ML19257B612

Rept to AEC on Reactor Licensing Program, June 1969

ML19257B612
Person / Time
Site:	Crane
Issue date:	06/30/1969
From:	Case E, Mangelsdorf H, Nyer W Advisory Committee on Reactor Safeguards, Atomic Safety and Licensing Board Panel, Office of Nuclear Reactor Regulation
To:
References
TASK-TF, TASK-TMR NUDOCS 8001170783
Download: ML19257B612 (54)

Text

I I ly

,a l'

REPORT TO'THE ATOMIC ENERGY COMMISSION ON THE REACTOR LICENSING PROGRAM by the INTERNAL STUDY GROUP June 1969 f

1916 359

a r

(

CONTENTS 2

PAGE I.

PREFACE 1

II.

INTRODUCTION 3

.g c:

III.

CONCLUSIONS AND RECOMMENDATIONS 7

A.

DEVELOPMENT OF REGULATORY CRITERIA AND STANDARDS RELATING TO SAFETT r3 B.

DIFFERING VIEWS ON REACTOR SAFEI. REQUIREMENTS 12 C.

SAFETY RESEARCH AS RELATED TO THE LICENSING OF POWER REACTORS 16 D.

RELATIVE EMPHASIS ON LARCE AND SMALL ACCIDENTS 21 E.

QUANTIFICATION OF SAFETY 25 F.

DEGREE OF STANDARDIZATION AND IMPOSITION OF ADDITIONAL SAFETY REQUIREMENTS 31 C.

CRITERIA FOR DECIDING WHEN TO BACKFIT AFIER ISSUANCE 07 A CONSTRUCTION PERMIT 37 H.

THE ROLE OF THE ACRS IN THE REGULATORY PROCESS 39 I.

TIMING AND STAGING IN THE REVIEW AND DECISION-MAKING PROCESS 45 st 1916'360

M k#

de t.+

vNN NV' IMAGE EVALUATION TEST TARGET (MT-3)

\\.0

• M Bu E EB Ej=i2 g,3 l.l f,

• EE I.8 1.25 1.4 1.6 4

6" MICROCOPY RESOLUTION TEST CHART

1. 4

$+%$

4%

• sifh/

t><

4

p i

4 4

$++<%> $e'e h

,_E E A<e 1,em TEST TARGET (MT-3) 1.0 lnm an is en p" u

's iLt l,l h,ES {g l.8 1.25 I l.4 1.6 I

6" MICROCOPY RESOLUTION TEST CHART 4%

4s k) p

• $[5h//h/

M)[4 y

/s 4

A

't II (>

/d <(f

+

TEST TARGET (MT-3) 1.0 5 m DM

!!!i} E I.l l 5 EM l.25

! l.4 1.6 I

6" MICROCOPY RESOLUTION TEST CHART

1. *i 4$ 4%

4#M

1. $$M b v

ix 7 <>d ew 0

i

.h

4+4%

g<s#A

$+-

4

,M eE Ev <e 1,em TEST TARGET (MT-3) 1.0 ERMEM E b$ E I.I i.'" lBM l.8 1.25 1.4 1.6 4

6" MICROCOPY RESOLUTION TEST CHART 4%

1. DS+

//p figAxxx)/#

Q4 of 3 <;,#

4

9

'#dkII*

k'k IMAGE EVALUATION NNNN TEST TARGET (MT-3) 1.0 l9FM EM f(( GE l,l

[ " !!!lde

'l I

.8 l.25

' l.4 i 1.6 I

i 6"

MICROCOPY RESOLUTION TEST CHART

1. • e4

+<4

• S

• Iff,h id,,>8 4

l -

3.

,a

(

PREFACE On July 8,1968, the Atomic Eacrgy Commission announced plans for an internal review of its regulatory program to help assure that procedures keep pace with the rapid expansion of the nuclear industry.

In this review, designed to be primarily technically oriented, the present process for the licensing of power reactors was to be examined frcm the standpoint of ef ficiency in the df acharge of regulatory responsibilities and compatibility with the commercial arrangements by which nuclear plants are purchased, designed, constructed and operated. The purpcse of the review was to recommend possible improve-ments in the licensing process, and to determine whether further d(tailed Commission study in any areas would be desirable.

. n:

i To conduct this review, the Commission named an Intnrnal Study Group drawn f rom the three principal components of the AEC regulatory system -

the staf f operation headed by the Director of Regulation (the regulatory staf f), the Advir -ry Committee on kcactor Safeguards (ACRS) and the Atomic Safety & Licensing Board Panel (ASLBP).

Members appointed to serve on the Internal Study Group were:

Harold G. Mangelsdorf, Chairman (Member, ACRS)

Warren E. Nyer, Vice Chairman (Vice Chairman, ASLBP, 1968)

Edson C. Case (Director, Division of Reactor Standards. AEC) 1917 001 John W. Crawford, Jr.

(Assistant Director, Division of Reactor Development and Technology, AEC)

David B. Hall (Merber, ASLBP; Chairman, ACRS, 1963)

Stephen H. Hanauer (Chairman,.ACRS, 1969)

Peter A. Morris (Director, Division of Reactor Licensing, AEC)

Carroll W. Zabel (Chairman, ACRS, 1968)

Marcus A. Rowden, Assistant General Counsel for the AEC was eppointed to serve as legal counsel and Ray G. Smith of the AEC's Division of Reactar Standards was selected as Technical Secretary.

The Internal Study Group met at approximately bi-weekly intervals during the period between July 1968, and April 1969. Discussicus were held with representatives of publicly and privately owned utilities, reactor manufacturers, architect-engineers engaged in the design and construction of nuclear power plants, and various industry associations.

Discussions were also held with Federal Aviation Administration repre-sentatives, two members of an earlier regulatory review panel and senior representatives of the AEC regulatory and other staff.

The Internal Study Group has arrived at a number of-conclusions, and has developed some related reco=mendations. The purpose of this report ie to present these conclu:sions and reco==endations and to discusa ihe reasoning behind them.

191700h

( -

(

II.

INTRODUCTION The Internal Study Group undertook its review of the reactor licensing process with a view toward making suggestions for ways to improve the process and its effectiveness in protecting the health and safety of the public, while at the same time minimizing the problems faced by the nuclear industry and the regulatory groups.

In addition to specific aspects of the licensing process, the Group considered the general questions of (1) the adequacy of the pro-tection of the health and safety of the public and (2) whether regulatory procedures and requirements have adversely affected the development of the industry.

The Group concludes.that the health and safety of the public k

has been adequately protected.

At the end of 1968, 17 licensed power reactors had been built and operated, nine of which continue in routine operation, with an accumulated experience of 88 reactor years. No member of the public has been exposed to radiation levels above permissible annual limits as a result of the operation of these licensed power reactors. Although this experience does not provide conclusive proof of safety, it does provide some indication that the health and safety of the public is being protected.

The general consensus among the industry representatives who talked with the Study Group was that there is a high degree of conservatism in nuclear plant designs and in safety reviews, and that this conserva-tism is at Icast as great as that of most other major industrial activities.

It was also agreed that this conservatism has contributed 1917 003 substantially to the good safety record of the nuclear industry and that it is not out of proportion to the needs of the industry.

The Study Group concurs in this view.

The groups and individuals who talked with the Study Croup were asked if they believed the regulatory bodies had required unnecessary or superfluous safety features. Few examples were cited of safety requirements that were ' elieved not to be needed.

A few persons ex, ssed the belief that future experience might show that some systems need not be required; however, the principal concern was not with the requirements themselves, but with the addad complexity and, perhaps, excessive redundancy that seemed to result and the uncertainty as to tha safety requirements and the timing for their imposition.

In general, the consensus of the industry repre-sentatives - concurred in by the Study Group - was that safety feature requirements have not been out of proportion to the need for the pro-tection of the health and safety of the public.

There was also general agreement that the licensing review process at the construction permit stage has not been a limiting iten in the time schedule for the construction of planta, although it could become so in the future.

In the latter connection, a mejor problem considered by the Group, to which some of its specific recor=endations are related, is the lack of correspondence in timing of decision points in the current licensing review process with decision points in the process of industrial procurement, construction and initial operation of nuclear plants.

This inck of cor4TTndence, 1

004 e

i k

together with uncertainties in regulatory requirements, has resulted in some hardship and f rustration to some elements of the industry and could become even more troublesome as the market for nuclear power expands.

The problem mentioned most by industry representatives was un-certainty and instability in regulatory requirements. Utilities faced with decisir - on the addition of generating capacity have not been certain of ultimate licensing requirements at the time of their selection of the type of nuclear power units to be installed and when they con-tracted for those units. There have been related increases in costs and changes in plant scheduling and manpower requirements which were not anticipated when utility selection and contracting decisions were 7

made.

Many of these unanticipated changes were the result of safety

\\

requirements impo. sed by the regulatory groups to maintain adequate margins of safety consistent with increases in reactor size and powar density. Neither the industry nor the regulatory groups had foreseen the extent of the provisions that would be needed to maintain these safety margins.

In the long run, the greater stability and predictability of regulatory requirements which both industry and the regulatory groups seek will depend on the development of comprehensive safety criteria, codes and standards.

While there has been considerable progress in that direction by both the Commission and industry, more needs to be done - a matter dealt with elsewhere in this report.

As the designs for the never and larger nuclear plants are evolving and as more comprehensive regulatory safety criteria are 1917 005

being developed, it is especially important that all groups' involved in the review process, and also those involved in the conduct of research and development programs directed toward resolution of potential safety problems, maintain effective co=munication with each other. The Study Group notes that the effectiveness of this communi-cation has steadily improved.

In particular, the Group supports.he current joint efforts at the staff. level within the Commission to define regulatory needs and to orient Commiselon-supported research programs toward resolution of those needs on a timely basis. These efforts should be continued and expanded to include active participation by the nuclear industry, partiet.arly by utilities.

The Study Group discussed at some length the cont'.auf ng problem of maintenance of an effective regulatory process in an industry marked by a rapidly developing technology and large increases in the number of applications to be processed. The Study Group believes that the AEC regul story staf f should continue to be the only regulatory body to perfo m a complete technical review of each reactor application. TheI regula.cory staff should have suf ficient strength - in manpower and other resources - to carry out in a timely fashion the activities necessary to assure that the regulatory process providen effective protection to the health and safety of the public.

Detailed discussion of many of these points will be found in the following sections of this report.

1917 006

(

III.

CONCLUSIONS AND RECOMMENDATIONS (t

1917 007

(

7-

A.

Development of Regulatory Criteria a,nd Standards Relating to Safety The lack of a comprehensive set of regulatory safety criteria and industry codes and standards relating to the safety of nuclear power.

plants contributes to the uncertainty concerning regulatory requirenents and to the length of time required to conduct regulatory safety evaluations.

CONCLUSIONS AND PICOMMENDATIONS 1.

Current efforts to develop and implement comprehensive regulatory safety criteria and industry codes and standards relating to the safety of nuclear power plants should be intensified consistent with the critical importance of such criteria codes and stand-ards for improving the regulatory process and benefiting the nuc/. ear power industry, 2.

There is an urgent need for substantially increased participation in and support of these efforts by all segments of the nuclear industry, especially the utilities.

3.

The ACRS should expand its participation in the develop-ment of regulatory criteria and standards relating to safety.

e m

=%

k 1917 0138

t 1

k DISCUSSION The Study Croup believes that significant benefits to the nuclear industry would result from the development and implementation of com-prehensive regulatory safety criteria and of industry codes and standards relating to the safety of nuclear power plants. Such criteria, codes and standards would contribute measurably to industry understanding of licensing requirements and would, at the same time, furnish an improved means for demonstrating that the necessary requirements have been met.'

Their use would result in better definition of the information to be supplied in license applications and so aid in reducing significantly the time required for regulatory reviews.

In addition, the Group believes a greater effort by both industry and the AEC to describe the

(

technical bases for safet'y requirements would be beneficial.

The Study Group in its discussions with industry representatives found substantial recognition of the need for, and importance of, hastening the development of regulatory safety criteria and of industry codes and standards relating to safety.

The Group found that the organizations affected are strongly interested in reviewing proposed criteria, codes and standards before they are put in effect.

It was less clear that such organizations rc: gnize the importance of parti-cipating in and actively supporting current standardsmnaking efforts within the nuclear industry.

~

The Commission has been strengthening its efforts on the develop-ment of regulatory safety criteria, codes and standards, including cooperative ef forts with professional standards groups, the supplier 1917 009

~

industry and the utilities. The nuclear industry, through the United States of America Standards Institute, professional societies and industry associations, has also taken some strengthening measures to develop codes and standards.

In addition, some companies have in-dividually recognized their responsibi.11 ties and are actively supporting these efforts.

Ilowever, despite the significant progress that has been made in the last few years, the Study Group believes that the rate of accomplishment by the Commission and the industry has not been fast enough to meet the needs of the rapidly developing industry.

While more technical information is needed before development of comprehensive regulatory criteria can be completed, the Group believes that the basic organizational strut _ures and technical capabi-lities for developing the needed industry safety codes and standards already exist. The urgent requirement is for all segments of the nuclear industry to recognize their vital interest in supporting such efforts and to implement that recognition through aggressive leadership and the furnishing of knowledgeable personnel on a high priority basis.

The Group is aware that the procedures associated with developing and promulgating regulatory safety criteria and industry codes and standarda have, traditionally, been time-consuming. This postpones their availability for use.

The effect of this traditional pattern is compounded by the time required to construct plants, with the result that current efforts will not be seen in operating nuclear power plants for many years. The Group believes that measures to reduce these delaya shculd bc t kan by the regulatot-groupe, by =tendards ching organizations and by the nuclear industry.

1917 010

_ 1o _

s

/

While emphasizing the responsibility the nuclear industry has in developing codes and standards relating to safety, the Group is aware of some of the difficulties that are involved.

These difficulties derive'from many factors, including procedures which require near unanimous agreement. to establish such standards or to modify them.

In vien of the nature of these factors, it is not realistic to expect that industry's ef forts can result in safety codes and standards which alone will be adequate to protect the public. Accordingly, there will continue to bef a need for the Commission to develop and promulgate supplementary regulatory requirements. However, to the degree that industry succeeds in adopting codes and standards which meet safety requirements, the need for additional requirements imposed by the

/

k Commission will be reduced.

1917 011 B.

Differing Views on Reactor Safety Requirements There are differing views among those in the nuclear industry, the regulatory groups and others, as to how the safety of nuclear power plants can best be provided. There are differences of opinion

.4.

on the degree of reliance which should be placed on the reactor system itself and on engineered safety features; the number of such

~

features required; and the kinds of failures to be considered. There are differences of opinion on whether, and to what extent, trade-offs can be made among the various safety elements. For example, can there be a reduction in the extent of duplication and layering of systems designed to limit the consequences of accidents if it is known that an effective quality assurance program has been applied throughout the s_

design and construction of the plant? Alternatively, can the per-formance specifications for reactor containment be reduced if there is adequate assurance that fuel melting cannot occur?

RECOMMENDATION The Commission should adopt the policy that the greatest emphasis and priority be placed on the application of quality assurance to the design, construction and operation of nuclear plants so as to achieve the exacting level of safety required.

DISCUSSION

~

The Study Group observes that the primary objective of the nuclear power industry is to build and operate afe, reliable and.

economically attractive generating plants. The Group is convinced that attainment of this objective requires rigorous application of quality assurance procedures in the broa de <t sense.

The achievement of an adequate level of safety for nucicar power plants is generally recognized to require defens&in-depth in the design of the plant and its additional engineered safety features.

The degree of emphasis on defense-in-depth in the nuclear field is new to the power industry.

In seeking reliability of t icty systems, there has been much attention in the nuclear field to reaundancy, diversity and quality control.

As a result of the evolution of designs, and the large number of new orders for nuclear plants, questions have been raised regarding

(

the proper balance among back-up systems with respect to the require-ments of basic plant design.

The Study Group endorses the defense-in-depth concept, but believes that the greatest emphasis should be placed on the first line of defense, i.e., on designing, constructing, testing, and operating a plant so that it will perform during normal and abnormal conditions in a reliable and predictable manner. This assurance of quality is obtained only if safety requirements are clearly and adequately defined, plant designs meet these requirements without excessive complexity, construction is in accord with design, and operation and maintenance assure continuing conformance with safety criteria.

The need for greater emphasis on quality assurance is supported by recent experience which has revealed a number of defects in reactor construction and deficiencies in mreting

(

1917 013 design objectives that have required correction in plants currently under construction.

A number of factors are included in the recoemended emphasis on quality assurance, all of which need attention to achieve the overall goal. Among these are independent reviews of design, construction, tests and operation; deliberate use of inherent safety features; pro-vision for effective inspection, maintenance and surveillance; accurate records; and performance evaluation.

The Group considers that industry has not demonstrated sufficient recognition of the contribution to safety that can be made by designing, constructing, testing and operating a plant so that it will perform during normal and abnormal conditions in a reliable and predictable manner.

Consequently, greater attention should be givcu by the industry to establishing the strong quality assurance programs that experience has shown are needed to achieve this kind of performance.

The Study Group believes that nuclear power plants designed, built, tested and operated in a disciplined manner with exacting standards of quality will provide an adequate level of safety.

Never-theless, abnormalities, deviations and acaidents may occur; and with no operating experience with the large reactors of new design, the Group does not believe that current requirements for engineered safety features are excessive or that trade-offs should be made at this tLse

, t' e various elements of design contributing to defense-in-depth.

at In

.cgard, most of those to whom the Group talked could offer no suggestion for elimination of the systems now provided.

(I 1917-014 4-

(

There was universal support for strong quality assurance programs among those to whom the Group talked.

There is, however, a wide vari-ation of understandir.3 and experience in this area. The Group conciars in the definition of quality assurance set forth in the " Nuclear Power Plant Quality Assurance Criteria," being developed for issuance by the Commission:

" Quality assurance comprises all those planned and systematic actions necessary to provide adequate confidence that a structure, system, or component will perform satisfactorily in service."

The term is used in the broad sense of applying quality assurance f(

throughout all phases of the design, construction, testing and operation of a nuclear power plant.

It includes quality control, which comprises those actions related to the physical chat a eteristics of a material, component or system and which provides a means to control quality to predetermined requirements.

In view of the wide variation in understanding of quality assurance and the need for increased attention to this matter, the Group believes it would be useful for the Conctission to adopt the policy of putting first priority and emphasis on quality assurance, as earlier define d, in providing nuclear plants with an adequate level of safety.

In sum, the Group believes that greater emphasis should be placed on providing sound quality assurance programs in the nuclear industry and that there should be no present reduction in the requirements for l

back-up or consequence-limiting safety features in c.urrent designs for 1917 015 water-cooled nuclear power plants.

C.

Safety Research as Related to the Licensing of Power Reactors i

Consistent with the Commission's two str.ge licensing process, construction permits fer large water-coolad power reactors have been issued on the basis that there is reasonable assurance that safety questions requiring research and development will be satisfactorily resolved prior to completion of the proposed facilities.

Because of the large number of these plants scheduled to begin operation id the next few years, special attention must be directed to assuring that the required research is being conducted and that the results will be availabic when needed.

RECOMMENDATIONS 1.

Applicants, reactor desigaers and the regulatory groups should continue to refine and formalize the existing practice of identifying at the construction permit stage for each nucicar power plant the safety issues requiring resolution by research, and the scope and schedule of programs exrected to provide the information needed to resolve these issues.

2.

Present AEC cfforts to document and review Commission and industry-sponsored safety research programs to determine whether these programs are adequate for timely resolution of safety issues for large water-cooled 7

jg nuclear power plants should be continued and expanded to include active participation by reactor designers and utilities.

3.

If necessary research programs are not being conducted, or are not sufficiently responsive to the identitied needs, alternative courses of action should be d sveloped and implemented by the AEC and the nuclear industry.

DISCUSSION Both the Connission and the industry sponsor nuclear safety research programs to support the development of power reactors.

Historically, the AEC has financed much of the safety research for water-cooled power reactors and it continues to support these safety

~

research programs. The large number of construction permu r for these power reactors which have been issued in the last several years does not imply there is a decreasing need for water reactor safety research.

Rather, because these construction permits were issued on the basis that planned research progracs would resolve certain safety questions related to these reactors and because new questions have resulted froa the increases in reactor power level and power density, there is an increasing need for safety research. As a result, the combined IIC and industry research program on s 4fety questions related to water-cooled power reactors continues to grow.

It is dif ficult to distinguish safety research necessary for the licensing of individual power reactors from basic safety research.

that increases the understanding of saf ety-related technical qdestions.

There is some tendency to classify in the basic category all research 1917 017 needed to resolve the safety questions identified for large water-cooled nuclear power plants and to place complete responsibility on the AEC for cor. ducting the necessary progrsms. An opposing view considers the technical questions to be related only to specific plant designs and believes that AEC financial support for water reactor safety research should be significantly curtailed, if not eliminated. Whatever the view and irrespective of who finances the programs to develop the information, it is indisputably the licensee's responsibility to provide the information necessary to resolve the techni.. safety questions related to his nuclear power plant.

Most of the present safety research ef fort is directed toward providing information concerning potential accidents having very low probabilities of occurrence. Analyses have shown that the safety features provided in current designs of water-coclo' power reactors are adequate to protect against these accidents. While the Study Group believes it is reasonable for some nuclear power plants to begin operation with the assurance of adequacy of safety features dependent' on analyses, it is important that the.se analyses be confirmed in a timely fashion by the results from planned research programs.

The first necessary step to this end is to identify clearly for each plant at the construction permit stage the technical questions requiring research, and the scope and scheduling of programs expected to provide the information needed to resolve the questions. The Study Group is aware that, in recent licensing cases, ef forts have been made to document not only the technical issues for the particular power 1917 018

_y_

k reactors and the research programs necessary to provide information to resolve those issues, but to document also the schedules of the necessary research programs. The Group recommends that such efforts be further' systematized so that licensees, regulatory groups and those responsible for conducting research can more clearly understand what specific information will be required prior to operation of power reactors and the timing necessary for developing this information.

A second step to assure that research is oriented to meet safety and licensing requirements is to document and review periodically the existing and planned programs. The Study Group notes that efforts in this regard are underway by the AEC's development and regulatory staffs.

Tae Group further notes that increased communication between the i

Commission's regulatory and development staffs has resulted in better coordination of the AEC-sponsored research programs with regulatory needs. These staf f level actions have been fostered and encouraged by the AEC's Steering Committee on Reactor Safety Research.

The Study Group believes there should be increased participation by the nuclear industry, particularly by utilities, in current AEC cfforts to relate the research programs in water reactor safety to the requirements and schedules of the licensing process.

Discussions, at both management and staff levels, should be arranged among appro-priate AEC, reactor designer and utility personnel who are conversant with industry research and nuclear power plant licensing and safety.

Through better appreciation of each other's problems, both technical and financini, and by frank discussion of the technical safety issues, r

t

} 9 } [- h } h a better definition of the necessary scope and schedule of the safety research programs can be developed. Possible courses of action if these needs are not being satisfied can also be deter-mined.

Closer cooperation between AEC and industry on safety research would improve the basis ' for determining which of the needed safety research programs should be financed by the nuclear industry and which by the AEC.

Development of appropriate financial arrange-ments at this time, when the AEC is still sponsoring a major portioat of the water reactor safety research program, will ease the inevitable transition to the time when industry must initiate and fund most of this program.

G w

O i917 020

(

D.

Relative Emphasis on Large and Small Accidents The Study Group considered the problem of whether detailed considera' tion of a few serious postulated accidents is diverting effort f rom needed study of less serious, more probable malfunctions.

CONCLUSIONS AND RECOMMI.NDATIONS 1.

It remains necessary to consider, in safety reviews, a wide variety of expected transients and postulated accidents.

2.

The design basis accidents presently used in the safety evaluation of large water-cooled power reactors should I

not be changed until convincing technical evidence is k

available that the change is justified.

3.

An integrated engineering approach is needed; safety features added to cope with one malfunction should be engineered so as to not unduly increase the probability or consequences of another malfuncti,on.

DISCUSSION Early safety reviews included postulated complete reactor failure and an all-enveloping containment sufficient to maintain off-site exposure at a tolerable level and thus to protect the public. All other possible accidents were considered to have contained consequences less severe than those of the maximum accident; therefore, the containment provided protection to the I

1"917 021 public for small as well as large accidents. The increase in the size and complexity of reactors since that time has required.

development of a system of engineered safety features which would operate in accident situations to prevent fuel failure and melting, to remove afterheat, and to cope with a radioactive, hydrogen-rich containment atmosphere in order to provide containmer.c or confine-ment of radioactivity for the protection of the public.

The regulatory system for safety reviews has evolved from one based on consideration of a single worst accident into one which considers a spectrum of expected transients and a number of postulated malfunctions and design basis accidents.

For the larger reactors, ontaining the worst accident considered credible. does not insure that a L credible accidents would be contained.

Since it is not permissible An the case of large reactors to concentrate safety reviews on the worst accidents and to ignore the lesser ones, a spectrum of accidents 1917 022 is considered.

Which postulated accidents should be used as the design basis for reactot safety? In the absence of wide experience with reactor accidents, calculation and judgment must be substituted for such

. experience.

It has been suggested that the design-basis accidents presently used should be revised to reflect a more nearly consistent and mechanistic view of what would actually happen. The Group believes that sufficient knowledge is not available to justify such a revision, and that current use of the non-mechanistic set of design-basis accidents presently employed provides a needed safety margin to allow

(

for unknown factors.

Changes in these design basis accidents should be made only when justified by convincing technical evidence.

In designing protective systems and safety features to deal with postul'ated malfunctions and transients, an integrated engineer-ing approach is needed to prevent interference of one safety feature with another, or with normal operation. A device added to increase safety in some postulated situation should not unduly increase the probability or cont.equences of any other malfunction. This possible result is a criticism the Group heard of the present design-basis accidents, i.e. that detailed attention to serious, highly improbable postulated accidents might result in such complex systems that overall safety might be decreased.

Neither the industry representatives heard by the Group, nor the Group itself, believe that this decrease in safety has actually occurred; and no examples were suggested of safety features that decreased overall safety. The Study Group believes, however, that continued attention is required to the engineering of safety features in a consistent, disciplined way so that safety.is in fact increased by their installation.

In this connection, it seems worth noting that the reliability required of a safety device depends both on the probability of need for its function and the consequences of its failure to function when needed.

Implicit in this statement is the knowledge that risk cannot be made exactly zero, but that an extremely low, acceptable value must be maintained.

Postulated events of sufficiently low probability need not be protected against, since the risk from them is 1917 023 n

~

negligible.

For serious potential accidents of low, but not negligible, probability of occurrence, protective systems with good reliability are required. Situations for which the anticipated rate of occurrence is relatively high (say, once every few years) require prot.2ctive systems of extremely high reliability in order to reduce the risk to an acceptably low value. These more frequent events - expected transients - play an important role in the assessment of the overall risk and may determine the reliability

/

requirements of reactor protcctive systems. The Group believes this matter warrants further consideration.

Attention should be directed toward expected transients and non-catastrophic malfunctions for another reason:

they provide

/

potentially useful indicators of incipient safety problems in an operating reactor. The rate of such incidents, and also of malfunctions in the protective systems and engineered safety features, should remain at a tolerably low level, consistent with design expectations.

Dis-covery of a malfunction rate higher than that reasonably expected -

in particular, of any upward trend in the rate with time - should be considered evidence that something is amiss. Fer example, it is beside the point that an abnormally high rate of unneeded reactor trips is in the safe direction.

Such a circumstance is a sygton of trouble in the protection system which may include concealed changes adverse to the functioning of the system. For this reason, the continuing sur-veillance of incidents and malfunctions discussed in Section III.I of j this report is important to maintaining the availability of safety devices to perform their functions if needed.

1917 024

_3_

z E.

Quantification of Safety It has beca suggested that a more quantitative approach be used in thd evaluation of riska associated with nuclear power plants both to ettempt to establish overall risks and to provide technigpes for comparative studies of safety systems. Such an approach, if it could be realized, would assist in reducing uncertainties in the licensing process, particularly in areas which have. bcen criticized for lack of clear, consistent requirements. The Study Group considered whether and to what extent the safety evaluation process can be quantified.

CONCLUSIONS AND RECOMMENDATIONS 1.

With existing techniques and knowledge, the total risks

'\\

to the public from nuclear power plants, although very small, cannot now he meaningfully expressed in numerical terms.

2.

Quantification techniques do show promise as a tool in comparative safety evaluation.

3.

Ef forts should be made to improve the collection of data needed to evaluate the reliability and causes for failure of safety-related systems in nuclear plants.

A cooperative ef fort by the AEC and the nuclear industry, particularly the utilities, probably will be reqpired to achieve such a program.

1917 025 DISCUSSION Two fundamental questions are involved in making safety evaluations for a nuclear plant:

1.

H'ow much safety do we need?

2.

How much safety do we have?

The answer to the first question is an evaluation of the degree

. - ~.

of risk that should be accepted.

It is inescapably a policy _ decision, even though the intriguing possibility is being raised that it ru0r be expressed quantitatively in comparison to risks that have been and are being accepted by the public.

The answer to the second question is an evaluation of the risk that ist being taken (presu= ably measured against an answer to the first question).

Since it deals with hardware rather than with judg-ments, it has the possibility in principle of being made on a more quantitative, objective basis.

Clearly, if this possibility could be realized, many of the uncertainties in the safety evaluation process would be reduced.

Historically, acceptable levels of risks for activities in society have not been established a priori, but have emerged in the form of acceptable practice (rather than as quantitative evaluations) over a period of time long enough to observe the interplay of costs, benefits and risks.

Primarily because of the risks involved in the low-probability, large-consequence accidents, the nuclear power industry and the regulatory bodies have followed a different course. They have

(

diligently attempted to consider hazards and to apply preventive measures prior to undergoing actual accident cyperience.

In so doing, they have so f ar eliminated, and hopefully vill continue to prevent, the empirical accident experience that has in the past provided the basis for evaluation of actual risk and for determination of the socially acceptable level of risk.

There are two important consequences of this circu= stance:

1.

Estimation of the actual total risk of nuclear power plant operation will be based only on extrapolation of experience data with small-to-moderate accidents and near-accidents. Neither the data nor the concepttal fra=ework now exists or may be attainable for such extrapolations.

2.

The acceptable level of risk -

"How much safety do we need?" -- will not be derivable from experience and will be a policy decision.

The answer to the second question - "How much safety do we have?" -- also involves evaluations not empirically determined.

But, in principle, an evaluation can be made.

Such evaluations are subject to two basic limitations.

First, no certainty exists that all f ailure modes are recognized, and thus the evaluation may not be conservative.

~

Second, and more basic to the usefulness of the approach, is the lack of probability information of the required detail and accuracy.

The availability of information does not appear adequate to permit a meaningful evaluation of total risk.

1917 027 C

Further development of techniques for comparative evaluation of risk appears to have promise of beneficial results and should be encouraged. Moreover, even with relatively sparse and imprecise failure data, the existing methodology can be of value in influencing basic desiga approaches or in comparing performance of subsystems.

If properly developed and applied, these techniques might be used to:

1.

Compare alternate safety systems and components of engineered safety systems.

2.

Measure the relative protection provided against several postulated accidents to help decide which should receive the most attention.

3.

Decide if the problems caused by the additional complexity from addir.g a safety system outweigh the advantages of that system.

4.

Measure on a uniform basis the relative gain in safety provided by an additional safety feature.

The nuclear industry and the regulatory groups have not used these quantification techniques in as systematic a way or to the degree they are used in some other industries. The Study Group believes that greater use should be made of these techniques and the development and application of these techniques for comparative analyses by the nucicar industry and by the regulatory groups should be encouraged.

1917 028 F.

Application of these techniques is dependent upon the availability of adequate data on failure rates and modes. Presently, only gross failures are reported. While the. lessons thus learned are invaluable, particularly in identifying previously unsuspected failure modes, much more data is needed.

Protective systens and eneineered safety features are teated rr gularly; and compilations of test data - successes as well as failures - are a largely entapped source of failure-rate data.

It is the utilitiec that are vital to such a program, since they operate the plants and collect the data.

Macover, it is the utilities who eventually benefit, in enhanced plant availability as well as safety, when lessons learned f rom f

failures and reliability studies are fed back into new designs and t

criteria.

Dile the proper use of quantification techniques may be helpful, a note of caution should be sounded concerning their misuse.

The Study Croup believes these techniques snould be used as a tool in achieving a sound engineering design. To rely on the use of these techniques as a substitute for or at the expense of a disciplined engineering approach to design with an associated strong quality assurance program would be a misuse of the techniques and could result in a decrease in overall safety.

~

Successful attempts to quantify safety appear to be limited presently to comparative studies.

It is not clear, however, what success might be expected from additional development of the technique and the possibility exists that total risk evaluation may be feasible 7

1917 029 in the future.

In any event, there have been only limited attempts to date at systematic examination of the data requirements for meaningful evaluation, or of the feasibility of alternate approaches.

The S'tudy Group believes such a systematic examination should

/

f be made.

(

In summaty, the Group believes that development should be continued on techniques for the quantitative assessment of risk.

However, before such techniques will become a practical tool in evaluating overall reactor safety, cuch work remains to be done in the following areas:

1.

Extrapolation fro = experience with small accidents to quantitative judgments regarding potential serious accidents.

2.

Identification of potential failure modes.

3.

Development of information on failure rates of equipment and probabilitiec of postulated accidents.

4.

Establishment of an acceptable level of risk.

1917 030 F.

Degree of Standardization and Imposition of Additional Safety Requirements As proposed reactors have become more nearly alike, regulatory requiremeata have tended to become more stable. Nevertheless. In a number of cases, safety questions not previously identified have arisen and their resolution has caused delay and increased cost.

A proposed different approach to reactor licensing would be certification of a reactor design and plant safety features outside the context of individual license applicction reviews. Duplicate plants could then be licensed without extensive review, except for site-related factors.

Changes in the certified designs would be

(

considered in a manner similar to the original certification.

Consideration of the value of a certification system requires, among other things, an evaluation of the extent to which current designs of large water-cooled nuclear power plants have become standardized and the degree to which the benefits of standardization can be realized within the present framework of the licensing process.

RECOMMENDATIONS 1.

Greater advantage of the current degree of standardization in reactor and plant design should be taken by applicants and the regulatory groups within the present framework of the licensing process in order to realize core of the benefits of this standardization.

1917 031 2.

A formal system for certification of details of reactor and plant design features outside the context of individual license application reviews should not be adopted by the Commission at this time.

DISCUSSION There appears to be a considerable degree of standardization by each major supplier in the conceptual designs and the nuclear

~

steam supply systems for current large water-cooled nuclear power plants. There is less standardization, and the technology seems less well developed, in the design of systems, such as emergency core cooling systems, that interface with the nuclear steam supply system.

There is even less standardization in the preliminary design of other engineered safety features for these large nuclear power

~

plants, such as containment, fission product removal systems, and emergency power systems.

Factors that hinder their standardization are (1) these features are site-related and thus subject to more variation; (2) they are usually designed by architect-engineer firms, which are more numerous than the companies designing nucicar steam supply systems; and (3) the utility influence on these plant design features is more pronounced.

!917 032 There has been considerable discussion within the nuclear industry regarding methods for taking advantage of the trend toward standardization in the design of water-cooled power reactors. The Study Group believes that some gains in this direction are evident now and that more of the benefits of the current degree of design standardization can be realized within the f ramework of the present licensing process.

For example, the present regulatory practice of conducting a single safety review for the identical design.

f eatures of twin nuclear power plants at one site could be extended to cover identical design features of all plants of the same class, taking into account only the different interaction problems and site considerations, and any new safety-related information.

The cooperation of applicants proposing duplicate designs is needed, since obtaining the benefits of such a procedure would require clear identification of features and their design bases which arn the same as those accepted for a previously reviewed plant. The Group believes there is a potential in greater use of such an approach for consi-derable savings in regulatory review time.

The Commission announced its willingness in December,1964, to

/'

f p

conduct informal _ reviews and evaluations of power reactor systems or major components in advance of the formal filing of an application.

This procedure has been used successfully on a number of occasions, with varying degrees of examination.

Reviews have ranged from initial, informal reactions of the regulatory staff to detailed safety evalua-tions by both the staff and ACRS.

Such reviews, conducted under present procedures, can provide some of the benefits hoped to be achieved by a formal certification process, such as the one discussed later in this section.

I917 033 The Study Group endorses the proposed' changes in licensing regulations concerning provisional construction permits developed by the director of Regulation and believes these changes will permit M

H

greater advantage to be taken of the current degree of standardization in water power reactors.

IY these changes are adopted, the extent to which specific reactor and plant features are approved by the regulatory 3roups at the construction permit stage will be more clearly defined.

/

aad modifications to the approved design of these features will not be imposed by these groups at the operating license stage unicas substantial additional protection, which is required for the public health and afety, would be provided.

Another method which has been proposed to take greater advantage of the current degree of standardization is for the AI ' to adopt procedures for approving (certifying) tne design of a specific type or portion of a nuclear power plant outside the context of an individual license application. This proposal would provide for establishment of joint AEC-reactor designer groups, apart from any license application proceeding, to identify all the accepted safety design features of a standard reactor and plant design.

The proposal provides that each team would then define and set forth in a document all the characteristics and bases of these accepted features which are important to safety.

The resulting documents would contain enough details con-cerning these features so that if the document were included as part or all of subsequent construction permit applications, the features described could be approved without further review by the regulatory groups.

This document could also be used at the operating license stage to justify acceptance of the certified design features upon showing that they had been built in accordance with its provisions.

1917 034

f The most important potential advantage of this proposal is the rapidity which it appears to provids for converting specific decisions made by regulatory groups into a basis that can be applied generally to subsequent licensing reviews of individual applications. The suggested procedure would achieve its purpose primarily by systematizing and organizing the steps in the licensing review process in such a manner that general applicability could be derived from regulatory decisiens.

There are several difficulties with this proposal. To begin with, one of the principal premises of the proposal is that prior acceptance of design features of facilities licensed for construction can be translated into general approval of sa'ety features. This vauld be f'

~

difficult at the present time, because of the limited number of detailed final designs that have been reviewed, and because of the net 2 to complete research and development programs substantiating design adequacy and, the need for confirmatory operating experience.

There are other disadvantages to the certification approach. The certification would represent an agreement on safety-related design aspects between the AEC and the reactor designer, rather than the reactor operator-owner.

Thus, this approach would depart from the underlying, philosophy of present licensing procedures which places responsibility for safety on the reactor operator.

Some of the utilities who met with the Group indicated a reluctance to accept a certification type arrangement unless they could become active participants in the discussions regarding acceptability of specific reactor design features. 1917 035

By enlarging these discussions to include any or all utilities,

complications would be introduced in establishing priorities, in obtaining agreement of all parties and in reaching timely decisions concerning.the acceptability of reactor design features.

It appears to the Study Group that formal adoption of cer-tifics _on procedures by the AEC might be more useful and practicadL in the future than at present.

It remains to be seen whether there will be substantial standardization of safety features which will encompass all areas of design, including containment and other site-related features.

Further, the results of operating experience with the larger power reactors and of research ar_ development programs, such as those concerning the effectiveness of presently designed emergency core cooling systems, would be available for consideration at that time.

In the meantime, the Group believes that further advantage of any standardization of the design of water power reactors can.be realized by applicants and the regulatory groups within the framework of the present licensing process.

+

(

G.

Criteria for Deciding When to Backfit After Issuance of a Construction Permit The imposition of additional safety requirements af ter issuance of'a construction permit (backfitting) has been dealt with on a case by case basis. While this approach permits maximum flexibility, it also creates considerable uncertainty for licensees.

Criteria would be helpful in reducing this uncertainty.

RECOMMENDATION The Study Group endorses proposed changes in the Comission's regulations developed by the Director of Regulation which will provide that additional safety requirements for a nuclear power facility for which a construction permit has been issued will be imposed by the I

Commission only if it finds that such action will provide substantial additional protection unich is required for the public health and safety.

DISCUSSION Most industry representatives with whom the Study Group talk.ed criticized the present practice concerning backfitting since it has led to uncertainty af ter issuance of a construction permit ns to what safety features a licensee may have to add or modify in crJer to receive an operating license without delay.

The Study Group believes that proposed changes in the Commission's regulations concerning backfittin', Cil alleviate some of this un-certainty. The proposed amandment will make it clear that additional

)h

- 37 _

safety requirements will be imposed by the Comission af ter issuance of,.:ena t. ction pensit only if it finds that such action will provide provide sut sts.ncial additional protection required for the public aealth and saicy. The amendment will not affect licensee responsibility for evaluating significant new information developed as a resulr. of experience with design, construction, testing and operation of a reactor or as a result of safety research and development programs and for recoc:nending any changes needed to protect the health and safety of the public. The AEC may still require information from licensees sufficient co provide an adequate basis for making judgments in particular cases, however, licensees should not consider such requests as a prejudgment of the issues.

One of the potential problems that might be encountered in imple-menting this criterion for backfitting would be a disagreement between the regulatory staf f and the licensee as to the safety requirements agreed upon at the construction permit stage. The proposed amendment to the Commission's regulations will minimize this problem by providing for development and use during reactor construction of a system similar i

V to the technical specification system presently being used during i

reactor operation. This new system will require delineation of the essential ele =ents of the design and specify that these cannot be changed af ter issuance of the permit without prior Commission approval.

Other design aspects can be changed at the licensee's discretion, subject to later review by the Commission at the operating license stage.

The Study Group believes that development and use of a system such as that outlined above would contribute to the stability of the licensing process. 1917 038

f H.

The Role of the ACRS in the Regulatory Process The Internal Study Group examined the role of the ACRS in the regulatory process with regard to the review of individual applications and the resolution of broader safety issues.

CONCLUSIONS AND RECOMMENDATIONS 1.

The ACRS constitutes a valuable resource of the AEC.

For optimum use of the Committee, its role in the regulatory process should be modified.

2.

The ACRS should be relieved of the obligation to review and report on all applications for power reactor construction permits and operating licenses.

The Committee should then gradually reduce its involvement in the reviews of individual app 74--**~r.:

and concentrate more on:

(a)

Safety issues involving a class of reactors, new concepts of reactor design and new approaches to accident preveation or con-sequence limiting safety features.

(b)

Evaluation of new data resulting from safety research and development programs and infor-mation gained during the construction an1 operation of power reactors.

l

=

(c) Development of regulatory criteria and standards relating to safety and the technical bases used in the regulatory review of individual applica-tions.

DISCUSSION The Atomic Energy Act presently requires the ACRS to advise the Commission as to the safety of each power reactor prior to the issuance of a construction permit and again prior to issuance of an operating license. This case by case review by the part-time advisory committee has produced valuable results.

In the Group's view, however, the relative utility of the present type of ACRS review must be viewed in the context of changing circumstances. As the number of similar plants and the relevant safety-related areas repeatedly reviewed have increased, the premises underlying regulatory approval of individual applications have been used increasingly to form the bases for regulatory criteria and standards. Although much remains to be done with respect to criteria and standards, significant progress has been made. With this background, and with the present depth and breadth of technical com-petence within the regulatory staf f, the Study Group believes the staff is in a position to and should perform more of the case reviews alone, without an ACRS review.

The consideration of the large volume of details inherent in an in-depth review of a particular application in best accomplished by a full-time, competent, professional regulatory staff.

1917 040 o

(

In difficult cases, in cases where novel design approaches are proposed, and in cases for which regulatory criteria are not available, specift: case reviews should continue to be made by the aCRS.

But apart from those cases, the Study Croup believes that it is in suxds areas as development of criteria and consideration of special safety issues, that a part-time expert committee can be most effectively and efficiently used.

It is the Group's view that the ACRS might better concentrate its efforts on safety issues involving a new class of reactors, new concepts of reaccor design and new approaches to accident prevention or consequence limiting safety features. The Committee should also be in a position to evaluate new data resulting from safety research

(

and development programs and information generated during the con-struction and operation of power reactors.

And, of progressively increasing importance, the Co=mittee should have sufficient time to make the maximum contribution to the development of regulatory criteria and standards relating to safety and the description of the technical bases used in the regulatory review of individual applications.

A change of this magnitude should not be undertaken abruptly.

The Study Group believes that if the requi_.

enabling legislation is passed the ACRS should gradually reduce its involvement in, individual applications and correspondingly concentrate its' involvement on issues which af fect overall safety and the criteria and bases on which the regulatory staff's safety reviews are made.

1911 041

_n

+

I.

Timing and Staging in the Review and Decision-Making Process A closer correlation is desirable between the timing,of industrial decisions underlying the planning and execution of nuclear power plant projects and the timing of related decisions in the regulatory review process. This is particularly true with respect to the timing of decisions on siting and proposed plant design.

RECOMMENDATION The Commission should explore the possibilities for revising the present regulatory review process with a view toward achieving one or more of the following objectives:

A.

An earlier regulatory determination than at present on the matter of site suitability.

B.

A phasing of regulatory design and construction approvals to correspond as closely as possible to

~

the normal industrial plant design and construction phases.

C.

An earlier construction permit decision than at present for reactors of established technology and generally standardized design, based on less documented design information in the application specific to the particular facility than is now required. 1917 012

~

~

(

uIscussIox Tae incustrial process of planning and achieving the production of power f rom a nuclear generating station involves a number of decision points.

These include the decision to build a nuclear plant; the choice of a site; the selection of plant size, type and supplierst the determination of the type of contracts and their execution; and the scheduling of construction so as to meet expected power needs.

The system for regulatory review of and decisions on power reactor license applications should provide the following:

1.

A sound technical review of the reactor site, design, construction and operation proposed by the applicant.

2.

safety criteria, standards and codes, or other bases, t

upon which well-founded and ticely regulatory decisions can be made.

3.

A procedure which defines the scope and timing of regulatory reviews - relatively inficxibly and predictably for reactors of established technology.

4 The means for public scrutiny of regulatory reviews and the timely opportunity foi those whose interests night be affected to have their views considered.

A lack of reasonable correlation between the timing of the decision points in the industrial process and the decision points in the system for regulatory review can have undesirable effects.

1917 043

.~

It may result in discontinuity in construction planning and extra costs for facilities; and, on a broader scale, it can be a possible hindrance to achieving the goal of economic nuclect power. Accordingly, an important area of consideration by the Study Group, to which much thought was given, was whether the decision points in the regulatory process could be better matched than at present with those in the industrial process.

Because the licensing process entails a review function, the '

timing of its decision points cannot correspond completely with the timing of industrial decision points.

Industry representatives were of the view, however, that an improved correlation between the respective decision points can and should be made. In this regard, suggestions were received that the construction permit decision be made earlier and that there be greater predictability as to what will be approved. This would be done primarily by stabilizing the safety requirements underlying issuance of a construction permit (a matter discussed in other sections of this report), by limiting the scope of review at the construction permit stage or by a combination of the two.

On a separate but related matter, the desirability of earlier oppor-tunity for public participation in the regulatory review process -

particularly with respect to the site - was also considered.

The Group believes that possible changes in tha directions outlined above merit further serious consideration.

In that connection, achievement of one or more of the following objectives appears desirable to the Group:

(a) an earlier regulatory determination f917 044

i

\\

than at present on the matter of site suitability; Q)) a phasing of regulatory design and construction approvals to correspond as closely as possible to the normal industrial plant design and construction phases; and (c) an earlier construction permit decision than at present for reactors of established technology and generally standardized design, based on less documented design information in the application specific to the particular facility than is now required.

The Group is not recommending any one particular course for achieving these objectives since any such restructuring of the regulatory review process should be preceded by a detailed explora-tion of relevant administrative, legal and other considerations.

The Com=ission, however, may find it useful to consider three variations t

to the present review process which were discussed by the Study Group as means for furthering the stated objectives. These variations are described in outline below.

A.

From the. standpoint of the public and the utility, a regulatory review of site suitability would be preferable before any large commitment has been made by the utility and before there have been any irrevocable changes to the landscape. Since the suitability of a reactor site cannot be judged completely independently of the nuclear plants proposed to be located thereon, it would appear desirable that the Commission consider whether adequate criteria can be developed for siting reactors of established technology.

1917 02$

With such criteria, a mandatory public hearing on site suitability for reactors of that type could be held very early. Site preparation and initial foundation work would commence after this early nearing, to the extent such preconstruction permit work is now allowed under Part 50.

A more detailed review of the proposed nuclear plant's design features would take place at a later time leading to the granting of a construction permit.

Notice of proposed issuance of the construction permit would be given and a hearing would be held before an Atomic Safety and Licensing Board if any party, including a member of the public whose interest might be affected, requested such a hearing.

Alternatively, notice of proposel issuar_ce and an opportunity for a hearing might be dispensed with in view of the prior site hearing. There also would be a notice of proposed issuance and an offer of a public hearing prior to the granting of an operating license.

The principal resulting improvement in the regulatory process would be the earlier consideration of site suitability. A public review at that time could be valuable to the utility in providing early identification of potential site-related problems.

It could also benefit those whose interests might be

(

1917 046

.s I

affected, such as local residents or representatives of interested States, by affording them an earlier opportunity to have their views considered on the safety questions involved.

However, it is not clear to the Group, based on its limited study, what the problems might be in developing criteria suitable for this approach.

B.

A possible variation to the changes in the regulatory process outlined in A., above, also appears to warrant further consideration for reactors of established technology. Under this variation, steps similar to those described.in the preceding paragraph would be followed until the completion of the mandatory public hearing on site suitability.

This hearing would be followed by a regulatory review of the design of the reactor plant; however, approval of construction would be in several stages rather than one stage. The several regulatory review stages would correspond as closely as possible to the actual industrial plant design and construction stages. There would not be an of fer of a.

public hearing on these separate regulatory construction approvals, but there would be a notice of proposed issuance and an offer of a public hearing prior to the granting of an operating license. To be feasible, this approach would }9l[ 04[

require development and approval of acceptable design interface conditions at the time of approval of any portion of the reactor plant design in order to assure compatibility of the several parts from a safety standpoint.

In addition to the potential advantages of earlier consideration of site suitability, discussed previously, this review approach would bring the regulatory decision points on construction authorization closer to the corresponding industria' decision points.

At the same time, however, the increased number of decision points would place special emphasis on the need

(

for making the regulatory decisions more predictable.

This could be accomplished if adequate regulatory criteria can be developed concerning the design interface conditions previously referred to.

C.

The correspondence between the regulatory and industrial decision points could be improved if the construction permit decision on acceptability of both the site and plant design could be made earlier in the review process than at present for reactors of established technology.

This might be done by requiring less documented design information in the application specific to the particular

~

f acility than is the case under present procedures. A reduction in such documented information would, however.

1917 048

~

i l

need to be compensated for by a corresponding increase in desi;n standardization.

Under this approach, a construction permit review for a reactor of established technology might be limited, for example, to the proposed site, general reactor characteristics, and engineered safety features. Handatory hearings at the construction permit stage might or might not be retained; but there would, in any event, be an opportunity for hearing at. both the construction permit and operating license stages.

An earlier licensing decision on site and design matters could reduce the difficulties inherent in 1

imposing regulatory requirements at a point in time af ter industry decisions and commitments have been made. However, to be fully useful, such a construction permit approval would have to be grounded on regulatory criteria, or other bases, suf ficiently definitive to give reasonable assurance of issuance of an operating license upon satisfactory com--

pletion of facility construction.

The feasibility of this approach would seem to depend on the possibilities for stabilization of the scope of the construction permit review.

For reasons discussed elsewhere in this report, these uafety reviews presently require a

~

substantial amount of information specific to the particular proposed reactor, flowever, the present trend toward 1917 049

_o_

r standardization of design in water-cooled reactors, the movement in the direction of establishing more comprehensive

. safety criteria, codes and standards, and the research information and operating experience which will. become available for the larger water-cooled reactors, may provide the basis for increased stabilization.

It is in this context that the possibility for an earlier construction permit decision might be considered.

An ndditional observation is in order concerning each of the above approaches, or others which might be considered for the re-structuring of decisional points in the review process. The public hearing phase of the safety review process has a bearing on the etming of regulatory decisions. Among those who spoke to the Group, opinion wac divided as to the need for or desirability of a mandatory hearing at the construction permit stage - although all agreed that there should, as a minimum, be an opportunity for a hearing. A hearing on an uncon-tested construction permit application does involve some delay in issuance of a construction permit (approximately six to eight weeks, under current procedures), and there is a question as to whether the safety benefit derived from the limited board review warrants. he time delay.

However, from the standpoint of public participation and understanding, the hearing does appear useful and the delay could be mitigated if tne hearing were held earlier in the development of the plant, as previously discussed.

(

1917 050 Dae further matter warrants comment in connection with the subject of hearings.

It was suggested t. the Study Croup that the role of the hearing board in an uncontested case might call for enlargemerit if there were a change in the present statutory require-ment for ACRS review of each construction permit application. The Study Group does not agree with this view. The lack of need for an ACRS review - premised, presumably, on the absence of substantial or novel safety questions and confidence in the competence of the regulatory staff - should not be grounds for expanding the review function of the board.

In this connection, it was recognized by

~

all who commented to the Study Group that hearing boards, by virtue of their ad hoc composition, their discontinuous service and the constraints imposed by the limited periods for which they sit, are not in a position to carry out a comprehensive technical review of V

individual reactor applications.

These circumstances further support the Group's view that the role of the boards should not be enlarged.

,51 -

1917 051