ML20151K495
| ML20151K495 | |
| Person / Time | |
|---|---|
| Issue date: | 07/27/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1683, NUDOCS 8808030198 | |
| Download: ML20151K495 (341) | |
Text
. . - - - - _ _ _ - - .
1 . . '
UNITED STATES O
. NUCLEAR REGULATORY COMMISSION
. m m e .
., . m m . = = = = = = = m as s ammess a m m ess= = = = = es m us e as e m m e = = = = = = = m e = m m m J
l ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
}
t In the Matter oft ) l 1
SUBCOMMITTEE ON DECAY HEAT I
)
REMOVAL SYSTnMS )
) !
1 !
l '
i i
I O i j l l
l 1
! f i
l l <
n
- N !
DATE: 12'y 27, 1988 *
.nyv ,, , , . . r4 y mt y4 t, :
.t 9 ,g ,n,, , f
.s 2 . , i r Washing ton , D .C .' i T ~
- - ' , , [ ', * ,
I LOCATION: s' :(
f) , ,
i <
l e *
, ; e n.' ;;
PAGES: 1 through 232 s,i :. i
- t c- *
- f; j ) f J
l,u%uf
. : t ,y r *'.. 7 /.
UO ! .1,%.std!:E.:.
= = se m m as se m en m e ss am m m em me ss o us w e m m m es m em an an as em as s an as e m l I
i l HERITAGE REPORTING CORPORATION 1 ep.wn-O 1120 L Seest. N.W., Sehe 400 i
l 690003019a uno7py Wan % D.C.30005 f l (282) O H 000 ,
l N76o$*' PNu I
l t PUBLIC NOTICE BY THE
() 2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S 3 ADVISORY C0KMITTEE ON REACTOR SAFEGUARDS 4
5 ;
6 ;
7 The contents of this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards (ACRS), !
10 as reported herein, is an unestrected record of the discussions i
11 recorded at the meeting held on the above date. (
12 No member of the ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors or [
t 14 inaccuracies of statement or data contained in this transcript. l
() 15 f 16 17 18 19 20 21 22 23 24 25 O t Heritage Reporting (202) 628-4888 Corporation j
1 1 UNITED STATES NUCLEAR REGULATORY COHHISSION 2 ADVISORY COKHMITTEE ON REACTOR SAFEGUARDS 3 )
In the Hatter of: )
4 )
)
5 SUBCOMMITTEE ON DECAY HEAT )
REMOVAL SYSTEMS )
6 )
7 Wednesday, July 27, 1988 8
Room 1060 9 1717 H Street, N.W.
Washington, D.C. 20555 10 l The above-entitled matter came on for hearing, l
11 pursuant to notice, at approximately 8:30 a.m.
12 BEFORE: HR. DAVID A. WARD 13 Research Manager on Special Assignment E. I. Du Pont de Nemours & Company 14 l Savannah River Laboratory Aiken, South Carolina 15 l
i
!! ACR SjtEMB.RRS_PPRE S ENT :
16 ,
I DR. WILLIAH KERR 17 g Professor of Nuclear Engineering d and Director of the Office of Energy Research 18 l:
Univesity of Michigan Ann Arbor, Michigan 19 pl 0l HR. CHARLES J. WYLIE 20 i Retired Chief Engineer i Electrical Division 21 Duke Power Company l Charlotte, North Carolina 22 ;
CONSULTANT.S.:
23 p I. Catton 24 ;f P. Davis b
25 11 o
I i
l HERITAGE REPORTING CORPORATION -- (202)628-4888
)
1 2
l 1 ACRS COGNIZANT STAFJ_M.EMB_EP,:
2 Paul Boehnert 3 URQ_S_TAFF PRESENTERS _:
4 R. Woods W. Hodges 5
i 6
7 8
l 9
10 11 12 l 14 I 15 i ;
- 16 t li t j 17 i I
l 18 l l
19
- I
- 20 i f
21 ,
- 22 I !
23 l{
li !
24 25 I
]
1 n ll HERITAGE REPORTING CORPORATION -- (202)628-4888 l
E - _ - _ _
3 1 R R Q C. E E D I_ U G S.
2 CHAIRHAN WARD: The meeting will now come to order.
3 This is a meeting of the Advisory Committee on Reactor 4 Safeguards, Subcommittee on Decay Heat Removal. I am Dave 5 Ward, the Subcommittee Chairnan. Other ACRS members in 6 attendance are Hr. Kerr and Mr. Wylie, and we have ACRS 7 consultants Mr. Catton, and I believe Mr. Davis will be here.
8 The purpose of the ineeting is first to continue our 9 review and in fact perhaps complete our review to the l
10 l Subcommittee from the standpoint of the NRC staff's resolution i
11 l for USI A-45, and second, to discuss the status of Generic 12 I-sue 99, which is concerned with loss of decay heat removal 13 in pressurized water reactors during shutdown conditions.
14 Paul Boehnort on my right is the cognizant ACRS 15 staff member for the meeting. The rules for participation in 16 ,
today's meeting were announced as part of the notice of the 17 0 meeting published in the Federal Register on July 5th and July b
18 13th, 1988. A transcript is being kept and will be made 19 4 available as stated in the Federal Register notice. I request 20 ;
that each speaker first identify herself or himself and speak 21 with sufficient clarity and volume so that she or he can be 22 readily heard, u
23 4 The first half of the meeting this morning will be 24 e concerned with the first issue, A-45. I think there really 0
25 1 isn't a lot of new material here. :taybe I will be surprised, h
HERITAGE REPORTING CORPORATION -- (202)628-4888
(
4 1 but from what I have read, most of what we are going to haar h 2 is the material that we have heard at least in the preliminary 3 form over the, within the last several months, but I think we 4 will, what we will hear is that this is now the staff's 5 official position on the AP-45 issue, and I think we will want 6 to take this to the Full Committee at the next meeting, and 7 the Full Committee will probably want to comment on it, and in 8 fact I think the staff is asking for and expecting some 9 comments.
10 HR. BOEHNERT: That is correct.
11 I CHAIRMAN WARD: From the Full Committee; the second i
12 issue, the GI-99, as I understand this is in the nature of an l 13 information briefing today. The staff has not reached a 14 resolutf.on position, and we may decide that we want to bring l
15 that to the Full Committee and comment. We may decide we 16 ,
don't want to at this time.
I i
17 Any o*her members have comnents they would like to l
l 18 ; make before we go to the agenda? Okay. I believe Roy Woods l
19 l is going to start off, so Roy?
I 20 MR. WOODS: I was going to start off by telling you q
21 what you received in your package. I don't see any packages 22 j in front of you thick enough, so I don't think an:rbody i l
C 23 h actually has the whole thing, but let me do that anyway. l l
24 What you don't have is six case study reports that l l L l
25 l we have done over the last seven or eight years that measure u
O j HERITAGE REPORTING CORPORAT. TON -- (202)628-4388 w
5 1 about 11 inches thick that we didn't send you. What you have received is a tummary report which was Enclosure B of the l 2 3 package that we have been circulating within the staff 8. o r 4 approval which is the final resolution package, proposed final 5 resolution package for A-45. That's a summary report of those 6 six case studies.
7 Then there is a summary of that, sort of a summary 8 of that. It is really almost as long as that because it also 9 lincludessome value impact work in the regulatory analysis i
10 L which was Enclosure A to that same package that we sent you, i
11 Continuing to summarize on the package we sent you, 12 j there was a Commission paper that summarizes the regulatory l 13 analysis, and then there was a sumnary of that which was a 14 h memo to Ed Jordon, the CRGR, and my first slidi really is a 15 summary of that summary. I have lost track of how many r
16 sunnaries there are, but we have got down to one slide from an f
i 17 L. ll-inch stack, so with that, I will put on my first slide.
18 Needless to say, there is a lot left out here if I've 1 i
- i 19 g summarized it that many times, j h l 20 [" (Slide) ;
l u
21 j HR. WOODS: Okay. Concerns on the decay heat l 22 removal reliability were originally raised, a *. least raised inj i
i 23 j this train of thought about THI. It got extended to include l 9 l 24 g special emergencies which we mean such things as fire, flood, I h
25 [ sabotage. It was recomrended that a high priority be assigned u
HERITAGE REPORTING CORPORATION -- (202)628-4888
]
6 1 to the efforts to improve tne DHR function, that as a result h 2 of that top item on this slide, unresolved safety issue A-45 3 was approved in late 1980 as USI A-45 of course.
4 The key questions that it not out to answer are do 5 the current regulations or the then current regulations 6 provide sufficient assurance that the risk from DHR fields is 7 acceptably low, and are risks to the DHR function and 8 operating plans cost beneficial? I am pretty much just 9 reading the slide here.
10 I The key conclusions were that indeed DHR failures 11 ;
are a Jignificant contributor to core damage frequency, and 12 j particularly to small break LOCAs and transients.
l 13 Vulnerakilities were fou.td to be quite plant-specific. We 14 fbasi<allydidnot succeed in finding a single generic fix that d
15 I wculd be applicable to all plants.
16 h We also found that the dedicated decay heat removal h
17 [ system that had been proposed was not cost beneficial- using h
18 h the, any normal methodology of doing cost / benefit analyses.
0 19 [ We really didn't go back and compare the designs to the, to N
20 h the, all of the requirements because the requirements are 21 design basis accident oriente.d, and what we had done was more 22 h of a PRA oriented analysis. Certainly we would have taken 23 . notice of glaring inconsistencies, but we didn't totally closel I
24 h the loop and go back and ask that question which would be l H !
25 related to the first one--do current regulations provide 4
t i
h d
HERITAGE REPORTING CORPORATION -- (202)628-4888 1
7 1 acceptable assurance? We arrived at risk, but we didn't 2 really look in detail at the current requirements.
3 DR. KERR: Could I conclude, Mr. Woods, that the 4 question one, that there does not exist an answer to key 5 question one? At least I don't see one on that slide. I 6 don't even see the question addressed.
7 MR. WOODS: Okay. You asked a slightly differer.t 8 question. A-45 didn't provide a direct total answer to that l
i 9 j question.
10 l DR. KERR: Did it provide an indirect total answer h
11 q to it?
o 12 0 MR. WOODS: Well, you will get an answer because I 13 will tell you what we think the quantitative risk is from DHR 14 h failures, and that's the risk that is, that is the end result
!i 15 p of the application of the current requirements, and if in your d
16 h judgment you think that is too high. then somewhere there is a h
H 17 problem, i i 18 DR. KERR: The question as I read it from your slide!
19 sufficient assurance. I don't know what that is, htalksabout f
20 1 HR. WOODS: I am prepared to tell you what I think h 1 ,
21 ;h, sufficient assurance is. l 22 DR. KERR: And acceptably low, und I don't know what 23 l' that is. !
24 MR. WOODS: That's what I am planning to talk about, h
25 h DR. KERR: Okay.
h 4
L
[
HERITAGE REPORTING CORPORATION -- (202)628-1688
l 8
1 HR. WOODS: What I am not going to tell you is I 2 looked back at all the detailed single failure requirements 3 and the DBA based regulations, and decided plant X is not in 4 compliance with this partic lar requirement, that I didn't do.
5 That's the thing that I have not answered, but I ' chink I will 6 be able to answer what really ought to be the question--is the 7 risk lov enough or not? That's what we are going to talk 8 about.
9 In fact, at the bottom of this slide--I guess I 10 would have to move it up if you want to use this--the key 4
11 f assumption that we made we think, the staff thinks, that core 12 damage frequency goal of about one times ten to the minus five i I l 13 per reactor year would be an appropriate goal to shoot for for 14 decay heat removal, related core damage frequency. I am 15 h prepartsd to go into some more detail on that later.
I 16 h We believe that that kind of goal is consistent with h
17 'h wh a t was done in USI A-44, which is Station Blackoht, and I
18 L
thero we used the goal of, the same goal, one times ten to thel l
19 minus five, for blackout related : ore damage frequency, and in 20 4 USl A-49, which is pressurized thermal shock, there we used a l 21 J slightly lower goal because there you are talking about more 22 a severe accident where you have already breached more of the l i
23 , barriers. You are talking about an event that has already breached the primary system, and we used basically five tines t
h 24 g
25 ten to the minus six: in other words, half of the one times il HERITAGE REPORTING CORPORATIOli -- (202)628-4888
9 1 ten to the minus five goal for that one, so we think what we g 2 have done here to consistent with what has been done before.
3 DR. KERR: When you started the investigation, you 4 hadn't decided what was acceptably low, but in the course of 5 the investigation, you concluded that it ought to be like 6 ; A-44? l 7 HR. WOODS: Yes, that is correct.
8 l DR. KERR: What does sufficient assurance mean?
l 9 4 HR. WOODS: It means th-t the numerical result you b
10 [ get is as low as you think it should be. I don't know how to b
11 9 give you a more definitive answer.
12 0l j
n DR. KERR: Well, acceptably low it seems to me is l 13 some measure of whether the risk is low enough or not, but the 14 l way the question, key question is worded, I wouldn't have
- f 15 h thought sufficient assurance was something else than was 0
I' 16 L; acceptably low.
! I am trying to understand the question.
17 HR. WOODS: I don't make that distinction. If we 18 P had seen the results of our case studies consistently come out h !
19 at one times ten to the minus five core damage events per 20 i' reactor year, and we thought that would be generally h
21 h applicable to all the plants, we wotad have concluded that 22 0 there is no further corrective actions needed, that the risk i
23 is sufficiently low. That's what we would have concluded. l i' l 24 h That is not where it came out. That is not what we concluded,j i
h 25 but that's what I mean.
H I
l' HERITAGE REPORTING CORPORATION -- (202)628-4888 i
10 1 DR. KERR: I am looking at a term that says 2 sufficient assurancs, and I, to me that means something like g
3 confidence or validity, not the number itself, the number is 4 the low. Maybe sufficient assurance in this context doesn't E have any significance. If that's the case--
6 HR. WOODS: I am at a loss. I don't appreciate the 7 difference somehow, so I'm not able to answer your question.
8 HR. CATTON: Does sufficient assurance apply some 9 l method of calculation, or way of going about it?
10 HR. HIliNERS: Yes, it does, and the regulationa, as l
h 11 0 es know, were deterministic. We are saying the deterministic 4
12 ! regulations from what we looked at probably don't give you f assurance that risk is ten to the minus five per year.
l 13 14 ll HR. CATTON: The answer to that first question is h
15 [ no?
4 16 11 HR. MINNERS: Not quite, because we, we can't--what 17 ;
we, when we did the studies, if you foun3 something that was l I 18 {
obviously not in accordance with the regulations, we would 1 l
19 j have reported it to NRR and compliant action would have been d
20 taken. We didn't find any of those things. But then again, 21 we didn't go back and look at the FSAR and the design, and the h
o !,
22 licensing basis of each plant and say hey, does this plant h l 23 h meet its licensing bases? So it may be possible the plants f
24 ,
don't meet their licensing bases. That wasn't our job, to l l
25 ' search it out, so we can't say positively. If that met the j i
. HERITAGE REPORTING CORPORATION -- (2021628-4888 t .
b 11 1 regulations, you are right, cut I think this is a pretty good h 2 feeling on our part that since we didn't find any claring 3 absence from what the regulations require, that you know, just 4 meeting the regulations is not going to get you down to the 5 ten to the rainus five.
6 DR. KERR: Well, I guess I'm puzzled by that 7 statement because I would have thought that one of the first I
8 things you would have wanted to do is to find out if these l
9 plants met the existing regulations, because it is a pretty 10 important question I think to determine what existir3
- l l
11 ] regulations make plants appropriate 1, safe, and apparently you h
12 H did not examine that in this study at all.
'i l 13 HR. HIhhEAS: In a systematic way, no, I can't swear L
14 that these planes meet all the regulations.
j 15 HR. WOODS: We only looked At six plants so when wo 16 [ had done this on the six plants-li HR. MINtiERS:
17 But the vulnerabilities that we found 18 are not things that where they don't meet the regulations, i
l' l 19 j DR. KERR: You did not attempt to determine whether (
I 20 they met tha regulations or not?
t l
21 [l MR. HINNERS: Not in ever detail, no, because I !
. \
22 don't think that was a w rthwhile effort. I think what we !
i l
23 did, we looked at, to the vulnerability, said these are the !
a 24 contributors to risk, and if we had, they met the regulations.
L 25 4 with that change in risk, then the answer is no.
d 11 lI HERITAGE REPORTING CCRPORATION -- (202)628-4888
12 1 DR. KERR: I, the reason I think it is import 1nt is l 2 because the answer we get in discussing with the people 3 responsible for 11.50 when we asked the same sort of question 4 which is do existing regulations take care of the 5 vulnerabilities that one encounters when one looks at severe 6 accidents, and the answer that we get is, as far as I can 7 understand it, is no, you have got to look at each plant 8 individually. You can't look at it on the basis of 9 regulation.
1 10 . Now in this case, it staas to me this is an I
11 It says that existing qextremelysigi.ificant finding.
And I think we ought 12 1[ regulations do not ensure plant safety.
l 13 to know whether they do or not.
14 HR. WOODS: I think we have kind of answered the i
15 i kind of question he is asking because what we did is we looked 0
16 g and we found the vulnerabilities and the vulnerabilities--
17 ! DR. KERR: You told me you didn't really lock to 18 find out that the vulnerabilities--
- l 19 i CHAIRMAN WARD: Maybe--I think what they said, h
20 j Warren seemed to say two different things. Maybe I h i 21 nisunderstood or maybe--but one thing he said was I think he h
22 y agreed that what they feand here was sort o .*; a subset of this h
23 !
general conclusion about 11.50. He said where they did find (
l 24 o some vulnerabilities in these sfx sample plants, they did not !
25 find that those were related to the failure to meet j i
(i h
HERITAGE REPORTING CORPORATION -- (202)628-4888
13 1 regulations.
l 2 HR. WOODS: That is correct.
3 CHAIRHAN WARD: So that's really the sinilar 4 conclusion.
5 DR. KERR: But I'm not sure that they .ooked very 6 c3refully. If they didn't m;ke the systematic study of what--
7 HR. WOODS: The kind of recommendations we would be 8 making seemed to be beyond the current regulations. So if-~
I 9 '
DR. KERR: If that's the case, it seems tv ne we are i
10 operating a lot of plants in a situation in which we are 11 convinced that the existing regulationn do not rake them safe d
12 q enough.
l 13 CHAIRHAN WARD: I think--Roy, correct me--but I 14 d think you went into this study assuming thst the plants all h
J 15 q met the regulations in this area.
16 , HR. WOODS: That is correct.
5 ii 17 4 CHAIRHAN WARD: You didn't have any reason a I I
h 18 suspect that they didn't.
19 '
HR. WOODS: Our objective was not to look for placesj a
20 p where they don't meet current regulations.
21 ! CHAIRMAN WARD: The one objective was, as you said, 22 to find out if meeting those regulations is sufficient to give 23 E you perfornance of decay heat removal systems at the level you I
24 ; want. and the answer seems to be--
g 25 HR. CATTON: Haybe.
o HERITAGE REPORTING CORPORATION -- (202)628-4888
14 1 CHAIRMAN WARD: Maybe; in some cases, no.
,r x
'ks ) 2 HR. WOODS: The answer seems to be we don't feel we 3 have full assurance that it does provide sufficient safety, 4 but we didn't find any particular places, glaring places where 5 it doesn't meet it. We weren't really specifically looking 6 for that, but we didn't, in the course of doing what we did 7 do, we didn't find such places. It was not the objective of 8 our otudy to do that. I think ic would have detracted from 9 what we are trying to do if we had done that.
10 CHAIRMAN WARD: But that's a key question you have i
11 l got here.
I 12 l HR. WOODS: It started out--I wasn't even involved il l 13 when it 1 tarted out so I am not really sure exactly what the 14 h main questions were, b it as the study evolved, it didn't seem b
15 g like that was the way to spend most of your time, looking at I
16 h that question. We decided we should look and see what h
h 17 [ thenvulnerabilities are, what the total risk, and whether or U
li not we could come up with fixes or corrective actions that f
19 would make it better, and whether or not those fixes,
]h o
20 h,
corrective action would be cost beneficial. Those are the 21 questions we were trying to answer. ;
e !
22 ( MR. MINNERS: I will take fire protection for an l
23 . example. I an probably going to get into trouble with taking I I
24 it as an exarple, but one of plants we studied had a l
% 25 L
1 q calculated vulnerability to fire because it did not have a i
1 HERITAGE REPORTING CORPORATION -- (202)638-4838
15 1 work quenching system in the cable spreading room, because it 2 had switchgear, and it just had--I shouldn't say just it has.
3 DR. KERR: Marty, is that mike on?
4 HR. WOODS: Turn it on.
5 HR. MIN!iERS: So the, that was a high vulnerability.
l 6 I Now does that aset Appendix R7 I have gotten various 7 expressions. The plant was licensed, so therefore it does 8 meet the regulations, but I think that's an area of 9 questionable, you know, where there is a vulnerability, and I 10 j don't know how to answer that question, whether it meets the H
a 11 j regulations or not, and I don't think we tried to do that.
I It 12 H wasn't fruitful, and if you are looking for some ganeric way 1
13 in which regulations are--they are generic requirements. We j 14 [ couldn't find a generic requirement that we cure all the h
15 h problems that we found.
I 16 j DR, KERR: The reason that I'm raising this issue is h
17 '
because it seems to me that we need to determine whether i 18 existing regulations make plants safe enough. Now when we l 19 asked that, I don't think we are asking necessarily is does !
l l 20 q the fire protection good enough? Is decay heat removal good j i
21 enough? Is something else good enough? What we are asking is 22 are the plar.t s safe enough?
23 !
Now it may be the plant is not sate enough becaust. 1 24 it doesn't meet the existing regulations, and therefore, 25 1 according to the risk analysis, the risk is too great. If i
il l'
[
HERITAGE REPORTING CORPORATION -- (202)628-488S
16 1 that's the case, then the answer is you don't have to change n
i K
x [) 2 the existing regulations. You simply have to make certain 3 that the plant is operating, the desiga is being operated 4 according to the regulations.
5 If the answer is it meets all existing regulations, 6 it is still not safe enough, then the answer is that to me, 7 you have to change the existing regulations or you have to 8 agree that something other than the existing requistions is I
i 9 I needed to make plant safe enough. I think this--
1 10 h CHAIRMAN WARD: But Bill, isn't that what is going h
n 11 p on? I don't see why you are not answering that.
d 12 l HR. WOODS: I think I have answered that if we start l 13 out and assume that these designs mest the regulations--let's 14 I make that assumption for the moment.
15 DR. KERR: I don't see how you can make that U
N 16 L assumption.
17 HR. WOODS: I can't even give a talk without naking .
I 18 l an assumption.
H 19 l CHAIRMAN WARD: One way is there is a whole body, 9
20 L you know, 80 percent of the activity of the regulatory agency i
21 ; is directed toward that.
22 DR. EERR: Every week almost we see examples of 23 systems in operating plants which, for exarple, do not meet i
24 the single failure criteria, and they sinply, they haven't l l
25 ,
been found. I think this is an honest effort and therefore it HERITAGE REPORTING CORPORATION -- (202)628-4888
17 1 seems to me in a study like this one needs a comparison 2 between what one determines the risk analysis and whether the
( 3 plant conforms to regulations.
4 ,
HR. WOODS: If we do assume for the moment despite 5 the objection that the plants that analyzed meet all the 1
6 current regulations. I'm prepared to tell you here that we 7 ; don't think that even making that assumption, even whether you 8 lmaketheassumptionornot, we don't think the plant as it i :
9 ) exists really has quite the assurance that we want to sea.
10 4 The risk is a little bit too high. Okay.
11 DR. KERR: Talking about total risk or--
12 HR. WOODS: This whole study is just decay heat d i 13 removal related failure, core damage frequency. That's 14 h correct. That's what we are talking about, the risk from that 15 f kind of failure. And that's all A-45 is, so please don't i
i 16 L broaden this to being total plant PRA.
l l 17 DR. KERR: I have to talk about a perspective about l 18 the total plant because that's what I an interestad in, i
19 HR. WOODS: Lcok at what I am telling you. That is ;
20 ' to how it affects the totel plant. but don't expect expect me 21 L to be able to tell you more than I did. I don't want to be i
22 dishonest with you and say we did more than we did. Okay, i
23 So I start out and I assume the plant meets the j 24 regulations, whether it does or not. and I conclude that the 25 plant needs more, and I look at the type of things that 'l l l
c i
)
I HERITAGE REPORTING CORFORATION -- (202)628-4888
18 1 would mean by more, and they seem to largely, the majority of g 2 them, be beyo.1d the regulations, so I think I am concluding 3 that the current regulations don't provide the assurance of 4 the one times ten to the minus five. That chain of logic. I 5 think I answered your question.
6 CHAIRMAN WARD: Yes. I don't know why you didn't 7 say that earlier.
8 1 HR. WOODS I had to think a minute.
J 9 j HR. HINNERS: You asked the question the other way il 10 y around.
11 , CHAIRHAN WARD: But further you concluded that you 12 can't come up with any change in the regulations, the 1
l 13 regulations of the deterministic form, but--let me finish.
14 That's the second point. You can't come up with any, you have d
15 haven't got any idea as to how to fix the regulations to close 16 0 the gap or to whatever you want to do.
17 And so third, what you are proposing is a risk based 18 evaluation to identify problems in specific plants that could i I
19 L be fixed. I 20 HR. WOODS: That's a good summary. Why don't I just t
21 , sit down and we will go on to the next speaker? That's I l
22 '
exactly where we are, because we were unable to fix a single l 23 regulation that would be applicable to all plants. That wouldI l
24 mean a generic fix. And we were unable to come up with the l 25 fix that would get us, that we could justify that would get i l
l HERITAGE REPORTING CORPORATION -- (202)628-4888
19 1 the risk down to the level that we would like to see it, and ll 2 that's why we are proposing an individual plant look. Okay.
3 DR. KERR: It is true, I think, thst you did nct 4 look at the other risk contributors? You just looked at this I
I 5 ;
one?
6 HR. WOODS: A-45 looked at risk contributors that ii 7 ! come from decay heat renoval failures and that, and the only
{
8 part of the decay heat removal failure, we didn't lock at 9 plants that come, that start from cold shutdown. That you 10 l hear about this afternoon.
11 '
DR. KERR: You are not telling me necessarily that .
I 12 l the plant itself is unsato, but just that the tisk !
l l l l 13 1 contribution from decay heat remeval and inadequacies is 14 F greater than you would like to seet 15 MR. WOODS: That's exactly what I am telling you.
16 And this afternoon, later this morning the industry will be 17 telling you that the risk from that type of failure they don't i 18 believe is as high as our results in$icate, so there is a 19 disegreement about whether or not the results, the risk, 20 really that kind of risk really is an high as we are saying it l l
21 ; is. l
\ l 22 DR. RERR: Your conclusion is besed on the numbers 23 you get from risk analyses?
24 HR. WOODS: Limited risk, i
25 DR. KERE: Almost entirely from--and the cora meat i h
h HERITAGE REPORTING CORPORATION -- (202)628-4588
20
) 1 frequency that you get?
2 H P, . WOODS: Core melt frequen.y is not determined.
3 The term I want to use, I want to use core damage frequency 4 because we really looked only at--
5 DR. KERR: I will accept core damage frequency.
6 HR. WOODS: Yes. Co r t (, c t .
7 HR. HINNERS: There are a number of factors; the 8 usual uncertainties with PRAs, and there are probably areas t
9 l that we have not seen, okay?
10 DR. KERR: And at least from what I looked at, PRAs, 11 1 it is not difficult to find a factor of ten difference in 12 h calculations in either sequences or total risk, when the 13 calculation for different groups both of whom one might judge 14 ,
are competent groups.
15 i HR. WOODS: On slide 9 you will a difference of a 16 factor of 31.
17 HR. HINNERS: The way you say that. Dr. Kerr, there 18 is an implication that people are using different numbers. I r
l l 19 think that I--we went through this comparison study. I think l l
20 our nur.bers were the same. Okay. But what happened in a lot '
21 of cases, they would point out to us that there was another 22 systen. Okay. They would point out to us the numbers were 23 different. We had different LOCA frequencies and so, yes--
24 DR. KERR: What I am simply observing, that it is j 25 not unusual to differ by a factor of ten, i
I I
HERITAGE REPORTING CORPORATION -- (202)628-48S8
21 1 HR. HIliffERS: But I don't want to leave the 2 implications that you got two different people who just 3 arbitrarily put two different numbers. There were engineering 4 reasons for the difference in the numbers.
1 5 DR. KERR: In fact quite the contrary; I am saying 6 { if ene does an honest, capable evaluation, two different 7 ] groups, it is not unusual apparently for them to differ by a L
8 d factor of ten in the number they get. !{ow that gives me a d
9 1 little bit of measure of uncertainty invol/ed in the process.
10 HR. WOODS: If you will let me get to slide 9, I can I I have 11 tell you where some of those differences come from.
12 L got a pretty good idea of what it comes from as Warren is 1
13 trying to do.
14 i DR. KERR: Independently of where they come fron, l l
l 15 what I an saying is, you know, if you do this, it is not 16 unusual to see this occur. And therefore, one might 17 conclude--I'm not yet ready to draw this conclusion--but one 18 r.ight, that ons should not take seriously differences by a i
19 factor of ten because that's within the uncertainty of the 1
20 proceJs. ,
21 HR. WOODS: I have no argument with your conclusion.lI 22 HR. HIlillERS: I think I have an argurert with that 23 conclusion. I think differences of ten is a significant 14 difference, and if there is that difference, you ought to go ;
25 . search out and find out who did it. I 1
e l HERITAGE REPORTI!iG CORPORATIOri -- (202)628-4888
22 1 MR. WOODS: I think we have done that.
g 2 MR. MINNERS: That is not just uncertainty.
3 HR. KNIEL: I agree with that, and the uncertainties 4 are a factor of ten. If you compare, you ccapare the two 5 different calculations and find out where the differences are, I
6 and the differences can be attributable to fundamental views l
7 l or tc lack of data on one side or the other.
l 8 DR. KERR: But you see now you are going to assume i
9 your calculation is valid and those peoples are not.
h 10 j HR. KNIEL: We are just objecting to the way you 1
11 h have posed, at least I am objecting to the way you posed the 1 ,
l 1 12 l factor of ten as to being that's the closest we can get in 1
13 terms of-- <
14 DR. KERh: I didn't sa) it van the closest you could 15 get. I would say it seems to be what one aces as not unusual 16 HR. KNIEL: I don't agrce with that. You con get 17 ,
factor of ten difference in a sort of a totallynewarewwheref 18 ; you give somebody a totally new kind of problem, and you will 19 get two people and they may get a factor of ten difference. l 1
20 We have been working at this now for years and think a lot of i
21 tne sequences and a lot of probabilities have been very well ;
i 22 established cnd accepted, and I don't think factors of ten in 23 those cases--
l 24 HR. WOODS: I am prepared to go into one pu ticular i I
25 example. ;
HERITAGE REPORTING COPf0 RATION -- (202)628-4888
23 1 CHAIRMAN WARD: Why don't we go ahead? Carl, w,uld l 2 you identity yourself?
3 HR. KNIEL: I am Carl Kniel from the staff.
4 HR. WARD: Go ahead, Roy.
5 HR. WOODS: I think I have spent fifteen minutes on 6 one slide and I have got ten slides. That's 150 minutos, and hli i
7 p1 that's too long.
8 This slide I have done. We did six case studies h
9 L which I have mentioned. We already covered what they are.
f 10 jl They are entirely related to decay heat removal failutes. We l I
J l
11 ? did linit them to--well, I have already told you it doesn't
- i
" include shutdown DHR failure risk, and it does include fire, 12 l 13 flood, seismic, insider sabotage. The only point I want to 14 make, we did not include the kind of sabotage that would j 15 involve external attack.
16 HR. HINNERS: The nunbers you see don't say that. ,
17 HR. WOODS: Okay. You are right. In Septenber of l l
18 '86, we came before the ACRS. I'n not certain at the nor.ent 19 if it was this Subcommittee or the Full Committee. That was a, I
20 predecisional draft of the regulatory analysis available at !
I 21 that time. It showed you six alternatives that we considered,l i
22 and it reccreended at that time a dedicated shutdown heat i i
23 removal system. You now see on this slide the sane six 24 alternatives, but we have a different raccreendation which is g f
.5 alternative 2. which is linited scope PRA.
h l HERITAGE REPORTING CORPORATION - (202)628-4888
24 1 (Slide) 2 HR. WOODS: Which wil1 be done, which we are g
3 recommending will be done as part of the severe accident 4 program, in particular the IPE program. The reason IPE here, 5 it says IPEE because there is a follow-up program to the 6 individual plant evaluation, which is IPE, which is individual 7 plant evaluation of external events, wnich comes later, so 8 here, on alternate two right here, that that's what this 0
9 y distinction means that was pointed out by the CRGR when we d
10 J took this to them about a couple of weeks ago. They said 11 truth in packagir.g would require us to say that the initial 12 d part of the IPE will not do all o' what we think needs to be l 13 h done. We have to wait for a follow-on part of the IPE which 14 1 is the part that will do the external events. So I am not I
15 1 quilty of non-truth in packaging again, t've told you that. {
i 16 These are the alternatives. I don't--do you want ce 17 to--I don't see any need in reading them, j 15 CHAIRMAN WARD: No. I think we have seen them ;
19 before.
I I
20 HR. WOODS: I can tell you what page in the reg 21 analysis these things are discussed if you want.
22 CHAIRMAN WARD: That's all right. l 23 HR. WOODS: Okay. Let's cove on to the technical 1
24 findings on the next slide.
25 (Slide) l l
HERITAGE REFORTING CORFORATION -- (202)628-4888
25 1 DR. KERR: Did the statement you made about external h 2 evunts imply that if A-45 is subsumed in the IPE program, that 3 in the first go-around it would not consider external events, 4 A-45 wouldn't?
5 HR. WOODS: A-45 won't exist anymore. The concerns
}
6 that were being pursued by A-45 will be being pursued in the 7 ! assignment, and it will not consider that type of concern.
8 j DR. KERR: The treatment of decay heat removal will a
l 9 1 occur in two phases, one of which is IPE which will consider l 10 the event, one of which will not?
i l 11 HR. WOODS: That is correct. ,
l 12 i CHAIRMAN WARD: Wait a minute. Decay breat removal l 13 isn't going to be considered separately anymore. It is just 14 going to become, I mean-- l 15 L HR. HINNERS: It is going to be part of the overal),
16 ! CHAIRMAN WARD: Part of the IPE: when they say A-45 l l
17 is subsumed. I am know sure I know what subsumed means. ;
18 Dh. KERR: I'm act, either.
19 CHAIRMAN WARD: But IPE is, he is just taking over 20 A-45. There isn't any specific A-45 subset of IPE.
21 ,
DR. KERR: The fact that it has a particular risk i
22 ,
will be unimportant anymore? We would loon at total plant t
23 risk and not worry about that fraction of decay heat removal? !
24 HR. HINNERS: That's right.
25 HR. WOODS: Subsumed means to take up into or HERITAGE REPORTING CORPORATION -- (202)628-4883
26 1 include in a large or higher class. Nobody else knew wLat it g 2 meant, either, so I looked it up. Okay, 3 Technical findings--we looked through the six case 4 studies, and if you wanted to characterize the studies with a 5 single number as they were perforned we have silice had sor.e 6 discussions with the indusery and in a later slide I will tell 7 you the modifications to this, but the case studies as they 9 hwereperforcedindicate that the decay 4 removal function 4
9 h failure core damage frequency is in ths sange of two to three 10 times ten to the minus four per reactor year, including 11 ;
internal and external causes, excluding the sabotage as an 1
12 external cause.
I l 13 It seemed that support systen failures like 14 ,
energancy power, component cooling water, et cetera, 15 contribute greatly to that core danage frequency.
16 Incidentally, PCM DHR, this term here, that appears throughout 17 all the stuff, and obviously it was originally meant to stand 18 for core melt, but we now really mean core damage. Core nelt 19 is a more severe event, somewhat less probable, j i
20 DR. KERR: Core danage means? l 21 MR. WOODS: Core damage reans basically that the ;
22 core stays within the primary system. That core relt-- ;
23 MR. MINNERS: It is the usual PRA assurption, loss -
24 of cool'.ng. That's all the FRAs calculate, j l
W 25 MR. WOODS: This reans you are not cooling the core i i
I
}
HERITAGE REPORTING CORPORATION -- (2021625-48d8
27 1 sufficiently, but it doesn't damage itself so badly that it 2 gets outside the primary system. It doesn't fail the primary 3 system. Core melt is what we mean by it is on the floor.
4 DR. KERR: And at what point does one enter the core 5 damage state?
6 HR. WOODS: Whenever you--I guess the way I would 7 say it is if yot; think that it is severe enough to the core, 8 what has happened, that you can't conclude you should start it 9 back up ano operate it, maybe it has been damaged. Maybe you 10 failed fuel.
11 DR. KERE: Anywhere from a little damage to complete 12 molten core that still is in the vessel?
13 HR. WOODS: Yes. That's my understanding of it. We 14 have several concerns with redt.ndir.cy, sharing of systems, 15 particular to support level, overall general arrangement of 16 equipment like insufficient physical separation of redundant 17 safeguard trains.
18 DR. KERR: Would that lead you to perhaps propose a 19 change in regulations which would not permit sharing?
20 HR. WOODS: Well, what we are proposing to do is to 21 have that kind of concern looked at en an individual plant 22 basis, and the -lant, that the licensee himself would be the 23 next person thu: would make such a recommendation.
~
24 DR. KERR: I am talking about the regulations which 25 on which we now depend to ensule plant safety.
HERITAGE hEPORTING CORPORATION -- (202)628-4888
28 1 HR. HINNERS: That couldn't solve the problem. That g 2 would solve part of the problem.
3 DR. KERR: In your view, that is a significant risk 4 contributor?
5 HR. MINNERS: On some plants.
6 DR. KERR: You mean that sharing in some plants is 7 okay and it is not in othes?
8 HR. MINNERS: It wasn't the only contributor.
9 DR. KERR: But it is always a significant 10 contributor?
11 HR. HINNERS: Never say always.
12 HR. WOODS: I didn't say always. We did six case l 13 studies. We found samples of places where sharing is a 14 concern. I couldn't say that sharing is a concern on all 15 plants. There is various degrees of sharing. I can't make 16 the general conclusion you are asking, but certainly it is 17 anticipated that some plants when they look at their system 18 they might find places where they think there is too much 19 sharing.
20 DR. KERR: It is not a problem that can be treated 21 by regulation.
22 HR. WOODS: At the present time; a direct answer to 23 your question, et the present time, A-45 is certainly not 24 recommending a regulation change in that area.
25 DR. KERR: Okay.
HERITAGE REPORTING CORPORATION -- (202)628-4888
29 1 HR. WOODS: Whether or not it happens in the future 2 as a result of the IPE, we will have to see. I don't know.
g 3 HR. MINNERS: There is--some of the concerns go l 4 beyond the numbers. This isn't directly in all of the numbers 5 that we are doing, but people looked at the plants. There is l
l 6 a lot of sharing closeness and interaction possibilities and l
l 7 that was a kind of qualitative thing which said gee, if you l
i 8 have a dedicated system that would take care of that problem, l 9 then you wouldn't have to worry about it, so your question 10 before whether it was purely on PRA numbers, I say ne.
11 Qualitative concerns such as this were also ec sidered in 12 deciding whether to have a dedicated system or not.
l 13 HR. WOODS: I think the main conclusions on this 14 slide, these last two, that everything is plant-specific, but 15 the vulnerabilities and the effects of the corrective actions, 16 so we were not able to say if every plant does this, 17 everything will be fine. We just simply weren't able to do 18 that.
19 (Slide) 20 CHAIRMAN WARD: I get the impression from something 21 you said or maybe Warren said, that while looking at this 22 sample of six plants, you oidn't get any indications, I mean 23 there wasn't anything that fell out indicating that this 24 particular regulation is a problem that needs to be changed, l
25 but you seem to have some thought that after the IPE, when I HERITAGE REPORTING CORPORATION -- (202)628-4888
30 1 these, a bigger sample of plants is looked at, that there may 2 be some indications of need for changes in the regulations h
3 that come out of that.
4 MR. WOODS: As I stand here, I'm not really prepared 5 to say that, but I certaily don't want to preclude some 6 licensee finding a, you know, a vulnerability that might 7 indicate a regulation change. I don't know what tney are 8 going to find. I am unable to say, of course.
9 CHAIRMAN WARD: Okay.
10 HR. WOODS: The previous slide showed you six 11 alternatives that we looked at. We looked at these 12 alternatives in the value impact way using three different l 13 methods which are shown here. Method A just looks at 14 population as the value of the fix, and the cost of 15 implementing the fix is the impact, so the value impact is the l
16 ' ratio of those two that come out in manrem per dollar I 17 suppose.
18 And the second method is basically the same, except 19 it also let's you take into account the, in the way stated 20 here, the off-site benefits and costs. I'm sorry--the 21 on-site, on site-benefits and, and those are the ones that we 22 can show you quite specifically, and hardened numbers 1
23 quantitatively.
24 Method C, we went somewhat beyond the standard g
25 methods and looked at the benefits to be gained from avoiding ,
HERITAGE REPORTING CORPORATION -- (202)628-4888
i 31 1 a moratorium if you were to have a severo accident and that sort of thing, so the results really--and I have got some l 2 3 back-up slides here that will give you the details of that.
4 Hopefully we will save time, not have to go to that, but 5 alternatives 2, 3 and 4 may be cost effective using method A 6 on certain plants and method A is the standard method.
7 Alternative 2 is the plant-specific evaluations.
8 Humber three is a, performing certain unresolved safety issue 9 and generic issue identified fixes, and alternative four is 10 feed and bleed and venting, just so you won't have just 11 numbers here.
12 Then, of course, going to method B where you take l 13 into account if the on-site benefits and cost savings would be 14 somewhat more cost effective.
15 Going on to the next slide, however, that's the big 16 however, those alternatives don't really get you down to the, 17 to the goal, the core damage frequency goal that the staff 19 would like to see that kind of thing reduced to. Using method 19 C which has taken into account moratoria and that sort of 20 thing, the dedicated decay heat removal systems might be cost 21 effective, but that's going considerably beyond the normal 22 methods that the staff has used in the past.
23 DR. KERR: In the context of the way in which you 24 are recommending that this be handled, this part of IPE, does 25 this have the significance it had before you made that
{
HERITAGE REPORTING CORPORATION -- (202)62e-4888
32 1 decision? Because if you now look only at the total plant 2 risk, rather than the risk contributed by decay heat removal, h
3 I am trying to put this in the context of that approach.
4 MR. WOODS: What these six case studies become are 5 basically examples of the way we would hope that the plants 6 would handle this kind of concern when they are doing their 7 IPE, and nothing more.
f 8 DR. KERR: Okay. Thank you.
9 MR. WOODS: Okay. Moving on, the staff of course 10 endorse--
11 CHAIRMAN WARD: The first one there. you are saying 12 alternatives 2, 3 a. 4 don't meet the staff's CEF goal.
l 13 MR. WOODS: Don't get enough risk reductions from 14 those alternatives to get you down to one times ten to the 15 minus five.
16 CHAIRMAN WARD: Okay, but in home cases you do get 17 enough risk reduction to be justified by either A or B. What g 18 that means is your ten to the minus 5th isn't justified by A 19 or B as cost effective.
20 HR. WOODS: If you set up a coal and say that's 21 where you would really like to see the risk, what it says to 22 me, I wasn't able to find a cost effective way to get to that 23 I goal. However, I was able to find cost-effective fixes, 24 meaning that basically the manrem that I get per dollar is in g
25 the range where you are allowed to spend that kind of money, HERITAGE REPORTING CORPORA?.' ION -- (202)628-4888
l l
33 1 but it doesn't, if I get, if I get a thousand manrem or person l h 2 rem reduction, it only costs me a dollar, then I can justify 3 it. However, it doesn't get me down to the ten to the minus 4 five.
5 DR. KERR: In your view, you can justify that even 6 if a plant already met the safety goals or the class of plants 7 already met the safety goals?
8 HR. WOODS: I don't know that I have any view on 9 that.
10 HR. MINNERS: What is your question again?
11 DR. KERR: Safety goals presumably set some 12 standard, and I am asking--
l 13 HR. MINNERS: These plants probably all meet the 14 same standard.
l 15 DR. KERR: I don't know. Let's suppose they do.
16 HR. WOODS: Are you talking about my ten to the 17 minus five, or the Commission's?
18 DR, KERR: The Commission's, of which you are a 19 part.
20 HR. HINNERS: We look at this question, and the 21 plants meet the safety goal.
22 DR. KERR: My question is in your view, should 23 cost / benefit analysis justify fixes even for plants that meet 24 the safety goal? Is that the current interpretation?
25 HR. MINNERS: That's the way I read the backfit HERITAGE REPORTING CORPORATION -- (202)628-4888 }
34 1 rule. 1 2 DR. KERR: Okay.
3 HR. WOODS: I am trying to decide where I am here.
4 Well, okay. I will go to the middle of the slide.
5 CHAIRMAll WARD: You are the staff endorses 6 alternative 2.
7 HR. WOODS: I have said that, so I am going on right 8 below that. Okay.
9 We had added a couple of items here just to flag 10 things that we are not recommending. Alternative 4, which is 11 feed and bleed, you can get a fair amount of risk reduction by 12 giving full credit for feed and bleed as an alternate cooling 13 mode.
14 The problem is it has an awful lot of reliance on 15 the operator knowing that he should initiate it and initiating 16 it in a timely way. It has to be done early, and there is a 17 disagreement between the industry reps and the staff on how 18 much reliance you should be able to credit the operator. They 19 want to give a lot more credit than we are willing to give, 20 and so we don't believe we get all that much safety benefit 21 from it. And therefore, we are not recommending it as tre 22 A-45 fix.
23 And as I have already said, alternative 5 and 6 24 which are the dedicated decay heat removal system wers not 25 cost effective unless you take into account such things as HERITAGE REPORTIllG CORPORATIOli -- (202)628-4888 ;
35 1 moratoria which we aren't prepared to do at this point.
l 2 So the plant-specific resolution down the bottom of 3 the slide is based on the fact chat the risk contributors are 4 plant-specific. The only way to identify that kind of risk 5 contributor is to do a plant-specific examination. The 6 effective corrective actions also is plant-specific, and all 7 that points toward the fact that you need an individual plant 8 evaluation before you can decide how or if you need to fix 9 that plant.
10 And since the IPE is going to do exactly that in the 11 broader sense, we are proposing to integrate these things with 12 the IPE.
l 13 DR. KERR: Let me make sure I understand. I think 14 Mr. Minners said that in your view, all of the plants examined 15 meet the safety goal. So when you talk about fixes, you are 16 talking about fixes that reduce the risk beyond that necessary 17 to meet the safety goal?
18 HR. WOODS: Now you have to define--when you were 19 discussing that before, I asked, and I never was sure that the 20 safety goal--the only official safety goci I am awsre of is 21 ten to the minus six for large releases.
22 DR. KERR: I thought I was quoting Mr. Minners.
23 HR. HINNERS: The health objective.
24 HR. WOODS: Yes, it meets that goal.
25 HR. HINNERS: It meets the health objective.
HERITAGE REPORTING CORPORATION -- (202)628-4888
~
36 1 HR. WOODS: It meets that by a long sight. The goal g 2 of 10 percent of cancers and whatever, we are way below that.
3 DR. KERR: So the suggestions you are making are 4 that the fixes be performed which will take the plant beyond 5 the safety goal criteria?
6 HR. MINNERS: They already are beyond the safety 7 goal criteria.
8 HR. WOODS: That is correct. I'm sorry. We did 9 consider that. I am too attuned to thinking in terms of the 10 core damage frequency and the health safety goals that 11 correctly point that out.
12 CHAIRMAN WARD: And as a justification for that, you 13 referred to the backfit rule, but if I look at your l
14 cost / benefit methods, and are both A and B, the, methods of A 15 and B consistent with what the backfit rule w>uld, would 16 suggest as appropriate?
17 HR. MINNERS: I think so. I'm not sure that the 18 Commission has spoken, but I think there is enough.
19 CHAIRHAN WARD: Probably clearly A is. B is 20 questionable.
21 HR. WOODS: B is your question point. It seems that 22 it certainly is part of the real world. I mean the on-site 23 benefits and costs, cost savings ire real, and we would g
24 propose to use them. Whether or not the Commission would 25 accept them, you know, that's up to the Commission to say. I HERITAGE REPORTING CORPORATION -- (202)628-4888
37 1 can't stand here and tell you. I '.hink I am on slide 7.
ll 2 CHAIRHAN WARD: Yep.
3 HR. WOODS: O! cay. Pete, did you have something?
4 HR. DAVIS: I'm sorry. I came in late, but I had a 5 question related to what just came up here. One of the 6 conclusions in NUREG 1292--
7 HR. HINNERS: What is 129, Pete? I haven't 8 memorized--
9 HR. WOODS: That's something we gave him.
10 HR. DAVIS: Do you know what it is?
11 HR. WOODS: Is it the reg analysis or summary 12 report? The summary, okay.
l 13 HR. DAVIS: On page 14-17, item 16, one of your 14 conclusions is that your study confirms one of the original 15 assumptions of the program which is that there are decay heat 16 removal vulnerabilities at existing plants.
17 Now if you are far below the health effects part of 18 the safety goal, you are basing your conclusion that 19 vulnerabilities exist on what? The core melt probability?
20 HR. WOODS: The core damage frequency.
21 HR. DAVIS: And why do you think that is a 22 vulnerability?
23 HR. WOODS: Can I get to slide 107 24 HR. DAVIS: Sure.
25 HR. WOODS: Okay. Let me--do you want to do that HERITAGE REPORTING CORPORATION -- (202)628-4888
38 1 now?
2 CHAIRMAN WARD: It will be all right.
g 3 HR. WOODS: I think I have covered slide 7 except 4 for maybe the--I have basically covered it one time or 5 another, so let's go on. I have said all that. That always 6 happens. You never know what order you are going to cover 7 things.
8 (Slide) 9 HR. WOODS: When you do one of these.
10 :
DR. KERR: I sympathize--
1 11 ! HR. WOODS: Well, it's part of the game. I guess 4
12 j what I am focusing on now is the fact that there was another, l 13 there was a reanalysis done by the industry, by NUMARC. I 14 quess it was sponsored by NUMARC, performed by Westinghouse 15 Owners Group. It was basically an industry sponsored 16 reanalysis of the Point Beach plant. The NRC reviewed it for 17 two reasons. I have on this s'ide--first of all, if it were, 18 if the NUMARC results were correct, if the PRA results were 19 correct, and if we could conclude that Point Beach is a 20 bounding plant, then we would be reaching an entirely 21 different conclusion here.
22 Our conclusion is individual plant evaluations are 23 needed. If the NUMARC results were correct, we would have to 24 conclude I believe tisat they meet the goal that we set up and W 25 therefore you don't need to look at it any further.
HERITAGE REPORTING CORPORATICN -- (202)628-4888
39 1 However--so we need to discuss whether or not we think the 2 NUMARC result is correct, and the staff does not think it is g
3 correct.
4 We need to have a discussion of why we don't think 5 it is correct and indeed the next two speakers will be giving 6 you more details of that. I have a summary here. The second 7 reason we thought that--
8 DR. KERR: Will they be telling us what they mean by 9 correct?
l 10 HR. WOODS: I'm sure < hey will. I would if I were l
- 11 giving it. And we think that in some sense, this is, this l
12 industry analysis approach is part of an IPE evaluation. I l
13 quess that's redundant--an IPE submittal, and we thought it 14 , would be extremely useful for us to review it and give some i
15 preview as to what type of review, what kind of opinions we i
16 would have of methods and assumptions and whatever that may be l
17 used by the industry in doing these evaluations, so for that l
lb reason, we also performed this review.
19 I think one of my more interesting slides--certeinly 20 some of them I have already had weren't very interesting, but 21 I think this one is pretty interesting. This is hopefully not 22 stealing too much of the thunder of the next speakers, but 23 this is the numerical results, the left-hand column. This 24 thing is not straight.
25 (Slide)
. HERITAGE RFPORTING CORPORATION -- (202)628-4888 L
\
40 1 MR. WOODS: This left-hand column is, well, this is 2 the different kinds of things that were evaluated. This is 3 the results of the one case study that the NRC did, one of the 4 six on Point Beach. So the internal, the internal failure 5 type core damage frequency is this, and these others, I'm not 6 going to read them ;o you, but they add up to--
7 DR. KERR: That's just the contribution of decay 8 heat removal?
9 HR. WOODS: That is correct. This is the decay heat 10 removal failure related frequency, and only operating type 11 events. It doesn't include cold shutdown events like when the 12 'evel gets down too low during cold shutdown operations.
13 That's a different issue. It is not included in here, so it 14 is only that kind of risk, and what we came up with 3.1 times 15 ten to the minus four, which is a factor of 31 above the goal 16 that we set up of ten to the ninus fivo, one times ten to the 17 i minus five.
18 The industry study--in fact, that's what this 19 number, well, sorry. The industry, EPRI, Westinghouse Owners 20 Group, sponsored a study, got these results instead, and the 21 number in parentheses here is simply our number divided by 22 their number. Reduction factor; and they get a reduction 23 factor of 31 and they came outright at our goal, one times ten 24 to the minus five.
W 25 We had several discussions with them. In fact, we l
l HERITAGE REPORTING CORPORATION -- (202)628-4888
41 1 had two meetings, public meetings. You have a rather lengthy h 2 package of their summary of that meeting, including their 3 document and whatever, and as a result of the meetings for 4 that specific plant, we were able to agree largely with 5 certain, certain of the things that they had. A lot of the 6 type of things we were able to agree with were okay. We have 7 now, we, the industry, have now looked specifically at this 8 very plant, and the plant-specific value here because we have, 9 these systems that you didn't credit us with or whatever, is 10 different from what you assumed. For that sort of reason, and 11 I can give you some more details of that, they are going te 12 give you great details, after we finished listening to them l
l 13 and reviewing their reports and whatever, if we were to redo 14 the analysis today, wo think we would come out with numbers 15 like this, so we onded up with a factor of whatever it is.
16 CHAIRMAN WARD: Three and a half.
17 MR. WOODS: Three and a half reduction; I think it 18 is 3.44, but it is still above the goal. It is still above 19 one times ten to the minus five that we think we ought to 1
20 reach. So the significance of this slide is that we still i
21 ! think even with, with all the different things we can accept l
22 l plant-specific wise, that it is above the, above where it 23 ; should be. They believe otherwise and they will tell you i
24 i that.
I 25 { DR. KEIT: There you have two different groups, both HERITAGE REPORTING COT.PORATION -- (202)628-4888
42 1 of whom are presumably competent and honest, ccme out with 2 numbers even after significant negotiation in order of g
3 magni ttid e .
4 HR. WOODS: That is correct, and a lot of difference 5 as I went through it, a lot of difference seemed to be the 6 credit you are willing to give for operator corrective 7 actions, and that's a subjective thing. I mean--
8 DR. KERR: I'm sure there is a reason for this 9 because there have to be, but I am saying that thess are two 10 groups, both of whom involved competent, experienced people in 11 PRA.
12 HR. WOODS: That's the very value of the thing, l 13 because what it does, it keeps you from just arguing well, I 14 think it is safe enough; well, I don't. It let's you aim at 15 the specifico. I think this is too much credit for corrective 16 actions. He doesn't. And that gets you right at the issue.
17 DR. KERR: It also tells me that if you take 18 seriously the fact that those people are competent and honest, 19 and I do--
20 HR. WOODS: I do, too.
21 DR. K3RR: That one simply has to be a little bit 22 leery of these numbers if you try to interpret them closer 23 than about an order of magni.tude.
24 HR. WOOD 3- I agree; no problem.
I 25 l HR. MINNERS: You keep saying that and he keeps HERITAGE REPORTING CORPORATION -- (202)628-4880
43 1 agreeing with you, but I don't. There is a real difference 2 between the numbers, okay.
3 DR. KERR: That is precisely the point I am making.
4 There is a very real difference even after people with 5 experience, competence and integrity, look at the m, the same 6 system.
7 H P. . MINNERS: Unfortunately people who are honest 8 and have integrity and all those good qualities sometimes come 9 out with the wrong answer. One of those answers is right.
10 DR. KERR: No. Both of them may be wrong. That's 11 the point I am trying to make. There is, it seems to me there 12 is an inherent uncertainty in the methodology.
13 l MP. MINNERS: I think, I think that those estimates 14 are probably somewhat around what the right answer is. I 15 agree with you. But the point I am trying to make is that 16 somewhere in there, there is a correct answer. When I listen 17 j to you, 1 get the implication, I get the inference that you i
I 18 know, any of these answers can be right.
19 The whole question is looking at the differences and 20 decide what is causing these differences between two people 21 who are struggling to find a right answer, and looking at that 22 say hey, where does the right answer probably lie? Not just 23 to say one guy did it this way, the other guy did it this way, l
24 ' so there is come, I think some of them are r i e:t t , and this 25 illustrates the question hcw much credit are you going to give l{ERITAGE REPORTING CORPORATION -- (202)628-4888
44 1 to operator recovery action? Okay. That's what this study g 2 reanalysis has done, okay, and I think people should very 3 carefully consider that and just don't say well, they have got 4 two right answers or one is as good as the other. That's not 5 the case. If there is a right answer, I don't know what the 6 right answer is. Okay. My guess is I think they shouldn't 7 give that much credit for operator action.
8 DR. KERR: Mr. Minners, I am not drawing my 9 conclusions based on just this one ctudy.
10 HR. HINNERS: I think they are all that way. There 11 are differences in all the studies, and I think you have to 12 look at differences. That's the interesting part of PRA, what l 13 the differences are, and those are the technical things you 14 ought to argue about, not justice mix them as two people 15 getting two numbers.
16 HR. WOODS: I didn't expect to have a religious 17 discussion here, but this--
18 CHAIRHAN WARD: That's what it is. Pete, do you 19 have any--
20 HR. DAVIS: It is I think important to recognize 21 that we don't know what the right answer is and we will never I
22 know what that answer is, and in fact, it will vary over the 23 lifetime of the plant and probably varies day to day.
24 l HR. WOODS: There is an absolute answer, but man is l
I 25 incapable of knowing it. -
HERITAGE REPORTING CORPORATION -- (202)628-4888
45 1 HR. DAVIS: I thi:4k the other thing that I found h 2 interest, something in your report, you said that your best 3 guess of the uncertaintxes is a factor of five to nine, which 4 I think would encompass the variation we see here.
5 HR. WOODS: That is probably not accidental.
6 HR. DAVIS: Yes. I want to say a little bit more 7 about that later.
8 MR. MINNERS: That is a theological discussion that 9 doesn't--and when you say there is a factor of five to nine, 10 that snything within that range is equally probable, I don't 11 thir.k this is that kind of distribution. I presume that there 12 is something in there which is a more likely number than the l 13 limits, okay? So even though we say the error da five to 14 nine, I would presume there is a reducing probability that 15 would get to the outer limits, that that's not the right 16 answer.
17 l HR. WOODS: Usual Gaussian curve picks up best 18 estimate and comes down hopefully symmetrically. Who knows?
19 CHAIRMAN WARD: Okay. Let's see. Roy, how much do 20 you have? You are about halfway through your slides.
21 HR. WOODS: I am basically through.
22 CHAIRHAN WARD: Is this your whole packet, six 23 slides?
g 24 HR. MINNERS: You want to do slide 10 for Pete?
25 HR. WOODS: I was going to have a more detailed HERITAGE REPORTING CORPORATION -- (202)628-4888
46 1 discussion of the ten to the minus five and why we chose it if 2 you would like, and then I'm prepared to quit.
g 3 CHAIRMAN WARD: But what I am trying--after you go 4 through this, are you prepared to sit down in favor of Mr.
5 Vine or--
6 HR. WOODS: Yes.
7 CHAIRMAM WARD: I just want to see whether we are on 8 schedule. We are doing fine. We have got, you have got until 9 ten o' clock.
10 HR. WOODS: Oh, okay. That's right. I had fifteen 11 minutes and fifteen minutes and it is all combined. I may sit 12 down earlier. Okay.
13 Lets--I can give you some more details on this slide 14 l now and as to where the differences are, and I guess the kind 15 i of thing we were able to agree with, they, based on the 16 plant-specific look, they were able to convince us that the 1
17 small break LOCA frequency was lower, the pipe break frequency 18 was lower than we had given this credit for in our study, and 19 that was worth about half of the, half of the change that we 20 made.
21 The plant-specific loss of off-site power challenge 22 frequency that was somewhat lower than we--we didn't give as 23 detailed a look at the data apparently for that site as they 24 [did, and that was worth a fair amount. I have the numbers, w 25 but that's hardly worth giving you the exact numbers. We gave i
HERITAGE REPORTING CORPORATION -- (202)628-4888
l 1
47 1 them quarter for obviously addition of a new station battery 2 that we, that either wasn't there or we didn't know about it 3 or whatever at the time when we did the study. We gave them 4 quarter to a second halon fire system and critical location.
5 There was factor of w1 assumed too much, too much pipe was in C a vulnerable location in the service water pump house and 7 therefore when wt looked at the smaller amount of pipe there 8 was a lower probability of it rupturing and filling out the 9 pump room so that's the kind of thing that they peinted out to 10 us in their detailed re-evaluation and we were able to accept 11 on review.
12 The kind of thing that we basically weren't willing 13 to accept was the much greater credit they wanted to take for 14 operator corrective actions.
15 DR. KERR: You got your numbers from consultants in 16 data, generic data or I mean--
17 ffR. WOODS: Our numbers were originally from one of 18 the case studies that, that Sandia did for the A-45 effort.
19 That's the left-hand column. The right-hand column is largely 20 from the name consultant helping us with the review. I think 21 it is, that consultant is Dave Ericson. lie will speak to you 22 later. That's where it came from, and on staff was 23 participating in that review and they agree with these 24 fnumbers.
25 DR. KERR: That's enough. I was just curious.
i i
HERITAGE REPORTIllG CORPORATIO;l -- (202)628-4888
48 1 HR. WOODS: That's where it came, from the staff and g 2 their consultants.
3 okay.
4 DR. KERR: Operators may have gotten a lot more 5 co:npetent since Ericson did his study, but you think probably 6 n. t .
7 HR. WOODS: Since Ericson--
8 DR. KERR: I am being--
9 CHAIRMAN WARD: Never mind.
10 , HR. WOODS: Okay. Well, the last slide I have is in 11 response to Mr. Davis' question. I put this together because 12 there has been a fair amount of discussion about where the ten l 13 to the minus five comes from, and my own feeling is if you are 14 going to do PRA study, you are going to end up with numbers, 15 and a logical question to be asked is okay, you got this 16 number, is it low enough or not? If I say gee, I think it is 17 not low enough, a logical question is what is low enough? So l
18 I have attempted to answer that, and that's what this slide l
19 tells you.
l l
20 HR. HINNERS: You want to switch your slido?
l l HR. WOODS: They have all got it in front of them.
21 22 I will do that. Thank you. That's the first item is what I 23 just said. The staff is currently considering the using and 24 proposing the more general core damage frequency safety goal I
( l 25 [intheorder, in the range of ten to the minus four per 1
h h
itERITAGE REPORTING CORPORATION -- (202)628-4888
49 1 reactor year, and that is not just decay heat. That is total 2 core damage frequency, so we have seen several PRAs that 3 indicate to us that the core damage, failure related core 4 damage, the decay heat removal related core daruage frequency 5 is a third to a half of the total, so what you would need then 6 is a total core damage frequency in the range of ten to the 7 minus four divided by say three. and that comes out to this, 8 three times ten to the minus five.
9 flow that's total t ruly a total DHR related, DHR 10 failure related core damage frequency. If you are going to 11 achieve that. you have got to consider the fact that the 12 quantifiable, in other words, the result of the PRA aally 13 doesn't include everything. You have got come other errors, 14 ;
some acts of commission. I could probably name some other 15 things that you really don't know how to properly put into a 16 l PRA, and you have got to put a safety factor on it if you I I I
17 will, so I propose a safety factor of about three, and what 18 ]thatgetsme to is about one times ten to the minus five, h
19 p which in the number that ought to be applied as a goal to the 20 seeable result, the quantifiable result of the PRA of the type 21 that we have done. That's where the ten to the minus five 22 comes from. It is related to the ten to the minus four that t
23 '
the statf seems to be using more and more.
a 24 tiow in a footnote down at the bottom, I point out ll 25 that if you are going to use--well, the only, I made tne error I
I N
HERITAGE REPORTI!lG CORPORATIO!! -- (202)628-4838
50 i 1 before of saying the only safety goal was the ten to the minus 2 uix for large ralease. Of course it isn't. It also includes h
3 !' the latent cancers and early and latent deaths and whatever, 4 but the only one in terms of the core is indeed the large 5 re. case, and that's ten to the minus six. And this ten to the 6 minus five, well, this ten to the minus four which I have now 7 telated to my ten to the minus five, goes with the laroc, the 8 large release frequency of ten to the minus six. Probably if 9 you assume, for example, a one to ten ratio between core melt 10 and core damage and the one to ten ratio between large 11 releases and core melt, there you could apportion that l
l 12 differently if you want, but it seems to be if you wave your l 13 hands a bit, and it seems to be consistent with the ten to the 14 minus six large relecse frequency also, so--and it also is 15 consistent with what we have done on A-44 and A-49 as I 16 L already pointed out, so if you don't like it, propose 1
0 something else, but then I warn you you have got to say why it i
17 18 is logical, and that's what I have donc.
19 i So that's what we use for this purpose, and for this i
20 ;
purpose only. We are not really in a position to say hey, 21 ACRS, or Commission, or whatever, you really ought to use this 22 anywhere else. We think maybe they need a goal and that is l
23 ! what we used for this, but we are not proposing it for the d
24 h IPE.
4 l\
25 CHAIRMAN WARD: In this, is your thinking here i
HERITAGE REPORTING CORPORATION -- (202)628-4888
51 1 consistent with what the people who are working on the safety g 2 goal implementation plant are doing? Or I mean this is one to 3 ten ratios, so forth. Is that--
4 HR. WOODS: It is one way you can get there I 5 believe.
6 CllAIRMAN WARD: I realize that.
7 HR. MINNERS: I talked to Wayne Houston who is 8 l working up the paper and it is generally consistent. I 9 haven't seen the latest paper. It is in the ballpark.
10 HR. WOODS: The ten to the minus four is consistent 11 with what they are working up and exactly the details of how i
12 l' you subdivide it might be a little different.
l l 13 CHAIRMAN WARD: What I am trying to find out, you 14 g would be a lot better off if Wayne Houston's work were o
15 11 j finished and on the street.
16 d{ HR. WOOD 3: Amen!
17 CHAIRMAN WARD: It is not, so are you discussing f
18 this with him or whoever is working?
19 HR. WOODS: Sure. lie is aware of this, yes. We 4
20 just felt we needed something now because we are doing this I
21 lnow, and that's what we have set up.
ll 22 HR. HINNERS: We agree this is fine, l
i 23 h HR. WOODS: I am prepared to sit down.
o 24 i HR. DAVIS: I think thic is a pretty good logic 25 H structure, and I understand how you arrived at the numbers. I l\
llERITAGE REPORTING CORPORATION -- (202)628-4888
52 1 think, though, that from whac I have seen, item 3, DHR failure 2 related to--is higher than one-third to one half because the 3 only thing you are leaving out are ATWS and large break LOCA 4 contributors.
5 HR. WOODS: Pressurized thermal shock and shutdown, 6 contributors to shutdown failure.
7 HR. Hill!IERS: And event V.
8 HR. DAVIS: The total of all those would be less 9 than the credits you are civing them here I think.
10 HR. WOODS: That may be. I don't think--I think it 11 is, I trink it is our--
12 HR. DAVIS: Especially if the primary liability 13 initiatives are implemented, which I understand they are, and 14 also the other initiatives related to the Q sequence are going 15 to push those way down I would think. I agree there is room 16 to argue that.
17 DR. KERR: Interestingly enough, you seem willing to 18 l make estimates of the rest of the plant's behavior in this 19 area on the basis of these five or six analyses, namely, that 20 a third to a half of the overall CDP.
21 HR. WOODS: Actually that is not where it came from.
22 ; It came from other PRAs that do include the entire statement b
23 of risk. I would have to look up what they are, but it didn't i In tact, it didn't come from 24 cc from the six case studies.
1 l 25 j that because we didn't, didn't do the other type of risk in h
it b
HERITAGE REPORTIliG CORPORATIOli -- (202)628-4888
53 1 these six case studies, g 2 DR. KERR*. But the other studies certainly haven't 3 been done the way these case studies were done.
4 HR. WOODS: That is correct.
5 DR. KERR: So I could--
6 HR. WOODS: It is an estimate. It is somewhere in 7 that ballpark. Thee is probably more than a tenth and 8 probably not all of it. Again--
9 HR. HINNERS: There is a lot of variability between 10 plants.
11 DR. KERR: Again, I don't disagree with some sort of i
12 structure of this sort as a yardstick. I'm not quite sure l 13 what would be useful.
14 HR. MINNERS: The reason that we went through this p
i 15 l rather quaint analysis is because the question was asked I 16 think fairly by industry is thera a problem? And that's the i
17 question that we are trying to answer. We said yes, compared i
18 l to a goal of ten to the minus five, there is a problem. Now l
19 4 people think the goal is ten to the minus four, there isn't a l
20 l problem, so it depends on what your standsrd is. So that's 0
21 through this, and I think it is--a fair question hl why we went 22 P and I think a central one that the Agency and the industry is 23 trying to decide is hey, what goal should we be striving for?
U 24 h It really hasn't been settled. And you know, ATWS was a b
25 perfect case of that we went out with a goal of ten to the l
l HERITAGE REPORTING CORPORATION -- (202)628-4888
54 i
1 minus seven, okay, and we finally moved the goal up until we h 2 solved the problem.
3 DR. KERR: Now putting this in the IPE, putting it 4 in a situation in which one is looking for outliers in a 5 population of plants which at one point at least was judged by 6 the Commission and its supporting staff as being adequately
? safe except possibly for some things that might be found in 8 the course of individual plant examination, my impression is 9 that your conclusion would say probably every plant is likely 10 to need some sort of fix in decay heat removal. That's not 11 the case?
12 HR. MINNERS: You can look at Peach Bottom. They l 13 have already done their IPE, and they are way down there, so I 14 don't see what they have to fix.
15 i DR. KERR: If you use, but you aren't going to use l
16 these criteria, though, the ten to the minus five. Is that a
17 ! what you are telling me?
I 18 j HR. WOODS: We are not. We can't tell the IPE h
19 '
people to use that criteria.
20 DR. KERR: Now wait a minute. I still believe we 21 are all part of the same aroup, the Nuclear Regulatory 22 Commission, and at some point it needs to have a policy on how 23 J it is judging plant safety. Maybe you can't tell the IPE 24 !j group, but you have sute influence within the Connission on i
25 l how we do analyses, and if the regulations are intdequate, l'
d HERITAGE R3 PORTING CORPORATION -- (202)628-4888
55 1 then we have got to use souething else, and it can't just be h 2 used by the pecp te lii A-45. I hope it has got to be used by 3 the people who are determining whether the plants are safe 4 enough or not.
5 CHAIRMA!! WARD: I think it is a good point. It 6 seems to me if you have decided that this is a reasonable 7 goal, your ten to the minus fifth let's say, reasonable and l 8 appropriate, and if you are turning over the A-45 issue to IPE 9 to do that in good conscience, you have to be assuming that 10 the IPE process is going to have a goal that somehow is 11 consistent with--
12 HR. WOODS: In good conscience you have to make 13 that.
14 HR. MINNERS: Conscience doesn't have anything to do 15 with it. It is a matter of what direction the Commission has 16 given to the staff. The Commission has only directed the I
17 hstaff at this point to require licensees to do an IPE. It i
18 I hasn't directed the staff to require licensees to fix anything I
19 or to meet any kind of a goal, and so that's the situation we 20 l are under, so you know, it is not a matter of conscience. It f
21 is a matter of what direction the Commission has given its 22
'li l staff.
23 HR. WOODS: I havt felt obligated to do at least 24 this and tell you about it and tell the users about it, but I i
25 can't make the Commission adopt it.
HERITAGE REPORTI?'G CORPORATION -- (202)628-4888
56 1 CHAIRMAN WARD: The IPE program isn't exactly in 2 good shape. I mean--
3 HR. MINNERS: But as a practical natter, okay, if 4 you have a core melt frequency less than ten to the minus six, 5 reduction of core melt frequency, it is unlikely that you are 6 going to get a value impact ratio that anyone is going to 7 accept. You are going to have to get above ten to the minus 8 five to get what is now considered to be a cost-effective 9 solution, so I think that we are a guing about what the goal 10 is isn't going to make any differenc'.. It is the value impact 11 ratio that counts. The goal is almost irrelevant.
12 CHAIRHAN WARD: I am not sure that, I'm not sure 13 that that's right because you are really talking about two 14 !different sorts of goals. One is kind of an absolute and 15 arbitrary goal which derives from the safety goal, and is the 16 ten to the minus fifth, and the other is the, is the 17 cost / benefit argument which is really independent of the ten 18 to the minus fifth.
19 HR. HINNERS: But you have control.
20 HR. WARD: Well, maybe, yes, you say that. I m.an l
21 it may be.
22 HR. HINNERS: That is my experience, and you know, 23 it isn't alway =. I mean I am telling you as a generality.
0 24 l HR. DAVIS: I have one final question, Mr. Chairman, 25 h if I could?
I HERITAGE REPORTING CORPORATION -- (202)628-4888
57 1 CHAIRMAN WARD: Sure.
HR. DAVI*: I hate ;o get into the number game.
l 2 I 3 will do it one more time.
4 One of the other conclusions that went along with 5 this one which concluded that existing p 4nts have 6 vulnerabilities was that these vulnerabilities could be 7 countered, and I had a little bit of a problem with that 8 because if I read this correctly, the modifications that you 9 were examining, if you implemented all of them. you would 10 reduce the core melt probability by a factor of two to six.
11 Now that would be insufficient to bring Point Beach, 12 for example, even with your revised r. umber of nine times ten l 13 to the minus five down--one times ten to th1 minus five, and 14 furthermore, these reductions are within your uncertainty 15 factors so I'm not sure that you can conclude that these 16 v6 .tages can be countered with the modifications you exsmined.
17 i Am I missing something in that logic?
18 '
HR. WOODS: You are just correctly pointing out what 19 i I have said many times, that we were unable to come up with h
20 h the cost benefits for justifiable fix or fixes that would get 21 down to the goal that we thought it should be at.
32 HR. DAVIS: The cost cannot be countered by a 23 uodification to examine. It looks to me like it is going to j 24 push you to the add-on system if you really want to get to 25 ten, to ten to the minus fifth, l
HERITAGE REPORTING CORPORATION -- (202)628-4883
98 1 MR. MINNERS: We have said that.
h 2 MR. DAVIS: We can't be countered in the normal 3 sense of modifying the plant.
4 HR. WOODS: We have conflicting requirements. We l
5 have the cost / benefit requirements and we have the safety goal I
6 or our own goal that don't seem to be totally compatible with l
)
7 each other, 8 MR. MINNERS: But Pete, that depends to a large 9 degree how much credit you are willing to give operator 10 I recovery actions. If you give them more than we do, you can 11 get duwn to the ten to the minus five and you probably get 12 there any time you want to.
l 13 MR. WOODS: That's the discussion you are going to 14 hear for the rest of the morning.
15 MR. MINNERS: That's a subst.antial discussion. I l
16 don't know how yoa answer that question in any objective way 1
17 ; from data. It is going to have to be a subjective judgment by i
18 people. It is a serious one.
l') MR. DAVIS: I understand that and I agree with you.
20 My only complaint was that you concluded that these i I 21 I vulnerabilities can be countered.
22 MR. MINNERS: Particular vulnerabilities that you li 23 find can be countered. We can't get to the goal unless we 24 h rely a lct on other recovery action.
25 ll MR. WOODS: We didn't mean what you read ta conclude i
I HERITAGE REPORTING CORPORATION -- (202)628-4888
59 1 that we could reach the ten to the .minus five with the fixes 2 that you are looking at. That's rot what that was meant to 3 say. You can, if you found a vulnerability, you can shortly 4 fix it and a lot of vulnerabilities you can fix in a cost 5 beneficial way, but they don't get you down to the goal that 6 we would like to see.
7 MR. DAVIS: Thank you.
8 CHAIRMAN WARD: Anything else? Okay. Thank you, 9 Roy. Let's take a 15-minute break now, and we will come back 10 with Mr. Vine.
11 (A brief recess was taken.)
12 CHAIRMAN WARD: Our next speaker is Mr. Vine of 13 NUMARC.
11 I HR. VINE: Gerry is going to be starting off here.
15 CHAIRMAN w'.'.R D : Mr. Neils will.
l 16 (A discussion was held off the record.)
l 17 f HR. NEILS: What I would like to do is just give an Il 18 I introduction to what we would like to present, and a little 19 i bit of our overview perspective of where we are comir.g from.
I 20 (Slide) 21 .
"R. NEILS:
, We are speaking for NUMARC here, as the h
22 C NUMARC working group which I chair on decay heat removal. I 23 have been here before a long time ago on the same subject.
p d
~
24 '
Obviously we have the people that did the work on inside 1.13.
J 25 !ObviouslytheWestinghouseOwnersGroup is represented here i
HERITAGE REPORTING CORPORATION -- (202)628-4888
60 1 because they have a great interest in this, and sharing the 2 cost of doing things here, and we are talking about Wisconsin 3 Electric's plant, so they are well represented here.
4 (Slide) 5 HR. NEILS: I think the answer to some of the 6 questions that were a little difficult to address here might 7 be found in Mr. Denton's letter to AEOD, Hr. Heltemes, which 8 provided specific direction to the A-45 program, and we would 9 like to extract ono quote from that which says, "A-45 is 10 directed toward determining the need for improvements in 11 current decay heat removal systems. In particular, it is 12 ; addressing the need for a dedicated decay heat removal 13 function."
14 In essence, that program was driven to answer that 15 question--is the dedicated add-on decay heat removal system 16 ,
justifiable as an answer to the A-45 question?
d 17 d (Slide) 18 HR. NEILS: And in our opinion, the A-45 case 19 studies worked very hard to answer that particular question, l
i 20 but not some of the lesser questions.
21 In our review of al.1 the case studies, we find that i'
22 h all address whether the add-on dedicated system is justified, h
23 i! All address that specific question. All -cuclude that the All have wide margins in 24
[hadd-on systems are not justified.
25 I that conclusion, and in our judgment, all the case studies 4
h j
HERITAGE REPORTING CORPORATION -- (202)628-4888
61 1 were modeled very conservatively so that the conclusions in 2 effect are unimpeachable. The flSAC 113 work, rework to the 3 Point Beach study was to quantify and illustrate the 4 conservatism in one of those studies, and we believe that same 5 deg:Je of conservatism from our review resides in all of the 6 case utudies.
7 DR. KERR: I have hope that you r. ave not concluded 8 from your experience of the nuclear power business that 9 conservatism ulways gives results that ate conclusions that 10 are unimpeachable.
11 HR. NEILS: 11 o . There are some things that are 12 l wrong, too.
13 DR. KERR: And indeed frequently. It is not unusual 14 that conservatisms are introduced in studies. I read the 15 conclusion maybe unimpeachable, but I am not impressed by 16 ftherebeingunimpeachablebecausetheyweremodeled i
17 conservatively.
l 18 ;
hR. NEILS: Fell, you have got to start to believe 19 in something, okay. And if there are wide margins in that 20 answer, and margins of conservatisms in the way the answers 21 l were derived, then I would be inclined to accept that 22 conclusion.
l 23 ji (Slide)
P l 24 l! HR. NEILS: Now looking at the case studies, they i I l
25 l really are a modified IPE, modified to the extent that they i
I HERITAL REPORTING CORPOFATION -- (202)628-4888
63 1 don't include external at all, sabotage; they don't include 2 large break LOCA and they don't include ATWS, but they do in g
3 our judgment include a heck of a lot more than one-third of 4 the total risk. That finally got the, caught your attention I 5 think between myself and my secretary--maybe it was a Freudian 6 slip, but I had to bring it with me because I had heard Dave 7 refer to the subject delayed heat removal, and that the DHR 8 people had to leave it in there for the time being, but we do 9 believe that the, that the A-45 is not an issue to be thrown 10 away and disregarded. We do believe that it can be properly 11 handled under the severe accident policy with IPE and/or PRA.
i 12 l We do believe that with that approach, the potential l 13 weaknesses car. be identified on a plant unique basis, and once 14 identified, can be addressed. From the case studies, I think 15 we agree with the staff that the studies illustrate that they 16 are short of the heroic add-on systems. There appear not to 17 ' be generic fixes that can be universally applied to to every 18 plant. To keep from complicating the DHR issue with the what 19 its of other issues, just about all the influences on DHR 20 vulnerability are the other influences in terms of loss of AC 21 power, fire, seismicity issues, et cetera, and these are all I
22 o being specifically addressed by other programs, too, and I b
23 think the value of the 1PE under the severe accident program L
24 [isapotentialplant unique identification of relatively low C
25 i cost, high risk reduction value items, b
i HERITAGE REPORTING CORPORATION -- (202)628-4888
63 1 DR. KERR: Let me take, see if I understand these 2 other influences. Let's take fire as an example. In what 3 sense is that being addressed by other programs?
4 HR. NEILS: By the Appendix R regulations, and the 5 Appendix R regulations are not universally completed in 6 compliance by all licensees at this time. It is one of the 7 difficulties, for example, presented to Sandia as to some 8 degreo a moving target.
9 DR. KERR: Are you convinced that Appendix R will 10 fix whatever vulnerabilities exist because o' fire?
11 HR. NEILS: To reduce them to zero, no, I um not 12 convinced it will do that, but I think when Appendix R--
13 DR. KERR: I didn't imply that they would be--yota 14 know, you point to the other programs, how they cover the 15 weaknesses, and I am saying do you think Appendix R will do 1 16 ll good job in reducing the fire risk to an appropriate level?
17 ll HR. NEIL3: Yes, il 18 DR. KERR: Because any impression is that there are ,
d 19 y a lot of people who think Appendix R is fairly highly flawed, i
I 20 HR. NEILS: Well, when Appendix R, when Appendix R l
21 lputsaplant in the condition where it can tolerate a fire, 1
22 that makes you assume everything in that fire zone is d
23 l! disabled, and still provides an alternative path to safe 24 shutdown, then I think it does a gceat deal to reduce fire k 25 h.j ri sk.
HERITAGE REPORTING CORPORATION -- (202)628-4888
64 1 DR. KERR: I think you know a lot more about this l 2 than I do because you are lot closer to plant. I am simply 3 asking if your Appendix R will take care of the fire problem?
4 HR. NEILS: Yes; reduce it to an acceptable risk 5 relative to the total risk.
6 DR. KERR: Thank you.
7 HR. NEILS: It botter because the industry will have 8 spent about--
9 DR. KERR: It doesn't make me feel good to say it 10 better because whether the industry spent a lot of money on 11 it, a facility, is a lousy rule. I don't know. I have heard 12 ltheopinionsexpressedthat it is not very good, but in your l 13 view, it is pretty good?
14 HR. NEILS: Yes. Having gone through compliance on 15 two plants, not one, but on plants that were commissioned in 16 I the early '70s, I think it has done a lot of good. I think in l
17 some cases we have spent monoy on things that don't do good, 18 but overall, ir effect I think it is a dramatic improvement.
19 HR. DAVIS: I have a related question, please.
20 HR. NEILS: If I might, you raised a comment back I
i 21 there at the consulting end of the table that the NSAC study l!
22 :j on Point Beach, you know, surely now I know where they got the l
23 ten to the minus five because that's the staff's goal. The 24 ,
staff's goal was invented to the ten to the minus five on DHR l
t 25 l risk quite some time after NSAC 113 was completed. If one b
I, HERITAGE REPORTING CORPORATION -- (202)628-4888
65 1 group relied on the other for the origin of the numbers, it 2 was the other way around, contrary to your suspicious.
3 CHAIRMAN WARD: Did you have a question?
4 HR. DAVIS: I wanted to ask a question on Appendix R 5 things.
6 Are you aware of any instance where implementation 7 of Appendix R may have been detrimental to risk? And I am 8 thinking about a case where putting up three-hour fire walls 9 around pumps would impose the need to have active room i
10 j ventilation to keep the pumps from overheating during decoy I
11 '
heat removal, or making--
l 12 i HR. NEILS: By saying some aspect of Appendix R had l
13 represented, when completed, represented an inconvenience to 14 be addressed fton time to time, but I don't think it has 15 introduced new safety problems.
16 l However, the process of implementing the 17 requirements of Appendix R may have incurred an interim risk l
18 4 during implementation because you are doing an awful lot of I
I 19 modification to the plant, a lot of rewiring, and you know, l
i 20 l you can fix, modify mechanical things, and like putting a cast 21 on a broken leg, but when you start doing major rewiring and 22 it gets you into the realm of brain surgery,
'lreconductoring, i
23 in my opinion, and I like for those things to be all done i
24 i instead of to be involved in them.
25 HR. DAVIS: Thank you.
j HERITAGE REPORTING CORFORATION -- (202)628-4888
66 1 HR. NEILS: On the NSAC 113 study, as I said, we 2 attempted to quantify the conservatisms we alleged were in the h
3 Point Beach case study because we believe that when ycu are in 4 risk space numbers and you wish to use those numbers for 5 research allocation decisions, as to whether you are going to 6 change the plant or do something very dif.ferent with your 7 software and your procedures and operators, which also 8 involves large resources, then one should have an obligation 9 to try to find the real world best estimate realistic numbers 10 j and depart to the fullest extent possible from the type of i
11 bounding calculations, conservative approaches that are more 12 j traditional in design Daais licensing analysis, and between l 13 our people that we have here today, we will try to illustrata 14 to you wherein we think those differences lie, ar'd hope to I
15 persuade you that this is not a big mystery, but rather l
16 ! technical differences that have an answer. Jackson?
l 17 MR. HAUGH: Good morning. I am Jack Haugh from the I
18 Electric Power Research Institute, Nuclear Safety Analysis 19 Center. Together with Gary Vine, I have been a project 20 manager for the NSAC 113 report. I would like to offer you a 21 l few preliminary considerations, and then I will ask Bill N
22 ! Parkinson, one of the principle investigstors, to get up and 23 4 address certain matters, and tren I will return to the podium l b 24 li to address the final matter end give you a fe9 perspectives i
25 of mv own.
l H 1
HERITAGE REPORTING CORPORATION -- (202)628-4888
t 67 1 Very briefly, without attempting to regurgitate all 2 that Jerry has said, I would like to draw your attention to 3 the fact that when the caso studies were in preparation, and 4 being published, various industry groups and committees 5 provided tons of verbiage by way of commentary on the 6 documents, generally of the nature that in the industry 7 viewpoints, the case studies were conservative and this was a 8 desire on the part of all, including the NRC statf, who 9 specifically asked us would you please quantify those 10 differences to give us a better perspective of what you're Il qfindingdifficultieswithandwhyyousaythone studies are 12 conservative, so we embarked upon that which ultimately b'.ame' 13 published as NSAC 113.
14 l Its purposes are as stated cr. the slide--to identify 15 fthoseconditionsinthecasestudy, and to quantify them by b
16 rneans of our best estimate analyses, and in so doino io focus 17 on ths same scope and plant model as was used in the case l!
18 i studies so that we could atternpt to as close as possible do a h
19 d one-for-one comparison that could illustrate differences that i b
20 0 result from input data, var'.ous resumptions made about the l I 21 l planto and So on.
\1 22 And then we also took our whack at trying to feel L
23 h out not only the initial core melt frequency estimates that 24 were calculated by the two analysis teams, but also to bring ,
l 25 h our same perspective to bear on the add-on SDA chart and again Yl HERITAGE REPORTING CORPORATION -- (202)628-4888
i
)
68 1 to provide to the NRC staff a documentable referenceable 2 quantified estimate of the conservatisms in the case study 3 reports.
4 (Slide) 5 HR. HAUGH: The next niides in your package I will 6 just touch on so briefly because they have been covered quite 7 adequately earlier by Roy Woods, and by Jerry also.
8 i In so doing these analyses, we have then had several I
9 technical exchanges as well as exchanges of correspondence on 10 various matters, disagreements between the analysis teams by 1
1
'1
, l which I mean, of *murse, the Sandia analysis team and the EPRI l
12 WOG analysis team. As a result of those technical exchanges, l
13 the Sandia analysis team revised its original estimate of the 14 core melt trequency for Point Beach. They have revised it l
15 downward from 3 minus 4 to 9.3 minus 5. The NSAC 113 estimate l
16 4 by way of comparison, has been 1.0 E minus 5. And an 1
17 interesting observation is, of course, that this includes the 18 I internal and external events as identified in the esse study.
19 (Slide) 1 20 HR. HAUGH: The principal contributors to the i
21 revised case study include several items--long-term Station ;
L I 22 I Blackout, anall break LOCA with failure of recire, loss of l u +
23 off-site power transients, and the standard common garden
(
24 .
variety reactor turbine trips assumed to be accompanied by j f !
25 1 stuck open relief valve. l l
i b
HERITAGE REPORTING CORPORATION -- (202)628-4888
69 1 The principles contributors to the external events 2 in the revised case study numerical requantification are 3 seismic sequences, and fire sequences. And with regard to 4 fire, we are talking specifically about the auxiliary feed 5 pump room and the switchgear room, and in terms of the revised 6 case study quantificatin, the seismic and fire sequences when 7 combined as the external events have a larger numerical value 8 than do the ensemble of internal events.
I 9 I (Slide) t 10 I HR. HAUGH: For your use during the presentation, I 11 j den't think we need to walk through all the numbers. These b
12 principle contributors to the revised case study 13 requantification of core melt probability are listed here and 14 you can refer to these during the course of the day as 0
15 L necessary.
L 16 I suppose thw most important thing to point out is 17 that the largest substantive difference between, at this 18 L point, between the N$AC 113 and r evised case study analyses 1
19 concerns the fire analysis. You will see great disparity in i
20 number, and we will be talking about rire at quite some length; i
21 during the latter portions of the presentation. ;
i 22 (Slide) l I
23 HR. HAUGH: The reasons, as Roy Woods mentioned l
24 earlier, are attributable to the choice of input data, various 25 L modeling assumptions that are made along the way, and h
!l HERITAGE REPORTING COP % ATION -- (202)628-4888
70 1 allowance for recovery actions. The differences can best be g 2 illustrated by focusing on those sequences that we have 3 identified as constituting some 90 percent of the risk, and 4 many of the comments and items that we will be bringing up and 5 discussing with you apply equally as well, however, to many 6 other sequences although with far less significance. The rest 7 of them are all in the noise level as far as we are concerned.
8 There are many things that we could continue to debate and 9 j argue in those other sequences, but it is really of little i
10 } value when they are in the noise level compared to these top i
11 level drivers at this point.
N 12 4 Yes, sir?
H 13 l DR. KERR: Point Beach I assume concludes that it is L
I; 14 now in compliance with Appendix R7 15 Y HR. HAUGH: I would like to ask Roger Newton from 0
16 Wisconsin Electric Power to address that question?
17 i MR. NEWTON: My name is Roger Newton from Wisconsin 18 Electric. We have been working with the NRC on Appendix R 19 h compliance for ten years now, and we have one remaining item 1
20 o to settle. !
.I l 21 DR. KERR: Hy question really has to do with the l I
22 risk calculations.
23 Does the risk calculation assure that all of the l i
24 Appendix Rs are in place? l 25 HR. NEWTON: No, it does not. That applies both to !
l i
h l HERITAGE REPORTING CORPORATION -- (202)628-4888
71 1 the work done by Sandia, and the work c'one by EPRI, and in my 2 discussion I will point out some of the items that are still 3 involved in the package.
4 DR. FF;RR : You are going to talk about that? I will 5 wait then. Thank you. Mr. Davis? !
6 CHAIRHAN WARD: Roger--but I don't want to let that 7 go with--are the items remaining to fulfill Appendix R 8 requirements estimated to be significant risk contributors? I 9 mean the PRA was done.
10 HR. NEWTON: The answer is yes, n
u 11 CHAIRHAN WARD: They are?
12 b HR. NEWTON: Yes. There is one item in particular a
13 that was identified in the ,landia study. EPRI study, that fire 14 j protection also identified it, and we have resolved how we are 15 b going to fix it with the NRC. We haven't finished all of the h
16 h details and we haven't implemented that, and I will be talking 17 J cbout that specifically in my portion.
18 '
M't . WARD: Thank you very much. !
i 19 , HR. DAVIS: Just for my own review of *.!is, you I 1 l 20 I prepared NSAC 113 which was the NRC's examination of Point I 21 ; Beach. Then you had a meeting with the NRC to discuss the 23 differences. The NRC then revised their estinate. ;
23 Did you do anything to your estimate after a meeting lj
'l 24 with the NRC?
25 HR. HAUGH: I':a prepared to discuss sore of that l
l HERITAGE REPORTING CORPORATION -- (202)62S-4888
72 1 with you a little bit later. Yes. We are gcing to give you a h 2 couple of new numbers.
3 HK. DAVIS: Now the report NSAC l'.3 that we have 4 now, this is your latest--
5 HR. HAUGH: That is the only published ve;sion of 6 our analyses. I will tell you later on in one particular aren 7 with regard to fire that we would now revise that number 8 somewhat, but that the bottom line difference is not 9 significont.
l 10 j HR. DAVIS: Thank you very much.
11 HR. HAUGH: At this point I would like to ask Bill I,
12 ! Parkinson from SAIC to come forward and discuss the internal l 13 events portion of the presentation.
14 q MR. PARKINSON: I am going to be talking about four t;
15 ll external events sequencas that continue to have a significant U
16 [ difference on how we have define what is a significant f
17 / difference between the NSAC 113 and the case study or the 18 ? revised case study.
19 There are two LOCA sequences--small break LOCA and 20 L transient induced LOCA, which I will talk about :1rst. And e
21 b followed by that there are two other sequences which are 22 essentially Station :kout sequences.
23 ! In the case of the small break LOCA, and one thing 24 that has been brought up here I think that's fair to say, is g 1 25 that we have had one detailed technical exchange so far. Some i'
i I
HERITAGE REPORTING CORPORATION -- (202)628-4888
73 1 of the information I don't feel we have completely 2 communicated and we expect and we are discussing some of (5at 3 now and we will be fellowing up on that. We do think that 4 sone additional changes will be made in the case study based 5 on these arguments.
6 Thu small break LOCA scenario that was examined is a 7 small, is a LOCA of two inches or less. Main feedwater in l
8 isolated by a safety injection signal. High pressure 1
9 injection succeeds and continues to succeed. There is an 10 li implication that is not addressed directly in the case study il 11 that RHR cooling is not achieved before or before the RWST is I
12 depleted. Subsequently there is failures in recirculation.
l 13 We will go over both of the dominant contributors to the 14 failures of recirculation, and they include operator failure l
h 15 [ to implement the switchover, and equipment faults which are h
16 h not recovered from which would prevent the switchover. -
U h
17 L subsequent core cooling if the switchover was prevented.
3 l 18 ? You c2n see in teres of the results that there has i
'1 19 been a Jignificant change from the case study to the revised ,
l 20 case study that has been due to the small break LOCA !
- l !
21 frequency, change thet has occurred based on NSAC 113. !
l I 22 Additionally, there is still some remaining s
l; 23 l difference, the majority of which, almost all of which is froml g 24 25 the 7,0E minus six to 5.8E minus seven due to the two points about operator failure to inplement somerecireandequipment,l i
h h
HERITAGE REPORTING CORFOEATION -- (2021628-4888
74 1 fat. lure to recover from those, g 2 Briefly I will go over this since we have generally 3 reached agreement in the area of small break LOCA frequency.
4 The case study started out with a value of 2E minus to. It is 5 a value that has been used in a number of PRAs, including 6 industry studies. It came originally from the ANOI rep which 7 in turn referred to a memo from Hurley to Eisenhutt about, 8 primarily oriented toward reactor coolant pump seal LOCAs.
9 These LOCAs range from a few GPH to I think in the case of 10 Robinson on the ordsr of a few hundred BPH or 400 GPH.
I 11 l When we looked at that experience, and when the L
12 h industry has or let's say the industry has looked at that U
l 13 experience, in terms of what those seal LOCA information was 14 ltellingus, one thing it particularly told us, as we all know, h
15 l there has never been a case where a plant has been required to b
u 16 so that in effect, by taking this number hl co to recirculation, 17 L of 2E minus two of small LOCAs that have occurred and saying j h
18 h implicitly that RHR cooling was not achieved, you are implying c
19 j then that this 2E minus two is always going to go to !
o 1 II 20 1 recirculation. Now that was obviously not true because i
b i 21 h today's experience level with numbers and reactors, we would 22 1 be expecting two accidents to go to recirculation every year.
23 '
We are not seeing anything like that, j 24 For that reason, this issue was raised and l
, I 25 understood, and the NSAC 113 value was accepted. We still l l
HERIT.'.GE REPORTING CORPORATION -- (2021628-4888
l 75 1 believe that there is some range in sanple use for small LOCA 2 frequency. One nay be more appropriate to injection faults 3 and one more appropriate to recirculation fault, but that's an 4 average between the two.
5 DR. KERR: What in the number? What does it 6 represent, the three times ten to the minus three?
7 HR. P ARKI!iS0!!: That number came from the Oconee 8 PRA. It is--
l 9 j DR. KEPR: What does it describe?
l l 10 HR. P ARKIN SO!!: Right now this describes in the way a
11 ;l it is implemented in NSAC 113, describes a small LOCA event n
12 -l which t_quAres injection and a small LOCA event which requires 13 recirculatior.. (
14 In tne previous technical excha:rge meeting on March 15 L 31st of this year, we p ovided some additional information 16 jl which in our opinion, indicated or our analysis of data 17 indicated that a number of about of E minus three was probably i
18 [ more appropriate for injection fault and there hasn't really f 19 been any recirculat$.on data.
20 DR. KERR: Hinus three is a small LOCA requiring i
21 g injection and it requires recirculation.
02 ,
HR. PARKINSON: That is true and that's how it 23 .
effectively turns out in the models. That by the way has bee::
1 24 applied in both our study and in their study.
25 DR. KERR: That's a =%s c of all small LOCAs. in I l
o' j HERITAGE REPORTING CORPORATION -- (202)628-4888
76 1 othar words? And the subset--
2 HR. PARKIllSOti: That 16 true. There have been, like 3 I said, there are more small LOCAs w'. ich have justified this 4 2E minus two number. They were quickly isolatable.
5 DR. KERR: Thank you.
6 HR. P AR KI!!S O'l: Okay. Of the remaining two issues, 7 there is one about operator again fails to implement 8 recirculation, and the equipment faults. They contribute to 9 about approximately 50 percent. Each contributes evenly to i
10 q the differences between the revised case study and flSAC 113.
11 In terms of the operator tailure to implement d
12 j recirculation, we had a discussion earlier about differences 1
13 1 in human errors or in consideration and treatment of human 14 h errors. We definitely do not agree with the case study h
15 hevaluationandithasnot been changed in the revised case
[
16 L study. ko have made many of these arguments before. I think u
17 '
it is important to look at the sources, however, of 18 information when making these conclusions about the various 19 : numbers in terms of the case study and flSAC 113.
20 Ll The the case study the value lE ninus three was 21 obtained from flUREG/CR 1278, the Swain handbook, flRC handbook 22 on human error probability. That particular value is 23 generally i.n the industry termed as a basic human error !
l 24 probability. It's has been argued in the revised case study l l
25 that this iis a generic value. It is not a generic value. It l
l l
l HERITAGE REPORTI?lG CORPORATIO!! -- (202)628-4888
77 1 is in, what we call in PRA a acreening value, meaning that 1 human reliability analysis is a complicated and costly task.
3 In general, the way PRAs are done, screening values are put in 4 initially, and if those values indicate in the first analysis 5 some risk potentially risk significant sequence, further 6 analysis is done. That further analysis includes, for 7 example, consideration of staffing in the control room, and I 8
l think you all know that there in duty technical advisors, Ros, 9 f other people in the control room that if you carry out this 10 d liUREG 1278 methodology, and apply it in a generic manner l
11 '
because these people are, there is a regulatory requirement 12 j for these people, that you would get a lower number than IE 13 minus three. In fact, we are going to presenc subsequently to 14 h you, and I apologize for not having this available now, to the e
15 p flRC our so-called reanalysis of considering these other 9
16 [ tactore as they typically are in 1278, which indicates a lower 17 valv? c'e n than ours.
18 DR. KERR: You are not claiming you have data that j l
19 demonstrate- 1E to the minus four, are you? You are making a l l
20 plausibility argument?
1 ,
21 L HR. PARKItiSoff: We are, our particular re.othodology !
22 for the choice of the value 1E minus four is based in large 23 part on simulator data. I think it is fair to say that that, 1 24 and that simulator data comes from the Army UP program which 25 is an analysis of LaSalle done by Sandia. That data has been HERITAGE REFORTI!1G CORPORATIOli -- (202)628-4888
78 1 combined to generatw a curve of the probability of human ortor g 2 as a function of time, under certain conditions, various 3 stress levels and things like that, so the certainly there is 4 data and simulator data, but there is also--I don't want to i
5 mislead you. There is interpretation involved in these 6 j numbers as well, but the basis, as you know, the root base l
7 l force, it is simulator data, and which is one reason why we l
l 8 think that our value is a more complete, a better value.
9 l Now in terms of, I mentioned that they have the, 10 j this basic human error probability. In NUREG/CR 1278, the 11 hi analysis of the basic human error probability was started 4
12 '
within the analysis, was continued and applied for a large l 13 LOCA switched to recirculation. In that particular case, and 14 L this is an example, in the NUREG 1278 how to apply this 15 , methodology, they calculated a value of 3E minus four, and
'5 16 that was, you know, we think we are, that's supportive of our 17 case because the large LOCA recire is a much more stressful 4
18 C situation, occurs much earlier, the our RWST depletion much !
19 l faster and so there is a lot of reason to believe that there f
20 would be higher chance of error in that particular case so we 21 L again think that our result is, is consistent with that.
l 22 Also again, our, the simulator data that we looked 23 '
at was looking at crew response, and it was also looking at 24 things which are more diagnosis oriented. It is our belief j t
25 that the particular failure modes identified in NUREG/CR 1278 l
HERITAGE REPORTING CORPORATION -- (202)608-4888
79 1 applied in the case study are more failures oriented toward lh 2 turning the wrong valves and things like that and we think a 3 diagnosis oriented analysis is more appropriate here, and 4 that's again the basis for our value.
5 "HAIRMAN WARD: New you are crediting the existence, 6 apparently the simulator experiments that are pertinent here 7 had EOPs available.
8 , HR. PARKINSON: That's true.
9 CHAIRMAN WARD: Now we have some indication that the o
10 ? quality of the EOPs existing in the hundred plants out there I
11 g is, varies over a wide spectrum from good to extremely poor.
12 q How would you characterize the EOPs used in this 13 set?
14 j HR. PARI NSON: Wisconsin may be able to better 15 characterize those, but my opinion is that they, that these 16 porticular procedures are very much standardized based on the i
17 Westinghouse emergency response guidelines, and so I would sayj l
18 this is, there is relative, relative consistency. Now of l 1
19 course LaSalle is a BWR, and there is, it is fair to say that l l ,
20 there is an additional simulator data being collected by EPRI, 21 and being interpreted. Now there is a nunber of prograns l
22 going on, so this event, the entire spectrun of all simulator l
23 ' data that is available, it is a linited set, and it does i 24 consider a variety of conditiens, and there are a range of I I
25 values. Depends on the type of action. .
HERITAGE REPORTING CORPORATION -- (202)6;S-4838 1
80 1 HR. CATTON: There was early work done by EPRI at
- h 2 SNUBBS using the simulator that showed that there was a rather 3 exchange of things that could occur, groups of operators that 4 didn't recognize things at all.
5 Is that data a part of this as well?
l 6 HR. PARKINSON: It is not. I am not familiar with i
7 that, to be honest with you, h
l 8 [ HR. CATTON: It was in an EPRI report maybe five 1 i l
l 9 years ago, six ycars ago.
l l 10 i HR. VINE: I think the important point to be made L
11 here is--Gary Vine from EPRI. I think the important point to l
12 o be made here is that the reference used by Sandia for coming l 13 up with the number of IE there in NUREG 1278 was to take that 14 NUP.EG and find out what NUREG 1278 says. The likely error 15 L probability for shifting to recire in a very rapidly F
16 developing large break LOCA sequence is 3E minus four, and 17 then you come up with a nunber three times worse than that for 18 a much slower and much more benign sequence. You are 19 inconsistent with the reference that you are using, and that's 20 the basis of the point we are making.
i I
21 HR. CATTON: Okay.
22 HR. PARKINSON: I guess to, to illustrate this with j 23 a little statement about what these basic human error {
g 24 probabilities mean, they are commonly referred to as BURPS, so 25 that gives you a little bit maybe of a perspective of how t I
i I
HERITAGE REFORTING CORPORATION -- (202)628-4888
81 1 final they are in a particular analysis, and this one 2 certainly left a bad taste in our mouth.
3 In terms of the next slide, the other major 4 difference between the two studies for this particular 5 ,
sequence is a consideration of recovery from recirculation i
6 faults. In the case study, or these recirculation faults 1
7 ' include failure let's say of a particular type of valves, for 8 I example, that might be in the sump recirculation lines.
\;
9 Things include failures in the RHR portion of the 10 recirculation since HPI is succeeding up to this time, and 11 s basically the question here is given these failures, given the l
12 ;i inability to immediately switch over to recire, what will the l J l 13 operators do? What will happen?
14 .
The case study in this particular case credited that 4
15 h the operators would in cases where motor operated valves had 16 failed, and that those valves had failed because of signal and 17 control fault, that the operators would go locally to the l
18 1 valve and open those valves and thereby establish the path forl l
19 recirculation and go ahead and implement that. ;
! l 20 The revised case study maintained that, that 21 ! particular analysis.
22
- MR. CATTON: Are operators exercised on this 23 particular event?
i 24 MR. PARRINSON: On recirculation? D e f in i t e. l y .
I I
25 o DR. RERR: Is the assumption that the stress of H
HERITAGE REPORTING CORPORATION -- (202)628-4888
82 1 simulator operation is equal to the stress one might have if l 2 one knew one was in an accident mode, or is the assumption 3 that the operator doesn't know he is in an accident mode?
4 HR. PARKINSON: I don't think I can really answer 5 that in terms of--I'm not a human reliability analysis expert 6 and I would really be reluctant to do that. My understanding 7 l is this is based primarily on simulator data, and that the I
8 l that, that question is a valid one.
0 9 h DR. KERR: Okay.
10 q CHAIRMAN WARD: Does Point Beach have a e
11 plant-specific simulator?
12 i HR. NEWTON: Roger Newton again--we do not, and we l 13 have not had ene in the past. We ha"e placed an order for one 14 1 and it is just in the early stages of manufacture right now.
15 We expect to have it operational in about a little less than 16 three years from now.
17 CHAIRHAN WARD: So you use, what do you use for i
18 sinulator training?
I 19 HR. NEWTON: We have trained on the "ion s.mulator l
20 and now we are training on the Kewaunee sinulator.
21 CHAIRMAN WARD: Ch. ,
l 22 HR. PARKINSON: I will point out ene thing, that in i
23 this particular Sandia report that I nentioned that had l l
24 sir.ulator data frcm LaSalle, local valve operation is one of .
1 25 the examples that they look at in that report, and that the l
}
HERITAGE REPORTING CORPORATION -- (202)626-4886
83 1 Swain handbook number is much lower probability than the 2 diagnosis nurrber for the operator actually determining that j 3 they need to perform this action, that chance that they walk 4 down and turn the wrong valve which is generally dealt with 5 with Swain, is much lower here. So these, the simulator does 6 cover that experience, does cover things like that, but you 7 know, certainly is not a wealth of information, and EPR has 8 made, is currently undergoing a pretty large program oL' t 9 simulator data and simulator, and analysis of that data.
10 flow liS AC 113, we basically went by the procedures.
l l 11 Okay. We fully credit Point beach ECA 1.1 procedures, which 12 is the procedure that deals with the case where you have l
13 problems with switchover to recire. In thst procedure the 14 first ching it says is to do what the case study, try to f I (
lb h perform the action that the case stud ~r said would you perform, 16 n arue ly , deal with local valve operations.
17 ,
The second stop in the precedure is to refill the ;
18 l RFST and try to maintain the level in tha, t and remember now 19 we are dealing with the small LOCA, few hundred GPH depletion 20 rates. We are not talking about huge depletion rates here, b and we have indicated the source of water in llSAC 113 and 22 amounts of water that are available.
23 But it is basically to go out and refill this RWST.
24 If you credit that additional recovery, basically again you 25 reduce this sequence to a very small value. One may argue HERITAGE REPORTIllG CORPORATIO!1 -- (202)628-4888
84 1 about what the probability is for that recovery, but there is g 2 multiple systems that are available to refill the RWST, and it 3 is in emergency procedures. It is in training. This is 4 something that is exactly the way the plant deals with it. It 5 is not near the artifacts of PRA analysis. This is the plant.
6 So we, based on that, we feel very strongly about 7 our particular estimate on operator recircs which that that 8 has not yet been agreed to and particularly also on that 9 recovery analysis should be based on what the procedures say 10 and that our representation in this sequence is a much better 11 representation of what the procedures say.
12 So you can see again our conclusions here. We l 13 believe the Sandia application of 1278 was incomplete. We are 14 going to try to provide one. We think ours is better because 15 it is based mare on simulator data.
16 (Slide) 17 0 HR. PARKINSON: Tne next event that we are going to 18 look at is something that's typically called and Q stuck open 19 relief. Valve basically these types cf scenarios are 20 transient induced small LOCAs.
21 Now you can see--I will jump down here and then we 22 wili go back to the scenario--that the case study has a value 23 of, originally value 2.5E minus five, went down by a factor of 24 seven. NSAC 11' .o t applicable. It may be hard to determine 25 truth, but we know the sequences is an untruth and this is one HERITAGE REPORTING CORPORATION -- (202)628-4888
I 35 1 like the fires I guess I would say that we are, you know, very ll 2 much opposed to.
3 Basically what this sequence says in the PRA 4 terminology, here is, T3 says that you have a normal type of 5 reactor trip, reactor turbine type trip. Main feedwater is 6 available. There is no primary / sect.ndary cooling mismatch of 7 any kind.
8 Now what the case study seys is that despits --
9 fact, there is still cooling in the primary system, that 10 miraculously a certain percentage of the time a safety relief 11 valve might open and then of course we have, we have some 12 reasonable amount of data or generic data on safety rel.ief j 13 valves failing to close. What they are saying, you have a 14 reactor trip, miraculously the safety relieve valves open. It 15 fails to close, and then you have a LOCA and you run into the 16 l same kind of recirc type faults that we discussed just 17 l previously.
18 Now originally the case study acsumed there was a 7 19 percent chance of opening following a trip. That was--and 20 they also assumed, by the way, that--this is a difference in 21 the way we view her the plant is actually operated--tnat the 22 PORVs were assumed to be blocked all the time, and we got 23 different, each of us got different information at the time so 24 there is some reason to believe that there is reasonableness 25 in both, or we feel comfortable in our situation.
HERITAGE REPORTING CORPORATION -- (2021628-4888
, I
86 1 HR. HINNERS: I'm sorry. Blocked open?
g 2 HR. HAUGH: That is an error. It should be blocked 3 closed.
4 HR. PARKINSON: Sorry about that. Blocked closed, 5 now--
6 DR. KERR: Why did one have to make an assumption 7 about that because the plant is operated both ways?
8 HR, PARKINSON: Well, I guess Dave can speak better 9 later to why the assumption was made that they were blocked 10 closed. My understanding of it is that they interviewed the 11 operators and they indicated that they were blocked cicsed 12 I most of the time. We--
l 13 KERR: That's enough.
14 l HR. PARKINSON: Okay. Now in terms of revised case I
15 study, subseqitent to our critique, the SRVs, the value was 16 changed. There wasn't a value referenced in the caso study 17 for the 7 percent. In the revised case study, there is a 1 1
1 18 percent chance of immediately opening following, SRV l
19 immediately opening following a trip. It is based on a very, 20 an old Sandia study, what is called RSSHAP--reactor safety l
l 21 study method applications program, and it was far a BWR PRA.
l l 22 I guess it is on'. of the, you know, we have discussed earlier 23 there is a major, there is a big difference in human
, 24 reliability estimates. There is another major difference 25 between the various studies, and it comes in translating what l HERITAGE REPORTING CORPORATION -- (202)628-4888
87 l 1 the'3 things mean to observable things in the industry.
2 If you take the normal number of, you know, seven h
3 reactor trips per year, 50 Westinghouse PWRs, multiply by, 4 that by 1 percent, this point up in this event tree indicates 5 that you get three or four safety relief valves opening on a 6 Westinghouse plant per year, and that's ridiculous. We 7 haven't seen that. We havin't seen anything like that.
8 DR. KERR: I guras we should ask Sandia at some 9 point why they assumed B&W data vsuld apply.
10 HR. PARKINSON: That's a good question. Now so 11 basically there is one big issue here, which is it doesn't 12 have anything tv do with operating experience or what we call l 13 observables, and that we feel that this should be tested 14 against observables, and for that reason delete each group 15 out.
16 Additionally, this is not comcon practice for a T3 17 type sequence or reactor turbine in the PRA community to 18 assume a 1 percent chance of a safety relief valve opening
[
19 regardless of status of the PORV, PORVs. It is not done in 20 '
NUREG 11.50. It is not done in industry PRAs. This 21 conditional probability in the 4550 handbook for bow to do a
.2 150 analysis states that this probability should be less than 23 1E minus eight, so clearly we do not think this is very well 24 founded.
"' 25 Now our basis, we went through this and we said J
HERITAGE REPORTING CCRPORATION -- (202)628-4888
88 1 based on thermal hydraulics analysis, based un experience, you 2 just don't get safeties opening. We quoted a Westinghouse 3 report which had a lot of experience that was motivated 4 primarily by Three Mile Island people became very concerned 5 about PORVs, and safety. We are--
6 DR. KERR: Do you mean safety?
7 HR. PARKINSON: Well, they dealt with safeties 8 because the FOhls were assumed blocked open, or blocked 1 hut.
9 Excuse me--blocked shut. We dealt with PORVs because w9 10 believed that safeties would not open regardless of t!.e 11 situation of the POTVs, or this, in particular for this 2-3 12 scenario. Now let me move this up a little more. In terms of 13 what we did--okay. What we said is that this does not exist.
l 14 The sequence is impossible.
1 15 We looked at other sequences, though, because there 16 are indications that PORVs will open. Obviously we all know 17 that, and we have seen experience on the order of I think it l
l 18 is .1 per year, so there definitely are indications PORVs will l
19 open. It is not a trivial thing to determine all those l
20 possibilities when PORVs will oper. so with that, we did, we 21 went to the next best things we could, which we looked at 22 these primary / secondary mismatch transients, which would be 23 the ones that would be likely to cause increase in pressure, 24 and conservatively assume that they would cause a PORV 25 openinc. Okay.
HERITAGE REPORTING CORPORATION -- (202)628-4888
89 1 I should mention, by the way, that this Westinghouse lh 2 experience that didn't show any safeties opening included, 3 about 40 percent of the data was both PORVs block d, so the 4 assumption of whether PORVs are blocked or not is not 5 significant to this conclusion at all, but to get back to 6 this, what we assumed was that a loss of off-site power could 7 potentially or would open both PORVs. They would have to be S reclosed and a loss of feedwater would result. We--that is 9 definitely conservative. If you use our numerics in 10 calculating that, we come up with a frequency of about one per 11 year, about seven or eight times higher, so this is clearly--
12 DR. KERR: Why do you do a PRA conservatively? I l 13 thought that was sinful, unethical, not professional?
14 HR. PARKINSON: Well, I think that if you 15 tried--this is a personal opinion, but I believe that if you 16 tried to do a PRA totally realistic, that you would bankrupt 17 any possible organization that wanted to do that, r.nd I think 18 ;
that the logical way you--I mentioned in the case of screening 19 values for human reliability, people select conservative 20 values, identify them in their documentation as conservative 21 values. If those particular things come up risk significant, 22 then an analysis, more detailed analysis is done, and that's 23 the approacy we took.
24 We were not doing a level one PRA here ourselves, 25 and so we do have to make conservative assumptions along the HERITAGE REPORTING CORPORATION -- (202)628-4888 i
90 f 1 lines. And you know given the, or in the technical exchange 2 for this, we have identified other specific ways in which I
) 3 PORVs could open, but I don't think we have a definitive model
)
I 4 at this point for failures in the plant which would cause 1
5 PORVs to open. We know if you close both MSIVs, you're more 6 likely to open a PORV. There is a higher possibility then, 7 but we cannot at this point make a one-to-one correspondence j 8 in a model.
)
l 9 (Slide) f f 10 HR. PARKItiSoli: I guess in terms of going on here,
)
[ 11 basically you know, our conclusion is that the revised case 12 study is not consistent is thermal hydraulics either 13 intuitively or calculationly. It is, it is not consistent 14 with the Westinghouse PWR experience, and finally, it is not 15 consistent with current PRA practice, so we believe that this 16 scenario should be thrown out, that we do believe that there 17 are other sequences, 2-3 or 2-1 sequences which could 18 potentially be more risk significant in terms of opening 19 PORVs, but it is certainly not this 2-3 sequence and it is 20 certainly not anywhere near this probability or our general 1
21 conclusions about the transient induced LOCAs are not anywhero l l
22 near their probabilities. l l
23 CHAIRMAll WARD: I mean you haven't, I mean do you l 24 propose to include those? Are those so small?
25 HR. PARKIllSoll: They are included in the report, and HERITAGE REPORTIllG CORPORATIOli -- (202)628-4888
91 1 I, to be honest, I can't recollect the exact numbers, but one ll 2 of our sigr.ificant contributors was loss of on-site power, 3 induced PORV opening with a subsequent Station Blackout, and 4 no recovery cf off-site or emergency power before the core 5 uncovered, and that was in the five times ten to the minus 6 seven range, and I think that was our dominant transient 7 induced LOCA leading to core damage. I um relatively sure 8 that was the case.
9 DR. KERR: Contribution by the, assumed in the 10 revised case study is about 3.6 times ten to the minus six?
11 HR. PARKINSON: That's right.
12 DR. KERR: And you are convinced that's much tco 13 high?
14 HR. PARKINSON: Exactly.
15 DR. KERR: Several orders of magnitude?
16 HR. PARKINSON: At least; like I said, the NUREG 17 11.50 is, dascribes this value that they used as more than six 18 orders of magnitude lower, so we are definitely off a little 19 bit. They are definitely off quite a bit.
20 HR. HAUGH: Excuse me, Hr. Kerr. This is Jack Haugh 21 apeaking again.
22 Our point is not merely that the revised case study 23 value is too high. We believe the sequence doesn't belong in 24 tha analysis at all.
]
25 HR. PARKINSON: One other point I should make--our, I
HERITAGE REPORTING CORPORATION -- (202)628-4888 1
92 1 the way we the model that, we move toward, which admittedly I 2 was conservative, was to look at primary / secondary mismatched 3 transients, and we believe that for the pressure to increase, 4 thermal hydraulics tells us this had been, has to be mismatch 5 between primary and secondary cooling so that the only 6 possible transients that could lead to a safety relief valve 7 opening would be primary / secondary mismatch transients, and a 8 reactor turbine trip is not a primary / secondary mismatch 9 transient, so there is no possibility that this could occur.
10 Any transient that could possibly under extremely adverse 11 conditions lead to a safety relief valve opening would not, 12 never be classified as reactor turbine trip.
l 13 Now along those lines, we are going, I will go 14 quickly through two slides that were presented at the 15 technical exchange meeting This just shows the diverse 16 mechanisms for pressure control. You can see sprays.
I 17 DR. KERR: Are you t:11ing me that a turbine trip 16 will not cause a pressure transient in the primary system, or l
19 it won't be big enough?
20 HR. PARKINSON: It will not be big enough, and 21 temperature not lead--the anticipatory trip that has been 22 added for reactor trip on turbine trip will prevent the 23 pressure from increasing.
24 DR. KERR: If it works?
g 25 HR. PARKINSON: Yes. If it worked. That's true.
HERITAGE REPORTING CORPORATION -- (202)628-4888
93 1 flow that is a, that is certainly good point. One would have 2 to consider the failure of that anticipatory trip which is an 3 ATWS event not included in our scope, and that that is 4 commonly done in PRAs, to include that particular failure mode 5 for this type of scenario.
6 Now basically these are the different ,
1 7 mechanisms--sprays, PORVs, reactor trip, and safety valves for l I
8 pressure control. And I will just go through this real 1 l
I 9 quickly. You can see here that we presented it, the technical 10 exchange meeting, measured plant response. I guess we didn't 11 make this particularly clear. At a Westinghouse PWR for a 12 pressurizer, this is pressurizer pressure response grid 13 disconnect from full power with no PORVs operational. You can 14 see the pressure does go up. After about 3.7 seconds, the 15 PORV set point is reached, but the PORVs don't open. They are 16 blocked closed here.
17 j You can continue up until you reach the reactor trip 16 set point, at which point the reactor trip signal begins or is 19 initiated and at about 7.1 secondo you reach the maximum 20 pressurizer pressure, so you can see that margin between the 21 safeties which are 2500, and the reactor trip set point, 22 effectively prevents the chance of a safety relief valve 23 openingin these particular cases, so you see how the pressure 24 can go up. This is a case where there isn't and anticipatory 25 trip scenario and that's why the pressure goes up past the I
!!ERITAGE REPORTIllG CORPORATION -- (202)628-4888
94 1 PORV set point, and once it increases to the point where the l 2 reactor trip occurs, shortly thereafter, the pressure goes 3 down.
4 DR. KERR: Has there ever been a case in which the 5 safeties were misset so that they opened at lower pressure?
6 HR. PARKINSON: There has been, I'm aware of one 7 scenario, and I'm not, I can't recall to be honest whether 8 that this was in power operation or not. It might have been 9 in hot stand-by, where safety spuriously opened and then 10 '
reclosed. But you know, we consider I guess that particular 11 event would be a potential small LOCA initiator or spurious 12 open. The answer to your question may be--Don, do you know?
l 13 On the case of--I do not.
14 DR. KERR: It is a rhetorical question. There 15 certainly have been cases in which they were misset.
16 Mr, PARKINSON: That's true.
17 DR. KERR: To assume that they will always open at 18 2500 strikes me as being a little extreme, but the problem is 19 small.
20 HR. PARKINSON: Yes. I guess we would argue that 21 that would be a small probability.
22 CHAIRHAN WARD: You are saying that the, there is 23 actually another answer. It is anticipatory trip after the,
, 24 caused by the turbine?
25 CHAIRMAN WARD: And in both that and this one, the HERITAGE REPORTING CORPORATION -- (202)628-4888
95 1 high pressure trip fail, that throws you into an ATWS event?
2 HR. PARKINSON: That's right. That's out of our lh 3 scope.
4 CHAIRMAN WARD: Now when the ATWS events are 5 analyzed--maybe you don't know the answer to this--ir., in the 6 Lottom line analysis of ATFS, is there credit given for 7 delayed manual trips in recovering from--
8 HR. PARKINSON: Sometimes; in that particular case 9 that we talked about there is, still would be the high 10 pressure trip, so it would depend on the nature of the 11 failure. If it is anticipatory trip and anticipatory signal, 12 we would expect the PORVs to open, but the anticipatory trip l 13 to occur at this point and the transient might appear much 14 like this particular case where then too, the pressure of the 15 primary system would go down and the shutdown occurred, so 16 the, when I mention this particular case about the turbine 17 trip, and the anticipatory trip, there are other trip 18 mechanisms available, both automatic t.nd manual.
19 In the manual, actually I am not that familiar. I 20 shouldn't say how much manual trip is considered in the ATWS 21 analysis on a PWR. It depents on the timing of the 22 pressurization transient. The next two internal events 23 sequences that we are going to discuss that still leave 24 significant differences between the revised case study and 25 NSAC 113 are primarily Station Blackout oriented sequences.
HERITAGE REPORTING CORPORATION -- (202)628-4888
96 i l
l 1 There is the what has been in the case study called 2 the long-term Station Blackout. That is an event where there h
3 is a total loss of on-site or of off-site AC power and on-site 4 emergency AC power.
5 However, that in itself does not cause core damage 6 in the PWR because you have auxiliary feedwater turbine driven 7 auxiliary feedwater pump which is AC independent,. It 8 continues to operate. However, you have to ensure that that 9 particular pump can operate for a reasonable period of time.
10 In the case of Point Beach, one of the 11 vulnerabilities I guess you would call it at Point Beach is 12 the, a small condensate storage tank, and the long-term l 13 Station Blackout sequence here results from the concern that 14 the auxiliary feedwater turbine driven pump will work.
15 However, the condensate storage tank in inventory will be 16 depleted ar.d that will then in the case, as the case study 17 assumes, that is, once that level is depleted sufficiently, 18 the auxiliary feedwater turbine driven pump will no longer 19 function, and the core will eventually become uncovered and be 20 damaged.
21 In the next scenario, which I will get to in a 22 minute, the short-term Station Blackout scenario, in that 23 particular one, the auxiliary feedwater turbine driven pump 24 fails immediately or is inoperable for other reasons.
25 Now you can see there has been creater than a factor HERITAGE REPORTING CORPORATION -- (202)628-4888
97 1 of three here change in the revised caso study to the case g 2 study, and that's due primarily to differences in IREP data.
3 The case study used the diesel general failure and eight-hour 4 mission time, used generic loss of off-site power data, and 5 considered recovery of off-site power, and they also 6 considered diesel generator repair but that had very little 7 effect--10 percent reduction, and that wasn't on specitic 8 scenarios. It was very small.
9 NSAC 113 used the, what we considered to be a data 10 base with higher integrity than the industry and the NRC we 11 believe do also, which is NSAC 108, which is an intensive look 12 at diesel generator failures. That particular data base does l 13 not specify particular hourly failure rates for diesels after 14 they successfully start.
I 15 Our review of the data base indicated in effect the 16 amount of time that was generally considered in the runs of 17 l the diesel generator was on the order of two to three hours, i
18 so we loosely described that as a two or three-hour ef2ective 19 mission time. That mission time is appropriate to this 20 particular case because the mean time to repair here for the 21 diesel or the mean time to repair for this situation is 22 getting off-site power back, and that generally, the mean time 23 for rr cuor off-site power is generally about one in three 24 quarte.s, two hours, somewhere in thst timeframe, so.
i 25 i CHAIRMAN WARD: So two to three-hour mission time, HERITAGE REPORTING CORPORATION -- (202)628-4888
i 98 1 you are saying that the, if it starts, it is available with a g 2 hundred percent assurance for two or three hours?
3 ' MR. PARKINSON: No. The NSAC 108 data base included 4 faults in the run mode, but they did not tabulate the number 5 of faults and the number of hours of successful operation.
6 Okay. So we, in using this data base, we make an implicit 7 assumption about what the average run, successful run time is 8 for these diesel generators, and our look at that is that it 9 is in this two to three hour time range for the avercqe run 10 time so that is in effect implicit by having a probability for 11 start failure and a probability for run failure.
12 In this particular case of the IREP data base, they l 13 took a run failure or diesel generator fails on a per hour 16 basis, and multiplied it by a mission time of eight hours.
15 l This calculation is inip?.icit in our analysis, and explicit in 16 theirs, but we, as 1 say, this affective mission time is 17 l appropriate given that off-sito power is recovered in about d
13 h that same timeframe or actua31y narlier than that timof r an.e .
19 CHAIRMAN WARD: Okay, so you are saying there is, 20 there is not an additional running failure?
21 MR. PARKINSON: The probability, no; there is, that 22 is included. That is included, and there is a probability for 23 that included in the NSAC 108 data base.
I 24 CHAIRMAN WARD: Oh, there is? Okay.
l 25 MR. PARKINSON: Yes. That is definitely included.
HERITAGE REPORTING CORPORATION -- (202)628-4888
99 1 Ne also used a plant-specific value for loss of off-site 2 power. There has been reasonable amount of experience at h
3 Point Beach, and onn fault, that number is roughly the same as 4 the generic, about a 25 percent reduction I guess is the,
)
5 about the value. You will see the numbers later.
6 Finally, we again like they did and like everybody 7 does included recovery of off-site power. We also looked at 8 refilling that CST once it was depleted, and therefore 9 maintaining the or allowing the auxiliary feedwater turbine 10 driven pump to continue to run, and then finally, some of the 11 Station Blackcuts were due to one AC bus being lost because 12 one diesel started and that degrading service water and l 13 subsequent failures in service water then failing to cool the l 14 i diesel generator. We looked at some load balancing of thene l
15 service water loads to the diesels. It didn't turn out to be 16 risk significant.
17 The status of this is we have reached relatively 18 l cood agroume.nt en the data. Aspects of the revised case atudy 19 is accepted. NSAC 113 loss off-site power diesel generator 20 data, we don't have agreement on our recovery actions.
21 CHAIRMAN WARD: Just one question--is the 22 plant-specific data on loss of off-site power at Point Beach 23 more favorable or less favorable?
l 24 HR. PARKINSON: It is slightly more inverable, but 1
25 it is very close. It is--the numbers are on the later slide, HERITAGE REPORTING CORPORATION -- (202)628-4888
100 1 but I think it is 8.4 versus 6.2 minus E minus two per year.
2 HR. DAVIS: Bill, I have a question. It is my 3 understanding that Westinghouse has in place now procedures 4 for rapidly bringing the primary system temperature and 5 pressure down in the event of this accident, and they do that 6 by manually opening the atmospheric dump valve on the 7 secondary side, and the reason for doing it is to avoid loss 8 of primary inventory from pump seal plants which are likely to 9 occur when you lose pump sealing injection and pump seal 10 injection cooling, loss of all AC.
11 Does that procedure, is that a procedure assumed to 12 occur when you look at the probability of this sequence?
13 HR. PARKINSON: I think it is fair to say that we 14 did not specifically analyze that procedure, or that the 15 effect of that procedure; I de not know what the auxiliary 16 feedwater flow rates would be in that situation. The, the l
17 normal auxiliary feedwater flow immediately post-trip I 10 ,
holieve is about 400 GPH, and the I guess hot stand-by type l 19 flow is on the order of 200 GPH. Do you want to make'-- l 20 HR. NEWToli: I would like--I think the proceduro you 21 were talking about is if you do have the Station Blackout with 22 a small leak that you know is going on, but the procedure is 23 addressed that if you have a normal behavior of primary I'
24 system, we don't expect significant seal leakage. Therefore,
% 25 we would just deal with return to power as the first choice.
HERITAGE REPORTI!JG CORPORATION -- (202)628-4888
101 1 If that can't be done, then you do the cooldown, ll 2 depressurization, and just retuce your seal leakage. If you 3 compound that with a small LOCA, you then take the more 4 serious actions that you have described.
5 HR. DAVIS: That's not my understanding. My 6 understanding is that the operator will be instructed to 7 immediately depressurize whether or not he has any indication 8 of a LOCA. And if that's the case, that may influence the 9 reliability of the steam driven aux feed train as well, credit 10 for recovery actions in these other areas.
11 HR. PARKINSON: Depends on what you mean by, mean by 12 the reliability. My experience indicates that turbine driven l 13 pumps will run at relatively low pressures, certainly 14 atmospheric pressures for steam, so that's another influence I
15 you are talking about. If the influence is flow I think we 16 have considered the maximum flow rate out of the turbine 17 driven pump. If your concern is auxiliary f eedwriter flow, re l
13 have considered tne maximum flow rate in our CST < depletion i
19 times, so I think we have that implicitly covered. That's ny 20 belief.
21 HR. DAVIS: You didn't assume that occurred when you 22 analyzed this?
23 HR. PARKINSON- . That is correct. Now again, in this 24 particular case, the big difference hara is whether you are 25 going to, is the reliability and the feasibility of CST HERITAGE REPORTING CORPORATION -- (202)628-4888
102 1 refill, primarily because I mentioned the service water g 2 loading thing didn't make a big difference.
3 CHAIRMAN WARD: Let me ask a question of Haugh or 4 Mr. Vine. We have until, let's see, 11:45 for the NUMARC 5 presentation. I have no way of telling whether you are on 6 schedule or not.
7 HR. HAUGH: WE have quite a bit more material that 8 we would like to bring to your attention. I shall endeavor to 9 curtail some of the long winded discussion on fires, for 10 example, to help bring this back.
11 HR. PARKINSON: I can move this I think pretty fast.
12 CHAIRMAN WARD: Maybe you better.
l 13 HR. PARKINSON: It is fair to say that the--excuse 14 me. It is fair to say the key issue is CST refill recovery.
15 We have different opin4. ors en whether it would be viable or 16 l not. It i s its toe procedutes f or the plant. Operators have l
17 l performed strailar types of actions of hooking up the fire 1
18 water systein to this in the sense of losing the hose 19 connections that would be involved. The CST rofill operation 20 has not been tested under Station Blackout conditions.
21 However, there is, you know, we do know where the 22 tools are. The operators are familiar where the fittings and 23 tools are. Emergency lighting, portable emergency lighting in 24 !available. We believe that additionally, the inventory of the 25 CST, the normal inventory is large enough to give them quite a HERITAGE REPORTING CORPORATION -- (202)628-4888
I 103 i
1 bit of time to do this, not the tech spec inventory which has )
2 been conservatively assumed in the case study, so we believe 3 this is very feasible.
4 Additionally one other point--recently Wisconsin 5 Electric Power has--
6 CHAIRMAN WARD: Let me just ask you a question about 7 that. You said the normal inventory is used rather than the 8 tech spec inventory, but I presume the plant would apparently 9 operate with, I mean does it have, is there a plant procedure 10 that keeps the inventory at some level higher than the toch 11 spec inventory?
12 HR. PARKIllSON: We actually--I guess it would be 13 better to say maybe operating practices would be the best 14 word. We went through the shift logs, and determined that 15 greater than 99 percent of the time--I believe this is the 16 right value--there was about 30,000 gallons rather than ten in 4
17 g the, or in that range or more in the condensate storage tank. j f l 18 q That particular value happenn to be a value which the ease 19 study indicates if that value was available, that this 1
20 recovery ac ion would be very feasible, so we think that whileI 21 the toch spec limit is 10,000 gallons, the normal operating 22 practices for the plant, you know, indicate that this is a 23 very feasible, feasible recoverf.
24 Wisconsin Electric Power wi.11 talk additionally, 25 too, about some plant modifications which would in effect HERITAGE REPORTING CORPORATION -- (202)628-4888
104 1 eliminate this concern anyway, so it is probably not worth all l 2 of us belaboring it, but I do want to mention one thing.
3 Wisconsin Electric Power has also recently reviewed its gas 4 turbine generator reliability data and we found a significant 5 difference in the last five years compared to the case rtudy 6 data.
7 We found about a 7 percent unreliability for the gas 8 turbine generator versus about a 20 percent unreliability if 9 you included that the long-term Station Blackout and the 10 short-term Station Blackout which we will discuss later would 11 ue reduced by the order of a factor of three. If you felt it 12 was unfair to use only the most, the last five years of data, l 13 and you combined everything together, it would still be on the 14 order of a factor of two, so there is a substantial reduction 15 in improved gas turbine reliability.
16 h (Slide) 17 ! HR. PARUINSON: In terms c7 the next sequence, I l l 18 j don't think it is particularly urote, for :x co go into this l'
19 in detail, but basically the differences between the case 20 ] study, or the revised case study of 4.9 minus, 6 and 7.7 minus l
21 7, we definitely expect to change in this particular case l
i 21 l because the principal difference in loss of off-site power 1
23 l frequency had already been credited for us but the principal 24 differences are une of new battery which we credited, very 25 c similar act credited under a seismic event, and looking to the e
li ll HERITAGE REPORTING CORPORATION -- (202)628-4888
105 1 next page diesel, generator failure probabilities, if 2 they--also that's already been accepted for long-term Station 3 Blackout and seismic. If those were included, this core melt 4 frequency would be reduced to roughly the NSAC value of 8.6E 5 minus 7. So there is a little inconsistency probably because 6 we didn't communicate clearly where all these things factored 7 in, but basically this value would again be flSAC 113 or around 8 the NSAC 113 value if things already credited on other 9 sequences were credited for this sequence.
10 And now I am going to turn it over to Jack and he 11 j will go over fires and seismic.
12 HR. HAUGH: I have two topics to diwuss with 13 you--fire evaluation, and seismic sequences, and I'm going to 14 arbitrarily handle the seismic immediately by saying you have 15 lacoupleofpagesinthure that talk about the principal d
16 differences that boil down to in the last group in this 0l l
17 portion of the presentation, that boi) down to differences l
18 I over the hazard curve.
l !
19 i Sandia has what we believe an unduly conservative i 20 hazard curve, which is the probability of exceedance of a 21 given earthquake at the peak ground acceleration. That curve 22 is based upon the Zion SSHRP studies, and then modified for, 23 for local soil conditions at .5. We have no problem with that l
24 j in and of itself.
1 25 l The difference is that to illustrate cor.servatioms ,
HERITAGE REPORTIllG CORPORATION -- (202)628-4888
106 1 we could not work from a Zion curve. We didn't have one, hI 2 EPRI studies have one, however, for the Braidwood site which 3 is near by Zion. And the EPRI and Livermore studies for the 4 hazard corvo, deep soil curve, differ, typically ours being a 5 factor of ten or more lower tnan the Livermore studies. We 6 would assume that going from Braidwood to Zion, Zion to Point 7 Beach, that a similar set of differences would hold.
8 However, we did not reduce the Sandia calculated 9 hazard curve by a factor of ten. We elect to use a factor of 10 two for less than three times SSE or factor of five, less 11 than, for greater than three times SSE. Anybody can question 12 either the one curve or the other curve. The point in there l 13 is no site-specific hazard curve. That will have to come in 14 the future. And right now you pay your money and take what 15 you bought. Okay. End of that.
ld
'I s 14 l There ore some seismic recovery aspects of it that I
17 Pe think are a little cetter than previously estimated that if prime.rily boil down to the fact that--we have it laid out 19 I there in agonizing detail--that there is a longer time 20 I available to time of the core uncovery such that if the 21 prin:ipal components driving these seismic frequencies, RWS1 1
22 l and CST tanks, were they to fail early, it really just doesn't 23 matter. We can make use of in place CST. We can align I'
24 auxiliary feedwater to service water system, do that within a, 25 I believe the service water system components can survive up i
HERITAGE REPORTING CORPORATION -- (202)628-4888
107 1 to a three times SSE earthquake level, which is the cutoff 2 limit that we applied. We did not apply that for greater than 3 that, and in so calculating, we have gone back recently to, 4 subsequent to our tech exchange meeting we have used our 5 fragility data. Everything that is done in the case study, we 6 used their hazard curve, not ours. Okay. So we believe that 7 that's clean.
8 The RWST recovery action of aligning to the spent 9 fuel pool, that I might simply add is the, actually the
! 10 modification that had been proposed in the original case study 11 for this fix. The whole business about how soon, what failure l 12 modes, that are involved in the tanks and so on, really just 13 doesn't matter any more. It has gone away because we have 14 cone back and looked and fcund that our original assessment of 15 s<sying wo only had 30 minutes to do this thing, it was 16 incorrect. We have gone back to the thermal hydraulics work I
17 q for. that formed the background document to the Weftirgbousa I
18 Owaers Group which set up everything that we r er ed t.o know , and il 19 shows that wo have considerably muen longer time.
If it I
I 20 believe it is 60 minutes or 90 minutes, and as far as wu are 21 concerned, this thing goes away. So I would like to just 22 finish seismic now unless there is, somebody feels that he 23 ;
just can't live without asking a question about seismic.
d 24 ) Good. I think I have offectively cut you all off on that one.
25 , Let s move on to fire. This is the longest portion 1
IlERITAGE REPORTIllG CORPORATIOli -- (202)628-4888
108 1 of the presentation, and I will ask your forebearance because gg 2 it is so long. This is probably the singularly most 3 significant area in terms of differences between the two 4 studies. You have heard bits and pieces all along the course 5 of today. Seismic, of course that is a big diffetence, as you 6 saw, but nothing so great as a couple of orders of magnitude 7 separating the various analysts, and well one might ask why 8 so? All right. I will try not to use all of the material 9 that is summarized word by word in there and try to confine to 10 pick up some time my discussions to this chart that we have 11 provided in your handout, but I would ask you as I am going 12 along to read the materials that correspond to these items.
l 13 All right. When one is attempting to conduct a fire 14 ,
risk assessment, you normally make use of a procedural 15 approach that says let us first address fires initiators, how I
16 l likely are we to have a fire in a given building in e. given 17 ( zone in a building? How sevore might that fire be? What does n
19 l this imply for requironents for fire supprossion, and ass *2 ming 19 successful supprossion of the fire, then what other recovery 1
20 actions might I need to take to prevent core damage, core 21 nelt? And then ultimately of course, grind all the numbers 22 together and come up with your estimate of core melt 23 probability.
24 ! DR. KERR: Excuse me. I would like to get a little I
25 l feeling for the context.
f l
HERITAGE REPORTING CORPORATION -- (202)628-4888
109 1 What I am hearing is an effort convince me that 2 there is now a standard method of treating fire in PRAs, or is 3 this your method for treating fire, or none of the above?
4 HR. HAUGH: This is what I would suggest to you is a 5 legical approach to follow going there. Generally speaking, 6 all of the PRAs that I have looked at attempt in some fashion 7 or other to at least follow this outer sequence of steps. It 8 is when you get down into the internals of it of how do you 9 apply and do things that the differences occur, and I think 10 that will be obvious through our comparisons along the way of 11 decay study approach and our own approach.
12 DR. KERR: Thank you.
13 HR. HAUGH: Beginning now with the building block 14 fire initiators, if we are concerned, for example, with the 15 sun feed pump room which is located in the auxiliary building, 16 {
Point Beach, one needs to consider generic fire frequency data i
17 for auxiliary Luildings. Then one should by our approach,
.S 8 that which I am recommending to you, one should not merely !
l 19 take a generic value specified somewhere for the aux l
20 l buildi.,gs. Ort needs to apply some engineering i
21 l interpretation, by which I mean that you have to see what are il 22 j the totality of the events that comprise that generic number, l
23 ! and which r >se events are actually applicable to your 24 given plant, the given fire zone within the room.
W 25 ,
For example, if the generic fire frequency is HERITAGE REPORTING CORPORATION -- (202)628-4888
110 1 composed of self-ignited cables, transfocmer fires as 2 separable categories, other electrical equipment by which I g
3 would throw the whole mess of motor control conters, 4 switchgear, circuit breakers, all that sort of thing, 5 separable again batteries and battery charger related fire 6 events, then other events related to pumps and motors, and 7 that sort of thing, and then a miscellaneous non-descript 8 collection of other things that include things like gas 9 explosions, you can immediately see that not all of those 10 events are applicable to the given room. The room that has no 11 pumps should not be considered a generic fire frequency that 12 includes pump related fires and vice-versa for any other item l 13 that may go there, so we would maintain that one should 14 perform a discrimination and a sorting into bins if you will, l
13 landnot have essentially one driver leg as an input signal but ir i j a collection of signsla coing into the analysis.
h 17 h Sepo.rately we would say that because you cannot 18 ,
c11minato the question of a transient combustible O re . that i
19 you should postulate a trannient combustible fire. Why 20 postulate that? Because all of the generic data base show 21 there have been no transient combustible fires certainly at i
22 least of any significant magnitude, within the aux buildings.
23 [Allright.
24 So having done this, now sorted according to the
[
e 25 h appropriate types of fire, fires to the individual zone also d
i HERITAGE REPORTING CORPORATION -- (202)628-4888
l l
111 1 have postulated a fire frequency. You can immediately say I lh 2 yes, I have taken the generic number, I have reduced it 3 some,that. I have given myself if you will some benefit to the 4 respecti/e room because of those fires that are not 5 applicable. I have also cone back in and added another term.
6 All right.
7 The next step we would suggest to you is that one B should .spply a weighting factor as appropriate to those given 9 types of fires because we are talking about initiators, not 10 potential for severity, that we have not yet come to that.
11 The case study--let me recapitulate for a moment.
12 The case study that used the generic fire frequency as l 13 reported in the Seabrook PRA did not perform this step, did 14 not perform this step, and went immediately to this step. In 15 so doing, it applied what it f elt was a reasonable scaling 16 factor, and that scaling factor was the amount of combustib.tv i
27 l material in the given fire zone to that in the whole building.i 18 It used cable loading. We would say that that cable loading 19 is an appropriate scalir.g factor for scaling of ignited cable 20 fires. It is inappropriate scaling factor for any other 21 i ec.ponent of the fire data base. It has nothing to do with I
22 i whether or not a pump goes on fore or catches fire and starts l
23 and can propagate to something else, for example. .
l o
c4 q All right. !
I l l 25 h CHAIRMAN WARD: A questien--the previous speaker HERITAGE REPORTING CORPORATION -- (202)623-4888 m c-
112 1 spoke about testing assumptions used in PRA, PRA against 2 observation.
3 HR. HAUGH: Yes, sir.
4 CHAIRMAN WARD: I think you said that the, you 5 postulated transient combustibles contributed to the fire 6 frequency.
7 HR. HAUGH: Yes.
8 CHAIRHAN WARD: How do you test that against 9 observations when you said there haven't been any?
10 HR. HAUGH: That is correct, and the st?.ndard 11 tachnique at that point is always to postulate an event.
12 Now the, if the collection of the commu: ty says you l 13 know, geaze, because there have been no events, throw that 14 whole bogus thing out, I would be probably equally well 15 agreeatic to that. However, I don't think I could come in i
16 here and say that well, here is a potential initiator that we 17 ,I can just throw away because we don't like it. Okay.
I 18 CHAIRMAN WARD: Not because you don't like it, but 19 because there is no er.perience to support it. If you could ,
i 20 ! reasonably expect to observe the experience, frequently--
21 HR. liAUGH: I would support your statement, Mr.
22 Ward. I would support your statement. However. I am telling i
23 : is standardly done, and the whole matter of transient
!! you this 24 i combustible fires, I don't want to get too lost in that, but W 25 fthathasbeenthe thing that has probably generated more heat l
l HERI... REPORTING CORPORATION -- (202)628-4888
113 1 than light in this whole discussion to date. All right. So g 2 let me move on.
3 Now having applied what we would consider to be 4 appropriate scaling factors, now we must address fire 5 intensity or severity of the fire.
6 The manner in which the caso study did this is by 7 assuming our friend the transient combustible fire--a trash 8 can filled with paper and stuff like that, relatively benign 9 sort of thing, and a ten gallon spill of acetone pooled fire, 10 without our getting all impassioned over that sort of thing, 11 l the approach that was followed was to say look, I need to 12 bound the energetics that might be associated with any of l 13 these kinds of fires here, so we will elect to perform some 14 computer code mixture fire growth and population calculations 15 using a computer code CMPBRN and we will start and model the 16 whole tning through the aceton pool fire and it will helt the 17 room up and ignite the cables in one region perhaps, and so 18 on, and maybe gas heating effects will ignite another set of 19 l cables elsewhere. It is all tied up in the modeling. All 20 right. That's what they did.
l I
21 Thtir approach again was to say look, we want to 22 just bound the range of possible energetics that might be 23 q associated with these events. All right. And we take some I
24 ! exception to that, that p'rticular approach as applied.
25 Again, to use a pump fire, tor example, if this is what we HERITAGE REPORTING CORPORATION -- (2C2)628-4888
114 1 were considering out of the generic data base, the appropriate 2 energetics for that would be say the amount of oil in the h
3 pump. It wouldn't be some arbitrarily speci.?ied whiz bang 4 that you ralk through the room and spill the ten gallons and 5 leave it there, and all of that sort of thing. And that in so 6 specifying these energetics, one should certainly take very 7 careful cognizance of procedures and requirements in the plant 8 that restrict transport, for example, of transient 9 combustibles and so on, but at this point I would like to 10 leave transient combustibles and move on.
11 But we would aluo say, however, that you should 12 ,
apply at this point conditional probabilities that relate to l 13 some of these assumptions that you have made. For example, if 14 you do go back and look at a pump fire, well, certainly the 15 the pump, if that pump is in q conditional probability is one, h
16 [ the room, okay. And you can apply that regarding the i
17 quantity of oil that you specify, cover it with that. If yor 18 were to look at one of those other transient jobs and you 19 analytically place the thing in various pc-tions of the room, i
20 j you should account for all of those things. All right, h
21 ! Moving along, at fire suppression, here is another 22 area that we really disagree with. We would certainly, we
[i 1
23 would--both studies agree you need to address the automatic l
24 suppression and manual suppression effectiveness. There was g
25 ; some differences on, early on in the original case study how 0
F p
ll HERITAGE REPORTING CORPORATION -- (202)628-4888
115 1 many halon trains that they credited, and all that stuff has h 2 been straightened out. Let's net wasto our time on that.
3 The real point of contention is what numerical value 4 do we apply to assign the probability of non-suppression? All 5 right. The case study follows very traditionally data that is 6 used in many of the other PRAs. It refers, references back to 7 ANI, American Nuclear Insurers, data for halon systems, and 8 these are acceptance tests. In other words, how many--the 9 insurers want a level of risk protection, and they require 10 that the, these halon systems have tull system discharge tests 11 and they measure concentrations all around the room, and how 12 long they maintain them and so on, and then make a l
l 13 determination. Either tnat system has passed or that system 14 has failed.
15 Now the acceptance tests are based upon a, within 16 that community, in the fire protection community, a d
17 l conservative approach. For example, if one were talking about 18 an acetone pool fire, three and a half vc ame percent, ic 19 suppresss that fire. You can look that up in the NFPA 20 handbook. However, it is required that the system be able to i
21 deliver 5 percent such that 4.9 percent which woulo have h
22 suppressed the fire is by definition s failure for an 23
[b h acceptance test, so we maintain that this data base is truly U
21 inapplicable. It really tells you nothing about the ability H
!I 25 ! to suppress at least certainly that kind of fire that I have i
l l
HERITAGE REPORTING CORPORATION -- (202)628-4883
116 1 just mentioned. All right.
gg 2 CHAIRHAN WARD: It is sure telling you something 3 about it.
4 HR. HAUGH: Point Beach gets 7 percent volume 5 percent in their room so they have got more. That system test 6 doesn't telling you what you really need to know. What you 7 really need to know is how effective is halon in suppressing 8 actual fires? All right.
9 When I say the case study, I don't fault them for 10 having done that which they did. This was standard. This was 11 traditional. What we have introduced is some new data which 12 i says--we don't have data from the nuclear power plsnts i
l 13 themselves, but we have a data base provided to us from the 14 Department of Energy, from some of their prod'.;ction l
l 15 ! facilities. Granted all the reers are not the sane, but ll 16 4 overall, you know, a room is a room is a room, and various i
17 I pieces of equipment are in there in rooms.
18 What we see is, when we look at their data base, is 19 that in 17 demar.ds f or halon suppression, they had 17 20 hsuccesses, and thogo fires were related to guess what?
I 21 l Electrical components, things of sort; even a few minor h
22 L transient combustibles, all that sort of thing.
i 23 h We would maintain given that and its correspondence Y
24 to the type of fires that are actually occurring in the plant, W 25 h that's your best source of data and you should use that data.
!l HERITAGE REPORTING CORPORATION -- (202)628-4888
117 1 Now having said that, we have 17 successes out of 17 ll 2 demands. We elected to postulate one failure in 17 demands.
3 Again, if we were to putzero in there, we would have told you, 4 you know, for non-suppression probability, thoro is--nobody 5 would have accepted it. Surely there is some thing, okay, so 6 we postulated one failure out of 17. That gave us a 7 non-suppression value far less than the case study value 8 Then we moved on to manual suppression, and, well, 9 i on this one here, I am going blank at the moment. I need to l
10 refresh my nemory.
11 CHAIRMAN WARD: That's probably good because you o
i 12 h just have a few more minutes.
l 13 HR. HAUGH: Suffice it to say we disagreed with that 14 l recovery. We also disagreed on racovery aspect. The most d
15 p important thing, they did credit the turbine driver Ond aux 16 feed pump in the switchgear room scenario, but we believe they h
17 should not because the new battery supplied the j 18 H instrumentation necessary for thet pump. That's the most h
19 j significant part.
I 20 l Now with that, gentlemen, I'll stop. ,
d I 21 L CHAIRMAN WARD: Any questions for Mr. Haugh? I e
I 22 y HR. HAUGH: Hay I respectfully submit to you, we hadt h
23 0 intended to have Wisconsin Electric Power address !
L 24 modifications ongoing at the plant so that they could help h
25 hanswersome of the questions that were raised in this b
i HERITAGE REPORTING CORPORATION -- (202)628-4888
118 1 morning's early genera; ;iscussion. We will do that at your l 2 sufferance, but if you tell us we don't have time--
3 CHAIRHAN WARD: Is that the last thing you have?
4 HR. HAUGH: That's it. Then we are quiet.
5 CHAIRMAN WARD: Mr. Newton, how long will it take 6 you to do that? Excuse me. How long will it take you to do 7 this?
8 HR. NEWTON: Ten minutes.
9 l DR. KERR: Could someone comment on the relevance of l
10 what we have just heard assuming that it is decided to subsume l'
i 11 ;
A-45 in the IPE?
l 12 j MR. VINE: Let me make an attempt to do that from l 13 the standpoint of NUMARC where I am working for a year.
14 NUMARC has undertaken an initiative to establish a l
i 15 )
severe accident working group, and this group has already 16 j met. There have been a number of meetings Letween NUMARC and U
n 17 the Commission with regard to severe accidents, and beyond the d
18 IPE process, what is going to be required out of the generic 1
19 i letter, NUMARC is going to be working with the staff to 20 address two issues that really go one step beyond the TDE, a l A
l 21 ] which both the NRC and the industry think are very important. j h
22 j one of them is severe accident management, and the other is h
23 3 external events.
24 I don't think that there is yet developed a position I
25 hl on the part of the industry with regard to how external events i
HERITAGE RFPORTING CORPORATION -- (202)628-4888
119 1 are going to be addressed or how we would like to see them ll 2 addressed. It appears that the NRC staff has already 3 developed that position, namely, they are looking toward a 4 second IPE or an IPEE as specified in the recent integration 5 plan for severe accident.
6 I think our conclusion from looking at this one 7 plant from Point Beach is that with the exception of seismic l
8 i which is still an open issue, and there is still a lot of work 9 being done on getting agreement on methodologies for seismic l
10 ; hazard curves, and we are getting very close to that, between l
11 ' both the industry and the NRC, with regard to external events i
12 l other than seismic, we would conclude from this study that l
l 13 there is reasonable evidence from the conservatisms we have 14 identified in the generic approach to analyzing those external 15 y events, that all of those external events are not significant 16 0l in l comparison to the other risks that have been identified.
!l n
17 p That includes fire. We are down at ten to the minus i
18 [ seven and lower. That includea external flood, and the I
19 lightning, wind, all of those other are very, very low, and I 29 , think we will be able to prove that without having to do I
21 [h individual plant evaluations for every plant in the country.
b 22 j Does that answer your question?
Y 23 DR. KERR: It doesn't tell ine exactly how you plan h
24 It tells me how you would d to deal with the A-45 issue.
25 propose to deal with external events generally. At least c
it b
HERITAGE REPORTING CORPORATION -- (202)628-4888
120 1 that's what I thought I was hearing.
2 CHAIRMAll WARD: There is no A-45 iscue.
3 HR. VIllE: The question of exactly how the staff 4 wants to resolve the residual questions that they have with 5 regard to vulnerabilities related to decay heat removal as 6 part of the IPE process is really a question that you have to 7 address to them.
8 DR. KERR: What I thought i had been hearing was a 9 lot of detail on difference between your analysis of decay 10 heat removal and staff's analysis of decay heat removal, and 11 if that's going to be subsumed in, into the IPE, I'm not sure 12 how relevant that is. That's what was puzzling me.
- J HR. VIllE : You are questioning the relevance of the 14 continued. dialogue to reach agreement on what the numbers are 15 for Point Beach. I think there is a good answer to that. The l
16 l stL*f has indicated to us, and you see reference to it in the I
17 [ staff's slides, that they viewed this dialogue on the 18 comparative analysis at Point Beach to be very important in i
19 terms of the defining how utilities might be expected to 20 i approach the particular calculations or the particular 21 analyses in their IPEs, be that how you deal with the small n
22 ,
break LOCA analysis or how you deal with an event Q. What 23 kind of assumption should the industry use when they make U
24 h their IPE analysis? Hopefully in the process of this dialogue h
25 hwewillcometo an agreement on what the right number is and u
a HERITAGE REPORTING CORPORATION -- (202)628-4888
121 1 then the industry can take that number and go with it without 2 having to debate it.
3 DR. KERR: The important thing about the discussions 4 is not the way it affects what used to be an A-45 issue, but 5 how it affects the general approach to IPE or whatever 6 analysis is done in the course of an individual plant 7 examination?
8 HR. VINE: Exactly. It seems to us the A-45 issue 9 is virtually closed with the exception of th2 commitment to 10 proceed on with the, with the IPEs, and with subsuming all of 11 the lessons learned out of A-45 into the IPE process that we 12 have addressed all the concerns that were present when A-45 l
13 i was first established as a safety issue.
14 l DR. KERR: Okay. Thank you.
1 15 l CHAIRMAN WARD: It seems to me it is, there is a 16 little bit more to it than that. I mean the staff has I l 17 I concluded that the A, the residual A-45 issue can be subsumed I o
18 I by IPE. It is also concluded that, for example, the initici f
19 0 central issue which was, should be is there a need for 20 p dedicated decay heat removal systems? Staff has concluded 21 [ that those aren't, there is not a need, but they have h
22 j<concludedthat baned on the six PRA studies or partly the risk 23 analy?co, and the conclusiona are, you know, based somewhat on h
24 Q the, on the quantitien, but they are kind of close, and I h
25 think what *;he NUMARC people are trying to say is it is really a
it
- t HERITAGF REPORTING CORPORA 110N -- (202)628-4883
l 182 1 not a close call numerically, that there are big conservatisms ll 2 or errors, biases in a given direction in the staff PRA, and I 3 think the NUMARC people are just trying to put another nail in 4 the coffin here to make sure that, for e;tample, the dedicated 5 decay heat removal system doesn't rise again. I mean I think 6 that's part of it, or that's what T. see.
7 MR. NEILS: That we have zeroed in on the fire. One 8 of the reasons we zeroed in on the fire issue, the treatment 9 of it in terms of core damage frequency, in the case study and 10 Point Beach, is that it involves an arena where you can 11 back-check against reality and see if your methodology in 12 predicting these probabilities does really connect with the l 13 real world, and we are talking here about initiating events 14 that go to the final core damage frequency. That spectrum of 15 initiating events is in the real realm of you know, plus or 16 minus a little bit of ten to the minus two per reactor year.
17 ! Now when there are that many reactor years of 18 experience there, the process should predict something close 19 l to what that has occurred. What the process would predict l
20 here as I understand, is on the order, thus far, of 35 to 40 21 major damage fires in reactor and auxiliary buildings, and 22 that hasn't happened. That's why we take issue with it.
ll 23 CHAIRMAN WARD: Okay, d
24 h MR. HAUGH: This is Jack Haugh speaking again.
b 25 d Since we have returned to fire, I think I owe it to the h
i l
l l HERITAGE REPORTING CORPORATICN -- (202)628-4888
133 1 Committee to, particularly since Hr. Davis asked the question lh 2 carlier, were we making any revisions to the study, we would 3 suggest that we didn't include everything that should have 4 been included in the generic fire frequency because we used a 5 different data base than did the case study, and had we used 6 their case study, and apportioned everything else as we would 7 do, not changing anything else, our fire risk numbers should 8 be approximately a factor of two higher than as shown.
9 CHAIPHAN VARD: Than you ahowed them in your 10 document?
11 l MR. d)"GH: Yes, that we would now be talking order l
12 roughly of *en to the minus seven for completeness.
l 13 CHAIRMAN WARD: Okay. Roger, we have used up about 14 ten of your five minutes, so we would appreciate it if you 15 just highlight this.
16 HR. NEWTON: I wanted to just talk about the 17 l integrity of j regulatory issues at nuclear plants, in i
13 I particular Point Beach, and because this got caught in the i
19 I middle of some of the old issues, sone were done in internal l
20 flooding blocked walls, Appendix R we were in the process of I
21 i implementing when Sandia came in and they didn't catch 22 i everything that Appendix R was having us do with respect to 1
23 j fire. In some cases, they backtracked and the revised case 24 study picked up some of the items, but even the one I have h
r 25 l mentioned already that we have yet to implement isn't in I
HERITAGE REPORTING CORPORATION -- '.202)628-4888
134 1 either of the studies.
l 2 Most of the THI backfits were caught. We were 3 integrating those together on feeduater system reliability 4 procedures. That caught most of those, and so on. The 5 equipment qualification 197 and training were pretty much 6 incorporated when we tried to integrate these in our plant.
7 We made those backfits. In a similar manner, we have new 8 issues coming up, and we feel it is important, too, that we 9 integrate and look at some of these issues together because 10 they are, they do interrelate. Station Blackout, or the loss 11 of all electrical power, is a driving force on decay heat i
12 removal and we have pretty much demonstrated that.
l 13 We have also pretty much indicated that decay heat 14 removal is probably one of the main components of the severe 15 accident considerations, and that's why it is being involved l
16 [ and being proposed to be rolled into that, and on top of that, H
17 we have got the seismic qualification of equipment issue in i
28 l our operating plants that we have to deal with at the same 19 l time.
I 20 (Slide) i 21 MR. NEWTON: That's coing on now. At the same time, 22 we have what I call plant-specific items that we have to add 1
23 h to this list of items that we would want to integrate, and of L
24 L major interest which has come up is the switchgear room h
25 L bypass. It is an Appendix R driven item. It is the last item l
i.
HERITAGE REPORTING CORPORATION -- (202)628-4888
r-129 1 that we have not settled with the NRC or implemented, and it r
2 is simply that this room has both trains of 4160 in close 3 proximity, and the measures for protection that we are 4 proposing under Appendix R was not acceptable to the NRC, so 5 we agreed to invoke the part of the rule that allows you to 6 come up with a solution to that, and that solution is an 7 alternative path around that particular zone or area that 8
l supplies safe shutdown equipment.
I 9 l This is allowed in Appendix R or in the regulation 10 and we will be implementing that to address this one 11 ; particular area, and it also showed up in the A-45 study as a 12 vulnerability area and penalty was attributed to it under the I
13 i fire risk analysis and also under our study as well. I will h
14 give you more specifics on what this will cover. Because this h
15 [ bypass needs a supply of power, we had to go out and revise U
16 f the supply coming in and that's just what the 136 is doing and l'
17 we had other plant-specific reasons for revising our 138 bus i
18 [ and this is a bus that is between the switchyard and the d
19 plant. This is the bus that our gas turbine hooks into, h
I 20 i We also have two low voltage stations transformers o !
21 that serve the plant in close proximity to each other. Both ,
22 < we and the NRC say this is vulnerability aspect, but it is not L
23 a safety issue because it is on the outside, because what we 24 : said, we should probably buy a third transformer and we looked h
25 j at what is the best way to use that third transforner. We lI l 4 !
I l HERITAGE REPORTING CORPORATION -- (202)628-4F.83
126 1 felt it was best to hook it up aa an active spare versus just ll 2 having it sitting thoru as a spare.
3 We are an older plant so battery life is and an 4 issue.
5 DR. KERR: What is an active spare?
6 HR, NEWTON: Active spare means that your hardwire 7 it into the plant as an alternative path into the plant, and I 8 have got a diagram that I can point that out to you.
9 Because we are an older plant, battery life we have 10 l to approach, and with Station Blackout, scoping of batteries 1
11 ;
and DC power is an issue, that we are trying to settle on what i
12 h to do about that so we have to roll this into the Station e
l 13 Blackout issue. We have had problems, as other plants have, 14 with their inverters failing and when they fail, they disturb 15 hI the instrument buses, and we are looking at static inverters L
16 1 that transfer to another AC source to supply the instrument 1
17 l bus, h
18 l The question is where do you transfer to? So we U
19 p need to come up with a reliable supply of where we have all of d
d 20 y these transfers. Our plant, we have eight major inverters b
21 n that we have uriving the two units, the four instrument buses a
i 22 g on a corporate-wide basis and for fire protection insurance 23 ' purposes, the company has looked at geting rid of PCB 24 . transformers. We have some in our plant. They are also on 1
25 safety and non-safety equiprent. We don't have a lot, but h
c HERITAGE REPORTING CORPORATION -- (202)628-4388
l 127 1 they are there. If we just replace the fluid in them, you 2 tend to downgrade the capability of the transformer so you 3 have that capability to do that or do you have to physically 4 put in a new transformer?
5 And we have increased our capacity of water 6 treatment plant and in that process we have added a water 7 storage tank of 400,000 gallons. We are saying that while we 8
should tilso get benefit for that storage tank for decay heat i
9 re noval purposes if we can design it properly, with the right l
10 pedigree, so we are looking at benefits from that 400,000 0
11 J gallon tank in decay heat removal space which will also 0
12 L benefit us in the severe accident policy. So those are items 13 outside of the regulatory issues that we have going on in our 14 plant that we also are trying to integrate.
15 : (Slide) 16 h[ HR. !iEWToli:
What does that result in? Well, this b
17 L results in the following proposed modifications. We will be h
18 L proposing to install switchgear which bypasses the switchgear h
19 ] room and provfdes an alternative path to safe shutdown 1
o 20 ); equipment, ar.d in essence it is another path to keep your 1r 21 plant, and then you look at safe shutdown of equirnent, that 22 j pretty ruch is equipnent needed to remove decay heat. ;
e i I
23 l We have started the process of modifying our 133.
24 e We plan to install a third, and we had to make a decision on h
n 25 j what pedi. gree do we put on it? We have elected safety-related s n 1
n i
d HERITAGE REPORTI!!C CORPORATICtl -- (202)628-4888 l
128 1 emergency diesel. This is to address Station Blackout. This 2 will be, have it own switchgear, its own starting batteries, h
3 its own cooling system, and its own building, so in essence it 4 is a diverse alternative source.
5 DR. KERR: Why did you decide on safety, on making 6 it safety grade or whatever?
7 HR. NEWTON: The safety grade part it was pretty 8 much an economic decision for our plant. Because we have two l
9 l dioscis for two units, our tech specs say that if one diesel I
10 is down for more than seven days, both units are shut down.
11 DR. KERR: Thank you.
12 HR. NEWTON: We want to be able to use that as 4
l 13 l alternative to either of those two.
14 HR. DAVIS: Do you have an estimate of what that 15 l modification will cost?
i 16 j HR. NEWTON: We have an estimate, a budgetary 17 j estimate. We have not gone out looking for a diesel yet. He f
18 also are going to be approaching the NRC with respect to how 11 h
19 q much of a pedigree do we really need to put on this third I!
20 diesel but still get credit for it? In other words, do we e
21 h need a QA diesel or can we in easence go out and buy I
22 j commercial grade but show it is in essence the same and we 23 don't have to psy f'.r twice or three times the price.
24 We may br- approaching the NRC to see if we can do it 1
25 more economically. We have budgeted on the more econom4c h
h HERITAGE REPORTING CORPORATION -- (202)628-4888
129 1 basis, and we are saying that this is around $8 million, but 2 it wilJ be in the range of eight to ten. If we have to go a 3 full-blown everything else, it will probably double that 4 number.
5 HR. DAVIS: What size is that?
6 HR. NEWTON: It is about 3 megawatts.
7 CHAIRMAN WARD: Haybe you better pick out just one 8 or two of these items.
9 HR. NEWTON: I think the other one of interest is 10 the seismic makeup water storage tank. We had to make a I
11 ' decision of do we want to make it just a tank or seismic tank?
12 We have, pretty much have said let's make it a seismic tank, 13 although that is initial investment. Now we can get more 14 credit for it in risk, in safety space later on if we have to 15 and we have to make a decision on that now. So that increases l
16 the cost, but it is something that you can only do up front, 17 so that will increase--instead of having 10,000 gallons 18 available per tech specs, we will have 200,000 gallons 19 available or something greater on a 400,000 gallon tank, and i
20 j we just have the other facilities that connected to the P
e \
21 N existing plant, i 22 That was our proposed modifications, j !
23 l DR. KERR: When you say seismic tank, does that mean) l 24 l a seismic up to your SSE, or what?
l' l
i 25 ll HR. NEWTON: We have not got the number that we are h
1 HERITAGE REPORTING CORPORATION -- (202)628-4838
130 1 looking for. Obviously it will be the SSE, and this question 2 of will we make it a what we call a standard safety grade SSE 3 design with that, that additional margin on it, or will we 4 pick some other number and do it another way, we will try to 5 look at exactly what it gives us.
6 The benefits covered, we covered them along the way, 7 but we plan to increase the reliability of the emergency 8 diesel generator system, and therefore the overall reliability 9 of the plant, and this is our reliability which then also 10 translates to the risk sspects as well. We take care of both 11 l at the same time.
1 12 j We plan to use these modifications to addresa both 1
13 A-44, which is Station Blackov*, and A-45, decay heat removal, l
14 or the severe accident policy. That will turn up in our 15 l opinion the same outlier high risk areas because we know we I
16 !! don't have much condensate storage capaullity there. We know f
17 I we are weak on the electrical supply with our particular
!i 18 [ plant, and this study showed the same thing.
19 CHAIRMAtt WARD: Are you using or have you thought 20 l about using the ISAP process? Are you familiar with the ISAP?
l L
21 '
HR. F.EWTON: No.
22 CHAIRMAN WARD: Integrated Saf ety Assess: tent n
23 Program? Is that what you call it?
l 24 HR. NEWTON: I think that is rore a back-look at do 25 plants meet current regulation, isn't it, versus a risk based 4
HERITAGE REPORTING CORPORATION -- (200)628-4888
131 1 analysis?
CHAIRMAN WARD: It is a means of using risk based lh 2 3 arguments in deciding how you are going to implement a whole 4 plate of different issues such as you have here.
5 HR. NEWTON: For example, we did, as an adjunct 6 analysis to what AIC did for us or for EPRI in our plant, we l 7 asked them to additionally model the electrical supply into 8 the plant, the aux feedwater into the plant, to make decisions 9 l on should there be other things that we want to do? The major l
10 decision that we were looking at is should we add a third and l !
11 h a fourth diesel, or just the third diesel? The study showed a
12 b that we got just about all the benefit from just adding the 1
l 13 third diesel and the fourth diesol really didn't buy us much 14 so we are really only planning to add one additional diesel t
15 versus two there. That standard point, we used that l
H 16 h methodology to help us make the decision.
l 17 h We also looked at the reliability of the aux o
18 L feedwater system, and determined that the water supply to thatl i
h 19 was the biggest contributor to risk. Therefore, when we fixed 20 h that problem, there was no need to do anything else with the !
d 21 system because the third diesel took care of the electrical i i
22 U concern part of the aux feedwater system, so we did use that i
23 as a basis for making some decisions on what, the i 24 '
modifications to make. ,!
25 [ DR. KERR: Was that decision heavily influenced by {
t i i
d HERITAGE REPORTING CORPORATION -- (202)628-4388
132 1 what one assumed about the common mode failure of the diosols?
ll 2 HR. PARKINSON: I am not sure about that.
3 HR. NEWTON: I don't think so. And with respect to 4 what we have set up here, we may buy another diesel that is 5 quite similar, yet it is going to be separate in all of its l
6 l support features, so common mode failure, from a system i
7 l
standpoint, we won't have it, so you just are down to common h
8 mode failure from a component wh.ch is not too likely.
9 ]j CHAIRMAN WARD: Okay. Thank you very much. Our h
10 t next speaker is Mr. Ericson. Dave, we have gone 16 minutes 11 j into your time.
12 . MR. ERICSON: That's all right. We will still l 13 finish on time. They have already said everything I was going 14 , to say, 15 J HR. WOODS: We have got 15 ninutes on summary. I 16 don't think we will need more than three or four minutes.
17 HR. WARD: I think there will be considerable i
13 l pressure to break for lunch at one. l t
4 19 HR. WOODS: I feel it myself! 4 I
l l 20 L HR. ERICSON: I said somewhat jokingly but also j l
21 . somewhat seriously I think Bill and Jack have covered most of ;
i 22 the areas that I wanted to mention as to why there were 23 diftetences or what was dene.
24 I would like to recap by just connenting briefly on i l
25 some of the factors. A small percentage of the differences is l i
l i
HERITAGE REPORTING CORPORATION -- (202)625-4888
133 1 really due to plant changes, new batteries that were installed 2 that we didn't have, that sort of thing.
gg 3 Most of the differences or changes relate to 4 methods, to assumptions, to data, and they have talked about 5 these at some length. I think one of the things we would like 6 to have you keep in mind, and those of you who have been 7 involved from day one will recall this--we set out to do a i
8 l relatively conservative based somewhat generic analysis. We i
9 i did not set out to do a very plant specific realistic best l
l 10 y estimate analysis. They did. We didn't. So there are bound 11 1 to be differences in detail and differences in interpretation.
h 12 n I think under assumptions we would say that we often L
l 13 made an assumption where they have gone back and dealt with 14 d the procedures in detail, with specific things in detail. In h
15 L the regulatory analysis, of course, it also does point out 16 J that these things were examined only in the context of Point 17 Beach. One again has to be careful about generalizing, 4
18 althovgh I guess I would say certainly the S2 argument I think 19 ,
are very clearly a general thing. We would agree very l I
20 I directly. l I
21 I think one of the interesting aspects of the-- j 22 CHAIRHAN WARD: What is the S2 argument? ,
l 23 L MR. ERICSON: Small break frequency; I don't think l 24 ,
there is any question about that. I think an interesting W 25 aspect of these results if you look at the original case f
H h
HERITAGE REPORTING CORPORATION -- (202)62S-4888
134 1 study, and then the revision, and I guess I would like to make 2 a semantic point--my colleaguan have frequently referred to gg 3 the revised case study. There is no such animal. There are 4 set of revised numbers that were, were, the sequences were 5 edited, but the case study was not revisited and revised I 6 will guarantee you, but the interesting thing is that if you 7 look at the way these broke out between internal and external 8 events, originally we had about a little over 40 percent l
9 internal, almost 60 percent external.
I i
10 t The EPRI and EPRI WOG analysis comes out to 25 b
11 4 percent internal, 75 percent external, and even in our, in the 12 J revised numbers we used where we did not take credit for as 1
l 13 hmuchoperatoractionastheydid, it comes out about 28 and 14 72.
4 15 p Now I think Gary made the comment here a little bit t
I 16 I ago that external is going away, and yet even at this number.
L 17 they still attribute 75 percent of that to external causes, t
i l
18 HR. HAUGH: Principally seismic. !
'l d
19 HR. ERICSON: I think, I think Bill covered all of 20 i the key differences here. The reduction in the frequency of l o l 21 the small break, I don't know whether you mentioned, as we 22 did, we did take, they provided some very solid data on ,
23 removing the component cooling water dependency for injection l l
l 24 which cranked in. It isn't a big contributor, but it i l
25 certainly is there, and again, then this discussion about f L
h' HERITAGE REPORTING CORPORATION -- (202)626-4888
135 1 operator action. We simply at this point have retained the 2 more conservative approach.
3 I think Bill has visited the T3, T30 at some extent.
l 4 We simply have kept the more conservative approach. The 5 arguments they have given you are certainly to be considered 6 ,
in very careful detail, and if one were going to sit down and I
7 l redo this, this very easily could show some change I think for l
8 l us.
l 9 !
(Slide) 10 DR. RERR: Do you really think it is lE to the minus 4
11 seven and zero?
i 12 HR. ERICSON: No.
13 DR. RERR: I don't, either.
14 4 CHAIRMAN WARD: They didn't find anything that you h
15 hhaddonenon-conservatively, did they?
L 16 [ MR. HINNERS: de had omitted something.
n 17
[ HR. ERICSON: This is one sequence that comes in, it 18 " is ten to the minus seven sequence as I recall, isn't it. l 4 i 19 Il Bill? TlO or something that did show up. I i
20 HR. PARKINSON: There were a few things, but none of L
21 g them were risk dominant.
22 y HR. ERICSON: I think Bill did mention this, and it 1
23 is, it is you need to keep it in mind, they did rot go back f 24 g and start with a, a classic clean sheet of paper. They have, -
25 h I mean they have, they haven't pretended. They have stateditl; i
[
HERITAGE REPORTING CORPORATION -- (202)628-4888
136 1 right up front. They took our fault trees, and excuse me.
2 Hanipulated them--play is not the right word. But with the 3 code that SAIC has, they are able to do that, and we have 13d 4 some long discussions about that, and it is a very effective 5 way of doing it, but if we left something out of the original 6 fault trees, they did, too, because they did not go back and 7 attempt to reconstruct all of those.
8 j Por the TlHLE, we have--it is new battery issue. We l'
9 j did not go back and attempt to resolve where that is, but we 10 i tentatively have accepted that. Clearly that they play a 11 major role in how you would handle the Tl sequence.
12 (Slide) 13 HR. ERICSON: For loss of feedwater, again, we have 14 hmaintainedthe--at this point we have simply left in the i
15 conservative assumption due to the PORVs, and that was 16 U primarily because of the data variability. You have to 17 0 renember we did this analysis, we did our plant visits and i
18 1 this sort of thing in late '84, very early '85. When we had d
19 i our technical experience, the data they presented was for '86 20 and '87, so the word we got when we visited the plant is, and I l
u 3 21 [ it is almost a direct quote, is yeah, they are blocked most of '
L 22 L the t i rt.e . And that's why we did it that way, t
23 ; Again, the CW dependencies come out.
l 24 HR. WOODS: For those people who don't speak the 25 L language, as I don't, in the back of Appendix D to the reg l h i U
U HERITAGE REPORTING CORPORATION -- (202)628-4888
137 1 analysis, the last page is a translator. T1HLE is loss of g 2 off-site power transient. I made him write it down for me so 3 you have got it.
4 HR. ERICSON: It is also in the set of vugraphs.
5 Those around the table, we have been talking this shorthand 6 for more years than any of us would care to admit, so I think 7 we are all on the same track.
8
! There was a question with the T2, the loss of d
9 l feedwater, about how the recovery differs, so we did not make 1
10 ;
the complete reduction, but again, E to the minus 7, 1E to the 11 j minus seven, is the difference significant?
4 12 h The others, the remaining HLE sequences and one l 13 other sequence are very low, but we chose to keep as a 14 [ conservatism the loss of bus leading to a trip. They have f
I' 15 l presented discussion in 113 that says that would not occur, a
16 ] For loss of long-term Station Blackout, I think we L
17 j have discussed everything here. We sinply at this point have 18 [ not agreed to the CST refill and some of the long-tern J l 19 q operator action, but clearly if one goes back and looks at L l i
o 20 h that recovery of off-site power, there ispotentialfor--thosel D
21 [ recovery times are getting very good at many of the plants, so l
22 [ it would be much lower number, or significantly lower number, j l
23 '
With respect to seismic, Jack covered that pretty l 24 . quickly, and I think he covered everything that we have, that ;
I 25 j is on here. The keep points being one, the hazard curve, and
!l 4
o HERITAGE REPORTING CORPORATION -- (202)628-4688
i l
138 1 I think when we had our meeting in March it was agreed that 2 that would await the seismic program that is now ongoing so we 3 did not pursue that any further.
4 Thest we disagreed on some of the post-recovery tines 5 and what night happen, and an item of technical disagreement 6 is the survivability of their particular RWST, but as Roger 7 has pointed out, there--and Jack, there are ways around that 8 that are already instituted. the drawing water from the spent 9 l fuel pool and this sort of thing.
I 10 h CHAIRHAN WARD: This issue, I hate to, don't like to i! !
11 hdismissthehazardcurvetooquickly. I d
12 HR. ERICSON: We are not. There's a whole ongoing
]
13 discussion between industry and the staff on what, what the 14 )curvesare for the various regions, and rather than for this l
15 4 group to attempt to resolve which is the best one for Point 1
1 l 16 <
Beach, at this point we will disagree. Obviously we believe 1
i 17 " we had a reasonable approach to the, constructing that curve.
l l 18 CHAIRHAN WARD: I guess I am not familiar with the l
19 staff. Can somebody surtmarize what the progran is?
l 20 HR. BECKNER: THIS is Bill Beckner from the staff.
21 There are a number of seisnic programs. Some of them are this 22 and t..at and the other, and I won't get into alphabet soup, l 23 but there are a large number of t h e:t . j i
24 We are stepping in with the IPEE which is another I
25 seistic program, and so right now, there is an effort to unify HERITAGE REPORTING CORPORATION -- (202)62S-4888
139 1 all the seismic efforts, and I think that effort is both on gg 2 the staff side and NUMARC side, but suffice it to say there 3 are a number of seismic issues, A-46 and A-17 in one of them, 4 and the eastern seismicity, all those programs I think 5 are a conclusion, and we are going to be trying to 6 comei a program that has industry look at seismic once.
7 And that's or. going with the staff as part of the IPE for 8 external events. We are trying to unify those programs.
9 CHAIRMAN WARD: Okay, but did I understand that one 10 product of this is going to be a, an agreed-upon, thoughtful 11 seismic hazard curve?
12 HR. BECKNER: I believe the staff has been l 13 discussing that for a number of years, and they are coming to, 14 hopefully they are coming to some conclusions.
15 HR. NEILS: Mr. Chairman, let's say the eastern 16 seismicity program, the staff, I think reality would suggest 17 that that is the critical program from which agreed-upon 18 seismic hazard curve site-specific would come, and it appears 19 that any differences which remsin between staff and seismic 20 owners group are likely to reach closure before the end of 21 this year.
22 CHAIRMAN WARD: Okay. Thank you.
23 HR. ERICSON: The fire issue, clearly no question
- 24 about adding a second nalon system in the aux feedwater room
W 25 we do disagree un the failure rate data for halon. My reading HERITAGE REPORTING CORPORATION -- (202)628-4888
~
140 1 of the DOE record that they cite says there is a significant 2 difference in that data. A number of fires are cabin enclosed 3 fires, computer related fires, with halon in small cabinets.
4 Jack says a room is a room is a room, and I think I disagree.
5 A cabinet is not an aux feedwater pump rrom. I Think this is, 6 there is a difference, and also I, I read the purpose of that 7 report somewhat differently.
8 It looks to me like the purpose of that report is 9 primarily to document how much flurocarbon has been released l
10 by DOE facilities into the atmosphere, 11 HR. HAUGH: May I respond? The particular report in 12 question is a report of ongoing surveys of fire protection at 13 the DOE facilities, and again in the early 1970's and it 14 started with water sprinkler systems and it has continued and 15 very recently they have added the halon systems to the whole 16 program, and document it. Indeed the report centains n.uch 17 discussion on the release of flurocarbons because t.hi s 1:; 3 18 I concern as we all knew for the Monte protocol, but that is not 19 merely a document that can be so characterized as just trying 20 to count up pounds of freon or halon.
21 C3AIRMAll WARD: Okay, but the 17 events are the, you 22 know, the systems or the spaces in which those 17 events 23 occurred, you know, really representative of the application 24 you are putting them through?
25 HR. HAUGH: They vary according to each individual HERITAGE REPORTIliG CORPORATIO!! -- (202)628-4888
141 1 fire occurrence that took place. Indeed some of them did lh 2 involve computer cabinets, and I don't--
3 CHAIRMAN WARD: Did any of them involve large rooms? I 4 HR. HAUGH: Yes--suppressing fire involved with the 5 explosion protection systems on vertical mills, some of the 6 machine areas, for example.
7 CHAIRMAN WARD: Okay.
8 HR. ERICSON: Well, just to close out, the ones that 9 were not discussed yet today, with the wind, the principal 10 problem there was the diesci stacks and that modification is 11 already in place, and again we are talking about differences 12 of order cf magnitude, but a number so small it doesn't really l 13 contribute much to the risk.
14 Internal flood, this was cae that we had a 15 substantial contribution. In the initial version of NSAC 113, 16 or in the NSAC 113, they get a much smaller number.
17 Considerable discussion revolved about the use of so-called 18 Thomas correlation for pipe failures. As we went back and 19 looked at that, we said yes, that's probably a reasonable 20 thing tu use. It is being used in other PRAs now, and c the 21 present state of knowledge, there is, there is no reason not 22 to use it.
23 We were entirely too conservative in our original 24 estimates of the failure frequency. We still have some
}
25 disagreement about how much pipe is involved, and what is the HERITAGE REPORTING CORPORATION -- (202)628-4888 j
142 1 proper length to use for, to provide line of sight, but it 2 does reduce that contribution of originally nearly almost ten g
3 t he minus four down to about ten to the minus six, so 4 s.gnificant change when we revised the frequency estimates 5 using the other correlation.
6 CHAIRMAN WARD: Could you go back to the one on wind 7 for a minute?
8 HR. ERICSON: Yes.
9 CHAIRMAN WARD: I am trying to get something a 10 little more general from this. But is the, is the reason for 11 your change because of the, just because of the modification 12 or--
l 13 MR. ERICSON: Primarily the vulnerability was the 14 loss of diesels because the stacks collapsed on it. Now as I 15 recall, we discussed at some length a quick fix was to go cut 16 the stacks off, and I think we credited that in our analysis, 17 that you if you lose the diesel because the stack had fallen, 18 stack has crumbled in the wind, then it is a simple matter to 19 go out with a torch and cut it off.
20 However, tor other reasons, I believe if I'm 21 correct, Roger, you had strengthened that stack, those diesel 22 stacks, and so we said well, clearly if they have been 23 strengthened, in other words, they have installed modification 24 that we had proposed in the case studies, the number therefore w 25 is down, the situation we analyzed no longer exists, and the HERITAGE REPORTING CORPORATION -- (202)628-4888
143 1 only other thing I had was to put up a summary of them one ll 2 more time, but I think, I think both Jack and Bill have 3 touched very specifically on the key differences, but the 4 point I wanted to make is that they indeed set out to do a 5 best estimate very real estate calculation, and we did net at 6 the time we did this.
7 HR. DAVIS: Dave, question that troubled me a little 8 bit--I thought that I read in some of the earlier 9 documentation on resolution of this issue that the studies you 10 were doing were supposed to be realistic estimates, not il conservative, and now you seem to be saying that your results 12 are conservative and even your revised results you think are l 13 still conservative.
14 HR. ERICSON: They are probably conservative, yes, 15 but I think, I think I have stood before this group on more 16 than one occasion and said we were doing a generic f
17 conservative analysis.
18 HR. DAVIS: That is not what the documentation says 19 for this effort that I have road.
20 Warren, do you still think the revised result is 21 conservative, or is that an unfair question?
22 HR. MINNERS: I don't really think I'm qualified to 23 answer. I guess if I had to take a decision on it I think it
] 24 is more on the conservative side. I--but what you have got to 25 renember is what we are talking about, you are talking about HERITAGE REPORTING CORPORATION -- (202)628-4888
144 1 how that particu?ar number was calculated. Okay. That's h 2 probably conservative. Estimate of core melt frequency may 3 not be conservative because there is lots of things left out 4 of that number. Errors of commissior are not included in 5 that, and that's what caused TMI, that you want to go back to 6 operating history.
7 DR. KERR: But Warren, you seem to be saying if you 8 l don't know to do the PRA, then you do the part you know how to 9 do and you do it inaccurately. I mean I don't understand 10 this. It seems to me you ought to do a PRA as well as you can 11 do it, especially if you are going to use it to make 12 decisione.
l 13 MR. ElINNERS: I don't think Dave is saying anything 14 different from that.
15 DR. KERR: He is saying I thought that it is 16 deliberately conservative.
17 MR. MINNERS: Because you have to make a choice when 18 you are only given a certain amount of money how much you can 19 do, and that's his limitation. They could have done a better 20 seismic analysis. Okay. We couldn't give him the money.
21 DR. KERR: Conservatism coesn't mean that. I mean 22 to me, conservatism means that you deliberately make something 23 worse than you think it is.
24 MR. ERICSON: I guess I am using conservative, Dr.
g 25 Kerr, in the sense that we used generic values. Therefore, in HERITAGE REPORTING CORPORATION -- (202)628-4888
145 1 some cases those may be for specific plants optimistic. They 2 may be on the non-conservative side, but in general, I think 3 the generic terms tend to be conservative with respect to what 4 you might find in a specific plant if you really dug into it.
{
5 HR. MINNERS: The fire analysis was simplified of 6 necessity, okay. And when that simplification was made, it 7 was done to try to be sure it was on a conservative siae 8 because you couldn't get the best estimate. You had to do 9 something simple so something simple gave you a more 10 conservative answer. And I think I agree with Jack, is that l 't you can spend everybody's resources and do a very good 12 analysis, but--
13 DR. KERR: You are going to be spending a tremendous 14 amount of the country's resources doing backfits if these 15 results tell you you have to do backfits, so you know, we are 16 making decisions that could--I mean maybe it is going to be 17 decided nothing needs to be done, but if somebody should make 18 a decision that rather significant changes in plants needed to J.9 be done, that would be extremely resource intensive, and it is 20 not, you know, it isn't, I don't understand the resource l 21 atgument entirely.
22 HR. HIN!!ERS: That i t, because you don't work with 23 the staff. Maybe the ACRS is given all the money they asked 24 for, but we aren't.
25 DR. KERR: If you don't have the resource to do a HERITAGE REPORTING CORPORATION -- (202)628-4888
146 1 job, you go to the Commission, you go to them and say we do ll 2 not have the resources to do this.
3 MR. MINNERS: I don't--all I netd is enough 4 information to make a decision, okay? I think the information 5 that was provided to do this thing is far above the i
6 information provided for most decisions, and well adequate for 7 making that kind of decision. Don't have a problem making the 8 decision.
9 Now I think again that even if we spent more money 10 and did a best estimate in every respect, okay, you would 11 still come back and say what about the uncertainties? So you 12 know, you are just getting to diminishing returns when you try l 13 to get these studies to perfection and you have to look at the 14 studies and make some estimate, yes, you do, have to make an 15 engineering judgment where these studies are conservative in 16 them or not.
17 DR. KERR: It is different, a simplified analysis 18 and being deliberately conservative it seems to me. What I 19 thought I was hearing was that this was a deliberately 20 conservative estimate.
21 HR. HINNERS: He did the simplifying. He said with 22 this analysis I csn't get best estimate because I'm not doing 23 it in a detailed enough manner. It is like the fire study.
24 They went further on. They got a better estinate than we g
25 probably did because they did a fancier analysis, so when we HERITAGE REPORTING CORPORATION -- (202)628-4888
_- ~_-
147 1 did the study, we said hey, we are quing to do a simplified ll 2 analysis, costs less money. We are going to be sure we are on 3 the high side, and not on the tow side when we do this IPE, 4 which I think is a reasonable judgment.
5 Now your question is how much on the high side are 6 we? I don't know.
7 DR. KERR: That is not my question at all. My 8 question is why you tried deliberately to be on the high side?
9 PRAs are supposed to tell you given the resources and 10 information, what is going on.
11 HR. HINNERS: I don't think it was possible to do an 12 analysis in the limitation we imposed on the contractor that l 13 gave you quote, best estimate. There are limitations in time 14 and resources. I am as dissatisfied with that as you are, but 15 I think that's what the decision-maker is faced with.
16 DR. KERR: There is another facet of this procedure 17 that puzzles me a little, and it is the first reanalysis that 18 I ever remember having seen at which all of the risks came out 19 to be lower than the first analysis, and violate the Davis 20 theorem!
21 HR. WARD: You don't understand the Davis theorem.
22 It depends on who is doing the reanalysis.
23 DR. KERR: That's a corollery perhaps, but I 2.' recognize that corollary.
25 CHAIRHAN WARD: And a subset of that has to do with HERITAGE REPORTING CORPORATION -- (202)628-4888
148 1 generic data. Wherever I see a specific plant analysis, they gg 2 always seems to use specific data for the plant where those 3 data are more, well, more favorable.
4 HR. DAVIS: They are always above average.
5 CHAIRMAN WARD: Everybody is above average except 6 where they cre average.
7 HR. CATTON: What does generic mean?
8 CHAIRMAN WARD: Its easy to get cynical about this.
9 HR. MINNERS: We think the numbers are presented to 10 cover the range of probabalistic values I think we say 11 somewhere, so that would say our number is, your number is 12 probably low. I think error probabilities of operating is l 13 something currently and only making a mistake one in 10,000 14 times, high and low; it is certainly beyond my capabilities.
15 HR. DAVIS: I would agree with that I think, but 16 Warren, I am not sure I accept your premise if you have 17 insufficient information or resources, that it is always 18 necessary to use a conservative number. I think you can use, 19 still use best estimate, best engineering judgment, e.nd that 20 doesn't necessarily mean that it is conservative.
21 HR. MINNERS: In the example of the fire study, I 22 think it was a reasonable approach. The next step to get a 23 better estimate would have been to put some of the conditional 24 probabilities and manipulate some of the other data which we i
25 didn't do because it would have cost--maybe we could have done HERITAGE REPORTING CORPORATION -- (202)628-4888
149 1 it, but we didn't.
2 HR. DAVIS. That doesn't mean your first try was lh 3 conservative until you deliberately made it that way.
4 MR. MINNERS: I think it was deliberately--I think 5 people were doing simplified analysis. They tried to make it 6 conservative because they didn't know how to make it a best 7 estimate. If we could have made it best estimate then we 8 would have done it.
9 CHAIRMAN WARD: I think, you know, there is kind of 10 an engineering cultural difficulty here, whenever engineers 11 simplify anything, you always simplify the in conservative 12 direction. We have got a new something we are doing here l 13 called PRA, and we can't, people really can't get out of that 14 mode of thinking. Maybe they shouldn't. I don't know. Dr.
15 Kerr thinks they should.
16 MR. CATTON: I always like to approach the solution 17 from above.
18 CHAIRMAN WARD: I'm not sure--can you explain that?
19 MR. CATTON: Sort of like a value problem. You keep 20 adding more terms to the matrix and the value keeps getting 21 smaller like the stability problem. There are lots of 22 theorems, too. They sound the same.
23 CHAIRMAN WARD: Okay. Well, we have spent a good 24 bit of time comparing these two analyses, and there is, you 25 know, that come up a little, not very much closer together, HERITAGE REPORTING CORPORATION -- (202)628-4888
150 1 but what, what, can somebody suggest 'that we are supposed to ll 2 conclude from this?
3 HR. MINNERS: Well, I think the bottom line is that 4 we think that real value lies somewhere between the two 5 estimates.
6 HR. WOODS: The real value was already included in 7 the analysis. There are things that aren't included.
8 CHAIRMAN WARD: You mean--and you conclude that 9 because your analysis has been purposely biased in the 10 conservative direction and you think their analysis is 11 purposely biased in the non-conservative direction?
12 HR. HINNERS: Yes.
l 13 HR. KNIEL: The real value lies in the fact the 14 analysis has shown that there are, that there appear to be 15 cost-effective improvements that can be made in the plant, are 16 being made without even pushing us at this moment, that have 17 shown up by doing that analysis, showed the value of doing 18 plant-specific analysis.
19 CHAIRMAN WARD: Value of the IPE process, not 20 quantitatively.
21 HR. KNIEL: Let the plant decide. There are some 22 appropriate things they might want to do. That's been 23 demenstrated in spades by this, by this.
24 HR. MINNERS: Don't put it in terms of who won and 25 who lost. It is not that kind of game. The game is to try to HERITAGE REPORTING CORPORATION -- (202)628-4888
151 1 identify plants, see if we can make corrective actions. I 2 think Point Beach is doing a lot of that and that's the whole 3 purpose of the exercise. I applaud them for doing that.
4 liow whether they have got to the goal or not is kind 5 of irrelevant. We are not playing that kind of a game. We 6 are trying to improve the safety of the plants in 7 cost-effective methds. The only, the reason that we put some 8 emphasis on these numbers is that because there was, the 9 original question that was asked, fairly asked was is there a 10 problem? And I think that, I think that the analysis shows 11 there are problems, and even the EPRI, the Westinghouse Owners 12 Group analysis showed that they made some changes, said there 13 are vulnerabilities that should be fixed, so the thing to look 14 at is try to identify not the bottom line. Let's not 15 concentrate on the bottom line.
16 HR. IISILS: I think that one thing that has been 17 clearly shown which neither the Sandia folks or the staff nor 18 we have any disagreement on, and that is that one, the manner 19 of the generic modification does not seem to be warranted here 20 because we don't know what it could be, and the one which was 21 considered, that is the dedicated add-on decay heat removal 22 system, is just profoundly not justifiable in accordance with 23 the rules we are working with.
24 CHAIRMAll WARD: It is more profoundly not justified 25 if I accept your analysis than if I accept the Sandia HERITAGE REPORTIllG CORPORATIOli -- (202)628-4888
152 1 analysis.
2 MR. NEILS: True, but get on to that level, we have l
3 agreed before we got there, okay.
4 The other thing I think has t=en illustrated here, 5 and that is that since the case studies sily are a modified 6 form of, say a shortenad form to SER, to some deg'ree of IPE, 7 and extended form to some degree also of an IPE because the 8 consideration of external events, but I think it, that it has 9 been illustrated that this is a good way, useful tool to catch 10 plant unique exceptions which may exist that haven't been 11 illustrated as yet in case studies, and Point Beach is one 12 utility that has illustrated what they are doing with it.
l 13 Now we hadn't really intended to alter IPE work in 14 the process here, but that turns out to be a benefit to them, 15 and I think I can speak for our own utility on whatever units.
16 We are about I think three fourths of the way through with our 17 IPE, and we are doing the same kind of thing. We don't even 18 know what the target is yet.
19 MR. NEWTON: Mr. Chairman, I had one more comment 20 with respect to the utilities selected and possible fixes.
21 The original eight. utilities that were selected for the study 22 were done by an overall survey of plant designs with respect 23 to decay heat removal, and that survey in itself selected g 24 plants they thought were somewhat vulnerable and they then, 25 analysis of the vulnerable plant, they passed more newer HERITAGE REPORTING CORPORATION -- (202)628-4888
153 1 plants. If you look at the age of plants selected, typically g 2 they are the older plants, Point Beach, and so on. So I think 3 the vulnerabilities we found in our plant are somewhat typical 4 of the age of the plant, but it also shows that the process 5 works.
6 I am saying this because if the new plant goes 7 through the same study, I would expect them not to find the 8 same things that we turned out because they are designed 9 differently. They have more storage capability for the kind 10 of state storage tanks than Point Beach has. They have the 11 rate that we don't have because when we were built we didn't 12 have those requirements, and so on down the line.
l 13 They have more batteries than we had to begin with 14 in the first place. They separated lE and the non-lE 15 batteries. You go down these items that as we move through 16 the '70s, the implementation of the regulations and the 17 expertise of the regulators on how they wanted it to be met 18 became more precise, and the plants built in features that 19 addressed these better than the older plants do, but the main 20 point is that it is working for the older plants, so it is 21 also saying that the older plants don't need the dedicated 22 system, but there are some areas that they may need to improve 23 on. New plants probably won't have to make many changes when 24 they look at it.
25 DR. KERR: Suppooe that you were buyino a new plant, HERITAGE REPORTING CORPORATION -- (202)628-4838
154 1 had some flexibility in buying a new design. Would you try to 2 get one with a dedicated shutdown heat removal system?
I 3 HR. flEWTO!!: You are asking my opinion on that. And 4 probably the new designs that I have seen have ways of dealing 5 with that aspect of safety that doesn't require a dedicated 6 system. I guess I would advocate those concepts versus a 7 separate dedicated system per se on what we are talking about, 8 adding on to our existing plants.
9 DR. KERR: Thank you.
10 HR. HIlillERS: I think you have asked the right 11 question, whether we have a sufficient basis to have the 12 industry spend resources, and that's the correct question to 13 have, an,d al} we are asking the industry to do at this point 14 is to do a study of their plants, so I think their resources 15 would be part of the, is commensurate with the basis we have.
16 The next step, to make the fix, will have to be judged at the 17 time we come up with those.
18 CHAIRMAll WARD: You know, it is popular to denigrate 19 the bottom line from PRAs, and say it is the journey that's 20 important, not the destination, all that sort of stuff, but 21 momehow, you know, I guess I can't see any of the, into the 22 mind of the Point Beach people, for example, why they decided 23 to do certain things, and not others. And I think unless 24 there is some sort of the pressure of a goal, of some sort of 25 quantitative or fairly specific concrete goal, it is not clear l
HERITAGE REPORTIllG CORPORATIOli -- (202/628-4888
l 155 1 to me how do you know when you see a vulnerability? And I lh 2 think you have--and that's what bothers me about the IPE 3 thing. You have done that. I mean whether we make fun of 4 your ten to the minus 5th or not, that is something out there 5 that people are going to test their judgment against, and I'm 6 not sure what the IPE process is going to substitute for that.
7 HR. KNIEL: I think doing the, your own IPE study of 8 your own plant you are going to see different sequences 9 leaving different numbers, and I think the individual 10 utilities aren't going to like some of the numbers they see 11 and they may see and they can connect some of those up and 12 make some simple fixes that would improve some of the numbers.
l 13 CHAIRMAN WARD: What is going to cause them not to 14 like certain numbers? What standards are?
15 HR. KNIEL: I think the standard is--
16 CHAIRMAN WARD: Some number?
17 HR. KNIEL: The numbers they look at are going to 18 be, you know, several times ten to the minus 5th or maybe ten 19 to the minus four or something like that. They are going to 20 look at numbers that are out of, out of place with respect to 21 other numbers that calculate for other sequences I think, and 22 they are not going to be happy with that.
23 HR. MINNERS: Under the rules we are playing by, I 24 think there are rational rules. D. is not so much what the g
25 goal is. It is how much safety you get for your buck. That's HERITAGE REFORTING CORPORATION -- (202)628-4888
156 1 what are the changes? And whether you reached the goal or not gg 2 is really not, is not the game we are playing.
3 CHAIRMAN WARD: Well, I don't think that that's 4 right, Warren. The safety goal is not a, for example, is not 5 a cost / benefit concept. I mean to the contrary. I mean the 6 safety goal says if this is good enough, and it is not 7 justifiable to spend one more dollar getting to that level.
8 HR. MINNERS: The safety goal is only a policy and 9 that was a rule.
10 CHAIRMAN WARD: What conclusion does one draw from 11 that?
12 HR. HINNERS: I follow the rules more than I do the l 13 policies. If they conflict, I follow the rules.
14 DR. KERR: I asked this question very care.tully. I 15 don't think it is clear that the staff does not believe the 16 safety goal is a rule.
17 HR. MINNERS: It sure is not.
18 DR. KERR: It does not provide any guidance to them.
19 I Apparently it doesn't.
20 HR. HINNERS: The ACRS says don't use it as a rule 21 and don't judge plants against it.
22 CHAIRMAN WARD: Well, you have mischaracterized what 23 the ACRS said.
24 h HR. HINNERC: No. If I have, I would like to W 25 understand what I have said wrong because that's what I think l HERITAGE REPORTING CORPORATION -- (202)628-4888
157 1 you said, 2 CHAIRMAN WARD: Well, I think the--I don't want to h
3 get into a long discussion of that. That's another subject, 4 but that needs to be, I wish the Agency was given a little 5 more priority to getting on with the safety goal 6 implementation because--
7 MR. VINE: Dave, can I ask for an opportunity to 8 make a couple of comments?
9 I would like to first of all comment on what the 10 overall conclusions were from the effort and what kind of 11 conclusions you might draw, and secondly on the safety goal, 12 and the whole question of numbers.
l 13 First of all, on the conclusions of the A-45 study, 14 I don't want to go over the points because I know you are 15 pressed for time, but if you take a look at the vugraph 16 package, the middle one, the thick one, that was primarily 17 handled by Jack Haugh and Bill Parkinson, the next to the last 18 slide there is entitled general comments on A-45 studies, and 19 I think this selection of seven points highlights what we 20 think are the major conclusions that wo drew out of our 21 experience of reanalyzing Point Beach as we were doing a very 22 thorough but more qualitative evaluation of all six case 23 studies, and I will reiterate just one or two of those points.
24 CHAIRMAN WARD: Would you mind sticking that up on g
25 tho--we do--have you got some wrap-up comments?
HERITAGE REPORTING CORPORATION -- (202)628-4888
158 1 HR. WOODS: It has all been said in the last ten 2 minutes. It has all been said.
g 3 HR. VINE: I would point out something that has been 4 said a numbar of times this morning, and most recently by 5 Gerry Neils, and that is that we have accumulated in this 6 industry now well over a thousand reactor years of operating 7 experience. And for those sequences of events weere we are 8 talking about the probability of initiator frequency, or the 9 probability of a component failure, that is down in the region 10 that is clearly evidenced statistically in that data base of 11 operating experience, we ought to be using the or rating 12 experience instead of ivory tower models that have no relation 13 to reality as the basis for our decision-making, and our 14 biggest problem in many of the areas of case study was where a 15 model was developed totally without regard to the operating 16 experience. And this took place many times, for example, in 17 the external events, lightning, for external floods, for wind.
18 Instead of going to the operating experience and asking how 19 often those things actually occurred, using that as the basis 20 for makiiin the decision, about what the risk levels are, we 21 created a model, an equation, and we really want to stress as 22 much as we can the importance of using real operating l 23 experience. Small break LOCA frequency, the number used there 24 ! says we have two snall break LOCAs overy year. That's not 25 true.
HERITAGE REPORTING CORPORATION -- (2021628-4888
159 1 The number on the frequency of very large fires that you draw out of the case studies, is simply not true. You lh 2 3 have got to berschmark the results of these analyses on what is 4 really happening in the industry, and that's one thing that I 5 would like to stress.
6 The other thing that I would stress one more time 7 and Dave mentioned that a large percentage of our risk in our 8 analysis as well as his was in the area of external events.
9 Again, I want to stress that our resulto enly showed seismic 10 l to be a significant contributor to that external event, and 11 that's a very tentative number based on our more or J ess 12 splitting the difference between what the NRC's position was l 13 at the time we did the analysis on the seismic hazard curves 14 and what the EPRI position was at that time. The two of then 15 are coming together. I don't know what the final number is 16 going to be. It is very likely to be I think a number lower 17 than what we had assumed 2or ours, which will be a l
18 conservatism in our seismic analysis, and therefore, basis for 19 I reducing that percentage contribution from external events, 1
20 but we would reiterate that all of the other external 21 events--fire, external flood, lightning and all of those, 22 l either based on this specific reanalysis, quantitative 23 reanalysis, or our very detailed review of how the case 1
24 studico modeled these, those external risks, are sieply not 25 l significant.
HERITAGE REPORTING CORPORATION -- (202)628-4888
I 160 ;
l 1 Now if you take a look at whert :s the l 2 perception of concern exists on external e. ants, you will find l
3 that a good share of that cane from a recently commissioned 4 and conducted study that the NRC did with, with Hister, Dr.
5 Buddance, and what his task was, was to do was to simply 6 review all the existing PRAs, find out what the contribution 7 from all of the external risks were, and then say does this 8 appear to be significant or not?
9 One of the major contributors to the Buddance study 10 was the A-45 program where he cited the numbers en external 11 flooding, on lightning and--not on lightning but on high wind 12 and tornado, some of the other major contributors, and so
! 13 these things perpetuate and we would argue that if we had our, 14 had our chance to go through all these other sequences like t
15 the external flood, that we would be able to show that those 16 i numbers are low, and I think the contribution and this 17 perception that all these external events were important would 18 go away down because we just don't believe there is any risk 19 there based on what we have seen in our basis.
20 DR. KERR: I must comment because I don't think, at 21 least our perception, certainly my perception of the seismic 22 risk does not come from budgets. It comes from a recognition 23 that we do not know very much about the frequency of large 24 lj carthquakes especially in the eastern United States.
25 f HR. VINE: I excepted seismic from that comment. I I
l HERITAGE REPORTING CORPORATION -- (202)628-4888
161 1 will talk about all the external risks other than seismic.
2 DR. KERR: I thought you said seismic.
g 3 HR. VINE: I would agree exactly with the 4 characterization we taard from the staff, that we are trying 5 to integrate all the seismic issues into one single analysis 6 so we will be done with it in one, in one effort, but for all 7 the other ones, it is not clear to me that there is any basis 8 in the A-45 program to say that we have to go out and do a 9 plant-specific fire analysis on every plant in the country or 10 a plant-specific external flood analysis on every plant in the 11 country. I think there is more than sufficient basis to say 12 there doesn't appear to be a generic problem there.
13 l HR. CATTON: How can you tell from plant data about 14 external floods? For example, you need real time to consider 15 a flood?
16 MR. VINE: You have to take a look at our letter 17 back to the staff a year ago on external flood.
18 MR. CATTON: I am asking--maybe I don't understand.
19 It seems to me that when you talk about a flood, you talk i
l 20 about a ten-year *lood or a hundred year flood or whatever.
21 That's over a period of real time. Your plant data comes in a 22 much smaller piece than that. You haven't had the history 23 time to come to that kind of a conclusion about external 24 ovents. You may have done fire. You have been talking about W 25 external floods, but certainly not external flood, wind and HERITAGE REPORTING CORPORATION -- (202)628-4888
(
162 1 those sorts of events, g 2 HR. VINE: We don't have the time now to answer 3 that. I would love to talk to you about external flood. If 4 you take a look at the--
5 HR. CATTON: I would like to hear your answer. I 6 happen to own a house in a canyon.
7 HR. VINE: We are talking about sequence in case 8 studies. We are talking about case studies sequences where, 9 for example, the Cooper assumption was that four dams would 10 successively fail on the Missouri River, and over a period of 11 three and a half or four days, all this water would come down 12 to the plant, and cause a core melt. The plant would be shut 13 down.
14 HR. CATTON: I understand the rationale. It is 15 based on analysis of how it was done betore, but not based on 16 e plant data because your plants haven't been around long enough 17 to make any conclusions about the external events.
18 HR. VINE: That's partially true for external flood, 19 but we can even get the initiator frequencies in the case 20 studies and show the frequency of flood is much lower than 21 what we assumed in the case studies as the initiator frequency 22 contribution to that risk, but I don't want to get off too 23 much on external flood. We don't have that much time.
24 I do want to stress that from what we have seen of W 25 the caso studios, in our analysis of Point Beach, all of those HERITAGE REPORTING CORPORATION -- (202)628-4888
163 1 other a.ternal events except for seismic are not significant 2 to the bottom line.
3 MR. CATTOll: You can't base it on plant history.
4 MR. VIllE: Part of that, you can't. You can for 5 lightning. You can partial 2y for the flooding, and you 6 certainly can for fire, and the, I guess the last comnent I 7 would like to make after, after having reviewed this slido, 8 has to do with the question of safety goals.
9 In the llRC prenentation, the number ten to the minus 10 5th was discussed, and I would like to make a couple of points 11 about that.
12 First of all, just to reiterate, that number first 13 appeared in the A-45 program after flSAC 113 was published with 14 its answer, what we thought the core melt frequency was.
15 The second point I want to make is I would agree 16 with Mr. Davis, and that is from everything we have seen from 17 PRAs to date, because the A-45 orogram has defined the decay 18 heat research to be all inclusive of everything, but with 19 respect to large break LOCAs and a couple of other very 20 incidental issues, that it is a large percentage of the 21 overall core melt frequency, and so we would say that as you 22 did that, the overall contribution from DRH is probably well 23 in excess of 50 percent and may be much higher than that of 24 that ten to the minus four starting point.
25 i The next thing I would 13ke to say is that if you do HEKITAGE REPORTIllG CORPORATIOli -- (202)628-4888
164 1 start with that three times ten to the minus 5th number, and 2 then cut that down to one times ten to the minus 5th, based gg 3 on--and this is going back to that, the vugraph--based on 4 operator error, and errors of could commission, a couple of 5 things need to be said.
6 The first thing is that operator error has already 7 been included in all the analyses of DHR risk, and second, to 8 assume that two thirds of all of the DHR risk is errors of 9 commission and other things that haven't been quantified I 10 think is, is a pretty large statement to make after all the 11 anal _ysis that has been done. That's a pretty sweeping 12 statement, and I would, I would guess that much larger 13 screen--
14 CHAIRMAN WARD: If I observe--
15 HR. VINE: THI, 16 CHAIRMAN WARD: THI-2 and Chernobyl I would say.
17 HR. WOODS: And WINDSCALE and SL-1.
18 CHAIRHAN WARD: They have had rather little to do 19 with the sorts of things that PRA tells us about.
20 HR. VINE: THI van not exclusively an error of 21 commission.
22 HR. WOODS: This was my summary. Thank you for 23 making my summary for me.
24 HR. VINE: Thank you.
W 25 CHAIRMAN WARD: Okay. Thank you, Gary. Could I ask HERITAGE REPORTING CORPORATION -- (202)628-4888
165 1 the staff a couple of questions? These might not be--because ll 2 we have time left. Under--let's see. In the EPRI design 3 requirements program or whatever that is called, one of the 4 things that they intended to do, and I am asking yta because 5 this was in cooperation with the staff, wha to somehow 6 incorporate the resolution of all USIs, maybe it was just 7 USIs, and generic issues, and now that A-45 has sort of 8 disappeared as a USI or been subsumed--we keep saying that 9 word--what does that mean? I mean how is EPRI going to deal 10 with that? Do you have any idea? Don't tell me that's up to 11 EPRI because I just want your opinion.
12 HR. HINNERS: Are you asking about how, what decay l 13 heat requirements will be for future--
14 CHAIRMAN WARD: I realize A-45 is not for future 15 plants. On the other hand, when EPRI and the staff developed 16 this kind of cooperative program of resolving all the USIs for 17 the future plant design requirements, how does that relate to 18 A-45?
19 HR. HINNERS: The short answer is we don't know yet, 20 but Tom King has appropriated for implementation a severe l
21 accident policy on future plants, and as part of that, he is 22 developing what requirements for decay heat removal would be 23 there for the policy, and when I talked to him he didn't have l
i 24 any specific requirements. He was not thinking of a dedicated l
i 25 system, but exactly what I was thinking of, I don't think they HERITAGE REPORTING CORPORATION -- (202)628-4888
)
l
- 166 1 l
5 1 have figured out yet. I ll 2 Does that answer your question?
- 3 CHAIRMAN WARD
- I think so. Okay. Any other l 4 questions, comments before we break for lunch? Okay. Let's I
5 come back at two o' clock.
6 (Whereupon, at 12:55 p.m., the meeting was recessed, i
- 7 to reconvene at 2
- 's0 p.m. the same day.) ,
1 8
9 10 11 l
12 h
14 l
15 16 I h
i 17 !!
I i 18 i
l 19 i i
20 l
l l
- 21 1
I i l 22 I
! 23 l l
l 24 g l h j 25 I I
i HERITAGE REPORTING CORPORATION -- (202)628-4888
167 i AFTZRNOON SESS ION 2:00 pum.
h 2 CHAIRMAN WARD: The next topic is a discussion of 3 the status of resolution of Ganeric Issue 99, loss of RHR for 4 PWRs during shutdown, and our presenter to start with is Mr.
5 Wayne Hodges, 6 HR. HODGES: I am going to talk a little bit about 7 what we are proposing to do. We have combined the issues of 8 the mid-loop concepts that you heard about about a year ago 9 with the Generic Issue 99, 10 Generic Issue 99 was being pursued by the Office of 11 Research, and the resolutions that they were proposing for 99 12 and the resolutions that we were proposing for mid-loop were 1
l 13 almost identical. And when we got together and discussed it, 14 we found that they really did dovetail very nicely, and we {
15 probably were going to be able to issue what we wanted to do 16 on the mid loop operation earlier than you could go through 17 the process for the generic issue.
18 As a consequence, we have agreed to combine the 19 resolution. I will talk about that in terms of mid-loop 20 operation. I won't say a lot about the Generic Issue 99, and 21 all of its ramifications. Gerry Hazetis from the Office of 22 l Research is over at the table, and if you have questions on i
23 l that aspect of it, he can answer those.
I
] 24 And also there was a probabalistic risk study cone 25 i for them by tre people at Brookhaven that formed part of the l
l 1
HERITAGE REPORTING CORPORATION --
(202)628-4888
168 1 basis for our cost / benefit analysis in our package to try to ll 2 justify what we are doing on the mid-loop operation and 3 Generic Issue 99, and if you have questions on that PRA, then 4 we will let Mr. Hazetis handle those as well. I had not 5 planned to other than to talk about the PRA, just kind of go 6 through the concerns and what we are doing.
7 (Slide) 8 HR. HODGES: Basically, we have got a concern that 9 ' operations during lecay heat removal system cooling contribute 10 a significant probability of likelihood of core damage, 11 (Slide) 12 HR. HODGES: Last summer, we issued Generic Letter l 13 j 87-12, which alerted the industry to our concern, and asked a 14 series of questions designed to say what are you doing now to, 15 that would try to avoid at.1 problem?
16 Despite that, and the number of events still 17 continue to occur, we had one at Waterford not too long ago.
18 Had one at Sequoyah that was really not, a concern because of 19 the low decay heat level and so it was nowhere el se to a 20 serious problem, and the one at Waterfird was not a serious 21 problem because they planned for it and took action, but the 22 fact remaint we are still having loss of decay heat removal 21 l during lower level operation, and we think that we need to try i
24 q to correct that.
I 25 : DR. KERR: What is neant by accident, i
ll HERITAGE REPORTING CORPORATION -- (202)628-4888
il 1
169 1 HR. HODGES: Yes. Accident is probably a strong g 2 word there. That's nothing was damaged.
3 HR. WARD: Wayne, is this rate just a continuation 4 of what has been typical for 20 years?
5 HR. HODGES: Pretty much.
6 CHAIRMAN WARD: Okay, and it is just despite the 7 interest and publicity, nothing has turned it down?
8 HR. HODGES: That is correct. Well, in fact, the 9 ene at Waterford was interesting in that they were actually 10 trying to do something about the problem. They had installed 11 some new instrumentation, but they hadn't checked it out yet.
12 They were still relying upon the old crudy instrumentation, l 13 and got in trouble.
14 There was one last month I believe it was at, one at 15 Palo Verde units where again they had what looked like a lot 16 of the right kind of instrumentation installed. They had two 17 trains of level instrumentation reading out in the control 18 room. Unfortunately, they had one common refer 3nce leg, and 19 while they were in the lower loop operation, they were doing 20 some maintenance work in the pressurizer and the maintenance 21 ,
man who was there had a, had a TV camera so they could report L
22 what was going on. He proceeded to place, to mount the TV 23 g camera, found a hole, stuck the camera 2n, he put the TV 4
h 24 !i camera on the hole, very nicely blocked the reference leg they l
25 l were using for the water level, and so they got bcth trains, i
HERITAGE REPORTING CORPORATION -- (202)628-4888
)
170 I
1 both trains of their level tracking very nicely--just wasn't 2 correct. And so there are things like that that you, need to 1 3 be planned out in advance, and what we are trying to do is l
4 just trying to make sure that a lot of planning is done before l
i 5 operation happens, that type of thing. We are saying the 6 events continue to occur. We haven't damaged any fuel yet, l
7 and I think in general we have been fairly far removed from 2 damaged fuel.
9 DR. KERR: What would be a" acceptably high rate?
10 Acceptable rate?
11 HR. HODGES: We would like to get it near zero, but 12 you never get to zero, but we would like to get it no more 13 than once every couple of years say. It is, it should be less 14 than one a year if you do it right I think.
15 CHAIRMAN WARD: Now are you talking--what you were s
16 just talking about? Were these specifically the mid-loop 17 l operation problems?
l 18 l HR. HODGES: Those were mid-loops problems at the 19 > lower level.
20 (Slide) 21 $
HR. HODGES: You esn get into trouble during 22 L nid-loop operation or lower level operations because of here 4
23 we said phenomena. It is really pressurization due to 1
24 hboiling, but we are using a more general term here. Can d
25 hpotentially 1 d to severe core damage in a shorter timeframe 1
l J
HERITAGE REPORTING CORPORATION -- (202)628-4888
171 1 than had previously been believed.
2 Most people were aware that as you lost the cooling, 3 you could heat up and boil and you had to try to do something, 4 but there was a general perception that you, the big pro 71cm 5 was boiling down the level. We found that that in many cases 6 is not the major concern.
7 Now at this point, I can do one of two things. A 8 year ago we talked about mid-loop operation and we showed you 9 lots of slides on it. I can either go back and summarize some 10 of that, or if you tell me that you remember enough about it 11 you don't want that. I can not do that. So what is your 12 choice?
13 DR. KERR: I am willing to assume if this is a 14 I problen, what are you going to do about it?
15 HR. HODGES: Okay.
16 CHAIRMAtt WARD: I think we remen5er.
17 j HR. HODGES: The last two slides of my package we 0
13 y try to go through that discussion if you wanted it, but if you 19 thirk you remember enough about it, we will charge ahead and d
9 20 Il not do that.
21 HR. WARD: Let's do that.
22 l] HR. HODGES: All right. Fine.
23 g (Slide) h 24 i HR. HODGES: We reviewed all of the respenses to 25 this Generic Letter 87-12 that were issued, and basically we
[
HERITAGE REPORTING CORPORATION -- (202)628-4888
173 1 had a set of 12 questions and so we broke down the evaluation l) 2 and responses into the 12 categorias, one for each of the 3 questions, and the bottom line is that none of the responses 4 we got were fully satisfactory, and some of the licensees' 5 responses were unsatisfactory in all 12 categories, and the 6 types of things we saw were a serious lack of understanding 7 and inadequate preparation for what might occur during lower 8 levol operation and just not really understanding the problem.
9 Some licensees appeared to not be takir.g corrective actions of 10 any kind, i
11 ! I must say, though, that the picture is better today i
12 lthanitwaswhenwegot the responses back in October. We did l 13 receive recently and you should have in some of your material 14 a letter f rom the Westinghouse Ow ars Group which describes 15 some analyses they did which indicates there can be a serious 16 problem primarily when you have got the, down in some legs in 17 some of the piping. This problem was identified separately 18 out of Diablo Canyon and I am not sure whether it was the 19 resident inspector or the people at the plant, but we heard 20 l about it from the resident inspector. It was identified 21 Il;separatelyat Diab?o by Westinghouse almost at the same time.
22 ,
Westinghouse was doing the analysis for tha Owners Group and 23 as a consequence, they have put out the letter to all of their 24 h owners and e"en to the other owners groups saying this could
)
25 i be a serious problem, watch out for it, so they are taking i
i HERITAGE REPORTING CORPORATION -- (202)628-4888
173 1 action and they are working toward developing some proceduros 2 to deal with it.
3 DR. KERR: EO you are, you have concluded maybe you 4 should get Westinghouse to send out your generic letters for 5 you? They might get more attention!
6 HR. HODGES: I don't know about mere attention, but 7 they are definitely moving, and my suspicion, and probably the 8 Westinghouse people in the audience would deny this, but it 9 may well be that they were trying to show the staff it was i
10 l wrong in what, seme of the things we said, and what they found 11 out was it was at least as bad as we were saying. It may be 9
12 worse. That may not be a fair statement for them, but that 13 was my perception.
14 (Slide) 15 HR. HODGES: Some of the individual licensees have 16 shown some excellent insights into what can happen. These 17 l included things like during the draining they had containment I
18 i closure. They have improved instrumentation. They make sure d
19 that they are monitoring the operation of decay heat removal 20 system. They make sure that they have got other back-up 21 l systems available to inject water, a numbrr cf things that we h
22 L considered the right kinds of action, but that has been some 23 for the most part isolated cases.
24 (Slide) 25 HR. HODGES: The kind of things that we see is HERITAGE REPORTING CORPORATION -- (202)628-4888
}
174 1 events continue to be initiated even though it is recognized 2 that it can be a problem. The planning to mitigate it has J been very poor in many cases.
4 The planning to prevent a release should the core 5 damage occur is often non-existent, and the analyses to try to 6 understand what can go wrong and how to deal with it has been 7 non-existent. Basically the plants have been operating in an 8 unanalyzed state where they just really didn't understand what 9 could go wrong.
10 t Some of the other things have been primarily poor 11 instrumentation, poor to non-existent procedures, lack of 12 training, and poor planning for the approach to mid-loop 13 operation.
14 y CHAIRMAN WARD: Could you give me an example of what 15 you mean by four?
i HR. HODGES: I'm sorry?
16 17 CHAIRMAN WARD: What do you mean by it is operating i
18 )inanunanalyzedstate?
19 HR. HODGES: Okay. Basically we have been going l
20 around fat, dumb, and happy, sayiro that if you lost your
\\
21 %n decay heat removal cooling, that there was plenty of time to ll 22 h restore cooling, the biggest concern being boildown, and if 0
23 1 you do the calculation on straight boildown, you have got an 24 order of four or five hours to take action. Nobody gets 25 I!terriblyconcerned.
h d
'i HERITAGE REPORTING CORPORATION -- (202)628-4888
175 1 If you look at, to the procedures that were in place ll 2 at a lot of plants, if you lost RHA pump procedures, it would 3 say, if it had procedures they would say start the other RHR 4 pump. Well, if you lost your RHR pump because you were 5 ingesting too much air because of the water level was too low, 6 as soon ac you start the other RHR pump, you would have lost 7 that one also, and there were no controls on how much 8 equipment could be out of service, so they weren't aware that 9 they were that close to baing in danger many times, and that's 10 what I am calling non-existent analyses.
11 At plants like Diablo Canyon we have seen that as a 12 result of the air ingested, that the level behavior didn't l 13 correspond witt. what you would expect from what you were 14 putting in. You could see that if you injected water, the 15 water level would actually drop. The indicated water level 16 would drop, or if you stopped injecting, it might go up, and I
17 I they didn't understand why. It gets to a redistribution of 18 some of the air that is in some of the piping in the system, 19 but nobody had really tried to understand what was happening, 20 and it was very confusing to many operators, so that's what I 21 mean by the operating in an unanalyzed state.
22 (Slide) 23 j HR. HODGES: Loss of decay heat removal has been a 24 j frequent occurrence, and roughly one third, and that number W 25 h[ could be updated maybe to 41 or so, have occurred during the l
HERITAGE REPORTING CORPORATION -- (202)628-4888 _ ______- - _
176 1 reduced inventory operation, so the mid-loop problem is not g) 2 the full loss of decay heat removal problem. That's what 3 Generic Issue 99 was discussing, but it does address a third 4 of the problem, the occurrences.
5 CHAIRMAN WARD: You say it addresses about a third 6 of the occurrences?
7 HR. HODGES: Yes.
8 CHAIRMAN WARD: Okay.
9 (Slide) 10 HR. HODGES: So what we are proposing to try to fix 11 the problem is really a two-pronged approach. The first one 12 we are calling the short-term or expeditious action which says l 13 that the, for the immediate term or short term, let's try to 14 reduce the likelihood of getting any off-site releasees should 15 you get any damage, and also do what we can in the short-term 16 to try to prevent the damage.
17 i And I will give you a list of things we are 18 proposing for that. We also, we want them to start on to a 19 longer term program which involves things like improved 20 instrumentation, improved procedures, analyses. Things that l
21 l cannot be done within a few months--to try to develop really
'l 22 l the way to avoid a problem all together, and then as they 23 j develop the improved analyses, the improved understanding, the h
24 ji things that would be required in the short-term like h
25 I containment closure, or minimizing accidents that perturb the HERITAGE REPORTING CORPORATION -- (202)628-4888 )
177 i system could be relaxed because we could have assurance you 2 would have a back-up.
3 (Slide) 4 HR. HODGES: Now for the expeditious actions, there 5 is, we have got seven identified. We want containment 6 closure, and by containment closure, here we mean prior to the 7 time that you would expect core damage, and so if you have got 8 no analyses at all to know when you might get there, when it 9 is fairly simple analysis but you have got none, then in our 10 l letter we would specify the timeframe that could be used and 11 it ranges from on the order of half an hour up to two and a 12 half hours, depending upon the design of the plant, things 13 like decay heat.
14 Once you have analyses to understand how long you, l
15 j how far you are removed from core damage, then the containment 16 closure requirement, how soon you have to be able to close it, 17 can be relaxed somewhat. Containment closure is a new term.
18 I We don't mean by that the old isolation that has been, is used I
19 when it is in, the plant is operating. All we are really 20 looking for is the capability if you do have the core damage, 21 to keep it bottled up, and so what we are saying is basically 22 one barrier, not a double barrier, and things like the i
23 equipment hatch that you would tighten enough bolts down that 24 j you could have the seal contacting all the way around. This b would give you a very high assurance that you were not going 25 l
HERITAGE REPORTING CORPORATION -- (202)628-4888
178 1 to get off-site releases. You don't have anywhere close to h 2 the energy levels associated with the melt under these 3 conditions as you would if you were in full power operation, 4 so it would be much easier to contain, and that also gives 5 them some flexibility in keeping the containment open while 6 they are doing operationc they need to do when they shut down.
7 We also want to have at least two--
8 CHAIRHAN WARD: You seem to have disposed of a 9 pretty complicated, what seems to be a really complicated 10 subject there with a few, a very brief paragraph.
11 HR. HODGES: Yes. It was not easy to get it to that 12 brief paragraph.
l 13 CHAIRMAN WARD: Yes.
14 Mr. HODGES: We have tried a number of definitions.
l I
15 Our major concern was this--we wanted to be able to make sure 16 that we could have the containment reasonably closed prior to 17 any core damage and yet have minimal impact on the operations 18 that have to occur while the plant is shut down, and so we are 19 I trying to strike a balance between having the safety with the 20 containment, and not being overlyburdensome, keeping it closed 21 down.
22 l CHAIRMAN WARD: So I mean what sort of procedure 23 would you, would the plant have? I mean is this a procedure
, 24 l!that woulu get, get the lid on loosely in ten minutes or ten
, L 25 h hours? What would cause an entry into that procedure?
b l
HERITAGE REPORTING CORPORATION -- (202)628-4S88
179 1 MR. HODGES: Okay. What would cause an entry into the procedure would be loss of decay heat removal. The pump lh 2 3 trips off, and then they would start into the procedure. Now 4 if they get it restored within half an hour or however long it 5 takes to restore it, they can go back out of the procedure, 6 but have at least started the procedure of trying to close the 7 containment.
8 Now the big, one of the biggest problems in keeping 9 containment closed is the equipment hatch. That's a very 10 large door, and if you insist upon the kind of closure we have 11 during plant operation, it could take a long time. It can 12 take anywhere from some plants eight or ten hours to get l 13 everything locked up the way it should be.
14 There is a couple of plants that have to chisel 15 l their bolts off the door and they ship it off so it will take l
16 ltwoweekstoget it on. And then there are other plants that l
17 apparently because of designs can get it on faster unan that.
l 18 We are not requiring that, the full containment 19 I isolation. We are requiring that the door basically be in 20 place, and if you had a pressurization due to a hydrogen burn 21 or whatever in the containment, generally the design of the 22 ,
door is such that you are going to push in against the sehl b
23 h and seal it tighter. So what we are saying is if you have got b
24 h it in place so it pushes against the seal, then you shouldn't b
25 have much blow-by around the seal, and any pressurization HERITAGE REPORTING CORPORATION -- (202)628-48SS
i 180 1 would tend to close it tighter, so we are trying to get a ll 2 definition that provides that the closure of the containment 3 without being overlyburdensome, and that was not an easy 4 definition to arrive at.
5 CHAIRMAN WARD: Yes. I mean you might use that 6 argument for normal operation.
7 HR. HODGES: You might. I don't think anybody wants 8 to try it at this point. Normal operation, you don't hhve 9 anywhere near the need to be going in and out. You know, you 10 have to have some time when you are moving fuel in and out, 11 when you are moving hardware in and out. There are a lot of 12 things that can happen during normal shutdown. They are going l 13 to be working on pumps. They are working on the primary 14 system. They have to be removing equipment in and out. Some l
15 i plants put railroad tracks through the doors and you have got l
16 ;
to be able to move those out of the way. There is a wide 17 l range of things that are done during that, during that 18 shutdown mode.
19 CHAIRMAN WARD: I am trying to figure out what the, li 20 F what the tradeoffs here are. I mean you have got an obviously 21 less tight containment. Are you tiguring the risk is
{
22 i acceptable because the tine in this condition is small or b
23 ll because the potential fission product release is small or 24 because the energy release from a damaged core is small?
i 25 HR. HODGES: The amount of fission product would l
l l
HERITAGE REPORTING CORPORATION -- (202)628-4888
181 1 probably be large because you would be uncovering basically g 2 the whole core and so you would expect to damage a large 3 portion of the core, but the entity associated with it is 4 going to be relntively small.
5 CHAIRMAN WARD: Why?
6 MR. HODGES: Because you're in cold conditions. You 7 don't have a lot of pressure, don't have all that. The systen 8 is at low pressure. You are not blowing down from a high 9 pressure so you don't have a large pressure force to push it 10 out of the containment. If you have got a hydrogen burn, you 11 might get some pressurization, and that's why we wanted the 12 thing to be sealed.
l 13 DR. KERR: This pressurization take place if it 14 occurs in the top of the vessel?
15 i MR. HODGES: Basically your vessel is open during i
16 mid-loop operation. You go into mid-loop operation for 17 several reasons, one of which may be to work on steam 18 t generator. Another would be to work on reactor coolant pump, 19 ! and generally or in many of the cases the vessel will be open.
20 DR. KERR: As I read the description of the Diablo i
21 l Canyon, was that the case in the Diablo Canyon?
l 22 ! MR. HODGES: No. They were not open. They were 23 l were there to work on the steam--
24 p DR. RERR: I was impressed by the pressurization.
1 25 top was, pushed water away from the--
lThe 1
HERITAGE REPORTING CORPORATION -- (202)628-4882
r t 182 1 HR. HODGES: Right that was our initial concern was 2 the preseurization pushes the level down in the core, and if 3 you have got s pump open for working on it, it pushes it right 4 out and you are in trouble. The Westinghouse analysis that 5 they did for the Westinghouse owners Group said that's not as 6 great a concern as we thought because what happens is the 7 steam clears the loop seal and you get escillitory behavior.
8 You have a longer period of time than we had calculated with 9 j our hand calculations, but they went a little bit further and 10 said if you have got dams installed in there, the 11 t pressurization can cause you to blow out the dam and now you l
12 l are in trouble.
13 DR. KERR: Suppose you had the PORVs open on the 14 pressurizer? You wouldn't.
15 MR. HODGES: At lower pressure conditions that is I
16 lusuallynot providing venting.
17 DR. KERR: That's what I thought.
l 18 !!R . HODGES: In fact, on some plants like a 19 Combustion plant, the limits thing would be not the PORV, but 20 l the pipe running over to the pressurizer, the surge line. It I
21 j would be fairly small.
}
22 h CHAIRMAN WARD: Okay. Well, I'm still trying to 23 h figure out why the risk is so much lower, you know, given a i
24 [ severe accident. I mean if I assume and what I have been 25 hearing is that tine probability of let's call it a severe n
I HERITAGE REPORTING CORFORATION -- (202)628-4888
183 1 accident or a core melt is, is about the same in terms of r
2 reactor week or something during shutdown us it is during 3 operation.
4 HR. HODGES: We believe--
5 CHAIRMAN WARD: Some reactors are only run half the 6 time and so maybe half the time--well, I guess these don't 7 have the containment open.
8 HR. HODGES: We believe that if you took no actions, 9 that the probability of getting out, release outside of 10 containment during the mid-loop type of operation is at least 11 , as high, if not higher, than a full power operation.
12 ,
DR. KERR: On the other hand, the actions to prevent 1
l 13 it is to take, is to get a good sized garden hose and start 14 lsquirtingwater.
15 HR. HODGES: For fairly small, it is easily done.
16 lIt doesn't take a lot of water, but you have got to be careful 4
17 h with that garden hose that you put it in the right place. And 0 If you put 18 q in some case you have to have hot leg in]ection.
I 19 it in cold leg, it doesn't work.
l 20 HR. DAVIS: With the containment sealed up, I don't 21 know how you are going to get the garden hose in there.
L 22 MR. HODGES: There are water sources that you can lI d
23 get to from inside the containment.
a A lot of people stay in there after you, 24 l HR. DAVIS:
il J
25 when you close it up.
l l
l HERITAGE REPORTING CORPORATION -- (202)628-4888
184 1 DR. KERR: He didn't say that you had to have a ll 2 containment closed, if I understood him. He said you had to 3 have a procedure so that you started closing. I don't know 4 what that .means.
5 MR. HODGES: If you calculate that you are going 6 to--say you have a manway on a pressurizer open. You have got 7 a reasonable vent, and so in thoce--and you shut down for a 8 while. You have got low decay heat. Hight calculate you have 9 got ten hours to boil down to get any damage. That's says in 10 that case, once you lost your RHR suction, within ten hours 11 you should have your containment closure.
12 l DR. KERR: First thing, you have to detect the P
l 13 problem, do a calculation, see how long you have.
14 HR. HODGES: It is fairly easy to look that up on 15 l the curve.
I i
16 i MR. CATTON: I think Wayne said he micht do that U
17 before in anticipation.
18 MR. HODGES: If you haven't done the analysis ahead i
19 k of time, then we are saying you don't have ten hours. You 20 have maybe two hours.
21 MR. DAVIS: Do you do anything with resetting the 22 h'!containment spray logic, assuring line-up of the spray trains 23 1 and water sources?
d 24 L MR. HODGES: We have not proposed that.
H 25 MR. DAVIS: You haven't considered that?
[
hl HERITAGE REPORTING CORPORATION -- (202)628-4888 >
185 1 HR. HODGES: Our biggest concerns are really trying h 2 to prevent the loss of decay heat removal, and two, if you do 3 1c:e it, making sure you put water back in, and finally trying 4 to keep the containment closed. Our short-term action says we 5 won't--at least in the short-term, all of the things you would 6 like to have to prevent is going to take a while to get in 7 place, and so emphasis is this strongly on the containment 8 closure, but as soon as they start getting some of that other 9 stuff in place, that backs off because that's not a simple l
10 thing to do and it in somewhat burdensomo.
11 CHAIRMAN WARD: Okay. Well, let's see. I am still h,
h' 12 l unsatisfied about why you are satisfied with this sort of l 13 half-baked containment closure deal. I mean why is a core 14 l melt entered into from shutdown low inventor" conditions less 15 severe than a core melt entered into from operating d
16 h conditions? I mean you don't have the blowdown of the prinary h
17 system pressure, but you do have the potential for--and I H
18 hl guess you don't have the potential for the direct containment d
d 19 [ heating problem.
I 20 HR. HODGES: You do have a potential for a hydrogen i
21 j burn. I 22 U CHAIRMAN WARD: Hydrogen burn, you have got what is l 23 h the, another threat everybody worries about with spurious l 1 l 24 accidents. That is core on the ficor, and concrete core l
1 25 I interactions and gas formation, that's just as big.
L l
HERITAGE REPORTING CORPORATION -- (202)628-4888
186 1 HR. HODGES: That would be probably just as big or h 2 close to as big.
3 CHAIRHAN WARD: Fission product inventory isn't 4 much.
5 HR. HODGES: I think you are starting to get into 6 the grass roots of likelihood. We are trying to strike a 7 balance of reason. We want some protectien. but we don't want 8 to be burdensome, to with--and we are trying to strike a 9 balance between providing a measure of safety we think is good 10 .
enough, and so that's all we are trying to do. It is not full l
11 l containment isolation. It is not perfect.
d 12 h (A discussion was held off the record.)
4 l 13 CHAIRMAN WARD: Well, you know, I guess it would be 14 j less burdensone en licensees if you permitted them to open a n
15 h containment once in a while during operation, too, but you i
16 have decided that's not appropriate, and I'm trying to see how J
17 J you have eado some sort of. you know, I nean you could require i
18 that containnent be absolutely buttoned up whenever they went
! l H l 19 L inte certain types of shutdown operations. l h ,
20 h HR. HODGES: That is correct. !
1 21 o CHAIRMAN WARD: What sort of reasoning have you gone 20 t through? Have you nade any numbers or-- ;
23 , HR. HODGES: We have not do: e d o ': t caleblation for !
l i
24 that. There were sene dose calculat 'nc h:. '.r o o k t a v e n i 25 for the Generic Issue 99. That tal s source tern d i HERITAGE REPORTING CORPORATION -- .)625 Ald8
I 107 I
1 and the energy level, but we haven't looked at specifically g 2 the mid-loop operation and fron various stages of it being 3 shut down. They made an assumption of being shut dovn what?
4 How many days, Jerry?
5 HR. MAZETIS: Three days.
6 HR. HODGES: Before they entered that; it can range 7 anywhere from one day to weeks. We haven't tried to do that.
l 8 , Their analysis showed it was a low energy level release.
I 9
l Based upon that and trying to strike a balance l 10 between delaying a restart by a couple of weeks, we decided l h i 11 ij t h a t if you had an assured containment closure as opposed to l I 12 i isolation, that was good enough.
13 HR. MAZETIS: This is Gerry Hazetis in Research. It 14 j is inportant to cephasize one part of that response, and that h
15 L is that Wayne hasn't gone through the whole list here, but to i
16 L answer the questien of risk, of course, there are two parts to U
17 '
it. There is the core damage frequency part, and then there i
18 is the back entrainment part, where the containment would get f
1 19 J involved, and there was no intent by the staff to fix only one L
20 l or the other, i
n 21 What Wayne tried to do in NRR, which was consistent I 22 with the approach on Generic Issue 99, was to look at the ,
t 23 l eeans available to reduce the core damage frequency to l I
24 p reasonable levels, and to at the same time, look at the l l
25 flexibility we have on providing more assurance of containmentj i
h HERITAGE REPORTING CORPORATION -- (202)628-48S8
188 l
1 closure, the total of which would give us what we want--a moro 2 acceptable total risk for these types of operations.
3 CHAIRMAll WARD: Well, right now we hear estimates of 4 the kind of the conditional probability of containment working 5 and doing, working as it should in accidents developing at 6 power, and those, you know, I have seen numbers rangi.ig from 7 10 percent to 90 percent.
8 What sort of conditional probability would you put 9 on containment functioning satisfactorily under the scheme you 10 l have proposed here?
11 ! HR. HODGES: We haven't.
12 CHAIRMAll WARD: But you are making some sort of i
13 judgment that it is not 10 percent?
14 HR. HODGES: Absolutely.
15 HR. HODGES: The things that drive you to the high l
i 16 ! probability of containment failure that you have seen there 17 [ have been primarily for MARK I containments, have been the h
18 ll things like the melt goes over and attacks the liner. That's b
19 [ where the liner is fairly close to the vessel itself, and the 20 h PWR generally is much further away, so that's not the same b
21 [magnitudeofproblem.
22 CHAIRMAtt WARD: Wouldn't have to do that under these 23 conditions because you already got it open. Wouldn't have to 24 melt through the liner.
25 MR. HODGES: But we are closing it. t 0 l L
h h
HERITAGE REPORTItiG CORFORATIO!! -- (202)628-4888
189 1 CHAIRMAN WARD: Okay.
HR. HODGES: That's why we are closing it. If you lg 2 3 did nothing, you are going to put something out, and we are 4 saying we want to try to prevent that, but while they are in 5 shutdown, you have got the equipment hatch open. You have got 6 work going on on a number of penetrations, so they may have 7 in, addition to the hatch open, some fairly large fraction of 8 the penetrations open. They have also got to get those 9 closed, put a flange or something on it. So we are saying you 10 don't have to have the double barrier and the double barrier 11 is there in case one of them fails. We are sayipo in this 12 case, one barrier is good enough. Just get it in place and be
, l 13 able to do it within X time.
14 CHAIRMAN WARD: Thats defense in depth.
h 15 ,
HR. HODGES: You have lost some of the defense in 16 l depth, but you started out, already when you started this I
17 event you had your containment open, and your primary system
[a i
18 1 open, so you had two of your normal barriers open before you d
4 19 h ever challenged your cladding, i h I 20 h HR. CATTON: Your forcing function is also much h smaller, so all the processes are sort of spread out. !
21 g L
22 d CHAIRMAN WARD: Part of the forcing function is, but 23 ha lot of it isn't it seems to me. I 24 HR. C ATION : The decay heat is way down, so whatever
[
i 25 [ happens is going to happen.
f I
I HERITAGE REPORTING CORPORATION -- (202)628-4888
190 1 CHAIRMAN WARD: "! '
. Jtill melted, you take the 2 water off it.
3 HR. CATTON: But the amount of heat generation is 4 much lower, so small heat sinks are going to do a lot of good 5 as far as time is concerned. It is going to be spread out so 6 you have more time to try to do something.
7 CHAIRMAN WARD: I'm not saying any explicit 8 tradeoffs here. I mean it is a lot of hand waving.
9 HR. HODGES: It is a--
l 10 HR. CATTON: It is not clear to me. If you dump it 11 on the concrete, it is going to attack it at the low decay 12 ! heat lovels, i
l 13 HR. HINNERS: Probably not a lot different than in 14 core melt accident. You don't dump it on the flaor in a full 15 power until hours later decay heat is down.
16 HR. CATTON: Here you are talking about a week l
17 hlater, aren't you?
h 18 f HR. HODGES: Not necessarily.
H 19 [ HR. HINNERS: Decay heat doesn't change that much 0
20 h anyway.
II 21 HR. HODGES: You could be in mid-loop operation 22 L within one day.
l 23 HR. CATTON: Haybe Dave is right. !
i !
24 o HR. HODGES: It is not normal, but you could be. It 25 L is normallf a few days, but you cculd be in mid-loop operation h
n ll f
HERITAGE REPORTING CORPORATION -- (202)628-4888
191 1 as early as one day after operation.
2 CHAIRMAti WARD: Well, it just--I'm just concerned.
3 You havo got a new awareness here that, looking at events and 4 analysis, that there is maybe a significant threat of core 5 melt during shutdown conditions. I'm just not sure whether 6 you have reacted to that quite as completely.
7 HR. HODGES: We are saying rather than go out and do 8 a lots of analysis, try to cover it, we think if you, if you 9 have a barrier to prevent the release, and we are detining 10 i that barrier as for the, something like an equipment hatch, 11 having it down, closed down, you don't have a gap around the I
12 seal. If you did have something like a pressurization, that 13 is going to help seal it.
14 CHAIRMAll WARD: That's true for every one of the 105
[
h 15 j containments out there?
16 h HR. HODGES: The ones I know about, yes.
17 DR. KERR: Implicit in it it seems to me is the L
18 d assumption that you have a lot of more time to do something j p
l!
19 ] and you have a good many more possibilities to do something 1
20 qi than you do in an accident which occurs in power, i I
21 HR. HODGES: Be careful with that. The worst l
22 g situation, you could be in core melt in one hour, in the worst I i 23 situaticn. But in that-- l 24 i HR. MIlillERS: Core melt or core uncovery?
25 HR. H0DGES: Core melt.
u HERITAGE REPORTING CORPORATIOli -- (202)628-4888
192 1 DR. KERR: I am saying the likelihood at this point 2 is unknown, but it is almost certainly low, that that will 3 occur.
4 HR. HODGES: I think now it is, because we have been 5 mede aware of what gets you there. Before it was not, I don't 6 tilink it was quite as rare because we didn't understand.
7 CHAIRMAll WARD: But I thought you said you think now 8 everybody understands it?
9 HR. HODGE9: I am saying--
10 CHAIRMAN WARD: The responses you got from your 11 generic letter.
12 HR. HODGES: The responses we got to the generic 13 l 1etter were not very good, but I don't think the industry had 14 stopped there, either. I say we know, for example, that l
l 15 Westinghouse at the expense of the owners went off and did a 16 q lot of analysis, and chey are the ones who said if you have 17 got the dam in the pipe, this is the worst situation, this is 18 [whereyoucanpressurize to the point you flow the dam out h
19 within 15 minutes and within an hour you can be melting the 20 core, so what they have told the utilities is stay out of that i
21 problem.
I Is that the case even if you have a head 22 g DR. KERR: j i
q L
23 on?
24 HR. HOD 9S: Flo . If the head is on, that doesn't j
25 h happen. ,
b f U
HERITAGE REPORTIllG CORPORATIO!! --
(202)62S-4888
193 1 DR. KERR: Most of the containments when you are h 2 doing this, don't you have a head off?
3 HR. HODGES: No, not necessarily.
4 DR. KERR: I mean most of the times when the reactor 5 is shut down for repair, refueling--
6 MR. HODGES: If they are refueling, the head is off, 7 but they are not always refueling. They may have been 8 operating and experienced a problem with a pump and they come 9 l down to fix the pump. They are in mid-loop operation now, i
10 q Because they have to lower the water level to get there, they I
- l 11 1 have to drain the steam generators, to put dams in to drain d
c 12 l: the piping to work on the pump, and they are wanting to get h
l 13 back in operation as soon as possible so they get into 14 mid-loop operation as soon as possible. If that's the case, 15 you will be there within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
16 h DR. KERR: How is that per year over the population 17 '
of PWRa? Do you have any idea? '
I 18 N HR. HODGES: We have got Roger Newton from the !
i-19 Westinghouse Owners Group. Roger, did you want to try to I u !
l 20 ;
answer that? }
o 21 L HR. NEWTON: Could I hear the question again? !'
I l
22 l DR. KERR: How many times does shutdown to repair a '
i 23 pump or something which requires mid-loop operation occur 24 during the course of a year for the population of PWRs just if {
l 25 i you have any idea, rough number? And is it once, ten tires, i a
i; HERITAGE REPORTING CORPORATION -- (202)62S-4SS8
194 1 once every ten years?
g 2 HR. l'EWTON : The name is Roger Newton. I am also 3 Chairman of the Westinghouse Owners Group. I don't know the 4 specific number, but just based upon what I perceive of what 5 is happening in the industry, our reactor coolant pump 6 aspects, we have had very good experience lataly. I don't 7 know that it has really occurred in any plant in the last few 8 years, so it is for the population less than one a year.
H 9 Also with respect to if Ke were doing that sort of I repair, in most cases where we come down quickly in a short 10 il 11 period of time, want to do a repair, most often it is looking 1
12 q at steam generator problems versus the pumps, but either case, 4
l 13 those situations, very seldom does a utility unbutton its 14 j containment equipment hatch or all of the other penetrations.
l 15 g Usually it is just the personnel one. And that we all have 16 cases where we have procedures for closing the personnel 17 , lock-up in case something happens during those times, so when l 18 there is a high risk, most of the time you Son't have the 19 containment open. It is usually during the refueling outage i
20 '
where you take several days to a week to get down to that ;
21 conditien, certainly when you have the equipment hatch off, so 22 I think they are 1 coking at the overc11 flow at which we 23 operate our plants as well to rake this containment closure a l 24 reasonable requirement. l l
25 DR. KERR: Thank you. j i
U PERITAGE REFORTING CORPORATION -- (202)6 S-4888
195 1 HR. HAZETIS: I believe Brookhaven looked at some of ll 2 the operating experience in thic area. Lewis?
3 HR. CHU: Based on the experience that Zion reported 4 in NSAC 84, -
?ghly they have reached a maintenance outage 5 what, about twice a year? And then .7 per year also they 6 reach the mid-loop in addition, so I would say it is roughly 7 three times a year.
8 HR. CATTON: Is the containment open or do they just 9 go in the personnel lock?
10 HR. CHU: That I don't really know, but I would l 11 expect if they are doing outage, probably the equipment hatch 12 will be open, but in the maintenance outage probably it is not l 13 open.
14 CHAIRMAN WARD: Okay, so then in the population of 15 PWRs, you know, we are only faced with this 40 or 50 times a 16 year let's say?
17 HR. CATTON: It also sounds like in the 18 circumstances where the decay energy is high, the cor.tainnent 19 is closed, that it is a week or so. When the containment is 20 open, it is, has been a week or so.
I I
21 HR. HODGES: And part of the, what we were doing 22 with the program is try to assure that, yes.
23 HR. CATTON: You do have time. Things will be 24 happening slowly when the containment is open, 25 HR. HODGES: We hope.
HERT. TATE REPORTING CORPORATION -- (202)628-4888
196 1 MR. CATTON: We hope.
h 2 CHAIRMAN WARD: Well, okay. I'm not sure the tech 3 specs assure that.
4 MR. HODGES: They don't.
5 MR. CATTON: It would be interesting to find out if 6 that is correct.
7 HR. HODGES: They don't necessarily now.
8 CHAIRMAN WARD: Maybe we ought to let you go ahead.
9 MR. HODGES: It is not surprising that took a lot of 10 discussion. We have had a lot of discussion both with 11 industry and in-house with that. That one is not a simple one 12 to handln.
l 13 The next one is after a short term, we want two 14 indications of reactor coolant system temperature. That would 15 typically be the core exit thermocouple in a lot of plants 16 once they have used RHR system. That is, we are saying that 17 is not good enough. If you lose that, you have lost your 18 temperature indication, so as long as the head is on the 19 vessel, we are saying we want two temperatures, and it is just 20 redundancy basically.
21 We also would like to have readout of the reactor 22 coolant system level in the control room, but for the short 23 term if they can't do it in the control room iith a direct 24 readout, we are allowing a local readout, and we also are g
25 saying for the short term, because it sakes them *ine to HERITAGE REPORTI!iG CORPORATION -- (202)628-4888
197 1 install the right kind of piping for a good instrument system, 2 we would permit the tygon tubing as long as it is verified to lh 3 be installed properly. For the long term, we don't want them 4 to use tygon tubing. There is too many problems with it and 5 we want to get away from t ., but for the short term, we are 6 not requiring the direct readout in the control room if they 7 don't already have it.
8 We also are saying any time that you are in the 9 mid-loop operation, do not per turb the reactor coolant system 10 deliberately. Most or a large number of the loss of decay 11 heat removal has occurred when they were doing operations that 12 affected RHR system, or the primary system itself and sc we
, l 13 are saying stay away from that. Do your planning ahead of 14 time so as you understand the actions you are going to take do 15 not perturb the system.
16 We want to make sure that they have got back-up 17 equipment available to provide water if it is neededif they 18 lose RHR system, and that has--and what wa are requiring there 19 are two systems, one of which has to be hot leg injection; and 20 on the hot leg, hot and cold leg enclosure, basically we are 21 just saying that there is an importance to the sequence in 22 which you do that, and follow the appropriate procedures.
23 And we also want them to have discussed with their 24 staff not just the control room operator, the implications of
[
25 the Diablo Canyon event and the relateu events and what they HEKITAGE REPORTING CORPORATIOM -- (202)628-4888
198 1 mean on operations. And it gets to be more--when you look at things like the San Onofre event, the maintenance people are ll 2 3 the ones who can get you in trouble, and also at the 4 Waterford, they had a reasonable installation of the tygon 5 tubing except it was too long, and it was kind of sprawled 6 across the floor and maintenance people walked around it and 7 said somebody is going to trip over that. They curled it up 8 and turned it over a railing and immediately you had lousy 9 instrumentation, so it is a matter of not just a control room 10 operators but all the maintenance as well being instructed 11 what are the implications.
12 Now for the longer term what we are calling the l 13 program--
14 CHAIRMAN WARD: Are you going to tell us how you are 15 going to get these things implemented?
16 HR. HODGES: Yes. For for the longer term, we are 17 going to what we call the program enhancements. These are 18 things where it takes time to purchase equipment or to write 19 procedures and get the training, and so that could take one or 20 two fuel cycles in order to get it in place on the plants.
21 And what we are saying, we want to improve instrumentation, 22 and here we want two channels of temperature juat as we talked 23 about before. We want two independent readout of the water 24 level in the control room, and if they have a situation there g
25 like they had at San Onofre, they already had the H2RITAGE REPORTING CORPORATION -- (202)628-4888 c
199 1 instrumentation installed, we want one common reference leg.
g 2 We wouldn't require that they go back and drill another hole 3 to get an independent reference leg, but we would require that 4 they have procedures to verify the reference leg is clear. So 5 it is a little bit of a compromise on the independence but an 6 assurance that it is working.
7 We also want instrumentation that gives them an 8 indication of the performance of the decay heat removal system 9 so they get alarms if they start to get fluctuations in the 10 pump current and alarms if the level goes out of range.
11 We need procedures on the proper way to conduct not 12 only mid-loop but all of the shutdown operations. Hid-loop is l 13 just one of those. And we are talking about in emergency 14 procedures. What we would learn from that is once you start 15 getting into trouble, whether it is a loss of decay heat 16 removal, pump or whatever, then they wculd determine the 17 emergency procedures that would specify the action it would 18 take, but to do those procedures they to have the analyses to 19 understand what can happen in a plant, so all the analysis 20 is--No. 4, that would really come before the procedures.
21 And equipment, all we are saying is make sure the l
l 22 equipment you have got is linoj up and available in case you 23 need it to provide water. W9 are not requiring new equipment, r
24 only that they have available sufficient equipment to provide 25 adequate cooling if needed, and then there would be a number l HERITAGE REPORTING CORPORATION -- (202)628-4888
200 1 of technical specifications changes that we see would be ll 2 needed as a consequence of this.
3 One would be the fact that typically when a plant is 4 shut down on decay heat removal, the tech specs requi.re a very 5 large decar heat removal flow, and this is not really 6 necessary. You don't need to put three or six thousand GPMs 7 through the core. You can get by with much less, and so we 8 would allow them to throttle back the flow but to have to 9 change the tech specs to do that, operate with lower flow 10 which would help on the problems with the vortex and the air 11 ingestion.
12 There need to be probably changes in tech specs to l 13 require some of this equipment to be available. Put 14 restrictions on how much equipment could be out of service for 15 maintenance. There need to be tech specs that would cover the 16 interlocks for the overpressure protection. This is one of 17 the other recommendations that is coming from the Generic 18 Issue 99. That was one of the major causes of loss of decay 19 heat removal at THI, a perturbation to the system, and so if 20 you can remove that interlock and replace it with some 21 conpensating procedures or whatever, you minimize the 22 perturbations to the system and you really have an overall 23 improvement to safety, but that also requires changes in tech g
24 specs, so we see a number of tech spec changes, many of which 25 will be relaxations over what we have normally required, but HERITAGE REPORTING CORPORATION -- (202)628-4888
201 1 should be in the direction of safety.
l 2 That's what we are going to be requiring. Now how 3 are we going to do it? We have in preparation now, it is 4 going through management concurrence, a generic letter to go 5 out to all the utilities to tell them to do it and giving them 6 timeframes to respond, outlining the expeditious action and 7 program enhancements. We have had meetings over the last 8 couple of weeks with all of the owners groups basically going 9 through the presentation that I have given you only in more 10 detail of what we were going to be doing, so they have been 11 made aware of what we are going to be doing.
12 In fact, some of the things have been rephrased to l 13 not be quite as burdensome and yet still get what we wanted.
14 We have taken some of their input, and we expect to go through 15 CRGR on this fairly shortly. We had hoped to have something 16 out by the end of July. It would probably be middle of August 17 now.
18 DR. KERR: And you will ask or tell them, require 19 them to do whatever it is?
20 HR. HODGES: The letter that will go out will not 21 require it. It will say this is what we recommend. What we 22 have discussed with our management is if we don't get the 23 utility doing that, we in very short order cou.'$ issue orders.
24 We are not making that threat to the industry, but it is q
25 there.
HERITAGE REPORTING CORPORATION -- (202)628-4888
202 1 CHAIRMAN WARD: I guess you just made it.
This is on the basis of some regulation ll 2 DR. KERR:
3 that is now being violated?
4 MR. HODGES: We are doing it from the safety 5 standpoint. It is, when you go through the regulation, '. c i s 6 not clear which ones apply here. Most of the regulations are 7 geared toward full power opetation, and you have to stretch 8 them to apply them here.
9 DR. KERR: So you don't think a regulation is 10 needed, though?
11 MR. HODGES: We don't think so.
12 HR. CATTON: As long as the stretcher doesn't break l 13 anything.
14 MR. HODGES: Right. Some people have proposed we 15 need, may need rulemaking here. I hope we can do it without 16 rulemaking. I think just on the safety basis, we can require 17 it, but to point to the regulations say Part 100, for example, 18 Part 100 is based upon full power operation.
19 DR. KERR: It seems to me we have all these 20 regulations supposedly to make reactors safer. We are finding 21 constantly that you need to do things to make them safer for 22 which there are no rules and we are not making rules and so 23 how does anybody find out what it is that one is supposed to 24 do to have a safe reactor? We just get all the correspondence 25 from the staff and go through it? You won't be able to tell HERITAGE REPORTING CORPORATION -- (202)628-4888
l 203 1 anymore by looking at the regulations.
MR. HODGES: Well, for the reactors that are in lh 2 3 operation today, we are going to be cover 4nc with this letter 4 and any follow-up that is aceded.
5 For the newer breed of reactors like the SP 90 or 6 the advanced light water reactor, in cur reviews of those, we 7 are putting words in there saying let's not put these band-aid 8 fixes on. Let's design it so you don't have a problem.
9 CHAIRMAN WARD: But ho,e are you going to do that.
10 Bring me a rock or by--
11 MR. HODGES: No. Already we have plants that have 12 less of a problem with this. The .3&V plants have very little l 13 problem with this for several reasons. One is they have that, 14 the higher level legs relative to the core. They have 15 internal vent valves that so you don't get the delta P effect.
16 As to some plants already are not as bad off as others. There 17 are things that can be done in design like installing vents in 18 the steam generators or places where you can inject nitrogen 19 to vent the steam generator so you don't have to drain down as 20 low. We are not trying to tell the industry how to do it, but 21 we know of several ways to do it, we think can be done, and so 22 we are just telling them to do it.
23 CHAIRMAN WARD: Well, why couldn't--I don't see why s
24 they couldn't be the subject of a regulation.
25 DR. KERR: You are certainly telling the industry HERITAGE REPORTING CORPORATION -- (202)628 4888
204 1 that they need to make temperature, RCS level and not perturb gg 2 RCS, get back up--at least I guess that's what you are doing.
3 MR. HODGES: For the plants that are out there now.
4 DR. KERR: Is that what you are going to tell new 5 plants?
6 MR. HODGES: No. New plants of the same design, 7 yes, but if you are talking about the next generation of 8 plants, we are saying they are not built yet. Now is the time 9 to fix them so you don't have this problem.
10 DR. KERR: How is it that you are going to tell them 11 what it is that they need to fix?
12 MR. HODGES: In our safety evaluation, we have put l 13 those words. We are reviewing the advanced light water 14 reactors, now the SP 90, the System 80 Plus, and the EPRI 15 requirements for the advanced light water reactor. In our 16 evaluation of those this is covered.
17 CHAIR!!AN WARD: It is covered by?
18 DR. KERR: By an SRP.
19 11R . HODGES: It is written into the safety 20 evaluation where it discusses RHP. operation and it gets into 21 problems that can occur with RH operation and it discusses the 22 kind of mid-loop concerns and it says that rather than go 23 through the things we would be doing in the generic letter, 24 design the plant so it is not a problem.
25 DR. KERR: SERs are written after the plant has been HERITAGE REPORTING CORPORATION -- (202)628-4888 _
205 1 reviewed. They are, they aren't, they aren't available to l 2 people before they design a plant, are they?
3 HR. HODGES: We are talking about plants that 4 haven't been built. They are not starting to build them in 5 the U.S. to my knowledge, but we are reviewing them, and so we 5 are saying when you build it, build it right. Don't have this 7 problem in thare.
8 The only other thing I had were those two slides 9 that showed the pictures, and that's it. Basically that is 10 it.
11 CHAIRMAN WARD: You said that about a third of 12 events that have occurred with problems with decay heat l 13 removal during shotdown have been during, have been events
- 14 where it was partial independent--I don't know if you would 15 call it mid-loop or not.
16 HR. HODGES: That is what we are calling mid-loop.
l 17 We are using that as a generic term to cover all of those.
18 CHAIRMAN WARD: Now you have talked mainly or only 1
19 about those. Does most of the risk in terms of core melt I
20 prebability come from that one third, or is there also risk 21 from the remainder?
22 HR. HAZETIS: To try to answer your question, Hr.
23 Ward, I will refer to the PRA conducted by Brookhaven and we 24 ranked CR 5015, and our follow-up reg analysis and an analysis
}
25 which I believe we sent to staff, ACRS. Brookhaven divided HERITAGE REPORTING CORPORATION -- (202)628-4888
206 1 their risk assessment up into three categories for shutdown.
ll 2 They called them the first category loss of cooling, and that 3 included those losses that you would get from losses of the 4 RHR system itself like air binding of the pumps, or closures 5 of the suction valves.
6 The other two categories were, they called the LOCA 7 category, and the loss of station power category, and when the 8 numbers were totaled on the sequences, they examined the 9 dominant sequences that showed up in the first category, the 10 lost cooling category, where that consisted of about 80 11 percent of the total core damage frequency after you totaled 12 up those three categories, so it was there also that Generic l 13 Issue 99 concentrated the effort to that, tried to examine where che cost-effective fixes could be, and it focused it,
) 14 15 turned out focusing on the areas that Wayne just went through, 16 instrumentation and the procedures in containment, so to 17 answer your question directly, yes, the risk numbers as they 18 turned out were concentrated in the loss of cooling areas of 19 mid-loop operation, and the reasons were what Wayne indicated 20 earlier. You have a reduced inventory condition, and you have 21 an open reactor coolant system and open containment.
22 CHAIRMAN WARD: Did the PRA address the containment 23 recovery actions?
24 HR. MAZETIS: Not explicitly; it addressed it only 25 in the sense of the sensitivity study, assuming three cases of HERITAGE REPORTING CORPORATION -- (202)628-4888
207 1 impact containment, open containment, or somewhere in the l 2 middle, and just to try to give the reader decision-maker a 3 sense of what value the containment was toward risk, and 4 obviously it was a large value.
5 CHAIRMAN WARD: A leaky containment?
6 HR. HAZETIS: Yen.
7 CHAIRMAN WARD: Is this your conclusion?
8 HR. HAZETIS: From the numbers, as I recall, the 9 numbers in the sensitivity study were, the mid-range number 10 was a 10 percent. Maybe Brookhaven can help me out on the 11 sensitivity study. Was it a 10 percent probability of open 12 containment? And an in tact containment, only one bound, and l 13 an open containment on the other bound, and with the 10 14 percent probability, the numbers were significant. Do you 1S remember, Louis, what they were?
16 MR. CHU: Basically we considered several cases. We 17 carried the probability, we developed t. simple containment 18 event sheet where we looked at equipment hatch, whether it is 19 open or closed containment, whether or not it is open or 20 closed, and containment spray system availability, and we I
21 assign such different sensitivity values to these top events 22 and then we calculate the risk accordingly, and in case, in 23 case of the containment is open, so it couldn't, the equipment 24 i, :h is open, then the consequence given a core damage is g
25 very high. If you '.cok at the other case, say the equipment il llERITAGE REPORTING CORPORATION -- (202)628-4888
208 1 hatch is closed, then the risk is reduced a lot. The detail h 2 is tabularized in our report.
3 CHAIRMAN WARD: I guess I am trying to figure out if 4 there is much difference be*. ween say the hundred percent open 5 case and the whatever, the 5 percent open case, or whatever 6 other cases you did?
7 HR. CHU: Maybe you can compare the case where the 8 equipment hatch is open with the other case where--
9 CHAIRMAN WARD: The penetration, yes; how would 10 those compare?
11 HR. CHU: I have to look at the table.
12 CHAIRMAN WARD: Is that in this?
l 13 HR. HAZETIS: No. It is NUREG 5015.
14 HR. CHU: It is on page 411. It looks like it is 15 effect of three reduction in the risk.
16 CHAIRMAN WARD: Where are you looking?
17 HR. CHU: Look at ' fable 4.6, the first row. The 18 first row says equipment hatch is open with probability one.
19 And then you compare it with another row where it is closed.
20 With 1 percent probability it is closed. And here this is the 21 penetration. It is open with one, probability one, so we get 22 294 versus a thousand.
23 CHAIRMAN WARD: Okay.
24 HR. CHU: Basically for the case that the equipment g
25 hatch is open we assure at least two, a PWR two release.
HERITAGE REFORTING CORPORATION -- (202)628-4888
l 209 i 1 MR. DAVIS: What kind of core damage probability did 2 you obtain?
3 HR. MAZETIS: Yes. The core damage frequency, the 4 three categories I mentioned earlier, the loss of coolant I
5 category was 4.3 times ten to the minus 5th per reactor year.
6 The LOCA category was .4 times ten to the minus 5th per l 7 reactor year, and the loss of station power category was .5 l
8 times ten to the minus 5th per year, for a total of 5.2.
9 MR. DAVIS: This assumes the plant is in shutdown 10 condition for how long?
11 HR. MINNERS: It is not a condition probability.
12 That is a probability per reactor year.
13 DR. KERR: It assumes the plant is shut down for 14 some period during the year. How much?
15 MR, CHU: That's the core damage frequency at 16 shutdown. Where it is accounted for, how frequently, you go 17 to different type of outages and how long you stay in that 18 cutage, and the outages sre divided into phases. Some phase, 19 the reactor coolant system filled. Some phase, the reactor 20 coolant system is in the mid-loop condition, and we have some 21 simple. thermal models determining time. l J2 DR. KERR: I'm sorry. You aren't addressing the l l
23 question I believe. What he is trying to find out is you use 24 the term per reactor year. It sounds as if you are assuming 25 that the plant is shut down for a full year in order to get HERITAGE REPORTING CORPORATION -- (202)628-4888
210 1 that number.
ll 2 HR. CHU: No. First we estimate how many times, how 3 many outages you have per year.
4 DR. KERR: How many did you estimate?
5 HR. CHU: There are three types of outages--for 6 buildings, .7. For maintenance outage, roughly three times a 7 year, two of them, they reached drain condition. .The other 8 one, they don't drain.
9 DR. KERR: What fracticn of the year did you assume 10 it was shut down? How many days?
11 HR. CHU: Given an outage, the time they spend in 12 shutdown condition, was estimating like 84 based on the NSAC, l 13 based on the Zion experience, and I think roughly 40 percent 14 of the time they are shut down.
l 15 I DR. KERR: So you assume that 40 percent of the year 16 they are shut down and that number is based on the reactor 17 being shut down 40 percent of the time?
l 18 HR. CHU: Right. That's the experience.
l 19 HR. MAZETIS: This explanation is on page 4 in 20 Chapter 3 of the NUREG, by the way. Those numbers that Louis 21 just gave you.
22 HR. CATTON: That is not very good availability, is 23 it?
24 HR. HINNERS: It is only 60 percent of availability g
25 on the average.
HSRITAGE REPORTING CORPORATION -- (202)628-4888
211 1 DR. KERR: But the answer to your question is no, it 2 isn't.
3 HR. MINNERS: That's their problem.
4 CHAIRMAN WARD: Make things worse.
5 DR. KERR: So if it were shut down twice that much 6 time, then the core melt probability would be about ten to the 7 minus five per year from continuously shut down reactor?
8 HR. DAVIS: I used to think they were safe to shut 9 down. Now I may have to change my mind.
10 HR. MINNERS: If they would shut down and stay shut 11 down; the problem is they go back up and increase the fission 12 product and come back down again, j 13 1 HR. HODGES: That's all I had unless you had more 14 questions.
15 DR. KERR: Indeed we also have conventional wisdom 16 they are safeer when they operate at 50 percent power than if 17 they cool down, and that may be that that is not so.
18 HR. HODGES: This does question some of the 19 conventional wisdom. It does.
20 CHAIRMAN WARD: When, what is, what is the status of 21 the generic letter?
22 HR. HODGES: We expect that it to be, well, we 23 haven't gone to CRgR. It is going through the bulk of our 24 management review. It has not completed management review 25 yet. We expect to send it hopefully this week over to CRGR g
HERITAGE REPORTING CORPORATION -- (202)628-4888
/
212 1 and then ask for an expedited review on their part. It is on 2 time. We were trying to get the thing issued by the end of gg 3 July. Hy guess, it will be the middle of August now.
4 HR. DAVIS: You may not know the answer to this but 5 I think it is true that the calculation of core melt 6 probability during shutdown is not required es part of the 7 IPE?
8 HR. HODGES: I think that is correct. I'm not an 9 expert on the IPE, but I think you are right. Most of what we 10 are talking about here would not be covered in the normal 11 PRAs, and a lot of concerns we have with nid-loop operations 12 would--the repressurization I don't think would be picked up l 13 in the Brookhaven study, so there were some problems there, 14 and it was, it act part of the problem, but I don't think it 15 got the ful; problem.
16 DR. KERR: Has any thought been given to subsuming 17 this problem as part of the IPE?
18 HR. HODGES: No, sir. We are trying to get 19 something done on this as expeditiously as we can.
20 DR. KERR: This is more serious than the general 21 problem of decay heat removal?
22 HR. HODGES: It may well be that if you fix this 23 problem, you may have done a lot to help that one also. I 24 don't know.
W 15 CHAIRHAN WARD: Oh, really? How is that?
HERITAGE REPORTING CORPORATION -- (202)628-4888
213 1 HR. HODGES: If the problem is loss of decay heat g 2 removal, I think that is what Generic Issue 99 was about was 3 the loss of it. Basically we are fixing that. You may be 4 good enough already. I don't know.
5 Obviously the, and I'm not an expert on A-45, but if 6 the study showed it was a very expensive fix to get the kind 7 of things that were needed there, then maybe you don't need to 8 do a lot more on a cost / benefit basis; and this may be much 9 more effective than what is planned there. I'm not an expert 10 there, so I should probably keep my mouth shut.
11 CHAIRMAN WARD: But tha things you are talking about 12 here aren't going to do anything to help the sequences that 13 are considered important in the A-45 studies?
l 14 HR. HODGES: No.
25 '
CHAIRMAN WARD: Okay.
16 HR. MINNERS: That's why these are two separate 17 j issues.
18 HR. HODGUS: What we are saying is that we think it 1
We want to get something fixed on it 19 is a serious problem.
20 fairly quickly and that's why we are taking the course that we 21 are taking.
22 DR. KERR: If you use the same logic, and the A-45 23 preblem are more serious, you want to get it fixed nore 24 quickly?
25 i HR. HODGES: Richt.
HERITAGE RFPORTING CORPORATION -- (202)628-4888
214 1 DR. KERR: And it's apparently well in hand enough 2 that we are going to wait until the IPE?
3 MR. HODGES: It would probably be prudent if I 4 didn't say much more.
5 DR. VERR: As you say--I guess I can't quite see why 6 this is so thoroughly separate from the issue of decay heat 7 removal. But, well--
8 MR. MIllNERS: It is different fixes.
9 CHAIRMAll WARD: I think the thing is they see a 10 bunch of nickle and dime fixes here and it is a hundted 11 million dollars a piece for the other.
12 MR. HODGES: We see the nickle and dime fix as being 13 effective for the problem we are solving and we see the 14 problem we are solving being a fairly significant problen.
15 DR. KERR: Is this required instrumentation going to 16 be, it is going to be safety grade?
17 MR. HODGES: llo . We want redundancy. We are not 18 going to require safety grade.
19 HR. WARD: Seismic resistance?
20 HR. HODGES: tio seismic resistance; probably control 21 grade, but we are not even going to insist on that as long as 22 it has got quality.
23 CH AIRM A!( WARD: Any other questions for Mr. Hodges I
24 or Mr. Hazeti on thin?
25 flow I guess I nean the Generic Issue 99 resolution i
l HERITAGE REPORTIllG CORPORATIO!1 -- (202)628-4888
215 1 still has to be developed, 2 MR. MAZETIS: Well, no.
h 3 CHAIRMAN WARD: Do you think, is this going to take 4 care of it?
5 MR. MAZETIS: It is our view in Research that what 6 Wayne just went through which in essence would be the generic 7 letter, whatever the format is would encompass the proposed, 8 would be adequate to resolve Generic Issue 99.
9 CHAIRMAN WARD: So you will write off Generic Issue 10 997 11 MR. MAZETIS: I'A not sure what the format would be.
12 Perhaps an internal memo to the EDO, but ye.1, that's the idea.
l l 13 MR. HODGES: The only real difference initially 14 { showed up between their proposed solution and our proposed 15 solution. The two different problems was I don't think we 16 initially included anything on the interlock for the 17 overpressure protection and that is, showed up to b? a major 18 contributor to the loss of decay heat removal, and so they had 19 that in their proposed solution for Generic Issue 99, but if 20 you think about it also it means perturbing the system and so 21 we definitely would just subsume it in, put the two together.
22 CHAIRMAN WARD: Are you giving up some overpressure 23 protection with whatever step you are taking? Sounded like i
24 l you might be.
I 25 MR. HODGE3: Not really; the solutions, I mean you HERITAGE REPORTING CORPORATION -- (202)628-4888
216 1 are using an alternative means of doing it rather than Paving 2 interlock.
3 HR. MINNERS: The purpose of interlock was not 4 overpressure protection. That's not the purpose of interlock.
5 Never was.
6 CHAIRMAN WARD: Was just what? Operating--
7 HR. MINNERS: The purpose of interlock was to assure 8 both valves were closed and you didn't have a LOCA.
9 HR. MAZETIS: During power operation.
10 HR. MINNERS: It was never for overpressure 11 protection.
12 CHAIRMAN WARD: It was event V protect. ion. Does l 13 that mean there is going to be a higher risk from event V?
14 MR. MAZETIS: I think the idea is that, that the 15 recommended approval of removal of ACI, of the autoclosure 16 interlock, would follow previous approvals of requests by 17 certain utilities where they have accommodated the potential 18 increane in risk during power operation with other features, 19 and--
20 MR. HODGES: I don't remember all the details. It 21 has been done for Diablo Canyon I know. There is a 22 Westinghou.se proposal that is generic, that looks at it. I 23 don't remember the details, Dave.
24 CHAIRMAN WARD: I kind of get the impression that g
25 last year's problem is, is being neglected.
HERITAGE REPORTING CORPORATION -- (202)628-4888
l 217 1 HR. HODGES: Let us get you an example of what is It has only been done for Diablo Canyon, and l 2 done there.
3 that's, that should help answer the question. I don't 4 remember the details. Rather than saying something that is 5 wrong, I don't want to tell you anything. I will get back and 6 give you a cooy of what has been done.
7 CHAIRMAN WARD: Okay. Anything else? All right.
8 MR. MAZETIS: Could I make a little clarification?
9 On the autoclosure interlock, the numbers as they came out as 10 far as cost / benefit somewhat supported removal of auto 11 interlock, but the, the risk reductions were not that 12 overwhelming. They are arguable at best, so I think
( 13 Research's view is while we endorse, we would endorse remo'ral 14 and talking just how the number came out, while we would 15 endorse removal of the autoclosure interlock, that probably 16 the rnmbers don't support requirement. It is not obvious that 17 they would support a requirement to remove the autoclosure 18 interlock.
19 CHAIRMAN WARD: Okay. But this is, this is the 20 recent, the PRA?
21 MR. MAZETIS: The PRA, right.
22 CHAIR!!AN WARD: And that's debiting whatever 23 increased risk there is for event V--
24 MR. MAZETIS: That is correct.
25 CHAIRMAN WARD: Okay.
HERITAGE REPORTING CORPORATION -- (202)628-4888
218 1 LA. HODGES: It has been considered. I know that.
g 2 We will show you the results of what was done at Diablo 3 Canyon, their study. I just don't remember the details.
4 CHAIRMAN WARD: Okay. Thank you very much. Let me 5 turn to the Subcommittee members now, and I would like to get 6 your thoughts--let's talk about Generic Issue 99 first and go 7 back to A-45--your thoughts on whether you think that the, you 8 know, what is being undertaken here by the staff is 9 appropriate, and whether they define the issue properly, and 10 whether the action they are proposing is appropriate.
11 Then also whether you think that we should make some 12 comment on it if you don't think it is fully appropriate, and l 13 in doing that, should we bring the issue to the, let the Full 14 Committee hear about it? So let me--Pete, could I start with 15 you?
16 HR. DAVIS: I guess I don't have any strong feeling 17 right now, Hr. Chairman. This is a pretty new issue, and I 18 haven't had a chance to study the Brookhaven document. It 19 appears like, it appears like it is appropriate, and I guess 20 that's all I have.
21 CHAIRMAN WARD: Ivan?
22 HR. CATTON: I think what Wayne had to say sounded 23 sensible. If indeed the containment is closed whenever they 24 have these kinds of circumstances, very shortly after the 25 pcwer has come down, part of the problem sort of goes away and HERITAGE REPORTING CORPORATION -- (202)628 4888
~
219 1 on the other hand, when the containment is open, some time has g 2 passed. The forcing function is lower, so I think his 3 approach is is sensible, although I would like to see the 4 generic letter.
5 CHAIRMAN WARD: Charlie?
6 HR. WYLIE: I agree. I think that what they propose 7 is a sensible way to go about the problem. I don't see it as 8 a big burden on anybody. I think it makes good sense as an 9 approach.
10 HR. CATTON: It is somewhat unusual actually.
11 CHAIRMAN WARD: Bill?
12 DR. KERR: In the face of all that wisdom, I don't l 13 know why I feel uneasy. I guess this is a problem, and '2 14 quess it is important enough that it should be picked out for 15 immediate resolution, but I don't know enough about what the 16 alternatives are nor do I know enough about what their actual 17 risk is.
18 That is not meant to be a criticism of the 19 ,
Brookhaven report. I'm sure given the things they assumed 20 their report probably makes sense, but whether that risk study 21 is representative of what is going on, I have no idea. And it 22 also puzzles me that this is not part of the decay heat 23 removal problen, so I'm uneasy about the problem not by 24 litself, but in the context of real reactors and real task that 25 reactor operators have of dealing with one more requirement.
HERITAGE REPORTING CORPORATION -- (202)628-4888
220 1 HR. WYLIE: I'm not sure it is in the same context 2 as a requirement. To me it seems like we have addressed the 3 probler, they are just saying take a look. When you are down 4 at this condition. it would be prudent for what you did. It 5 doesn't cost that much.
6 DR. KERR: I do not have the temerity to disagree 7 with such a triumvirate. I simply expressed some uneasiness.
8 CHAIRMAN WARD: I have got some, too, Bill. You 9 know, this issue has surfaced or the rock, this rock has been 10 turned over, whatever the mixed metaphors, but you know on the 11 one hand we hear that there is pretty significant risk from 12 the loss of decay heat removal and -hutdown and it does have l 13 some characteristics that are different from loss of cooling 14 and power, but I'm not sure they are powerful enough to 15 compensate for this drastic difference in defense in-depth 16 that's available or required to be available, and I really 17 haven't heard the satisfactory argument.
18 HR. WYLIE: You mean the calculated risk seemed to 19 be fairly high as well (s core melt frequency?
I 20 CHAIRMAN WARD: Yes.
21 HR. WYLIE: But the time olement, you have got 22 better control of this situation.
23 DR. KERR: What we have got was the, on the risk 24 probability of core nelt and it was not very much less than g
25 one tre get fron all the other contributors.
I HERITAGE REPORTING CORPORATION -- (202)628-4888
221 1 HR. CATTON: The magnitude I think is, just thumbing 2 through this--
h 3 DR. KERR: Ten to the minus five.
4 HR. CATTON: Actually EPRI has published a couple of 5 reports what three, four years ago?
6 CHAIRMAN WARD: '83.
7 HR. CATTON: Really sort of appealing for this kind 8 of action. One has to wonder why the staff took so long to 9 respond to a problem that was pointed out by the industry.
10 CHAIRMAN WARD: Well, the industry, it was an 11 industry record. The indust *y had every opportunity to 12 respond.
l 13 HR. CA1 TON: Maybe the staff has to say the industry 14 doesn't do anything.
15 CHAIRMAN WARD: It sounds that is why, doesn't it?
16 ,
HR. DAVIS: I wanted to ask you why this problem has 17 only been considered in the context of PWRs? Because it is l
18 '
not a problem for boiling water reactors? And wh; is that the 19 caso?
20 HR. HINNERS: I don't think they run a mid-loop 21 operation.
22 HR. CATTON: I think there is an NSAC report of BWRs 23 and it ic a problen there as well if I romenber right. I
~
24 ' don't recall. I think the author of the report was here.
l t
25 Wasn't it you?
HERITAGE REPORTING CORPORATION -- (202)628-4888
222 1 HR. VINE: Yes.
2 HR. CATTON: Are DWRs a problem?
3 HR. VINE: It is a very different problem. It is 4 not related at all to mid-loop, and what we found in the cas3 5 of BWRs was .aore an issue of such things as drain downs 6....
6 something as thet is easily donc, handled by procedures, 7 administrative controls, but it is not related to t!:u kinds of 8 sequences that you are looking at under Issue 99:
9 HR. NEILS: Don't pump down the reactor inventory.
10 CHAIRhAN WARD: What do you think we ought to do?
11 Is this something that the Full Committee could, probably 12 shouid hear about?
13 HA. WYLIE: I would think so.
14 l DR. KERR: I agree.
15 HR. CATTON: Shouldn't yau have the generic letter 16 in hand when you tell the Full Committee about it? It sounds 17 to me it is like a draft that we Pave floating around.
18 CHAIRMAN WARD: That might be a gcod time to go to l
19 <
the Full Conmittee. When will the gener'.c letter be, whun l
20 would you be willing to share a draft?
l 21 HR. HODGES: I don't know the, what is the normal 22 l procedure? Do they normally show it to you before going to 1
23 fthe CRGR?
24 CHAIRMA.I WARD: It is up to the branch chief.
I 25 l MR. HODGES: I don't have any problem with that, as i
i HERITAGE REPORTING CORPORATION -- (202)628-4888
223 1 soon as it is signed out by, signed out by CRGR.
ll 2 CHAIRMAN WARD: Well, I guess so. What we could do 3 is get the generic letter to be read and then ask you to come 4 in and give this sort of a presentation, just be a little 5 $arther along. Okay. Why don't we do that? Well, I guess it 6 isn't absolutely essential we schedule that for the August 7 meeting, but in fact it probably ought to be the September 8 meeting if I understood your schedule on the generic letter.
9 MR. HODGES: I don't see how it can get out before 10 the middle of Augut.t.
11 CHAIRMAN WARD: Why don't we schedule--you know, I 12 think we ought to schedule a couple of hours at the August, at l 13 the September meeting, if I can talk our Chairman into it.
14 j HR. HODGES: Sure.
15 HR. CATTON: I think it would be helpful when they 16 '
they separate the two kinds of incidents, the ones l do that, 17 occut when they just come aown from power as contcasted lthat 18 to when they spend a week or two. They are very different.
19 CHAIRMAN WARD: I think it would be a good idea also 20 to, if we could have a summary presentation on the risk 21 analysis that you have done so that the, set the context so we 22 understand it a little better.
23 MR. MAZETIS: Sure.
24 '
CHAIRMAN WARD: I would like te turn now to A-45, 25 and ask our consultants, Subcommittee members, the sacs set of HERITAGE REPORTING CORPORATION -- (2021628-4888
224 1 questions and start at either end. Start with you, Bill?
g 2 DR. RERR: I think it is logical to handle the A-45, 3 former A-45 issue as part of our IPE and it seems to me the 4 studies have been, that have been dooe will make it, wel? at 5 leant will point toward some methods of treatment that can be 6 looked at, and I think including it as part of the total risk 7 picture rather than an isolated statement is a good thing to 8 do.
9 I am puzzled as to how the treatment of the seismic 10 part of the analysis that were made is going to be separated, 11 but I guess that remains to be seen, so that I suppose at this 12 point, that letter would probably comment on whether we agree l 13 that it makes s'.nse to put this in as part of the package, and 14 with some caveats which are not significant, I would endorse, 15 I would endorse what I think is being done.
16 CHAIRMAN WARD: Yes. Charlie?
17 MR. WYLIE: I would agree with that. I assume the 18 letters that we have written 'ichin the last few months 19 regarding the recommen'.tions e the ACRS on closure severe 20 accident issue, the integration ta the IPE and so forth, 21 j would pertain to tnis subject as well. We wrote a letter July I
22 ! 20 ano one in May that are related to this.
23 LR. KERR: At one point, my impression was tb '
24 there was some serious disagreement among the staff a' the I,
25 l'possibilityof a dedicated decay heat removal system. I don't d
4 h
HERITAGE REPORTING CORPORATION -- (202)628-4888 \
225 1 know whether that staff member is still around.
g 2 CHAIRMAN WARD: I think there was.
3 DR. KERR: Whether we need to mention that; I simply 4 remember it.
5 CHAIRMAN WAPD: I think one staff member who had a 6 Mey role in A-45 took the position that the dedicated add-on 7 system should be required not on the basis of quantitative 8 analysis, but just based on juogment about a range of things 9 that could only be done by subjective--
10 DR. KERR: I mention it not because I share in the 11 view, but he did a play a prominent part in the views of the 12 person who had a good bit of responsibility in the l 13 investigation of it.
14 CHAIRMAN WARD: Charlie, did you have anything more?
15 MR. WYLIE: No.
16 CHAIkMAN WARD: Ivan?
17 MR, C ATTC'N : I think the letters that you have 18 ; already written particularly on IPE seem to cover what has I
19 ! happened today. They are going to full A-45 and IPE. It is 20 sort of like the stockmarket, but it has been discounted 21 { already.
22 l On the other hand, it is sort of unsettling to see j
l 23 fthe way the process goes. You spend years fooling around with i
24 it, and I know this person you are referring to, and I happen 25 to have a lot of respect for his judg.ent,m and you just sort I
\
HERITAGE REPORTING CORPORATION -- (202)628-4888
226 1 of grab hold of this thing and it pops out of your hand like a g 2 greased olive, greased pig. It is gone. It is going to fall 3 into IPE. What is coing to happen with IPE when they don't 4 know what to do with it? Are they going to go on now to a 5 full scope PRA before they can make the decision? Thin means 6 ten years downstream or what? It seems to me for some of the 7 time you ought to bite the bullet.
8 CHAIRMAN WARD: Pete?
9 DR. KERR: You sort of need to know what bullet you 10 are biting.
11 MR. CATTON: After what, four years of this, you 12 sort of--
13 CHAIRMAN WARD: Six and a half.
14 HR. CATTON: You sort of think you know the 15 vulnerabilities. The way to do them, it costs a lot of money, 16 but there are ways to cure it. Then there were arguments that 17 gee, this is just a tip of the iceberg. There are other il
) 18 I plants out there that are much worse. We really ought to do 19 something more serious.
20 i CHAIRMAN WARD: Well, I think, you know, the i
l 21 l dedicated add-on system got a lot of consideration, and 22 l partially from some stimulus from the ACRS early on, and also d
23 because we are observing what some of the Europeans were i
24 doing--not all the Europeans, but some of them. I we,uldn't J
25 l say most of t h e .3 , certainly not in numbers of installed I
i HERITAGE REFORTING CORPORATION -- (202)628-4888
227 1 megawatts, not by a long shot, 2 But I think, you know, it was considered and a lot h
3 of study, of the Sandia study was dedicated really to figuring 4 out whether that should be done, what kind of case could you 5 make for it? Well, the answer came out I think in the 6 negative, that you couldn't. It wasn't justified in 7 quantitative or simple engineering terms, that if you wanted 8 it, it had to be justified in some sort of political terms or 9 really--which is why the Europeans are putting it in.
10 HR. CATTON: Politics is real even here. You know 11 that.
12 CHAIRMAN WARD: I don't know, but you don't think we l 13 as a committee should make a contribution in that way?
14 HR. CATTON: Whenever you take the ratio, or the l
15 cost / benefits valuem impact or whatever you call it, we know 16 there is orders of magnitude uncertainty, at least order of 17 magnitude uncertainty. That Switzerland study we heard about 18 where they took a group of people that did a simple study of 19 i action feedwater system, and even talked to the people who l
20 l were building it and they were a tactor of 30 different in the i
21 ! unreliability, and that's just one little piece this whole l
22 thing, so I think that Bill's observation about a factor of 23 ten is res1. It may be more than a factor of ten. It may be 3
l 24 a factor of 30 or 40. Maybe that 30 is a real number, and 25 l. expected differences. If that's the case, multiply some of F
1 i
HERITAGE REPORTING CORPORATION -- (202)628-48S8
228 1 those numbers by 30 and things that are not cost effective g 2 become cost effective. I'm not too much in favor of value 3 '.mp a c t .
4 DR. KERR: The reason I am a little leery of 5 dedicated independent decay heat removal system is that I 6 would guess that after they are installed and in operation for 7 ten years we would discover that they have the same warts that 8 the other systems have, that they have common mode systems, 9 and they aren't really as independent and dedicated as we 10 thought.
11 CHAIRMAN WARD: Yet another one.
12 DR. KERR: If I could take a piece of paper and l 13 determine that would cure the problem, I would be enthusiastic 14 about it 15 HR. CATTON: The trouble is I think too often we 16 just take a piece of paper.
17 CHAIRMAN WARD: Pete?
18 HR. DAVIS: I endorse the decision to put A-45 in 19 the IPEs. I am concerned about the things that have already 20 been brought up of external events. It is not clear to me 21 either how that is going to be handled since that's not part 22 of the IPE scope, and yet of the six plents, over half of the 23 risk comes from external events on the average, and I an also 24 L concerned about the delay, but I don't think anything can be U
25 ! done about that.
HERITAGE REPORTING CORPORATION -- (202)628-4888
i l
229 1 CHAIRHAN WARD: I guess I would like to make a ll 2 couple of comments. I think pushing this into the IPE process 3 is probably okay. I think the Committee has to encourage that 4 as a general proposition, but right now the IPE is just sort 5 of a big empty box, and I'm not sure what it going to be done, 6 and I think there was, I think the attempts to make some 7 quantitative judgments in A-45 provided some sort of 8 discipline to the process and some sort of hard, you know, 9 stepping stones for judgment that we don't have yet in the 10 IPE.
i 11 j I mean I think there is a tendency in the staff--and 12 Bill, you alluded to this a couple of weeks ago--I think there l 13 is a tendency in the staff to develop some sort of a process l 14 for doing something and then hoping that scue kind of good l
15 '
Mill fall out of the process, and without having particular 16 goals or aims for what the process is intended to do, other 17 ;
than optimistically something good is going to happen, and it 18 seems to me that right now that's kind of what IPE is, and l
19 ! there is probably some, some truth in that, that some good i
20 does come up, but I don't think that's the optimum way of 21 regulating or should be the only regulatory tool because I 22 l sort of see a pattern developing that every one of these l
23 !! issues that come up, whether it is the bunch of things we call b
24 severe accident issues or the decay heat removal issues, a 25 bunch of, number of plants are examined and samples of the HERITAGE REPORTING CORPORATION -- (202)628-4888
[
230 1 population, and the staff tends to throw up its hands and say g 2 gosh, we can't draw any general conclusions about this at all 3 but we are just going to have to examine esch one of the 4 plants and then hope somebody, either the owners of the plant 5 or the project manager for NRC or somebody is going to think 6 of something good to do with that plant result of the 7 examination. I mean it almost seems like that is becoming the 8 regulatory process.
9 I guess the last point is I'm disappointed that 10 there hasn't beer. any kind of traditional regulatory safety 11 insight that has come out of this whole A-45 study. Even 12 though, you know, the main thrust conclusion is that the plant l 13 has to be examined and each plant has to be examined in detail 14 because they are so different.
15 It seems to me that there might have been some, some 16 general conclusions drawn. You know, I hate--we have been 17 pushing the risk perspective for, the Conmittee has been 18 pushing the risk perspective as a useful way to evaluate 19 safety and even regulate, and I don't really want to back down 20 on that, but I don't think that the risk perspective is 21 lcapableof taking over a hundred percent of the job of the 22 1j safety evaluation and regulation. And well, for example, INPC i
23 in its SOER program looked at decay heat removal events in a
h 24 ll some depth and came up with some sort of relatively simple l
25 common sense not risk based arguments for things that plants HERITAGE REPORTING CORPORATION -- (202)628-4888
231 1 should, tor things that savald be done, and I really think we 2 continue to need bcth of those perspectives, and I 3 don't--maybe the staff feels that 111P0, for example, has taken 4 care, has taken care of that, that simpler more direct S experience reaction perspective, but I don't know. The staff 6 seems not to desire, not to have any control over that or 7 really participate in that. I guess I question whether that's 8 the proper way to continue. Kind of rambling--any ott'T 9 concise profound statement?
10 MR. WYLIE: Are you, what about the, to the Full 11 Committee?
12 CHAIRMAll WARD: Oh, yes--I think, I really think we 13 need to spend, to let the Committee, the Full Committee spend 14 a little time with this. Paul, we have something scheduled?
15 MR. BOEH!!ERT: Two hours.
16 CHAIRMAll WARD: Do you think that is enough? Bill?
l 'i DR. KERR: I would suggest at least it would--maybe 18 three.
19 ! MR. MIlillERS: Do you want both song and dances?
l l
20 l DR. KERR: I don't think we want the--are you 21 I talking about the--
l 22 MR. MIlillERS: Do you want tlUHARC and us, too?
23 CHAIRMAll WARD: Yes. What do you think on that?
24 I Maybe that isn't particularly useful to hear that?
i 25 b DR. KERR: I would say not, h
HERITAGE REPORTIllG CORPORATION -- (202)628-4888
232 l 1 CHAIRMAN WARD: Okay. I think the only way that is l 2 being used is I, the point I think is correct that was brought 3 up in that this has kind of given IPE, whatever is going to 4 be, this may be a precursor of, you know, the wave of the 5 future and whether you want, whether you want the Cammittee to 6 see some of the realities of what that process is going to be 7 like before they get too enthusiastic about IPE?
8 MR. CATTON: It might be helpful.
9 CHAIRMAN WARD: I don't know.
10 MR. CATTON: Whether is there wide disagreement.
11 CHAIRh.!v WARD: I think it might be--why don't we 12 plan on three hours, and try to work out some reasonable l 13 combination of the pure A-45, battle of the PRAs?
14 DR. KERR: We are now scheduled for two, Paul?
15 MR. BOEHMERT: Yes.
16 MR. WARD: Sort of arbitrarily.
17 DR. KERR: I just--we can work until 10:00 p.m..
18 CHAIRMAN WARD: We still have two hours to lunch! I 19 think we are finished, as a matter of fact. Does anyone else 20 have anything they would like to say?
21 Gentlemen, thank you very much.
22 (Whereupon, at 3:45 p.m., the meeting was 0
23 i t adjourned.)
h P
24 h 25 j h
l HERITAGE REPORTING CORPORATION -- (202)628-4S88
1 CERTIFICATE
() 3 This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the matter of:
5 Names Advisory Committee on Reactor Safeguards, Subcommittee on Decay that Removal Systems 7 Docket Number:
8 Place: Washington, D.C.
9 Date: July 27, 1988 10 were held as herein appears, and that this is the original 11 transcript thereof for the file of the United States Nuclear 2
12 Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewriting by me or under the direction 14 of the court reporting company, and that the transcript is a 15 true and accurate record of the foregoing proceed gs.
0 16 /S/
MW" I
cu i
17 (Signature typed): Catherine S. Boyd f
18 official Reporter 19 Heritage Reporting Corporation 20 21 22 23 24 i , .
i 25 i
() Heritage Reporting Corporation (202) 628-4888
O RECOMMENDED Filial RESOLUTION OF USI A-45 "SHUTDOWN DECAY HEAT REMOVAL REQUIREMEllTS" j
PRESEllTED TO THE ACRS SUBCOMMITTEE FOR DECAY HEAT REMOVAL l
JULY 27, 1988 O
~
l i
i l ROY WOODS, SENIOR TASK MANAGER j DIVISION OF REACTOR AND PLANT SYSTEMS OFFICE OF flVCLEAR REGULATORY RESEARCH !
l I
O
C
SUMMARY
' UNRESOLVED SAFETY ISSUE (USI A-45) APPROVED DEC, 24, 1980 O (SECY-80-325)
KEY OUESTIONS:
D0 CURRENT REGULATIONS PROVIDE SUFFICIENT ASSURANCE THAT RISK FROM DHR FAILURES IS ACCEPTABLY LOW?
- ARE IMPROVEMENTS TO DHR FUNCTI0tl IN OPERATING PLANTS COST-BENEFICIAL?
i KEY CONCLUS10f1S:
l Q DHR FAILURES ARE SIGlilFICANT CONTRIBUTORS TO.C0RE DAMAGE FREQUENCY FROM SB-LOCAs, TRANSIENTS '
I i i VULNERABILITIES AND CORRECTIVE ACTIONS, COST / BENEFIT RATIOS ARE PLANT SPECIFIC l DEDICATED DHR SYSTEM NOT COST BENEFICIAL l
DESIGNS NOT COMPARED TO "CURRENT" DBA-BASED REQUIREMENTS i
- KEY ASSUMPTIONS: ;
- CONSISTENT WITH A-44 Arid A-49
O ll l
USI A-45 SCOPE SIX CASE STUDIES, DHR FAILURE RELATED PRA'S (SUNilARIZED IN ENCLOSURE B)
LIMITED TO SYSTEMJ NEEDED TO RESPOND TO TRANSIENTS AND SMALL BREAK LOCAs EVALUATED SUCH SYSTEMS' VULNERABILITY TO FIRE, FLOOD, SEISMIC, INSIDER SAB0TAGE i
l.
l.
I i
9 ,
j I
O CONTENT OF REGULATORY ANAliSIS 1
APRIL, 1988 DRAFT REVISION DESCRIBES SIX ALTERNATIVES AND PROPOSES RESOLUTION WITil ALT. # 2 (PLANT-SPECIFIC ANALYSIS)
ALTERNATIVE 1 - NO ACTION COULD BE ACCEPTABLE If HRC ANALYSIS RESULTS ARE OVERLY CONSERVATIVE (EPRl/WOG: POINT BEACil)
ALTERNATIVE 2 - LIMITED SCOPE PRAs SEVERE ACCIDEllT PROGRAM IPE/IPEE O ALTERNATIVE 3 - SPECIFIED SYSTEMS MODIFICATIONS USis AND GIs ALTERNATIVE 4 - DEPRESSUR12ATION AND COOLING PWR - FEED AND BLEED ,
BWR - CONTAINMENT VENTING I
ALTERtlATIVE 5 - DEDICATED HOT SHUTDOWti CAPABILITY ALTERNATIVE 6 - DEDICATED COLD SilUTDOWN CAPABILITY I I
l l
O l 1
I TECHfilCAL FINDINGS O
FREQUENCY OF CORE DAMAGE DUE TO DilR FUNCTION FAILURE (P(CM)DHR) AVERAGES 2 TOPER 3 xR-YR (INCLUDES 10-4 l I
INTERNAL AND EXTERNAL CAUSES) l SUPPORT SYSTEM FAILURES (E.G., EMERGE!!CY POWER, SERVICE WATER, COMP 0NEllT COOLING) CONTRIBUTE SIGNIFICANTLY TO P(CM)DHR REDUNDANCY CONCERNS AND CONSIDERABLE SHARING OF SYSTEi1S, PARTICULARLY AT SUPPORT SYSTEM LEVEL FOR SOME PLANTS 4
O
- CONCERNS WITH OVERALL GENERAL ARRANGEMENT OF EQUIPMENT FROM A SAFETY VIEWPOINT, E.G., LACK OF INDEPENDENCE, SEPARATION j
& P!!YSICAL PROTECTION OF REDUNDAllT SAFEGUARD TRAINS [
FIRE, FLOOD, SEISMIC, SAB0TAGE RISK CONCERNS i !
t RELATIVE IMPORTANCE QF. yj)LNERABILITIES H PLANT-SPECIFIC EFFECTIVENESS 01 CORRECTIVE ACTIONS Bi REDUCING P(CM)DHR E ;
PLANT SPECIFIC !
l 4
l t
'_____________ i
REG, ANALYSIS O
DIFFERENT APPROACHES TO VALUE-lMPACT ANALYSIS VALUE-lMPACT ANALYSIS PERFORMED 3 WAYS:
A. VALUE TERM: AVERTED DOSE TO POPULATION IMPACT TERM: COST OF IMPLEMENTATION !
B. VALUE TERM: SAME AS NETHOD "A" EXCEPT AVERTED !
ONSITE DOSE INCLUDED ;
IMPACT IEBRE: COST OF IMPLEMENTATION LESS AVERTED ONSITE COSTS C, SAME AS METHODS A a B PLUS THE SAVINGS FROM O SPECIAL CONSIDERATIONS (E.G., SAB0TAGE,
~
MORATORIUM, RESOLUTION OF OTHER GENERIC ISSUES, UNQUANTIFIABLES) !
1 RESULTS: (
I !
- METHOD A - ALTERNATIVES 2, 3 a 4 MAY BE COST-EFFECTIVE METHOD B - ALTERNATIVES 2, 3 a 4 MAY DE MORE !
?
COST-EFFECTIVE !
O _
i I
REG. ANALYSIS (CouT'D)
O DIFFERENT APPROACHES TO VALUE-!MPACT ANALYSIS r ALTERNATIVES 2, 3, a 4 DO NOT MEET Ti1E STAFF'S CDF G0AL OR REDUCE SAB0TAGE RISK i
METHOD C - ALTERNATIVES 5 a G MAY BE
- COST-EFFECTIVE AND REACH CDF G0AL STAFF ENDORSES ALTERf1ATIVE 2, PLANT-SPECIFIC ANALYSES,
. ON FOLLOWING BASES:
1 N0 GENERAL RESOLUTION 4
- ALTERNATE 4 (fab) - T00 MUCH RELIANCE ON O OPERATOR j - ALTERNATE 536 (ADHR) - NOT COST EFFECTIVE UNLESS USE METii0D C (MORATORIUM, ETC) WHICil GOES DEYOND VALUE/ IMPACT METHODS PREVIOUSLY
- USED FOR US!s/Gis i
i PLANT SPECIFIC RESOLUT10fi
- MANY RISK CONTRIBUTORS ARE PLANT-SPECIFIC l
- ONLY WAY TO IDEliTIFY DHR VULfiERABILITIES IS THRU PLANT-SPECIFIC EXAMINATIONS
- - EFFECTS OF CORRECT!VE ACT!0fiS ARE PLAfiT-SPECIFIC j - CDF AT MANY PLANTS ABOVE STAFF-SELECTED C0AL O 6
f
?
I O iMeteseNTATiON DeTAlts l
1 (ALTERNATIVE 2) {
l NRC PROPOSES TO REQUIRE A PLANT-SPECIFIC EXAMINATION TO i t
IDENTIFY VULNERABILITIES TO SEVERE ACCIDENTS (Tile IPE j PROGRAM) i IPE WILL INCLUDE, BUT NOT BE LIMITED T0, DliR FAILURE i
~
RELATED CORE DAMAGE EVENTS j
SEPARATE, DEDICATED "A-45" EXAMINATION WOULD BE REDUNDANT I l O
- WE CONCLUDE TilAT A-45 SHOULD BE SUBSUMED INTO IPE/IPEE i i l !
INSIGHTS GAINED FROH SIX CASE STUDIES AND EPRI-WOG l ANALYSIS (PLUS NRC/SANDIA REVIEW, SEE APPENDIX D
- f TO ENCLOSURE A) WILL BECOME EXAMPLE FOR LICENSEES, f
l I I
f l
t
- i
!O 7 l
i !
i i
i 4
O ((UMARC/EPRI/WOG PRINT BEACH PRA
- (APPENDIX D TO ENCL A) i (
SUMMARY
OF MARCil 31, 1988 MEETING WITil NUMARC)
NRC REVIEWED FOR TWO REASONS:
- 1) ANSWER POSSIBLE CLAIM THAT ALTERNATIVE 1 IS JUST!FIABLE IE NUMARC PRA IS CORRECT, AND IE PB IS A "BOUNDING" PLANT (THEN ALT. 1 WOULD BE ACCEPTABLE)
WE ARE NOT CONVINCED THAT EITHER OF THE AB0VE ARE CORRECT
- 2) EXAMPLE OF NRC REVIEW 0F AN IPE, ll!GilLIGHTS DIFFERENCES BETWEEN NRC AND INDUSTRY:
METHODS
!!UMERICAL ASSUMPTIONS O 8
, O EPRl/WOG STUDY RESULTS !
(NSAC - 113) !
1 CORE DAMAGE FREQUEllCY PER YEAR .
SOURCE OF RISK @_C EPRI/WOG (RF)* REVISED NRC (RF)*
INTERNAL 1.4E-4 2.6E-6 (54) 2.5E-5 (6) l SEISMIC 6.1E-5 7.4E-6 (8) 4.1E-5 (1,5)
FIRE 3.2E-5 6.3E-8 (500) 2.2E-5 (1.5)
INTERNAL FLOOD 7.7E-5 1.0E-8 (7700) 9.8E-7 (79)
EXTERNAL FLOOD 1.9E-8 1.0E-8 (2) - - >
WIND 4.0E-6 1.0E-8 (400) 1.7E-7 (24) l
]O 4
tion 1Nina 3.se_a 1.0E.8 (6) _ _
i
! TOTAL 3.1E-4 1.0E-5 (31) 9E-5 i
) (3.5) !
i i'
1
'8 EDUCT!0f4 EACTOR COMPARED l 4
To "NRC" !
1 9 :
O I
1 I . :
l l
O CDF G0AL FOR APPLICATION TO A-45 l
l MUST ANSWER QUESTinN - IS DHR FAILURE RELATED CDF HIGli Efl00GH {
TO JUSTIFY RECUIRIfiG PLANT SPECIFIC ANALYSIS?
STAFF CURRENTLY CONSIDERING PROPOSING MORE GENERAL USE OF OVERALL 10-4/RY YR CDF*
BELIEVE DHR FAILURE RELATED CDF 1/3 TO 1/2 0F OVERALL CDF l
THEREFORE, NEED TOTAL DHR FAILURE RELATED CDF 3x10-5 j O :
TO ACHIEVE THAT, NEED CUAtlTIFIABLE DfiR FAILURE RELATED CDF 1x10-5 (CPERATOR ERRORS, ACTS OF COMMISSION...)
i l IflTENDED ONLY FOR A-45. NOT IMPLIED THAT IPE SHOULD ADOPT.
1 ;
- i COMPATIBLE WITil STAFF'S A-44 AND A-49 G0ALS l l I f
[
- ColdSISTEf4T WITH LARGE RELEASE CDF OF 10-6/RY FOR EXAMPLE IF 1:10 RATIO BETWEEll CMF At4D CDF, At:D 1:10 RATIO BETWEEll !
LARGE RELEASE FRE0 AND CMF (
E O l t
i I
O ~
O O
~
DVERVIEW T ALTERHTIVES 04 GDEltlCAl' SIS - IW1 (GIST WR P-IU1 - AVERAE Sl]E)
EXTUlT T IM W M30lT COST T IN W M NilI COST K R KRSRI-RBi P(oi) 1011. DOSE WFSilE WF + GISITE GROSS IEI W/ GROSS W/IEI IlllTIAL % lillIIAL % IWACT IM'ACI IM*ACT IM'ACT ALIERETIVE VALIE REDICTION VAllE BEDICTIGi $ _$ ($ M R P-IUi)
~
2 2.2E-4 75% 3.7E3 62% 9.4E6 SE6 4100 2180 3 2.2E-4 10% 3.7E3 11% 0.50E6 -0.52E6 1370 10 COST 4 4.8E-4 61% 8.3E3 61% 7E6 -6.2 E6 1390 10 EOST 5 4.T-4 94% 8.3E3 91 1% 06E6 4fL6 8400 5330 6 5.7E-4 95% 9.9E3 94% 9tE6 716 10,140 7520 Ikite:
Alt. 2 G 3 Assanne ICli Alt. 4, S 4 6 Asstarc No I:Gil Al t. o Assantes + 20* for Cold Sinitikum
1
-O -
O O 0WERVIEW W ETERETIVES 31 GEIERIC 1% SIS - BR (0)ST K R P-IU1 - A K HAGE SilD EXIDif & IlfRMMNT COST F l&l0VDENT COST KR PERSGi-IU1 PIO1) P0lli. IDSE WFSIIE WF
- GISITE GOSS IET W/GOSS W/IET
!!1111AL I I!!!IIAL Z IM W T. lif%CT IN%CT IN%CT ALJENEllvE VAllt 101ET101 VAllE IGICTIG4 $ $ ($ PER P-M M) s
~
2 2.2E-4 541 2.3E4 185 1 13E6 9E6 1260 870
- 3 2.2E-4 41 2.3E4 ist 0 28E6 -0. I'56 300 10 COST
- 4 2.67E-4 30E 2.7E4 31% 1.1E6 2.7E6 120 to COST S 2.67E-4 84% 2.7E4 8881 RE6 UJE6 3400 3020 4
I 6 3.5E-4 841 3.E4 84% 84E6 7?i6 2090 2200 i
Not c:
) Alt. 21. 3 Asusne Gmt. Vent j Alt . 1. 5 f, 6 Assisne No Cont. Vent Alt . 6 Asumes + 2(l* l'or Cold Sinitikuu 1
i
)
1 .
PWR Core Melt Probability by Vulnerability -
Base case with Recovery Genera 11:ed Comparable Plant A Plant 3 Plant C Plant D Vulnerabilities Rob, Cent. Prob, cont, Prob. Cont. Prob. Cont, AWS h rbine Pump 1[-i 2% SE-6 2% 4E-7 1% 7t-6 4%
Diessi Cenerators 4E-7 1E-6 2% (( 1 3%
valve common Moda 1E-5 3% 25-5 11%
HP!S & RWST Valves 4t 6 24 6t-6 3%
) Recirculation Switchover 1[-1 7% 9t-7 1%
LP!/R System 1[-1 5% 4t-6 3% . }I-1 6%
f.P Pump Cooling 1[-1 4%
CCW System 1E-6 11 Service Water System u-4 1%
SIS & Manual Actuation ,
Seismic Cab & Racks M 16% 65-5 33%
Seismic RWST & CST 1[-1 3% W1 6% 11-1 17% 1[-1 0%
Spray Fire Header Rupt. It-5 25%
Fire AN Pump Room ((-1 4%
Fire a160 Sw Cear P:n 1[-1 6%
Fire Cable spreading Room St-5 3 r% 4t-5 59% 6t-6 3%
Wind Chimney Collapse It-5 10%
b*ind DC Exhaust Stock at-6 1% 5t-6 3%
Flood Safety Systems M M1 3E-0 4% 7E 6 4%
Lightning DC Power it-7 tt-7 2t-7 Unspecified Vulnerabilities 1t-5 4% 5E-5 19% 6E-6 4% 2E-5 10%
Total Core Melt Prob. 3.1E-4 2.4E-4 7.tt-5 1.8E-4 O
f-3
O O O FWR Internal and SPeclat EmerEency Core Melt Probabilities Fleet B Plant C ptsnt D InitiatinE Fleet &
Prob. EL of Total Prob. 1L of Total Prob. 1L of Total
- rob. 1L of Total Event 45 7.1E-5 32 1.4E-5 19 S.SE-5 49 Internal 1.4E-4 IS F.3E-6 3 1.3E-5 18 7.3E-5 41 seismic 6.1E-5 3.3E-5 10 F.5E-5 33 4. 4 E- 5 59 5.SE-6 3 Fire Internal F1oed F.JE-5 25 NA NA BBA 4.6E-5 20 3.2E-6 4 7.2E-6 4 External Flood 1.9E-S 11 . 1. 6 E-1 5.3E-6 3 Entreme Wind 4.Os-6 1 2.4E-5 2.0E-7 1.sE-7 LI EhtninE 5.8E-8 2.6E-6 1 .
TOTAL 3.13E-4 2.36E-4 7.48E-5 1.79E-4 Internal 1.4E-4 45 F.)E-5 32 1.4E-5 19 a.st-5 49.
Special EnerEency 1.JE-4 55 1.6E-4 68 6.0E-5 81 9.IE-5 51 u
l' i
i BWR Vulnerabilities and Proposed Modifications l
Plant t l Plant F Vulnerability Modification Vulnerabi g Modificatien ,
f Failure of 2 of 3 Add a 4th Diesel Failure of both Add a 3rd Diesel I Diesel Generators Generator Diesel Cenerators Generator l
Failure of both Add Deditated Failure of both Add Dedicated t Station satteries Battery to at least Station Satteries lattery to at least i and subsequent 1 DC or add 3rd and subsequent 1 DC or add 3rd !
failure to flash station Battery failure to flash Station Battery the DC fields the DC fields [
Failure of Diesel Additional DC Failure of R3 Add Bypass Line t Generator Jacket Cocling Water Pump Closed Cooling with NC Manual l Cooling or crosstle the Water System due Valve existing Cooling to FTO of 2 NC MOVs i Water Lines l Failure of 123 VDC Automatic transfer Flow Diversion Add a second isola.
Circuit Breaker of Loads or failure of R5 tion MQV that auto- l Control Power Battery Chargers closed Cooling natically closes in j Water system an accident "
Fire in Control gnhance procedures Flow Diversion Add Automatic Actu-O- Room or Cable for operating the failure of RS stor to another Spreading Room safe shutdown Pump Service Water Blocking MOV to systes due to prevent loss of 2 frilure of an Cooling Loops
!selation M0V Seismic Failure seismically gusti. Fires in Cable Add a one-hour Fire of Station fy and upgrade tryansion Room Barrier around HPCI
- latteries sattery Racks and & R33W Power Cables Supports l
Seismit failure of Add Seismic tuternal flood Develop Procedures A160 VAC Buses postraints at top for Safe Shutdown of Sua Cabinets for very high flood crests Seismic Event Add or strengthen supports & braces to several Lepor-tant components O
a-c
s ,
l' i l
- l rwn Internal and special Emergency i
- Core Melt Probabilities
, .I i
l I niti,a ting Plant ! Plant P __,
i Event Probability % of Total ProbabilitV % of Total 4
j Internal 9.9E-5 50 2.9E-4 66 Seisnie 4.3E-5 42 8.1E-5 it i 1 F Fire 1.32-5 7 1.12-5 3 t Internal Flood NA ,
NA .
j External Flood 9.4E-8 5.0E-5 12 Extreme Wind 1.4E-7 3.4E.6 1
) Lightning 1.7E-6 1 1.82-6 1 I
! TOTAL 2.0E-4 4.4E-4 !
l !
t I Internal 9.9E-5 50 2.9E-4 66 !
I i I Special !
Energency 1.0E-4 50 1.5E-4 34 .!
i l
, I 4
l 1
f I !
l !
l 4
r 3 ;
1 n -i
l A
o o l 4 l
)
l l ISSUE a
) OPERATION DURING DECAY HEAT REMOVAL SYSTEM COOLING IS A SIGNIFICANT CONTRIBUTOR TO THE LIKELlHOOD OF A RELEASE DUE TO A CORE DAMAGE ACCIDENT 1
i l
.i, e
8- ~
~O O O
)
l
)
A. EXPERIENCE ACCIDENTS CONTINUE TO INITIATE AT AN UNACCEPTABLY HIGH RATE. TWO REPORTED IN MAY. NUMEROUS PUBLICATIONS AND M Et- I IN GS HAVE NOT LED TO SOLUTION.
. O O O 1
B. PHENOMENA i
PHENOMENA IDENTIFIED WHICH POTENTIALLY LEAD TO SEVERE CORE DAMAGE IN SHORTER i TIME THAN PREVIOUSLY BELIEVED. OTHER
- "NEW" PHENOMENA AFFECT THE REACTOR COOLANT SYSTEM (RCS), DECAY HEAT REMOVAL (DHR) SYSTEM, INSTRUMENTATION AND OTHER l EQUIPM ENT.
i l
O O O
!I l
D. GL 87-12 i,
- a. NO RESPONSES FULLY SATISFACTORY.
! SOME LICENSEES UNSATISFACTORY IN EVERY ONE OF 12 CATEGORIES EVALUATED. SERIOUS LACK OF
]
l UNDERSTANDING AND INADEQUATE l PREPARATION FOR OPERATION IDENTIFIED.
SOME LICENSEES NOT TAKING CORRECTIVE i
1 ACTION OF ANY KIND.
I i ,
l
l O O O 9
l D. GL 87 12 cont l
1 l' b. INDIVIDUAL LICENSEES HAVE SHOWN EXCELLENT INSIGHT INTO SELECTED 1
l AREAS SUCH AS RCS DRAINING, -
1 1
I CONTAINMENT CLOSURE, i
INSTRU M ENTATIO N, DHR SYSTEM l O P ERATIO N, OTHERS. IN FORMATION IS l' NOT EFFECTIVEL~Y SHARED.
!l
!i
O O O E. SERIOUS DEFICIENCIES l 1. EVENTS CONTINUE TO INITIATE WHICH l HAVE POTENTIAL TO BECOME SERIOUS l 2. MITIGATION PLANNING TO PREVENT CORE DAMAGE IS O t- 1 EN POOR i 3. PLANNING TO PREVENT A RELEASE I
SHOULD OORE DAMAGE OCCUR IS OFTEN
- NONEXISTENT t
I
- 4. ANALYSES OFTEN NON EXISTENT. PLANTS l OPERATED IN UNANALYZED AREAS WHERE IMPLICATIONS NOT UNDERSTOOD.
l
! 5. MANY MORE ....
es- -- - - - - n-- + u w -L.-
O O O I
F. REDUCED RCS INVENTORY OPERATION
- 1. LOSS OF DHR A FREQUENT OCCURRENCE
- 2. ONE THIRD (40) HAVE OCCURRED DURING REDUCED INVENTORY OPERATION l
l i
}
i-L __ . - - - _ _ . _ - _ . . . _- - -
O -
O O
.j A. APPROACH
/
- 1. SOME REDUCTION IN CORE DAMAGE
) LIKELlHOOD -
"EXPEDITIOUS ACTIONS" l ACCOM PLISH IM M EDIATE, EFFECTIVE REDUCTION IN LIKELlHOOD OF RELEASE
]
! IF CORE DAMAGE ACCIDENT OCCURS I
l 2. SIMULTANEOUSLY INITIATE ACTIONS 4
l WHICH TAKE A LONGER TIME TO DEVELOP l
"PROGRAMMED EN HAN CEM ENTS" i
i
~
l .
! ENHANCEMENTS BECOME AVAILABLE l
\ - - .* !! _ - - -
O O e B. EXPEDITIOUS ACTIONS
- 1. CONTAINMENT CLOSURE
- 2. RCS TEMPERATURE
- 3. RCS LEVEL
- 4. DO NOT PERTURB RCS
- 5. BACKUP EQUIPMENT
- 6. HOT AND COLD LEG CLOSURE
- 7. DISCUSS DIABI CANYON IMPLICATIONS WITH OPERATIONS STAFF C. PROGRAMMED ENHANCEMENTS I
- 1. INSTRUMENTATION
- 2. PROCEDURES
- 3. EQUIPMENT f
j 4. ANALYSES l 5. TECHNICAL SPECIFICAT!ONS
E M T R E O O T H - C S
Y N
~' s S
('~. =_
T _
N _
A _
i.
L _
O O
C 4 R
O T
C A
E R
ON TO AI LT
/ R OL TE CS UC R AS R I ME UN CN CO D
L T E W
RT ON P TA CL M EE RV 1
C k EO RC
/\
l R (
U R G O \
T
,)c I M A B A
F E R 7
)f E
T N S E J \
C F
y O
/ l E
N iZ i 7 P"
I L PA N_ -
2 2
RS E
C TG OE HL pa
~
O O O R-R _0SS BE -
A V :: O R BOILING INITIATES l- PRESSURIZATION INITIATES w
168 ,
,/
- - COLD SIDE SG MANWAY OPEN m
I i /
' ' - A MID-LEG ELEVATION Z I -_____
' ' ' ' OF HOT LEG O 106- ! ! - ,- ._. X_._. BOTTOM 6- ! !
< l I
y l 164 - j w j j TOP OF CORE x
.A_._._._._._..______. . _ . _ . _ _ _ _ _ _ . _ . _ _ _ _ _ _ _ . _ .
~~
6- ! !
Z l I
< 102 - - -
1 I
_1 . .
Q_ l I v i i RCP OPEN FOR REPAIR
' ~
100 -
L1J l I y .
HOT SIDE SG MANWAY OPEN w ! !
J 98 - ! !
I I O- .
W ! ! MID-CORE ELEVATION
- .___.L_i_. ._ 1 .___._____________._ .
~
96 I I I I r 2ee e se tee 1se TIME, MINUTES gS
INDUSIRY RESPGNSE TO USI A-45 DECAY HEAT REMOVAL PRESENTATION TO THE ADVISORY COMMITTEE ,
ON REACTOR SAFEGUARDS DHR SUBCOMMITTEE JULY 27, 1988 l
i BY O
NUMARC EPRI WISCONSIN ELECTRIC POWER C0.
WESTINGHOUSE OWNERS GROUP WESTINGHOUSE ELECTRIC CORP. l l
1
I O
INDUSTRY SPONSORS / PARTICIPANTS NUMARC GERRY NEILS, NSP (W. G. CHAIRMAN)
EPRI JACK HAUGH GARY VINE SAIC BILL PARKINSON WEST. DON PADDLEFORD !
WEP R0GER NEWTON HARV HANNEMAN 1 l
STAN GU0KAS WOG WARREN ANDREWS ,
i E-_. . ___ __ ___ _ ___ _ _ - _ _ _ _ _ . _ _ _ _ _
S . UNITED ST ATEs
~% NUCLEAR REGULATORY COMMISSION g
.[' Tv ASHINGTON. D C. 2055%
y l
~ '
,/ February 10, 1986 k.,a...
MEMORANDUM FOR:
C. J. Heltemes, Jr. , Director Office of Analysis and Evaluation of Operational Data FROM:
Harold R. Denton, Director Office of Nuclear Reactor Regulation
SUBJECT:
AEOD'S REPORT ON DECAY HEAT REMOVAL PROBLEMS AT U.S. PWR The Office of Nuclear Reactor Regulation has reviewed the revised version of AEOD's, "Case Study Report - Decay Heat Removal Problems at U.S. Pressurized Water Reactors," transmitted in your memorandum of Decemoer 23, 1985. This report contains a number of findings and conclusions, and presents six requirements for consideration by NRR.
As mentioned in your memorandum, peer review comments previously sent to you have been resolved. The infomation in the report contains many useful and constructive coments. We do not believe that your recommendations are A-45 directly applicable to the resolution of USI A-45, as you suggest.
is directed toward determining the need for improvements in current decay heat removal systems. In particular, it is addressing the need for a dedicated decay heat removal function. One alternative resolution of A-45 is to identify and correct the vulnerabilities in current designs, such as those discussed in your report. However, until A-45 is complete, specific O generic vulnerabilities to decay heat removal are being pursued separately.
For example, improvements in auxiliary feedsater system reliability are being proposed through Generic Issue 124.
Therefore, we believe that it would be appropriate to use the information in the report in the resolution of Generic Issue Nc. 99, "RCS/RHR Suction Line Interlocks." This issue is specifically concerned with loss of the RHR system While this issue was originally limited to during cold shutdown or refueling.
the loss of RHR due to inadvertent closure of :he RHR suction valves, the scope will be. expanded to include consideration of all of the failure modes discussed in your report.
Tracking of the resolution of this issue and therefore your six recomendations, wjll be done in the NRR Generic issues Management Contro,1 Systen.
l n Harold R. Denton, Director Office of Nuclear Reactor Regulation
Enclosure:
AE00 Report on Decay Heat Removal Problems at U.S. PWRs
Contact:
H 8. Holz, NRR/DSR0
~
r A-45 IS DIRECTED TOWARD DETERMINING THE NEED FOR IMPROVEMENTS IN CURRENT DECAY HEAT REMOVAL SYSTEMS. IN PARTICULAR, iT IS ADDRESSING THE NEED FOR A DEDICATED DECAY HEAT REMOVAL FUNCTION.
O O O~ -
A-45 CASE STUDIES BY SANDIA ALL ADDRESS WHETHER ADD-ON, DEDICATED DHR SYSTEM ARE JUSTIFIED.
ALL CONCLUDE THE ADD-ON DHR SYSTEMS ARE NOT JUSTIFIED.
ALL HAVE WIDE MARGINS IN THESE CONCLUSIONS.
ALL CASE STUDIES WERE MODELLED VERY CONSERVATIVELY, SO THAT THE CONCLUSIONS l ARE UNIMPEACHABLE.
NSAC 113 WAS REWORK OF POINT BEACH TO l ILLUSTRATE CONSERVATISM IN THE CASE
! STUDY.
O O O THE CASE STUDIES .
WERE "A MODIFIED IPE".
ILLUSTRATE THAT POTENTIAL DHR WEAKNESSES ARE PLANT SPECIFIC, NOT GENERIC.
ESSENTIALLY EVERY *OTHER" INFLUENCE ON DHR RELIABILITY IS COVERED BY "OTHER" NRC PROGRAMS
- APPENDIX R
- A-44 SBO
- SEISMICITY PROGRAMS, ETC.
VALUE OF IPE IS PLANT UNIQUE IDENTIFl-CATION OF LOW COST /HIGH VALUE POTENTIAL IMPROVEMENTS.
1 NUMARC WORXING GROUP DN
. DECAY HEAT REMOVAL CHAIRMAN: GERRY NEILS C,-
NSSS DESIGNS REPRESENTED CASE STUDY DHRTSG NAME, COMPANY h GE CE B&W PLANTS HEMBER GERRY NEILS, NSP X X CHAIRMAN JEFF JEFFRIES, CP8L X X (NSAC T.F. CH.)
ROGER NEWTON, WEP X POINT BEACH (WOG CH.)
DAVE HELWIG OR X GEORGE BECK, PECo (BWROG DHR CH.)
OALANLADIEU, YANKEE X X
MIXE MEISNER, LP8L: X X X TURKEY POINT /
DON JAMES OR MIXE SCH0PPHAN, FP8L ST. LUCIE (CEOG REP)
LARRY TAYLOR OR TED ENDS, APat X X ANO-1 GREGG SWINDLEHURST, DUKE X X XAVIER POLANSKl, COMMED X X QUAD ClT1ES DON REEVES, NPPD X COOPER X
OGARY VINE, EPRI i (NSAC STAFF SUPPORT)
~
BRIEF HISTORY OF NSAC-113 0 . DEC 1985 DHRTSG MEETING; REVIEW 0F PB & QC DRAFTS FEB 1986 COMMENT LETTERS TO SANDIA FROM EPRI. AIF MAR 1986 DHRTSG MEETING. REVIEW 0F PB/DC COMMENTS; DISCUSSION OF TP, COOPER. NRC REQUEST FOR INDUSTRY ANALYSIS JUN 1986 EPRI INITIATED REANALYSIS OF POINT BEACH.
SUPPORTED BY WOG AND WEP OCT 1986 FIRST MEETING OF NUMARC WG; ENDORSED PB REANALYSIS EFFORT FEB-APR 1987 ALL SIX CASE STUDIES DISTRIBUTED FOR FINAL REVIEW O -
MAY 1987 NUMARC (C0UNCil.) INCORPORATED AND CHARTERED JUN 1987 EPRl/NUMARC DHR WORKSHOP. NEW ORLEANS.
NUMARC REVIEW COMMENTS ON ALL SIX CASE STUDIES. PB REANALYSIS RESULTS DISCUSSED.
JUL 1987 NRC (B. SHERON) LETTER TO NUMARC (G. NEILS)
REQUESTING MORE INFORMATION ON PB REANALYSIS:
SUGGESTED MEETING OCT 1987 NSAC-ll3 (DRAFT) FORWARDED TO NRC BY G. NEILS NOV 1987 A-45 PRESENTATION TO ACRS DHR S.C.; NSAC-ll3 DELIVERED TO ACRS JAN 1988 A-45 PRESENTATION TO ACRS DHR S.C. ON NSAC-113 O GLV:3867NS8
e
@ * **
- eq#e
~. g UNITE D 8 TATE S
/ NUCLEAR REGUL ATORY COMMISSION g pr A94t pdGTON. D. C. 706s 5 JUL 2 0 W
'd \ e . . . . .
Dr. Corald Neils, Chairman NtFJutC Working Group on DHR Northern States Power Company -
41$ Nicollet Hall Minneapolis, Minnesota 55401
Dear Dr. Neils:
I received a copy of your June 22, 1987 comunication to Dr. David Ericson on the subject of our USI A-45 Program on Shutdown Dacay Heat Removal Requirements. Sandia National Laboratories (SNL) is in the process of stuctying.. ,
these co ments in detail; however, we are particularly interested to your indication that a separate PRA of one of the $NL Case Studies was sponsored by EPRI, the Westinghouse Owners Group, and Wisconsin Electric Power and concludes that core melt risk is about ten times lower than the SNL Case Study for the same plant. Since we perceive that this industry-sponsored study represents your quantification of the dif fering views outlined in your June 22nd letter, it would assist us in our deliberations to better understand your technical basis for these differences. Therefore, we ask that you identify the major items in your recent PRA study which contribute to the factor of ten difference in core melt probability, and present the technical basis for the value(s) selected (e.g., referenceable operating experience data base, human factor studies, corrponent reliability data, external event initiating frequencies, etc).
We also wish to acknowledge the creation of the NUMARC Working Group to study the DHR issue, and we look forward to interfacing with members in the near future. Since our draft Regulatory Analysis on USI A-45 is still pre-decisional, perhaps we can consider a first meeting to focus on our review of the items you will identify as key contributors to the factor of ten difference in the core seit probability between the two PRAs. We appreciate the technical attention that NUMARC hat apparently devoted to review of the siX Case Studies, as indicated by Enclosure 1 of your June 22nd letter, ond we intend to work with SNL to consider your coments.
Sin rely,
- 44* D.) * ~_
Brian W. Sheron, Director Division of Reactor and Plant $ystems of fiee of Nuclear Regulatory Research cc: D. Ericson l O
l
- Y 9 I
- Northem States Power Company
- 414 Neonei Man Minnespoks, Konesota $5401 lenephone (612) 330-55oo U
October 28, 1987 Dr. Brian W. Sheron Directot Division of Reactor & Plant Systems Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Coc=ission Washington, D.C. 20555
Dear Dr. Sheron:
In response to your letter dated July 20, 1987, I am pleased to forward a draft copy of a docu=ent entitled "EPRI/WOG Analysis of Decay Heat Removal
. Risk at Point Beach." This study, sponsored by EPRI and the Westinghouse Ovners Group, was prepared by Science Applications In:errational Corporation and Westinghouse Electric Corporation with the assistance of Wisconsin Electric Power Cocpany, the owners and operators of Point Beach. The NWARC Vorking Group on DHR has followed and endorsed this effort.
The primary purposes of this study were to provide a best-estimate analysis of DHR risk at a selected USI A-45 Case Study plant and to quantify the bs) differences discussed in our June 22 com ent letter on the Case Studies.
" The results of this Point Beach reanalysis, as they nov stand, indicate an approxicate factor of thirty reduction in core-melt frequency for the sequences included in the scope of the NRC study; an approximate factor of seven reduction in the offsite consequences of these sequences, over and above the core-melt frequency reduction; and an approximate 50-400% increase in the estimated cost of the various backfit proposals evaluated in the NRC study. The EPRI/WOG findings indicate that the core-melt frequency estimate for Point Beach (1.0 x 10*3 per reactor year) is a f actor of ten lover than the core-melt frequency target in the NRC's Safety Goal. The EPRI/WOG study, like the NRC study, also concludes with a very high degree of confidence that an add-on, dedicated SDKR system would not be cost-beneficial for Point Beach.
We would be pleased to meet with you and se=bers of your staff, as suggested in your letter, to discuss the methodologies, technical bases and findings contained in both studies. We have provided you with this draft report prior to publication to allow suf ficient time for your staff to familiarize themselves with the EPRI/ WOC reanalysis in advance the meeting. In antici-pation of that meeting, EPRI and WOC are continuing to double-check the models used in their analysis against the final, as published, numerical values used in the NRC study. Although some small changes in the comparative estitastes of core melt frequency could occur in so=e instances, we expect the O
V
Dr. Bricn Shsrcn g
Octcbor 28, 1987
. Page 2 of 2 l
overall results and conclusions of the EPR1/WOG study to remain essentially unchanged. In the meanwhile, we would be pleased to schedule a meeting for I
C'/]
~
the first autually convenient opportunity.
Sincerely,
- 2 C H Neils Chairman
- NUMARC Working Group on DHR CHN/vf cet B. Lee, NUMARC NUMARC Working Group Members T. Speis. NRC K. Kneil NRC A. Marchese, NRC D. Ericson, Sandia National Laboratories R. Newton, Wisconsin Electric Power Co.
W. J. Parkinson, SAIC D. F. Paddleford, Westinghouse Electric Corp.
A. Ladieu. Chairman WOG Analysis Subcom=ittee J. Taylor EPRI ,
W. Lay =an, EPAI
() 7. Marston, EPRI G. Vine, EPRI J. Haugh, EPRI :
O
., .-,-..,--,,--,-...n- - - - - - - . - - - - - - . , - -,, . - , - _ - , , _ . , , . , , - - - , - . , , , ,
O O O TAP A-45 SHUTDOWN DECAY HEAT REMOVAL ANALYSIS q
REVISED ESTIMATES FOR POINT BEACH I
1, T
i l David M. Ericson, Jr.
1 ERC INTERNATIONAL
! for l SANDIA NATIONAL LABORATORIES and U. S. NUCLEAR REGULATORY COMMISSION I
ACRS SUBCOMMITTEE ON DECAY HEAT REMOVAL July 27,1988 i
l a
l O O O l
i i
i 1
SUMMARY
OF RESULTS i
i i
I i
P(CM.) /RX-YR SNUNRC ORIGINAL ESTIMATE 3.0 E-04 i
EPRl/WOG REANALYSIS 1.0 E-05 SNUNRC REVISION 8.9 E-05 t
~ ~
O O O FACTORS BEARING ON REVISIONS SMALL FRACTION OF CHANGE (<20%) DUE TO PLANT CHANGES MOST OF CHANGE RELATED TO:
METHODS ASSUMPTIONS DATA EXAMINED ONLY IN CONTEXT OF POINT BEACH
O O O COMPARISON OF SEQUENCES
\
i P(CM) PER RX-YEAR SEQUENCE ORIGINAL EPRI/WOG REVISED l S2MH1H2 4.7 E-05 5.8 E-07 7.0 E-06
) S2MD1D2 8.7 E-06 9.5 E-08 9.2 E-07 S2MXD1 5.7 E-07 1.0 E-08 1.0 E-08 I
RATIONALE-ACCEPTED 3 E-03/YR FREQUENCY FOR S2 l AGREED WITH REMOVAL OF CCW DEPENDENCY l
l DISAGREED WITH 1 E-04/ DEMAND OPERATOR ACTION l FAILURE, RETAINED 1 E-03/ DEMAND l
1 1
i
~ ~
O O O COMPARISON OF SEQUENCES (CONT!NUED)
P(CM) PER RX-YEAR SEQllENCE ORIGINAL EPRI!WOG REVISED T1MLE 6.7 E-06 7.7 E-07 7.7 E-07 RATIONALE-ACCEPTED PLANT SPECIFIC Ti FREQUENCY TENTATIVELY ACCEPTED CREDIT FOR NEW BATTERIES, OUANTITATIVE VERIFICATION REQUIRED E
i
O O O COMPARISON OF SEQUENCES (CONTINUED)
P(CM) PER RX-YEAR SEQUEN_Cf ORIGINAL. EPRl/WOG REVISED T3OH1H2 2.5 E-05 0 3.6 E-06 T3OD1D2 4.6 E-06 0 1.8 E-07 RATIONALE-ADOPTED REVISED VALUE FOR O (0.01 VS 0.07), DID NOT AGREE O = 0.
AGREED V'!TH REMOVAL OF CCW DEPENDENCIES
O O O COMPARISON OF SEQUENCES (CONTINUED)
P(CM) PER RX-YEAR SEQUEN_CX DRIGINAl, EPRl/WOG REVISED l T2MOH1H2 3.5 E-06 1.9 E-07 5.0 E-07 T2 MOD 1D2 6.6 E-07 4.1 E-08 4.1 E-08 RATIONALE:
DID NOT CHANGE T2 FREQUENCY (1.0 VS 0.92)
RETAINED COMSERVATIVE ASSUMPTION PORVs BLOCKED DUE TO DATA VARIABILITY ADOPTED REVISED VALUE FOR O (0.01 VS 0.07),
DO NOT AGREE THAT O = 0.
AGREED WITH REMOVAL OF CCW DEPENDENCIES
O O O COMPARISON OF SEQUENCES (CONTINUED)
P(CM) PER RX-YEAR SEQUENCY .OMIGINAL EPRI/WOG REVISED T2MLE 6.6 E-07 1.0 E-07 6.6 E-07 RATIONALE:
DID NOT CHANGd T2 FREQUL.:CY (1.0 VS 0.92)
NOT CLEAR HOW MFW RF.COVERY DIFFERS, ADDITIONAL INFORMATION REQUIRED l
O O O COMPARISON OF SEQUENCES (CONTINUED) l P(CM) PER RX-YEAR SEQUENCE ORIGINAL EPRl/WOG REVISED T5MLE 9.1 E-07 1.3 E-08 9.1 E-07 T4MLE 6.2 E-07 0 6.2 E-07 T2MLH 2.0 E-08 1.0 E-07 2.0 E-08 T10D1D2 1.0 E-08 1.0 E-07 1.0 E-08 1
i RATIONALE:
1
) OPERATOR ACTIONS (RECOVERY) ASSUMED BY EPRl/WOG NOT ADEQUATELY JUSTIFIED
} RETAIN A-45 CONSERVATISM, LOSS OF BUS LEADS TO TRIP i
1 l
1
~ .
O O O COMPARISON OF SEQUENCES (CONTINUED)
P(CM) PER RX-YEAR i SEQI).ENCE ORIGINAL EPRI/WOG REVISED 1
LTSB 3.6 E-05 5.4 E-07 9.9 E-06 1
l t
, RATIONALE:
! ACCEPTED LOWER Ti FREQUENCY AND LOWER DIESEL GENERATOR FAULT VALUES DID NOT CREDIT CST REFILL AND LONG TERM OPERATOR ACTION
. ASSUMPT;GNS REGARDING RECOVERY OF OFFSITE POWER COULD REDUCE ESTIMATES l
I l
O O O l COMPARISON OF SEQUENCES (CONTINUED)
I
! P(CM) PER RX-YEAR 1
i SEQUENCE ORIGINAL EPRF/WOG REVISED
{ SEISMIC 6.1 E-05 7.4 E-06 4.1 E-05 i
l l
- RATIONALE:
i !
AGREED WITH CREDIT FOR ADDED BATTERIES i
i j DISAGREED WITH EPRl/WOG RECOVERY POST-OUAKE l DISAGREED GN SURVIVABILITY OF RWST f
I
) '
i O O O
- COMPARISON OF SEQUENCES (CONTINUED)
)
P(CM) PEE RX-YEAR l SEQiLENOE ORIGINAL EPFil/WOG REVISED i FIRE 3.2 E-05 6.3 E-08 2.2 E-05
) ,
i l RATIONALE:
] AGREED WITN CREDIT FOR SECOND HALON SYSTEM IN AFW ROOM j r
D;SAGREED WITH LOWER HALON FAILURE RATE l'
I i j l
.l !
O O O COMPARISON OF SEQUENCES (CONTINUED)
P(CM) PER RX-YEAR GE_QLLEN_Cf ORIGINAL. EPRf/WOG REVISED WIND 4.0 E-06 1.0 E-08 1.7 E-67 HATIONALE:
r AGREED WITH CHANGES RELATED TO STACK MODIFICATION i
C.-_ 1 e _ _ _ -. . --- m - _2 - - 4 e - _ -%n* _ a -_ -.e .
9
- O O O COMPARISON OF SEQUENCES (CONTINUED) s P(CM) PER RX-YEAR SEQUENCE ORIGINAL EPRI/WOG REVISED INTERNAL 7.7 E-05 1.0 E-08 9.8 E-07 FLOOD RATIONALE:
AGREED WITH USE OF THOMAS CORRELATION USED PIPE LENGTHS PRESENT AND IN LINE-OF-SIGHT REVISED INITIATING EVENT FREQUENCY CONSISTENT WITH OTHER REPORTS ,
i i ,
.i
_ _ --- --,,v- - ,- , , _ _ --,,-,- -. - . - - , , , . - - - - - . _ , .
j O O O i
SUMMARY
i
- P(CM) PER RX-YEAR ,
l SEQUENCE ORIGINAL EPRI/WOG REVISED .
) INTERNAL 1.3 E-04 2.5 E-06 2.5 E-05 EVENTS EXTERNAL 1.7 E-04 7.5 E-06 6.4 E-05
! EVENTS !
4 TOTAL 3.0 E-04 1.0 E-05 8.9 E-05 ,
l l
! l
.i I
l 1
l i
l i
i t i
O O O l
I l
! DOMINANT SEQUENCE DEFINITIONS 1
j S2MH1H2 - Small-break LOCA with subsequent loss of main feedwater and failura of emergency core cooling in recirculation.
! S2MD1D2 - Small-break LOCA with loss of main feedwater and failure of emergency core j cooling in the injection mode.
]
l S2MXD1 -
Small-break LOCA with failure of emergency core cooling in the injection mode and I failure to achieve secondary blowdown.
4 i
i
_ - _ _ - - _ _ w--e w - - - - --*--- --=------------e- -- - - - --+ ,,-----w.--*n -- ---- -- - -- , -
- , - - y e m - m ---- w-
! O O O I
i l
l l
- DOMINANT SEQUENCE DEFINITIONS (CONTINUED)
T1MLE - Loss-of-offsi'e-power transient with failure of auxiliary feedwater and feed and
, bleed.
J l T2MLE - Loss-of 'eedwater transient with failure of auxiliary feedwater and feed and bleed.
! T5MLE - Loss-of-DC-bus transient with failure of auxiliary feedwater and feed and bleed.
l i
j T4MLE - Lossef-AC-bus transient with failure of auxiliary feedwater and feed and bleed.
T2MLH1 - Loss-of-feedwater transient with failure of auxiliary feedwater and failure of
) emergency core cooling in the recirculation mode.
l i
I i
1
-.----------,----_.----n-~r------ -, ,, - . - - - - . - - - . - - - , , - - - - - , - - - - - ~ - - . , -
O O O DOMINANT SEQUENCE DEFINITIONS (CONTINUED)
T3OH1H2 - A transient followed by a stuck-open relief valve (transient-induced LOCA) and failure of emergency core cooling in the recirculation mode.
T3OD1D2 - A transient followed by a stuck-open relief valve (transient-induced LOCA) and failure of emergency core cooling in the injection mode.
T1QD1D2 - Loss-ofeffsite-power transient followed by a stuck-open relief valve (transient-induced LOCA) and failure of emergency core cooling in the injection mode.
~
O O O t
l I
- DOMINANT SEQUENCE DEFINITIONS (CONTINUED) i I
T2MOH1H2 -Loss-of-feedwater transient followed by a stuck-open relief valve (transient-induced j LOCA) and failure of emergency core cooling in the recirculation mode.
I i T2 MOD 1D2 -Loss-of-feedwater transient followed by a stuck-open relief valve (transient-induced j LOCA) and failure of emergency core cooling in the injection mode.
i l'
~
O O O 1
DOMINANT SEQUENCE DEFINITIONS (CONTINUED) j LTSB - Long-term station blackout caused by a loss-of-offsite-power transient and failure j to recover offsite power with subsequent failure of diesel generators.
I 1
I i
4 l
. - , . . , , . _ - _ -_ . _ - - - . - - - , . _ - - , - _ . - _ _ _ _ , _ _ - - . , . _ - - . . ,,. _ ,. --. <~- ,.__ ,_ _ _, . ,_ ---. .. __.._ _ ,. _-._,, _ -_-_,,.._.-_ ,_ __.
t 1
!O INDUSTRY RESPONSE TO USI A-45 DECAY HEAT REMOVAL 4
j PRESENTATION TO THE ADVISORY COMMITTEE i ON REACTOR SAFEGUARDS i DHR SUBCOMMITTEE j JULY 27, 1988
\
!O BY i NUMARC
! EPRI I WISCONSIN ELECTRIC POWER C0.
l WESTINGHOUSE OWNERS GROUP l WESTINGHOUSE ELECTRIC CORP.
l
!O i
l
[
)
O l
- INDUSTRY SPONSORS / PARTICIPANTS :
i f l
(W. G. CHAIRMAN) l J
1 >
i EPRI JACK HAUGH I
, GARY VINE
. SAIC BILL PARKINSON
)O l
WEST. DON PADDLEFORD !
)
WEP ROGER NEWTON
! HARV HANNEMAN STAN GUOKAS !
! WOG WARREN ANDREWS l 1 i l t i !
! I 1
0 l l
. UNITED STATES 9 p',m j NUCLE AR REGUL ATORY COMMISSION usmNotos. o c. osss g l February 10, 1956
(},..../
v MEMORANDLM FOP: C. J. Helterres Jr., Director Office of Analysis and Evaivation of Operation 41 Data Harold R. Denton, Director FROM:
Office of Nuclear Reactor Regulation
SUBJECT:
AECD'S REFORT ON DECAY HEAT REMOVAL PROELEMS AT U.S. FWR5 The Office of Nuclear Reactor Regulatien has reviewed the revised version of AE00's, "Case Study Report - Decay Heat Renoval Problems at U.S. Pressuriced Water Reacters " transmitted in your memorandur of Decem:er 23,19S5. This recort contains a nur.ber of findings and conclusions, anc presents six requirerents for consideration by NRR.
As rnentiened in your recorandum, peer review cc-rents previously sent to you have beer, resolved. The inferration in the re; ort contains many useful and constructive corn nts. We do not believe that your recem. endations are A-45 directly applicab a to the resolution of USI A 45, as you suggest.
is directed toward deternining the need for ircreverents in current decay heat rencval systems. In particular, it is actressing the need for a dedicated decay heat reroval function. One alternative resolution of A 45 is to identify and correct the vulnerabilities in current design}, such as O those discussed in your report. However, until A-45 is complete, specific O generic vulnerabilities to decay heat reno.al are being pursued separately.
For exar le, improve ents in auxiliary feedanter system reliability are being proposed through Generic Issue 124 Therefore, we believe that it would be appropriate to use the information in the report in the resolutien of Generic Issue N:. 99, "RCS/RhR Suction Line Interlocks." This issue is specifically conce-ed with loss of the RHR syste :
during cold shutdown or refueling. While this issue was originally limited to the loss of RhR. cue to inadvertent closure of the RHR suction valves, the secte will be. expanded to include consideraticm of all of the failure modes discussec in your report.
Tracking of the resolution of this issue and therefore your six reco rendations, will be dere in the NRR Generic Issues Manage ent Contro1 System.
l k Harold R. Centon, Director Office of Suclear Reactor Regulation
Enclosure:
AEOD Fer:rt on Decay Heat O Rescval Proble-s at U.S. FWRs O
Contact:
H. S. Hol:, h::/CSF0
O O .
O~
l A-45 IS DIRECTED TOWARD DETERMINING i THE NEED FOR IMPROVEMENTS IN CURRENT DECAY HEAT REMOVAL SYSTEMS. IN ;
i PARTICULAR, IT IS ADDRESSING THE NEED FOR A DEDICATED DECAY HEAT REMOVAL FUNCTION.
4 O O O
A-45 CASE STUDIES BY SANDIA ALL ADDRESS WHETHER ADD-ON, DEDICATED DHR SYSTEM ARE JUSTIFIED.
ALL CONCLUDE THE ADD-ON DHR SYSTEMS ARE NOT JUSTlFIED.
ALL HAVE WIDE MARGINS IN THESE CONCLUSIONS.
ALL CASE STUDIES WERE MODELLED VERY CONSERVATIVELY, SO THAT THE CONCLUSIONS ARE UNIMPEACHABLE.
NSAC 113 WAS REWORK OF POINT BEACH TO lLLUSTRATE CONSERVATISM IN THE CASE STUDY.
I
O O .
O 1
! THE CASE STUDIES ,
s l -
WERE "A MODIFIED IPE".
l -
lLLUSTRATE THAT POTENTIAL DHR WEAKNESSES .
ARE PLANT SPECIFIC, NOT GENERIC.
ESSENTIALLY EVERY "OTHER" INFLUENCE ON
! DHR RELIABILITY IS COVERED BY "OTHER" l NRC PROGRAMS
- APPENDIX R
- A-44 SBO i
- SEISMICITY PROGRAMS, ETC.
l I
VALUE OF IPE IS PLANT UNIQUE IDENTIFI-I CATION OF LOW COST /HIGH VALUE POTENTIAL IMPROVEMENTS. ,
NUMARC WORKING GROUP DN DECAY HEAT REMOVAL
(] CHAIRMAN: GERRY NEILS NSSS DESIGNS REPRESENTE_D CASE STUDY DHRTSG NAME, COMPANY W E CE B&W PLANTS MEMBER GERRY NEILS NSP X X CHAIRMAN JEFF JEFFRIES, CP8L X X (NSAC T.F. CH.)
ROGER NEWTON, WEP X POINT BEACH (WOG CH.)
DAVE HELWIG OR X GEORGE BECK, PECo (BWROG DHR CH.) .
l ALAN LADIEU, YANKEE X X
MIKE MEISNER, LP&L: X :
00N JAMES OR X X TURKEY POINT /
. MIKE SCH0PPMAN. FP&L ST. LUCIE i (CEOG REP) i l
LARRY TAYLOR OR TED ENDS, AP&L X X- AND-1 GREGG SWINDLEHURST, DUKE X X XAVIER POLANSKI, COMMED X X QUAD CITIES DON REEVES, NPPD X COOPER ;
X
, GARY VINE. EPRI (NSAC STAFF SUPPORT)
. BRIEF. HISTORY OF NSAC-113 O -
DEC 1985 DHRTSG MEETING: REVIEW 0F PB & QC DRAFTS FEB 1986 COMMENT LETTERS TO SANDIA FROM EPRI, AIF :
MAR 1986 DHRTSG MEETING, REVIEW 0F PB/0C COMMENTS; DISCUSSION OF TP, COOPER. NRC REQUEST FOR INDUSTRY ANALYSIS ,
j JUN 1986 EPRI INITIATED REANALYSIS OF POINT BEACH.
SUPPORTED BY WOG AND WEP OCT 1986 FIRST MEETING 0F NUMARC WG: ENDORSED PB REANALYSIS EFFORT FEB-APR 1987 ALL SIX CASE STUDIES DISTRIBUTED FOR FINAL f REVIEW O -
MAY 1987 NUMARC (COUNCIL) INCORPORATED AND CHARTERED ,
JUN 1987 EPRl/NUMARC DHR WORKSHOP. NEW ORLEANS. !
NUMARC REVIEW COMMENTS ON ALL SIX CASE !
STUDIES. PB REANALYSIS RESULTS DISCUSSED. !
JUL 1987 NRC (B. SHERON) LETTER TO NUMARC (G. NEILS) !
RE0 VESTING MORE INFORMATION ON PB REANALYSIS:
SVGGCSTED MEETING r
DCT 1987 NSAC-ll3 (DRAFT) FORWARDED TO NRC BY G. NEILS .
NOV 1987 A-45 PRESENTATION TO ACRS DHR S.C.: NSAC-113 DELIVERED TO ACRS JAN 1988 A-45 PRESENTATION TO ACRS DHR S.C. ON NSAC-ll3 {
O :
GLV:3S67NSS l I
I
/ #e,,
UMTE D STATES NUCLE AR REGUL ATORY COMMISSION q
[ 9t At* NOT ON. D. L. so6ss
(...../ .
JUL 2 0 567 Dr. Gerald Neils, Chairman NUMARC Working Group on OHR Morthern States Power Compa y -
414 Nieollet Hall Minneapolis, Minnesota 65401
Dear Dr. Neils:
I received a copy of your June 22, 1987 comunication to Dr. David Ericson on the subject of our U$1 A-45 Program on Shutdown Decay Heat Removal Requirements. Sandia National Laboratories (SNL) is in the process of studying..
these cocents in detail; however, we are particularly interested in your indication that a separate PRA of one of the SNL Case Studies was sponsored by
, EPR1, the Westinghouse Owners Group, and Wisconsin Electric Power and concludes that core melt risk is about ten times lower than the SNL Case Study for the same plant. Since we perceive that this industry-sponsored study represents your quantification of the dif fering views outlined in your June 22nd letter, it would assist us in our deliberations to better understand your technical basis for these differences. Therefore, we ask that you identify the rajor items in your recent PRA study which contribute to the factor of ten difference in core melt probability, and present the technical basis for the value(s) selected (e.g., referenceable operating experience data base, human factor studies, coeponent reliability data, external event initiating frequencies, s etc).
We also wish to acuowledge the creation of the NUKARC Working Group to study the OHR issue, and we look forward to interfacing with members in the near i future. Since our draft Regulatory Analysis on U$1 A.- 5 is still
< pre decisional, perhaps we can consider a first meeting to focus on eur review of the items you will identify as ley contributors to the factor of ten We appreciate difference in the core relt probability between the two PRAs.
l the technical attention that NUMARC has apparently devoted to review of the six Case Studies, as indicated by Enclosure 1 of your June 22nd letter, and we intend to work with SNL to consider your coments.
Sine rely, 44*D./ %
Brian W. Sheron, Director Division of Reactor and Plant Systems Of fice of huclear Regulatory Research cc: D. Ericson O
Northem States Power Company (a /
uwawivvww ste Teephcm t6th 3E SMO october 28, 1987 Dr. Brian W. Sheron, Director Division of Reactor & Plant Systets Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Co==ission Washington, D.C. 20555
Dear Dr. Sheron:
In respense to your letter dated July 20, 1957, I ac pleased to forvard a draft copy of a document entitled "EFR1/WO; Analysis of Decay Heat Receval Risk at Point Beach." This study, sponsored by EPRI and the Vestinghouse Cvners Group, was prepared by Science Applications In:ernational Ccrporation and Westinghouse Electric Corperatien with the assistance of Wisconsin Electric Fover Company, the owners and cperaters of Point Beach. The N WARC Working Group on DHR has followed and endersed this effort.
The prie.ary purposes of this study were to previde a best-enticate analysis of DHR risk at a select.e4 USI A-I.5 Ca se Study plant and to quantify the
~
differences discussed in our Jur,e 22 ccement letter on the Case Studies.
The results of this Point Beach reanalysis, as they nov stand, indicate an approximate factor of thirty reductien in core-ceit frequency for the requences included in the secte of the STC study; an approximate factot- of seven reduction in the offsite c on s e q ,;e nc e s of these sequences, over and above the core-ceit frequency reduction; and an approximate 50-4001 increase in the e s t is.a t e d c o s t of the varicus backfit preposals evaluated in the NRC study. The ETR1/WOG findings indicate that the core-relt frequency esticate for Toint Be ach (1.0 x 10'8 per reactor year) is a f actor of ten lover than the core-ceit frequency target in the NEC's Safety Goal. The ElrR1/WDC study, like the NRC study, also concludes with a sery high degree of confidence that an add-on, dedicated SDKR systee would not be cost-beneficial for Point Beach.
Ve vould be pleased to meet with you and techtte of your staf f, ss suggested in your letter, to discuss the esthodelegies, technical bases and findings contained in both studies. We have provided you with this draft report prior to publication to allev sufficient tice for your staff to familiarire In antici-theeselves with the EPR1/WDC reanalysis in advance the meeting.
pation of that resting. EFR1 and V0G are continuing to double-check the codels used in their analysis against the final, as published, numerical values used in the NRC study. Although seee small changes in the cceparative estiesten of core celt frequency could occur in seee instances, we expect the 3
(V
' Dr. Pricn Sheren Octobor 28, 1987
. Page 2 of 2 O' overall results and conclusions of the EPRI/WOG study to remain essentially In the meanwhile, we would be pleased to schedule a meeting for unchanged.
the first mutually convenient opportunity.
Sincerely, f d' C H Neils Chairman ,
Ntt.GC Working Group on DHA CHN/vf cc: B. Lee, N"MARC NLHAEC Verking Group He: bars T. Spets NRC K. Kneil. NEC A. Marchese. NEC D. Ericson, Sandia National Laborateries R. Nevten Vis:ensin Electric Power Co.
V. J. Farkinsen, SAIC
- 0. T. Faddleford Vestingheuse Electric Corp.
A. Ladieu Chairman VOC Analysis Sut;e==ittee
! .1. Taylcr. EPKI V. Laysan. EFK!
O T. Marsten. EFL1 G. Vine. ETK!
J. Haugh. EFFI 1 l O
1 L _ _ _ _ _ . .- - ___ _ _ _ _ _
o O
EPRl/WOG ANALYSIS OF DHR RISK AT POINT BEACH PRESENTATION TO THE ADVISORY COMMITTEE ON REACTOR SAFEGUARDS DECAY HEAT REMOVAL SYSTEMS SUBCOMMITTEE JULY 27. 1988 O
BY J. J. HAUGH, EPRI W. J. PARKINSON, SAIC AND D. F. PADDLEFORD. WESTINGHOUSE O
OBJECTIVE OF NSAC-ll3 .
O REANALYZE THE DHR RISK AT POINT BEACH. BY BEST ESTIMATE METHO AND DATA. USE THE A-4S CASE STUDY OF POINT BEACH AS THE BASELINE AND POINT OF COMPARISON.
SPECIFIC OBJECTIVES ,
- 1. QUANTIFY THE CONSERVATISMS IN THE A-4S CASE STUDY. AND DEMONSTRATE THESE CONSERVATISMS AND LIMITATIONS CAN BE CORRECTED BY BEST-ESTIMATE ANALYSIS.
2, CONDUCT THE REANALYSIS WilH THE SAME SCOPE AND PLANT MODE USED IN THE CASE STUDY. 50 DIFFERENCES IN RESULTS CAN BE LIMITED TO DIFFERENCES IN INPUT DATA. SUCCESS CRITERI A. AND CONSTRAINTS ON NON-SAFETY EQUlFMENT AND HUMAN PERFORMANCE (PROVIDE FOR EASY SIDE-eY-SIDE COMPARISON),
O
- 3. DEMONSTRATE QUANTITATIVELY THE PARTICULARLY SIGNIFICANT SHORTCOMINGS IN THE CASE STUDY TREATMENT 0F EXTERNAL EVEN AND THE DEDICATED DHR SYSTEM.
- 4. RESPOND TO NRC REQUEST FOR DUANTITATl H CRIT 10VE OF CASE STUDIES. PROVIDE BETTER OVANTITATIVE BASIS FOR NRC REGULATORY ANALYSIS.
- 5. PROVIDE QUANTITATIVE BASIS FOR NRC/NUMARC DISnlSSIONS ON A-4 RESOLUTION.
O GLV:3S57NS8 l
l )
l i O l POINT BEACH CORE MELT FREQUENCY ESTIMATES l 1 !
l AS A RESULT OF TECHNICAL EXCHANGES BETWEEN EPRl/WOG AND !
NRC/SNL. THE ORIGINAL CASE STUDY ESTIMATE OF CORE MELT j FREQUENCY (CMF) DUE TO DHR AT POINT BEACH HAS BEEN !
I LOWERED FROM 3.0E-4 TO 9.3E-5 PER REACTOR-YEAR !
1 i i
THE NSAC-113 ESTIMATE IS 1.0E-5 PER REACTOR-YEAR i .
[
THESE TOTAL CMF ESTIMATES INCLUDE CONTRIBUTIONS FROM [
INTERNAL AND EXTERNAI. ("SPECIAL EMERGENCY") EVENTS O !
i i
i O
I PRINCIPAL CONTRIBUTORS TO REVISED CASE STUDY CMF .
O THE PRINCIPAL INTERNAL EVENTS CONTRIBUTING TO THE CMF ESTIMATES INCLUDE:
LONG TERM STATION BLACK 0UT (LTSB SEQUENCES)
SMALL BREAK LOCA WITH FAILURE TO IMPLEMENT SUMP RECIRCULATION (S2MH1'H2' SEQUENCE)
LOSS OF 0FFJ IE POWER TRANSIENTS (TIMLE SE0VENCE)
REACTOR / TURBINE TRIP TRANSIENTS INVOLVING A STUCK OPEN RELIEF VALVE (T30H1'H2' SEQUENCE) 0 THE PRINCIPAL EXTERNAL EVENTS CONTRIBUTING TO THE CMF ESTIMATES INCLUDE:
- 7,clSMIC SEQUENCES FIRE SEQUENCES (AFW PUMP ROOM A!!D 4.'60V SWITCHGEAR ROOM)
- THE CONTRIBUTION DUE TO SEISMIC AND FIRE SEQUENCES IS LARGER THAN THE CONTRIBUTION DUE TO AI.L INTERNAL EVENTS O
i l
l
- COMPARISON OF PKINCIPAL CONTRIBUTORS n ($)
EVENT REVISED EPRl/WOG t
CATEGORY CASE STtDY CASE STUDY NSAC-ll3
! TOTAL INTERNAL 1.3E-4 2.9E-5 2.6E-6 LTSB 3.6E-5 9.9E-6 5.4E-7 S2MH1'H2' 4.7E-5 7.0E-6 5.8E-7 TIMLE 6.7E-6 4.9E-6 7.7E-7 T30H1'H2' 2.5E-5 3.6E-6 N/A SUBTOTAL 1.lE-4 2.5E-5 1.9E-6 i
TOTAL EXTERNAL 1.7E-4 6.4E-5 7.5E-6 SEISMIC 6.lE-5 4.lE-5 7.4E-6
(]) FIRE 3.2E-5 2.2E-5 6.3E-8 l
l SUBTOTAL 9.3E-5 6.3E-5 7.5E-6 4
'J0TAL 3.0E-4 9.3E-5 1.0E-5 i
PROBABILITY OF OCCURRENCE PER REACTOR-YEAR l
(
1 l
(E)
L ___
~
REASONS FOR DIFFERENT CMF ESTIMATES THE CMF ESTIMATES DIFFER IN SEVERAL RESPECTS:
CHOICE OF INPUT DATA MODELLING ASSUMPTIONS ALLOWANCE FOR RECOVERY ACTIONS THE DIFFERENCES CAN BE ILLUSTRATED CONVENIENTLY BY FOCUSSING ON THE PRINCIPAL CONTRIBUTORS TO THE REVISED CASE STUDY CMF ESTIMATE MANY OF THESE DIFFERENCES APPLY TO OTHER SEQUENCES NOT COVERED IN iHIS PRESENTATION (E.G., S2MDlD2, T2M0H1'H2' AND T30DlD2) $
ALTHOUGH WITH LESSER SIGNIFICANCE O
. SMALL-BREAK LOCA AND FAILURE TO IMPLEMEN1 SUMP RECIRC O S2MH1'H2': SBLOCA + MFW FAILURE + llPRS a LPRS FAILURES t CASE STUDY: 4.7E-5 REVISED CASE STUDY: 7.0E-6 I
NSAC-113: 5.8E-7 PRINCIPAL DIFFERENCES:
- 1. SBLOCA FREQUENCY CASE STUDY: 2.0E-2 BASED ON LEAKS <2-IN. DIA: DERIVED FROM ANO-1 IREP (MURLEY MEMO: ISOLABLE LOCAS 4
DOMINATED)
P NSAC-ll3: 3.0E-3 BASED ON LEAKS <2.0-IN. DIA DERIVED FROM O OCONEE PRA AND INDUSTRY EXPERIENCE. CREDITS ISOLATION PRIOR TO RECIRCULATION AT ~20 HRS:
SUMP RECIRCULATION NOT REQUIRED FOR ANY LOCAS EXPERIENCED S0 FAR L
STATUS: REVISED CASE STUDY ACCEPTS 3.0E-3 SBLOCA FREQUENCY
NSAC-ll3: 1E-4 BASED ON POST THI ERGS /EOPs AND STAFFING REQUIREMENTS AS WELL AS LONG TIME TO DEPLETE RWST l
STATUS: NO AGREEMENT
4 SMALL-BREAK LOCA ANL FAILURE TO IMPLEMENT SUMP RECIRC ,
(CONTINUED)
O
- 3. REC 0VERY FROM RECIRCULATION FAULTS CASE STUDY: CREDITED LOCAL VALVE OPERATION TO ESTABLISH RECIRC FLOW PATH IN SOME INSTANCES :
REVISED CASE STUDY: SAME AS ABOVE NSAC-ll3: FULLY CREDITS POINT BEACH PROCEDURE ,
ECA-1.1: INCLUDES LOCAL VALVE OPERATION AND RWST REFILL (BUT NOT APPLIED WHEN OPERATOR FAILS TO IMPLEMENT RECIRCULATION.)
STATUS: NO AGREEMENT J
O EPRI/WOG CC cuSIONS THE CASE STUDY TREATMENT OF OPERATOR FAILURE TO 1 SWITCHOVER TO RECIRCULATION IS INCOMPLETE. A COMPLETE APPLICATION OF NUREG/CR-1278 WOULD YlELD A LOWER PROBABILITY OF FAILURE. THE NSAC-113 METHODOLOGY IS BASED ON SIMULATOR DATA I
NSAC-ll3 FULLY CREDITS THE POINT BEACH PROCEDURES THAT ADDRESS RECOVERY FROM EQUIPMENT FAILURES DURING SWlTCH0VER III i
. . - - - . . . . - . - , . - - - - - , - ~ , - - . - - - , - - - - -. . , - - - - - - - - - - - . - - .
1 EVENT 0: STUCK-0 PEN RELIEF VALVE O
T30H1'H2': REACTOR / TURBINE TRIP (MFW AVAILABLE) +
SRVs FAIL TO CLOSE + HPRS & LPRS FAILURE CASE STUDY: 2.5 E-5 ,
REVISED CASE STUDY: 3.6E-6 NSAC-ll3: NOT APPLICABLE
^
PRINCIPLE OlFFERENCES:
CASE STUDY: SRVs ASSUMED TO HAVE A 7% CHANCE OF OPENING IMMEDIATELY FOLLOWING TRIP:
PORVs ASSUMED TO BE BLOCKED OPEN O CONTINUQUSLY REVISED CASE STUDY: SRVs ASSUMED TO HAVE A 1% CHANCE OF ,
OPENING IMMEDfATELY FOLLOWING TRIP BASED ON SNL RSSMAP STUDY FOR A B&W PWR NSAC-113: EVENT 0 SEQUENCES DO NOT EXIST FOR REACTOR OR TURBINE TRIPS (T3) AT PT BEACH: NE!THER PORVs NOR SRVs WILL BE CHALLENGED BASED ON WESTINGHOUSE PWR OPERATING EXPERIENCE EVENT Q CONSERVATIVELY INCLUDED THE OPENING 0F BOTH PORV3 IN ALL LOSS OF 0FFSITE POWER (TI) AND LOSS OF MAIN FEEDWATER (T2) SEQUENCES O
STATUS: NO AGREEMENT
t Ol i EVENT 0: STUCK-OPEN REllEF VALVE (CONTINUED) l EPRl/WOG CONCLUSIONS: l 1 i THE REVISED CASE STUDY IS NOT CONSISTENT WITH L
THERMAL HYDRAULICS AND OBSERVED W PWR EXPERIENCE, AND IS INCONSISTENT WITH CURRENT PRA PRACTICE-(1.E., ,
i BOTH INDUSTRY AND NUREG-1150 STUDIES) i i ;
! 5 i
1 ei i
)
i l
f l 9 l
O' DIVERSE ECM!ilSMS FOR PRESSURE CONTR0_
ECENI91 SET PRESSURE ESIGN TRAf4SIENT l
EATERS 2250 PSIA NORMAL OPERAT10fi SPRAY 2275-2325 10% ET LOAD CMfE PORV 2350 50% STEP LOAD REDUCTION i
REACTOR TRIP 2400 TOTAL LOAD REJECT 10i1 WtTY VALVES 2500 LOAD REJECTION W0 lhTDI ATE TRIP
- O
! O EACH ECM1191 ESIGtED TO PREVENT OPERATION OF TEXT ECMilS1 FOR ITS
- MSIGN RA'E OF TRANSIENTS O TE ECANIS*o' ARE ESIGtED WITH REDUfMICY AND Wt SUBSTANTIAL l OVERCM ACITY (IN STARTUP TEST SPRAY WAS AT 0JATE TO PREVENT PORV OPENING FOR 25% LOAD REDUCTION, OfE PORV ANQUATE TO PREVENT HIGH PRESSJRE TRIP, ETC.)
i O
e MEASURED PLANT RESPONSE at a Westinghouse PWR PRESSURIZER PRESSURE, PSIG 2410 - - - - - - - . - - - - - -
2385 < - - - - - - - - - - 1 I
I l l 2335- ------ I I I
i I l
1 l l l 3 I i l 1 .
l 2;0 l
, 4.0 l
,6.0 l 8.0 TIME, SEC.
e 7.1 SEC (MAX.
PRESSURIZERPRESSURE) 5.7 SEC (HIGH PRESSURE REACTORTRlrDEMANDED) 3.7 SEC (PORV SETPOINT)
Figure 1 PRESSURIZER PRESSURE RESPONSE FOR NET LOAD TRIP (GRID DISCONNECT) FRD FULL POWER WITH N3 PRESSURIZF.R PORVS OPERATIONAL
1T LONG TERM STATION BLACK 0UT O
CASE STUDY: 3.6E-5 REVISED CASE STUDY: 9.9E-6 NSAC-113: 5.4E-7 PRINCIPAL DIFFERENCES:
CASE STUDY: -
USED IREP DATA FOR DIESEL GENERATOR FAILURE AND 8 HOUR HISS 10N TIME USED LOSS OF 0FFSITE POWER FREQUENCY BASED ON NUREG-1032 GENERIC DATA REC 0VERY INCLUDED OFFSITE POWER AND DIESEL O GENERATOR REPAIR NSAC-ll3: -
USED NSAC-108 DATA FOR DIESEL GENERATOR FAILURE (WITH 2-3 HOUR EFFECTIVE MISSION TIME)
USED LOSS OF 0FFSITE POWER FREQUENCY BASED ON PLANT SPECIFIC DATA REC 0VERY INCLUDED OFFSITE POWER REFILLING CST AND BALANCING SERVICE WATER LOADS TO DIESELS STATUS: -
REVISED CASE STUDY ACCEPTS NSAC-113 LOSS OF 0FFSITE POWER AND DIESEL GENERATOR DATA NO AGREEMENT ON RECOVERY ACTIONS O
LONG TERM STATION BLACK 0UT (CONTINUED) 0 EPRl/WOG CONCLUSIONS:
- CST REFILL RECOVERY ACTION IS REASONABLE; HOWEVER, PLANNED POINT BEACH HODIFICATION WILL PROVIDE ADDITIONAL 200,000 GALLONS OF CAPACITY, ELIMINATING THIS CONCERN SERVICE MTER RECOVERY IS NOT SIGNIFICANT TO THE FINAL CMF ESTIMATE RECENT GAS TURBINE GENERATOR RELIABILITY DATA FROM POINT BEACH INDICATES LTSB FREQUENCY SHOULD BE REDUCED BY A FACTOR OF 2 TO 3 O
O
LOSS OF 0FFSITE POWER <
(Il TIMLE: LOSP + MFW FAILURE + AFW FAILURE + F8B FAILURE (PREDOMINANTLY SHORT TERM STATION BLACX0VT)
ESTIMATES: CASE STUDY 6.7E-6 REVISED CASE STUDY: 4.9E-6 NSAC-il3: 7.7E-7 PRINCIPLE DIFFERENCES:
- 1. USE OF NEW STATION BATTERIES CASE STUDY: ANALYSIS CONDUCTED PRIOR TO INSTALLATION OF NEW BATTERIES NSAC-ll3: INCLUDED OPERATOR ACTION TO USE NEW BATTERIES TO START DIESELS AFTER INDEPENDENT OR COMMON O CAUSE FAILURE OF STATION BATTERIES STATUS: REVISED CASE STUDY HAS NOT YET CREDITED NEW BATTERIES IN THIS SE0VENCE (NEW BATTERIES CREDITED IN REVISED CASE STUDY SEISMIC EVALVATION)
- 2. LOSS OF 0FFSITE POWER FREQUENCY CASE STUDY: 8.4E-2 BASED ON NRC GENERIC ESTIMATE (NUREG-1032)
NSAC-ll3: 6.2F.-2 BASED ON PT. BEACH SPECIFIC DATA STATUS: REVISED CASE STUDY ACCEPTS THE NSAC-ll3 DATA O
LOSS OF 0FFSITE POWER (CONTINUED) .
3 0
- 3. DIESEL GENERATOR FAILURE PROBABILITY CASE STUDY: 3,8E-2 BASED ON IREP GENERIC DATA
)
NSAC-ll3: 2,2E-2 BASED ON NSAC-108 i STATUS: REVISED CASE STUDY HAS NOT YET ACCEPTED NEW DIESEL DATA IN THIS SEQUENCE (BUT CREDITED IN ,
LONG TERM STATION BLACK 0UT) .
EPRl/WOG CONCLUSIONS:
INCORPORATING NSAC-ll3 DATA AND RECOVERIES ALREADY CREDITED IN OTHER REVISED CASE STUDY SEQUENCES REDUCES THE REVISED CASE STUDY CMF TO 8.6E-7 O O
. FIRE EVALVATION O
CASE STUDY: 3.2E-5 REVISED CASE STUDY: 2.2E-5 NSAC-113: 6.3E-8 PRINCIPAL DIFFERENCES:
- 1. FIRE INITIATOR DATA CASE STUDY: -
USED AUX BLDG GENERIC FIRE FREQUENCY DATA REPORTED IN SEABROOK PRA; ALL REFERENCED FIRES ASSUMED TO APPLY TO AFW PUMP ROOM AND SWITCHGEAR ROOM.
AUX BLDG GENERIC FIRE FREQUENCY DATA (2) RAT 10ED BY THE AMOUNT OF COMBUSTIBLES (CABLE LOADING) IN THE AFW PUMP ROOM AND SWITCHGEAR ROOM TO THAT IN ENTIPE AUX BLDG, NSAC-113: -
USED GENERIC AUX BLDG FIRE FREQUENCY DATA REPORTED IN LIMERICK FIRE STUDY (BUT NO REDUCTION FOR CABLE QUALITY)
GENERIC FIRE FREQUENCY DATA PART10NED BY FIRE TYPE (1.E., AS CABLE FIRES AND ELECTRICAL PANEL FIRES) AS APPROPRIATE TO EACH FIRE ZONE: APPLIED SCALING FACTORS APPLICABLE TO EACH FIRE TYPE.
O
FIRE EVALVATION (CONTINUED) 9 POSTULATED A FREQUENCY FOR LARGE TRANSIENT COMBUSTIBLE FIRES STATUS: -
NO AGREEMENT ON PARTITIONING OF INITIATORS BY TYPE AND ZONE EPRl/WOG CONCLUSIONS:
THE GENERIC DATABASES SHOW NO EVIDENCE OF LARGE TRANSIENT COMBUSTIBLE FIRES OF THE SORT MODELED IN THE CASE STUDIES.
A LARGE TRANSIENT COMBUSTIBLE FIRE SHOULD BE POSTULATED AS A SEPARATE INITIATOR, $
NOT ALL FIRES IN THE DATABASE CITED BY THE CASE STUDY ARE APPLICABLE TO THE AFW PUMP ROOM AND THE 4160 V SWITCHGEAR ROOM, THE GENERAL METHODOLOGY FOLLOWED IN NSAC-113, WHEREIN FIRE FREQUENCY DATA ARE APPORTIONED BY FIRE TYPE AND SCALED ACCORDINGLY, YlELDS A MORE REALISTIC FIRE INITIATOR MODEL O
i
FIRE EVALUATION (CONTINUED)
]
G THE ORIGINAL AND REVISED CASE STUDY ANALYSES OVERESTIMATED THE INITIATING EVENT FREQUENCY DERIVED FROM ITS CITED DATABASE, AND NSAC-113 UNDER-ESTIMATED THE INITIATOR FREQUENCIES NSAC-113 SHOULD HAVE USED THE SEABROOK FIRE INITIATOR DATABASE--
HIGHER IN!TIATOR FREQUENCY AND MORE FIRE TYPES APPLICABLE TO BOTH ROOMS
- 2. FIRE GROWTH CASE STUDY: -
COUPLED TRANSIENT COMBUSTIBLE FIRES O TO GENERIC FREQUENCY DATA TO BOUND FIRE ENERGETICS (USED COMPBRN 1983)-
-USED TRASH CAN AND 10-GA'. ACETONE P0OL FIRES NSAC-113: -
APPLIED A CONDITIONAL PROBABILITY ON TRANSIENT SOURCE (0.3 FOR AFW PUMP ROOM AND 1.0 FOR SWITCHGEAR ROOM)
DID NOT PERFORM COMPBRN ANALYSES STATUS: -
NO AGREEMENT ON NEED FOR CONDITIONAL PROBABILITY O
FIRE EVALVATION (CONTINUED) ,
O EPRl/WOG CONCLUSIONS:
THE MECHANISTIC FIRE GROWTH MODELING IN THE CASE STUDY SHOULD NOT BE ACCEPTED MECHANISTIC FIRE GROWTH MODELING FOR EACH INITIATOR TYPE SHOULD USE REALISTIC ASSUMPTIONS I
PROBABILISTIC FIRE RISX ESTIMATES SHOULD INCLUDE CONDITIONAL PROBABILITIES FOR THE ENERGETICS OF THE FIRE--INCLUDING LIKEllH000 0F t
OCCURRENCE, SIZE / INTENSITY AND LOCATION O
- 3. HALON SUPPRESSION CASE STUDY: -
CREDITED TWO-TRAIN HALON SYSTEM IN l I
SWITCHGEAR ROOM CREDITED ONE-TRAIN HALON SYSTEM IN AFW PUMP ROOM HALON SYSTEM FAILURE ESTIMATED AT 0.2 PER DEMAND BASED ON ACCEPTANCE TEST DATA REPORTED IN MILLSTONE PRA.
REVISED CASE STUDY: -
CREDITS TWO-TRAIN HALON SYSTEM IN AFW PUMP ROOM O
FIRE EVALUATION (CONTINUED) .
O NSAC-ll3: -
CREDITS TWO-TRAIN HALON SYSTEM IN AFW PUMP ROOM AND SWITCHGEAR ROOM HALON SYSTEM FAILURE ESTIMATED AT 0.06 BASED ON DOE HALON SYSTEM RELIABILITY DATA FROM ACTUAL FIRES i AT DOE FACILITIES RATHER THAN ACCEPTANCE TESTS STATUS: -
REVISED CASE STUDY INCORPORATES SECOND HALON TRAIN IN AFW PUMP ROOM NO AGREEMENT ON REMAINING DIFFERENCES O
EPRl/WOG CONCLUSIONS:
=
AUTOMATIC SUPPRESSION DATA SHOULD BE :
BASED ON ACTUAL FIRES. THEREFORE, USE OF THE USDOE HALON RELIABILITY DATA CITED IN NSAC-113 IS MORE APPROPRIATE O
.i FIRE EVALUATION (CONTINUED)
G
- 4. MANUAL SUPPRESSION CASE STUDY: -
2.0E-1 AT 30 MIN AND 1.0E-1 AT 60 MIN: USED MANUAL NON-SUPPRESSION PROBABILITY CURVE REPORTED IN SEABROOK PRA USED ALLOWABLE TIME CALCULATED BY COMPBRN ANALYSES NSAC-ll3: -
1.5E-1 AT 30 MIN AND 4.0E-2 AT 60 MIN; USED MANUAL NON-SUPPRESSION LOGIC MODEL FROM MILLSTONE PRA SUPPLEMENTED BY LIMERICK FIRE STUDY CURVE USED CASE STUDY ESTIMATES OF g ALLOWABLE TIME STATUS: -
NO AGREEMENT ON NON-SUPPRESSION VALUES EPRI/WOG CONCLUSIONS:
THE NSAC-ll3 ESTIMATE OF MANUAL SUPPRESSION RELIABILITY IS MORE APPROPRIATE O
FIRE EVALVATION (CONTINUED)
O
- 5. RECOVERY CASE STUDY: HUMAN ERROR CONTRIBUTION TO TDAFW PUMP FAILURE ESTIMATED AT 0.1 DURING AFW PUMP ROOM FIRE.
NO CREDIT GIVEN FOR TDAFW PUMP IN 4160V SWITCHGEAR ROOM FIRE NSAC-ll3: -
CREDITED TDAFW PUMP IN AFW PUMP ROOM AND 4160V SWITCHGEAR ROOM FIRES:
CREDIT GIVEN TO NEW BATTERIES HUMAN ERROR CONTRIBUTION TO TDAFWP FAILURE ESTIMATED AT 0.03 O STATUS: -
NO AGREEMENT EPRl/WOG CONCLUSIONS:
TDAFW PUMP OPERATION AFTER THE FIRE SHOULD BE CREDITED IF INSTRUMENTATION REQUIRED FOR OPERATION EXISTS. THEREFORE, THE NSAC-113 ESTIMATE OF TDAFW PUMP RECOVERY IS MORE APPROPRIATE DUE TO THE NEW BATTERIES.
O
SEISMIC SEQUENCES 9
CASE STUDY: 6.1E-5 REVISED CASE STUDY: 4.1E-5 NSAr.-113: 7.4E-6 PRINCIPAL DIFFERENCES:
- 1. HAZARD CURVE CASE STUDY: -
GENERATED A SEISMIC HAZARD CURVE BASED ON ZION (SSMRP) AND CALCULATED LOCAL S0ll COLUMN EFFECT. ,
NSAC-ll3: -
SEISMIC HAZARD CURVE WAS REDUCED $
FROM CASE STUDY VALUES BY FACTORS OF TWO FOR <3*SSE AND FIVE FOR
>3*SSE BASED ON EPRI VS LLNL EVALUATIONS OF BRAIDWOOD STATUS: NO AGREEMENT
- HAZARD CURVE BEING ADDRESSED SEPARATELY--SEISMIC MARGINS RESEARCH O
SEISMIC SEQUENCES (CONTINUED) 0
- 2. RECOVERY ACTIONS CASE STUDY: -
NO CREDIT GIVEN FOR SEISMIC REC 0VERY ACTIONS.
NSAC-ll3:
- ONLY ALLOWED CREDIT FOR RECOVERY ACTIONS UP TO 3*SSE, E.G., USE OF SERVICE WATER SYSTEM IN PLACE OF CST AND SPENT FUEL POOL IN PLACE OF RWST.
STATUS: NO AGREEMENT ON CST AND RWST RECOVERY ACTIONS.
REVISED CASE STUDY STATES THAT RECOVERY SHOULD NOT BE CREDITED AT LESS THAN 30 MINUTES.
- 3. RWST FAILURE AFFECTING RECOVERY CASE STUDY: -
CALCULATED TANK FAILURE DUE TO BUCKLING AND ANCHOR PULLOUT.
NSAC-113: -
RWST NOT EXPECTED TO Fall CATASTROPHICALLY (INSTANTANE0USLY)
AT LOW ACCELERATIONS SUCH THAT ABOUT 30 MINUTES ADDITIONAL TIME AVAILABLE FOR REC 0VERY (TOTAL OF 60 MIN TO CORE UNC0VERY)
RWST RECOVERY ONLY CREDITED IN SMALL-SMALL LOCA SE0VENCES (<1.5-IN
^)
O
SEISMIC SEQUENCES (CONTINUED) ,
STATUS: -
EPRl/WOG REEVALVATION INDICATES 120 MINUTES AVAILABLE TO CORE UNC0VERY (90 MIN W/0 RWST) DURING SMALL-SMALL LOCA SE0VENCES
- 4. NEW SEISMIC i BATTERIES AFFECTING RECOVERY CASE STUDY: -
ANALYSIS CONDUCTED PRIOR TO INSTALLATION OF NEW BATTERIES.
REVISED CASE STUDY: INCORPORATES NEW BATTERIES NSAC-ll3: -
INCLUDED NEW SEISMIC I BATTERIES DURING RECOVERY (CONSIDERED HUMAN FAILURES)
STATUS: -
REVISED CASE STUDY CREDITS TFE NEW g BATTERIES IN SEISMIC SEQUENCES EPRl/WOG CONCLUSIONS:
THE MOST SIGNIFICANT NL3ERICAL DIFFERENCE IN CMF RESULTS FROM THE HAZARD CURVE USE OF SERVICE WATER FOR AfW WATER SUPPLY IS PART OF DESIGN BASIS AND TRAINING: IT WlLL SURVIVE SEISMIC EVENTS <3*SSE AND THEREFORE SHOULD BE CREDITED FOR CST RECOVERY RWST REC 0VERY IS BEllEVED TO BE APPROPRIATE BUT NOT SIGNIFICANT TO THE FINAL CMF RESULTS O
GENERAL COMMENT
S ON A-45 STUDIES e
- 1. DEDICATED SDHR FAILS ALL COST BENEFIT MEASURES BY A WIDE ,
MARGIN.
- 2. MANY OF THE REASONS FOR HUCH LOWER RISK AT PT BEACH GENERALLY APPLY TO OTHER CASE STUDY PLANTS. ,
- 3. WITH POSSIBLE EXCEPTION OF SEISMIC, ALL OTHER "EXTERNAL RISK" FACTORS SELECTED FOR ANALYSIS BY A-45 CASE STUDIES SHOWN TO BE INSIGNIFICANT.
- 4. BEST-ESTIMATE ANALYSIS IS ESSENTIAL FOR CREDIBLE USEFUL RESULTS. ANY ADDITIONAL MARGIN (IF NEEDED) SHOULD BE ADDED AT END OF ANALYSIS. IMPORTANT LESSON FOR IPE PROCESS.
- 5. U.S. NUCLEAR POWER PLANT OPERATING EXPERIENCE IS THE O BEST SOURCE OF CREDIBLE DATA FOR BEST-ESTIMATE ANALYSIS, AND BEST FOUNDATION FOR "DEFINING THE PROBLEM'.
(E.G., A-44)
- 6. WHERE POSSIBLE, THE VALIDITY OF THE ASSUMPTIONS USED IN THE ANALYSES SHOULD BE TESTED AGAINST OPERATING EXPERIENCE
- 7. A-45 HAS NOT YET ACCOUNTED FOR SAFETY IMPROVEMENTS FROM OTHER NRC AND INDUSTRY PROGRAMS (IN PARTICULAR. SB0 AND APPENDlX R ).
O
A-45 RESOLUTION VIA OTHER PROGRAMS ,
9 MAJOR SOURCES OF RISX IN PROGRAMS ADDRESSING THESE SANDIA STUDIES RISK SOURCES STATION BLACK 0UT A-44, NUMARC
- CONTAINMENT RELIABILITY, NUMARC, IDCOR, NUREG-1150 FAILURE MODES, CONSEQUENCES
- PWR AFW REllABillTY, BWR INPO, NSSS OGs H.P. CORE COOLING SBLOCA EPRI, INP0, NSSS OGs PCS LOSSES (SCRAMS) INPO, NSSS OGs, NSAC, AE0D l
DC POWER RELIABILITY NSAC, INP0 O
VALVE PERFORMANCE EPRI, INPO, AE00 FIRE PROTECTION APPENDIX R, INPO
- INTERNAL FLOODING, LIGHTNING NSAC, INPO, A-17 SEISMIC RISK EPRI, S0G, SOUG, A-46, A-17 SAB0TAGE UTILITY SECURITY PROGRAMS, FITNESS FOR DUTY RULE O
l l
l 1
0 FIRE RISK ASSESSMENT PROCEDURE FIRE GENERIC FREQUENCY POSTULATE TRANSIENT I l INITIATORS APPROPRIATE TO BLDG. COMCUSTIBLE FIRE l l 4 FREQUENCY j l SORT BY APPLICABLE FIRE TYPES l APPLICABLE TO EACH FIRE 2ONE IN BLDG.
p I APPLY WEIGHTING FACTORS APPROPRIATE I I
TO EACH FIRE TYPE IN EACH FIRE ZONE l FIRE CALCULATE FIRE ENERGETICS APPROPRIATE l
SEVERITY a
TO EACH f!RE TYPE IN EACH FIRE ZONE l
APPLY CONDITIONAL [ PROBABILITIES FOR l O I l
ENERGETICS APPROPRIATE TO EACH FIRE j TYPE IN EACH FIRE ZONE
- -)
FIRE EVALUATE SUCCESS OF AUTO. SUPPRESSION SUPPRESSION I l 4 ;
l EVALUATE SUCCESS OF MAN. SUPPRESSION M-
- - - -- --. - - - - - - - l RECOVERY EVALUATE RECOVERY l p _. _ _. _ _. - - -. 4 - - - -
g EVALUATE C H f L___ -- - - -- -- - -a l
l O
O 1
1 INTEGRATION OF REGULATORY ISSUES AT A NUCLEAR POWER PLANT i
O 1
1
- ROGER NEWTON l WISCONSIN ELECTRIC POWER COMPANY l
l 0
INTEGRATION OF REGULATORY ISSUES AT A NUCLEAR PLANT I. OLD ISSUES A. INTERNAL FLOODING B. BLOCKWALLS C. APPENDIX R, FIRE PROTECTION D. TMI BACKFITS
- 1. Auxiliary Feedwater System Reliability
- 2. New Instrumentation
- 3. New Emergency Operating Procedures E. EQUIPMENT QUALIFICATION F. REGULATORY GUIDE 1.97 G. TRAINING REQUIREMENTS C) i
Ih"I'EGRATION OF_ REGULATORY ISSUES 4
AT A NUCLEAR PLANT II.,NEW ISSUES A. EMERGEhCY DIESEL GENERATOR RELIABILITY B. STATION BLACKOUT, A-44 C. DECAY HEAT REMOVAL, A-45 D. SEISMIC QUALIFICATION OF EQUIPMENT IN OPERATING 'LPITS, A-46 E. SEVERE ACCI;.
- h ' '. 2 C)
.~
_bmA6m.,-,- 5 = * = - - - im'E--------,M-m uA-~@mmu* '
-As-
=,aA---,- - -,-ma--6 m--:,-,eaaema -- -
a uaa-en-aoa-,.,, _-As 4
0 a
I d
i I
4 I
I 1 l 4
4 i
J (
i l
i
- b d
i f
I P
1 l
1
'\ ,
t I
1 I l I
I IP I
i l i
I i
i !
- i V
5 i
L
, ,,----- , .., , -. ,-,,-, ... . .--,-, -,_ . . - - .- - -- - - - - -. - .- - - - -- - .~-- --- - ----- .------- - - - -., ------- --- - - -
INTEGRATION OF PLANT SPECIFIC ITEMS l
ITEMS AT POINT BEACH NUCLEAR FLANT
- 1. APPENDIX R 4160 SWITCHGEAR ROOM BY-PASS
?. 13.8 KV EUS MODIFICATIONS
- 3. LOW VOLTAGE STATION AUXILIARY TRANSFORMER O COuSIDERAT10NS
- 4. BATTERY LIFE
- 5. INVEATER STATIC TRANSFER SWITCHES
- 6. PCB TRANSFORMER REPLACEMENT
- 7. WATER TREATMENT PLANT STORAGE TANK O
1 ,
PROPOSED MODIFICATIONS h
- INSTALL SWITCHGEAR TO BY-PASS 4160 SWITCHGEAR ROOM
- MODIFY 13.8 KV BUS SECTION I
- INSTALL A THIRD SAFETY RELATED EMERGENCY DIESEL GENERATOR (EDG) WITH ASSOCIATED SWITCHGEAR, BATTERY, COOLING, AND BUILDING.
- MODIFY THE 4160 VOLT SYSTEM TO FACILITATE THE NEW EDG -
O
- INSTALL THIRD LOW VOLTAGE STATION AUXILIARY TRANSFORMER t
INSTALL A SEISMIC MAKE-UP WATER STORAGE TANK (MUWST)
INSTALL AN UNDERGROUND CONNECTING DUCTBANK BETWEEN THE UNIT 2 FACADE AND NEW EQUIPMENT {
l LOCATION OF THE NEW EQUIPMENT BUILDING WOULD BE IN THE AREA CURRENTLY OCCUPIED BY WAREHOUSE 2
BENEFITS
- INCREASE THE RELIABILITY OF THE EDG SYSTEM AND THEREFORE THE RELIABILITY OF PBNP.
- ADDRESS UNRESOLVED SAFETY ISSUE (USI) A-44, '
STATION BLACKOUT.
i ADDITIONAL ONSITE EMERGENCY POWER SOURCE OR i ADDITIONAL DC BATTERY CAPACITY REQUIRED.
ADDITIONAL AC INDEPENDENT WATER SOURCE b REQUIRED.
- ADDRESS USI A-45, DECAY HEAT REMOVAL.
INCREASED RELIABILITY OF ONSITE POWER 4 REDUCES RISK.
ADDITIONAL SEISMIC AC INDEPENDENT WATER SOURCE REDUCES RISK.
4160 VOLT SWITCHGEAR ROOM BYPASS REDUCES CJ RISK TiUE TO FIRES. ,
. _ _ _ _ . .