ML20033E046
| ML20033E046 | |
| Person / Time | |
|---|---|
| Issue date: | 01/29/1990 |
| From: | Taylor J NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO) |
| To: | |
| References | |
| TASK-PII, TASK-SE SECY-90-031, SECY-90-31, NUDOCS 9002080106 | |
| Download: ML20033E046 (29) | |
Text
i.
1I k
- I RELEASED TO THE PDR
/
-\\
e I
J/rMo dL-
'g,,,,,/
date '
infats l
POLICY ISSUE January 29, 1990
{lgf0fm8100)
SECY-90-031 g:
The Comissioners From:
James M. Taylor Executive Director for Operations Ssubject:
EVENT SEVERITY SCALES FOR COMMERCIAL POWER
. REACTOR FACILITIES Eurpose:
To inform the Comission of the status of the International development of an event severity scale for commercial power reactor facilities.
Discussion:
The proposed development of event severity scales for comercial power. reactor facilities and the staff position.regarding such scales were presented to the Comission via SECY-89-266. The-impetus for numerical severity scales in those countries which have,either developed or expressed interest in such systems is to improve communications with the public and media regarding events occurring at nuclear facilities.. The activities of individual countries such as France and Japan and the plans of organizations such as the International Atomic Energy Agency (IAEA) and the Organization for Economic Cooperation and Development (OECD) Nuclear Energy Agency (NEA) were discussed in the previous comission paper. The staff position regarding severity scales is that while development of such systems may be warranted for those countries needing to improve public and media comunications, the adoption of a severity scale in the U.S. should not'be pursued since the existing U.S. systems for event reporting are well established and include extensive and open comunications to the public and media.
A joint IAEA/NEA meeting was held in October 1989 to discuss the experience and use of event severity scales. The meeting was attended by 45 representatives from 17 countries and 4 international organizations (including Edward L. Jordan as the U.S. representative). A general consensus was reached in favor of the development of an international severity scale and the ideal characteristics that should be incorporated into the scale.
Such characteristics as format, parameters to be utilized in event classification and the approximate number of N evels within the scale were agreed upon. Details of the
~
QO 2l06 $ )
p
Contact:
iu Gary Zech, AE00 Ay c
49-24154 m
he' 1
4, -
.;-9,7 y, m 2
<1 meeting are provided by the U.S. representative's trip report (Enclosure 1) and the. IAEA/NEA progress-report (Enclosure 2).
The U.S. representative presented the U.S. position that c
, continued involvement in the development of.an international severity. scale was desired but that the U.S. does not plan to adopt such a scale for routine use.
3 A. joint IAEA/NEA Advisory Group on the International Severity Scale has been formed and will meet in late January 1990.
Gary Zech, from the Office for Analysis and Evaluation of-Operational Data, will represent the U.S. during this session.
This advisory group will develop a draft' scale which will be provided to member countries for use during a trial period.
3 After the trial period of approximately one year, IAEA and NEA 1
will review the experiences, revise the scale if necessary and endorse the final scale to member countries for general use.
Note that the staff will continue to hold the Commission i.
approved position that the current NRC event classification system best fits U.S. needs and there will be no commitment to test the draft scale during the trial period. The staff continues to believe, however, that it is to our best interest to participate and to provide representatives to the international organizations curing the development and implementation periods for the international event severity scale.
mes M. T or xecutive irector for Operations
Enclosures:
1.
U.S. Representative's Trip Report 2.
IAEA/NEA Progress Report l
gg
- t, ENCLOSURE 1-November 17, 1989 MEMORANDUM FOR:
James M. Taylor Acting Executive Director for Operations ROM:'
Edward L. Jordan. Director Office for Analysis and Evaluation of Operational Data
SUBJECT:
FOREIGN TRIP REPORT - EXPERIENCE IN DEVELOPMENT AND USE OF NUCLEAR INCIDENT AND ACCIDENT SEVERITY SCALES Enclosed is my trip report for the subject meeting held in Paris, France on October 23 thru Octccer. 26, 1989.
Original Ei;r.ed by:
E.L Jord:n Ecwarc L. Joraan. Directer Office for Analysis ano Evaitation of Operational Data i
Enclosure:
As stateo cc/w enclosure:
.H.
'E. Beckjord, RES R. Bernero', NMSS H. Denton,.GPA R. Eraay, SEC Distribution:
j AE00 R/F Suoject File ELJoraan CJHeltemes TNovak LSpessard 0FC :0:
0 NAME d boy an
~
~
DATE :11/ / 7/89:jr.
i l
l OFFICIAL RECORD COPY
Wh g:1);
]
N g.
r f-I L
hg
- )
. TRIP REPORT ABSTRACT i
November ~ 17, 1989 1
Ir
-3fficial Traveler:
E.
L.' Jordan
' Travel To:
' Paris. France Office:
AEOD Beginning On:
October 23, 1989 Until:
October 26,-1989 i
4 Meeting Title and/or Affiliation:
Experience in Development and Use of Nuclear.
j Incident ano Accident Severity 5cales
~.r;sni:ec By:
'AEA/NEA
- ESTRACT
'The October meeting was neld at the OECD offices in' Paris between 50 represent--
atives from 19 countries IAEA, NEA, WANO and CEC.
Since this was a Joint IAEA anc OECD/NEA meeting it was remarked this was the first time Soviet bloc s
nations'naa participated in a meeting at OECD offices.
The U.S..was' the only nation that expressed the-position that we do not plan to adopt a new severity scale if-an-international scale is developed thru IAEA/NEA efforts.
Other' representatives are dedicated, compelled or resigned to-cevelopment of an international scale and implementation in their countries.
From the comments of. the other representatives an international scale socears inevitable.
An international scale would most likely be a modification of the existing French scale.
In addition to the basic position of not planning to adopt a new scale, the U.S. has raiseo the following points regarding development and-implementation of an international scale.
l.
The purpose and use must be restricted to communication to media and public.
There remains a strong tendency among participants to consider-it for wider purposes including technical uses between countries.
The criteria are designed for relatively promot determination (hours to days) and thereby do not hold up for more scientific safety significance determinations.
w y j: :.
lc.
b*
i f ;
'2.
The upper portion of the scale must consider Emergency Preparedness Provisions in criteria and or examples to prevent conflicts.between
- media notifications and emergency response notifications.
3.
Detailed numerical criteria for radiation dose or radioactivity j
releases may be impractical because of National-and site specific regulatory values.
Use of national values with reference to applicable well known standards may be a practical approach.
(This area is the one with the greatest remaining uncertainty and-
)
aisagreement.)
i 4
There is a need to minimize minor adjustrents of a scale value for a j
given event.
Only significant changes so.ould be made.
(Participants 1
interested in wider applications recommend maintaining a continuing
- listing of severity thru various levels of analysis.)
5.
A severity scale should not be used as an " indicator" of relative.
national safety performance since the_ criteria are not closely related to safety or risk especially at the lower end of the scale.
Arguments:were presented by the U.S. representative that a numerical scale for the high or accident end is unnecessary since-the media and public will relate to TMI, Windscale and Chernobyl accidents as the points -
offcomparison.
Application of a detailed low end of the scale may be misleading since a determination of risk or safety significance requires longer term analysis.
Such analysis must first consider risk at the facility and then risk with respect to other facilities.
The U.S. is caught up in a laudable international' movement to communicate reactor operating events more openly to the public.
Since U.S power reactor events are already openly communicated we find ourselves in the g
dilemma of supporting more open public communication in the international community while disliking the mechanism of an international severity scale-i because of potential adverse impacts of adoption in the U.S.
4 The participants recommended another expert meeting in February to propose a scale for testing purposes.
A subsequent' larger meeting is planned to review that group's product for voluntary trial use.
Strong NRC representation should continue in these meetings to assure that if an international scale is promulgated that it is logically developed and that flexibility remains regarding its use by member nations.
" ~
w
-ENCLOSURE 2-LOCDE OECD ORGANISATION DE COOPERATION ET
.-b OROANISATION -FOR ECONOMIC.
DE DEVELOPPEMENT. ECONOMl00ES 4
COOPERATION. AND_ DEVELOPMENT AGENCE POUR L'ENERGIE NUCLEAIRE/ NUCLEAR ENERGY AGENCY atrtatscs
- =,7 Tel. 45 34 82 00 EN/S/2764 Paris, 6th December 1989.
To:
Participants in the Joint IAE&/ IGA Technical Coenittee Meeting on Experience and Use of Nuclear. Incident and Accident Severity Sealms Please find enclosed the report from the Joint Technical Committee meeting which was held in' Paris from 23rd to 26th October 1989.
This report has been prepared on the basis of the documents drafted by the three_vorking groups.during the meeting and includes-all the material which was contained in those documents, although a significant reorganisation and.some expansion of the text were necessary.
This report vill constitute the principal working document for.the forthcoming Joint IAR&/NEA Advisory Group on the' International Severity Scale, which will be held in Vienna on 29th January'to 2nd February 1990.
Invitations to attend the Advisory Group are being issued by Mr. Thomas to a restricted group of experts.
in its present form, if you have any major comments to it please co them to se within the next few days. This might-give rise to some amendments-in the report in' time for the Advisory Group meeting.
Yours sincerely, u
'p7 O.Jiari Attached:
As stated c.c. Mr. B. Thomas Mr. B. Tomic i
National Delegations to the OECD f
- ,.?
- r;
..q
,ll 4
i
i; i
OECD NUCLEJL EIGEGY INTERNATIONAL ATOMIC ENIEGT AGENCY AGENCY t
iil v
i:
1
.i PROCERSS REPORT ON THE DEVELOPMENT OF AN j
INTERNATIONAL SEVERITY SCALE FOR EVENTS OF SAFETY' SIGNIFICANCE r
't JOINT IAEA/NEA TECBNICAL C000GTTER MEETING j
PARIS, '23RD - 26TH OCTOBER 1989 o
b i
e
2 l
q reassono Although nuclear operators around the world have generally achieved-very high standards of safety, from time to time, safety-related events do occur at nuclear installations. The vast majority of such events have no radiological significance and their implications for continued safe operation are very slight.
However,;it is very difficult for the media and public to make judgements about the seriousness of such events and very small incidents-have sometimes caused unnecessary concern.
~
Because of the difficulties. involved in explaining complex technical issues, the nuclear community has been concerned to explore possible new mechanisms for explaining to the media and public the significance of incidents and accidents at nuclear plants, in order to put_them into better perspective..It has been suggested, for example, that if a suitable severity i
scale could be developed to classify events according to their health and safety significance, this vould be an important aid to better understanding.
Several countries have already given consideration to the development of such a scale and a number of versions are now in use, mainly on a trial basis. It is videly appreciated, bovever, that the proliferation of different scales is to be avoided if at all possible, since this may lead to confusion.
The IAEA and NEA have therefore both begun the process of attempting to explore the possibility of making available a single severity scale to describe the safety significance of incidents and accidents within countries and between countries.
It is intended that this should have the videst possible international agreement and could be used in all countries, where appropriate.
This paper represents the views of a joint IAEA-NEA Technical Committee meeting to explore the issue. The meeting examined the purposes and ideal characteristics of a scale, begun the process of examining the technical basis which might underlie it and has recommended a programme of further discussions and consultation to an agreed timescale which, it is believed, should lead to the development of a suitable scale which would be videly acceptable.
p.7 f.
3-b 1.
Ilmt0 DUCTION-
?
^!
e The activities towards development of an international severity scale started in 1987, when a preliminary study to examine various technical f
questions in connection with severity scales was carried out within the NEA.
This study was endorsed by the NEA Committee on Radiation-Protection and i
Public Health (CRPPH) in November 1987, and, on this basis, the NEA called in 1988 an international expert group meeting to make further progress on this subject.
The discussions of the Expert Group were concentrated on the principles and criteria needed for the establishment of a severity scale and possible forms of.the scales with reference to their potential for successful h
application internationally. The. findings of this group have been_ discussed more broadly within the NEA CRPPH and the Committee on the Safety of Nuclear L
Installations (CSNI).
Also, in its 1988 report, the IAEA's " vise men's group" h
(an international body monitoring the IAEA's scientific programme) suggested strong and immediate IAEA involvement in development of a severity scale.
The step forward in international development was the IAEA-Consultant's Meeting (CM) which was convened in April 1989 to discuss the experience of development and use of severity scales. This group addressed in more detail the ideal characteristics of the scale, possible approaches to the development of the scale, and-serits and disadvantages of producing a severity-scale.
The report of the meeting, which also contained descriptions of the scales in use or under development in some countries at that time, was videly' distributed to interested individuals.
Findings and recommendations of the CM vere presented and discussed at various IAEA meetings, including the L
September 1989 Technical Committee Meeting of national IRS (Incident Reporting System) co-ordinators.
Issues on public communication of reactor events, including the national and international severity scales, were discussed during the meeting on the' Scientific Programme for Nuclear Safety organised in conjunction with
~
the IAEA's General Conference in September 1989.
Top regulators and experts from many IAEA Member States attended the meeting and the majority of them expressed strong support towards development-of an international severity scale.
Finally, IAEA and NEA jointly organised the Technical Committee Meeting on Experience in Development and Use of Nuclear Incident and Accident Severity Scales, which was held in Paris in October 1989, with the objective to exchange experience on development and use of existing severity scales and to establish the fraaevork for the development and introduction of a severity scale acceptable internationally.
Due to high international interest in the topic, the Technical Committee Meeting was attended by 45 experts from 17 member states and 4 international organisations.
A list of participants is attached as Annex I.
The experience of development and use of existing scales in different countries, issues to be addessed in the development of an international scale, and the reports of NEA Expert Group and IAEA CM vere presented during the Technical Committee Meeting.
Based on the presentations and subsequent discussions, three major issues relevant to the establishment of an international scale vere identified.
These were:
l
~
- ,g
',o (i)
Ideal' characteristics and purposes of an international severity scale k
-(ii)
Technical criteria and general form of a multi-dimensional.
severity scales (iii)-
considerations and recommendations for establishment and validation of an international severity scale, including provisions to ensure consistency in its application..
Vorking Groups were formed to address each of these issues. The Vorking Groups were asked to discuss their respective topics in detail on the basis of presentations and plenary discussions, and also to review respective chapters of the IAEA's CM report. Based on detailed discussions,-each Vorking-Group was asked to draft conclusions and recommendations for consideration by the IAEA and NEA on its respective topic. Vorking Groups' discussions, i
conclusions and recommendations are reported in subsequent chapters of this document.
~
b d
e---
'~
,,,..x 5
i
- 2.
PtRPOSES AND IDEAL CR&RACTERISTICS OF AN INTERNATIONAL Sg7ERITY SCALE 1
- 2. l' Purpose,s_ofd he Scale, The prime reason for adopting a~ severity scale on a national basis is-to facilitate prompt communication with the media and the public on newsworthy
. events in a vay which vould aid understanding and place incidents and accidents in their proper context.
In this respect, it is likely that the vast majority of events which would be classified on any severity scale would be incidents of a minor nature. A scale vould have auch less importance in classifying relatively severe accidents.
In this context, a severity scale vould not be the most important means of providing public information, since better and more extensive information vill be provided by existing emergency arrangements.
On an international basis, a unified severity scale would, again, be primarily used for prompt communication between governments and the media and public on the significance of events which may have implications which cross national boundaries.
In this context, the benefit of an international scale vould'also be to supplement the information to be provided to governmental organisations in application of the Convention for Early Notification of nuclear incidents and accidents. This would allow the significance of any incident or accident to be understood in a consistent way internationally.
In fact, a proliferation of different national severity scales could introduce-confusion of public opinion and difficulties in international communication of-information.
In conclusion, it was agreed by the participants that the purpose of an international severity scale is twofold, namely (i)
To enable announcements on incidents and accidents to be made by the operators or authorities which give the media and the public a rapid and clear assessment of seterity on an internationally
-agreed basis; (ii) To ensure that the severity of reported incidents and accidents can be understood in a consistent way internationally.
L It is to be stressed, however, that, in order for an international scale to succeed in fulfilling these purposes, its objectives and scope should be kept clear and limited.
It was, therefore, agreed that such a scale should not be intended for use in detailed analysis of safety performance.
Such analysis is achieved by different systems and criteria applied within organisations and l
nationally.
Nor should the scale be seen as suitable in any way to make judgements about hov safety performance compares between different countries.
This should be clearly stated when introducing such a scale.
2.2 Ihe characteris_ tics of_an "id_eal", severity _ scale Although it is unlikely that any severity scale vould be able to meet all of the criteria set out belov, it is useful to list the requirements of an
" ideal" scale before going on to discuss a scale which might be considered in practice.
The participants recommended that the international severity scale i
should ideally have the following characteristics.
The scale should:
l-
y..
w
-6' a
(1)
-Be easy to use and understand.
i Although it is important that any scale.should have a sound
[
E and defensible scientific basis, it is of considerable importance that the search for technical soundness should not conflict with the requirement for simplicity and
. intelligibility. This requirement applies to nuclear' b
operators and regulators, to the media and to the public..
t (ii).
Be capable of being quickly applied to all incidents and.
~
accidents.
Whilst there vill. undoubtedly be problems in assessing the' severity of some incidents and accidents, the. application of the severity scale should not be a significant cause of delay in making an announcement, even though the event may have later to be reclassified. However, a-balance aust.be shown
[
~
between simplicity (ease of application) and detailed criteria, in order to reduce the likelihood of reclassification at a later date which would undermine the i
l confidence in the scale.
(iii)
Be based.on clear technical criteria to promote consistency.
It should be recognised that the scale vill need to be applied by technical people describing a technical event.
4 Satisfactory application can only be achieved by use of technically agreed criteria.for the positioning of an event on the scale, suc.h that any operator or regulator can classify an event in an overtly consistent manner. Although, from time to time, incidents or accidents may occur which are close.to the boundary between different scale points, and over which there may be subsequent debate, it is judged important that the '
scale should minimise the number of occasions when such conflicts could arise, and should be such that there is the maximum consistency in the classification of events.
(iv)
Promote credibility by having a sufficient number of levels to ensure that some incidents are placed at levels higher than the lowest point.
Any scale must provide good resolution of incidents and accidents of various sizes.
In this respect, it is essential that insignificant incidents should be well separated on the scale from those which may have had potentially serious consequences (e.g., THI), or those which were of considerable radiological severity (Chernobyl).
It should be noted, however, that this requirement to separate events of different soverity by an adequate margin may potentially conflict vith the need to minimise the number of scale points so as to achieve the above objective of allowing confidence that different operators vould be capable of classifying an incident or accident rapidly at the same point on the scale.
> 4 4
w, :
7-
/
P (v)=
_ Not have so many levels that.the endia and the public cannot q
- understand the difference between them.
The-vast majority of events to be cir.ssified vill'be incidents of minor severity.1-It is important.that-a reasonable number of events should be classified at the lover.end of the scale, in order for the use of the scale _to be appreciated and'its operation to become familiar to the media and general public..
llovever, if the scale is not able to distinguish between events of'a more minor nature, there_any ba a tendency to place every incident at the lowest level on the scale. The scale vill than quickly lose credibility with the media and the public.
If, on the other hand, it is so finely described-technically that it has a large number of levels, the media and the public vill not appreciate the significance of the different levels.
(vi)
Reflect the severity of different. types of incidents and accidento by being based on a set number of diversa criteria.
It vill.not be possible to construct a single technical criterion which vill be suitable for assessing the severity of-_
i a range of different types of event in different types of plant. Furthermore, the criteria vill need to indicate when the potential effects of the event must also be taken into account in assessing the severity of the event.
(vii)
Not conflict with provisions in energency plans established by national authorities.
r The scale should not cause confusion by, for example, describing an event as of relatively minor severity if the event is actually causing an alert under emergency procedures.
For emergency. planning purposes, notification to the public has_to be done at a very early stage and usually before a-level in-the severity scale could be defined.
In order to avoid confusion with those emergency provisions, the level could in the first instance be determined based on the emergency actions decided, even if the event has to be reclassified aftervards.
r t
(viii) Be capable of being applied to reactors, other nuclear installations and transport.
Although understandably-a detailed severity scale may initially concentrate on incidents and accidents in nuclear reactors, it should also be capable of being used for all the other events - in other types of nuclear installations and involving the transport of radioactive materials - on which it l.
may be necessary to make statements to the public.
l This list of proposed characteristics was made recognising that it is what a scale should contain in ideal circumstances.
It is recognised that not l.
every point can'necessarily be fully satisfied, but the list vill hopefully provide a useful check on the merits of a proposed detailed scale.
t I
- 9s.,
a...
8 s,,
9
- 3.
TE G NICAL CRITERIA FOR T B DEFINITION OF AN INFERNATIOBEL SEFERITY een n A fundamental question to be resolved in order to develop'a severity-
'r scale is the choice of attributes which are to be measured.-.The participants discussed inssone detail this question. Their considerations were largely based on the reviv of the severity scales currently being used, or developed, in some countries.. The principal conclusions and recommendations of the group are reported in the following.
1 E
. There are many attributes of an incident or accident which may be regarded as significant and which could have a role in the formulation of a severity. scale. The more important include:
3 (a)
The amount of radioactive material released; (b)
The off-site impact; (c).The on-site impact; (d)
The potential for aore serious consequences; (e)
The economic loss due to damage or contamination of the affected'-
installation.
It appeared rapidly clear that the severity of an incident or accident cannot be properly described by any of the above-mentioned attributes if each of.them is taken separately.
Any incident or accident presents, in fact, several different aspects which are relevant in terms of public concern. The participants, therefore, agreed that a multi-attribute scale - of a matrix type - based on a set'of defined criteria appears to provide the best basis
=
for a severity scale for purposes of public information.
l The participants also agreed that the best compromise between the
. requirements of simplicity of the scale and that of being sufficiently representative of the actual degree of severity of incidents and accidents could be achieved by a scale based on the following three attributes:
off-site impact; on-site impact; impact on the defence-in-depth principle.
In discussing the actual meaning and scope of the chosen attributes, the following considerations were made by the participants.
a) The off-site impact can be expressed as:
the amount of radioactive material released to the environment;.
1 the dose to individuals in the surrounding population.
7 i
4.
1 s,e 9
-The amount of radioactive material released is a concept which is readily understood, and'it is also amenable ~to estimation in a relatively short period of time following an incident or accident, although the quality of the estimate would vary with_the particular
{
circumstances of any release..However, expressing the seriousness of an event in-' terms of release size has the disadvantage'that it does not necessarily correlate well.vith the potential' for hara, since this vill be dependent upon the composition,-direction and duration of the release.
One possible way of expressing the release, at least in the lover range _of levels in the scale, vould be to classify releases in terms of-appropriate sub-sultiples and-multiples of the discharge limits established by the licensing.
authorities.
Empressing the classification in terms of dose from an incident or accident has, in principle, considerable potential for forming the basis of a scale, since it provides a good surrogate for hara and is amenable to.estination with acceptable precision and in a reasonable timescale following an accident.
Bovever, the' quality of any estimate vill depend on the nature and extent of environmental monitoring and requires the formulation of an agreed protocol for the estimation of doses as well as the definition of reference groups of exposed people and exposure pathways.-
The participants felt that either the release from the plant to the environment or the dose to individuals in the surrounding population or an appropriate combination of these two parameters should be the basis for classification in terms of off-site impact.
Bovever, no conclusion was reached on which of the above options abould be preferred.
u b) The parameters of importance for the definition of on-site t
l impact could bei exposure of vorkers,-internal radioactive leaks, failure of barriers, including degradation of the core.
The group thought that exposure of workers is an important subject of interest'to the public, although it was specified that in general'only exposures related to incidents should be taken into account for the purposes of classification in a severity scale.
These kinds of exposures, even if they are not accompanied by an off-site impact, can be the result or'the symptom of malfunctions or incidents caused by bad plant design or unsafe plant operation j
vhich may be of interest to include in a severity scale.
As far as radioactive leaks and failure of barriers are concerned, t
the participants felt that they could also be considered as degradations of the " defence-in-depth" and could, therefore, be classified under either "on-site impact" or under
" defence-in-depth".
In conclusion, the participants agreed that the classification according to the "on-site impact" attribute should be based on the three parameters considered above, namely, failure of barriers, internal leaks and exposure of workers.
-~
10 j
s.
, _4,.
L c) It ~is r; gard:d es important that some parameter should be included -
which may reflect the seriousness of an incident, even though the radiological consequences were, in the event, lov.. The accident at Three Mile Island exemplifies this issue. Although it is difficult to' develop. a scale ' capable of accommodating this feature,- and its inclusion vauld rely upon qualitative rather than any quantitative
' factor, there are many serits in attempting to include the potential for more serious consequences within a scale.
'i The potential for serious' consequences outside the' plant is i
expressed by the erosion of_the provisions of the
" defence-in-depth" principle.
The concept of " defence-in-depth" is centred on the existence of multiple levels of protection against an accidental release, I
including successive barriers preventing such a release.
It also
'C includes consequence aitigation sensures to protect the public and the environment from hara in case these barriers fail or are not fully effective.
The participants concluded that " defence-in-depth" vas the third necessary attribute to be considered for severity classification purposes.
d) The threshold for classifying events into the scale was an-important consideration by the group.
In reviewing the-contribution of the three chosen attributes;to the definition of
. )
this threshold,_or bottom level of the scale, it was agreed that this botton line is only dependent on parameters ~ relevant to
" defence-in-depth".
The "off-site impact" and the'"on-site impact" do not appear to influence the threshold for the classification of events in the scale.
The general description which was suggested for the threshold for classification of events into the scale was the exceedance of :
regulatory limits and conditions established for the normal operation of a plant. Examples of instances include i
(1)
Failure of a safety train without timely corrective action; (2)
Failure of a safety train upon actual demand (excluding deliberate deactivation);
i
=(3)
Operation outside the regulatory licence limits and conditions.
Most of the participants found the French scale approach appropriate to define the threshold for an international scale.
This threshold seems to be between level 1 and level 2 of the r
Japanese scale. Events below the threshold would be classified as "belov scale".
It is recognised that news media releases may occur related to events which are outside the scale due to instances such as:
(1)
Outage to repair only non-safety related equipment:
(2)
Industrial injury to plant personnel vithout radiological implications.
l
-1 3
11 1
- Bovever, the participants telt that there any be a number of events
'd which are 'balov scale" or outside the scale amongst which each national-authority any have its own criteria for saleeting those which should be reported to tho'public.
e) The group considered the proposal. to include the economic. impact as one of the attributes of the scale.. Bovever, it was agreed that F
the economic impact is not relevant as an_ attribute since the scope-of :the scale is limited to prompt communication to the public on
~
1 the safety relevance of events. The economic impact, although it i
any be an important element in judging the severity of an event,.
could introduce confusion in the classification, because there exist no clear relationship between costs and safety significance and it vould be perceived by the public differently from safety and health impacts. Also, the economic impact would in most cases be
,'l very hard to estimate in the short. ters.
Moreover, the group considered the opportunity to take into account, as criterion for classifying an event, the lessons learned' from the event. Lessons learned have, however, a different connotation and.are not systematically related to the safety significance of the events. ~Therefore.'it was suggested not to be used as a criterion. This parameter would be included,: when appropriate, within the "defance-in-depth" and the aon-site impact" attributes.
Finally, the advantages and disadvantages of. including energency protective actions in the scale vere considered.
The group agreed that the upper part of the scale should be, if possible, consistent
-with emergency planning provisions and allow the classification of.
t major nuclear accidents.
Descriptions should be added to the higher levels of the scale to give examples of possible emergency actions which could be expected.
e) The number of levels to be adopted for an international scale was also considered. The severity scales currently in use have, respectively, six levels in the French scale and eight levels in the Japanese scale.
A number.between six and eight was regarded as t
appropriate.
The French and Japanese scales were examined in some-detail, as a possible basis for an international scale.
It was agreed that the
" accident" part of the scale (the three upper levels) might require to be expanded to include an additional level in order better to separate accidents of considerably different safety significance or severity (such as TMI and Chernobyl).
The participants did not reach a conclusion on this issue. which vill require further consideration.
Similarly, the adequacy of having the " incident" part of the scale (the three lower levels) only divided into three levels was not addressed by the group and may require further consideration.
l 1
- k. ;
. c, 12:
n di sb p
f) The definition of'the levels (rows of the matrix) was discussed.
L The group believed that guidelines need to be defined for this L
purpose but recognised the difficulty this may cause because of.
L, differences in national regulations and practices' including safety and impact assessment methods. However, the group felt that,- in i
assigning to an incident or accident a level'of classification l' '
according to each of the attributes of the scale, a different veight should be' attached to each attribute to take account of the l-different impacta-that the various attributes have in terms of I
public concern. -This implies that the classification levels L
available fer each attribute vould not necessarily encompass the whole range of levels in the scale.
It appears important that, within a particular level, the safety.
significance is reasonably similar for the different attributes.-
It was also felt by the group that the various intermediate levels of the scale should be, where possible, related to and checked against vell-recognised health and safety standards.
The participants did not discuss in detail the criteria for the establishment of the classification levels in the international scale. Bovever, as a contribution for further work to this.
purpose, an example.of possible levels of classification is
. suggested in the following:
Events exceeding technical specifications but without consequent leakage of radioactivity beyond designed barriers in quantities exceeding permissible levels for normal-operation.
Events exceeding technical specifications and with consequent release to the environment not exceeding permissible levels.
Events within the design limits but without consequent release exceeding permissible levels outside boundary.
- Events beyond design limits and with severe fuel damage but vithout consequent release exceeding permissible levels outside boundary.
Severe accidents and consequent releases requiring countermeasures to be taken according to emergency plans.
Catastrophic accidents with consequent releases that are beyond the scope of emergency plans.
g) Finally, the group considered that the presentation of the scale, besides its content, was also very important.
In this respect. it was felt that, for presentation to the media and the public, the international scale should be based on simple and easily intelligible descriptions of its levels; however, for its application by the operators and regulators it should be supplemented by detailed criteria of use and examples of application.
(,
13'
'4.
StrFING UP AN INTERHATIONAL SEVERITY SCALE There was consensus among participants that the provision of an 4
. international severity scale is essential.
The basic requirement for the scale stems from the fact-that this vill give a clear picture of events to the i
public and the media should help to minimise misinterpretation of events.
Other benefits which vill accrue from a scale vould be to provide intelligible information to governmental organisations under the Convention for Early Notification, allowing unambiguous interpretation by the countries affected.
The participants noted that some countries, e.g. France and Japan, have established scales and are using them.
Several other countries are developing severity scales, but they are not_used in a formal manner for public notification.
It was, therefore, agreed that an international scale should be established and implemented quickly, otherwise more countries vill find it necessary to introduce their own domestic scales, with the disadvantages previously noted.
Discussion was centred around the following three points:
(i)
How to establish an international severity scale.
(ii) Bov to validate it once this severity scale has been established.
(iii) What type of guidance and training should be provided to the countries to assist the operators, the regulatory authorities and'the media on the interpretation and use of the scale.
one proposal discussed by the participants was that the IAEA and NEA should convene a panel of experts to prepare a recommended severity scale l:
which should then be made-available to Member countries for application.
It vas also. agreed that the proposed scale should be validated.for consistency
(
and lack of ambiguity. For this purpose, it was felt that-the use of a collection of data on past events, including data extracted from the Incident
-Reporting System (IRS of NEA and IAEA), to validate the scale for events at nuclear power plants was desirable.
Member countries should be encouraged to contribute to this validation by providing additional data'as well as l'
information on other notable events (besides those reported in the IRS) that vere or could be made public.
In the opinion of the participants, it vill also be useful to prepare 1
guidance on the use of the severity scale, in the form of a user's manual, to 1
L avoid obtaining different results due to different interpretations of the scale.
Many examples of application to various types of events vill be useful reference material for those who use the scale.
The IAEA and NEA should distribute the result of the above described validation to the Member countries as part of the guidance relating to the scale.
In order ter ensure consistency and prevent undue delays, it was considered beneficial that the same group of international experts produce the scale, validate it and provide initial guidance notes for its application.
i l'
L I '..
^t 3.:
'3 14 The participants agreed that the-international severity scale and its guidance notes should be initially tested in Member countries for a= trial period _of about a year.- The role of the IAEA, NEA and national authorities i
during the trial period of application was discussed by the participants.
In particular, the importance of reviewing the implementation of the scale was stressed.
For this-purpose,-it was felt that the NEA and IAEA should monitor the progress in the use of the scale during the trial period and review its results before recommending permanent application of the scale.
For implementing the scale, it was recognised that the decision on the degree of the involvement of the different parties concerned is a matter for the national authorities and organisations involved-in introducing the scale.
The involvement of the press (and other media) during the trial use of the-severity scale was also discussed. The general view was that such involvement is also a matter to be dealt with at national level, but, at the launching of the severity scale, an appropriate and timely press conference organised by the IAEA/NEA (in collaboration with national authorities) vas suggested as a useful initiative to give the necessary publicity to the scale.
Finally, the participants stressed the need to update the guidance notes from time to time and to offer appropriate training on the interpretation and use of the scale.
For this purpose, it was suggested that the IAEA and NEA should consider offering such a training for operators, regulatory bodies and other parties involved.
j In conclusion, the participants offered the following recommendations:
al
_Establishe_ent of_the_Internati_onal_ Severity Scale 1.
NEA and IAEA should develop an internationally agreed severity scale and' provide sufficient underlying criteria to ensure consistent operation of the scale in different countries.
2.
Based upon the exchange of information in October _1989, it is judged that international experts could-provide a draft scale early in 1990.
NEA and IAEA are invited to convene an Expert Group to achieve this.
3.
NEA and IAEA should subsequently convene a Technical Committee, in -
Spring 1990, to agree the scale, criteria and guidance. NEA and IAEA should then offer the scale to Hember countries for a trial period of one year, during which time NEA and IAEA should review progress in the countries villing to participate in the trial.
4.
NEA and IAEA are invited to agree with Member countries the text of a press communique announcing the trial of an international scale, and agree the timing so as to aid the launch in certain Member countries.
5.
At the end of the trial period NEA and IAEA should reviev the experience before finally endorsing and recommending the scale to Member countries for general use.
l
" j,:
,s, 15 5
i b)
!alidation_of the Scale l
NEA and IAEA'are' invited to obtain consent from Member countries to 6.
make available a selection of events from IRS records and other notable events to allow validation.
7.
.It is recommended that, as far as practicable, the same group of international experts produce the scale, validation and initial ~
. guidance notes.
f c)
Gui,d_ance Note,s 8.
NEA and IAEA should obtain from Member countries a selection of notable events that can be made public and be embodied in guidance notes, to act as reference marks.
9.
NEA and IAEA should provide these guidance notes in the form of a user's manual.
10.
NEA and.IAEA should update the guidance notes from time to tira, as requested by Member countries, for subsequent agreement.
d)
Training.
11.
NEA and IAEA should consider the training needs of operators, regulators, media, and arrange for vorkshops and courses to this q
purpose.
-(
i
16
,s.
- s.
(
s 5.-
SUlst&RY CONCLDEIONS, l
-1.
There was' general agreement that the formulation of a severity scale by an expert group, which could be made available to countries as a recommended international scale, would be strongly supported.
2.
There was general agreement about the purposes and ideal characteristics of such a scale.
In particular, the main purpose was seen to be to give the media and public a rapid and clear assessment of the severity of nuclear incidents on an internationally agreed basis.
3.
Significant progress was achieved in defining the characteristics which such a scale should have.
It was agreed that it should be a multi-dimensional scale based on defined criteria. The attributes (columns.of a matrix) on which to base the scale, should be off-site radiological impact, on-site radiological lapact and defence-in-depth.
It was also agreed that economic criteria and the ability to learn lessons from incidents and events are not in themselves suitable attributes to include in a scale. It was concluded that the levels (rows of the matrix) might be encompassed by six to eight points. Furthermore, the upper levels of the scale should, if possible, be consistent with emergency planning provisions and descriptors should be i
added to the highest levels of a scale giving examples of expected emergency.
actions.
L 4.
The way that the scale is presented was regarded as important.
It was l
agreed that for the general public and media the scale should be based upon L
simple intelligible descriptions of the levels of the scale, but, for the use of operators in determining the classification of incidents and accidents, detailed criteria with examples would be required.
5.
-It vas agreed that a group of international experts should.be convened by the NEA and IAEA to provide a draft scale early in 1990.
Following this, the NEA and IAEA should convene a technical committee to agree the scale criteria and provide guidance. The scale should then be offered to Member i
countries for a trial period of one year, during which the NEA and IAEA j
experts should review progress in the countries villing to participate in the 1
- trial, j
6.
Because of the importance of public acceptance and understanding, the NEA and IAEA should agree with interested countries the text of a press
.=
communiqus announcing this trial and should agree the timing of such an announcement,-so as to aid the launch in Member countries.
7.
At the end of the trial period, the NEA and IAEA should review the experience of those participating in the trial and would, if found to be suitable, finally endorse and recommend the scale to Hember countries for general use.
8.
It was agreed that it was important to provide validation of the scale, and the NEA and IAEA vere invited to obtain the consent from Member countries to make available a selection of events from IRS records and other l
sources to allow such validation. This should be carried out by the same group of intenational experts who produce the draft scale.
Ei, <
17'
},79 -@,.:
9.
~
... It was agreed that guidance notes to act as reference marks for:the scale should be produced'and.these guidance notes should~be. compiled by the.
NEA and IAEA in: the form of a user's manual.- These should be updated ~ from
- time-to time, as requested by Member countries.. The NEA andLIAEA~should
. consider training needs and make proposals for workshops and courses in order to ensure the consistent _and proper' application of.the scale in Member
. countries.
J
)
?
b G
7
'?
L l
l-p 7
l.
l-l l-,
- > K 18-AM I LIET OF PARTICIPANFS INTERNATICIIAL A'ItlIIIC ENERGY AGEIICT l
Ale OECB/ NUCLEAR EMBEGY AGENCY
'J0 INT TECllitICAL COWETTEE NEETIIIG ON EIFERIIIICE i
AND USE OF E=' Lwa --. anm g.f,.- a savuun FAf **
I 23rd - 26th OCTOBER 1989 OECD, Paris, France
. Scientific Secretaries:
O. Ilari (NEA)
B. Thomas (IAEA)
NATRMAN:
DR. R.R. TATLOR BRIGIUM Mr. B.A. DECKERS Vincotte 157, Avenue du Roi i
1060 Brussels Mr. Ph. de GREVE Minist4re du travail Service de la sicurit6 technique des installations nucl6aires rue Belliard 51.
A-1040 Brussels CANADA Mr. M.R. RH1AUME t
Centrale Nuclsaire C.P. 360 Gentilly Qudbec G0K IGO
'CZECB051AVAKIA Mr. L. NAMESTEK State Nuclear Safety Inspectorate
[
c/o Mrs. Marxbauerova L
Czechoslovak Atomic Energy Commission SlerskA 9 Prague 9 i
wr 19 i
FEDR &L REPUBLIC OF G BMANY
.Mr. K.J. K01TB0FF Gesellschaft fQr Reaktorsicherheit Schwertnergasse 1 D-5000 K61n l' Dr. R.D..VENDLING Federal Ministry for Environment, Nature Conservation and Nuclear Safety, P.O. Box 12 0629 D-5300 Bonn 1 FINLAND S
Mr. B.A.O. REGNELL Imatran Voima Oy P.O. Box 112 i
01601 Vantaa FRANCE 1
Mr. J. BREUIL Service Central de SQret6 des Installations Nuclsaires-Minist6re de l'Industrie et de l'Amsnagement du 1
- i Territoire a!
99,, rue de Grenelle i!
75700 Paris Cedex
.j Mr. R.B. COULON IPSN/DPS Commissariat & l'Energie Atomique
)
B.P. No. 6 92265 Fontenay-aux-Roses
.I Mr. M. DEBES.
l Electricit4 de France Service de Production Thermique i
13-27 Esplanade Charles de Gaulle 92060 Paris La Defense Codex 57-
- j Mrs. M-C. DUPUIS Laboratoire d' Etude du Facteur Humain IPSN/DAS Commissariat & l'Energie Atoalque B.P. No. 6 92265 Fontenay-aux-Roses Cedex v
=. F p.
'W
. Mc, Wh c
-4 20
'l o.
.t Ms. M.-CONTE I
IPSN/DAS'
~ t Section of Regulation, Documentation, Foreign Relations y
Commissariat & l'Energie Atomique B.P. No. 6 91265 Fontenay-aux-Roses Codex 8
Mr. H.C. GUIMBAIL Electricit4 de France Direction Gendrale 32,-rue de Monceau 75008. Paris
\\
ITALY i
Mr. S. BEHASSAI ENEA/ DISP Via Vitaliano Brancati, 48 j
I-00144 Rome 4
JAPAN-Mr. M. HARIMA Nuclear Power Safety Information.Research Centre of NUPEC j
3-17-1 Toranomon Minatoku, Tokyo l
Mr. M. HATTORI Stience and Technology Agency 2-2-1 Kasumigaseki Chiyoda-ku, Tokyo Dr. S. KONDO University of Tokyo Department of Nuclear Engineering 7-3-1 Bongo Bunkyo-ku, Tokyo-r Mr. K. NII Nuclear Power Safety Information Research Centre of NUPEC 3-17-1 Toranomon Hinatoku Tokyo L
HEZICO i
Mr. S.M. VALLER CNSNS r
Av. Insurgentes Sur 1806 Col. Florida Del. A. Obreg6n D.F. 01030 tl.
r
l l :a r.
EA Mr. A. de J00DE Ministry of Social Affairs and Employment
{
Nuclear Safety Department Balen-van Andelplein 2 Voorburg v.
Mr. M. NASIM Director Nuclear Safety & Radiation Protection Pakistan AEC P.O.-Box 1912 Islamabad I
i' SPANI L
Mr. L. ANT 0LIN
.j C.N. Trillo
)
Rosario Pino, 14-16 28020 Madrid Mr. J. GIL Consejo de Seguridad Nuclear Justo Dorado, 11 E-28040 Madrid SUIDEN
'Mr. S.R. CARLSSON National Institute of Radiation Protection
~
Box 60204 104 01 Stockholm Mr. E.T.R. NILSSON Swedish Nuclear Pover Inspectorate
[
Box 27016 S-102 52 Stockholm SUITZERLAND Mr. P. BUCHER Swiss Association for Ator.ic Energy P.O. Box 5032 4
CH-3001 Bern IRIION OF S0VIET SOCIALIST REPUBLICS s
Mr. V.F. FEDULOV Division Head Science and Engineering Centre State Committee for Supervision of Nuclear Power Safety (SCSNPS)
Moscov
fu_
22
., y ;.. * *.'s 1
fe
.Mr. S.A. LEON 0Y
'I c/o USSR Ministry for Nuclear Power K-74 Moscow 103074 7 Kitaisky Proyesd V
i UNITED KINGDON I
Mr. P.K. BASU Nuclear Installations Inspectorate h
j
-Baynards House i
L 1 Chepstov Place h
London V2 4TP l'
Mr. B.C. CARPENrER h
~
UK Atomic Energy' Authority l:
11 Charles II Street l
London SV1Y 40P Mr. V.R. GOODVIN 1;
British Nuclear Fuels ple p,
Sellafield, Seascale Cumbria, CA20 1P Dr. R.R. TAYLOR
.i o
CEGB t
L Courtenay House 18 Varvick Lane London EC4P 4EB 1
Mr. P.A. V00D800SE i-L Nuclear Installations Inspectorate St. Peter's Houso b
Stanley Precinct D
Balliol' Road l
Bootle Mersey:ide UNITED STATES Mr. E. JORDAN.
Office for Analysis and Evaluation of Operational
'I
1 l
COMMISSION OF THE EUROPEAN COMMUNITIES u
Mr. G. BARBERA Joint Research Centre Ispra Establishment 1-I-21 Ispra (Varese)
Italy Mr. J-P PELE Division Sorets des Installations Nuclsaires 200, rue de la Loi 1049 Brussels 1
i 23 DifRH&TICR&L Attst!C 3m57 MiWKT Mr. B. TBONAS Division of Nuclear Safety Mr. B. TONIC Division of Nuclear Safety 90RLD ASSOCIATION OP NUC12AR OPERATORS Mr. S. SILBER 8 ERG 262a Fulham Road London SV10 9EL United Kingdoa ORGANISATION FOR BCONOKIC Co-OPERATION AND DETE1APMElfr Environment Directorate Mr. H. SMETS Nuclear Energy Agency Dr. O. ILARI Deputy Head Radiation Protection and Vaste Management Division Dr. G.D. McPHERSON Bead. Nuclear Safety Division Mr. K-B. STADIE Deputy Director, Safety and Regulation l
Mr. C. VIKTORSSON Radiation Protection and Vaste Management Division t
__