this time.

In this critical period, no patent bacterial infection was evident; th e lungshadow found on the X ray on day 38 was suspected to be alocalized fungal pneu-monia an d treatment with amphotericin B was started.

51


62/88

101oX

I IIL-3r I I ~IGMCSF

Erythrocytetransfusions f IT m imurnnnTMii im in m H IT

111

10(a)

0 10 20 30 40 50 60 70 80 90 100Daysafterexposure

110

I IL-3 1GMCSF

1 2,x,

81

10

ThrombocytetransfusionsinTf? nin nrir imi r t v f v

(b)0 10 20 30 40 50 60 70 80 90 100 110

DaysafterexposureFIG. 8. Counts during thewhole clinical course ofARS, andtheir dependence on treatmentwith g rowth factors: (a) erythrocyte an d reticulocyte counts; (b) thrombocyte counts;52


63/88

io-3 0 10 20 30 40 50 60 70 80 90 100 110Daysafter exposure

10-=

(d)

10 20 30 40 50 60 70 80 90 100 110

Daysafter exposure(c) neutrophil counts; and (d) lymphocyte counts.

53


64/88

From the beginn ing of this period, a gra dual increase in infec tion of the liverstarted. B iochem ical tests show ed only an insign ific ant rise in transaminase andcreatine ph osphakinase, wi th s ignificant ly progressed hyp oalbum inaem ia, wh ichrequired intensive album in t ransfusions.6.2.3. Clinical course days 40-70)

During this period, the general state of the patient was relatively stable, andhis mental state had even improved; his fever after day 50 was neither as constantnor as high as earlier.

The neutrophil counts in the blood showed a plateau, wi th small var iat ions(between 400 and 1000 X 109/L). The lymphocytes also showed a plateau, but onlyat a lower level. Thrombocytes were still at a low level and repeated thrombocytet ransfusions were necessary. The erythrocyte counts cont inued todecrease slightlyduring the first days of this period; th e ret iculocyte wave that had appeared in theprevious period lasted up to the end of the second month and had completely dis-appeared by day 62 (Figs 8(a) to (d)).

Bone marrow exam inat ions were done on days 44 and 62 and demon strated anincrease in cellular i ty (photograph 18).

Gastrointestinal syndrom e ma nife stations began to decrease after day 40;however, anorexia w as persistent and there were some episodes of vomi t ing . Thepatient received partial parenteral feeding.

Some signs of hepatit is, revealed by biochemical findings, were constant andbecame more evident by the end of thisperiod.

The oropharyngeal syndrom e was character ized by only a sl ight at rophy of themucus. Epithelization w asevident inmostof theskinlesions;how ever, the new skinwas atrophic and dark brown, atrophy being more marked on the left side of thebo dy .Phlebitis in a vein of the left leg (vein saphena) persisted. The skin of the leftleg washyperhaemic , wi thoedemaof the underly ing tissues and superficial erosions(photograph 19); pain was exacerbated dur ing thisperiod. A skin biopsy showedfibrinoid swelling in the skin blood vessels and moderate dystrophy of theappendices.

X ray examinations carried out on day 51 showed slight amelioration of theright lung lesion. Tom ograph y on day 54 showed increased in filtration in the leftlung, and some tissue destruction was suspected. Some time later (day 58), clinicalsigns of bilateral pneumonia appeared. An X ray examinat ion performed on day 67showed an en largement of and an increase in the density of the right lung shado wan d the appearance of a new shadow in the lower lobe of the left lung. X rays onda y 70 again showed some signs ofpneumonia which, however , gave no indicationof th e etiology (causal agent) of the lung injury. During this t ime, there were noclassical symptoms of radiation pneumonitis. This was also confirmed by the nega-tive results of repeated transcutaneous and transbronchial lung biopsies. However,th e patient received all the necessary treatment for pneumonia .54


65/88


66/88

(6) At the end of this period, 3 months after exposure, signs of renal fai lureappeared.

6.2.5. Clinicalcourse(days 102-113)Anopen biopsy of the left lungw as decided upona nd carried out on day 102,

but no infection was revealed. The patient withstood the operation relatively well,bu t 2 da ys later (day 104) his respira tory con dition suddenly wo rsened, resulting ina typical acute adult respiratory distress syndrome.

The general condition of the patient deteriorated, with a sudden collapse onda y 104, accompanied by a rise in body temperature. Alterations in the ma in m etab-olism process were observed and acidosis appeared. During this time, the patientdemonstratedev ident symptom s of respiratory insuff ic iency, with hypoxae mia, chestpains and a slight cough, although there was uncertainty as to their specificcause.

Active therapy, including hyperbaric oxygenation following artificial lungvent i lat ion (ALV), proved to be ineffect ive, even harmful ; hypoxaemia increasedrapidly and, as a resul t of AL V, a pn eumo thorax developed and new symptoms ofsevere pneu mo nia appeared. A severe respiratory syndrom e, intoxication andhyperosmolar state with increasing renal insufficiency led to encephalopathy ofmixed genesis.

The patient's blood pressure decreased sharply on day 112, and he died thefollowing day whi lein ahy poxaem ic coma, main ly because of respiratory distress.

6 .3 . M EDICAL TREAT M ENT AND A REVIEW OF ITS EFFECTIVENESSThe specialized hospital in wh i ch th e patient w as treated, th e clinic at the

Institute of Biophy sics, has m any years of experience in the management of radiationinduced injuries. The main object ivesof the treatment were to reduce the period ofbone m arrow aplasia, m aintainin g circulating blood cells at an acceptable level; toavoid infection and haemorrhaging; to protect the gastrointestinaltract (GIT) fromas much harm as possible; and, finally, to support th e metabolic funct ions.

As soon as the patient was admitted to the Moscow hospital he was placed inan ove rpressurized isolated room that was providedwith sterilized air. From the firstdays of hospitalization he was given intensive ant i- infect ion treatment, includingclassical antibioticsas well as antiviraland antifungal drugs. During th e course ofthedisease, preciseantibiotherapywasadapted toactual clinical findings. From thestart, the patient also received enteral selective decontamination. On day 6, he wasplaced under total parenteral nutrition, providing 2500 kcal/d. On the first day ofadmission (day0), immunoglobulinand heparin were provided.

Informat ionon dose level and its distribution are of prim e importance for somecrucial therapeutic decisions. Inthis case, it was recognized early that th e dose w as56


67/88

very high, around 10 Gy ormore. From the initial informat ion received from thepatient and from a description of the accident circumstances, the physicians couldnotexclude some non-uniformityofexposure. Thiswas in factlaterdemonstrated,witha factor of about 1.5shown between the least exposed and the most exposedparts of the body. Nevertheless, becauseof the high level of dose, the relativelyheterogeneous dose distribution did not constitute a major parameter for decisionmaking.

After longa ndcontroversial discussions with foreign specialistsin aninterna-tional consultation held on 3 November 1991 (participants: A.K. Guskova,A.E.Baranov,H.P.Jammet, R.P. Gale and A.B utturini ,as well as T.M. Fliednerby telephone) it wasdecidednot toperformaBMT, either from thehaploidenticaldonor (the patient's sister) or from the donor found through the French Register,because the hazards, as judged by the physicians responsible for the patient, out-weighed the benefits. Therefore, the decision was made to stimulate haematopoiesiswith growth factors and to carry out replacement treatment with classical trans-fusionsof red blood cells andthrombocytes(T-leukocytesdepleted), the frequencyan dvolumeofwhich were accordingly adjusted.Thetotal numberofperfused cellsw as very high, far above that usually found in an ordinary hospital department.Transfusions started on day 6 and were necessary during the whole course of theillness.

Itshould be emphasized that, insuchsevere cases, the need for repeated trans-fusionsof redblood cellsandplateletsisvery high. These provedto bevery effectiveas the patient overcame the haemorrhagic syndrome and intensive bleeding.

Before anydecisionwasmadeonwhetheror not tocarryout a BMT it wasdecided to stimulate haematopoiesis by growth factors. Two growth factors wereused: GMCSF, which is a common stimulator of myelomonocytic progenitors, andIL-3, whichis alargespectrum stimulator of themost primitive stem cells.

Treatment with GMCSF (slow intravenousinfusion (1 1-4 p ig-kg 1-d~ ' ) ) w asstarted on day 1 and a further 6 p i g - k g ^ - d 1 was initiated on day 2. However,treatment was interrupted on day 6 because some clinical signs appeared that mayhave been drug related rather than due to radiation. These possible side effectsincluded fever, oedema, pain, arthropathies and thrombophlebitis, whichcould havebeenthird grade GMCSFcomplications.For this reason, GMCSF was replaced byIL-3, whichwasinjected (slow intravenous infusion (10/ig-kg^-d 1)) from days6to31.As the general state of the patient improved, and as the supposed side effectsdecreased to a level consistent with radiation exposure, GMCSF was again used fromdays 16to 39.Consequently, thepatient received both growth factorsfrom days 16to 31.

Regarding the haematopoietic system, the two waves of reticulocytes thatappeared in the circulating blood on days 34 and 82 (each lasting about 1 month) canbe interpreted as follows: the first wavewas directly related to the use ofgrowthfactors, according to the bone marrow examinations, while the second was due to

57


68/88


69/88

(2 ) More or less pronounced epilation, with newhair follicles appearing on therightsideof theface, while some regions of thebody appeared to bealmostnormal:right scapula, upper partof theright arm, right part of thechestandtheabdomen, and the right thigh.

(3 ) Deepandlimitednecrosisin thesacrum region, related toprolongedde

the radiological accident at the irradiation facility in nesvizh

Documents