ML22230A157

Tran-M780517: Briefing on Anticipated Transients Without Scram (ATWS) (Open to Public Attendance)

ML22230A157
Person / Time
Issue date:	05/17/1978
From:	NRC/OCM
To:
References
Tran-M780517
Download: ML22230A157 (1)

Text

RETURN TO S

• """ "'"c,,o EGRETARIAT RECORDS f "1 Transcript of Proceedings

\ .,.1-,, ~o l NUCLEAR REGULATORY COMMISSION BRIEFING ON ANTICIPATED TRANSIENTS WITHOUT SCRAM (ATWS)

(open to Public Attendance)

May 17, 1978 Pages 1 - 38 Prepared by :

C. H. Brown Office of the Secretary

DISCLAii*\i:R This is an unofficial transcript of a meeting of th2 United States Nuclear Regulatory Commission held on May 17, 1978 in the Commission 1 s offices at 1717 H Street, N. W., Washington, D. C. The meeting was open to public attendance and observation. This transcript has not been revie1*1ed, corrected, or edited, and it may contain inaccuracies The transcript is intended solely for general informationa1 purposes~

As provided by 10 CFR 9.103, it is not part of the forma1 or informal record of decision of the matters discussed. Expressions of opinion in this transcript do not necessarily reflect final determinations or beliefs. No pleading or other paper may be filed with the Commission in any proceeding as the result of or addressed to any statement or arg'..1ment contained herein, except as the Com~ission may authorize.

1 1

UNITED STATES Of AMERICA 2

NUCLEAR REGULATORY CbMMIS.SION 3.

4

.BRIEFING ON ANTICIPATED TRANSIENTS 5

WITHOUT SCRAM(ATWS) 6*

7 (Open to Public Attendance 8

9 10 Corrimissioriers' Conference Room Room 1130 11 1717 H Sbreet, N.W.

Washington, D. C.

12 Wednesday, May 17, 1978 13 14 The Commission met pursuant to recess, at 11:00 a.m.,

15 Joseph Hendrie, Chairman, presiding.

16 17 PRESENT:

18 Chairman Hendrie Commissioner Gilinsky 19 Commissioner Kennedy 20 ALSO PRESENT:

21 J. Hoyle J. Kelley 22 L. Gossick A. Kenneke 23 s. Hanauer R. Mattson

.24 E . Case

s. Levine 25

(

.- 2 1 \

P R O C E E D I N G S 2.

CHAIRMAN-'HENDRIE: This is one of those ,occasions 3

of joy which come-s really too rarely to the Commission, 4'

a briefing on Anticipated Transients Without SCRAM. In this 5

case, Mattson*, .-Hanauer,: I don't know whether to inciude you in 6

the*.guilty partyi***L'ee, or not. Why don't.you *sit bae;k a 7

little bit 8

MR. GOSSICK: I intend to stay out of the iine of 9

fire.

10 CHAIRMAN HENDRIE: and I'll *ask these other 11 fellows what they are up to and who is going to make the, 12 presentation.

13 MR. GOSSICK: Dr. Mattson will lead: *of,f followed 14 by Steve.

15 MR. MATTSON: We a~e here to give you a status 16 report today on ATWS and ~ot to ask you for a decision.

17 We have sent some paper up to you,~including a 18 NUREG report,. the number is 0460 and I will probably refer to 19 it that way this morning, and a copy of some remarks that 20 I gave to the public meeting on the 19th of April and to 21 the ACRS Subcommittee on ATWS on the 20th of April.introducing 22 the report and laying out a plan of action for bringing ATWS 23 to a conclusion.

24 I'd like to summarize that briefly today, give 25 you a flavor of where we are headed and where that will

3

\

\

\

l eventqally inv6lve you in the decision process, and give 2

you a sumrn:ary overview. of some of the: techn.:j.cal things

.3 that are said in: the r,eport .

4 First, I ,ought to highlight that the technical 5 report, DSS,.Division of .Syste!Ils Safety, staff report.

6 *It is not yet a position. for .the Office of* Nuclear .Reactor 7 * 'I*. I' ,, !:** *. *.

Regulation.

• That report is a 'reevaluation 'of the. staff*'

8 position on ATWS. It traces through* the history *of WQrk on 9 this technical subjeet;:, commencing in the late 1960s* when 10 the issue w~s identified.as a safety concern by the ACRS.

'

• I 11 In tracing that.history i t marches by some rather important 12 ~ilestones ~nd I will. list a couple:

13 One was the production in 1973, by.the Regulatory.

14 S'taff of the AEC of a document* known affictipnately as 15 WASH-1270, which one can think of as the five+/-year-ago 16 .predecessor of this NUREG report that you have. in front of you 17 today.

18 That>Jwas *followed in 1974 and 1975 by a series 19 of reactor vendor rep9rts proposing analytical methods for 20 iTWS and a seri~s bf *staff. status reports on those analytical methods.

21 22 Also, about the same time frame, the reactor safety 23 study was published, *in 1975, *wASH-1400. The.in 1976 there 24 was a major contribution to the body of knowledge on ATWS by 25 the Electric Power Research Institute, EPRI, surveying and

4 drawing conclusions from the* body of data on.the frequency 2

of transients and

,' the. probability .'of failure* of SCRAM .

3.

This reevaluation of the staff position on ATWS

,4 was directed by Ben Rusche, that is, ordered to commence by 5

Ben Rusche the f6rme~ Direct6r of NRR about a year ago~ The 6

.tebhnical work *which formed the *original. basis for the report 7

was led by

• S.teve

• Hanauer drawing upon experts from -- throughout 8

NRR and Research and a couple of consuitants. Principatly, 9

though, the experts were from the Division of Systems Safety.

10 Last*fall, .he turned i t back to NRR, i t was assigned 11 to the Division of Systems Safety where we have managed the 12 technical development, rev,iew and report drafting since .that

• 13 .time.

14 In our work over the last six months, *there'.have

• been

,15 major contributions from other sectors from the agency, I 16 have mentioned three. The Office of Research and the 17.

Probabilistic Assessment Branch has pulled a major ore in 18 helping us in our application of probabilistic assessment 19 .techniques. *The Division of -- .the Environmental Division 20 of.NRR, Harold Denton's division, is in the main responsible 21 for the value impact assessment which. we* think is one .of the 22 more thorough going and complete of any yet done by the staff.

23 And the Division of Operating Reactors has supplied us with 24 technical insight and considerations on the backfitting issue 25 and will continue to do so in some major way in the course of

5 1 the next several months.

2 The new report is intend~d to form the. technical 3 basis for a final policy decision on ATWS. We don't consider 4 it yet to be the final technical word, for example,* changes 5 are occurring t?day and will occur over the next several 6 weeks in response to the process~s which is in motion to 7, review the report, namely, the ACRS review and the Regulatory 8 Requirements Review Committee review.

9 We would expect by the time the report reaches you 10 for a decision, with a proposal for implementation, it will 11 be sufficient for reaching a final decision on ATWS. We 12 believe that the technical issue has drug on about long 13 enough. I think the Chairman's informai remarks carry that 14 same message. We believe *that by and large the dialogu~

15 between government and industry is complete on this issu*e.

16 There is continuing discussion going on as we try to explain 17 and make clear the words in our report so that people may 18 more effectively comment on it.

19 The principal additions of this report to the state 20 of knowledge on ATWS are several. They include the use 21 bf probabilistic .techniques, the contexting of the probability 22 of:*severe consequences arising for ATWS, relative to the 23 probability of severe consequences from other causes which 24 are provided by the reactor safety study.

25 It is yet a another review of the available data on

6 1

these very I.wow probahili ty events and a* final contribution

• 2 6:15 va.lue :.impact 'assessment.

3

'

• In the media* we have been reading that it i's the 4

industry.' s yiew that ;not much has changed in the years that 5

have co~~* since the iast technical*report, with the qualifi~

6, cation'.:that these things* have 'been added, we* would agree 7

that the technicai' position in the *main has.remained.relatively

. *a stable. That indicates.to us that i t is time for a d~cision.

9 I woui~ like to burn to some slides. Tom, are 10.

~pu, badk there and' can give me Slide-No. l?

11 (Sl'ide) 12 Briefly the proces~ ~hat we have s~t in motion 13 for reaching this final decision on ATWS *is described on*

14 that ~hart. A public briefing has been held, the ACRS 15 Subcommittee meeting has been held, the RRRC':s 'first meeting 16 on the subject was yesterday, we hav~ another ACRS Subcommittee 17 meeting on the 26th, we would expect several more Subcommittee 18 meetings.before we reach the full ACRS for comments.in June 19 or July. We are here with the Commission breifing today, we 20 would expect to get back to you-about August with .a Commission 21 paper recommending a course of action.

22 One other miles tone we have to accomplish :bef cire *2 then 23 is RRRC recommendations to the Director of NRR which he will 24 take into consideration in formulating the Commission policy 25 paper with his recommendation for action.

7 In looking ahead to .the likely content.of that 2 . recommendation,. i t is my opinion at the moment and*. ~t remains

,*' r , ' * ' * *

• 3 unchanged.through ,several weeks of public. discussion of this 4 subject now, that rulemaking is the, most likely course of 5

• action td~finally resolve this i~sue. lhere ~r~ pros a~d cons, 6 *both sides t9 that. , I also believe .tha:t it is reasonable.* to 7 expect tI:iat* such, rulem~king co.uld be a~c.omplished in the course 8 of about two years, 'with i:qiplerrientation expected after that.

9 COMMISSIONER GILINSKY :, Two-year rulemaking?

10 MR *. MATTSON: We have a leg up, 'Commissioner. It is 11 . a li:ttle shor.ter, than one would expect from ---

CHAI.RMAN HENDRIE: Are you start led . at the shortness ,

13 .the abruptness or.the lorigness qf it?

14 MR. MATTSON: L.took it to be the fo'rriler, bu't 15 perhaps i t is the latter.

16 The'average is somewhat longer than two years, but 17 we have in the rep0rt, the draft of a proposed rule.

18 COMMISSIONER GILINSKY: The average for what?

rulemaking,? .

19 MR. MATTSON: For rulemaking, from beginning to 20 end.

21 COMMISSIONER GILINSKY: For NRC rulemaking?

22 MR .. MATTSON:

• I don't believe Mr. Minogue is in the 23 room, but I think his rule of thumb is in the order of three 24 years.

25

8' 1

CHAIRMAN HENDRIE: Does that include that petition 2 that we finally dismissed that came in in.'68 or '70 or 3

something like .that?

4 MR. MATTSON: That would* include a number of 5 things, including ACCS which I'm sure weights the data.

6 COMMISSIONER GILINSKY.: You can get a CP license 7 sooner than that.

8 (Laughter) 9 CHAIIDII..AN HENDRIE: They are trying to fix that.

10 MR. MATTSON: So I guess the principal thing that 11 I would like the Commission to draw from that chart is 12 that we will be back to you some time the end of the summer 13 for more indepth and less status-oriented discussions. That 14 is, we will be looking for a decision.

15 I'd like, having dwelt on the procedural aspects 16 .of where we have been and where we think we are going, to 17 turn quickly through an overview of what ATWS is and what 18 we are saying about it in this report.

19 If we could move on to slide 2, I don't spend a lot 20 of time here 21 (Slide) 22 =- but just so we are all talking about*the same 23 thing, ATWS is composed of two elements. First an anticipated 24 transient, events expected to occur one or more times during 25 the life of a plant. Some examples are loss of feedwater,

9 1 tur.bine trip, loss of off-site power and so forth.

2, The second.half of ATWS is the failure of*the 3, shutdown system, the SCRAM system. The failure ,to function.

4 The.mystical common mode failure appears in that element of 5 the ATWS formula.

6 In the next slide --

7 (Slide) 8, -- we show a.little bit more about t.he nature of 9 these elements. Why is there an ATWS pro?lem? Well, we 10 are concerned that the frequency of ATWS events, having 11 severe consequences, by severe consequences we mean things 12 ranging from public -- releases of radioactivity leading to

'~

13 doses in excess ,6f~part~100 guidelines ~anging up to events 14 of a more catastrorhic nature as described in the reactor 15 safety study. And the frequency of such events with severe 16 consequences shown here s.implistically as the frequency of 17 anticipated transients with potententiil for severe 18 consequences, that is, not all transients, if there were no 19 SCRAM would lead to the same consequences.

20 Times of probability fo SCRAM failure given 21 the .antpcipated transient with the potential for severe 22 consequences. Now, the debate over the years which has 23 raged on all three points.

24 First, what's an acceptable value for this 25 frequency of ATWS events with severe consequences? I ' l l spend

10 1 quite. a lot of time talking about that in . a few minutes.

2 For shorthand, we say that's ,our safety goal~ .. How safe do*

3

• we -want* thepiant~*to be for the severe ATWS events?

4 Th*e s*ec.ond element, the mj,ddle cclumn,what

• value 5 0f the freq~~ncy of antiblpated trin~ienis with potential

• 6 fo~ severe conse_que~ces should be used. in this formu],a .. ' That Is 7 an -*area, on-which' today, we feel there is 'much less controversy 8 than).in the past. The numbers used in the reactor safety 9 'study; the numbers used in the EPRI document, the numbers 10 proposed. by the i*ndustry through* th~ *years, and the number 11 proposed here are traceable to fairly good data at this point, 12 given the operating experience with the number 'of plants in 13 the* country. * 'Not a lot of controversy, over pumber.

14 What value the probability of failure of SCRAM 15 should be used, relatiyely at least, much more qontroversy 16 in that area. And then, in the bottom line, what are we 17 saying in this .report? A safety goal of 10 to the 6th.

18 For a f~equency of ATWS events with severe consequences 10 19 to the minus 6 *~*.: .

20 COMMISSIONER GILINSKY: Let me ask you about that.

0 21 Now, this is for the entire class of these events?

22 For each particular type of event? I mean, it doesn't seem 23 to me to be a very w.e.11 defined concept.

24 MR. MATTSON: The types of events for which we are concerned are defined in the report.

25

11 1 COMMISSIONER GILINSKY: And this is for_ the 2 entire class?

_3 MR. MATTSON: For the entire class there'are some 4 that are important. For those that are important the 5 pn:>babil.i, ty of yielding at ATWS with consequences .i,n excess 6 of part 100, we propose should be shown to he less than 7 10 to the minus 6th for each reactor.

8 Now, the next slide will turn to i t if I would just 9 -summarize the next two points on this one.* That is something 10 I do want to go in to.

11 The value of 6 for the frequency of anticipated 12 transients with potential for severe con5-equences comes from 13 operating data. And the probability of *scAAM failure comes 14 from a variety of statistical and relia,bil.1.t:y assessments 15 with several orders of'magnitude difference between the 16 staff position and the historic industry position, principally 17 owing to the weight given'by the staff to the potential for 18 common mode failures. But when you perform the multiplication 19 up in the top line and compare the two numbers, you that .we-~.

20 have, and the number is more precisely stated in the report, 21 at present an average probability of occurrance of an ATWS 22 wit.h severe consequences for reactors operating or under 23 design in the United States, of 2 times 10 to the minus 4th 24 and a safety goal of 10 to the minus 6th. And the propos~l is 25 to bring the design capability of the facilities into conformanc

12 with the safety goal.

2 COMMISSIONER GILINSKY: Now, how does that relate 3 to the numbers in t.he reactor. safety study?

4, MR. MATTSON: The reactor safety study looked at 5 two specific reactors. The Surry* Reactor. ~ndi,~~E:~~li-~~~§~_t_qm*--

6 Reacto~~ . The Surry Reactor;*~*~r~~s~rize~ wate~ ieaot~r.

7 The reactor saf~ty* study and the st~£f in NUREG. 0460*.

8

• essentially, show the Surry reactor, a Westinghouse pressuriz~d 9 water reac~or is capable of meeting the 10 to* t.he minus 6th 10'. goal t.oday. ',' '

11 I will get on a little bit later to saying wh~~

12* we think generically about tlle,various class(;;!S of facilities, 13 but subject to confirming analysis and a little more detailed 14 look at Surry, :~.it's 'fairly apparent to us at this_ poi~t that

'I ' '

15 the Surry reactor would meet the goal and. that -othei We~tinghouse 16 reactors would.be similarly close.

17 COMMISSIONER GILINSKY: What about the other.* one?

18 MR. MATTSON:"*p~~~h-Bo*Etom_;reactor

--*"* -- --- - - - **-- - --- -- we say in 0460 19 that there is new information available irt some.of these areas since th~ reactor safety study. The reactor safety 20.

21

• study showed ATWS to be an important contributor to the

22. overall risk of severe consequences portrayed by the. reactor safety study. we*say.it is an even more _important contributor.

23 COMMISSIONER GILINSKY: \_*-sw~p~}:hie:

So does this ....

24 --... -~- . **.- -- -** -~- -

13 1

MB,. ,MATTSON: Swamp is the wrong word; but we can 2

bring down the overall risk -- the o.verall pr.obabili ty., o_f

• 3 s*ev~r~ c~nsequenc~s .*. sig~ificantly by fixil)g. the 'ATWS 4

problem on boiling water reactors,, including Peach Bottom.

5

. COMMTSSIONER GILINSKY .: ,

• Well, let !Jle ask you , --

.6 is your estimate. here different .than *the one in the. reactor 7 safety study?

8 MR .. MATTSON: Yes, slightly higher.

9 COMMISSIONER GILINSKY: Slightly higher?

10 MR.'MATTSON: Yes.

11 MR. LEVINE: .. Let me say a word.

12 The problem is one of the fact that ther*e* are other ways of shutting down the reactor besides~:this* SCRAM'system 14 which were* considered i:n: the reactor safety study and which 15 are in fact considered in the ATWS report. These -numbers 16 that you see there represent an incomplete statem~nt in the 17 total situation. There is another fac~or of 10 6r 106 18 in probability reduction when you look at the. backup shutdown 19 systems, which Roger, I'm sure will be getting.to in his 20 discussion.

21 The numbers we had are about the same for this 22 portion -- ATWs:*was defined as the anticipated transient 23 and the failure of the normal SCRAM system. Beyond tha.t there 24 is an additional probability to get the reactor shut down 25 which other additional things have to bear.

14 I

COMMISSIONER GILINSKY: So why is there a problem?

2 MR. LEVINE: Well, there is a question now of how 3.

close do you come -to the 10 of the minus 6, when you consider 4

the other mechanisms.*

,5 COMMISSIONER GILINSKY: .So you are saying just a 6

little bit away or ---

7 MR. LEVINE: The number 10 to the minus 4 is a

• s factor, but 100 away from 10 of the minus 6 and the question 9

COMMISSIONER GILINSKY: Yes, but the question is 10 MR. LEVINE: -- how close do you_get there with 11 these backup shutdown system.

12 MR. MATTSON: The problem 13 MR. HANAUER: I guess we had better go into it.

14 The problem is more complicated. In the first 15 place, the Peach Bottom reactor has and had during the safety 16 study evaluation some of the equipment which we are planning 17 to require for all boiling water reactors.

18 The second point is that the Peach Bottom reactor 19 has the 20 CHAIRMAN HENDRIE: Another way of saying that, 21 Steve, is for the run of BWRs the ATWS situation is considerably .-

22 less cheerful than i t is at Peach Bottom.

23 MR. HANAUER: Well, there are two classes of BWRs.

24 Some have this additional equipment and some don't.

25 The Peach Bottom reactor was in the more favorable

15 1

group.

2 COMMISSIONER GILINSKY: So i t was not a fortunate 3 choice.

4 MR. MATTSON: In this problem .

. .5 MR .. HANAUER: Yes ..

.6

• The other point is *that whereas the reactor safety 7 study made prpperly for safety studies an evaluation qf the 8 chance that these additional systems would bring the plant to 9 a satisfactory situ~tion, the look we have taken is in some 10 respects more conservative ~han that in the safety study~ and 11 we come out with the same -- I'm making this up as I go along 12 we come out with the same functions ;--we:'.are. asking for more 13 performance and reliability in these funcitions than was 14 provided at P~ach Bott6m.

15 Therefore, we.would evaluate that Peach Bottaro 16 needs faster equipment, to us, confirms and partly not 17 confirms to us the safety status evaluation. I didn't say 18 that too well. I have to warn you it's a fairly complicated 19 subject which I will go into in whatever detail you ne~d.

20 COMMISSIONER GILINSKY: Let me ask you this: _suppose 21 we haa not taken Peach *Bottom, but one of this.other class*

22 of reactors. Would we then have come out with numbers which 23 are significantly larger or the probability of these events/

24 which are significantly larger that we obtained on the reactor 25 safety study?

16 1 MR.. HANAUER.: 'Yes, sir.

2 COMMISSIONER GILINSKY: By what, like a factor of 3 100 or something?

4 MR. HANAUER: More like a factor of 5.or 10.

5 MR. LEVINE: By the way,, this information was given 6 to the regulatory staff, even before. the study was done, 7 that is, the fact that" not all plants had used this equipment 8 and +/-hat made a significant difference in the results.

9 MR. HANAUER: We have known this for a long time.

10 MR. MATTSON.: Yes.

11 If we could go .to the next slide.

12 (Slide) 13 He we will talk a*little bit about how we get to 14 the safety'objective that we have proposed.

15 The interest, of course,'is what's an acceptable 16 objective for protecting the public from expbsure from 17 radiation. We hav~ proposed that 18 COMMISSIONER GILINSKY: Let me ask you again about 19 this 10 to the minue 6th. You may be getting to it, I see 20 it at the bottom of your slides, but the::thing. that concerns 21 me here is that it is not clear how many items you stuff 22 into this package that it is suppose to be when you add it up, 23 10 to the minus 6th, there is a certain amount of arbitrariness?

24 MR. MATTSON: Yes, that's what I was going to talk about.

25

17 l COMMISSIONER GILINSKY: You are going to talk about 2 'it?

3 MR. MATTSON: Yes.'

4 COMMISSIONER GILINSKY: Very good ..

MR. MATTSON: WASH-1270 set.a safety goal of 10 to 6 the minus 7th.and contained a rationale that if the:ire. were 7 l0*such causes of severe consequences -- potential causes 8 of severe consequences of this nature, then a goal of 10 to 9 the minus 7th for each of the 10 consequences would yield 10 overall, a safety level of 10 to the minus 6th probability 11 of severe consequences. 10 times 10 events at 10 to the minus 12 7th.

13 The reactor safety study came along after WASH-1270 14 and said that the core melt frequency, was somewhath,ighe:1; than 15 ha~ been generally understood be;Eo:1;e, but desc~ibed a 16 spectrum of consequencE!s which was much different than had 17 generally been understood before.

18 The core melt frequency of 5 times .10 to the minus 19 5th was spoken to in the reactor safety study without 20 catastrophic consequences of the sort that. have been held in 21 mind prior to that time.

22 In light of that information, in using the same 23 rationale that was in WASH 1270, the staff could not continue 24 to support 10 to the minus 7th for one of 10 contributors, 25 rather 10 to the minus 6th for one of 10 contributors for

18 1 *an overall probability of 5 times 10 to the minus 5th made 2 much more sense. Now,. the factor of 5 in there is an 3 i:nteres~ing number

• als,o. One might -argue that you could 4 make the ATWS objec~i~e _5 times 10 to the minus 6th, but 5 the 5 actually goes away when you take the ATWS cdntribution 6, out of .the overall cor_e melt frequency and when you make 7 some estimation for* improvements in safety for future 8 reactors rela.tive to the estimation from the reactor safety 9 study ..

10* So the.number for core-melt frequency excluding 11 ATWS is somewhere in the range one to 3 times 10 to.the minus 12 5th, we say in the report.

13 *COMMISSIONER GILINSKY: So you are saying you will 14 bring it down to the level of other* events?

15 MR. MATTSON: Yes.

16 COMMISSIONER GILINSKY: That's your objective?

17 MR. MATTSON: Yes. And ther.e are several other 18 events of the order of 10 to the minus 6th or slightly higher 19 which contribute in the main to that 1 to 3 times 10 to the 20 minus 5th over~ll probability of core melt. And this 21 brings ATWS into line with. those, ~ut does not go as far 22 as proposed in 1973 to go to 10 to the minus 7th.on the basis 23 of the insights gained in the o~erall ris~ by the reactor safety study.

24 25 CHAIRMAN HENDRIE: 10 to the minus 7th, we bandied

1 those numbers about and they didn't look too cheerfully 2

in those days. I !eally shouldn't have,picked that up, 3

stuck in*the 1270 .-- it was much too simplistic as*sort 4 of a ~is~ussion ~nd it had and continues to have the 5 , ~~traordinary dis~dvantage that it is an absolutely undemonstr-6 ~ble and you are not a heck of a lot better off, but not 7 . notably'* so. 10 .to the minus 6th is still something that you 8 are *not going to demonstrate for plant experience.

9 MR .. ,LEVINE: That factor 10 is a very important 10 difference. ,*.

11 CHAIRMAN HENDRIE: But you are getting a lot close 12

• to

• the demonstrability of it ..

13 COMMISSIONER GILINSKY: Let me ask you, Roger.

14 Suppose* iomeone comes ilong and recalculates the reactor 15 safety study and discovers that what you think is 10 to the 16 minus 5 is really 10 to the minus 4. You then suggest that 17 the ATWS be knocked down in the order of magnitude?

18 MR. MATTSON: The Commission would first have to 19 gi~e me the same kind of implicit guidance that I had from 20 the Commission today on the reactor safety study, that is, 21 that it is generally sound and generally* describes an 22 acceptable risk, because that is implicit in my use of 23 the reactor safety study to change 10 to the minus 7th to 10 to the minus 6th.

24 25 COMMISSIONER GILINSKY: Where does that come from,

20 1 has 'the Commission ever said that?

2 MR. MATTSON: Well, the Atomic Energy, Commission 3 said that in 1974 and the i~formation has never been or 4 the guidance have never been overturned.

5 COMMISSIONER GILINSKY: Well, T don't think there 6 is guidance from this Commission on that.

7 MR. CASE: Well, it is more implicit than explicit 8 guidance. The implicit guidance is that a well done study ---

9 . MR. HANAUER: We didn't stop reactor licensing in 10' view of that - -i~:f:qrmat,i.on ,i,n -pa,rticular:.

11 COMMISSIONER GILINSKY: I am getting a curious set 12 *o:t: arguments.

13 MR. MATTSON: Well~. i t is also the corner stone 14 that you must have, in order to do probabilistic licensing 15 assessments at all-~-

16 COMMISSIONER _GILINSKY: I remembe,r testimony along 17 the lines if they weren't safe we would have shut them down 18* wouldn't.we have?

MR. HANAUER: Yes.

19 20 COMMISSIONER GILINSKY: I don't think that's a good argument.

21 22 I mean, you cari consider certain levels of safety to be a6ceptible or tolerable for a period of time, but_you 23 can have quite a few goals. You know, we just got through 24 a session on safeguards in which we are tolerating a situation 25

21 l'

we would not in the long .'run consider respectihle.

2 *I'm not* saying there is ~nything. wr*ong. ne~essarily, or *that 3 that should riot' be the guida.nce.

MR.* HANAUER: There is no. notion *.that:... i t is the 5 last word either .. The~~ are-~ it is, first of.alli the 6 best information. available. And the sec~nd place, i t impelle.d*

7 neither the staff,rior the Commission to conclude that there 8 was *anything seriously wrong with reactor safety.

9 So in the spirit of using the best available 10 information, we have ha.sea. our .goal to some considerable 11 .extent, on the results.of' the study .. If the study came out 12 very different, if the next study or some recalculation came 13 .out very different, we would have to rethink this.

14 .MR. MATTSON: It is .important to ~ecognize that the 15 approach to developing licensing cr+/-teria and the requirements 16 that we are proposing to put on the design of reactors do 17 not flow in any direct quantitative fa~hion from the reactor 18 safety study number or from the 10 to the minus 6 safety goal*.

19 What we have done is examine the general safety 20 goals stated in probabilistic quantative terms in WASH-1270, 21 said it too low, it needn '.t be that low and** move.it up in 22 order of magnitude. But*then we are still stuck, as I will 23 describe later, with being unable to state in licensing 24 .criteria probabilistic methods, totally p~obab:i,listic methods 25 for deciding how a system should be configured, what components

22 1

i t sho~ld contain, what safety stress limits i t might 2 be designed for for pressure:*cqntaining components, what 3 operator reliability can be relied upon,** et cetera, et 4 cetera.

5 So we have fallen back to a det.erministic licensing q approach which has in i t an element of conservatism. When 7 you move from the probabilistic to* *the determin.istic, no matter 8 how much you say you are doing i t with realism as we say in 9 this report, there is an element of conservatism which 10

• provides a hedge against changes in that. 5 times 10 of the 11 5th number. The additional hedge we have looked at is in 12 our review.of the reactor safety study,.through the years of 13 its availability and in the context of ATWS,. we would conclude 14 that the number will likely _move down and now up; that 5 ti.mes 15 10 of the minus 5th is high.

16 COMMISSIONER GILINSKY: Well, I think that if indeed 17 you are using or guiding yourselves by some implicit 18 Commission approval, I am wondering about a document and 19 I think that is something the Commission ought to take up.

20 MR. MATTSON: I think that that's a very important 21 point you raise, and I personally would *recommend that perhaps 22 the time has come for that.

23 Any probabilistic assessment in reactor safety 24 hinges upon the implicit understanding of the general acceptabil ty 25 of the reactor safety study results. That's the only benchmark

23

'\,

I I 1

2 MR. HANAUER: This is one of*our first forais

.3

• hito a*n explicit *use of* numbers iri establishing safety goals 4 ' '

and ther.efore ,* i t is riot_..,,,.surprising that new considerations of 5 .this sort_ s~ould come up._.

6 COMMI$SIONER G~LINSKY: *. Well, I think the Commission 7 . ought 'to gbr~-, guidance on:Utt.

• _,8 MR. MA'i'Y.rSON: . , That is what we -wiB. be looking for 9 the Commission mo.st for, when we cornea back to you for a 10 decision, is how safe is safe enough for ATWS? Is it required 11 to b~ fixed~ot~i~n~t~it required to be fixed? If i t is, to 12 what deg1;ee? And that's the question_ -- .that',s the name of*

13 _

• this game. That' 1 s* what 'this is all about.

14 MR. :HANAUER:* ,That will be an issue during the*

15 rulemaking als6. You don't have tp decide that in any final 16 way in putting out a proposed rule for consideration.

17 MR. MATTSON: Let*me briefly summarize the kinds 18 of solutions to tl].e ATWS issue that have been proposed arid 19 discussed in the past. They are fairly straightforward.

20 You_ e_i ther decrease the frequency of anticipated 21 transients which is impracticable to demonstrate maintenance 22 of that throughout the' life_ of the 'plant . '

and probably over::-,.:::.*:::'.~:*

23 constrains the operation of the plant for other reasons to 24 the point that i t is just not thecptacbical way to proceed.

25 Decreasing the probability of SCRAM failure,

24 1

intellectually, intuitively and an acceptable way to proceed, 2

but.you have to demonstrate such a low probability.of SCRAM 3

failure in order to achieve the safety objective~- and.this 4

is what the Chairman referred to earlier -- i t becomes a 5 number that is beyond the realm of technology to underwrite,.

6 numbers in the range of 10 to the minus 5th,. 10 to the minus 7 6th reliability.

8 The third.option is to decrease the severity of 9 ATWS event~. Said another way, provide mitigating ~ystems 10 which assure *.that the probabiii ty of* severe co~sequences is 11 sufficiently low.* That's our preferred approach. It has*

12 been the preferred approach of the staff for some years~

13 The fourth option, some combination of the three, 14 for example, if some cost effective improvement in rods and*

15 drives could be proposed in combination with some.lesser 16 mitigating system than would accrue if you d.idn't decrease 17 the probability of SCRAM failure, there might be fruit in that 18 area. No one has proposed it, so*we haven't examined i t in 19 detail, i t doesn't seem cost effective.

20 CHAIRMAN HENDRIE: The last one has the difficulty 21 of involving the probability matters in *the regulatory 22 position in a very direct way?

23 MR. MATTSON: Yes, sir.

24 CHAIRMAN HENDRIE: For all that you look at the 25 reactor safety study here in the course of the staff's

25 1 rec6nsideration for the 99th t!me of this problem.

2 You eventually come down to the foll,o'wing judgments:

3 (a)* something ought to be done about ATWS, i t is the staff's 4 judgment, I take it, that i t ain't good enough as i t stands.

5 MR. MATTSON: Yes, sir.

6 CHAIRMAN HENDRIE: (b) you conclude that the best

.7 way to fix it is .to implement consequence limiting measures 8 rather than trying, as we started out in '73 to improve SCRAM 9 systems or equivalent.

10 Now, as in othei places where the staff has said 11 well, here's an accident sequence or type of accident 12 sequence that we think is a problem and ought to be fixed.

13 You now propose to class ATWS as a design basis accident.

14 MR. MATTSON: Yes, sir ..

15 CHAIRMAN HENDRIE:-- which means that the plant 1,6 designs ought to be such that they can deal with that class 17 of accident~ with consequences within the Part 100 ~uidelines.

18 In providing consequence limiting measures there is rio 19 consideration of how probable the design basis accident is.

20 MR. :t>'.IATTSON: In the proposed licensing system, that 21 we have, yes.

22 CHAIRMAN HENDRIE: As in the classic approach, you 23 just say it will be declared as a design basis accident. We 24 know it is low probability, never mind, but nevertheless, we 25 have declared it a design basis accident and that means you have

26 1.

1 got to* have something in the plant *that takes care of it.

2 So that the approach here comes back finally, for 3 all' the talk while we consider the place of.how this fits 4 with the reactor safety study and those considerations of 5 the staff, really does come back in the classical modes and 6 in effect or dir~ctly saying,*we believe ATWS.should be 7 a design basis accident. If i t is a design ba;is accident, 8 why we don't look at the combined probability consequence 9 spectrum and pick some acceptable combination of probability 10' and consequence for the fix; we say design basis accident 11 fix it.

12 MR. MATTSON: Yes, sir. With one exception --:-,

13 CHAIRMAN HEND~IE: Well, No. 4 up here would come 14 away from.that and say, no, what we would do wouid in fact 15 be to do: a probability consequence analysis and then fix some 16 acceptable place on that and say ~t has to be at least that 17 good. But that really does pull you into the risk analysis 18 business.

19 MR. HANAUER: There is something, however, that I 20 would suggest needs to be said.

21 .Every word you said is exactly* right, Mr. Chairman.

22 CHAIRMAN HENDRIE: That's the advantage of being Chairman.

23 (Laughter) 24 25 MR. HANAUER: But I would point out what we already

27

• 1 know, and that ip that the selection of the design basis 2

. accidents, a'nd the . sel~ction of, -- i°:n: det.aiL.,:,w.nich: .asstimptdcons 1

.3

'one shouid \make and whi,ch values and parameters one uses in 4

evaluating them are choices from a spectrum, and that these 5

choices h~ve

• been made in the past. based on unq:uantifiecL

.'6 engineering j{idgment and which *probabilities :and risks were 7

somehow involved. They were one of the ingredients even 8

though we really didn't know how to do i t in a quantitative 9

• _w~y.

10, The difference is that this time we have explicitly 11 taken some numbers and sh.own, in a much less than optimum way 12, how we have .used some quantitative ~isk and probability 13 ideas in choosing the accident, the assumptions to be made 14 in the evaluation, a~d in at least 6ne case, the ~alrtes of 15 the parameters to be used in the evaluation. And that's the 16 difference.

17 MR. MATTSON: The one exception to that was the same 18 exception I had to your statement.

19 '(Slide):_.**

20 If I *could turn to the next slide, we wiil talk 21 about briefly, the proposed licensing requirements. _ Quickly, 22 they treat*radiological concerns, primary system integrity, 23 fuel integrity -- you can read them as well as I can.

24 The exception that I want to talk about is the mitigat 25 ing system desigri: and the draft of the rule as i t currently

28 1

exists in.the report proposes an, option on the reliability or 2

availability of that'mitigating system~

', 3 It says you can proceed one of two ways,, Mr. Designer.*

4 The first way is the old way, you'.can make it safety grade 5

and do all of the things to gold-plate that system that go

'6

~long :with the definit;i.on of the word "safety grade". That 7

includes IEEE-279, that includes seismib qualification, 8

environmental qualification, all of those things required 9

pur~uant to our. regulation for safety grade equipment.

10 Or, you can show a reliability -- an unreliability 11 for that system of 10 to the minus 3rd or better. And that 12 10 to the minus 3rd system is good enough for. a mitigating 13

ystem.

14 Now, that makes our licensing requirements a bit of 1-5 a mix of the deterministic and of the probabilistic and it 16

• is fair to say at this point that mix is causing some 17 controversy. It caused some with the ACRS, causing some with 18 the Ratchet Committee and ---

19 CHAIRMAN HENDRIE: Well, let me tell you as a member

, 20 in good standing of the American Society of Mechanical Engineers 21 as well as other learner societies, I just make it Category I 22 safety system and my deep intuition tells me I would save an 23 awful lot of design in work and analysis just going down.the 24 classic route.

25 MR. MATTSON: That may very well be true.

29 1

. CHATIIRMAN HENDRIE: . 10 to the minus 3 or better for --

2 that's pretty good reliabili-i:y ~*

3 4, I wonder; if my car starts 9 9 9 timeS"

• out ::of every 5, l_, 000 or .':what else t~at I use in the ordinary, life works 999 times.6ut of 1~000 or better.

6 MR. MATTSON: We. ,think 10 to the minus 3rd'..is

• 7 achievable and can be demonstrated. There is some controversy 8

on that stibjept ---

9

. ,MR. CASE: ,Well, in fact our judgment says safety

. 10 grade does that, but as you point out, i t is so.metimes 11 difficult to prove that.

12 CHAIRMAN HENDRIE: Yes. It is going*to be shaky to 13 get there.

14 MR. MATTSON: As we go through the next slide ~hich 15 briefly characterizes what we would be looking for in mitigating 16 systems 17 CHAIRMAN HENDRIE: Before you go away from that, 18 please, since we are having a briefing and don't have to 19 face proposed rules, regulations or whatnot -- if in its 20 wisdom the staff judges that a rule::,.and an appropriate guidance 21 is the proper set of actions for the Commission to take, may 22 this time we please h~ve the rule contain the essential 23 elements, acceptance criteria in summary for~ and that all of 24 the rest of that good* stuff be in regulatory guides _: ihitead of 25 regulations?

30 1 MR.** MATTSON: That is exactly the proposal*

2, The primary acc.eptance cri t.eria *. would be 'in the 3 regulation and t,h'e. evaluation model~. would be in s*upporting 4 regulatory guid~s~

5. CHAIRMAN HENDRIE~ I know he .has been dea'iing with

.6 Appendix K for a long time and 7 MR~ MATTSON: We have-learned the same iessoI)..

8' ' Okay, if I could go back .to the next slide ..

9 (Slide) 10 The mitigating system we are talking about,* again,.

11 this just summarizes what I have already said. The classic 12 safety grade requirements -- one exception there: the expected

• 13 natural phenomena we would be willing to speak to the OBE 14 instead of the SSE in terms seismic qualification for this mitigating system.

15 16 Of course, able to withstp.nd the ATWS-environmental.

17 conditions inside a*containment, diverse, automatic, relative 18 to othe;- sys terns. Some credit for operator . action: there is.

19 a. magic number in the report aimed at 10 minutes, L *believe, 20 with some support in ANSI standards work and a general movement in our safety work toward that 10-minute* criterion.

21 And then the bottom two line:s, either safety grade 22 or reliability analysis with a demonstrated unreliability of 23 10 to the minus 3~d with 50 percent confidenc~.

24 CHAIRMAN HENDRIE: Well, at least you didn't hang up 25 in the 95 percent confidence level. There is some solace in

31 1 that. ,

2 MR. MATTSON: That's an important f~ature of what.

3 we are trying .. to do in the evaluation models is to say we want

• 4 to use realistic evaluation models for a. design basis accident.

5 We are having some difficulty with people.believing _it. :--' :.,

6 esr~ci~lly sine~ there are some notabl~ exceptions t6 that 7 ~ith moderator temperature *coefficient and use of service level 8 C limits for the pressure vessel.

9 But, in the main, the initial conditions and the 10 other parameters are treated realistically and we agree with 11 that,. and the industry evaluation models that we have been 12 reviewing over the last years.

13 If we could turn to the last slide.

14 (Slide) 15 At the bottom line, what do we think this safety 16 goal and these requirements will lead to in design of plants?

17 As I said earlier, with Westinghouse, some plants we think 18 will.make i t and other plants will need circuitry changes, 19 combustion engineering and,:Babcock ~ and Wilcox, the other 20 two PWR manufacturers, the same thing plus some relief valve 21 additions.

22 In the case of the boiling water r~actors, those 23 two lines are a little bit difficult to understand, let me try 24 it. It is recirc-pump trip for the boilers, plus a high 25 capacity boron system, got to pause there a moment ---

32 1

COMMISSIONER GILINSKY: That's a new system.

2 MR. MATTSON: Yes .

.3 General El~ctric has been in with thi staff 4

informa,lly and has ora_lly pr~posed an. automat:Lc boron syst?m o:f: some higher capacity than as previously been discussed

,6 for ATWS con~iderc;1_tions.

7 We continue to put th~ lin~ there at the bottom 8

of the reliability, the high pressure make-up*system, the 9

high pressure coolant injection system and t_he reacto:t. core

• 10 isolation cooling system on current BWR'designs. It is likely 11 that if the capacity is su_fficient and the:*quality .of the 12 high capacity boron system is sufficient,* that the . concern 13 with the reliability of present high-press~re make-µp systems 14 go away, that with the safety grade, high capacity automa,tic 15 boron system and with the existing high pressure make-up 16 capability, and with recirc-pump trip that the boiling water 17 reactors can be found to meet the safety goal.

18

, That doesn't say I c~n do i t tomorrow, there is a 19

• fair amount of technical work., evaluation model, things to be 20 done, but i t is our estimation of whe+e we would come out in 21 the future, but that would be the case. I think General 22 Electric shares that estimation.

23 We need, at this juncture, in the development of 24 the position and consideration of these things*by ACRS and 25 the RRRC to hear from General Electric on that subject. I am

33 1

informed that we will in the course* of* its presentation to 2

the ACRS wh_ich is scheduled for like, * *Jul:y:.

3 The ACRS Subcommittee', 1 .should have mentioned 4

earlier, is going to provide two days. for an oppo,rtuni ty for 5

industry 'to comment. the NUREG report., One day will be 6

,spent with two of the reactor vendors and the At?m.ic Industrial

,7 Forum, the other two vendors plus EPRT will make' presentations 8

and respond to que*stions from the Subcommittee.

9 That completes my presentation, .if you. have further 10 questions?

11 MR. KENNEKE: May I ask one question.

12 As your thinking is now would you haye the safety 13 goals numerically defined in the rules, or would that 14 appear 15

• MR. MATTSON: It would not appear in the* rule at all.

16 It would be in the technical ba~is and be used in the, 17 value impact assessment and would be used much the same way 18 as* I have used it.today, to context what we are trying to 19 achieve, but i t would not appear in the regulation~. It wohld 20 be deterministic criteria matching the <;reneral framework of 21 those regulations to that.

22 CHAIRMAN HENDRIE: To the extent that ~e are trimming 23 here on some sort of an overall safety map, we are trimming up 24 a corner that has long been in controversy. In effect, the 25 ATWS approach of the staff has always been in the context, that

34 1 as you build more plants, why it is inevitably true you 2 get increased reasons to trim some of these things out.

3 What sort of a'gradation of requirements do you see reasonable 4 and possible from very old plants to construction permits?

5 MR. MATTSON: Th~re are some alternative~ available.

6 The a~tiactiveness of the p~os* and cons, if you will, 7 of ~.those various alternatives,. I believe will be one of the 8 primary subjects for consideration by the Ratchet Committee 9 and by the ACRS.

10 The report does not go to that subject to any large 11 degree. When we start from where we are at today, I can tell*

12 you what I see we will be considering.

13 CHAIRMAN HENDRIE: Well, rather than discuss it at 14 length, let me I think it would be useful for the staff 15 to consider that sort of approach,--

16 MR. MATTSON: We are doing that.

17 CHAIRMAN HENDRIE: -- to the extent that older 18 machines that had been operating*for a long time don't have 19 fully upgraded ATWS provisions, they indeed represent a 20 small local risk element, I guess, but in the overall scheme 21 of things it doesn't loom very large and against the costs.

What you'.have is typically, ver often, is a safety yield per 22 dollar which rises as you come into latter plants, and what it 23 means for some of the earlier plants is, yes, you have got some 24 safety, but boy, it's very expensive to implement it. And 25

35 1

I think* i t is a fair proposi~ion.

C0MMI$SIONER GILINSKY: Well, you ~pply that

' principle every time you pao~fit:.,\:*~oo .*

'.I 4

MR. MATTSON: Yes.

5 CH.A;I:RMAN HENDRIE: Yes, 'and what we are ,talking a:bout here is. again, -- 'i~ fact, 'if people* don't or.der some plants, why the whole r~gulation here will be backfitted.

8 (Laughter) 9 MR. MATTSON: One way to come at this gradation and 10 the alternatj.ives for gradation of requirements, gradation 11 versus age of plant is to look at the contribution to the 12 cumulative probability of core melt, that is, th_e total 13 core melt frequency* from all causes for various levels of 14 ATWS fix, if you will.

15 Just to be simple, if I said 10 to the minus 6th 16 for all future plants and 10 to the minu* 5th for all past 17 plants, I can sum the contribution to the core melt frequency, 18 the total core melt frequency .from let's say the hundred older 19 plants versus the contribution at another safety .objective 20 for a hundred new *plants and show the difference for no 21 backfit versus backfit to va~ying levels~**

22 Again, that's the kind of thing that depends fairly 23 strongly on using the reactor safety* study, and the numbers 24 which can be derived on core. melt frequency from the reactor 25 safety study. That is using them in a relative risk mode,

r 36 1

that is, you are comparing increments and deriving changes 2 in the overall risk from the incremental changes in the 3 pieces* rather than saying that the reactor saf~ty study is 4 an absolute measure of risk.

5 CHAIRMAN HENDRIE: I. think in that context there is 6 much less*question ~bout it.

7 Buppose it weren't labeled WASH-1400 to reactor 8 safety study, but were a set of publications and ba6k-up 9 ,file reports, and MIT reported nuclear technology, a journal 10 arti~le or whatever, staff wouldn. 1_t hesitate to .use the -pr,ipcip es 11 laid down there as well as some of the relative numbers.

12 MR. *MATTSON: There is one other principle that is 13 being brought to the backfit question, and that is the principl 14 which underlies the systematic evaluation program.* The basic 15 outline there is to look at the plant generally, at a number.

16 of areas, and then make a .decision overall on the safety of 17 the plant, not piecemeal on 175 issues.

18 That same concept can be brought to _bear for the 19 ATWS question for the older facilities, i t is in the context 20 of overall safety*and i t is an important contributor.

21 CHAIRMAN HENDRIE: I recommend that approach for

• 22 for the older plants, because i t would make little sense.

23 On a particular older plant, to fix ATWS and thereby make 24 a 10th of a percent difference in the overall risk situation, 25 that is, there may be a~.iirimher of the older plants in which

37 1

ATWS doesn't represent the same proportion of the core melt 2

risk as .it ~oes in the Surry, Peach Bottom tases as calcuiated 3

in WASH-1400. And ther~ it ~ak~s progressively less sense 4

to fix ATWS on a sort of a blind-across-the-board basis, or*

5 another way of putting* it is just as I did before, that in 6

those cases. your safety yield per dollar of in.vestiment is 7

going way down.

8 COMMISSIONER GILINSKY: There are several measures, 9

you know, there is overall safety for all of the reactors, 10 and then there is that particular reactor.

11 MR. CASE: Yes, we are talking in that context.

12 COMMISSIONER GILINSKY: And you have to look at 13 several ways. I think you also have to be cauctious about 14 using this safety study. I want to repeat that point, 15 assuming kind of an implicit Commission support for certain 16 numbers on the basis of not having shut down reactors.

17 You could also argue that none of these reactors 18 have the ATWS fix and we have not shut that down. So why 19 institute it now. I think it is just an overcritical way 20 of looking at it.

21 MR. CASE: You must appreciate that the biggest 22 criticism that industry made against our ATWS position of 23 three or four years ago was WASH-1400 came out and you haven't 24 taken this account into account in your deliberations. We 25 were forced to do this because -- and it was a fair question

38 l

that they ~aised.

2

. COMMISSIONER GILIN.SKY: Well, you obviously want 3.

to take into account work that has been done,: but I, think 4*

you want to be careful about imputing to the Commission

• S ,,

some guidance concerning overallr~sk levels. I.think that is something the Commission really hasn't addressed.and 7

something I think it should address.

8 CHAIRMAN HENDRIE: Okay. Well, thank goodness 9

this was a briefing and not a decision.

10 I can simply say thank you very much.

11 (Whereupon, the meeting was concluded at 11:55 a.m.)

12, 13 14 15 16 17 18 19 20 21 22 23 24 25

J