ML20197F467
| ML20197F467 | |
| Person / Time | |
|---|---|
| Issue date: | 05/10/1988 |
| From: | James Shea NRC OFFICE OF GOVERNMENTAL & PUBLIC AFFAIRS (GPA) |
| To: | Stello V NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| Shared Package | |
| ML20197F457 | List: |
| References | |
| NUDOCS 8806130265 | |
| Download: ML20197F467 (73) | |
Text
_
p ' ' '. . .
' i- s UNITED STATES lh / i8 NUCLE AR REGULATORY CCMMISSION
'. ASHINGTON. O C. 20555
. * %gn,.9x >- f ,E
. . . . . ./
MAY I 01ggg r*
f' MEMORANDUM FOR: Victor Stello, Jr.
Exccutive Director for Operations FROM: [haamesR.Shea, Director
/ International Programs
/ Office of Governmental and Public Affairs
SUBJECT:
hRC COMMENTS Ch DRAFT 1AEA CONSULTATIVE DOCUMENT, "THE APPLICATION OF THE PRINCIPLES OF RADIATION FROTECTION TO SOURCES OF POTENTIAL EXPOSURE:
TOVARDSAUNIFIEDAPPROACHTORADIATIONSAFETY(
The enclosed draft IAEA repert is forwarded for NRC staff review. C.oments
, _should be provided_ by lune 10 to Maroe Mahy, GPA/IP. In addition, inter-e agency discussions with 00E, EPA and others could take place in mid-June
- to prepare consolidated U.S. coments to the IAEA by June 30.
GPA/IP has infomally contacted NMSS regarding this report. As the review
' progresses, GFA and E00 should consider whether the document raises policy questions of interest to the Comission.
Enclosures:
- 1. IAEA letter dated 4/7/88
- 2. DOE memo dtd 4/14/88
- 3. IAEA letter dtd 3/15/B8 w/enci 4 DOE note dtd 4/28/88 cc: H. Denton, GPA H. Thompson, NMSS R. Cunningham, NMSS I
8806130265 PDR ORG
% PDR 3
I CC---O t, J c c 5
.a I
1
)
ti-ENCLOSURE 1 i
l J
- . 2 ;_ .: .. ; _ . . r:- ,O g y , ;;;n :,;,. = _
h INTERS TIONAL ATOMIC ENERGY AGENCY ACESCE INTERNATIONALE DE L'ENERCIE ATOMlQUE ME*MH APO2 HOE ATEHTCT80110 ATOMHOR 3HEPryH ORGAN 15MO INTE',h AC10N AL DE ENERGIA ATOMIC A,*
hf w AQs *?.ur! A A151 S, P 0. tox ICC A 1400 V!1hh A. AUSTRIA #
TEl.1X 14 W1 ' AJL.L J th ATOW Y.fBN A. F ACGOd 1143 322 2H164. TZ1.DH0h'E. (222)2360 nw.utmw, _mw=:s J1-Ac-496.o: less-o4-o7 sir, I have the honour to inform ou that the International atoale Energy Agency plans to engage a casultaatifrea 18 42 July 1988 to assist the secretariat in the preparatten of a working document for an Advisory Croup to be held from 17 - 21 october 1984 on the Application of the Prinelples of Rediation Protection to sources of potential tacposure. This Working Document
- will be prepared based on the earlier dra.f t which has been circulated as a j
Consultative Document to all Mee.ber States for comments. The task of the consultant will involve reviewing the comments and incorporating them as eppropriate. Into the draf t to make it ready for consideration by the Advisory Croup f rom 17 - 11 october 1944.
The Agency initiated the prograaste on the Application of Radiation protection principles to the Sources of potential tacPesum in October 1985 by conven!ng the first Advisory Group meeting on the subject. The second Advisory meeting wa.a held from 19-23 January 1987 which prepared a consultative document for circulation newng the IEmaber states for comments.
The Advisory Croup meeting from 17 - 21 October 1984 will be the third and the final Advisory Group meeting to finaitse the document for publication in the IA1A tafety series.
I should be grateful if your Govet'nannt would necinate an expert who could be designated by the Secretariat to et.-ve as a consuitcat for the purpose just mentioned.
TtA Agency will bear the actual cost of 1.he round trip by the most direct route to and from Vienna, listited to ocorday class by air or first class call (in double sleeper), and peevide hla with a subsistence allowance at the cate prevailing at the time of the meeting. This allowance at present amounts to As 1,921 for each r.ight while in Vieran in connettien with the meeting.
The Resident nepres6ntative e.f the tmited states e ! Amstice to time Internatiot il- Er v r Montt obe rs t eine ry +
A-1194 Vien-
, .e 3 oc .
. I j
The tickets, if tifte per. nits, will be issued by the travel 44ent of the Agency. Plasse note that any expense incurred in connection with the travel aey only be relabursed if necessary evidence (receipt, bill etc.) is provided.
Fle484 note that compensation is not payable by the IAEA for any das.44e to or loss of the participants' personal property or for any illness, inj ury or death attributable to their relationship with the IAEA water the present a rransec.e n t . However, while travelling under the aatherity 2And at the request of the IAXA, the des 1& nated experts wl11 be covered under the IAJ.A insurance policy for. Inter alla, permanent total disability or death up to an a . cunt of US$$0.000 in case of an accident related to euch travel.
It would be appreciated if your Coverreent's nomination could reach the Agency as soon as possible. Subsequent correspondence will be exchanged between the scienti.f te secretary of the meeting, nr. J.U. Ahmed of the Division of helear safety, and the eitport.
Acceptg Eir, the assurances of my Qhast eq@ggi.ga.
F n- g, i
onstan i v
- , Deputy Director General Head of the Dwpartment of
- '. Sueleat Snergy and S36!ety f or DIRECTOR GEW12AL y _ . . ... .. .
I g --. - - - - , - - - - - - - - , , - - - - - - - - - - - - , - , .--,------,,--,,-,-,~e -
---. ----v-,,v--,-,a
i V.*
I
\
ENCLOSURE 2 4
,--- - - - - . -. - , - - . - . - - - - - - - . _ , . -,-------ev._,--., n
.v'% .
r-1 Department of Energy WasNngton DC 20585 s April 14, 1988 MEMO FOP: J. Fitzgerald, EH-35 R. W. Wood, ER-70 C. Welty, EH-21 S. Rosen, NE-14 C. Cooley, RW-41 ,.
4' SUBJFCT: Call for Comments on IAEA Radiation Protectiot Document - Friday, June 10, 1988 Attached for reuiew by comment by June 10 (HOS and/or Field as appropriate) are copies of a draft IAEA Consultative Document, "The Application of the Principles of Radiation Protection to Sources of Potential Exposure: Towards a Unified Approach to Radiation Safety." As the cover letter indicates, U.S.
consolidated comments are due in Vienna by June 130, and the final version is the subject of an Advisory Group meeting scheduled in October. (Participants in the previous Advisory Group meetings are identified on pages 56-69).
Please note the statement in the forward. It appears to me that ethe IAEA is moving into sienificant new territory. The brinciples being developed are directed beyond planned operations to accidental conditions.
In your comnents I would appreciate your addressing not only the content of the document but also on underlying assumptions and the extention of principles developed for one area of circumstances into a different area. Please also let me know 6 about areas in which to inter- and intra-agency differences of views. I do not yet know what other agencies are reviewing the document besides DOE and NRC. I would expect that a U.S. reply will have to factor in EPA views. Early coordination will be helpful in considering guidance that may be required for whoever attends the October meeting.
Thanks for your help.
i Barbari Thomas, IE-13 (586-6188)
Attachment:
a/s cc w/att:
P . B ru s '.1, IE-13 R. A. Bowen, DP-332
- w. Shepard, DP-222 W. Declercq, OES/ State R. Hauber, IP/NRC
, "TERN A710NAL ATOMIC ENERGY AGENCY
'E INTERN ATION ALE DE L'ENERGIE ATOMlQUE
. aYHAPO2 HOE ATEHTCTBO 110 ATOMHOil 3HEPTHH QRC ANISMO INTERN ACION AL DE ENERGIA ATOMICA 4 ACR A.MERSTRA551 !, P O SOX 100. A.1400 VIENN A At'STR!A TELF.X l 1:643. CA3LE INATOM VIENNA. FACSLMILE 43 ::: 230184 TII.IPHONE (:::):360 Inisn' tit [OIt[tal'dirtat<t cocosIaeia S t Nu'ta!Yt$N J1-AG-5/,8.2 Cire. ,
15 March 1988 Sir, I have the honour to forward to you the attached Consultative Document on "The Application of the Principles of Radiation Protection to Sources of Potential Exposure: Towards a Unified Approach to Radiation Saf ety" f or coceents f rom relevant institutions in your country. The coveents arising from the coneultations will be considered by an Advisory Group meeting scheduled from 17 - 21 October, and the final document.will eventually be published in the IAEA Safety Series. Please note that the attached document is only available in English.
I should be grateful if the cements on the Consultative Document are sent to the Scientific Secretary, Mr. J.U. Ahmed, at the Division of Nuclear Safety, IAEA, P.O. Box 100, A-1400 Vienna, by 30 June 1988.
Accept. Sir, the assurances of my highest consideration.
k[h Attachment Maurisio Ziffersro Acting Director General 1
COPY COPIE : COPY COPIE : COPY : COPIE : COPY COPIE : COPY : COPIE : COPY l
l l
'r.
ENCLOSURE 3 J
J O
O 4
4
- , , . - , - ~ - - - - - - - ,n -wev., ,,,__n-.,- - - , - , ---mv _,,-_v.w_,, -wn-->,,---,-,, -gw,,,,,. em.-~,ww-,-~w-eyv,,
825es y.
. s The Application of the Principles of Radiation Protection to Sources of Potential Exposure:
Towards a Unified Approt:h to Radiation Safety e
e J
J I
COWSULTATIVE DOCUMENT l
l FIBRUARY 1988 l
/
P TABLE OF CONTENTS i
10 REWORD /
PREFACE CHAPTLR 1 h*
SCOPP OF THE REPORT ,
CHAPTER 2 BASIC UNDERSTANDING Uncertainty Probability Risk Limits Obj ec tives ,
Goals Sensitivity
\
Craphical representations CHAPTER 3 -
- , IFTRODUCTICW
". Characterization of Sources of Radiation Erporures The need for a Unified Approach to Radiatian Safety Recent international developments connecten with a broad approach to radiation safety.
CHAPTER 4 BASIC APPROACHES TO SATETY AND PROTECTION Current Principles of Radiation Protection A Possible Extension of Current Radiation Protection Principles Current Principles for Saf ety Assessment Examples of Current Eafety Practices CHAPTER 5
- LIMITAT10R OF INDIVIDUAL RISK
The Derivation of a Risk Limit Risk Limits f or Potential Exposure of Individual Members of the Public Risk Limit for Potential Exporure of Workers Derivation of Design Constraints from Risk Limits !
Comparisen with Duclear Safety Criteria l l
CHAPTER 6 I IMPROVING SAFETY FURTHER (CR SAFETY OPTIMIEATIOW) l l
CHAPTER 1 l LIMITS OW SOCIETAL RISK l l
l i
l l
~
.- -(.
)
l FOR EWORD t
This technical docueent on the Application of Principles of Radiatien Protection to Sources of potential Exposure is published as a cottsultativer document for Mirculation among the IAEA's Member States and 4FproFriate organizations and experts in the field for corrents.
f *.
J In 1982 the IAEA published the revised Basic Safety Stanbards'for Radiation Protection (Safety Series No. 9), which was sponsored jointly by the / DEA. the International Labour Organisation, the World Heal'.h Orgytitation and the Nuclear Energy Agency of the Organisation Jf Economic Co-caeration and Development. The Basic Safety Standards rare based on the reconrandations of the the international Commission on Radiological Protection (ICRP), which t<ere issued in 19 ? ? (ICRP Publication No. 26). The relevant reconnendation of the ICRP requires c(epliance with tM,e system of dose limitation, which consists of three interrelated components, namely the justification of a practice, the optimization of protection and individual e dose limitation. This system hse been adopted internationally and by most national organizations.
f The radiation protection principles developed by the 1CRP are consistent and coherent for wpplication to exposures which are assumed to occur with a probability of unity, usually during normal operation of an installation causing exposure. The situation is rather different for exposures that are not certain to occurt that is, exposures that may or may not occur, with a given probability. This usually refers to a:cident ( !*
s conditions where the probability of exposure is less than one. So far no lf international or national organization is known to have prepared f#
reconnend4tiors on safety for radioactive sources having the potential to cause exposure to radiation that are consistent with the principles of radiation protection applied to actual exposures rekulting from the normal '
j i operation of installations causing exposures. The ICRJ dose limithtion '
systee cannot be applied in its current form to control sourt<,s '. hat may or >
l 1
may nbt give rise to radiation doses.
I i
3 l
't ,- 11 e
The IAEA initiated in 1983 an expendsd progr:mma of rediction ,
protection, which includes a component on the. application of radiation protection principles to sources potentially c'ausing ,cxposure. The aim of this activity is to develop guidelines for a yryified approach to the application of radiation' protection principles )both to radiation exposures occurring with certainty and to exposures thaty are not,lcerta inI, tp",4 occur.
The programme started with an Advisory Grou? meeting irt get ibar F985. This was followed by a meeting of a group of cont /ultants in April 1986 to
\
complete the first draft of a document, whith was then circulated among the participants of the Advisory Group for edhronts. The second Advisory Croup meeting was held in January 1991 to prepale this consultative document. A third Advisory Group meeting is scheduled for September 1988 to finalize the document for publi:stion in the IAEA Safety Series.
The Agency wishes to convey its thanks to all the particidents in the Advisory Group and Consultants Meetings for their contributions towards the completion of this work.
1 /
e t
i r l l' Y ,
s
/
9 t h'
< /-
i a
i
\
- ,. . ~ ,
t
, g f
PRETACE Over recent years there have been substantial international discussions on t,he so-called ' interface' between the systems for the regulation and control of normal (i.e. planned) and accidental exposures to ionizing radiation Thesehavetakenplacebecauseof,forexam{ie, the need to establish procedures for assessing the safety of radioactive waste
. repositories, and the desire to consider the interaction between_tfoutine exposure of workers and requirements for safety related maintenance or ,
inspections.
This dia'ogue is continuing and can be vrpected to have f ar-reaching consequences for radiation protection and nuclear s.fety. The present consultative document has been produced as a contribution to this dialogue and as a means of helping to reach a consensus on the current state of thinking before proceeding to produce guidance and/or recommendations in a
, Safety Series report. Comments are invited and should be sent to Mr. J.
, Ahmed. IAEA, P.O. Box 100, A-1400 Vienna, by 30 June 1988.
I 3
i
Scction 1 Scope 1.01 This document discusses radiation safety principles that could be applied to sources actually or potentially causing exposure. The document analyses the yRp's radiation protection principles for exposu s presumed to occur with certainty and their extension for dealing with exposures that are not presumed to be certain to occur. It also examines how a unified approach to radiation safety could be developed in the light of the various radiation related safety criteria proposed.
1.02 The International Commission on Radiological protection.(ICRP) has developed a consistent and coherent set of radiation protection principles
[1] which apply to exposures that are presumed to occur with c rtainty; the source of such planned exposures can therefore be assumed to be' under control . Such control applies to the level of exposure itself and not to the probability of its ectual occurrence, since this is assumed to be unity.
~
The ICRp's principles have been adopted internationally, including their
[ *. incorporation into the joint IAEA/1LO/WH0/0 ECD /WEA Basic Safety Standards for Radiation protection (BSS) (2), and by most national organizations.
1.03 The situation is rather differer.t for exposures that are not presumed to be certain to occur. The ICRp principles do not apply directly to such situations, and although radiatien safety objectives _have been developed at the national level for some sources - - notably nuclear reactors _,- no international consensus has yet emerged on safety principles for sources potentially causing exposures in general. This document is not intended to discuss all the details or ramifiestions of the unified scheme proecsed I
The tenn ' radiation safety' is uset in this consultative document to encompass all scientific disciplines engaged in the safety of sources involving exposures to ionizing radiations; they include the radiation protection and nuclear safety disciplines.
2 The +== enntrol is used in this document to mean the act or fact of exercising restraint, rather than checking, testing,or verifying.
3 ILO, International Labour Organisation; WHO, World Health Organization; OECD/ MEA, Nuclear Energy Agency of the Organisation for Economic Co-operation and Development.
l
.-. .-... -. ~ _ _ _ - . . ..- ..- . . . . . . . . _. . .
i
-2 I herein. Instead, the princip1as and fundamental ideas of the proposed unified scheme are presented. After the consultation process and appropriate review and editing, the document is intended to be published in the IAEA Safety
)
Series.
- q.
4 s
S e
e 0
J e
f I
I l
l l
.. . . . . - . . . - - . - _ _ . . . . _ . . . . _ , _ _ _ _ . . , . . . . _ _ . _ . _ _ _ . ._ _ ._,_.____..-. . _ . ~ _ _ _ . . _ .. ,
. 3 Section 2 BASIC UNDERSTANDING 2.01 There are confusing f actors in discussions of exposures that are not 1*
certain to occur. One such f actor is that some terns, such as uncertainty, probability, risk, limit, objective, goal and sensitivity, are used with different connotations by different authors. To minimize this confusion before proceeding with the analysis of the present situation, this section defines how certain te_rms are used in thi_s. document.
It is important that the reader be clear on how these terrs are used to avoid misunderstanding the intent or the implications of the proposals presented.
Unc e rtainty *
, 2.02 Uncertainty is a concept expressing the incompleteness of the knowledge e of or the imprecision in, an estimate or assessment. Any assessment of
, radiation erposure is subject to uncertainty, and it is tempting to assume that if the probability of exposure is less than one, the uncertainty in the asserrment of radiation risk will be greater. Within the general uncertainty there are usually several dif f erent contributors. These include:
imperfect knowledge of the system to be analysed; the limited degree to which a sedel of the system simulates reality;
- imperfect knowledge of a model paremeter; the intrinsic variability of quantities; and incoepleteness of the knowledge of exposure scenarios and pathways.
2.03 Some types of uncurtainty are amenable to mathematical treatment, but others are not. Some aspects of modelling uncertainty, therefore, may be dealt with only by making cautious, but reasonable, avrumptions. The propagation of uncertainty through the entire assesrment proceos is a dif ficult but necessary task, since point estimates of quantities do not carry suf ficient information about the uncertainty assoetated with the estimating process.
- 4 -
2.04 lt should be noted that uncertainties exist independently of the assessment method used. Whether modelling is performed 'deterministically' or
'probabilistically', the uncertainties are present. probabilistic approaches explicitly account for uncertainty, since probability theory is intended f or situations of incomplete knowledge. Moreover, uncertainty can .in'jf act be expressed mathematically through the concept of probability.
Probability 2.05 There are two customary interpretations of the concept of probability.
(a) ' Probability' in its interpretation as the mathematical limit of the relative frequency of occurrence of an event (the frequentistic _ _
int e rp re t a t ion ) .
/ (b) ' Probability' in its interpretaticn as the degree of belief that an event will occur (the rubjectivistic interpretation) .
~
2.06 Cenerally, the f requentistic interpretation of probability is used to quantify uncertainty arisins from possible stochastic variation, whereas the subjectivistic interpretation is used to quantify uncertainty arising from inaccurate knowledge of deterministic (' fixed') values. Both types of probability can be quantified on the basis of sample evidence (and are thun termed ' objective'), if ruf ficient experience is available, or on the basis of e xpert judgement (and are then termed ' rubjective'), if there is no, or not sufficient, statistical information.
2.07 in the f requentistic interpretation, the probability, either objective or subjective, is usually termed ' estimates'; while in che subjectivistic inte rp re tat ion , it is urually termed ' confidence levels' or ' confidence intervals'. In practice there will of ten be a mixture of the f requentistic and rubjectivistic interpretations of the probability concept. It will, however, almost always be possible to classify with respect to the dominant type of uncertainty. If this is not possible, the contribution of each type of uneartainty has to be modelled separately. Dncertainty' analysis must usually cope with quantities which are subject to uncertainty of both types.
To allow meaninstu t .8nterpretation of the results of an analysis, it is essential to discriminate betwerm both types of uncertainty in the course of the analysis.
5-2.08 it should further be noted that the subjective probability is not an arbitrary value assigned to the probability. Subjective probability values are assigned on the basis of the information available, which includes scientific knowledge, expert judgement and historical experience.
f' l'
2.09 The validity of subjective probabilities are dependent upch maintainin6 coherence and consistency when assigning probability values to events.
Coherence simply means that all of the probabilities are assigned in i conpliance with the rules of the calculus of probacility. These state that the assignment of probabilities is coherent only if the complement of an event with probability p is assigned a probability of (1-p); that events which occur with a higher frequency have larger probabilities; and that if event 'a' is ecre probable than event 'b', and 'b' is more probable than 'c', then 'a' is more probable than 'c'. Information is available in the literature of l probability'and statistics on the calculation of probability and the concept of coherence (3).
J 2.10 Moreover, the use of rubjective probability is acceptable as long as the quantitative value assigned through 'best estimates' or ' engineering judgement' is consistent with the quantitative value of the relative frequency in situations where more information is available. Thus, the probabilities !
assigned for various events will be consistent and continuous, and low probability events can be integrated with higher probability events into a complete analysis of the options under consideration. l l
Risk 2.11 The term ' risk' has been used previously with several different connotations and definitions. For example, the Pssic Safety Standards for Radiation Protection (2) adopted the term risk as introduced by ICRP (1),
which is precisely defined in the IAEA Radiation Protection Glossary (4) as
"...the probability that a ;iven individual will incur any given deleterious stochastic ef f ect as a result of radiat ion exposure". Moreover, the ICRp has recently recommended that' its definition of risk should be extended to nake it cler.r that it not only includes the probability of serious detrimental health effects given the dose, but also the probability that an event gives rise to a
dose, or more strictly, to a distribution of doses (S]. However, the IAEA's NUSS (6) documents use the term risk "in the general sense of a combination of probability and consequences of an event rather than restricting it to a quantitative consideration of probabilities and consequences". ' Risk' has airo been used in the literature in ics colloquial meaning, and also W" as equivalent to ' consequences'. In some of these definitions, the*tett incorporates arrects related to the perception of importance for various outcomes, and even of attitudes to the source of the risk, but none of them seems entirely appropriate, however, and all are difficult to describe and use in a consistent manner for the purposes of quantitative risk comparison.
2.12 For radiation saf ety purposes there is a need for a single, simple mathematical definition of risk that can be used in all situations of radiation exposure. Thus, it is tempting to use the extended debinitions of risk previously recommended by ICRP. Risk is therefore defined in this document as the probability that a serious detrimental health effect will 4
occur to an exposed individual. Mathematically, this definition can be expressed as:
R = p(D).p(efflD) where R is the risk associated with an event which has a probability p(D) of giving rise to a dose D and p(efflD) is the probability of a serious detrimental health ef fect arising from the resulting dose D. However, many processes give rise not to a unique dcae but to a distribution of doses.
Therefore, more generally, we may define the dif ferential risk due to a tiven event i as:
dR =p p(D) p(efflD) . dD where dR is the probability of a serious health effect occurring as a reruit of a dose between D and D + dD, pg can be the product of several probabilities of events leading to the exposure, including the probability of occurrence of the initial event and, p(D) is the normalized probability density distribution of delivered dose (i.e. its integral is unity) given that
. ?.
the event has occurred. Whence the probability of incurring a detrimental health effect as a result of any given dose dae to a given event i is:
R = p .f p(D).p(efflD) . dD l'
The total risk is in principle given by R = 1 - R(1-R ),
where'R{is risk due to the event or process i. If R are rmall and evenis are independent then total risk is simply the linear rum of the risks of all events:
R = 2 R.1 = { p. . fp (D) p(efflD) . dD 2.13 The term ' serious detrimental heelth effect' used in the definition uf s
risk includes:
- non-stochastic ef forts that can lead to death; 4
- fatal cancers;
' '- - major hereditary ef fects in the exposed individual's next two generations of descendants.
I Whether or not these three types of effects should be treated as equivalent
- may depend on the application of the risk concept and is discussed further in Sections 4 and 5. The dose-response feinction p(ef flD) is discussed in more detail later in the document in para 2.20 and in Section 5.
Societal and collective risk: detriment 2.14 Even in cases where the risk to any individual due to a particular radiation source is deemed to be acceptable, it may be that the predicted hat %
to a population is considered unacceptably large. There is therefore a need to express clearly the concept of predicted harm to a group of individuals.
Such a concept has been referred to in the literature on safety as ' societal
, risk' or ' collective risk'. However, risk as defined in this document is a
probability, which is not an extensive quantity, and individual risks cannot ,
j simply be summed to deduce a risk to the group. (Section 4 refers further to this matter). The ICRP introduced, and the IAEA adopted in the Basic Safety l Series Standards, the concept of health detriment to express the expected harm l l
to a population from the biological effects that can appear after.an i e xpo su re . Detriment is defined as "the mathematical expectatio'n tif the harm (damage to health and other effects) incurred from the exposure of individuals or groups of persons in a human population to a radiation source, taking into account not only the probabilitics but also the severity of each type of deleterious effect" (4).
LLaits, objectives and goals 1
l
'. \
2.15 A safety limit is the /alue of a safety related quantity which is not l r
4i
\, to be exceeded (4); this is assumed to occur with a given degree of i
~
confidence. A safety 1imit, therefore, is not a safety objective but is just
, a constraint on the objective. Depending on the actual problem, a limit may be forwolated as a curve (e.g. a limit line), a set of values or a single value.
2.16 A safety ob_ieetive is usually a qualitative expression of a destro (e.g. ' doses rust be kept as low as reasonably achievable') but it can also be 4 a value of a quantity which has been forwulated as a target or a numerical bench-mark of performance (as opposed to a limit which is not to be exceeded). 7hus, objectives are normally more restrictive (in terms of dose incurred or aucunt of radioactive materials permitted to be released) than the
, applicable linits. Depending on the particular circunstances, quantitative objectives can also be expressed as a curve, a set of values or a single value.
2.17 The word Lqa_(
s is sometimes used as a quasi-synonym of 'obj ective' .
However, some national authorities have used the expression ' safety goal' as a qualitative _ expression of the level of protection and safety that the nuclear industry should strive to achieve. In this context, a goal is supported by quantitative objectives.
9-Sensitivity 1
2.18 The concept of sensitivity is normally used in modelling as the relation between changes in output quantities and changes in input quantities for the model. It shows whether and how strongly the components ,of the model influence the results of the estimation. Sensitivitystudiesare{performedto assess the effect of variations or changes to the components or input data for the model in question. Because of the uncertainties inherent to probabilistic approaches, which arise f rom both the imperf ect knowledge of and the stochastic nature of the processes, the sensitivity analysis is a useful tool for qualitative investigation of models used in these approaches and for helping to judge their acceptability. Sensitivity analysis and its results can be an important factor in decision making processes, mainly in cases for which the uncertainty range is rather high.
Graphical representations a
J
+
2.19 Some of the concepts explained before con be represented graphically.
A discussion of some typical graphical representations follows.
Functional Curve 2.20 A functional curve shows in graphical form the mathematical relationship between a dependent and an independent variable. For instance, a dose-response function curve is obtained when p(ef flD) is plotted against D 1
(Sec Fig ,1), where p(efflD) is the probability of a serious detrimental I health ef fect for an irradiated person or his or her progeny, given that the person has received a dose D.
Probability density function curve 2.21 Figure 2 illustrates the special case of the function curve for which the probabilities of the occurrences of events or instances of items with I dif ferent values are plotted against those values. For instance, the well known normal (Gaussian) bell-shaped curve, when normalized, is a probability density function curve.
..,s._ .. , - , - -
Cumulctivo probcbility density function curvo i 2.22 Figure 3 illustrates the integral of a probability density function which shows for each value of the variable considered the probability of occurrence of an event or the instance of an iten correspondins.to,that value f
or a lower value. - 1 p(x s a) = j' p(x') . dx' For instance, the integral of the nornal (Gaussian) curve is a sismoid from.
zero (indicating that the probability of obtaining a value louar than the lowest possible value is zero) to one (indicating that the probability of obtaining a value equal to or lower than the highest possible value is unity).
Complementary cumulative probability density function curve (CCDF) 2.23 Figure 4 illustrates the complement of a cumulative probability density J
. function. For each value of the variable studied, this shows the probability of not obtaining an evert or itus with that or a lower value. For instance, risk curves with the probability of the event on the y-axis and consequence on the x-exis can be depicted in CCDF format. Figure 5 is a risk curve in CCDF format.
p(X i b) = f.,P(X') . dX' LLait curve 2.24 In some cases, a limit is not set to one single quantity but to cambinations of quantities. In this case, a curve can be used to illustrate which combinations of values are acceptable. Such a curve can, but need not, refleet a mathematical relationship between the quantities involved. For instance, a risk curve in CCDF format could be chosen as a limit curve (see Fig. 5) but Fig. 6 shows a LLait curve which is not derived from an explicitly given mathematical function.
1
-1 10 g'. .
1
-3 ~
$ 10 .
O '
. 1
~L '
~
> .l 10 -
w i
.c 10
-5 8
- a. <
1 10-6 ,
l l
1 10' -
~
20~ 10 ' 10~ 10
~
10'I 1 10 Dose (Sv)
Tig.1 : Individual probability of severe harmful effect versus l radiation dose.
1
- 12 l
(x) q*
4*
4 Probability denstty of qcantity x 0 x Figure 2: The probability density (t:nction
- 1. .
P(x 'd a) - - - - - - -
l l
l l
a x Figure 3: The cumulative probability density function
f* ,
(. \
l 1
l i
)
l l
i
\
- * ~ - ~ ~ ~ - ~ ~ ~ ~ ~
~
, P(xff a) I l
l l
l 1
t I
i i
I
& X Figure 4: The complernentary cumulative probability density fun :: r i
1
1 J
-3 .
,. - e ,
, i r ..
^ ted rs -
e k %
1 %
o thacceptable risks
~ %4 ~
f 1e"
[ '\
t ec i -e -
._ <\% -
3 P
\s%
g,,-.
Acceptable risks c s
,,.,' I I l,i s i to 108 tes , ga N. Mmber of. Fatalitics Fig. 5 Canparison of proposed collective safety goals i
l
,, . . . . .- .- _ . -. - .- - - . . . . . - - - - - - - - - ~ - - - - ~ - - - - - - - - - - - - - - -
Section 3 INTRODUCTION - FRAMWORK OF THE REPORT f*
3.01 National and international standards for protection again}'t exposure to ionizing radiation are based on the system of dose limitation recommended by the ICRP. These recommendations cover all situations in which radiation exposure of people is assumed to occur and the source can be controlled. The ICRP basic system of dose limitation, however, cannot be applied in its current form to control sources that may or may not give rise to exposures .
More precisely, the ICRP dose limitation system applies to doses that are presumed to occur with certainty, but not to situations where th,ere is a potential for a dose but not a certainty. '.
6.02 As a simple example of expoouces that may or may not occur, consider
/
, the potential exposure pathways from a radiation generator enclosed in a endiotherapy room. On the one hand, people will be exposed to the radiation beam that penetrates the shielding. The levels of exposures and the associated radiation risk depend on variables euch as source characteristics, the shielding attenuation factor, the location of exposed persons and the exposure time. On tt:c other hand, these rooms usually are equipped with interlocks to prevent inadvertent exposures. These interlocks might fail, l l
with the result that someone might enter the radiotherapy roc: when the generator is operating. Such ' potential' exposures are not certain to occur, but have a probability of leading to doses end, therefore, also entail a radiation risk.
3.03 The risk due to sources of potential exposure can only be analysed using probabilistic concepts. In *.his report, ruch sources are called
' sources of probabilistic cxposura' . The orpression 'probabilistic exposure' is preferred to ' accident' since the occurrence of events can be foreseen to l
which probabilities of less than unity can be assigned, yet which would not I generally be considered to be unforeseen accidents. Examples of. this type of event would be movements in the water table around a radioactive waste repository, or the thef t of a vehicle transporting a radioactive source, or the misuse of an irradiation source. .
3.04 Furthermore, at nuclear power plants, for example, a so-called .
' routine' exposure may result from a number of separate i j dents. If these incidents are reasonably frequent, there may be a tendency to designate them a s ' n o rma l,' , whi le if they are infrequent, they may be included in the
'_a c c id ent ' category. The dif ficulty of classifying events as noreg1
_ q or accidental can be overcome by using the terms ' planned exposure _sf[ and
' unplanned exposures' . If a maintenance worker is exposed in performing .a job
> necessitated to deal with plant malfunctions (which, at the design stage, were envisaged to occur), he or she receives a ' planned exposure'; but if a corponent containing radioactive material is damaged and an intake of radioactive material by a worker results, he receives an ' unplanned exposure' 3.05 Ef fectively all sources give rise to normal exposures and could give rise to probabilistic exposures. The relative importances of thh two modes of exposure may dif fer enormously for different sources but in principle both modes should be considered for all sources. In the abstract it is not always J
, obvious how to make the distinction between the two modes of erporure, but it is alwsys possible to resolve the isrue for any particular source. This problem is discussed further in Sections 3 and 4.
3.06 Sume procedures for the assessment and control of probabilistic exposures at nuclear installations, and it. particular at nuclear power plants, have been developed in parallel to, and to some extent separately from, the basic recommendations of the ICRP on radiation protection. procedures for assessment and control relating to waste disposal also started to evolve separately but have now been tackled by extending and developing the ICRP i basic recommendations to deal with the particular problens of wastes.
Suggestions have been made for a unified approach to control for all these f
areas of concern, with common principles for dealing consistently and l coherently with routine exposures and potential exposures. This document suggests how to start working towards such a unified approach to radiation j safety.
Characterization of sources of radiation exposure 3.01 Radiation sources can be characterized according to their potential for harm. For exposures presumed to be certain arising from normal operation, the l
relevant quantities are the dose to the most exposed individual _and the total r '
l
, collective dose from the scure_e. For the low doses expected in normal operation the incremental dose received by individuals is assumed to produce a proportional incremental hat?. and the collective dose is theref e a measure of the total expected harm. For potential exposures there can also be identifiedaprobabilityofindividualharmandaprobabilitydikributionof consequences. These two features, the probability of individual harm and the probability distribution of consequences, characterize the source from the point of view of radiation safety. Other source features, ruch as the radioactive inventory of the source and its potential for dispersion and transfer to man, are subsidiary features of its classification.
3.08 it should be noted that these two features will in general have time distributions which might be very dif ferent according to the typ's, of source.
For example, exposure arising from the generator enclorure can only occur at the time of operation, while a radioactive waste repository is a source of !
$ probabilistic exposure over a time-scale that extends into the distant
. . future. A change in the safety level applied to the source will invariably introduce a change in these features and eventually in their time distribution.
The need for a unified approach to radiation safety 3.09 A unified approach to radiation safety is needed not only to enhance the coherence of the current situation but also because there are practical P
dif ficulties in applying radiation safety principles consistently and logically for a wide range of sources and exposure scenarios. Notably, in the l use of radioisotopes and radiation sources, there are few safety criteria for limiting the chance of incidence of accidental exposures. This contrasts with the number of recommendations, standards and criteria that are in use for the same sources but whose scope is limited to exposures presumed to occur with certainty only. On the other hand, there is a growing problem of inconsistency be, tween radiation protection and safety considerations in cases, such as ,forguelear power plants _, for which dif ferent principles apply *.o ._ _
exporures prerumed. to be certain and to potential exporures.
In fact, in some
, cases an increase in protection against exporures presumed to be certain implies a decrease in safety for potential exposures, and vice versa. Only a coherent and consistent approach to radiation safet) in general can cope with I these situations.
- 1S -
Recent international developments connected with a broad approach to radiation safety 3.10 in 1983 the IAEA published a report giving recommendations on criteria for underground disposal of solid radioactive wasteJ (7). The.rkhiation protection requirements given in the report are of tuo types: tho'se for use during the operational phase of the disposal facility and these for the post-operational phase. For the former phase, the document recommends that the ICRP system of dose limitation in its standard form should be applied for normal operating conditions. For accident conditions, the document recommends that national authorities establish appropriate standards. Two requirements are given for the post-operational phase: one is a version of optimization of protection, but taking into account the probability of doses being received; the other relates in a similar wsy to individual limitation and implies that the risk to an individual in the future shall not be greater than is now e regarded as acceptable for individuals.
s 3.11 An expert group, set up jointly by the Radioactive Waste Management Committee and the Committee on Radiation Protection and Public Health of the Nuclear Energy Agency of the Organization for Economic Co-operation and Development (OECD/NEA), has published a report (8]. This OECD/NEA study deals with all forms of radioactive waste, including uranium mill tailings, and with disposal options which may have radiological implications over extended periods of time. The NEA document also concludes that the appropriate form of ILmit on individual detriment is in terms of risk, where the document defines risk as the product of the probability of exposure and the probability that the dose received will have serious health ef fects. The WEA group agreed that the optimization principles should apply to waste disposal, but particularly emphasized that they should do so within the broader context of decisions on waste management.
3.12 In 1982, the ICRP established a Task Group on the Application of Basic Radiation Protection Principles for Radioactive Weste Disposal. The recommendations resulting from the work of this Task Group have now been published (5). The ICRP also states that the dif ficulty of applying a
system based on dose limitation can be surmounted for the disposal of solid radioactive wastes by moving to a system based on risk, in which tne probability that a person is exposed to a dose and the probability of hatt as a result of that dose are combined to give the ' risk'. This measure can be compared with limits on risk, by analogy with current limitson(dese. The t
ICRP recommends a numerical value for the individual risk limit.
3.13 Over recent years the approach to safety in the nuclear power industry has moved on from its early emphasis on the deterministic evaluation of safety j systems for compliance with engineering standards to more direct incorporation of the results of probabilistic safety analyses (pSAs) [9,10). In examining l the more recent developments with respect to nuclear safety, for comparison with radiation protection principles, it is essential to focus en the more l general probabilistic safety criteria (PSC) related to public health rather than the techniques and results of the probabilistic safety assessments j themselves. In 1986, the IAEA held a Technical Consmittee meeting on the
, Status, Experience and Future prospects for the Development of probabilistic Safety Criteria. The presentations at this toesting, as reproduced in the document issued, are a useful sumary of the current position.
3.14 More general reviews can also be found in the proceedings of a seminar y held by CECD/NEA to discuss problems of consistency between radiation, waste management and nuclear safety (11). The seminar, organized in April 1985, i 1
discussed the different views expressed by experts on radiation protection, nuclear safety and waste management. Surprising for the experts was the recognition of how far apart the three disciplines are and how waJeh reconciliation of theory and practice is needed. Improved capabilities for probabilistic safety analysis (pSA) appeared necessary for achieving optimum protection. The need was emphasized for comprehensive safety assessment of long-lived radwaste and use of the results in the management of radwaste hazards.
3.15 The OECD/WEA Comittee on the Safety of Nuclear Installations (CSHI),
acting through its subcomittees, addressed the entire range of current nuclear safety issues in 1985. Its subcomittee on licensing concluded that there was a need to review the results of many years of studies of the source term and severe accidents in respect oi their implications for regulatory
approaches,and held a special meeting on this topic in June 1986. The Senior Group of Experts on Severe Accidents developed a new status report and emphasized its preference for the prevention of severe accidents rather than l the mitigation of their consequenced. The principal working groups continued their activities in their respective fiel'is. Evaluation of operating s.
experience and feedback and the co-operation of the incident repodting systems of the IAEA and the OECD/NEA were given much attention.
3.16 The OECD/NEA Committee on Radiation protection and Public Health
( C RP PH ) confirmed its support of the activities of CCHI and the Waste Management Committee (RWMC), particularly with regard to occupational exposure and radwuste disposal. For improved communication with the ICRP a flexible mechanirm has been set up to aid in the interpretation and application of ICRP recommendations and to provide input f or future revisions. An expert group report en the implications,of nuclear safety requirements for occupational
'. il exposure wun prepared for publicaticn in 1986. It mainly reviews the
.~. ,
l\
e consequences in terms of doses of such activities as in-service inspection,
- - - - - . (
- _ equipment backfitting and maintenance.
3.17 There is a growing interest in evaluating the actual and potential radiation exposures associated with the transport of radioactive materials, which requires shipments throughout the world by all modes of transport. In 19 7 7 the IAEA undertook a programme of assessing the risk due to the transport of radioactive materials worldwide. The first step was to provide its Member States with a simple method to assess the effects of such transportation. In 19 79 Sweden of fered to assist by developing a model for transport risk assessment. Under the aegis of the Agency a joint ef fort by Sweden and the USA has led to the development of the computer code INTERTRAN, which provides a system for assessing the effects of transporting radioactive materials (12). In 1985 the Agency convened a Technical Committee on the assesrment of the Radiolos!. cal Impact of the Transport of Radioactive Materials. The proceedings (13), a report of the Technical Committee (14) and an IAEA technical document (151 based on them which provides a rtatement on the radiological impact of the transport of radioactive materials are available j from the IAEA. Further IAEA sponsored work on INTERTRAM is in progress.
3.18 Following the Chernobyl reactor accident in the USSR the IAEA Board of Governors requested the Director General to establish an erport working group I
to consider, over a longer period, additional measures to improve co-operation in the field of nuclear safety, including vays and means of further elaborating nuclear safety standaeds. Following such a decision, an Expert Working Group on International Co-ope. ration in Nuclear Safety and Radiation
_p rotection was convened by the IAEA in November 1986. The Working , Group ,
composed of 175 experts f rom 14 MemberStatesandobserversfroml~five international organizations recomeq1e1 that a comon safety culture should be strongly encouraged _by_.t_he IAEA, which may become a fonam in which saf ety experts f rom dif ferent cultures - such as_ radiation, protection experts _, ,
f nuclear safety experts and waste management experts - can meet in order to ,
j discuss - inter alia - common safety principles and criteria and even a common I The expert group also recognized that methodologies on
, safety _ language.
probabilistic safety analysis have developed significantly and endorsed the preparation of guidelines on ruch methodologies and additionslly'. supported the IAEA promotion of and assistance in the use of ruch analysis in Member States.
j 3.19 The IAEA International Nuclear Safety Advisory _ Croup (INSAC), in
( reporting on poet-Accident Review Meeting on the Chernobyl accident, has recommended that the 1AEA should strengthen its wor'x in promoting, assisting
,and facilitating the use of probabilistic ssfety as,se,stm,ent (pSA), by reviewing the techniques developed in Member States for the use of pSA, assisting in the formulation of guidelines for its- use and helping Member States to appi) ruch guidelines in order to enhance safety in all nuclear power plant operating modes.
l l
1 1
I l
i i
e Section 4 l
BASIC APPROACHES TO PROTECTION AND SAFETY
. f j*'
Current principles of radiation protection 4.01 The basic principles of radiation protection stem from the system of dose limitation recommended by the ICRP. The main feature of the system can be expressed as three interrelated components identified by the terms:
' justification of practices'; ' optimization of protection'; and ' individual
- me limitation'.
J;stification of practices
, 4.02 The first component of the system of dose limitation, tot 9eed justification, provides that no practice shall. be adopted unless its introduction yields a positive not benefit. A practice , ruch as the generation of electrical energy by nuclear fission or the radiosterilization of medical products, means the sum of all processes, industrial operations and actions associated with that activity which produces the benefit. The net benefit should be determined by assessing both benefits and ef forts, including the possible harm due'to radiation, resulting f rom the introduction of the practice.
4.03 Although the justification of practices is a fundamental concept of 7
radiation protection, decisions of this type usually have _many_ aspects,no,t
, direct _17 concerned with radi,ation protection and radiation protection is only one input to the decision making process. For example, the decision whether a country should proceed with a nuclear power programme would be taken at the highest political and governmental level, and would include considerations such as the availability of other fuels, the need for electrical energy, the strategic advantage of a diversity of pow-r sources, and the relative capital and operating costs, as well as the overall health and safety Laplications in normal operation and in potential accidents. Once the decision has been made '
to proceed, then practical radiation protection rests with the other two ! )
components of the system of dose limitation.
l l
I
,; Optimization of protection 4.0# The second component of the system, known as the optimization of protection, requires that the radiation protection applied to'sgsource of exposure must be optimized in order that all doses be kept as lo'w as is reasonably achievable 'ALAPA), econemic and social factors being taken into account. Optimization requires an evaluation of the various possible options for protection and a judgement of their different features against preference criteria. The feature. o which these criteria apply include the achievements in protecticn, such as reductions of the doses involved and favourable changes in their distribution in time and level of dose, and the ef forts, such as costs and difficulties, involved in achieving such protection. ,Since some of these criteria may be in conflict with others, evaluation for other than the simplest problems will require some kind of decision aiding technique to dif ferentiate betwaen alternative options from the point of view of radiation protection. One particular technique recommended by ICRp is that of I cost-benefit analysis -_-
(16), but it has been emphasized that this is only one s
\
way of quantifying some of the inputs to the optimization decision. Other l techniques, such as multiattribute analysis, are also being investigated by (
IC Rp .
i Individual Dose limitation 4.05 The third component of the system requires that the doses incurred by individuals should not exceed the dose limits recornended for the appropriate circumstances by the Commitsion. For occupational exposure the ef fective dose equivalent limit is 50 mSv in any one year, with additional overriding linits on the dose equivalent incurred in and committed to individual organs and ti s ru e s . For the ef fectiva dose equivalent for members of the public the Coweniskion gives two values in its 1977 Recossr.andations (1). The Comission's present view is that the principal limit is 1 mSv in one year. However, it is permissible to use a subsidiary dose limit of 5 mSv in one year for some years, provided that the average annual effective dose equivalent over a lifetime does not ekceed the principal limit of 1 mSv in a year (17). The limits for the public apply to a particular, althoush extensive, set of sources defined by ICRp to be all those not f alling within the categories of exp6ture to naturai radiation er oJ exposure as a patient to medical uses of radiation.
- 24 -
l l
4.06 The numerical values of the dose limits are based on a judgerent by the ICRF that an appropriate comparison is with the levels of risk associated with industrial operations generally regarded as ' reasonably safe'. Information on I these industries was reviewed by the ICRP, leading to the view 'thprt such I
~
industries had average rates of death to workers in the range 10 to 10~
per year. Bearing in mind that the dose limit marks the lower boundary of an unacceptable region and that everage doses are well below the limit, the ICRP decided on the numerical value for workers given earlier. Similar considerations were applied for members of the public, but also taking account of other f actors, such as the existence of special subgroups, including l children, the longer poter.tial exposure time and the involuntary nature of the i e xpo su re . Using accepted risk f actors for general populations, t,he current for members of the public of 1 mSv in a year' implies a l lprincipaldoselimit -5 constraint on the annual risk to a level of less than about 10 . These
, judgements by the ICRP have been implicitly adopted by practically all
> national and international orgsnizations, including the IAEA, and by national
[ '
authorities which have adopted the numerical values for the dose limits recomended by the ICRP.
4.07 Conceptually, the dose limits apply to the total dose to an individual. Most applications of radiation protection principles deal with particular sources, and unless it can be shown that the contribution from all other sources is negligible, the dose from any one source must not be alluwed ktobecomparablewiththeentirelimit. It was originally suggested by the L IAEA in Safety Series No. 45 and the same procedure was recommended by the !
ICRP (5,18) and confirmed by the IAEA in Safety Series No. ?? (19), that this allowance for other sources can be formalized by using a source upper bound j rather than the dose limit. This source vyper bound naast serve as a design i constraint for a source, and, conceptually, would therefore be the dif f erence between the individual dose limits and the sununation of other local exposures, regional and global contributions and an allowance for exposures due to future l activities. The ICRP has given guidance (5) on the fundamental principles which the source upper bound should observe but has lef t det ailed evaluation for particular circumstances to national authorities or international J
agencies. The IAEA has initiated a programe on the preparation of a Saf ety cuide on the Principles and Methodologies for Calculation of Global Upper )
l Bounds. i I
l
- 25 -
A possible extension of the current radiation protection principles 4.08 The ICRP dose limitation system is based on a number of radiobiological assumptiens. In employing a risk bas.ed system, which admits the probability of doses exceeding the dose limits an: entering the 'non-stochastic' region, i'
it is particularly important to specit y these assumptions. Figur&*1 shows h>w the individual probability of a 'stvete httmful effect' varies, given the.
standard set of assumptions, at dif'eiant 11vels of dose or, more precisely. t
(
levels of dose increment, since all ;1c sco' are additions to the background i level of radiation. An underlying st:plifying assumption is that all health effects represented by the curve in F g. 1 are equally 3erious. The curve shows three distinctive parts which could be described as follows:
(i) At levels up to a fraction of a sievert, only stochastic e'ffects occur. These include fatal cancers in the irradiated individual and severe genetic effects in the sacc*eding generations of descendants of
$ the irradiated individual. In this range, it is astueed that to any
. increment of dose correrponds a proportional increment in the probability of an effect for dose levtPL higher than the background dose level and independent of the dese' rate. This is not necessarily an exact representation of radiobiological data, but should be seen as a simplifying assumption which means, inter alia, that uncertainties pertaining particularly to low dosas and low dose rates are not shown.
The relationship of probability s (oc v4sk) to dose is therefore assumed to *;3 linear in this range and the slope of thi _surve (i.e. the risk factor in this region) 1s, for the sake of simplicity, saken to be
~ ~
10 sv f.
(ii) For doses that exceed O.. Sv. deliv'arst in a short period of time, non- stochastic ef f ect,s may occur. These effects are of increasing severity with increasing radiation dose. This portion of the curve 5
approximates to a sigmoid relationship. As in the stochastic range, the curve is an approximation and the exui t shape depends on a number of factors. If relevant, some rueh factors (ruch as the dose rate) could be taken into account for a particuler scenario.
\
2
- 25s-t'
< s (iii) Finally, at doses hither than about 5-10 Sv. delivered in a short period of time, p'ractically all irradiated individuals will suffer an '
acute radiation syndrome und eventually die es a consequence o'[ the irradiation. Again, the ensa.t shape of the curve would depend on ,
factors ruch as the availability of supportive treatment yp to
- approximately 15 Sv. AbovethisdoseJevel,allirradiatehpersons I
would die. As a conservative appronimation, therefore, the relationship can be asrumed to approach asymptotically a__probacil_ity_of unity f or doses higher than about 5-10_,Sv,. i l
4.09 The ICRP has made some progress towards solving some of the ;
difficulties in achieving a unified approach to radiation safety for the purposes of extending its radiation protection principles to cover the probabilistic arpects of the disposal of solid radioact3.ve wastes *-(5},. the ICRP, recognizing, as noted earlier, that the underlying basis of the dose
~
, limitation system was a judgement of thofacceptabilityorotherwiseofan ~
- implied rante of risks to individuals, has recommended a risk Limi_t of,10,
, per y.sar to be applied to probabilistic events. Thus the safety analysis of a waste repository should be able to show comp 1/'.ance with the dose _lipit for norsql events and with the risk limit for ptbbabilistic events.
i 4.10 There has been mch discussion of the exemption f rom regulatory _ concern, of sources of routine _a,xp_osurejiving ply, trivial doses. The 4_eneral r
'g ? > conclusion ,is thatyyuch exemption. t is poss_i_ble.ifj oth .the maximum,ind_ivi.d,ual s
, dosehthe collective dos.e are. bot _h very_ ems 11, and the sourcs has been properly defined. It might be possible to apply a similar principle to
- sources of potential exposure but investigations of this approach have not yet I been reported.
Current principles for safety assassment 4.11 As has been mentioned before, almost all sources have the potential to cause accidental f:rposuran and their safety level should be saaessed against I criteria that take account of th's probcbilistit nature.'of those potential e xpo sur e s . However, in. practice west of the development of ruch_ criteria has i taken place in the context of the safety eve.luation of nuclear power reactors, )
1 so this section necessarily concentrates on the cri'.oria being forrautated in l
. s l this area. l 1
i e -, ,_ _ _ __
- a ? .'
4.12 The study of nuclear reactor saf ety developed f rotr
- hat of conventi.onal ' .
. ~ . _ _ _
- safety and initially inherited the essentially detet ninistic concep *-_- . . _' that_
discipline. Engineering safety standards were set either as a rew.}t of experiraents and tests o7 more tubjectively using engineering judgement. An exampleofsuchadeterministicapproachinearlynuclearreactogsafetywas the so. called,maximun. cred_ible accident,' criterion. This wv'i sk? ect=d c n the basis of engineering judgemei.c of the worst accident that could ressenably bw expected to occur and the engineering features were designed to cope with it.
The criterion then ey lved.to.the concept of ' design basis accident',
maintaining. however, the initial approach.
4.13 It became apparent, however, that there remained some probability of accidents occurring and not being controlled by the safety-related engineering systems. These accident sesnarios and their causes, probabiltties and consequences were studied using the technique of probabilistic saf ety assessment (pS Al. The results of the most general pSAs can be expressed in many ways but essentiall'y they give the probability M stributions of the 1* .
various outcomes, includingthepublichealthimplicaj. ions, in tems of_
expected numbers of prompt deaths and injuries, t. umbers of potential oe' tyed .
cancers, genetic effects and non. fatal effe. cts, resulting from single un nts, initiators of accidents, or from a combinati)n of various accident scenarios.
S 4.14 in order to evaluLte the results of__ pSA, it is necessary to establish criteria for comparison . Unlike f o radiation protection for normal '
[
exposures, there has been no unified system of criteria internstionally recoteended and nationally adopted. Proposals have been made by many authors for criteria of various types based on dif fering judgements; sotne of thes_e have ueen or are being incorpora_ted t into regulatory recuirements or gui$ance _
( 20,,11,2 2 ) . In this report it is only useful to concentrate on the most general probabilistic safety criteria (PSC) related to public health and safety.
4 15 There are several different starting points for the judgement of tho level of risks either to be achieved or got to be exceeded. These include the following:
J
t
/,
- The most expos'ed individual should bear no significant additional l risk to life and health.
- Society should bear no significant additional risk to the life and health of any of its individual components.
- Risks ~and expectation of ham arising from nuclear powerhlants should be less than or equal to the risks and harm due to viable f
competing technologies.
4.16 These starting points can be used to derive several different kinds of criteria. The most useful foms for this comparative study are:
- Limits on individual risks that should not be exceeded:
Objective,s for individual risk at which designers and regulators should aim; i
- Limits for ' societal risk', expressed in terms of relationships b M en probability and numbers of people affected, in various ways; j - objectives for ' societal risk'.
There may also be general criteria of a dif fe.ent fom; for example, a requirement to balance the efforts for improving safety against the benefit achieved in tems of risk reduction.
4.17 For comparison with radiation proteccion principles, the most useful fom of criterion is that of a limit on individual risk. Objectives can be derived from the radiation protection system, for example as design targets, but only for specific situations as a result of a particular or generic
{ optimization study within the appropriate constraints. These objectives are not generally expressed in tems of individual risk, although the possibility of deriving comparable targets and expressing them in such a way could be further investigated.
4.18 However, a form of PSC is that expressed in tems of a ' limit on societal risk'. The critation is usually a figure expressing the relationship b een the frequency of occurrence 4
- scenario and the magnitude of the_
,consequenees expressed as the_ numb 3 t, fatalitie_s Exemples of such criteria ate shown in Tig. S [9,101 A comm. f eature of these criteria is that the f requency decreases more rapidly than the size of the consequence increases.
In this document an attempt has been made to make a comparison of sociosal l
l
risk with radiation protection principles although there is no direct analogue (see Section 1).
Examples of current safety pr tices conforming to current radia i,on protection principles .
4.19 Several regulatory authorities have adopted radiological criteria for the safe design of nuclear installations, in direct analogy with the individual risk limitation based on the principles of the ICRP, Thus, in this United Kingdom, the Nuclear Installations Inspectorate (NII) (23 ) expresses its principles as a set of ' assessment reference levels' that can be summarized as follows:
\
"The done equivalents received by the public from direct radiation or release of radioactive material due to accidents arising from a discrete f ault sequence which is judged to have a frequency of
'. occurrence greater than once in a reactor lifetire (about 30 years) should be no more than 1/30 of the appropriate annual dose limit; for s l frequency less than once in a reactor lifetime but greater than once in l a reactor progranne (of about 100 reactors), no more than the annual dose limit; for a frequency less than once in a reactor programne, no more than the appropriate Emergency Reference Level of dose (ERL)".
There is a further requirement that the occurrence of any accident sequence leading to exposures in excess of the ERL be made as remote an eventuality as reasonably practicab,le. The principles are further discussed in Section 5. )
4.20 Another - similar but conceptually dif ferent - example of such criteria is that adopted in 1986 by the United Otates Nuclear Regulatory Commission (USWRC) [24). The USNRC has established quantitative objectives. These are based on the principle that nuclear risks should not represent a significant addition to other societal risks. Thus, the risk to an average individual in the vicinity of a nuclear power plant of prompt f atalities that might result I from reactor accidents should not exceed 0.1% of the sum of prompt fatality I risks resulting f rom other accidents to which members of the population of the USA are generally exposed. Furthermore, the risk to the population in the
vicinity of a nuclear power plant of cancer fatalities that might result from the operation of the plant should not exceed 0.1% of the sum of cancer fatality risks resulting from all other causes. Mean value estimates are used to assess consistency with these objectives, l'
4.21 Recently, generaldesigncriteriaforfuturenuclearpoweriplantshave been developed in,I_t_aly [25). They can be summarized as follows: for normal operations of the plant and for transients occurring with a frequency of more than once during the plant life time, the effective dose equivalents to individual members of the public must be less than or equal to 0.1 mSv/a. For
~
accidents and trenstents with an annual probability of between ) x 10 and
~
10 , the effective dose equivalent must be less than or equal to 5 mSv per event. For accidents and for all the design basis accidents with annual
-3 -4 probabilities of between 10 and 10 , the effective dose equivalent must be less than or equal to 0.1 Sv/ event. The annual probability of exceeding
-5 -6 thu core coolability shall then be 10 /10 .
e J
4.22 An example which corresponds closely to the aerumptions of Fig. I comes from Argentina [20). The competent authority estimated that an annual
~~
-6
( s ource ) risk upper bound of 10 for accidental exposures arising f rom a nuclear power plant would be consistent with the current approach to radiation safety. Since it would be very dif ficult to identify all possible accident sequences, the authority asrumes that a tenth of the relevant sequences have to be identified, and has assigned an annual (scenario) risk upper bound of 10' to them. As each sequance may reruit in dif ferent doses, a criterion curve was adopted, which is a relationship between the annual probability of occurrence of the sequence and the expected individual dose, each point of the curve representing a constant level of 10" risk; the criterion curve enforced by the competent authority in the year 1979 is shown in Fig. 6. The authority has isrued a regulation applying that criterion to accidents (26) and an additional regulation on ' failure analysis of risk evaluation' (27). ,
The main features of these regulations are the following:
1 (1) The applicant for s . clear power plant licence shall identify relevant f ailure sequences wt. in the event of their occurrence, would deliver a radiation dose to members of the public.
1 l
! l I
l 1
i . . _ . _ _ . .- .-
31 -
(2) The probability of occurrence of each f ailure sequence, as well as the corresponding radioactivity of released radionuclides, shall be assessed by means of using event and f ault trees. A list of criteria that must be taken into account for the assessment includf
- the following:
- The failure analysis shall systematically encompass all foreseeable failures and failure sequences, including the common mode failurea, the f ailure combinations and the situations exceeding the design basis. .
- A f ailure or a f ailure sequence may be selected as representative of '
a group of failures or of failure sequences. In such a case, the f ailure or f ailure sequence to be selected f rom the group shall be that delivering the worst consequetices and the analysis shall take into account the rum of the probabilities of the f ailure or f ailure sequences in the group.
The analysis of failures or failure sequences or of any part thereof shall be based on orperimental data as f ar as is possible. If this cannot be done, the valuation methods must be validated by means of appropriate tests.
(3) The doses to the critical group that would result from the release of radionuclides due to a failure or failure sequence shall be assessed by accepted methods. The assesrment shall take into account the meteorological conditions of dispersion at the site and their probabilities. The assesrment shall not take into account the eventual application of countermeasures, even if they are foreseen in the emergency planning.
(4) The annual probability of occurrence of any f ailure sequence, if plotted as a function of the resulting effective dose equivalent assessed as indicated earlier, shall result in a point located outside the non-acceptable area of Fig. 6.
Section S LIMITATION OF INDIVIDUAL RISK t'
$.01 Following the discussion in Section a. it seems reasonab'leTt o focus on the idea of a limit on individual risk as one necessary, although not sufficient, requirement for a unified approach
.-~
to radiation safety in general
- and, pa _cularly, to the control,of probabilistic exposures. It was explained that the ICRP has made a start towards developing this concept in the context of radioactive vaste disposal ($1. In this Section the possibility of developing the concept in more general terms ic explored.
5.02 The essence of the ICRP proposal (S) is to achieve compatibility with the current system of dose limitation by specifying an overall limit for the risk to any individual, and to deal with the questice of the extent to which e safety should be improved below this limit by extending the concept of optimization to include the consideration of all risks to the exposed population. While this is conceptually attractive, and the two ideas of individual limits and optimization are developed and elaborated in this Section and the next, it is necessary to recognize that "societal risks" have I
{ been considered by the nuclear safety cotmaanity in terms of criteria f ramed as
' societal risk limits', or ' societal risk objactives'. This arpact , which would be outside the direct extension of the requirements of individual dose
{limitationandoptimization,isdiscussedinsection7.
5.03 The use of probabilities for evaluating the radiation safety of some practices notably nuclear power plants, presents some practical problems which need to be solved for a successful implementation of the methods and for the subsequent decision making process. Some of these problems were not easy to solve under the mainly deteministic framework with which nuclear safety evaluation initially evolved in many areas of the world. These problems are particularly relevant for the case of limitation of individual risks, and include the following:
l l
l l
- 33 -
(4) There are 1nany uncertaintites involved in PSA. The consequent lack of confidence in the result should be reflected either in the destee of conservatism to be used in establishin5 the relevant objective or when j comparing the results with the objectives. For instance, the use of risk ;
limits as a saf ety objective rather than as a constraint to the 9 bjective y.ay l
present difficulties in this context. It should be clast therefhethatrisk
, limits can neither be interpreted ,as. objectlyes .n c_ars_ste,ls_but ,just as a boundary _of _a forbidden region _and they sh;uld inerrporats the necessar;'
conservatism to cope with the expected uncertsinties, l (b) Currently, there are no standardized tools svailabh for performing probabilistic assessments. 1.arge variations in the tmhs car be obtained for the same situations if the methodology and the bout. h y/ conditions for the I analysis are not specified in sufficiant detail. Thies sap Troduce the undesired situation that two analyses might show etiNnet u::mpliance or non-compliance with a defined objective. The art r.tJxm tt Mi3s pec51em usems ,
I e to be ti.e development of standardized probabilist.it safett enesessment procedures in paralle__1 with the_estaklishment sf T j.ybabiAstut safety f critsria, and to incorporate them into the relevani rwsu'bsrt3 mms . This howevei, would tNn counter to the regulatory phlbesphty in-----~~ -
me countries l which adopt a non-prescriptive spproach.
(c) Another problem is t5at cf measurability ent as c e ounta'bilit y of risk.
For exposures expected to occur with a probabi3ity of unhy, the use of the quantity termed the ' effective dost equivalent' wus reemamended by ICRP and established in the IAEA/WH0/ILO/NEA(CECD) Basic Safety Standards for Radiation Protection as an indirect measura of the risk incutmedt by exposed persons.
Such a quantity is '_nessurable', alhit 'by say of sthe t- tielated physpal quantities and various assumptions and hypotheses, uno t.hrerefore can be accounted for in relevant records with legal sistuv. The situation is rather dif forent for exposures expected to occur with a pt tiha.bi,E.ct. y lower than one:
the ef fective dose equivalent to be incurred if the expersure is actually delivered does not laessure ths risk (since there is a chance that the exposure does not occur) and no other ' measurable' and ' accountable' quantity does not exist. The probability of the exposure or the combinatiwn of probability and g cannot be ' measured' in the instrunwrntal sense and sury not be accepted as quantities for record purposas, and a posteriori compliaree would not
- 30 -
therefore be legally ' demonstrable'. There are ways of solving this legal
_ _ . - - -- . __ _ _ . - ~ ____ _
problem but they have not been explored yet (see para. 510).
5.04 in view of the afore mentioned problems some nuclear safety experts have expressed caution concerning the application of a risk limit $ ion system
~
to nuclear power plants. They prefer to focus on general probabiFisti.c objectives. Other radiation safety experts have felt that in spite of practical dif ficulties the es,tablishmen_t of a, basic _ phi _lo_sophical_f ramework of
, risk limitations should be encouraged. The eventual maturation of presently l available methodologies and systems will solve the present practical problems.
The derivation of a risk limit
=-
5.05 The ICRP system of dose limitation has been used as the be is for a rigorous control system applied to the doses to both workers and members of the public due to normal operations. The system for control of risks from e
e accidents, although less unified worldwide, has nonetheless been applied in a
. similarly rigorous manner. The attempt to unify and co-ordinate the control system should not be misinterpreted as implying that one or other source of risk has been insufficiently controlled in the past. Wor does the introduction of a new type of criterion for risk evaluation imply the discovery of new risks. These have existed and been recognized throughout.
5.06 The system of dose limitation is well established, but it desis only with events or processes which are nssumed, to occur with certainty. It seems to be rational that the introduction of a system of radiation safety for the class of events or processes which have a probability of occurring of less than unity should be done in a way which is consistent with the current system of dose Limitation, since the latter is widely used and found to ta satisfactory in its domain of applicability.
5.07 A possible method of incorporating accident situations into a risk based system of radiation protection would be to define a total individual risk limit for the combired eum of normal and accident events or proc _ esses..
However, this approach has two disadvantages: firstly, it could be construed '
to i;mply an allowab1e trade-of f of risk between normal and _ accident situations. It is not clesr that_the public or the nu_ clear communit.y would- - - - -
.)
39 -
favour such a trade off. Secondly, such a method would involve changing the curret{ system of dose limitation for normal events.
5.08 The simplest way of inco{porating accident events or proces.ses. in.to a.
__-_g
~
risk based system of radiation protection is to define separate limits for normal and accident situations. In this way, the current system of . dose limitation remains intact (as a special case of a more general risk limitation) and the radiation protection framework is extended to allow the treatment of accidental events or processes.
5.09 The ICRP was aware of the existence of probabilistic events when it set the numerical value of the dose limit for routine situations, so that
\
specification of a separate risk limit for accidents does not imply a cause.
for a corresponding reduction in the dose limit. Furthermore, the doses received by members of the public f rom normal operations very rarely approach 4 the dose Ibmit, and the low precision in predicting future situations does not
. warrant the refinement of modifying the dose limit (and hence the risk limit) for normal operations to accommodate the recognition of a risk limit for events of a probabilistic nature.
5.10 There i g further reason for_ maintaining some separation between the
, dose limitation and risk limitation syste7g. _A system for the_c_ontrol,of risks has to be implemented in a different manner from the current system for
' ~ '
control of doses, inter alia, because while it is possible to measure,or
, assess and record _ doses as they are received and to use this record to check for compliance, there is no equivalent method for measuring erposure to risks _
(see para. 5.03 (c). The system is, therefore, to be applied as a_ pro _syective _ i
.desi$-
- n to.ol. .In many cases checking for compliance will not be po._ssib.le, although for some situations par _tial checkin6_could be ba_ sed on, accumulate _d, e xpe ri_ enc e For example, complience.w__it..h a. r..isk limit may be verified to some
, extent at the plant level by checking that components and systems perfom with the necessary reliability.
5.11 In surenary, then, the simplest method of incorporating accident events or processes into a risk based system of radiation protection is to define _a.
separate risk limit for events of a probabilistic nature and to retain the
____._t._ __ -
I l
l 1
I l
l
- 36 -
current dose limits for normal operations. For waste disposal, ICRP has .
suggested sei. ting the risk limit for accident-- -situations to be of the sare
[ me&nitude as that implied by the dose Limit for normal developments._
I Risk limits for potential exposure of individual members of the,public
. l g'*
5.12 the current ICRP dose 1tmit of 1 mSv per year for members of the publi:
corresponds to a constreittt of 10 on the risk incurred in a year. For routine exposure below the dose limit, the only health effects that have to be considere'd are cancers and hereditary effects. If the dose limit now is converted into a risk limit, the constraint on dose itself is removed; other health effects such as acute death from high doses must therefore be taken into account. Thus the entire range of the dose effect curve in Fig. I has to be considered. It is_not_necessary__in_ principle to give all types of health effects,_the same weight; if, for instance, the loss of person-years were used as a wei5hting function, then acute death from high doses would have mor_e_
s weight than death fron,, cancer after a latency _ period. However, g the sabe simplicity it seems reasonable in the context of limits (not objectivei
, treat all health _ ef fects represented by the curve of Fig. 1 as equally serious; it is then possible to specify a single numerical,value_es a risk limit to apply to all events of_ _a_ probab.i.listic nature.
f , . _ _ _ _ _ _ - .
5.13 For consistency _with the general safety standards of dose limitation a
-5 risk limit of 10 in a year is propos & This limit would apply to the I
individual risk in the most highly e.x. posed cr_itical group from all sources of potential, exposure except natural radiation or medical administration of radiation to patients. It is important to be aware that the risk limit is the lower boundary of the region of unacceptable risk; a risk below this limit is not necessarily judged to be acceptable.
5.14 Since a member of the public can be at risk owing to more than one facility or practice, the ICRP introduced the concept of the source uppe,r _
bound [28) which is reconnended by the IAEA for limiting releases of radioactive materials into the environment (19). Thus, in addition to the ;
ri i limit, which ref ers to an individual, there needs to be a source related Eskupperboundwhichlimitsthe_individualriskcomin3 from a single _ source,
.e. a single facility or practice. The risk uppse bound is apportioned from l l
l l
the risk limit, i.e. is choa.in as some fraction of the latter. The fraction to be used may be different in different circums t anc es ; thus e smaller fraction may be chosen in waste disposal than in nuclear power plant,s owing to our uncertain knowledge about the variety of sources to which an individual 3.sy be eyposed in the future.
A risk upper. bound, allocated _to,y* source. is
- +
to be used in' design and regulation of a particular f acility in the same
, manner as current dose upper bounds are used. R enables accident scenarios to be consistently treated in the safety analysis.,. ,,
5.15 in order to assess a f acility on the basis of a risk upper bound, one may, in principle, proceed as follows: the set of scenarios leading to actual or potential doses is analysed, yielding a set of dose distributions or - in the discrete case - a set of doses. The dose distribution or thy doses together with their probabilities are converted to risks using the equation:
R=[R = 1p f p(D) p(efflD) . dD 1 The sus of the risks over all scenarios is compared with the risk upper f} bound.
The doses to_the_ individuals in ,a critical group have to be used in showing compliance with.a risk criterion. It could, however, be over-conservative to add the maximum risks if the critical groups are I I
different for the different scenarios. .
In this case one would first have to '
identify which of the several critical groups for the single scenarior is the I one with the highest overall risk, and then use that group to show compliance with the risk upper bound. _
5.16 In practical application of the risk upper bound method, some simplification may be introduced. In fact. design engineers tend to group scenarios into categories defined in terms of the physical characteristics of I l
the facility (i.e. those leading to f ailure of a nuclear power plant containment) or the commonality of consequences (i.e. those leading to increased groundwater flow from a waste repository), or defined in terms of a range of probabilities. Thus itjmay not always be easy to identify all the
_ relevant scenarios, to analyse them .ad to add their corresponding risks. In this case a further apportioning, in addition to that montioned earlier ,may be used, wh(ch consists in apportioning the_ source related_ risk upper bound to the risk upper bound related to a single category of sc.enarios.__for the r NJ
- 38 -
particular sour _ce. If one already has a f air idea of tbs totality of relevant '
scenarios pertaining to a category, an apportioning is also possible for each acenario.
5.17 In performing this apportioning one has to make sure that ,not too large f
a fraction is sesigned to a single scenario; sud it would also b(.p ruden t to assign only a small fraction to a single scenario in due regard of possible scenarios still unidentified. If careful judgement is used, apportiening to single categories or scenarios may be a useful tool to facilitate the probabilistic assessment of a source.
5.18 Compliance with a risk upper bound, either for a source or even for a single scenario, can also be shown by means of a criterion curve, which allows theuseofdosedistributionsordosesandprobabilitiesdirect1hwithout having to convert them to risks.
An example of such a criterion curve is shown in Fig. 6. The two axes represent the snnual doses from all scenarios j and the probability in one year that an individual receives a given dede. For the sake of argument, a risk upper bound of 10 in a yese is assumed in Fig. 6. We shape of the curve is derived from the dose ~ef fect relationship
~
, of Fig. 1. Every point on the curve represents the same risk, that is 10 in a year. The curve starts at the point where the probability is one (i.e.
occurrence is assumed to be certain) and the dose is about 10 Sv, because the risk is then 10 . Doses below 10 Sv correspond to a risk of less
~
than 10 and therefore automatically comply with the risk upper bound. The first part of the curve (log-los graph) is a straight line of slope 45 ; it represents the region of doses up to about 1 Sv whe e the risk h e to a given l dose is assumed to be proportional to that dose. Thus a dose of 10 Sv occurring with a probability of 10 induces the same risk as a dose of 1 Sv
-4 occurring with a probability of 10 . From doses of 10 Sv or greater, for which death is certain to occur, the risk no longer depends on the dose but only on the probability. Since the risk upper bound is assumed to be 10 in a year, the criterion curve is then horizontal at a probability of 10' '
in a year. In the short region between 1 Sv and 10 Sv the two straight parts of the curve are joined by some smooth path whose exact shape is irrelevant.
If the probability distribution (given that its integral is nomalized to 1) or - in the simplified case of discrete doses - the number pair of annual dose and annual probabil;ty of that dose falls evetywhere below the criterion curve, then compliance with the risk upper bound has been shown. If the risk l l
l l
- 39 -
e upper bound refet s to the source, then the sum over all scenarios ha-. to be used for the dose distribution; if, however, the risk upper bound has already been apportioned to a single scenario, then the dose distribution of that scenario is to be entered into th diagram with the criterion curve.
f*
T*
1 ,
N E3 -2 10 -
5$ UNACCEPTABLE REC 10ii .
x
>- .4
< 10 .
~
> Eh$ I
=
- y l td '
dg 10
-6 . . l/ . I E5 l 85 25
)
~ ~
10 ' 10 1 10 ANNUAL DOSE (Sv) _
l
. ~.. .-' . , -
i 1+ ,' $ * ) j) \ ,' a Fig . 6: CRITERION CURVE CORRESPONDING6TO AN P FROM ANNUAL RISK CONSTRAINT OF 'C ALL EVENTS l l
l Risk limit for potential exposure of workers I
5.19 Experience indicates that routine exposures, particularly exposures -
resulting from repair work, are the dominent occupational exposures, at least in the nuclear industry. However, the consideration of exposures'from repair ,
1 activities as ' routine exposures' is a matter of convention since those i exposures include stochastic contributions. Although accidental exposures of
_ _ _ . - _ . . . . _ _ _ , =__ __, . . - .
- 40 -
worket s occur that somstimes lead to doses in . excess of recommended limits, gtances,its which dose limits are exceeded and accidental _expnsures_are greater than routine exposures are rare. _W hen exposures are.less_than__the dose limit, they are usually recorded as part of the routine oClyp.ational dese.
I' 5.20 Perhaps because acciden' al exposures have not amountedt'o{a' substantial part of total occupational exposures, accidental exposures of workers has not been addressed by ICRP or the IAEA in the derivation of occuphtional limits.
However, for some workers, it may be desirable to establish risk limits,for exa yle, for those involved in the use of radiation sources, such as for
, industrial radiography. Following t.he approach described in the preceding section for exposures to the public (and assuming that the ICMP was aware that accidental exposures would occur above the recomended 0.05 Sv annual dose
. ; limit), one can establish a risk limit for accidental occupationa'1 exposures. _
-4 This corresponds to an absolute limit on totiti risk of 5 x 10 per year.
As with routine occupational exposures, a conservative approach to -
e -
s implementation should ensure that this limit is generally met by a significant
[ '
safety margin. The risk limit from routine exposures can be applied to the problem of accidental exposures, as the following example illustrates.
1' ., Derivation of design constraints from risk limits 5.21 A risk limit can be applied in safety analysis and transformed into an I equivalent limitation on familiar quantities such as the frequency of failure. For example: consider the case of a door to a radiation area that is fitted with an interlock to prevent entry when a radiation field is present.
,k This arrangement could be found, for example, in a nuclear power plant, in an accelerator laboratory or in a radiotherapy room in a hospital. _ Assume that j v\l , if the interlock fails, a person entering the area receives.a unique dose
/ l D . The use of the concept of risk in designing the interlock and in s': ' establishing radiation protection procedures is illustrated as follows.
The risk corresponding to the scenari0 (interlock fails, person enters area, instrumentation intended to confits the absence of a field f ails, person receives dose) is given by R = p(D) p(efflD )
as defined in Section 2. Assuming that all doses are in the stochastic region
. 48 -
for this example, the risk reduces to:
R. = p,.k.D 1 1 i where k is the dose conversion factor, p istheprobabilityofgbescenario being realized and D. is the doze delivered. The probability p'i ef the 1
scenario being realized is the product of three terms: the probability that the interlock fails (p g ), the probability that a person enters the area 1 (p ) and the probability that the instrumentation does not detect the presence of the radiation field (p ):
" l p
g
=p g p , . p, e
If a risk limit R is established for the scenario (for example. by apportioning the total risk limit for accident scenarios), then adequate ',
safety is determined by asruring that: ,
l
" j p .p,.p ,.k.D s R g
whence p
f p,.P, i R /(k.D) s i
5.22 An analy st who wishes to establish design constraints inr the interlock (
l may reason as fol.'rws: at some time during a working year a person will definitely enter the interlocked area, or even reveral people or one person j more than once. Therefore, the probability p, of entering = 1. If it is further assumed that the instrumentation will f ail and chat the person entering will definitely not use a survey meter, then p, = 1 also. Under these assumptions, p = R /(k.D) g It may be seen, therefore, that the risk _ limit for the pathwny can be
,, reduced to a design constraint. This last equation sets an upper limit on the probability of f ailure of the door interlock. The designer can then use this constraint in further considerations.
- 42 5.23 The possitility of instrumentation (cilut o, such es m:chenical f ailuro >
of the survey meter or even of its being misread, can be decounted for in the analysis by defining p more precisely as the probability that the meter j does not disclose the prese:nce cf a radiation field (either because the meter
! is not used, or bacavse it is reading low or being misread). ThisIillustrates l
) that the system c4n deal with human factors on an equal basis with) mechanical '
considerations.
Comparison with nuclear saf ety criteria 5.24 There are many nuclear saf ety criteria in use in dif ferent countries.
As ex mples three gets of criteria, described in, the following have been selected for analysis. These range from one set that is extremely close to the ideas proposed er.rlier in this Section to others that, althoug framed in a .tfferent way, have some elements in common.
J
, 5.25 in some cases, as described in Section 3, safety criteria have been established in terms of doses for each scenario or category of scenarios.
These are called Level 3 to distinguish them from Level 1 pSAs, which focus on accident sequences from an enalysis of plant design and operation; and Level 2 pSAs, which also censider the plant consequences and the response cf the mitigation systees, but stop short of assessing the public health s
implications. This is a national criterion (of the USNRC). In other j countries pSAs are not divided into these three arbitrary levelr. In these cases the reverse procedure to that described in Section 4 under Current principles of safety assessment can be applied to build a bridge between these criteria and the risk limit.
)
3 5.26 in carrying out an example comparison of nuclear safety criteria with I the extension of the ICRP system described earlier in this Section, the most
! striking example is that of the Argentinian authorities described in Section 4 ,
I l
' under the Exampics of current saf ety principles. The criterion curve adopted 1
, by the Argentinian authorities (see Fig.6) is precisely the same as that I d erived f rom the risk limit suggested by the ICRP.
I 5.27 Probabilistic criteria for individual risks have been adopted for regulation of reacter safety by the US puclear Regulatory Commission (24).
I The USNRC has adopted the following two quantitative objectives:
(1) The USNHC notes that the 0.1% ratio to other risks is low enough to support the expectation that people living or working near nuclear power plants would have no special concerns due to the plant's proximity, and that an additional risk which exceeds 0.1% would not necessarily constitute a significant additional risk. 'Thg'. p romp t f atality objective refers to risks within 1 mile (1.6 km) *of the site boundary; the cancer objective applies to the average risk within 10 miles (16 km). In addition to these two criteria for individual risk, the USHRC is also considering an objective that the probabilty of an accident leading to a large release should be less than 10' per reactor year.
(2) In quantitative terms, these objectives are that prompt fatality f
1 '
risks should not exceed 5 x 10 per year, and that cancer risks 6 ~~
should not exceed 2 x 10 per year. Analysis by the USWRC indicates that the cancer risk objective is unlikely to be a controlling f actor.
$ -5 3 ese values are compatible with th,e 10 _per_ year risk limitj ICRP )
when source apportionment is considered.
5.28 Another example is given in the following with reference to the UK Nil c.iteria quoted earlier. The three basic criteria ot' the UK N11 principles can be represented by a log-log plot of f requency versus dose limits, as shown in Fig. 7, and the f requency/ dose relationship for wholc body doses, as shown in Fig. 8. Ustng either a constant risk criterion (a constant product of l frequency and maximum dose; Case A) or a risk aversion criterion (a reduction by 10 in the product of frequency and maximum doses for doses above the ERt; Case B) a relationship corresponding to Fig. 9 can be obtained [23).
This provides vaiues of p(D) referred to in Section 2.
5.29 The limit lines on dose are transformed to individual risk diagrams usies the factor relating health hazard to dose, equivalent to P(efflD). This f actor can be obtained from a full scope Level 3 PRA, based ideally upon all event sequences and the entire range of meteorological conditions. However, limitations on the information available about all event sequences generally l restrict the application of this approach, and in this case only an average value of p(ef flD) over all values of D was obtainable. This enabled the individual risk curve for f atal cancers to be evaluated for that plant which just meets the UK W11 limit line, i
i 1
- 44 .
5.30 in particular cases where information on discrete fault sequences is available for given health hazards the compliance check is made at an earlier stage than that for individual risk. Figures 7 and 8 illustrate typical situations in which consequences of individual event sequences siV,ing doses in
- 4 a particular dose band show compliance with the limit line. The WK N11 principles require not only that each sequence frequency (the X in the figures) must be below the corresponding Limit line but also that the sureed frequency (the o in the figures) in each band must also lie below the limit line.
5,31 Expressed in this way, and bearing in mind the assumptions and judgements needed to express it in this way, a direct comparison can be made between these criterion curves and the criterion curve derived in' Fig. 6. The coherence in concept and in overall structure between the criterion curves is
, enc ouraging .
4 e
6 i
I
i 1
. 45 -
l l
l 2
l 1
is l L i .
j ,.
4*
p ,
1 eo j ,,1 .
sGd,
. ,% *,4 . ~
r / l=
3 v
A*.v i*
r
=,,' .
y =* v , < .
. =1 *
- y s
' v g , 8
=
' s V .s ,
- e , '.
- esa e W6e MAIger~f li,+.it reference level Figure 7 NH assess ent reference hvels for
$ discrete fault sequerces I
7" A
S I -
W '
E *
l P
s l l 1
ji. - je,
. 1 6 '
Y X'
= is -
=
C A
i e 1
)- '
{,,.sw' wJ wl w' l
6 cose egaivalen: Sv l Fa y e 8 !GI assessrent reference levels for discrete fault seg.iences, expressed in tems of Wie body and thyrotd dose egaivalents.
1 1
l 9
4 i
i
-. '.j' i t -
i
_ Case A
... Case B I, ,.1 ,.
No. rders to dose band 'g d E 1 .
4 s 5 l
- j %p - 1 l
3 t p . i ';
- i. . . ,
- s. I E th*-I l '
C b...$
. L......I-p: c' f
.c>
, 1 2.2 r i or 1rdividaal risk of fatal cancer for one year of reactor cperation Tipre 9 : Fregaency of accidents versu individal
' risk of fatal cancer, i
49 Section 6 IKpROVING SAFETY FURTHER (OR SAFETY OpTIMILATION) f 6.01 husuring that no single individual will incur an unduly high probability of narm because of potential radiation exposure is a necessary but not sufficient condition for ensuring the appropriateness of the level of
. safety of a radiation source. The question rem 4 ins whether that level should v be improved further by taking into account, for instance, that a high number of individuals incurritig an acceptably low probability of hat-n may stR1 represent an unacceptably high expectation of harm. _
For exposures assumed to occur with certainty, the ICRP has recommended - and the IAEA has adopted the requirement that the radiation protection app)'ed to the source must be optimized in order to ensure that all doses be kept as low as reasonably
, , achievable. (This requirement generally leads to individual dones being well
, below the individual dose limits.) Thus, gtimization of protection is
_ spy 11ed to the source of radiation, in contrast to 2do,se limita_ tion which is
~
applied to an individual. -
6.02 It is clear that the full assesrment of the consequences of accidents includes consideration of the number of people af fected and the levei_of, hat.n to them, and of all efforts including costs, required to improve safety. This aspect is sufficiently Close LO the ideas involved in optimization of protection that it is sens_ible to examine how the concept might be extended to a risk based system by analogy with the treatment in Section 3.
6.C3 The concept of optimization involves the choice of the most appropriate level of protection, taking into account a number of factors, the major one being the total t. arm to the exposed population, represented by the collective dose, and the costs of protection. It is recognized, however, thet other factors may be taken into consideration, such as whether the doses are received by workers or members of the public, the average level of individual dose or even the distribution of doses. Thus the process of uptimization can
be thought of es a decision aiding technique. Thoro uilt bn other f actot s
- that enter the final decision, some of which have nothing to do with raitiation protection, so the result of the optimization study can be seen as a partial input to the final decision (see Fig. 10) 6.04 In extending this system it ismostusefultofocuson.t1I[v 'deeision aiding' concept of optimization and to extend the number of factors incorporated to include the probability and consequences of potential e xp o su re s . In this extension it may not be hvipful to make the assumption about the equivalence of various. types of health ef f ects that was adopted for simplicity in Section 3. In particular it is probably helpful to deal separately with t_he consequences in tetus of..n..on.. .stochasti.c ef f ects ,.
perticularly acute death, and with stochast.ic ef fects. It may also be us,:ful to consider other human related aspects, such as _ the number of people who need to be evacuated or the area of farmland contaminated beyond a particular level.
a 6.05 A basic principle of optimization is that the inputs should be a
, , realistic rather than deliberately pessimistic. This would apply to cptimization for probabilistic events as such as to normal operations. If this rule is followed and if the best, estimates in the risk assessment are used, then it might be considered proper to take account of post accident countetweasures such as evacuation and food restrictions in assessing the most realistic results. These countermeasures could be ignored, as a deliberately '
pessimistic choice, in assessing the maximum individual risk in Section 3.
6.06 The relevant quantity sugt,ested by the ICRP and adopted by the IAEA for optimization of protection against exposures assumed to occur with certainty is the ' detriment', defined as the expectation value of harm for the group of people af fected by the source of radiation. For exposures which have a probability of occurrence of less than unity, the use of the concept of detriment may not be straightfotva _, as the following example shows: consider en accident sequence which has a low probability P of occurrence and which has a high consequence C if it occurs (and of course no consequences if it does ;
not occur). The expectation value of harm is given by the product pC. If p l
il very smat1 and_C_is_very_large, the. detriment _will,be_of an intettediate value which does not adequately numerically represent the _ _
i9
, i Selection of practical options
$r
\/
Identification of' Otb r radiation safety / i j , g g ,,,,
factor T U
l
. o, l
Decision-aiding technique l O 1 i
Results of g optimitation study i
V V 1 l
Decision on option Figure 22: Outline of radiation safety inputs to apeir 1: stim.
. procedure.
situation, which is that there are wither no consequences or major ,
/' , ' , '_"_'**"*""*
l i the consequences is not evidenced,to_t_he_shcjjion maker and theref ore is not able to be included in the decision making process. Thus, for these or any ,
other conditions under which the expectation value of the consequen.ces -~7- .
differs significantly from any one of the possible magnitudes of the_conse 3ences, detriment may not be a useful quantity for evaluating engineering options.
6.01 In addition to the pt oblem of what quantities to compare, there is the problem of how to include in. the cosparison process quantities or pref.erences which are not expressed in comensurate units. Such preferences, which should explicitly be accounted for, include: the degree of risk aversion _for higher consequence accidents; social costs for restrictions or inconveniences, morbidity and mortality; and the reist.ive weighting (i.e. thedegrkeof relative importance) of the various manifestations of health effects.
)
j 6.08 The problem of comparing quantities which are not directly lin_esrly _
, comparable can be addressed using utility functions and decision theory.
Preferences for quantities of differing types are expressed using a utility function which prescribes how the dif fering types of quantities are to be combined for the purposes of comparison. The resulting utility functions are then processed by a decision mechanism to arrive at a 'best under the circumstances' (i.e. optimized) option. ,possible implementations of this approach are being discussed in the literature ( 2_9 ,3 0J . The use of this class of decision aiding technique is being discussed by a task group of the ICRP but no publication is yet available from the group.
6.09 Wuclear_saf ety criteria have ,in _some _ cases been expressed as _ obj ectives in the fots of relationships between y ob_ ability,and size of consequene,e. If these are correctly framed as objectives rather than limits, then the implicatio ist be that a design corresponding to_the objective is an optimum d e s is,n.
The corrolary is therefore that such_ societal risk objectives should result from e_ither specifinorgeneric optimization studies, rather than being set up 'a priori'.
, ~
- 51
, i Section 7 -
J Nhl LIMITS CW SOCIETAL RISX
(
y JU V '
1.01 Societal risk limi8.s have been proposed or utilized f or any aspects of radiat. ion saf ety. Perhaps the earliest and best known recornmendatioris are those which apply to the probability and consvguences of reactor accidents, as illustrated in Fig. 1. While it is not the intention of this report to make a recormenJstion regarding the application of societal risk criteria, their potential use in radiation safety should be considered.
7.02 Societal risk limits do not follow directly from the principles of protection developed by the ICRP and extended to accident risks,in Sections 3 and a. Conceptually, however, societal risk limits may have a l'egical connection with the ICRP principle of justificattor,.
~
J
- - With respect to planned exposures, the ICRP concept defines limits only ;
'. for individuals. The detriment for society, which consists of costs and of health detriment, is then considered in the optimization' process required to be performed under the constraint of the individual limits.
However, a limit for the_ total detriment caused by a given practice,
_results from the justification requirement which states that this detriment must not exceed the benefit from the practice.
The detriment to be considered in this context is not per se restricted to planned exposures only but may - at least in principle - include probabilistic exposures.
The existence of intangible costs and ben 2 fits may make such a i l
justification analysis complicated. It would therefore be helpful to i 1
--define _ app _or,tioned limits for specifi_c conte _ibutions.to the.t_otal
_ detriment caused by the practice. Such contributions are the health j
,d_etriment caused by the planned radiation exposure of members of the public, planned oc,cupational exposures, and expected exposures among _the public as well as of employees caused by accidents.
- 52
. For one of these contributors,_namely_.the .hedth detriment incurred by members of the public, limits have beers fot nulated in a few specific Cases.
- In one case the total number of fatalities which must not be 9 ,xceeded I,g during the whole duration of this practice was specified (31).
- With respect to this ilmit, it would make no dif ference whether the number of f atalities would be expected to be caused by few accidents wash
. with major consequences or by more accidents each with lesser consequenceJ.
- If one wents to_give different we,ights to each totality depe'iyding upon, the total number of fatalities in one accident, the methodology as outlined earlier is therefore insufficient. One possible way of ,
s
, introducing such a risk eversion f'ictor is the definition of a limiting-
'.* CCDF curve, as shewn in Fig. 5. -
. - ~.__. _ ..~..-.
1 W
l
- 53 RETERENCES' 1
[1] It'TKkNAT10NAL COMMISSION ON RAD 10 LOC 1 CAL PROTLCTION, Reconcendat ions of the International Concission on Radiological Protection, ICRP Publication Wo. 26. Pergamon Press, Oxford and Now York (1977).
[2] INTtkNATt0NAL ATOMIC ENERGY AGENCY, Basic Sefety Standardg.for Radiation Protection: 1982 Edition, Safety Series No. 9, {AEA, Vienna (1982).
[3] APOSTOLAX1S , C. Probability and Risk Asscstment: The Subjectivistic viewpoint and some suggestions. Nuclust Safety, Vol. 19, No. 3, May-June 1978. l
[4] INTERNATIONAL ATOMIC ENERCY AGENCY. Radiation Protection Clossary.
1AEA, Safety Series No. 76, IALA, Vienna (1986).
(5) INTERNATIONAL COMMISSION ON 3AD10 LOGICAL PROTECTION, Radiation Protection Principles for the Disposal of solid Radioactive Wastes, Publication No. 46 ICRP, Oxford and New'rork (1985). -
l (6) IFTERNATIONAL ATOMIC ENERGY AGENCY, Conduct of Regulatory Re9teh and Assesraent During the Licensing Process for Wuclear Porze Plants, IAEA 4
Saf ety Series No. 50 SC.C3, p.14. , /LtL*., Vienna (1980).
.., (1) INTERNAT10WAL ATOMIC EVERCY ACENCY, Criteria for Underground Disposal of Solid Radioactive Wastes, IAEA Safety Series Wo. 60,.1AEA. Vienna ;
(1983).
[8] WUCLEAR EWERGY .tCENCY OF THE ORGANISATICW FOR ECOWOMIC Co-0PERATION AND DEVELOPMLWT, Long Tern Radiation Protection Objectives for Radioactive .
Waste Disposal, Report of a Group of Erports, OECD'WEA, Perly (1984). '
I
'\(9] HICS0W, D.J., Wuclear reactor safety goals and assestment principles.
} Wucl. Saf. 26 it1985)1.
l(10) IKTERNAT10WAL ATOMIC ENERCY ACEWCY, Status , Experience and Future s Prorpects for the Development of Probabilistic Saf ety Criteria, Report of the Technical Committee Meeting,1986. I AEA TECDOC Wo. , IAEA, Vienna (1986, 1987).
[11] FUCLEAR ENERGY AGENCY OF THE ORCAWISATION FOR ECONOMIC Co-OPERATION AND D EV ELOPMENT , ' Seminar on Interface Questions on MUelear Health and Safety.' Paris 16-18 April 1985, OECD/NEA.
[12) INTERNATIONAL ATOMIC ENERCY ACENCY, INTERTRANs A System for Assessing the lupact f rom Tranrporting Radioactive Kat arials , IAEA-TECDOC-28 7, IAEA, Vienna (1983). (i (13! IWTERNATIONAL ATOMIC ENERCY AGENCf Technical Committee on the Assesraent of the Radiological 1eloset from the Tranport of Radioactive Materials. Proceedings TC-556 (Addendum), IAEA, Vier.na (1985). i I
l t
/
lle) INTERN ATION AL ATOMIC ENERCY ACENCY , Chairman's Report of the Technical ,
Committee on the Assearment of Radio 19 steal Irpact from the Transport '
of Radioactive Materials. TC 556, IAEh, Vienna (1985). l j' l (15) INIERNATIONAL ATOMIC EWERCY ACENCY, Asnessment of tpa Radiological Impact, of the Transport cf Radioattive Materials,1<WA! TECDOC- 398, IAEA, Vienna (1986).
fL (16) INTERNATIONAL C0KM1SS10K OW RADIOLOGICAL PROTECT 101), Cost.kenefit Analysis in the optimization of Radietion Prc.tection, ICRP Publication No. 31 Pergamon Press, Oxford and New York (1983).
I I
[17) INTERNS.TIONAll COMMMISSION ON RADIOLOGICAL PROTECTION, Stat:went from
\ the 1985 Faris Meeting of the ICRI", Ann. ICRP 15 3. Pergamon Press.
Oxford and Nev York (1985). l l (
F18) INTERNATIONAL C0KM1SS10N ON RADIOLOGICAL PROTECT 10N,i?.-inciples of Monitoring for the Radiation Protection of tie Population, ICRP e Publication No. 41[ Pergamon Press, Oxford and New York (1984).
a.
(19) INTERNATIONAL ATOMIC EMERGY AGENCY, Principles for Limitinh yeleases of Radioactive Ef fluents into the Environment. 1996 edition. Satety Series Wo. 17, IAEA,+ Vienna (1986). ,
/ [20) COWIALEZ, Abel J., "The regulatory use of probabilistic safety 1141ysis in Argentina", Internatior.a1 Meeting on Thermal Wucivne Reacter Tafety.
- Chicsso, 1982. /' ,
y r f 121) S N 1,2 E K , J.H., An Integrated SafetyjCool Concept, Proceedings af the 1.f.EA Tchnical Comma. tee on the CLitus Experienen and Futur,= Pt-ospects for the Development of ProbsLilistsc Safety Criteria, IAEA.,Viennt (1986). WUREC. 0800, on saf ety Coal Policy Statement (48FR10)12) .
122) HER MAJESfY'8 WUCLEAR INGTALLATIONS INSPECTORATE, Safety As'sessment Principles for Nuclear Power Reactors Hesith and Safety Executive, (July, L982j.
(23) UNITED KIWCDOM, NAT10WAL RADIOLDCICAL PROTECT 10W BO ARD, Procedtces to Relate the Wil Safety AsJestment Principles for Wuclear Reactor; to Riek, WRPS.R180 (1985). ,
[24) UNITEDSTATESWUCLEARR3bOLATORYCOMMISSION,SafetyCoalsforthe operation of Duclear Pever Ple.nts, Policy Statement, 4 August 1986, USWRC, Wdahington, DC (1986).
[25) SEWASSA1, S. et al, in Optimization of Radiation Frotection Proc.
IAEA CECD/WEA Symp., IAEA, Vienna (1986).
[26) 'COMiSION WAC'CNAL DE ENERCIA ATOMICA, BUENOS AIRES, AltCENTINA, Consejo Owsor para el Licenciamiento de Instalaciones uvcisates. 'Criterios T*dio16sicos relativos a accidentas'. Norma Calin no. L.1.3 CWJA, Suenos Aires, 1979. 2y l
( -
. . i (2?) COMISION NACLONAL DE ENERGIA ATOMICA, BUENOS AIRES. (Argentina).
Consejo Asesor para el Licenciamiento de Instalaciottes Nucleares.
'Andlisis de fallas para la evaluacidn de riesgos'. Not-a Ca t itt tio .
3.2.2. CNEA, Buenos Aires, (1980)
(28) INTERNATIONAL COMMISSION ON RA0lOLOGICAL PROTECTION, A Compilation of the Major Concepts and Quantit ies in Use by ICRP, ICRP Publication ho.
42 Pergamon Press, Oxford and New York (1984). -9
- r (29) BENINSON, D, and Gonzalez A., "Optimization in relocation decisions",
Optimization of rat'lation protection Pro . lAEA OECD/NEA Symp., IAEA Vienna, (1986).
[30) WEBB, C. et al., "Development of a general fr.mework for the practical implementation of ALANA Optimization of Radiation Protection Proc, Symp. lAEA OECU/NEA, Vienna, (1986).
[31) UNITED STATES ENVIRO >NENTAL FROTECTION AGENCY, Environmental Standards for the Mana$ement 4 .d Disposal of Nuclear Fuel, High. Level and Transuramic Westes , 40 Code of Federal Regulations, Part 191 UShPA, Washington, DC.
n N
' o.
Ik
!{
't i I
i l
rll
?
List of Participants 1st Advisory Croup Meeting. 1-10 October 1985 Ahmed, J.U. International Atomic Energy Agency (Scientific Secretary) Division of Nuclear Safety Wagramer Strasse 5 g.
1400 Vienna -
4*
Austria
\ Cool, D, Nuclear Regulatory Crymmission Divison of Fuel Cycle and i daterial safety Washington DC 20555 USA Coulon, R. CEA/IPSN/DPS/SECP BP6 92260 Fontenay aux-Roses France ',;
Gonzalez, A. RNACE S.A.
(Chairman) Casilla de Correo 1589 Duenos Aires j Argentina l
.~ (Presently at the Internstional Atomic Energy Agency l Division of Wuclear Safety l Wagramer Strasse 5 l 1400 Vienna Austria)
Martin, D. Atomic Energy Control Board P.O. Box 1046 Ottaws, Ontario K1P SS9 ;
l Canada Moberg, I. Wational Institute of Radiation Protection Box 60204 S-104 01 Stockholm Sweaen Shinichi Bugs Deputy General Manager l Bionssasy Division Department of Health Physics JAERI To Kai-mura Waka gun Ibaraki-ken 310- 11 Japan I
51 -
Sinha P4y, H.K. Scientific Officer Bhabha Atomic Research Centre Bombay 400085 India
-__- Un ru h , C . H.
Battelle Pacific Northwest Laboratory (*
T' P.O. Box 999 Richland, WA-99352 USA Waight, P.J. Radiation Scientist Prevention of Envirotunental Pollution Division of Environmental Health World Health Organization Avenu Appis CH-1211 Geneva 21 Switzerland .
-- Webb, C.
National Radiological Protection Board Didcot 01 11 ORQ i England U.K.
2nd Advisory Croup Meetinx. 19-23 January 1987 Ahmed, U.
International Atomic Energy Agency (f isnttfic Secretary) Division of Wuclear Safety Wagramer Strasse 5 1400 Vienna Austria Benassai, S.
ENEA DISP Italian Direcorate for Nuclear Safety and Health Protection Via Vitaliano Brancati 48 Roma - 00144 Italy Gonzalez, A.
EMACE S.A.
(Chairman) Cadilla de Correo 1589 Duenos Aires Argentina (Presently and the International Atomic Energy Agency Division of Duclear Safety Wagramer Strasse 3 1400 Vienna Austria)
Hock, R. Bselinerstresse 295 , ,,
D.605 offenbach Federal Republic of Get9'any 1
l Koeberlein, K. Cesellschaft fur Reaktorsicherheit mbH D-8046 Carching Federal Republic of Getyany 1
l Madhvanath, U. DivisionofRadiologicaI.
Protection Bhabha Atomic Research Centre Bombay India Martin, D. Atomic Energy Control Board P.O. Box 1046 Ottawa, Ontario KlP SS9 Canada Wiederer, U. twiss Nuclear Safety ',*
Inspectorate CH-5303 WUrenLingen Switzerland b
. Peirmattei, S. EN EA. DIS P 1
. Italian Direcorate for f Nuclear Safety and Health l Protection {
Via Vitaliano Brancati 48 i Roma - 00144 Italy t-Riron, J. Nuclear Installation Inspectorate Barnards House 1 Chepstow Place Westbourne Crove London W2 4 TF United Kingdom Valentin, J. National Institute of Radiation Protection Box 60204 )
S-104 01 Stockholm l Rweden V6rBss, L. Head of the Division Institute for Elactrial Power Rosmarch Budapest V., Zringy u. 1 H-1368 Budapest Hungary l
l l
l
_. _ _ _ _ _ _ - - _ - _ - . _ 1
. . . - . . _ _ . - . _ . - . . . - . . . _ . . . . . - - . _ . = . . . . .. . ~ . . ~ - . , . - . . -
. . . . i
. - - - Webb, C. National Radiological Protection Board Didcot OX 11 ORQ England U.K.
4.,
Whipplc. C. Electric Power Resear,ch Institute '
3412 Hillview Avenue P.O. Box 10412 Palo Alto, CA 94303 USA
'Consu lt a . for the Task Webb. G. National Radiological Protection Board Didcot OX 11 ORQ
~
England 'N U.K.
,~_ 'a K
I J
l i
l l
.- ._.-.;.-... . . . . . - . ............__,._-._..-.__...._..__...~...,....,__._.-..,_..-.;._.-._....~_.~.___
, . . i 4
. 'j '
\
ENCLOSURE 4
/,
4 i
b
(
{
l l
l l
l
,,, , ,, - , . - , , . , , .