ML20247A033
| ML20247A033 | |
| Person / Time | |
|---|---|
| Issue date: | 09/06/1989 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1753, NUDOCS 8909120025 | |
| Download: ML20247A033 (341) | |
Text
F 00k3F /7s3?
J - 34
- l O~ . !
UNITED STATES -
NUCLEAR REGULATORY COMMISSION RGEN A _
ADVISORY COMMITIEE G4 REACIOR SA?EGUARDS In the Matter of:
SUBOOMMITIEE Qi EXTRDE EXTERNAL DHEIOMEIR TRO4 4g DElfrE B. hMITE RE NRN ORIGINAL. TO B. WHITE, ACRS=IO10 :: g,,,,gjj---
Thanks!
Barbara Jo White 4-t409- A M E 8 Pages: 1 through 273 Place: Bethesda, Maryland
~~
Date: Septenber 6,1989 6~R,C,l'
- f4 g
'A}0 oA e;10860py 4 5temn
~
70d6 Uf6 Of Ele C0millidd(T i
O HERITAGE REPORTING CORPORATION OpidsfReporters 1224 L Street, N.W., Suke 600 Washington, D.C. 20005
- c. . v m.:
gog1_.g.._m 202) 6M T-27s3 c t.c L-__-_-_____-___-__-__
- . < l p .7 -
O L/i 1- PUBLIC NOTICE BY THE-
> 1 .2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S 3- ADVISORY COMMITTEE 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 uncorrected' record of the discussions
- 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 respon.sibility for errors or /
.g T)s 14 inaccuracies of. statement or data contained in this. transcript.
15.
16 17 18 19 20 21
.22 23 4
24
~
25 l b Heritage Reporting Corporation (202) 628-4888
n 1
i , rw.
i; f )1'
(( .14 UNITEDLSTATE6. NUCLEAR REGULATORY COMMISSION t{
2 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 3- )
In the' Matter-of: )
4 }
}
c 5l SUBCOMMITTEE ON EXTREME. EXTERNAL- .)
PHENOMENAL )
'6 )'
7 Wednesday, September 6, 1989
~8 Room P-110 L 9 7920 Norfolk Avenue.
Bethesda, Maryland 10
-The above-antitled matter came on for hearing, 11 pursuant to notice, at 8:35 a.m.
12-BEFORE: DR. CHESTER P.'SIESS f (a { 13 Professor Emeritus of Civil Engineering University of Illinois 14- Urbana, Illinois
~
- 15. ACRS MEMBERS PRESENT:
11 6 - MR. CARLYLE MICHELSON Retired Principal Nuclear Engineer 17 Tennessee Valley Authority Knoxville, Tennessee 18 and Retired Director, Office for Analysis and l- Evaluation of Operational Data h ~19' U.S. Nuclear Regulatory Commission Washington, D.C.
20 DR. IVAN CATTON 21 Professor of Engineering Department of Mechanical Aerospace and Nuclear 22 Engineering School of Engineering and Applie6 Science
- 23 University of California L .
Los Angeles, California 25'
2
.~
I r
)
1 MR. CHARLES J. WYLIE Retired Chief Engineer 2 Electrical Division Duke Power Company 3 Charlotte, North Carolina 4 PAUL G. SHEWMON Professor, Metallurgical-Engineering Department 5 Ohio Statr- University Columbus, Ohio.
6 ACRS COGNIZANT STAFF MEMBER:
7 A. Igne 8
NRC STAFF PRESENTERS:
9 A. Murphy ;
10 L. Reiter j G. Bagchi )
11 N. Chokshi l R. Cenneally l 12 i
("N s
(-) 13 l
14 15 16 17 18 !
19 20 21 22 23 24 25 HERITAGE REPORTING CORPORATION -- (202)628-4888
3 a/
'~
v 1 P_ E O_ C E E D_ 1 11 G S_
l 2 CHAIRMAN SIESS: The meeting will come to order.
3 This is a joint meeting of two ACRS subcommittees, the 4 Subcommittee on Extreme External Phenomena, and the 5 Subcommittee on Severe Accidents. I'm Chet Seiss. I'm 6 Chairman of the Extreme External Phenomena Subcommittee. Bill 7 Kerr is Chairman of the Severe Accident Subcommittee, but he 8 will not be able to be here today, but the other ACRS members 9 are present--Mr. Wylie, Mr. Catton, Mr. Michelson.
10 The cognizant ACRS staff engineer is Al Igne sitting 11 on my right.
12 The purpose of the meeting is to get a briefing from 13 the NRC staff on the status of the IPEEE program. That is the 14 individual plant examination for external events.
15 The rules for participation in today's meeting were 16 announced as part of the notice in the 7ederal Register on 17 August 18th. The meeting is being conducted in accordance 18 with all the provisions of the Federal Advisory Committee Act, 19 the Government and the Sunshine Act, and we have had no 20 written or oral statements from members of the public.
21 As usual, we request that each speaker first 22 identify himself or herself and use the microphone so that the 23 reporter can get what he or she says.
() 24 I would like to make a few introductory comments.
25 Then I'll ask other members of the Subcommittees if they have HERITAGE REPORTING CORPORATION -- (202)628-4888
4 7'
8
\~ - anything to to say.
1 l l
2 As you will recall, the individual plar a t examination j
)
3' program is a part of the I guess implementation of the 4 Commission's severe accident policy statement. That statement i
5 covered a number of things. It said we thought that, they 6 thought the plants were safe as they were now, but they needed 7 to take care of those things, needed to resolve all the .
8 outstanding unresolved safety issues, and high and medium 9 priority generic issues, and then--Mr. Michelson is laughing.
10 And then in addition, there was to be an examination 11- at each plant to look for, I think the official statement now 12 says look for vulnerabilities. I have also seen a statement 13 that says look for plant-specific or plant unique 14 vulnerabilities, and I think we have used the term look for 15 outliers.
16 Now the IPE or so-called internal events is covered 17 by generic letter, is it? I think we just, somebody said we 18 have gotten a copy of it. I haven't seen it.
19 MR. BAGCHI: 8820.
20 CHAIRMAN SIESS: It is a generic letter that tells i 21 them what they are supposed to do, but it also says that the 22 individual plant examinations for external events is not 23 covered and will be done in the 'uture, but I think there is I) 24 some provision for integrating .*. As I see it, going through 25 two complete cycles didn't make a whole lot of sense, but the i
HERITAGE REPORTING CORPORATION -- (202)628-4888 Q__=___-________.
5 ac 1- external events that we are' talking about are earthquakes,
'2 .high winds, floods, I think explosions from near-by 3 . facilities,'and fire.
4 Now that's, fire is within.the plant where they are.
5 External events has never been quite clear, and floods 6 inc1'udes.both floods from' rivers and lakes and oceans and so 7 forth, .and floods internal to the plant, so'the term external 8 has to be understood to really mean not things that you cover 9 normally I guess.
10 Now I think what we need to keep in mind, and I hope 11' the staff will clarify it, is that the objective of this thing 12 ~is to look for tF.ese so-called outliers or vulnerabilities or 13 plant-unique vu2nerabilities.
14' The objective is not to show that earthquakes are 15 , dominant contributors tu risk. We have had lots of PRAs that 16 reach that conclusion,.although I'm not quite sure what
= 17 - dominant means; that other PRAs that say floods are the 18 dominant contributor to risk, and others that say fire.is the 19 dominant contributor to risk, and others that say human error 20 are dominant contributor to risk, so I guess I'm not quite 21 sure what is a dominant contributor, but presumably in the 22 IPE, we are looking for things that may or may not be a 23 dominant contributor to risk, but in some way they represent a (f 24 plant unique vulnerability.
25 Now that term has been used by the staff, that plant HERITAGE REPORTING CORPORATION -- (202)628-4888
V ,
6
[ /~y
~
' ~ unique. I don't know what'it means--individual plant unique:
$~
12' or type of plant unique. That could be site unique
'3 vulnerabilities. Local flooding is found to be very site 4- dependent, but the IPEEE, I don't think we want tx) confuse it 5' with' things like the seismic margin review plants east of the
!6 Rockies. ~That's covered presumab1'y by A-40, USI A-40 I. guess-p
- 7. it was, so what the staff is going to talk about I hope is the p 8! vulnerability issue 'and they promised at another meeting I 9 attended'to come up with a definition of vulnerability.
10 If they have, we look forward to hearing it. The staff that is working on internal events, the last I read, had 12 not come up with a-definition of vulnerability. .They had 13 simply said that the individual utility will identify 14 vulnerabilities and define them, which can give us at a 15- mrximum what is 83' definitions which the staff.wi13 then 16 synthysize into something else,rso if you guys have come up.
17 with the definition of vulnerability, we certainly would like 18 to hear it.
19 We are going to have the staff in tomorrow afternoon.
20 for-the Full Committee to tell the rest of the Committee-what 21 they are planning to do. As T understand it, you have not yet 22- a final document on the IPEEE. This is exploratory right now 23 to.give us some idea in which directions you are going and i 24 why. Am I right?
2i: MR. SHAO: Right.
HERITAGE REPORTING CORPORATION -- (202)628-4888
i* 7 I -
1
~
' CHAIRMAN SEISS:
And then we would.like to pass that l
l
~2 knowledge on to the Full Committee by having the' staff come in-3- and talk about it, and I.believe on Friday, we are having
-4. NUMARC'in and they are going to cover among other things, ;
5 their reaction to the staff's proposals so far as they know 6 them.
7 MR. MICHELSON: How far do they know them, Chet? l 8' CHAIRMAN SIESS: Well, the--
9 MR. MICHELSON: We got the paper just now. l
- 10. CHAIRMAN SIESS: 1 attended a meeting the staff had 11 with NUMARC f to talk about everything but seismic, and that was j 12 the first they knew about it.
13 On the seismic stuff, I think NUMARC is much more up- 1 14 to date, and EPRI has also been involved I believe in the 15 seismic, and the staff will, we will ask the staff to give us 16 what they think NUMARC knows at this point,Lwhat documents and
-l 17 so forth.
18 MR. SHAO: NUMARC is here today.
19 CHAIRMAN SIESS: They are here, but they didn't want 20 to talk to the Full Committee at the same time you did for 21 some reason, but they are free to comment, time permitting, 22 today.
23- MR. MICHELSON: They might like to read it before j (k 24 they talk, too.
25 We can't even talk. We just got the material.
HERITAGE REPORTING CORPORATION -- (202)628-4888
i 8
)
1 CHAIRMAN.SIESS: We have got a-stack of stuff'here
-2. that none of us have seen, and we are not going to get it read
-3 by: tomorrow, but then since this is to sort of introduce it, 4 we are hoping'that the staff will orally try to summarize it, 5 some of the paper we have, and-tell us what they think we 6 -should be looking at.
! 7 Cther comments from the Subcommittee? Charlie? Do 8 you have any?-
9 'MR. WYLIE: No.
10 CHAIRMAN SIESS: Ivan?
11 -DR. CATTON: I defer. I haven't read the paper.
12 CHAIRMAN SIESS: Oh, well.
s 13 MR. BAGCHI: Dr. Seiss,-I think in order to put 14 things in perspective, the conference paper, that may be one 15 place you want to look for a summary.
16 CHAIRMAN SIESS: You can tell us what is in this 17 pile and what you think we might start with if somebody wants 18 to start reading.
19 Carl, some more--
20 MR. MICHELSON: Do you intend to have a subcommittee 21 meeting at such time as we have had an opportunity to digest 22 this and a number of other pieces of material which we have 23 never seen yet?
() 24 MR. SHAO: That would be a good idea.
-25 CHAIRMAN SIESS: I think right now the staff is HERITAGE REPORTING CORPORATION -- (202)628-4888
. -9 h
"~ ;.1 - looking-for' reaction.
2- MR. SHAO: Right now we are looking for reaction.
3 About.two; months, we will--
'4 MR.1MICHELSON: Reaction mup* mean off the top of 5 the head'then because we haven't gotten the material. Off the
'6 top of the head. reactions are not always too wonderful, l .
7 either.
8 CHAIRMAN SIESS: That's"what I was going to ask.the 9^ staff. You'really don't expect a letter, do you?
10 MR. SHAO: . Whatever reaction you have.
11 CUAIRMAN SIESS: The thing is the=only reaction you-12' can get that has any' validity--I shouldn't use that word.I 13 guess. You are going to get reactions from individuals, and-14 you are entitled to take those for what they are worth. You-15 have to evaluate how knowledgeable the individual is and how 16 much influence he might have on the Full Committee..
17 MR. SHAO: What I intend to do is in about'two 18 months, they can--we will produce to you, at that time we will 19 give some kind of introductory--
20 CHAIRMAN SIESS: Today you.will get some reactions 21 from the five people present. Tomorrow, there will be five 22 more maybe, and you may get some other reactions. None of 23 those are any'more than individual reactions. If a member of
) 24 the Full Committee that is hearing the presentations tomorrow 25 and perhaps even hearing the NUMARC presentation on Friday, if HERITAGE REPORTING CORPORATION -- (202)628-4888
=_ _ _ _ _ - _ - _ _ _ _ _ _ _
h.l ... ...
10
- 1. .there!_is something that the Full Committee reaches a consensus ,
-2 on, it is obvious that we think that we 1 ought to give to you 3t as official advice,'we will try to do that, put it in a
- 1. 4 letter.,
5 MR. SHAO: Okay.
l 6- CHAIRMAN SIESS: But you do realize that the 7 comments you get from individuals are just that. They may be 1
i
-8 very valuable. May be very knowledgeable individuals, but they-9- are still not ACRS positions, so whether you get a letter or 10 not will depend on whether we have heard enough and understood 11 .it well enough to reach a consensus.
.12 MR. SHAO: As I say, we are always willing to come 13 back.
14: CHAIRMAN SIESS: But we certainly, when you finalize
- 15. this, then we expect to get something. What will you do, a 16' generic letter?
17 MR. SHAO: A generic letter.
18 CHAIRMAN SIESS: We certainly expect to review the 19 generic letter by the Subcommittee and Full Committee.
20 MR. SHAO: That would be great.
21 CHAIRMAN SIESS: Daul, comments? Okay. Then we 22 will turn this over to the staff, and Larry Shao will be the 23 leadoff on it. We have got a lot of back-up behind him there,
() 24- and I assume others are going to be making presentations. You l
25 have a revised tentative schedule here.
11
'(
' ,1 MR. SHAO: My name is Larry.Shao. I am the Director 2 of the Division of Engineering in the Office of Research. I 3 talso happen to be the Chairman of the NRC Steering Group on
'4 ' External Events, but today we cre going to present to ACRS all 5' the external events except fire ~. I think the fire was 6 presented to ACRS a couple of months ago.-
7 MR. MICHELSON': Not from the viewpoint of IPEEE;-we
-8 haven't talked about how you are going to do a fire analysis 9' yet. It was promised I think by the~end of this year, but-10 not, not ever presented yet, to my knowledge at least.
11 MR. SHAO: Let me put it this way. The fire 12 positions will be, are being developed by NUMARC, and it will
.gs -
13 be reviewed by NRC. When the positions are developed, we will
- 14. come to ACRS.
15 CHAIRMAN SIESS: Okay, but it is important that we 16 have not, we have heard stuff on fire for the last letter we 17 wrote on the fire says that we have not yet heard the IPEEE 18 approach and we are waiting. The fire is still an open-issue 19 CHAIRMAN SEISS: Will still be open after today.
20 (Slide) 21 MR. SHAO: The severe accident policy statement did 22 not differentiate between internal and external events. As 23 Dr. Siess, said the IPEEE generic letter that was sent to the
-( I 24 industry in October 1988 only covered internal events for 25 mention. External events will come later, mainly because L
9- HERITAGE REPORTING CORPORATION -- (202)628-4888
.e 12 i 1 there was no' developed methodology to perform IPEEE for 2 external' events.
3 (Slide).
4 MR. .MICHELSON: Can you tell me briefly by l
5 definition what an internal and external event is? -Since I
- 6- have heard so many definitions, now I'm confused by your. slide L
L 7 because I' don't know what you are really say.
? MR. SHAO: Okay. Maybe I come later on to what is 9 external events. From that vugraph,~what is anything other l
10 than these--
11 MR. MICHELSON: I hope you will tell me what an 12 external' event is.
O 13 MR. SHAO: Yes.
14 MR. MICHELSON: Because external floods and internal 15 floods and so forth confuse the issue because many times you 16 say you are' including internal floods in the internal events.
17 MR. BAGCHI: I don't know if it requires a lot of 18 deliberation on this.
19 MR. SHAO: Let's go to the vugraph.
20 MR. BAGCHI: Internal and external, simply internal 21 to the system or external to the system.
22 MR. SHAO: It was a question about floods, external 23 event was--
) 24 MR. MICHELSON: We are doing floods obviously.
25 MR. BAGCHI: There was some confusior. about floods, HERITAGE REPORTING CORPORATION -- (202)628-4888
13 1 l ri .-
J H :;+
' / ( )~ - 1- but we had workshops. In the workshop we had said internal i JL flood is going to be considered under IPE. l 3 MR. MICHELSON: External floods is going to be i
4 . considered, .too, isn't'it?
l 5 -
MR. BAGCHI: For external events. ;
I 6 MR. SHAO: Internal floods.
7 MR. BAGCHI: Handled in internal IPE.
8 MR. MICHELSON: He will clear it up later.
9 MR. SHAO: . Yes. Okay. Realizing the shortcomings,-
1 10- NRC found the External Events Steering Group. The mission of l
.I 11 the Steering Group is to.make recommendation to NRC senior 'l I
12- management regarding what is the role of external' event O 13 related and accident policy statement, and to develop l
14 ' methodology for external event IPE, to integrate'all NRC- !
l 15 external event programs. There are quite a few of them, and !
16 to recommend neither research nor technical ~ assistance 17 program, i
18 Okay. What the what are external events? j l
19 (Slide) l i
20 MR. SHAO: Okay. These are external ;
i 21 events--earthquake, internal floods, external floods also. j 22 This is, there is confusion. Internal floods is also external ,
i 23 event. It is being considered by the generic, internal
() 24 generic letter, so I didn't put it on here. There was 25 confusion at one time, but I think it is clear now.
HERITAGE REPORTING CORPORATION -- (202)628-4888 ;
= _ _ _ _ _ - - _ - _ _ - - _ _ _ - l
t 14 l k 1L MR. MICHELSON: 'Well, is'this now your definition is L 2- ' supposed to' clarify.what'you mean'by an internal-versus 3 external events?
4' MR. SHAO: To be exact, I think. internal flood is-I 5 ' external event.
6 MR. MICHELSON: Well, I don't know. 'Then we are-7' being' exact. What by definition, by what policy.is it going 8 to be an internal event?
9 CHAIRMAN SIESS: I think it is a mistake, Carl, to 10 try to separate them.
.11 MR. MICHELSON: I do, too.
. ._ 12 CHAIRMAN SIESS: The IPEEE ccvers everything, and I O- 13 think what is confusing us is that the implementation of the-14 IPE has been assigned to two-different groups of people, and-
- 15. one group was'given what they would, presumably somebody 1 16 called internal events, and they have considered flooding as
'17- part of that, so this group didn't have to consider internal 18 floods. They just looked at external floods.
19 Now external floods really says if the water gets up 20 to a certain level, it is going to get in the plant, but what 21 it does after it gets in the plant? Whose job is that? And 22 now an internal flood, whether the water comes from the 23 Mississippi River--
()! 24 MR. SHAO: That's our job, this group. This generic H25 letter will cover the water level.
HERITAGE REPORTING CORPORATION -- (202)628-4888
! 1 L
15
,s J )
1 CHAIRMAN SIESS: External flood?
2 MR. SHAO: Will be considered by this generic 3 letter.
4 CHAIRMAN SIESS: Not just to the point when it I 5 reaches the sil of the door and gets a foot higher, but it 6 will be followed all the way through that door into the plant 7 to see what will happen? )
1 8 MR. SHAO: Right. Yes; all the way.
9 CHAIRMAN SIESS: And the internal floods, the same 10 thing? '
11 HR. SHAO: Pipe breaks, internal floods.
12 CHAIRMAN SIESS: Different source of water; it is a
- ~n
\- 13 very artificial subdivision, isn't it? They are different in 14 terms of the initiating event, but in terms of the 15 consequences, they are the same.
16 MR. SHAO: They might be the same ones you get into 17 the building.
18 MR. MICHELSON: How are you treating external fires?
19 I guess internal here means within the building.
20 MR. SHAO: Within the building.
21 MR. MICHELSON: How about one outside on the wall?
22 Is that an external fire?
23 MR. SHAO: Anything within the site, within the 24 site.
25 MR. MICHELSON: Outside the side the fence then; who HERITAGE REPORTING CORPORATION -- (202)628-4888
K 16 jfK b ' (f is treating-external fires since it--in some cases, there will-1 L.
12 .be forest right up t'o the site boundary and site boundary is 3 'very.close to the building. j
- 4. CHAIRMAN SIESS: That has never.been a design 5 . basis--
6 MR. SHAO: ' If you want to consider--
l 7 MR. MICHELSON: A lot of this has, some of this has l
8' -never been a design basis. ;
1 9 MR. SHAO: You want to really create a landslide, l 10 everything.
11 MR. MICHELSON: I am just trying to' find out. ,
12 MR. SHAO: External fire, we are not treating. If 1 13 you want to really treat external, I think there are 14 hundreds--them landslides, j 15 .MR. MICHELSON: I am trying to figure out what you I
16- are' thinking of.
17 MR. SHAO: Mainly internal fire within the plant.
18 CHAIRMAN SIESS: What is the basis for that 19 limitation, Larry, when the severe accident policy statement 20~ says look to the-plants for--each site plant?
l 21 MR, SHAO: If you loob at the last column, i
22 ' transportation accidents and others, suppose for some reason p 23 the plant is very close to forest. It will be covered by
( 24 forest. Suppose you build plant near air airport? That will 25 be covered by other.
!f 17-l^~ J:
1 1 'Mt. MICHELSON: Won't it be covered-by-. transportation
, fif there is'a. railroad in.that forest? Definition of others 3 'in here; it was not there at all.
4' .MR. SHAO: We intend to offer others. For instance, 5c many plants--
6 MR. MICHELSON: Are you going-to' revise your 7 document to include external fires and the others act in there 8' now?
9 MR. CHEN: Mr.-Michelson, I think we have--
10 CHAIRMAN SIESS: Your name?
11 MR. CHEN: My name-is John Chen with Research. The 12- Lother external event, all the external events we are 13 considering. So far we talk about internal events, okay. We
'14- cover internal event, and in addition to that' internal' event,
'15 we have internal flood, which has been covered by IPE, but
- 16. anything else that's considered as others, or external event, 17 have all been covered, but we just highlight those important !
18 things on these slides, some.others, other things,-for 19 instance, external fire, you talk about the forest fire. This 20 has been covered. Lightning, that has been covered. All been.
21; covered in our program. ]
s L 22 MR. SHAO: Example, Three Mile Island, airplane, th<
23 airplane' crash, that's covered. And that's covered in others.
) 24 MR. WYLIE: You say you covered lightning. Where?
'25 MR. CHEN: We cover lightning. It is like this. We HERITAGE REPORTING CORPOR7. TION -- (202)628-4888 )
18 W-
~ 1 '- .have .a NUREG' report,'CR'5042. In that' report, we assess what 2 is external. event as important, could have a potentia 1>
3 contributor to severe core damage. We assess lightning itself
.4 because"it is-a~very short. duration event. That's also been'-
5, covered by the Station Blackout.
6' Therefore, we feel in'the severe accident. situation, 7- we already cover that potential' problems, and the' potential 8 for the core damage is not as significant. Therefore, in our 9 final recommendation, this lightning would not be addressed.
10 CHAIRMAN SIESS: That's a generic conclusion.
.11- MR. CHEN: Yes.
121 CHAIRMAN SIESS:- I think the object of the game was 1 13 to look for plant unique' vulnerabilities.
14' Is there some plant out there at some site out there
.15 where lightning can do something that is an unusual 16 vulnerability?
17 MR. CHEN: For the lightning.
18 CHAIRMAN SIESS: Not generic; general we covered by 19- the regulations, the reg guides, the standard review plan.
20 Presumably what we are doing in the IPE is going looking for
.21 those things that might be an outlier.
22 MR. CHEN: I think you are right there, that point 23 of view, whether it is a potential outlier or potential
( 24 vulnerability, or whether based on our evaluation we feel that 25 lightning itself would not cause the problems provided the j HERITAGE REPORTING CORPORATION -- (202)628-4888 4
i l- 19
/~T r
Station Blackout is incorporated. Without a Station Blackout, 1
2 then we see there is a potential problems for the lightning.
3 But with that--
4 MR. WYLIE: Wait a minute. But Station, you are 5 saying lightning affects the power supply and the Station J
6 Blackout rule covers it? Is that what you are saying?
7 MR. CHEN: Yes.
8 MR. WYLIE: There are other things that lightning 9 does. If you look at the LERs that come through where plants 10 get knocked off the line. Lightning effects gets down into 11 the control systems of the plant, the instrumentation systems 12 of the plant. That's a vulnerability.
(
13 MR. CHEN: The vulnerability is whether you are 14 going to lead to core damage.
15 MR. WYLIE: If it ending up in the instrumentation 16 systems of the plant, you are getting pretty close to core 17 darage.
?b MR. CHEN: No. What you said, if you assume nothing 19 or you cannot resume the station power--
20 MR. WYLIE: I understand that, and the blackout rule 21 covers that. Yes, I understand that.
22 CHAIRMAN SIESS: Suppose it louses up some 23 instrumentation? Suppose it puts a surge through? Suppose it
() 24 does something to some solid state stuff? Suppose it fuses 25 some fuses or are there any, are we sure that there is no HERITAGE REPORTING CORPORATION -- (202)628-4888
'll 20 i i plant out there where lightning can have any effect other than 2 knocking out off-site power?
3 It seems to me that the IPE requires that either 4 from PRAs that have looked at vulnerabilities and found that j 5 certain plants have certain vulnerabilities to lightning and 6 that we eend out a list of does your plant have these? That's 7 what you are doing in other areas, but you know, you are 8 giving me a generic answer, and what we are supposed to be 9 looking at in the IPE are the non generic, the specific plant 10 unique vulnerabilities. How do we know that there aren't any 11 unless we--we might know, we might say in all the reactor 12 user's experience there has never been lightning that did 13 anything but knock out off-site power, which isn't true.
14 MR. CHEN: That essentially what, the study results 15 shown in the NUREG/CR 5042, that conclusion.
16 MR. SHAO: Bob Murray is here. Do you want to say 17 something? You did some study.
18 MR. MURRAY: Come up here?
19 MR. SHAO: Yes.
20 CHAIRMAN SIESS: Get anywhere there is a microphone.
21 MR. MURRAY: Bob Murray from Lawrence Livermore 22 Lab--our study that John Chen and Larry referred to, was say 23 supplement to, to NUREG 5042, and it went through--
() 24 CHAIRMAN SIESS: What was 5042?
25 MR. MURRAY: It was the effect of external events on 1
1 HERITAGE REPORTING CORPORATION -- (202)628-4888
't' 21 f,
[N?)
ap
.1 nuclear power plants. The first-report covered earthquakes, 2 orLnot--it' covered fires, floods, winds,'and transportation 3 accidents. Supplement 1 covered earthquakes, and. Supplement 2 4 covered what'is. called lother external events. One of these 5 was the effect of lightning, and the study examined the 6 effects from all the LERs, tabulated all the events for all L 7 the plants, and it came to the generic conclusion that it did 8 cause effect of power outage.
9 CHAIRMAN SIESS: And that--
10 MR. MICHELSON: That's not the question. Did it 11 cause more than that?
- 12 MR. MURRAY: From the LERs, the only thing that was L
O . 13 shown was the power.
14 MR. MICHELSON: I don't know which LERs you can 15 ' read. I have read them where they circulated right into 16 certain control systems in the plant. That was the last year 17 and a half, two years.
l .18 MR. WYLIE: This has occurred for over more than one 19 year where the effects of lightning have shown up in the 20 instrument and control systems now. I don't know whether you 21 saw those or not.
22 MR. BAGCHI: This is Goutam Bagchi, NRR staff. Can 23 I. interject one other thought? It is possible that we could I). 24I include a checklist of external events where the basis, 25 licensee would check off and say that this doesn't apply. It I. HERITAGE REPORTING CORPORATION -- (202)628-4888 e-
E-
! 22 .l b f ... -
11 'doesn't-have to be a su'bstantial effort, but simply go through 2 _ thel 11st and say thattit doesn't apply to that particular 1
3- -plant.
4 -CHAIRMAN SIESS: Isn't that the general approach of 5 the IPE?
6 MR. SHAO: Yes. We will eventually have, come out 7 with you have to address all external events that could be 8 vulnerability of plant, including stating why you.can exclude !
9 these things. I 10 NR. MICHELSON: Why would you say it doesn't apply?
l 11 .Is that because you don't have-the system, or because you.have
-12 done the analysis and found it is a non-problem?
O 13 MR. BAGCHI: hit is non-problem.
14 MR. MICHELSON: I. thought IPEEE was going to provide- j l
15 some kind of guidance and so'forth on how you do some of these 16 analyses.-
17 You are just going to provide a checklist of 18 problems to look at, or are you going to provide more than 19 that?
20 MR. BECKNER: I think probably the best thing to do j 21 is tell you what we did. It sounds like there may be some l
22 disagreement, but there is, in the. report that Bob Murray i 23 mentioned, there are large number of external events,
() 24 including extreme cold, and things like that. Of course, a 25 lot of these things affect off-site power which we have i
HERITAGE REPORTING CORPORATION -- (202)628-4888
I 23 l ,.
2 1 already covered meteor strikes, the whole gamut of things that 2 you can think of.
3 The attempt that was done at Livermore was to try to 4 get rid of some of these things so we would not have to 5 consider them. Meteorite strikes were eliminated. Severe l
l l
6 weather was eliminated because its primary effect was thought 7 to be the loss of off-site power, and the idea was to try to L 8 limit the scope of what would go out there by trying to check 9 things off and say hey, on a generic basis, obviously we don't 10 think that they are worthy of conducting examination. I guess 11 what I am hearing here is there may be some disagreement with 12 regard to the lightning, whether or not we screen.
l (~'
13 MR. MICHELSON: Severe weather, too, you have got to 14 look at. The one system for sure you need is some means to 15 contain the feedwater during blackout. Now that is providing 16 the feedwater is in any way affected by the event, severe 17 weather, be it lightning or whatever it would be, if it can be l
18 affected by that event. Also there is the possibility that 1
19 you will have both the blackout and the loss of the auxiliary, ]
i 1
20 in which case you are in deep trouble. )
l 21 MR. BECKNER: I guess we have got a comment in ACRS j 22 that you don't necessarily agree with our screening because 23 you go back--
() 24 MR. MICHELSON: You have got to look at the whole 25 problem.
""*^""*"""" c " ""^ * " -- <202>628-4888 j
24
's 1 CHAIRMAN SIESD: I think it is a process of 2 screening that. You mention severe weather. I recall two, 3 three years ago, we had an unusually severe winter. A number 4 of plants that didn't think they were in cold regions had all 5 sorts of things freezing up. they never lost off-site power, 6 but they lost certain water supply. I don't remember what 7 they were, but they were effects other than on off-site power 8 due to freezing.
9 MR. BECKNER: I think we have got a comment. Maybe 10 we should re-think that.
11 CHAIRMAN SIESS: Am I right that the general 12 approach to this thing is to try to give the licensee some
(-) 13 guidance as to the kinds of vulnerabilities that have been 14 found and the things they should Inok for?
15 MR. BECKNER: That is generally correct, and again 16 our initial work we did at Livermore was an attempt to try to 17 get rid of things. We didn't want the plate too full so that 18 they--you can't do everything, and I think, like I said, it 19 was a screening process, to try to get rid of things that 20 didn't appear to be important enough to warrants search, from 21 PRA insight again.
22 CHAIRMAN SIESS: I think there is a basic problem 23 with that approach in that by its nature that tends to be
() 24 generic.
25 MR. BECKNER: That's true.
HERITAGE REPORTING CORPORATION -- (202)628-4888
25
_j .
\- 1. CHAIRMAN SIESS: The whole IPE process is l 2 non-generic. It is looking for the user plant unique 3 vulnerabilities.
4 MR. BECKNER: I agree with you, but the other 5 alternative is not limit the scope of the effort. That's the 6 problem, to have no bounds on it.
7 CHAIRMAN SIESS: Suppose the effort was a PRA?
~8 Would the PRA take into account these things that we are 9 talking about--lightning?
10 MR. BECKNER: Some PRAs have, and some haven't.
11 MR. MICHELSON: Have you ever seen lightning in a 12 PRA? I haven't. I don't look at that many. Have you ever 13 seen lightning in a PRA? I have never heard of it. There are 14 a lot more important things that are being eliminated from the 15 PRA just to keep it simple enough to manage.
16 HR. BECKNER: It is a matter of trying to_get your 17 arms around this thing. That was the attempt.
18 CHAIRMAN SIESS: That's what bothers me, because I 19 thought your screening approach had as its basia a way to go 20 at this thing without a full PRA, go out and look for the 21 kinds of vulnerabilities that have been found.
l 22 Now we have talked about systems interactions, and 23 presumably the USI, but basically what bothers me is that it
() 24 seems to me detracting the attention of the licensees from 25 these plant unique things.
HERITAGE REPORTING CORPORATION -- (202)628-4888
26
(^T
\ 1 MR. SHAO: The general screening criteria for any_
2 event, let's say if the power is less than 10 to the minus 5, 3 because if that event screen out, okay--
4 MR. MICHELSON: How do you determine 10 to the minus 5 5 when you haven't included it in a PRA?
6 CHAIRMAN SIESS: That's an initiating event. That's 7 a frequency.
8 MR. SHAO: That's frequency of initittting event.
9 That's how people license, because if you license when you 10 design a plant, there are so many things you have to design 11 to, but for any event less than 1G to minus 5, people don't 12 design for it.
13 Airplane crash, the frequency is less than 10 to the 14 minus 5. The same way with others, so mainly it is very 15 difficult for this task to say address everything here, but we 16 as a general criteria say most events, you say anything is 17 less than 10 to the minus 5, we don't have to worry anything 18 more than that. We put airport near the location. Airplane 19 came, better consider this in your plan.
20 MR. MICHELSON: I think lightning is probably well 21 within <.he ground.
22 MR. SHAO: I feel I want to have somehow the scope 23 is managable because we can ask for the moon, and then you say
() 24 take about two, three, five memorandum to do the PRA analysis.
25 CHAIRMAN SIESS: There is a possible pitfall in I
! HERITAGE REPORTING CORPORATION -- (202)628-4888 L_-____-_
[.-
h1 s s 127 ,
'J
}l f \ 1 cutting off initiating events at some frequency.
2 MR. SHAO: 'That's what we were going do do..
3 CHAIRMAN SIESS: But the possible pitfall is that:we
, -4 have never done.a nargin' study very. broadly on anything but- l 4
5: seismic.-
'i 6 In other words, we cut something-off-at l'O.to the: I
- 1 7 minus 5, but if, probability of the consequences of the event 8: is just a little bit beyond the 10 to the minus 5 level is 9 catastrophic, that wasn't a good cutoff, see. If we design 10 the plant for 10 to the minus 5 flood, and the flood is one
'i
~
11 foot higher than that, it is going to cause us severe 12 -accident.
l kO 13 MR. SHAO: Then no good. We should go one step 14 further, 10 to'the minus 6. j 15 CHAIRMAN SIESS: We have done this on seismic. We !
i
- 16. worry about being on the edge of a cliff in seismic, and we l 17 did the seismic margin studies, and we are pretty well !
16 satisfied that generically we have got margins, although ,
R 19 obviously some. plants have specific plant unique 20 vulnerabilities at that seismic level. This has been one
. 21- characteristic of seismic PRAs. You always find something l 1
)
22 that is a vulnerability and it is usually easily fixed, but we !
i l
23 - have never done that kind of study for some of the other j
'( ) 24 things, like flood we have, but not--I think if we went back 25 to the SEP, where we found several plants that were vulnerable
-HERITAGE REPORT.TNG CORPORATION -- (202)628-4888
__ 1____________. _-
28 11 to floods, plant unique vulnerabilities, Nine Mile Point is an 2 example, as.I recall, some of them were easily fixed. Some 3 'weren't, but that was a vulnerability. Not all plants, not 4L all of the SEP plants were vulnerable plants,.so it was a 5 plant--
6 MR :. BECKNER: The examination of the plant.
7 MR.:SHAO: We are go_ag to look at it. So our-6 general. approach is we--mainly it is based on PRA results and 9 maximum external events or for other external events, we have 10 general. guidelines like the frequency of initiating events.
11- _For_anything greater than initiating events, you have to 12 consider,~but we cannot come out with guideline for every 13 external event. There are so many of them, 30, 40 of them.
l 14 CHAIRMAN SIESS: You don't, but we have got I guess 15 two' concerns. One is what guidance you are going to provide f 16 people. Surely your job and our job is to look at it.
L 17 The other concern is what people are going to do.
18 The object of the game is to find out if there are any plant
-19 . unique vulnerabilities that make that probability of core melt
.20 higher than we thought it was, and that's something that every 21 utility ought to be more concerned about than we are because 22 it is their plant and their investmert, so whether or-not you 23 give them guidance, the question I guess is going to be what
() 24 they look for in spite of what you tell them.
25 MR. SHAO: In the past, according to my experience,
__________________________I- 1
l 1
29
.l" '
l k
- 1. except lightning, in the transportation area, like, for 2 example, explosion at Three Mile Island, airplane crash, !
3 Turkey Point, airplane crash, these are unique to the p.. ant 4 because they are initiating events higher than--pretty high, 5 and be surprising lots of plants have specific design, 6 specific external events built into the design, and I 7 think--let's go through this.
8 These are good comments. That's the reason we come 9 here, and these are lots of good comments, but you also should 10 see on our side because external events, they are so numerous 11 you cannot tackle. You have landslide I can think of 30 12 external events tilat can hit the plant.
13 CHAIRMAN SIESS: You missed my point. I think your 14 conclusion is that there are so many possible others that you 15 can't think of them all, which is probably a valid conclusion, 16 and I'm not sure it is ycor job to think of them all, but I 17 think we are going to want some assurance that when the 18 licensees do the IPEEEs, or the IPEs for that matter, that f
19 they think of some of these things.
20 MR. SHAO: That's good comment, and we will think 21 about it.
22 CHAIRMAN SIESS: We can't tell them exactly what to 23 do, and we have set up--now they are supposed to bring you a
() 24 record and you will decide whether they have done a good job.
25 MR. SHAO: Bill, did you hear that comment?
HERITAGE REPORTING CORPORATION -- (202)628-4238
,_ _ - _ _ - _ ~
L, 30 i 1~ MR.'WYLIE: Larry, just.one more comment on the u 27 subject of lightning--it is the effects of. lightning that's 3 the concern, not the lightning strike itself.
4 MR. SHAO: ~The effects of lightning.
5 MR. WYI,IE : -And'also it is plant-specific.in that 6 .the plants are designed different. They are different ground
.7 systems. Some are inadequate. Some are-better than others, 8 and there is also differences in the way the buildings-are 9 constructed and the shields that is provided to' protect 10 against the effects of lightning, so it is plant-specific.
11 MR. SHAO: Okay. Lightning, definitely we talk 12 about.it. Any other external event you people have.in mind, L
13' maybe you think we should give little more emphasize?
l .i 14 MR. WYLIE: We have got a whole file. I will be 15- glad ~to make it available to you.
16 MR. SHAO:- Okay.
.17 MR. MICHELSON: Maybe it will be useful if you tell
~
18 us the kind of thing you have already eliminated. I think you 19- mentioned a couple of times you are not--I have no problem
- 20. with that. What other kind of things aren't you considering 21 or you already, or you eliminated generally?
122 MR. BECKNER: Bob, did you want to try to address 23 .that?
) 24 MR. MURRAY: Bob Murray--volcano except for the 25 Trojan site; sinomi, except for the California coast sites.
HERITAGE REPORTING CORPORATION -- (202)628-4888
. = _ _ _ _ - - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _
h 31 q
^ 1 Diablo Canyon--those two have been eliminated generically.
I i
2 The guidelines are still under development, of 3 course, and it, the purpose was to look and find the ones that 4 were the--to concentrate on the first, but it didn't say that 5 the, it could ignore anything else, and transportation, for 6 example, they were still to look at highway, rail, barge, 7 aircraft, and pipeline. It may be for a particular plant, 8 that it wasn't a concern. It was nothing there.
9 Our worry was that during the original design, 10 certain transportation modes may have been used and they may 11 have changed over time. New highway may have been put in or 12 larger transportation quantities may be there, so the utility 13 would be asked to examine and see if it is important, and if 14 not, we could throw it out. If it was important, then we 15 would have to do something to examine his plant.
16 MR. MICHELSON: Of course you could keep the volcano
'17 in, too. You write it off immediately, so why do you 18 eliminate it generally from the discussion, or do you? Or 19 maybe I misunderstood. I thought you are going to eliminate 20 the--generically you weren't even going to look at them on an
-21 individual utility basis.
22 MR. BECKNER: I think that's right.
23 MR. MICHELSON: Why do you eliminate any of these?
() 24 It takes about one minute to decide if volcanos are a problem.
25 MR. BECKNER: I think it is just, again is a matter HERITAGE REPORTING CORPORATION -- (202)628-4888
~o- -1. of trying to bound the thing, that we would have difficulty
- 2. putting out'a request to examine for everything.
l 3 MR. MICHELSON: You eliminated aircraft also?
4 MR. BECKNER: No.
5 MR. MURRAY: Aircraft is considered transportation. .g i
6 MR. MICHELSON: Meteorite, volcano, what else?
7 MR. BECKNER: Severe weather--I am not sure if~there 8 is any more; sand storms.
9 MR. SHAO: Landslides. l l
10 MR. MURRAY: Would be examined uniquely.
11 MR. BECKNER: Sleet, freezing rain, severe cold.
12 MR.'MICHELSON: You eliminated that generically?
b 13 MR. BECKNER: Except for impact on off-site power, l
I 14 yes. j 15 MR. MICHELSON: You think that's a safe 16 justification?- There is plenty of experience with severe 17 weather affecting on plants themselves. All have been minor
'l 18 so far admittedly. They are not--certainly exist.
19 MR. BECKNER: That was done based on review of 20 . existing PRA and'I believe on LERs. I 21 MR. MICHELSON: PRAs don't have severe weather in 22 them necessarily. j 23 Have you seen a severe cold PRA?
"( ) 24 MR. BECKNER: No. l 25 MR. MICHELSON: Effects of severe cold are included HERITAGE REPORTING CORPORATION -- (202)628-4888 ]
L. 33 i
t
^s 1 in the model. I
)
2 MR. BECKNER: I think we have also performed LER 3 rearches. !
4 MR. MICHELSON: You have got quite a few severe -l 5 weather LERs, cold weather particularly. About three, four j 6 years ago, we had such a cold winter I was reading a number of 7 them.
i 8 HR. BECKNER: Like I said, we have examined the 9 LERs, and we could go back and make sure that we have done i
10 that. )
11 CHAIRMAN SIESS: How do you that? Pick the LERs on 12 freezing pipes? And that's what happened a few years ago.
13 When you eliminate that--let's say there were 20 LERc from 14 14 different plants that it was cold enough to do something that 15 caused a problem.
16 How can you say, what basis would you have for 17- saying that if it got 10 degrees colder, that some plants 50 18 miles further south with a different design wouldn't have <ad
'19 a problem?
20 MR. BECKNER: I don't think we can get into that 21 detail.. Obviously we don't get that information.
22 CHAIRMAN SIESS: Then isn't this what the plant 23 owner should be looking at? Shouldn't they be looking at
()
' 24 their plant to see if they have got a plant unique 25 vulnerability, that to subfreezing temperatures at Turkey HERITAGE REPORTING CORPORATION -- (202)628-4888
34 0 1- -Point? Maybe that is such probability I can move it up to 2 Hatch.. You know, maybe that type of plant, vintage plant has 3 never had a subfreezing temperature, so I don't see how we can-4 eliminate these things generically based on the historical LER 5 record when we are'looking for outliers.
6 MR. BECKNER: I think it is completeness versus can 7 you get your hands around 4t?
8 MR. SHAO: Maybe t.tese are good comments. We are 9 going to maybe, we try to include, maybe let's think about 10 this.
11 MR. BECKNER: We can gSre you a report, Carl.
12' CHAIRMAN SIESS: I am not asking you to eliminate or 13 not eliminate the cold weather. What I am saying is that the 14 utilities should be told look for vulnerabilities to 15 subfreezing wer.ther, and if Turkey Point comes back and cays 16 chances are one in ten million it will ever be subfreezing 350 17 miles south of Miami, that may be an acceptable answer, but 18 somebody else may come back and say we have looked and we
/
19 don't see anything can happen, but they are supposed to look.
20 It is not your job to tell them not to look. You 21 should help them.
- 22 MR. BECKNER: We are not telling anybody not to 23 look. What we are doing is we are not just, we are not
() 24 developing methods or telling people specifically to go out j 25 and look for cold weather.
t ly HERITAGE REPORTING CORPORATION -- (202)628-4888 L
35
.f" k- 1 MR. SHAO: I think the Bill's comment, we are not 2 developing methodology for every external event. We are 3 developing methodology to major external events. For'other 4 event, we can put in general sentences.
5 CHAIRMAN SIESS: You can, where you can develop a 6 methodology that you will accept as meeting the requirements 7 of the severe accident policy statement, fine, but where you 8 can't develop a methodology or where the 3uv'~ L111Ly is so low 9 that you don't want to spend the effort, I don't see anything 10 wrong telling the licensee that this is something you ought to 11 look at.
12 MR. SHAO: We agree with you.
13 CHAIRMAN SIESS: But don't, for goc.Aess sakes, 14 don't tell him don't bother.
15 MR. SHAO: Okay. We agree with the comment. Until 16 we have really facts that can be screened out something, 17 otherwise they should be included.
18 CHAIRMAN SIESS: This assumes the staff is going to 1
19 accept your comments in its reply. l 20 MR. SHAO: Right. Okay, i 21 DR. SHEWMON: And the ACRS, and that the ACRS will 22 accept the common sense reply. !
23 CHAIRMAN SIESS: The odds of that--
() 24 MR. SHAO: Okay. Any more questions on this slide?
25 Okay. This is the composition of the Steering Group. I'm the HERITAGE REPORTING CORPORATION -- (202)628-4888
7 36 l- l L
l' Chairman. .The_ member ~is T. Novak, Wayne-Houston, and Bakher 2 having transferred to NMSS; Jim Richardson, and Goutam Bagchi, 3- Executive Secretary, and under the Steering Group there is 4 three_ subcommittees. The Seismic Subcommittee is chaired by 5 Leon Reiter and Andy Murphy. The Fire Committee is-chaired by EL Conrad McCracken, and the High Wind, Flood and others, L 7 Subcommittee are chaired by D. Jeng and Bill Beckner.
1 H
8 (Slide) 9 MR.- MICHELSON: Just for_ clarification, I think-you 10 said a little earlier that the internal flooding was a part.of 11 the basic examination of systems although it was an external 12 event.
13 Which of these committees or subcommittees handle 14- the internal flood?
15 CHAIRMAN SIESS: These are all external committees?
16 MR. MICHELSON: Your--by definition, you are 17 excluding the internal flood because of the other--
18 MR. BECKNER: Internal flooding--
19 CHAIRMAN SIESS: Everything else.
20 MR. BECKNER: Internal flooding was included as part 21 of the original core method. It was in this core methodology 22 so that that went on a long time before the Sur, committee got 23 together.
() 24 MR. MICHELSON: There is a whole lot of stuff that 25 is not included in their guidance system interaction.
n I
HERITAGE REPORTING CORPORATION -- (202)628-4888
37
' - 1 MR. BECKNER: We have included the insights from 2 A-17, in the internal events IPE.
3 MR. MICHELSON: Who is doing that? You say "we."
4 MR. BECKNER: That has been done. That's done under 5 my branch, and in cooperation with some of the people.
6 MR. MICHELSON: That guide has already been issued?
7 MR. BECKNER: That has been issued, NUREG 1335.
8 CHAIRMAN SIESS: I think we can call this the IPEEE 9 would be the independent plant examination for everything 10 else!
11 MR. MICHELSON: Now were the original--do you think 12 that guidance is adequate?
13 MR. BECKNER: What I referenced was the IPE guidance l
14 document which makes further reference to ssme information the 15 A-17 people gave you. I don't have those references off the 16 top of my head.
17 MR. SHAO: That's a good onel 18 MR. MICHELSON: Is there another subcommittee of 19 ACRS looking at 1335 now?
l l 20 CHAIRMAN SIESS: I haven't the slightest idea.
21 MR. MICHELSON: The basis of current knowledge, that I
i 22 is still a reasonable guidance? !
23 CHAIRMAN SIESS: I don't even know ahat 1335 is, but
() 24 the Severe Accident Subcommittee is supposed to be looking at 25 the IPE generic letter that we haven't gotten yet. I got a 2
( HERITAGE REPORTING CORPORATION -- (202)628-4888 o___ _ _ _ _ _ _ _ _ - _ _ _ - - - -
38 I")
\ 1 note last week it is out and they are going to supply us with 2 copies as soon as we get it.
3 MR. MICHELSON: That generic letter includes the 4 analysis of tne systems themselves and the IPE?
5 MR. BPCKNER: For internal flooding?
6 MR. MICHELSON: Yes. j 7 HR. BECKNER: Yes.
8 MR. MICHELSON: We need to look at that document.
i 9 CHAIRMAN SIESS: Yep. l I
10 MR. MICHELSON: Because it overlaps the external.
11 Once the external flood gets in the door, I guess the internal l 12 flood people take over or something. I don't know, but I
('T
\~ 13 would like to see that guidance.
14 CHAIRMAN SIESS: Don't change the color of water as 15 soon as it gets'inside the door. l 16 MR. MICHELSON: Muddy water is probably just as bad. ;
l 17 Could be worse.
18 MR. SHAO: Okay. Industry has found similar groups 19 to address the external events. The industry effort is led by 20 NUMARC. There are two working groups. The first one is i 21 called Seismic Issues Working Group, and the other one is 22 Severe Accident Working Group.
23 The Seismic Issues Working Group is headed by Bill
() 24 Lindblad. Bill Lindblad is the president of Portland General 25 Electric Company.
HERITAGE REPORTING CORPORATION -- (202)628-4888 :
j
39 p
4
\ The Seismic Issues Working Group is responsible for 1
2 all seismic issues, not only the IPEEE, but all seismic 3 issues.
4 The Severe Accident Working Group is chaired by 5 Cordell Reed. I think maybe you people are familiar with that 6 group. The Severe Accident Working Group is responsible for 7 accident management, and all other external events other than 8 seismic.
9 (Slide) 10 MR. SHAO: What are some of the key issues related 11 to external events? We have discussed for a long time there 12 are many, many possible sources of external events, so we have 13 to get our arms around it.
14 For these external events, there are large 15 uncertainty of frequency. Here we are talking 10 to the minus 16 5, 10 to the minus 6--in many cases, just not enough data.
17- Also the plants are designed to various criteria and i
18 the extent of protection for the external events beyond the l
19 design basis are really unknown for many plants. How much 20 margin do they have?
21 And also the past PRAs indicate risks due to 22 external events can be high, especially seismic, fire, and 23 hurricane for certain sites.
() 24 Aside from PRA, there are no other methodologies f or 25 evaluating external events beyond design basis in the seismic N_______-_____-___-__ __ _ _ _ _. ._
r 40
~ (3 I area. There are many, many programs.
2 CHAIRMAN SIESS: That's not true, Larry. That's not 3 true. It may be true for non-seismic, but you have got the 4 seismic margin approach for seismic. It is not a PRA.
5 MR. SHAO: We do have approach, but we have EPRI 6 approach. We have NRC approach. We are looking at, we are 7 looking at--
8 CHA1 AMAN SIESS: We have got, we have approaches for 9 evaluating seismic beyond the design basis.
10 MR. SHAO: Yes, wo do.
11 CHAIRMAN SIESS: We have had it for a long time.
12 MR. SHAO: But we try to tie it down right now.
13 CHAIRMAN SIESS: You haven't formalized it, agreed 14 on somet)ing.
15' MR. SHAO: EPRI use their approach.
16 CHAIRMAN SIES6: We are a lot farther along on 17 seismic than we are on anything else.
18 MR. SHAO: Agreed. In seismic margin, we have 19 worked on Catawba, Yankee. Right now we are working on Hatch 20 right now, the Hatch.
21 CHAIRMAN SIESS: We have been working on seismic now 22 for 20 years or some of these other things you started last !
l' 23 month, fou know.
() 24 MR. SHAO: You are right. l 25 MR. MICHELSON: I have a little problem with, you HERITAGE REPORTING CORPORATION -- (202)628-4888
41 Lg 1 know, as soon as we get.into this a ways, we start talking L
i 2 beyond design basis, and events beyond the design basis. The 1 3 -inferenceLis we haveLalready handled the events within the 4- . design basis.
5 I thought part of the IPE process, though, was going 6 to be to go back and see if we really have handled events that 7 are within our dnsign basis already but which we hadn't 8 -analyzed in-an adequate proper depth.
9 For instance, fire is within our design basis 10- already I think. Fires of certain durations at least are.
11 Like a one-hour fire or three-hour fire, but we haven't 12 necessarily examined what happens when we get a design basis
~13 fire such as the effect of heat smoke migrating and initiating 14 other fire protection equipment, things of this; sort, so I 15 thought the IPE process would go back and pick up those kinds 16 of things to see where we are at because those haven't been 17- -done, are a not in a PRA presently. PRA doesn't know how to 18- calculate--
19 MR. SHAO: I agree.
20 MR. MICHELSON: Thermal and smoke spectrum around 21 the building and things of this sort; you can't look to PRA 22 for this.
23 Is IPE going to look at this?
I ) 24 MR. SHAO: Yes.
25 MR. MICHELSON: So it isn't just beyond the design HERITAGE REPORTING CORPORATION -- (202)628-4888
42
_rg
( '1 1- basis we are going to look? We are going to look at things 2 that were within the design basis but had been inadequately 3 analyzed in the past?
4 MR. SEAO: Agreed.
5 MR. MICHELSON: Isn't that correct? That was the 6 impression I had of the IPE process, but you shift gears.
7 Right away you start talking about beyond the design basis. I 8 wasn't sure if the other part got lost or what.
9 CHAIRMAN SIESS: I think a lot of what the staff has 10 done in here has been influenced by the history of the 11 approach on seismic, because in seismic, there was hardly 12 anything that hadn't been analyzed. I mean there is one thing A
k- 13 we did. We analyzed plants for seismic. Whether it was right 14 or wrong or we knew how to analyze them, boy, did we.
15 When you get into other areas, it is not quite the 16 same thing, as Carl points out. There are things that are 17 within the design basis that either weren't analyzed properly 18 or considered properly, or the scope was too narrow, but a lot 19 of the presentation of the approach seems to be along the 20 seismic lines where it is beyond the design basis, the margin 21 thing, and IPE is not just margins. There is probably one 22 thing that worries you most in this business. Is there an !
23 initiator out there that can be a major contributor to severe ,
() 24 accident that we have overlooked?
25 MR. SHAO: That was not in the original design l
HERITAGE REPORTING CORPORATION -- (202)628-4888 )
l
j .3 43 1 basis. 'The design is wrong, but it was not in the original q: 2 design basis.
- 3. CHAIEMAN SIESS: The design basis--
~4 MR. SHAO: Was wrong.
5 ' CHAIRMAN SIESS: It wasn't complete. .The design 6 basis for Trojan didn't include a volcano. It turned ~out it L7 was all right.
l 8 MR. SHAO: Okay. I change the wording.
9 CHAIRMAN SIESS: You know, what we worry about 10- completely in PRAs, it's missing an initiator, but that's also' I
11 what a designer worries about all the time, and you're a l i
5
-12 structural engineer. I'm a structural engineer. Did we miss a ]
O. 13 load? -Did we. overlook some loads? We have got this thing i
14 -designed for 14 different. loading conditions with factors of. j 15 search of two and a half, but if there is a 15th loading 16 condition we haven't considered, forget about it. You know, 17' and that's the big problem. It is not the only one, but if we 18 run into trouble, it is going to'be overlooking an initiator, 19: not overlooking the consequence I think.
20 MR. MICHELSON: I'm not sure I agree with that, 21 Chet. Even in the case of seismic, what we are overlooking so 22 far is the vulnerability.of small pipes through interfacing ,
23 from equipment which is not seismically qualified and'which
() 24 wasn't analyzed for the effect of those impacts in terms of 25 the feedwater or whatever. We went all the way through this i
l HERITAGE REPORTING CORPORATION -- (202)628-4888
44 ,
- ' 1- on A-46, and we simply haven't done it. l 2 CHAIRMAN SIESS: We thought about it.
3 MR. MICHELSON: We thought about it.
4 CHAIRMAN SIESS: We may not have done it. It must 5 be puts in seismic walk-downs, seismic margin studies, and 6 those things.
7 MR. MICHELSON: It was not put in the walk-down.
8 The walk-down said look for physical impact. Don't look for 9 the release of the water from the tank, for instance, that 10 flows-over. That was the classic one I gave the Committee on 11 several occasions. Even the Committee didn't support it.
12 CHAIRMAN SIESS: I would have to look back at the 13 Indian Point seismic interaction study to see whether they 14 really did include that.
15 MR. MICHELSON: Hopefully utilities are smart enough 16 to look at what happens when a tank falls over and turns over 17 its contents and not just on its physical impact, but the rule 18 on A-46 was physical impact only.
19 MR. BAGCHI: I respectfully submit you are ignoring 20 a lot of work that has been done by the seismic qualification 21 utility group, by the research staff, and the margin on 22 non-category 1 piping is substantial.
23 MR. MICHELSON: Read your A-46 record, and I think
() 24 you will understand what I am talking about.
25 CHAIRMAN SIESS: A-46 is only one part of the story.
HERITAGE REPORTING CORPORATION -- (202)628-4888
45 4
c.
k~ 1 MR. MICHELSON: That was the final resolution. j i
2 CHAIRMAN SIESS: A-46 was not the final resolution 3 of seismic margin or seismic vulnerabilities, which has been 4 attacked--
5 MR. BAGCHI: I am not convinced there is anything 6 that additional needs to be done. I understand that there may 7 be some plant-specific unique outliers of vulnerabilities in 8 terms of interaction with safety systems and other things that 9 impact in this case, but other than that, I think substantial 10 work has been done, and I am convinced it--
11 MR. MICHELSON: I didn't say substantial work wasn't 12 done. I am just saying this is an area that has yet to be t
13 examined from the seismic viewpoint.
14 CHAIRMAN SIESS: I am not sure that's right, Carl.
15 It may be true that A-46 did not go that far, but the seismic 16 margin studies and other things that aren't a part of A-46--
17 MR. MICHELSON: Did they look at 1 inch pipe, for 18 instance? Two inch pipe? They looked at 6 inches or larger.
19 CHAIRMAN SIESS: I don't know who they are. We have 20 only done a couple of seismic margin reviews so far. Just ,
21 what they looked for in those, but the point is that the 22 seismic margin, the seismic vulnerability, is somewhat broader 23 than A-46. There may still be gaps, but it is
() 24 compartmentalization. A-46 is one part o' t. A-40 is 25 something else. IPEEE is somet hing else. We have too much HERITAGE REPORTING CORPORATION -- (202)628-4888
46 1 -
'D 3 11 stuffidivided up here, the names attached to it, m<
2' MR.LCHOKSHI: This is Nilesh Chokshi on the seismic 3 margin program--
~
4 CHAIRMAN SIESS:' Get a.little closer tx) the mike.
5 MR. CHOKSHI: You will hear more about how these 6 small pipe breaks are handled. In the margin review we have 7 small1LOCA and look.at the mitigating systems.
8 MR. MICHELSON: We are not looking at LOCAs. This' 9 is'outside of containment.
p 10. MR. CHOKSHIt I think thu way it is handled, one 11- part is the walk-down. 'One--rigorous, and you are looking at 11 2 a lot of other things than A-46.
13 MR. MICHELSON: Yes. I am only addressing A-46, my 14' comment. What originally you said, you haven't done a
15, everything yet on seismic.
16 For instance, in A-46, we create a long debate.
- 1. This I know for'a fact. We haven't considered the contents of 18 tanks when they fell over in A-46. Whether you did it 19 somewhere else for in seismic tanks, I don't know.
20 CHAIRMAN SIESS: A-46 is going to be reviewed by 21 Charlie's subcommittee when? We can get a meeting scheduled.
~22 And'I think, I hope that this IPEEE thing is somewhat broader 23 than individual, you know, I know that the severe accident
()L 24 policy statement says that unresolved safety issues will all 25 be addressed, but I don't think you are accepting A-46 HERITAGE REPORTING CORPORATION -- (202)628-4888
L 47 4 4
resolution as handling seismic IPEEE. Am I right?
1 2 MR. SHAO: No.
3 CHAIRMAN SIESS: If somebody says we have addressed 4 A-46, that's not an answer.
5 MR. SHAO: The A-46 only covered the equipment part 6 of it. We covered much broader.
7 MR. MICHELSON: For the parts covered by A-46, do 8 you consider the A-46 walk-down complete now and IPE doesn't 9 have to pick up any more?
10 HR. SHAO: IPE, we still have to do the IPE. We 11 take part of the credit for A-46, but not all the credit.
12 MR. MICHELSON: For instance, this time through you 13 will look at water coming out of the tank, a non-seismic tank 14 falls over and doesn't impact the vital equipment, but the 15 water does impact vital equipment. You will look at that I 16 under IPE even though this might not have looked at it under 17 A-46 because it wasn't a part?
18 MR. SHAO: Exactly.
19 MR. MICHELSON: Then I will watch for it in the 20 guidance. That's a good answer. Should have gotten it to 21 begin with.
22 MR. SHAO: We will cover seismic margin later.
23 CHAIRMAN SIESS: I think we might be putting too
'( ) 24 much emphasis on the guidance. I hope licensees do more than 25 the staff tells them to do.
HERITAGE REPORTING CORPORATION -- (202)628-4888
l-48
. ,m
' i '"'J - 1 MR. MICHELSON: We have to assume they do only what 2 the staff tells them to do.
3 CHAIRMAN SIESS: I won't assume that because that 4 would be assuming once the staff issues the generic letter we 5 all say--
6 MR. SHAO: I am supposedly talking for ten minutes.
7 CHAIRMAN SIESS: I don't think anybody is going to 8 be satisfied that we have met the requirements of the severe 9 accident policy statement for the results of the IPE was also 10 come in and somebody looked at them, issuing the generic 11 letter 12 a piece of paprir. May define a lot more paper, but 12 okay.
\- 13- MR. SHAO: Okay. I won't go very fast. I am 14 supposed to talk only ten minutes. Now I am here about an 15 hour.
16 CHAIRMAN SIESS: That's all right.
17 MR. SHAO: The main objective of a severe accident 18 program is to identify vulnerability, but in order to identify 19 vulnerability, it is very important to know how the plant was 20 designed to. Let's say for seismic loading, depending on the 21 vintage, the plant was designed to various criteria. For 22 instance, before 1962, no seismic loading were considered in 23 plant design.
() 24 MR. WYLIE: Are you speaking of license application?
25 MR. SHAO: License applications; they apply for HERITAGE REPORTING CORPORATION -- (202)628-4888
49
(
,-,)
- 1 operating license. From the 1963 to 1969, uniform building 2, codes or equivalent static method of 1150 was used for design.
3 From 1970 to 1975, so-called Housner response, small damping i
4 values were used in the design, and the criteria for it 5 varies.
6 For 1975 to current, we call current criteria, we 7 use Reg Guide 1.60 and Reg Guide 1.61, and all the equipment, 8 electrical equipment, is qualified in accordance with IPEEE 9 344, so-called current correct criteria.
10 Seismic re-evaluation effort, earliest plants was 11 under evaluation program. Also we did a couple of seismic 12 margin methodologies on Main Yankee, Catawba, and we are 13 applying at Hatch.
14 CHAIRMAN SIESS: Larry, what plant do you think 15 vulnerabilities, vulnerabilities--as I recall, when we looked 16- at Big Rock, which wasn't designed for any, anything, they had 17 one pipe support in the whole plant. They went back and I 18 think they were going to spend up to a million dollars looking 19 at seismic vulnerabilities and fixing them, and I don't know 20 that I have seen the report on just what they did.
21 I am wondering how many vulnerabilities they fo'tnd 22 in that plant say as compared to Indian Point with a building 23 which was designed seismically. Buildings would bang
() 24 together. The ceiling was going to fall down and kill 25 everybody in the control room. Interesting scenario--Big Rock HERITAGE REPORTING CORPORATION -- (202)628-4888
50 L i
/ ') i
^/ 1 gct much worse than Indian Point, not having been designed--
- ]
2 MR. SHAO: I think people spend more effort in i
3 Indian Point. l 4 CHAIRMAN SIESS: Did they? I don't think they did a 5 PRA on Big Rock, did they?
6 MR. REITER: Yes, they did.
7 CHAIRMAN SIESS: What was the seismic contribution 8 to risk?
9 MR. REITER: I don't remember.
10 CHAIRMAN SIESS: Couldn't have been more than 11 dominaat.
12 MR. SHAO: Anybody know about PRA on Big Rock?
O V 13 CHAIRMAN LIESS: What else did we have? 362.
14 MR. BAGCHI: Yankee-Rowe.
15 CHAIRMAN SIESS: How did it come out in the SEP? I 16 know we were worried about dams failing in earthquakes and 17 stuff. They did make modifications to the plant. I'm not 18 sure exactly what they did.
19 Would they have been called vulnerabilities or just 20 sort of a general, you know, things that when they were 21 fixed--I would say vulnerability is something that when you 22 fix it, the severe accident--
23 MR. SHAO: Anybody want to answer.
() 24 MR. REITER: I don't want to go into 25 vulnerabilities, but again the idea of the SEP was what we HERITAGE REPORTING CORPORATION -- (202)628-4888
51 m
1 considered were all these design bases were too old, wanted to 2 look at them. We came up with quote, what we call, might call 3 surrogate design basis. The plants had to demonstrate using 4 realistic modern criteria they could meet the design basis, 5' and I think one of our projects at Livermore was to collect a 6 list of all the fixes that we had, SEP, have that information.
7 I don't know if we have it with us--i order to make sure to 8 learn from that experience.
9 CHAIRMAN SIESS: What did we learn besides 10 vulnerabilities of the buildings that weren't seismically 11 designed?
12 MR. REITER: I don't know know if we ever came out 13 with generic lessons from that.
14 CHAIRMAN SIESS: Even plant specific, I can't think 15 of any.
16 MR. REITER: We had a copy.
17 CHAIRMAN SIESS: The point was there might have been 18 plant-specific vulnerabilities at Big Rock. There were also 19 plant-specific vulnerabilities at Zion and Indian Point, and 20 what is the one we did where the staff, we did the seismic 21 margins on? You have got a tank--you know, and I am looking 22 at Larry's list up there, and the last two items, difference !
i 23 between Housner and Reg Guide 1.60 response spectrum is a j
() 24 pretty small increment compared to the difference in the first 25 two items on, or the first one and the third one.
HERITAGE REPORTING CORPORATION -- (202)628-4888 !
_2_______--__--___--_---_--__ ;
52
.p. ;/
'~
1 The equipment qualification differences between 75, 2 present and after, might be more important. I don't know, 3 although they still didn't anchor a lot of equipment.
4 MR. SHAO: One, 2, 3 is covered by 146, 5 MR. REITER: The point, this line, I think it makes )
6 a good point is that the family plant out there is extremely 7 diverse in terms of what they started out and what has been !
8- done to them and that we have to come up with criteria which 9 deal with diverse families of plants, seismic design plants 10 also which have, high seismic design plants have been looked 11~ at in the context of other programs, and I think that's the 12 point of the slide.
I 13 CHAIRMAN SIESS: It is a large variety. That's not 14 your job to look at them. It is the utilities' job to look at 15 them.
16 MR. REITER: It is our job to take that into account 17 and come up with criteria.
18 CHAIRMAN SIESS: I guess so. I'm not sure why, but 19 you know, to me what is important is not the diversity of the 20 design basis for the plant, but the differences in the 21 capacities of a plant, and just on a rather crude empirical 22 basis, I haven't seen a lot of evidence that the old plants 23 had any more susceptibility to plant unique vulnerabilities I ) 24 than some of the more recent plants.
25 MR. BAGCHI: I think Dr. Siess' point is well taken.
HERITAGE REPORTING CORPORATION -- (202)628-4888
53
,e
"' 1 It is the capacity that, the resistance of the p'. ant that's of l
2 concern here, and based on the Main Yankee margin study, we 3 notice that although it was designed for .10G at least for the 4 safety systems, the required systems, we found value to be up 5 around .21 G.
G That doesn't mean to say that some modifications 7 weren't made. They were made, but relatively inexpensive.
8 And with that--with that, even with the .10G design, it came 9 up with the value of .21, and there is a message there.
10 CHAIRMAN SIESS: I think and I would say that the 11 plant designed to equivalent static is just as likely to have 12 the equipment anchored as one designed to Reg Guide 1.6. If 13 you go get a force, whether you anchor the equipment, if you 14 don't anchor it, it is a design error. Most equipment that 15 isn't anchored really isn't anchored. It is not that it is 16 inadequately anchored. It is the bolts aren't there.
17 MR. SHAO: That's why we're going to have walk-down.
18 CHAIRMAN SIESS: To get seismic robustness or i
19 sturdiness into a plant, the difference between your second 20 item and your last item is not all that big.
21 MR. CHOKSHI: Generally the plants are similar 22 plants, similar to what you, you know, look for the things in 23 our new plants.
() 24 CHAIRMAN SIESS: We have a lot of block walls in 25 seismically designed plants. They aren't any better than II HERITAGE REPORTING CORPORATION -- (202)628-4888
_ _ _ . - _ _ _ _ _ _ _ - - - _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . )
F l 54
,;D
J
'- 1 plants not seismically designed.
2 MR. MICHELSON: How did you treat the 3 non-seismically designed equipment when doing these analyses 4 for seismic margins? Did you look at that equally to all the 5 safety-related equipment?
6 MR. CHOKSHI: If it is within the event tree or 7 fault tree--
8 MR. MICHELSON: 'How about if it gets into the event 9 tree since they are not safety-related?
10 MR. CHOKSHI: If there is any importance in 11 correction--
12 MR. MICHELSON: You have gone through and determined 13 all of your potential interactions and put them into the PRA 14 then?
15 MR. CHOKSHI: Right.
16 MR. MICHELSON: You have done that.on the basis of 17 structural interaction then?
18 MR. CHOKSHI: Well, physical interaction, system 19 interaction, and the boundaries.
20 MR. MICHELSON: Release of contents of the 21 equipment, that kind of interaction?
22 MR. CHOKSHI: Yes, if it is possible.
23 MR. MICHELSON: If it is possible--do you mean if it 24 is possible for a tank to split open or fall over and break?
-( )
25 CHAIRMAN SIESS: I'm sure you are getting an answer.
HERITAGE REPORTING CORPORATION -- (202)628-4888
55
(^-
1 MR. MICHELSON: I'm not sure.
2 CHAIRMAN SIESS: He is asking about PRAs. And have 3 you done any PRAs?
4 HR. SHAO: Yes--PRA experts.
5 MR. MICHELSON: Would you give me a PRA in which you 6 examine the seismic margins and that included this non-seismic 7 equipment and this rupture?
8 MR. CHOKSHI: Yes. For example, Diablo Canyon would 9 be a good example of seismic PRA that was done all kinds of I 10 possible--
11 MR. MICHELSON: You expect to see those kinds of 12 interactions?
I k' 13 MR. CHOKSHI: Yes. They are present. You will see.
14 MR. MICHELSOMt If they are present, but what do you 15 mean by that? If the guy thought of it? Where is the 16 documen-ation that shows he has gone through the process and i
17 he has eliminated that, he hasn't eliminated that? Is that 18 documented?
19 MR. CHOKSHI: Yes. If you go to Diablo Canyon--
20 CHAIRMAN SIESS: You might have trouble. Diablo 21 Canyon has a long-term seismic program. We had how many 1
22 volumes there for the PRA? And since the PRA, that was 23 issued, the staff has probably requested and received ten l
() 24 times that much paper. The staff has been reviewing -
25 '.ndividual event trees, fault trees and so forth now.
HERITAGE REPORTING CORPORATION -- (202)628-4888 w-____-_
d -
56' 1- :MR. CHOKSHI: There are three volumes.
2' ~ CHAIRMAN SIESS: They can get it for you.
- 3. .MR. MICHELSON: What I would really like to see are a ,
4 -the' guidelines you used for such.an analysis.
5: CHAIRMAN SIESS: It is not that simple.
6- MR. MICHELSON: You start out'and do I guess withoutL 7- the guidelines.
8 MR. CHOKSHI: You know, there are tremendous 9 information on interactions.
10 MR. MICHELSON: How did you say a 1 inch pipe, which 11 is normally a no never mind, but it might be located in an 12 . area where it is very vital. Control cabinets underneath that-
) 13 is wetted down with the 1 inch pipe. fails and the 1 inch' pipe 14 may fail by itself. It is a non-safety pipe going to the 15 -water cooler over in the corner, might fail by itself, or it
~
16 might fail because of a piece of equipment has fallen against 17' it and 1 inch pipe is not hard to rupture.
18 Is that sort of thing included? Several of these 19 might be happening during a seismic event.
20 CHAIRMAN SIESS: Carl, we are having a day and a 21 half on PRAs next month, and I have asked, I suggested that we 22 get some seismic PRA experts, and I'm sure that if Bob
, 23 Kenneally were here or some of those people, he could' answer
( 24~ those specific questions.
25 MR. CHOKSHI: Let me give an example. On Main HERITAGE REPORTING CORPORATION -- (202)628-4888
l
. 57
~.
.1 Yankee we did the margin review.
2 CHAIRMAN SIESS: That is not a PRA.
3 MR. CHOKSHI: We used PRA type techniques, and in 4 the generic letter, there was 1 inch pipe, and that was 5 examined and looked at for the impact on--
6 MR. MICHELSON: Was it safety-related pipe to begin 4
7 with?
8 MR. CHOKSHI: I have no idea. It was fire pipe.
9 MR. MICHELSON: Fire protection, which is not 10 seismic.
11 MR. CHOKSHI: So depending on the location, we have 12 tried to look at those. 1 13 CHAIRMAN SIELS: Next olide.
14 MR. SHAO: Next slide--
15 CHAIRMAN SIESS: I think we are finished with tt' 16 if I can remember that far back.
17 (Slide) 18 MR. SHAO: There are two available methodologies for 19 seismic IPEEE, seismic PRA and seismic margins. All.those 20 seismic PRA gives a lot of insight.
21 The problem with seismic PRA is large uncertainties 22 in seismic curves sometimes can be as large as three orders of ,
1
^
23 magnitude, another methodology is seismic rargin, and we are )
l
() 24 going to cover in a little bit more detail in the Subcommittee j 25 discussion. -
HERITAGE REPORTING CORPORATION -- (202)628-4888
i 1
58
. ,.m
~> 1 MR. MICHELSON: One of the things that puzzles me on i
2 this business is after our long discussions on A-17, system I
3 interactions, and you know, the long harangue with the staff 4 about how do you understand how you are going to handle system 5 interactions, I think the general conclusion was that it is 6 something we really don't know how to put our hand around yet.
7 If that really is the conclusion, then I guess we 8 are saying system interaction really isn't a part of this IPE 9 process because we don't even know how to handle systems 10 interaction, don't even knou how to define it for sure. Is 11 that right?
12 In other words, what is your viewpoint on system 13 interaction effects in the IPE process?
14 MR. CHEN: In the system interaction effect, we are 15 going to examine v -c is potential impact.
. That is one 16 possibility.
17 MR. MICHELSON: Physical impact?
18 MR. CHEN: Spatial interaction--that's a physical 19 impact. Also we talk about the contents. If you have a sort 20 of fire protection, they have water. If the water came down, 21 enuse pipe break, what the impact is going to be on the 22 adjacent equipment or facilities; all those things we are 23 thinking about to examine to identify whether there is a
() 24 potential problem, and then we examine whether that effect 26 much that impact. .
HERITAGE REPORTING CORPORATION -- (202)628-4898
b 59 O'
1 MR. MICHELSON: So you are going to include what? l l
2 MR. CHEN: Yes. !
l i 3 MR. MICHELSON: Let's go one step further in order i 4- to really do that. For instance, if I have fire protection, 5 and I have to release, and they get down into solid state I 6 control cabinets, how are you going to analyze the effect of 7 water on the solid state control cabinets? !
8 MR. CHEN: That is next. We are try to see if there 9 is w potential for that or try to see whether there is any 10 potential to avoid, to make sure that process would not 11 happen.
12 MR. MICHELSON: Are you including that as a 13 requirement in the IPE then?
14 MR. CHEN: In the guidance.
l 15 MR. MICHELSON: IPEEE, and I think generally, in 16 those complex cabinets, there is no way to understand what 17 really could happen, so you are saying the alternative is to 1
18 assure it can never happen?
19 MR. CHEN: Yes. .
l 20 MR. MICHELSON: That may be better protection 21 against water intrusion and smoke intrusion and heat intrusion 22 and so forth.
23 MR. CHEN: No, it does not to that extent apply to (h 24 everything, but we are thinking about how there are several ,
25 things to prevent it from happening or to evaluate its impact HERITAGE REPORTING CORPORATION -- (202)628-4888
L 60 1 and to determine how significant it is; if we cannot, really 2 cannot do that, then determine whether it is easy to,'just 3 prevent it.
4 MR, MICHELSON: Can't protect ags. inst it.
5 MR. CHEN: Yes. You also provide--
6 MR. MICHELSON: It would be interesting. That will 7 be in the guidance that they must address those kinds of 8 issues, either show that there is no effect on, acceptable 9 effect, or alternatively, to protect against.
10 MR. CHEN: We are hoping to have tnat.
11 MR. BAGCUT: We are going to declare A-17 is 12 included in the integration of all the generic issuen for
('N
\/ 13 IPEEE. Absolutely : ou can see it.
14 MR. MICHELSON: You remember A-17 finally was 15 resolved by saying we are going to worry about most ef this 16 under arnther_ program, and if that's, you are telling me that, 17 that's a different answer than I thought I just he'rd a because 18 that's the way--
~
19 MR. BAGCHI: In IPEEE we are supposed to include 20 this, supposed to integrate this.
21 MR. MICHELSON: A-17 finally decided it didn't know 22 -how to handle the problem and it delayed until the future. It 23 is going to be re-identified es a number of unresolved issues.
() 24 CHAIRMAN SIESS: I don't agree with that conclusion.
25 Let's go ahead.
HERITAGE REPORTING CORPORATION -- (202)628-4888
L 61 D
<)~
1 You said something that bothered me. IPEEE, does 2 that include the extent to which the plant has addressed the 3 UIs and GIs as well as the search for vulnerabilities?
4- MR. BAGCHI: There are specific UIs and GIs that are 5 supposed to be included and integrated in the seismic--not 6 every one of them. A-17 happens to be one of them.
7 CHAIRMAN SIESS:. The IPE includes the UIs and GIs.
8 MR. BAJCHI: Not every one of them.
9 CHAIRMAN SIESS: That's a part of the IPE?
10 MR. BAGCHI: Yes, sir.
11 CHAIRMAN SIESS: I thought the IPE was an individual 12 plant examination. Some of the UIs and GIs can be addressed 13 without examining the plants, can't they?
14 MR. BAGCHI: I am not talking about all the generic 15 recollection. I am talking about specific A-40, A-46, a-17.
16 MR. SHAO: Some specific ones.
17 CHAIRMAN SIESS: You just referred to the ones that 18 relates to the search for vulnerabilities, not to the rest of 19 the policy scope.
20 MR. BAGCHI: A-17 happens, A-17 happens to be one of 21 them.
22 CHAIRMAN -SIESS: Larry, we need a break. Can we 23 take it about now?
24 MR. SHAO: Sure.
25 (A brief recess was taken.)
HERITAGE REPORTING CORPORATION -- (202)628-4888
62
' 1 . CHAIRMAN-SIESS: We will continue now.
2 MR. SHAO: The next item ic fire. Why is IPE for 3 fire necessary? Mainly because the industry PRAs have shown 4 fire contributes to up to 50 to.60 percent of core melt 5 frequency. Also the criteria does not adequately address 6 certain fire issues that were identified by Sandia fire 7 scoping studies.
8 CHAIRMAN SIESS: Larry, does anybody know how much 9 of that 50 to 60 percent contribution is due to plant unique 10 vulnerabilities?
11 MR. SHAO: I didn't hear you, i I
12 CHAIRMAN SIESS: How much of that 50 to 60 percent !
i 13 contribution i:s due to plant unique vulnerabilities? I l
l 14 MR. SHAO: Plant--okay. Look at the next vugraph .
I 15 here. )
16 CHAIRMAN SIESS: That just tells me that every plant 17 they have looked . .
18 MR. SHAO: Every plant look a little bit different.
19 Okay.
l 20 CEAIRMAN SIESS: Put the next slide on then.
21 M?. KELLY: My name is Glenn Kelly. I am with ti.e 22 staff.
23 The results.of the PRAs have normally been that j 1 ) 24 while occasionally there is a specific area that was a 25 vulnerability, had a large vulnerability, what we would ,
HERITAGE REPORTING CORPORATION -- (202)628-4888
63 s- 1 -probably today define as a vulnerability or expect a" utility 2 to define as a vulnerability, most of the time the fire damage 3 is coming from areas fairly consistently in all the plants, 4 coming from the control room, from the cable spreading room, 5 from switchgear, and it is only occasionally where you will i 6 get a place where perhaps there is one plant had transformers ,
7 that had had a potential that if they had failed, that the 8 fire in that area could have knocked out a lot of very vital !
9 equipment, and that was a particularly strong problem that we 10 felt was, probably would be considered today a vulnerability.
11 CHAIRMAN SIESS: Well, now I look at that figure up 12 there, and Indian Point 2 which is a, if I'm not mistaken, a 13 large four loop Westinghouse plant, is a hundred times more 14 vulnerable than the contribution of fire or a hundred times 15 greater than for Zion, which is again a rather similar basic 16 plant.
17 Now is there something plant unique about Indian 18 Point 2 that makes the probability of core melt from fire a 19 hundred times greater than Zion?
20 MR. KELLY: Well, you can have several things. No.
21 One, you can have different people performing the reviews 22 which will make the difference. The other thing that you 23 can--
24 CHAIRMAN SIESS: That's not very encouraging.
f( ) J 25 MR. MICHELSON: That's one of the biggest factors in HERITAGE REPORTING CORPORATION -- (202)628-4888 3
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . --_ __ _ J
I 64' m
E( j .
2 1: :how you'do a. fire PRA, what you include _and what you exclude 2 and so~forth.
z3 MR. KELLY: There were some specific areas at Indian 14 . Point, and I think--it has-been many years. I remember there 1
5 was a steam tunnel, that.there was an area in there where he 6' had a fire that knocked out quite a bit of very,-very 7- important equipment, and there were a number of things, that l
8 al'so what happens sometimes, some plants when they are 9 performing a PRA, when they determine that the scenario that 10 it looks like it~is going to really knock up the core damage i 11 frequency, they will fix it and it won't end up being 12 reflected into the PRA. I believe in.the case of. Indian' Point' O- 13 that their PRA, this was perhaps--the risks, you are doing the 14 LER, see, were before they made-- I 15 MR. SHAO: That's a good one. Sometimes then'when
-16 they do a fire PRA, they see some vulnerability they fix right' 17 away and then the number come down after fixes.
18 CHAIRMAN SIESS: Big Rock is no worse than Indian 19 Point.
20 MR. SHAO: The number is higher, but percentage is 21 the same. Big Rock is 10 to the minus 4. Indian point, but--
22 ~ ' CHAIRMAN SIESS: Seismic differences; I just 23 wondered what you can conclude. I mean the fact that fires
() 24 contributes 50 to 60 percent, I don't know how pertinent that
.25 is to the IPE.
HERITAGE-REPORTING CORPORATION -- (202)628-4888 x___-____---_ _ _ -
65
,~
\~ 1 MR. SHAO: Look at Peach Bottom. Peach Bottom is 2 very recent one. This 1150, and Livermore has, EPRI had 3 Hi t , the fire can be 20 percent to 72 percent.
4 CHAIRMAN SIESS: If I compare Peach Bottom to 5 Limerick, they are about the same on fire, but the percentages 6 are different, and I don't know what good, I don't know what 7 the percentages are telling me.
8 MR. SHAO: Percentage, even though they are the same 9 number here, the percentage can change. Livermore and EPRI, 10 look how much they can change.
11 CHAIRMAN SIESS: What is the significance of the 12 percentages?
13 MR. SHAO: Mainly show why this particular e, at is 14 mportant.
15 CHAIRMAN SIESS: Has this got sonsething to do with 16 your definition of vulnerability?
17 MR. BECKNER: No, it does not. Obviously what is 18 dominant can change depending on other things. You can look 19 at either the fraction it contributes or look at absolute 20 magnitude.
21 CHAIRMAN SIESS: It seems to me that the Zion, for l
22 example, 1.8, 10 to the minus 6, and if all other initiators 23 were also 1.8, 10 to the minus 7, that would be 50 percent, 1
() 24 but I really couldn't get too excited about it at those L 25 levels.
HERITAGE REPORTING CORPORATION -- (202)628-4888
Yi ?'" '
66;
-~ '
1 MR. BECKNER: That's1true. .
~2 -MR. SHAO: But only--if you remember the internal
.3' IPE! generic letter, I think the utility.had to report any-4 ' event that's.more than 5 percent of the' total? Something like 5 that.
6 . CHAIRMAN SIESS: This is for the IPE.
7 MR. SHAO: IPE generic letter, they have to report 8 any event.
9 . CHAIRMAN SIESS: Which leads me back to the~ question 10' of what is a vulnerability?
11 MR. CHEN: This is John Chen.
12 CHAIRMAN SIESS: If it' contributes more than 5
- 13 percent?
14 MR. BECKNER: No. That.is definitely not the 15 definition of vulnerability.
16 . CHAIRMAN SIESS: Are we going to play this game by 17 me offering definitions.and you telling me no, or is somebody 18 going to offer me, is somebody going to define a vulnerability 19- for me?
20 MR. BECKNER: At this point in time, we are not 21 going to give a numerical definition of a vulnerability.
22 CHAIRMAN SIESS: Can you give a verbal definition?
23 MR. DECKNER: Without a number, potentially, yes.
(f - 24 CHAIRMAN SIESS: That's verbal.
- 25 MR. BECKNER: It's--I am not going to do it right I
i HERITAGE REPORTING CORPORATION -- (202)628-4888 i
- _ _ _ _ _ = _ _ _ .__ . _
67.
<^3 2 1 now. I think we have talked about, we have talked about that 2 we might get some guidance on the vulnerability. While I 3 think that we have been very far along, we do not want to 4 define it as 10 to the minus X or something like that. We 5 definitely do not want to do it that way.
6 CHAIRMAN SIESS: But how do people know what to look 7 for if you don't know how to define it? Right now you have, 8 the other generic letter is asking the utility to identify 9 vulnerabilities, and I think it says something about 10 justifying their category or something. Is that the way we 11 are going to go?
12 MR. BECKNER: The generic letter for the internal 13 events requests that they tell us first of all, information, 14 some reporting level.
15 Then it asks for what, their definition they use for 16 fixing things or for calling it a vulnerability or whatever.
17 And then thirdly, it tells us to, it requests them 18 to tell us what they actually did, what they fixed or what la they didn't fix.
20 We are basically leaving it up to the utilities.
21 That is consistent with what has happened with PRAs. People 22 who perform PRAs, they find things you--I don't know what you
- 23 'want to call them, outlier or vulnerability, or just something
() 24 that they were uncomfortable with and they fix them. That 25 historically happens with the PRA, and that's the same process HERITAGE REPORTING CORPORATION -- (202)628-4888 p
68 A-
~ l' here.
2 CHA13 MAN SIESS: But the PRAs have not been mandated 3 by the Commission, and the IPE is something that has been 4 mandated. I guess you can call a policy statement a mandate, 5 and when you say you are going to leave it up to the utility 6 to define a vulnerability or identify them, does that also 7 mean that you are going to accept what they find and what they 8 do about it?
9 MR. BECKNER: Basically we are going to review it.
10 There is no regulatory requirement for the IPE other than 11 perform it. There is no requirement to fix any given level or 12 do any specific thing. All they have to do is to tell us what 13 they did. That's the requirement.
14 CHAIRMAN SIESS: If they come back and say we didn't 15 find any vulnerabilities, you have to accept that? If they' 16 come back and say, gee, we found a vulnerability, the tank 17 could fall over, but we are not going to do anything about it, 18 you accept that?
19 MR. BUCKNER: We would request a little bit more 20 information as far as what their logic was. If we disagreed 21 with their conclusions, we could not force a fix under the IPE 22 progran. We have to go through our existing regulations if we 23 wanted to require them, to make a fix.
() 24 CHAIRMAN SIESS: Just the regulations, you mean 25 50.109?
HERITAGE REPORTING CORPORATION -- (202)628-4888
69 k-:-
1 MR. SHAO: Right.
2 MR. BECKNER: Unless it was that we determined they 3 were not in compliance with an existing regulation. Then it l
4 would be an enforcement issue. j i
5 MR. MICHELSON: Are you going to provide some kind 6 of guidelines as to what you expect that they will have to !
7 include in their fire PRA when they conclude that the fire may 8 be a non-problem for requiring them to fix it?
9 MR. BECKNER: I think that's what the fire 10 subcommittee is working on.
11 CHAIRMAN SIESS: 1.ey don't have to have a fire PRA.
12 MR. MICHELSON: No, but if they are going to use PRA
['i
(/ 13 as their basis for excluding things, they have to include fire 14 in the PRA.
15 MR. BECKNER: Whatever method they used to examine 16 fires, the subcommittee, I think they are going to report 17 eventually here. We are' working on that.
18 CHAIRMAN SIESS: They are excluding things simply on 19 frequency without a PRA?
20 MR. BECKNER: That's true 21 MR. MICHELSON: You have do determine--
22 CHAIRMAN SIESS. You can arbitrarily--the staff has 23 already done it. They have excluded things on the basis of
.() 24 frequency.
25 MR. MICHELSON: I don't think fire is in that I
HERITAGE REPORTING CORPORATION -- (202)628-4888
c
, 70 m
[1- category by any means.
'2- CHAIRMAN SIESS: No.-
3 MR. MICHELSON:' The.IPE proceca was dealing with
~ ~
4 'beyond designLbasis events-at least in part. I'm sure we.can 5~ think of others as well, but in the. case of beyond design 6 basis, are you going to look at beyond design basis fires?
-7' MR '. BECKNER:- I would assume so.
8 MR. MICHELSON: Will that be a fire in an area 9 designed for one hour that burns for two hours or something?
10 MR. SHAO: Let's let Conrad answer that.
.11 - MR. McCRACKEN: Conrad McCracken--we are not looking
'12 'for a simple definition of a design basis fire and'trying to 13 look at the two-hour fire in a one-hour area.
14- .Everybody has had to do a fire hazards analysis as 15 part of their plant, as part of their Appendix R resolution, 16 so they have.had to identify their fire loading, and the very i 17 vulnerable equipment in given areas. All right. They are not 18 going to suddenly assume that two times that fire loading 19 exists because it doesn't. They know it is there.
20 MR. MICHELSON: As you recall, of course, during the 21 core analysis, you assume the fire mitigation equipment 22 worked.
23 Now what happens if the fire mitigation equipment It is not non-qualified.
~( 24 doesn't work? It is non-seismic.
25 It is low-grade stuff. What happens--and you look at fire, L . HERITAGE REPORTING CORPORATION -- (202)628-4888-
=.
71 J%
() 1 people's records on fire protection, it isn't a hundred 1
2 percent certainly a fire protection system will even come on.
3 MR.'McCRACKEN: That is correct.
4 MR. MICHELSON: So in doing these kinds of analyses, 5 are you going to look at fires that burn beyond one hour or 6 three-hour barrier, going to look at a fire that burns beyond 7 three hours on the barrier, its effect on adjacent equipment?
8 MR. McCRACKEN: Plants have to make an assumption 9 they can put out a fire, have to justify that assumption. If 10 they are saying they can put out a fire by automatic system or I
11 fire brigade response or a combination of the two, they have 12 to show in fact that they can respond to fire brigade and do A- 13 what they need to do with that kind of equipment. That's part 14 of the, of the, of what they have to demonstrate, but we are 15 not going to assume that a three-hour fire barrier disappears.
16 MR. MICHELSON: No. We are talking about a fire 17 that burns beyond three hours. You don't know what it does 18 beyond three hours, just like a beyond design basis seismic 19- event. You look at what happens and you get a higher--you 20 look at what happens when you have a fire that burns longer
.. 21 than you thought it would burn.
22 MR. McCRACKEN: No. We do not assume that any fire 23 is going to burn and burn through a three-hour fire barrier.
() 24 MR. MICHELSON: Or one-hour barrier, which are much 25 more vulnerable.
HERITAGE REPORTING CORPORATION -- (202)628-4888
, i k <
i
.72 I l (~h l
-l 1 MR. McCRACKEN: One-hour barriers have automatic !
1 2 suppression.
]
i 3 MR. MICHELSON: Yes. That's what didn't work.
l 4 Whatever reason, the seismic event, local fire-as a {
l l
5 consequence of the seismic event, do you look at that? l l
6 Non-seismic fire protection--
I 7 MR. McCRACKEN: The issue of seismic and fire we i
8 addressed in the water procedures. I went through that the I 9 last timing.
l 1
10 DR. SHEWMON: Conrad, what, who collects information 11 on fire frequency and fire severity and how long they, you 12 know, what is--you get data on--are we seeing examples of l
O- 13 where fires indeed burn longer than the design basis and how 14 often that happens?
15 MR. McCRACKEN: The procedure we have now, and I 16 have forgotten the NUREG number, but in the emergency response 17 procedures, any fire that is greater than ten minutes, they 18 have to report, and that's one of their response categories 19 which activates the emergency response, so that's getting a 20 fire reported that occurs, lasts longer than ten minutes.
21 Any fire that disables, under 5072 and 73, a 22 safety-related component is reportable no mattgr how long it 23 lasts. If it last 30 seconds, disables a safety-related
() 24 system or component, it gets reported, so we have LERs that 25 tell us every time we have lost a saf ety system component. It HERITAGE REPORTING CORPORATION -- (202)628-4888
73 g' /^)
1~ is rare that you see a-fire that lasts as long as an hour.
2 That's an unusual fire. There have been very few.
3 DR. SHEWMON: This data has been collected over how l
4 long?- Five years? Ten years? i 5 MR. McCRACKEN: The data has been collected probably 3 6 closer to.15 years now, and that data base is currently being 7 updated. Research just updated the data base on all the LERs.
8 That's there. Now industry--
9 DR. SHEWMON: Unlike seismic events then that may or 10 may not come in my lifetime for many of these plants, we do 11 have a fair amount of information on fire frequencies?
12 MR. McCRACKEN: Yes. I think this is a good data sO\
() 13 base.on fire frequencies, fire frequency.
14 MR. SHAO: That's right.
15 DR. SHEWMON: Mixed up at least in my mind. Thank 16 you.
17 MR. MICHELSON: They had a recent zirconium fire 18 that burned more than one hour.
19 MR. McCRACKEN: But that fire burned because they 20 specifically elected to let it burn while they were there, 21 knowing what equipment was in the area and what they were 22 doing. They chose that instead of using fire to, water to put 23 it out.
() 24 MR. MICHELSON: Maybe I read a different report than 25 you read. They first put on 150 pounds of chemicals on it, HERITAGE REPORTING CORPORATION -- (202)623-4888
74 1 CO2. Brought in another 150 pounds of CO2, didn't put it out.
2- Brought 150 pounds of chemicals, didn't put it out, and then 3 finally decided they would have to put water on it.
4 MR. McCRACKEN: That is correct, but they were 5 there. The fire brigade was there. They knew that the fire 6 was under control. They knew where it was limited, and they 7 were going with progressive actions to put it out, the last 8 action being putting water on an electrical fire. That was 9 the conscious decision. It didn't burn out of control, nobody 10 knew what was going on. They knew where it was controlled, 11 where the boundaries were, and the fire brigade person in 12 charge made a conscious decision to do what he was doing.
13 MR. MICHELSON: Part of the problem was they didn't 14 know what all electrical equipment was still energized, were 15 afraid to put water, and thar's what it was, but this is the 16 way it is in any area of the building when you are dealing
~
17 with electrical fires, and one-hour limit is not necessarily--
18 MR. McCRACKEN: But an electrical fire is rarely 19 going to be a full involvement fire, going to destroy a 20 building or go through three-hour period. They are going to 21 burn for a long time, but they are in a controlled area.
~
22 DR. SHEWMON: You are implying that it would burn 23 through a one-hour barrier and go out of control. He is
() 24- trying to tell you, as I understand it, that the probability 25 of that was about zero.
HERITAGE REPORTING CORPORATION -- (202)628-4S88
75
,m .
+
N-)' 1 MR. MICHELSON: No. The question is whether the 2 -fire goes out of control is the question, whether the control 3 circuits which might be vital to continuing moving of the core 4 which go through a conduit with a'one-hour barrier right off, 5 the fire will survive beyond the hour. That's the question, 6 not whether the fire will go out. I'm not worried about the 7 fire. That area is lost.
8 MR. McCRACKEN: That is exactly what you do in your 9 vulnerability search. You find whether you have areas that 10 exist like that.
l 11 MR. MICHELSON: My original question was and still 12 is do you go back now and look at what is in these conduits in j O 13 the unlikely event that you have a fire that exceeds one hour?
l 14 Is that a part of this beyond design basis examination? And 15 we keep saying we are doing it,'and in this case, I am not l 1
I 16 sure you have said you will go beyond one hour.
17 MR. McCRACKEN: In that particular fire, the answer 18 would be no, but they have an alternative method of shutdown i i
19 which if that fire burns, everything over and above would not 20 have caused that plant a difficulty.
21 MR. MICHELSON: I am talking about the principle, j 22 In principle, are you going to go back and look at your fire 23 vulnerabilities and fires that go beyond an hour, see whether .
() 24 or not, how important that conduit might have been? ,
25 MR. McCRACKEN: Not unless the data base shows that l HERITAGE REPORTING CORPORATION -- (202)628-4888
\~/ ' 1 that is a realistic fire.
2 MR. MICHELSON: What is this--the data base doesn't 3 show a high, you know, very large earthquakes are necessarily 4 very probable, either, when we are looking at--
5 MR. McCRACKEN: That's because you haven't got a lot 6 of experience with a lot of earthquakes. We do have a lot 7 experience with fires in nuclear power plants.
8 MR. MICHELSON: This is really not looking at beyond I 9 the design basis? l 10 MR. McCRACKEN: Not in the, not in the context that i 11 you are asking the question, no. i
)
12 MR. MICHELSON: That's the context that fires beyond !
-(
( 13 one hour is beyond the design basis. The rest of it we have i
i l
l 14 looked at. i 15 MR. McCRACKEN: We don't have a design basis fire.
16 MR. MICHELSON: I thought we did in Appendix R. ;
17 MR. McCRACKEN: No.
18 MR. MICHELSON: How did you ever do your analysis I 19 without a design basis for knowing how big it was and how long i
i' 20 it burns?
21 CHAIRMAN SIESS: Set up criteria for fire protection 22 and criteria for walls and so forth, i
23 MR. MICHELSON: The loading, the loading has been l
() 24 defined, and the duration.
25 CHAIRMAN SIESS: You can design without the fire HERITAGE REPORTING CORPORATION -- (202)628-4888 !
[ .rn.
77 ll . criteria if you just arbitrarily set barriers and spatial 2 requirements and water requirement. !
l 3 Larry, let's see your next slide.
4 MR. SHAO: Okay. This is my last slide.
5 (Slide) 6 CHAIRMAN SIESS: Not in my book. You have skipped I 7 the one that said fire IPEEE.
8 MR. SHAO: Oh, yes.
9 CHAIRMAN SIESS: That's the one I want. Now let me 10 point something up here to the Subcommittee. Mr. McCracken 11 has said that as we all know, he presented most of everything 12 he had on the fire IPEEE at the Subcommittee meeting two or 13 three months ago and again at a Full-Committee meeting, and 14 the Committee agreed at that time that it 1>oked reasonablo 15 but that we would reserve any final comment until we saw the 16 actual guidelines. I guess that's the right word for the fire 17 IPEEE.
18 Mr. McCracken has advised me that he has nothing to ;
19 add at this time to what was presented to the subcommittee and ;
20 and the Full Committee in those meetings. He is prepared to 21 repeat what he said, and I would like to suggest that that's 22 redundant, and that we let him go home.
23 MR. MICHELSON: When do we get the guidelines?
() 24 CHAIRMAN SIESS: Whenever we get them for the rest 25 of this. I think they are all coming at the same time. Is !
HERITAGE REPORTING CORPORATION --- (202)628-4888 l
78
"~ that right?
1 2- MR. MICHELSON: What point is that?
3 CHAIRMAN SIESS: What is your deadline for 4 . guidelines?
'5 MP. SHAO: Sometime in November, December.
- 6. CHAIRMAN SIESS: Generic letter.
7 MR. SHAO: What time--you guys going to get a 8 calendar?
9 CHAIRMAN SIESS: Sometime this year, end of this 10 year.
11 DR. SHEWMON: I wanted to ask then those being 12 developed by NUMARC, that is by the end of calendar '89?
13 MR. SHAO: Right.
14 DR. SHEWMON: Schedule.
15 CHAIRMAN SIESS: NUMARC is on the same schedule you 16 are?
17 MR. BAGCHI: Yes, sir.
18 MR. McCRACKEN: For the fire, they are on the same 19 schedule we are. As I told you last time, there are a couple 20 of things that may be miss, there may be a couple of tables 21 missing from the guidelines. By the end of December, the l
22 guidelines will be well enough developed that they can be 23 reviewed fairly thoroughly.
() 24 CHAIRMAN SIESS: Okay. We will wait until we see 25 the guidelines on fire. We have heard which direction we l
HERITAGE REPORTING CORPORATION -- (202)628-4888
,,--e -ve- - - -
$N ,
x :?? -
. 79 EU fthink.they are going, and I think we did provide you some p
- ajs 2_ comments in that letter, so it is the last we will hear of QD -3 fire today and now sps to other things.
4- (Slide) I C
5 MR. SHAO: This is really the catchall. The 6 -important ones are high wind, flood, tornados, transportation, ,
< 7 and others.
i
'8- CHAIRMAN SIESS: High winds is hurricanes? i 9 MR. SHAO:' High wind, hurricanes, can be hurricane, 10 tornado, tornado missiles, flood, explosion, toxic gas, 11 aircraft, shipping accident, and so on. !
12 MR. MICHELSON: What is a beyond the design basis !
L/^'Y . l k/ 13; high' wind? Do you have a certain design basis wind-- l 14' MR. SHAO:- Right now the most wind designed is a l L15 hundred miles per hour; for the tornado, designed for 360 j 16 miles per hour, i
17 MR. MICHELSON: We aren't going beyond the design 18 basis then in this case?
1 19 MR. SHAO: Okay. Now the thing is you hear later on 20 for the tornado is designed for 360 miles per hour. If you 21 look at probability of that tornado, it is 10 to the minus 7.
22 It is already in the severe accident realm, so you hear later
- 23 on any plant designed to the current criteria which is tornado i
) '24 of 360 miles per hour, which is probability of 10 to the minus l 25 7, this plant is creeping outright automatically.
HERITAGE REPORTING CORPORATION -- (202)628-4888 1
pyr" ey" 3
(hK, @ 80 N ', ' V , 4 C -'
1 j Y- N^
1: CHAIRMAN SIESS: Where did you get that' figure 10 to l
[ ,
V!M -
h "
2 the'minus-77 ;
3 MR. SHAO: 'How it-is from the reg guides.
4- CHAIRMAN SIESS: I thought you said it was zero i i
Si once. !
1 I 4
6~ MR.-SHAO: Based on-- 1 l
A 7 CHAIRMAN SIESS: I don't think anybody has ever seen 8 a tornado'that. strong, have they?
9 MR. CHEN: That's based on the study by the ANSI 10 2.3.
- 11 CHAIRMAN SIESS: That was extrapolating? l o 12 EMR . CHEN: That is extrapolating. It is not say you I
-(~Y 'actually, you have seen it.
i A/ - 13.
- 14 CHAIRMAN SIESS: Phenomenologically some people have
- 15. decided that.
l16 DR. SHEWMON: Do you think if you came back to a l 17 mini-tornado, that higher than a wind storm, are you, is there !
, 18- something then, what about a 200 mile tornado, 150 miles an i
.19 hour tornado? The probability of that would be higher or-- !
20 MR. SHAO: Oh, that would be higher.
21 DR. SHEWMON: Or is 10 to the minus 7 the 22 probability of a tornado hitting any plant in the Midwest?
23 MR. SHAO: Plant, for instance, let's say a 24 -building, a recent plant, Limerick, whatever, it is designed
. i( f
- 25 to 360 miles per hour, 10 to the minus 7. Okay. We have 200 HERITAGE REPORTING CORPORATION -- (202)628-4888
_=_ _ _ _ _ _ _ - - _ _ - _ - - _ _ - - _ _ -
81 1
N' 1 miles per hour. The design is okay right there.
2 DR. SHEWMON: The probability of a tornado hitting a 3 plant like Byron is probably higher than 10 to the minus 7.
4 MR. SHAO: Maybe 10 to the minus--Byron is designed 5 to 10 to the minus 7 tornado.
6 CHAIRMAN SIESS: I think'there is a confusion 7 between the probability of tornado hitting the plant and 8 probability of tornado existing.
9 Is it 10 to the minus 7, the probability of a 360 10 mile an hour wind, a tornado, or probability of a 360 mile an 11 hour1.273148e-4 days <br />0.00306 hours <br />1.818783e-5 weeks <br />4.1855e-6 months <br /> wind at a site?
12 MR. SHAO: Probability of wind and at. site. Can be 13 even higher even at each plant.
14 CHAIRMAN SIESS: The property of the plant can 15 sustain even lower than--
16 DR. SHEWMON: Even though the probability of 360 3
17 miles an hour wind at Byron might be 10 to the minus 7, it is 18 designed to take that, that tornado of that strength, is that li right?
20 MR. SHAO: Right.
21 CHAIRMAN SIESS: As T recall, straight winds govern 22 for the low probability--the higher probability, the straight 23 winds tend to dominate?
() 24 MR. CHOKSHI: You are right. For the plants which 25 have not bema designed for the tornado, the straight winds HERITAGE REPORTING CORPORATION -- (202)628-4888
i; 82 I.
L' -
1 because of the higher frequency will go under core damage 2 frequency.
3 CHAIRMAN SIESS: If they had been designed for 4 tornado, straight winds dont--they haven't been designed for 5 anything. Hundred mile an hour wind has a high probability, 6 it will be up, it will be more damaging, more governing on the 7 plant in terms of' risk than the 260 mile an hour.
8 MR. MICHELSON: Well, presumably the plants have 9 already been designed for those major sources, is that 10 correct?
11 MR. SHAO: No, not necessarily For current plants, 12 all the current, every plant designed to current criteria.
13 Nothing can be screened out, but some other plants designed in 14 the 1960s or early '70s.
15 CHAIRMAN SIESS: The IPEEE, the look for 16 vulnerabilities, doesn't that include let's say a plant that 17 was finished yesterday?
18 MR. SHAO: Yes.
19 CHAIRMAN SIESS: That might have been designed for a 20 360 mile an hour tornado, but some part of the balance of 21 plant wasn't, and when you do a real risk assessment, you find 22 out that that balance of plant c6n go out as a contributor.
23 MR. SHAO: Even for current plant, we ask for
() 24 somathing. That is included.
25 CHAIRMAN SIESS: That's why I think tl.is discussion HERITAGE REPORTING CORPORATION -- (202)628-4888
83
. (N}' 1 of what it is designed for doesn't mean that it has the 2 resistance to it everywhere at the same time, same. level of 3 ' risk.
4 MR. SHAO: Let me give exmmple. The plant was 5 designed yesterday, you design for 360 mile per hour.
6 CHAIRMAN SIESS: So somebody could have made a 7 mistake.
4 8 MR. SHAO: Designed a year ago. All of a 9 ' sudden--and somebody bring another power plant next tour.
10 They putting in high stack 400 feet high, and that stack is 11 not designed for tornado, and when that stack fail, it may hit 12 the other plant, vulnerable to that failure or criteria cover 1 13 this kind of thing.
14' CHAIRMAN SIESS: You are going to look at the risk 15 in terms of core melts. As a result of these phenomena, they 16 were designed for it.
17 MR. SHAO: Okay. Let's defer this discussion until 18 later on when we give you detailed presentation on this.
19 CHAIRMAN SIESS: The utility can point to design 20 basis as being suitably carried out, reason for not looking 21 further on sotie of these things. That' a legitimate argument.
22 You don't necessarily rule it out because it was part of the 23 design basis. People do make mistakes.
()' 24 MR. SHAO: We are not going to check every 25 calculation yGu nake in the stress reports.
HERITAGE REPORTING CORPORATION -- (202)628-4888
w.-_ _ _ _ _ . _ _ - . - . -
+. .
M 84
- f~)
'\' 1 CHAIRMAN SIESS: Indian Point was designed for an 2 earthquake, too. Somebody didn't taxe into account the 3 hammering effect between two buildings, and so it turned out 4 that was a--
5 MR. SHAO: The hammering, the hitting of the two 6 buildings, is not in the design basis for SSE.
7 CHAIRMAN SIESS: I'm sorry, but the earthquake, it
-was designed for earthquake.
9 MR. SHAO: The Indian Point, when--the two buildings 10 hit only when occurred earthquake equal to two or three times 11 SSE. During the, at the design basis, the two buildings 12 doesn't hit. You go two or three times SSE, then the O. 13 buildings start hitting.
14 CHAIRMAN SIESS: But the point is you don't have to 15 limit yourself to the design basis when you look at risk.
16 MR. BECKNER: We agree with you, and we have 17 included it.
18 MR. SHAO: I think we will have a detailed 19 discussion on this wind, tornado and others design approach.
20 CHAIRMAN SIESS: Some design basis events, you think 21 the probability of exceeding them is exceedingly low?
22 MR. SHAO: Yes.
I 23 CHAIRMAN SIESS: Design basis earthquake is what?
() 24 You figure 10 to the minus 3, 10 to the minus 47. Design !
25 basis tornado you think is 10 to the minus 7? j l
HERITAGE REPORTING CORPORATION -- (202)628-4888
J l
85 l l- '
4- ' ' 1 MR. BECKNER: Right.
2 CHAIRMAN SIESS: That's the difference. Go beyond i
3 the design basis when it looks-- I 4 MR. SHAO: If you design properly, if the design is l 5- done properly, the tornado design is already in the severe l'
! 6 accident.
7 CHAIRMAN SIESS: You know, that's just common sense. I 8 I am not going to look at external river flooding et Byron.
9 MR. SHAO: No.
10 CHAIRMAN SIESS: What, a hundred feet above the 11 river on the bluff?
12 MR. SHAO: Right, i
i v 13 CHAIRMAN SIESS: But you would at Twin Cities.
14 Okay.
15 MR. SHAO: Okay. Any more questions on this 16 vugraph?
17 These are--the wind can cause loss of off-site 18 power.
19 CHAIRMAN SIESS: Aircraft, you are sticking to your 20 previous criteria of aircraft outside a certain zone? It is 21 background?
22 MR. SHAO: Aircraft, I think our criteria would be 23 frequency.
() 24 CHAIRMAN SIESS: This is the way we have been 25 designing them. You made out your runways standing out from HERITAGE REPORTING CORPORATION -- (202)628-4888
86
.them, and if anything was outside.that range, it was 1
2 considered. background and backgtound was what, 10 to the minus 3 7?
4 MR. BECKNER: Typically I think the SRP is using 5 If.ke the figure of 10 to the minus 7 or something, and it is l
6 for--we will be using the same criteria.
I 7 MR. SHAO: Frequency less than 10 to the minus 6.
8 CHAIRMAN SIESS: If you don't do that, every plant 9 is vulnerable to an earthquake.
10 MR. SHAO: Yeah. Okay. I'm done.
I 11 CHAIRMAN SIESS: Have you ever looked at the concern 12 the Germanr had on aircraft crash? For those plants, I mean 13 we had a couple of plants designed--TMI.
1 l 14 MR. SHAO: Jet pir.nts, which is a high missile.
15 CHAIRMAN SIESS: They were concerned about the 1 16 impact of the aircraft producing some very high frequency 17 vibrations which could affect things like relays. Anynody l
18 ever look at that?
19 Do you know what--not just the structural effects, 20 but they were worrying about developing a high frequency 21 vibration which we worried about a few times.
22 MR. WYLIE: You mean if a crash?
23 CHAIRMAN SIESS: Yes. Jet engine going in there.
() 24 MR. SHAO: The Germans usually, because they have a 25 small country, they have a lot of jet engines--I remember I
l HERITAGE REPORTING CORPORATION -- (202)628-4888
87 p'
'- 1 United States have only one site that they may have jets l
l 2 engines.
3 CHAIRMAN SIESS: Three Mile Island was designed for 4 Boeing 707 at Zion. The design concerned fire from a business 5 jet, and I think going into Milwaukee.
6 MR. SHAO: Buriness jet is not hard. There is a 7 real softness.
8 CHAIRMAN SIESS: There is a question of fire. The 9 crashing, the fuel spilling all over the plant, and they ended 10 putting in the labryinth entrances at certain places. I think 11 there is some--that's the only one I know that did that. I 12 hate to think of the n'imber of times I have flown over 13 Dresden.
14 MR. SHAO: Three Mile Island was designed for--
15 involving, designed for airplane crash.
16 CHAIRMAN SIESS: Okay.
17 MR. SHAO: Okay.
18 MR. MURPHY: I'm Andrew Murphy.
19 CHAIRMAN SIESS: Is that the rest of the handouts 20 the rest of the day?
21 MR. MURPHY: I can't see. This would be for, the 22 handouts for the seiFmic portion. That will take us until 23 10:15!
() 24 DR. SHEWMON: No sarcasm!
25 MR. MURPHY: I'm sorry. My name is Andrew Murphy.
HERITAGE REPORTING CORPORATION -- (202)628-4888
89 0 1 I'm the branch chief. of the Structural and Seismic Engineering ;
i 2 branch, and more important at this staga, I'm the Co-Chairman l 3 for the Seismic Subcommittee of, for the External Events 4 Steering Group. I 5 Just to let you know who is developed with the )
a 6 Steering Group, we are starting off with a couple of 7 seismologists leading this parade, followed by some structural 3 engineers, structural engineers with strong background in PRA, 9 a systems engineer, and some more structural engineers, so 10 that you can see that we have got a reasonable mix of people 11 that are addressing this problem.
12 CHAIRMAN SIESS: What part of that am I supposed to 13 get comfort from?
14 MR. MURPHY: I'm not sure I am offering you comfort, 15 just letting you know what the ground rules are.
16 (Slide) 17 MR. MURPHY: And our presentation at this stage, the 18 whole package that you do have in front of you, we will start 19 off, we will all present a short overview followed by Nilesh 20 Chokshi, who will have a discussion of the PRA issues. Roger 21 Kenneally will talk about the seismic design margin or seismic 22 margins methodology to give you an overview of what that 23 process or that methodology is all about. Nilesh Chokshi will
() 24 then talk about the containment issues, followed by Leon 25 Reiter, who will give you our feelings about the review level HERITAGE REPORTING CORPORATION -- (202)628-4888
l 89 i
( 1
\- 1 earthquakes for this process.
2 (Slide) 1 l
3 MR. MURPHY: At this stage, I think I want to j i
4 reiterate that this is a progress, definitely progress
> 5 reports, this is a progress report. I f elt that I could start 6 off with two unequivocal statements at this time.
7 The first is that seismic is definitely an external 8 event whether it is inside or outside of the plant boundary.
9 The second one is probably more important. We are 10 not proposing to provide a definition of vulnerabilities. We 11 are going with management policy at this stage that we are 12 leaving the term vulnerabilities as an item for the individual 13 utilities to address. We certainly will address what they 14 have chosen for a definition when we come to the review.
15 Okay. I want to say at this stage the primary 16 purpose of the IPEEE is, the seismic portion of it, is for the 17 plants, the utility, to identify specific vulnerabilities 18 within that individual plant. There will be generic 19 guidelines and guidance provided to help them pick out the 20 specific vulnerabilities within the plant.
21 The second iten would be for the plant owners, the 22 utilities, to develop a better understanding cf the plant 23 response to a seismic event.
() 24 The Subcommittee has agreed that at this stage, the 25 guidelines will be provided and that those guidelines will HERITAGE REPORTING CORPORATION -- (202)628-4888
90 0 1 indicate that either a. seismic PRA or a seismic design margins 2 program, either the NRC or the EPRI version, will be a 3 sensible way for them to carry out this program.
4 We further decided that we will use, for ourselves, 5 we will use the HCLPF--high confidence low probability 6 failure--number as guidance on this. In other words, we will 7 request, require that the utility, if they have selected to do 8 a PRA, that we will require that this number be converted to a 9 HCLPF for the plant to aid the staff in comparison, to make 10 comparisons that the HCLPF respond used as a comparative for 11 the staff.
12 The important criteria that we need to (#ddress for O
\~ 13 doing the IPEEE examination is the review level earthquake, 14 and this will be the earthquake that would be required, the 15 staff would require on the utility to examine their plant to.
16 We went to look at this or set this review level 17 earthquake high enough to cover, uncover plant vulnerabilities 18 if any exist out there, but we want to also set this to 19 moderate this by indicating that we want to set it low enough 1
1 20 that we minimize the senpe of the examination. We don't want 21 to overly-burden a utility with the examination process.
22 DR. SMEWMON: Explain what dimension you are in when 23 you say high enough because I tend to think in terms of G, and
() 24 if we set this thing at .6G, that's high enough that will 25 never, so that you must be using some inverse of G when you 1
"""' " " """ " " c "" "^ * " -- <202>e28-4888 E_ _ - -
91 rT
- l' 'say high enough to uncover vulnerabilities.
2 MR. MURPHY: Well, we will address that in effect in 3 the next bullet.
4 MR. MICHELSON: Does high enough mean beyond the 5 design basis?
i 6 MR. MURPHY: Yes, sir, very definitely.
.l 7 DR. SHEWMON: Don't hide it and solve it that way.
8 MR. MURPHY: I wasn' t hiding i t. I wasn't sure that 9 you could see the bottom part of the slides, which in part i 10 answer that question. l 11 What we are considering now for the review level 12 earthquake is f~or that number to be established by the staff.
13 Originally we were thinking about the generic level for the 14 eastern United States and then some plant-specific levels for 15 the plants west of the Rockies.
16 Currently what the staff is thinking is for plants ,
17 that have elected to use the seismic design margin process, 18 establishing three bins. Depending upon which of these bins 19 the plant falls in, we are talking about requiring a seismic 20 margins quality walk-down. The utility would be required to 21 carry this walk-down out, specify in effect how they did it, 22 what they looked at, and report back to the NRC. At that 23 stage, they would nominally be finished with the process.
() 24 CHAIRMAN SIESS: What does the word quality mean in i 25 there? Qualitative?
HERITAGE REPORTING CORPORATION -- (202)628-4888
92 o- 1 MR. MURPHYr No. It's a quality rather than--it 2 would be quantitative in there would be things reported, but 3 what we are saying is the quality, the level, the goodness of 4 the walk-down would be more on a par with what we are seeing 5 in the seismic design margin walk-downs than what we saw with 6 the PRA walk-downs. We are saying that the design margin 7 walk-down is a better walk-down than the PRA walk-down.
8 CHAIRMAN SIESS: And that differs from a seismic, it 9 differs from a quantitative one that you don't develop the 10 fragilities for componants? You just go look at them, see if 11 they are anchored down?
12 MR. MURPHY: In effect that's the tendency of it.
13 Okay.
14 For the second bin, we would be requiring, 15 requiring, our guidelines would suggest a seismic margins 16 walk-down, seismic margins review at the .3G 1evel.
17 For the third bin, it would be a seismic margins 18 review at the half a G 1evel.
19 CHAIRMAN SIESS: Now the bins are different plants?.
20 MR. HURPHY: The bins would contain the names of 21 different plants.
l 22 CHAIRMAN SIESS: You are going to tell us what
, 23 plants go into each bin?
l
() 24 MR. MURPHY: Today, no.
25 CHAIRMAN SIESS: On what basis would you decide? l l
l L_____.--_______---__--___-_.---_ i
i 93
- f~y.
'= 1 What would be the criteria? .
2 MR. MURPHY: That is one of the items that we are 3 developing. This morning you will see the thinking behind two 4 -criteria that we thought that we have come up with so far as 5 to binning the entire onsemble of plants.
6 CHAIRMAN SIESS: Some plants would be three tenths 7 and some for five tenths?
8 MR. MURPHY: That is correct.
9 MR. SHAO: And some would be reduced. l 10 DR. SIESS: You are working on criteria which would 11 be which?
I 12 MR. MURPHY: That is correct. What we are looking O. 13 for is to develop a number of criteria. We have got two in l l 14 hand right now which may or may not work. We would like to i
15 develop a number of them, maybe five, six, seven. i 16 CHAIRMAN SIESS: Why would there be two levels?
17 MR. MURPHY: There wouldn't be two levels. There 1
18 would be two different criteria for deciding how a building l
19 ended up in a particular bin.
20 CHAIRMAN SIESS: If I take all the plants east of 21 the Rocky Mountains, some will be looked at for three tenths 22 and some for five tenths?
l 23 MR. SHAO: Based primarily on hazard.
l
() 24 CHAIRMAN SIESS: South Taxes, you look for le>k at 25 for three tenths.
l HERITAGE REPORTING CORPORATION -- (202)628-4888 L_---_---_-_
'O 'l u
94
' 1 MR. SHAO: Minimum prngram; Florida plants are j
~2 minimum.
5 CHAIRMAN SIESS: Five tenths would be--
4 MR. SHAO: Maybe like New England areas.
5 CHAIRMAN TIESS: Okay, and West Coast.
6 MR. SHAO: This is talking about east of the Rockies 7 only right now.
8 MR. BAGCHI: I like to interject one thought. We 9 are working pretty closely with NUMARC. They have gone 10 through the PRA seismic evaluation program. They are doing 11 some work. We haven't heard all of it. We would like to 12 identify a fair anount of flexibility in the way we are i
l l 13 looking at this thing. We are going to wait, incorporate some l
14 of their thoughts, and then our initial thinking is that we 15 will bin plants into different, three cetegories that Dr.
16 Murphy has pointed out.
17 MR. SHAO: I think with Mr. Reiter you will hear--
18 CHAIRMAN SIESS: I get your point. It is based en 19 hazard. This is only for the design margins. If they choose 20 to have seismic PRA or choose to do one, then they de a PRA?
21 MR. BAGCHI: Right.
22 MR. MURPHY: That is correct, and it would be our 23 intent to provide specific guidelines on whrt hazard curves r () 24 are used in carrying out that PRA.
25 CHAIRMAN SIESS: I still don't really understand HERITAGE REPORTING CORPORATION -- (202)628-4888
95 1 that. As an alternative to?
2 HR.. MURPHY: Let's start off by making it simple, 3 that we are thinking about requiring a seismic walk-down. Our 4 thought is that the seismic--
5 CHAIRMAN SIESS: That I understand. It is when you 6 put margins in there that it confuses me.
7 MR. MURPHY: Okay. There are a cet of prescriptions 8 if you want for how to carry out a good seismic margins 9 walk-down or walk-down in the seismic margins program. Okay.
10 Those prescriptions provide us with more information 11 and more insights than what we as a group think we see in a 12 typical seismic PRA walk-down, and what we are saying is we 13 would like to see the better walk-down carried out, a more 1$ thorough walk-down carried out.
15 CHAIRMAN SIESS: Is there a distinction between a-15 seismic margins walk-down and what you might call a seismic 17 vulnerability walk-down?
18 MR. BAGCHI: No.
19 MR. MURPHY: I would say the same thing not using 20 either one of those terms.
21 CHAIRMAN STESS: If the margin is low enough, it is 22 vulnerable?
25 MR. BAGCHT: Right.
() 24 CHAIRMAN SIESS: You are going to have to excuse me, 25 but I'm having trouble, and I guarantee you that the Full HERITAGE REPORTING CORPORATION -- (202)628-4888
n 96
,,-s-
- f k 1 Committee will have trouble talking about vulnerabilities 2 without a definition.
3 MR. MICHELSON: I'm not sure when a seismic margin 4 means if you just do a walk-down. Seismic margins--
5 MR. MURPHY: It is a whole--
6 MR. MICHELSON: Calcu'.ating whatever--
7 MR. BAGCHI: The Committee on this, if you go to 8 EPRI, NRC or EPRI, seismic margins program, the documents are 9 out there. There is a component which discusses what kind of 10 walk-down it is. There is a procedura. It is written down.
11 What we mean is the walk-down that follows that kind 12 of a procedure, and he is identifying this as seismic margins l
l 13 type of walk-down.
14 CHAIRMAN GIESS: You don't have to have fragilities?
15 MR. BAGCHI: You don't have to have fragilities. We 16 just discussed whether or not based purely on hazard you could l
17 decide to do nothing. That was one of the proposals on the 18 table.
1 19 CHAIRMAN SIESS: So it is looking then for 20 interactions, looking to see if the anchored equipment is
, 21 anchored? Looking to see--
l l
22 MR. BAGCHI: Absolutely--looking for outliers.
23 MR. MICHELSON: If you do a walk-down of a given
() 24 area, you may decide a rupture of a particular pipe in that 25 area is perhaps a no never mind because of the consequer.ce in HERITAGE REPORTING CORPORATION -- (202)628-4888
97
) 1 that' area, but you have to integrate that with all the other l 2 areas that aight experience small pipe ruptures at the same i
3 time unless you are saying somehow an earthquake never breaks 4 more than one pipe in the whole plant irrespective of size.
l 5 Maybe you have got a rule like that. !
l 6 MR. CHOKSHI: There are two things simultaneously.
i 7 One thing in the low hazard that-- j P MR. MICHELSON: Are you looking at the whole plant 9 when you lock at the hazard in a particular area?
10 MR. CHOKSHI: No. What T think we are saying, that 11 the welk-dewn is the, one of the major parts of doing PRA 12 margin. Most cets out cf doing a walk-down. You can define i 13 the problems like land or age which are the typical problems 14 found, tanks, walk-down will take care of a number of those 15 problems, if there are any problems--
16 MR. MICHELSON: Decide whether or not you need to 17 reinforce the anchorage on the plant, decide based on what 18 happens in an area you don't need to, but do you remember what l'
! 19 happens in all the other areas at the same time you are out 29 there doing this piece by piece?
21 MR. BAGCHI: Mr. Michelson, please bear with us a 22 little bit. In the margins program, the EPRI margins program 23 anyway, it is purely a deterministic approach. You define
() 24 such, and the equipment in that is well defined. Everybody 25 knows what they are. Every plant will have to develop that HERITAGE REPORTING CORPORATION -- (202)628-4888
L 98 1 list. When you walk down, you walk down the-census path.
'2 MR. MICHEI ;N: Now you have missed my point. My 3 point is how maty failures are you assuming? Hcw many are you 4 going to wipe out when you have an earthquake?
5 CHAIRMAN SISSS: This is the problem--just a minute.
6 Carl is asking aoout the consequenc?s of a seismic, 7 seismically-induced failure of a component system or 8 structure.
9 Seismic margins only determines whether it will 10 fail. If nothing fails at 3 tenths G, if that's your test 11 earthquake, you are home free.
e 12 MR. BAGCHI: That is correct.
- 13 CHAIRMAN SIESS: If something does fail, then you t
^
14 have got to decide whether you are going to do something about 15 it. This does not address that part.
16 HR. 14ICHELSON: Where is that addressed?
17 CHAIRMAN SIESS: That would be a part of the IPE as 18 to whether the utility considered it a vulnerability.
19 MR. MICHELSON: They will integrate all the 20 failures.
l 21 What is the assumption if you decide that a ;
22 particular component can't fail at that G level? Do you j 23 consider that every plant and every component in the plant l l
() 24 that can fail at that same G 1evel fails at the same time? Is 25 that, that would be the worst case? How many, if less than HERITAGE REPORTING CORPORATION -- ('02)628-4888 4
99 LA- 1_ that, hou you decide how many less than that failed?
2 MR. BAGCHI: Hopefully there is only one or two 3 components that are critical. That would be the plant.
4 MR. MICHELSON: You missed the point. What--Chet 5 just explained to me first of all, you decide what can fail.
6 You haven't decided how important it is yet, and you defined 7 that 50 things can fail at that G 1evel.
8 Now in determining how imporcant f a13 nre is at that 9 G 1evel, do you integrate the effecta of all 50?
10 HR. CHOKSHI: I think to answer that question you 11 have to, let's talk about success path. The contention is.any 12 component success path can fail. You assume that successful
( 13 is not--
14 MR. MICHELSON: You will then include all 50 15 failures?
l l 16 MR. CHOKSHI: That's richt.
17 MR. MURPHY: If they are in the path.
18 MR. CHOKSHI; You need to fix the component to 19 make--
20 MR. MICHELSON: That's what you have to do and 21 decide you can fix a few and make some of the other drop out 22 after that point, but you have to start out assuming they all 23 fail.
i
() 24 CHAIRMAN SIESS: That's pretty much what big Rock 25 was going to do--look at the consequences of an earthquake, i
HERITAGE REPORTING CORPORATION -- (202)628-4888
G 100 1 and find'out which was most likely and most important and fix 2 it, and they were going to keep fixing things until they spent 3 a million dollars and stop.
4 MR. CHOKSHI: One shutdown path.
l 5 MR. MICHELSON: In dete'rmining the effect of this 1
6 structural failure that you are looking at, the 50 of them, l
l 7 you have to chase these system interaction effects offall the 8 failures, include that in your PRA?
l 9 CHAIRMAN SIESS: You have got to remember what he 10 told us at the beginning. He has got a committee made up of 11 seismologists and structural engineers, and all they can tell 12 you is it is going to fail and they have got to get the l
l 13 systems people to find out whether it is important.
l l
l 14 MR. MURPHY: We have got a systems man.
15 CHAIRMAN SIESS: You have one systems man?
16 MR. MURPHY: But he is very vocal!
17 CHAIRMAN SIESS: Again, the consequences of whether 18 it is important enough to constitute a vulnerability is 19 something that thn utility is going to decide and tell the 20 staff what they decide.
21 Now when you did Main Yankee, the only thing that 22 came in below the--what do you call the earthquake?
23 MR. MURPHY: Review.
() 24 CHAIRMAN SIESS: Was the tank.
25 MR. CHOKSHI: No, Main Yankee fixed about eight HERITAGE REPORTING CORPORATION -- (202)628-4888
I f' 101 l-i.r s)
. I things before they even finished their margin review, and they
)
2 were given few more itesc but the--caused the plant to be at )
3 .121G. l 4 CHAIRMAN SIESS: They had some other things below 1
4 5 .3.
6 MR. CHOKGHI: They fixed a number of things.
7 MR. MICHELSON: Now what you told me so far sounds !
l i
4 good. When looking for failures, structural type failures, is ]
l )
l 3 there any cutoff point? Do you look at all pipes from quarter j 10 inch on up, or do you have a cutoff in sizes that you look at?
11 Tanks, whatever? Do you look at everything? Is that right?
i 12 MR. bACCHI: Everything is supposed to be included 13 in that. ,
l l 14 M2. PICHELSON: Now when you go to look at things I
.15 like the freon compressor on essential chilled water system, i 16 you are going to look at the copper tubing that in the--the l i
17 failure of that cne piece of copper tubing fails the !
18 compressor Because you are draining the, you use the gas.
19 MR. CHOKSHI: I think--
20 MR. MICHELSON: You look at that level?
21 MR. CHOKSHI: No. On the components like that, if 22 you have enough experience data, the compressor, we have 1
1 23 enough data.
() 24 MR. MICHELSON: You have got seismic data on the 25 compressors? ;
I HERITAGE REPORTING CORPORATION -- (202)628-4888
102 k 1 MR. CHOKSHI: Yes.
2 MR. MICHELSON: For the particular plant?
3 MR. CHOKSHI: Not for the particular plart.
4 MR. MICHELSON: It is the particular plant who puts j
5 the instrument tubing in, not the manufacturer.
6 MR. CHOKSHI: Similar tubing and instrumentation 7 exist at other plancs.
8 MR. MICHELSON: You better think about that a while 9 end look at it and look at craftsmanship. I have scen some 10 that I wouldn't--I can do better. If you crimp a tube, for 11 instance, in the process of putting it in, it is, the 12 vulnerability is quite different. If you don't support it
( 13 properly, it is quite a bit different.
14 I just wondered did you look at that level, though, 15 or just say look at the compressor and the manufacturer's 16 data, whatever it might be? It won't cover that. l 17 MR. CHOKSHI: Wo look at whatever--walk-down is 18 based on experiences like electrical cabinets. We look a lot !
l 19 more in detail when we open up the cabinets on the 20 transformers and things like that. I don't know how far we do !
21 the compressor. Like diesels, ne look at all kind--small l 22 tubing. !
l l
23 CHAIRMAN SIESS: Do you think you can take a given
() 24 plant and if you locked at it at three tenths G, at five 25 tenths G, do you think you could find vulnerabilities at five j l
HERITAGE REPORTING CORPORATION -- (202)628-4888
103 h
\~' 1 tenths G that you didn't find at three tenths G7 2 MR. MURPHY: Yes.
3 MR. CHOKSHI: Yes. ]
4 CHAIRMAN SIESS: There is something that is not a 5 strong--that's a vulnerability. l I
6 MR. CHOKSHI: You will be looking at more 7 components, but it is possible you will define more conoonents )
8 within the capacity of that range. !
9 CHAIRMAN SIESS: Is a vulnerability something you !
l 10 didn't expect or didn't predict? Or if somebody walked !
11 through the plant and says you guys did a lousy job on this, i
12 if you ever had an earthquake, this will fall down, is that a !
13 vulnerability, or unexpected?
l 14 HR. CHOKSHI: There are a number of ways I think one 15 can look at one thing. As you walk down, you find something ;
i 16 which is left out, you fix that.
1 17 Other thing is you go through the analysis, find a )
3 18 core damage frequency and you find there are two or three 19 components which contribute to the core damage frequency. Now ;
20 that does not necessarily mean one vulnerability, but because ]
21 it brings your attention to see if there is anything can be l 22 done, if you want to do it.
23 In that context, I think the margin review, looking
() 24 for the core damage point of view and what are the components 25 important to that scenario, HERITAGE REPORTING CORPORATION -- (202)628-4888
104
'~ 1 MR. MICHELSON: These probabilities in PRA tre based 2 on normal operating experier.ce, not experience under seismic 1:
3 conditions. Your whole picture changes.
4 CHAIRMAN SIESS: He wasn't limited. He was simply 5 looking at fault trees or event trees.
6 MR. MICHELSON: Ending up with the core damage 7 probabilities.
8 MR. CHOKSHI: No. The PRA then, I'll do the core 9 damage frequency.
10 MR. MICHELSON: Look at what the cut sets are.
11 HR. CHOKSHI: Margin, essentially I am looking at
- 12. that, and not really pointing--
13 MR. MICHELSON: You are not looking at the--
14 CHAIRMkN SIESS: Really looking at that.
15 MR. MICHELSON: How do you handle the problem of, 16 for instance, I have a seismic qualified relay which was i
17 tested and showed chatter at certain G 1evels and so forth.
18 Now I am going to do a walk-down in which I am assuming some 19 higher G 1evels.
20 What do you assume about how that relay behaves 21 under those conditions?
22 MR. CHOKSHI: We have a, I don't think we have 23 brought any material, but we have a, the pre-margin method
() 24 which addresses it in general.
25 MR. MICHELSON: Do they know how to extrapolate test HERITAGE REPORTING CORPOR ATION -- (202)628-4888
105
(~') .
ik / ' 1 data to much higher G values in terms of things like relay 2 chatter?
3 MR. CHOKSHI: Yes. Also have to look at systems 4 review to identify relays which will cause the frequencies in 5 the chatter--both aspect of the problems.
6 MR. HICHELSON: Is there a document I can read about 7 how to extrapolate relay, you know--
8 MR. BAGCHI: It is not extrapolating. It is not 9 extrapolating data when it doesn't exist. It is trying to 10 identify what the fragility level is for a certain class of ,
11 relay.
12 MR. MICHEhSON: How do you do that when ycu define 13 fragility as the chatter of the relay?
14 MR. BAGCH1: There are test data.
15 MR. MICHELSON: For the much higher G values?
16 MR. BAGCHI: Yes, sir.
17 MR. MICHELSON: Okay. If they don't have it, what 18 do they do then?
19 MR. BAGCHI: W e l '. , in that case, they will have to 20 replace the relay. That's the extreme bottom line, or in 21 other cases, they would have to show that that relay chatter 22 is not of major consequence.
23 MR. MICHELSON: But they have to show that in terms
() 24 of all relays in a plant they can be chattering at the same 25 time because--as opposed to fire, it a common across the beard J
HERITAGE REPORTING CORPORATION -- (202)628-4888 l
= -_ _ _ _ J
i l} 106
) 1 and they have to chatter all of the relays that are l
2 potentially vulnerable.
3 MR. BAGCHI: One bit of information, they are doing 4 this, this seismic margins program, and I understand that they 5 have spent just about 2700 manhours looking at those relays 6 that could chatter. Now they are narrowing that down to about 7 nine relays.
3 hR. MICHELSON: Assuming simultaneous chatter?
9 MR. BAGCHI: Simultaneous chatter, yes.
10 CHAIRMAN SIESS: When you cet down to 9--I remember 11 seeing down to a couple of dozen.
12 MR. MUti?HY: I think they have reduced it further.
i ) 13 MR. BAGCHI: I called yesterday to find out where 1
1 14 they are.
15 CHAIRMAN SIES3: When we look at components 1
16 equipment that has been qualified by testing, how much of the l
17 qualification has been carried to a higher level of excitation 18 than, than let's say design basis? Hasn't EPRI developed a l
l 19 data base on stuff?
l l 20 MR. PAGCHI: Yes.
l 21 CHAIRMAN SIESS: How much of that goes up into the 22 three tenths G type?
23 MR. BAGCHI: Quite a bit goes up more than that.
(} 24 CHAIRMAN SIESS: You mean equivalent to three tenths 25 G earthquake, which might be five tenths?
l HERITAGE REPORTING CORPORATION -- (202)628-4888
107 1 MR. MURPHY: We have heu quite an extensive 2 component fragility program in the Research Office for the 3 last few years, and a lot of that equipment is qualified up to 4 the half, the half G, the SG and 10G 1evel.
5 CHAIRMAN SIESS: This is the Brookhaven facility?
6 MR. MURPHY: It is the Brookhaven work.
7 CHAIRMAN SIESS: Some of that is actual testing?
8 Some is loohing at existing data?
9 MR. MURPHY: Yes, sir.
10 CHAIRMAN SIESS: And most is electrical equipment?
11 MR. MURPHY: At this stage.
12 MR. MICHELSON: Most of this work has been done with 13 safety-related relays, too, hasn't it? All the data, I think 14 most of your testing has been only on safety-related 15 components.
16 MR. BAGCHI: That is correct.
17 MR. MICHELSON: Now how do you do that evaluation of 18 plant response when you are dealing with non-safety-related 19 relays? And as an example, mercury wetted relays such as they 20 found at Peach Bottom recently which turns out drips on the 21 diesel generator if you shake the relay because it was 22 supposed to trip the diesel in order to have the CO2 properly 23 mitigate fires in the diesel, so as of the seismic event, you
() 24 trip all four diesels and turn on all the CO2.
25 MR. BAGCHI: That's an outlier.
HERITAGE RE"ORTING CORPORATION -- (202)628-4888 L --- . _ _ _ _ _ - _ _ _ _ _ _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ - - - - _ _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ -
I 108
.fi kJ l' MR.'MICHELSON: How do we find those kind of things !
2 which are perhaps much more--
3 CHAIRMAN SIESS: They found it, Carl.
j 4 MR. MICHELSON: Yes, one case.
1 5 MR. MURPHY: How many cases were there? l 1
i 6 MR. MICHELSON: There are at least two I read.
i i i- ^/ .MR. BAGCHI: That is really the hard question. I l I
8 would not have believed that mercury controlled relays are 9 still being used in a nuclear power plant, but there are. )
10 MR. MICHELSON: I wouldn't have, either. Now how do 11 we get those out by this process?
12 MR. BAGCHI: By looking in detail. I think you will 13 be impressed if you look at the study. They have spent 2700 1'
14 manhours looking for those types of relays.
I 15 MR. MICHELSON: This will do the comparable level of 16 examination?
17 DR. SHEWMON: You get--
1 18 MR. BAGCHI: With that kind of level of examination, 19 I don't think too many projects are going to be looked at. It 20 is going to have to be more streamlined and more directed, 21 more focused. I don't know how we are going to get there, but 22 it is a big concern right now.
25 DR. SHEWMON: 2700 man powers being spent on it, I
() 24 would feel a little bit better if you tell me that you felt 25 that they had come up with criterion that were indeed going l HERITAGE REPORTING CORPORATION -- (202)628-4888
109
- 1 through end finding the mercury relays, or saying that indeed !
2 do'you.have these, and if you do, will you get rid of them and I 3 prove that they aren't in the sense in the circuits? Is that i 4 also a corollary of what you say about the 2700?
5 MR. BAGCHI: I just wanted to share with you what 6 kind of effort is going into it.
7 DR. SHEWMON: You have--would you answer my question 6 now?
9 MR. BAGCHI: Yee, Absolutely; it is an obvious 10 question. We are going ta: Ecck for outliers. We are going.to 11 try to develop methods with 13ss than zero in on the critical 12 components and outlaws things that should not be there.
13 DR. SHEWMON: That will come out of the 2700 hour0.0313 days <br />0.75 hours <br />0.00446 weeks <br />0.00103 months <br /> 14 study and be here by the end of the year or something?
15 MR. BAGCHI: It depends. Everything is a 16 compromise. We cannot a priori tell you exactly what kind of 17 criteria we will end up with. We will try to look for these 18 things and you will see what.
19 DR. SHEWHON: This is somebody else spent the 2700 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br />. What manyear or whatever it is, what are they going to 21 come up with?
22 MR. BAGCHI: You cannot--generally if you go through 23 this process, you will find all the outliers. We will do the
() 24 best we can.
25 CHAIRMAN SIESS: Why should you be doing the best HERITAGE REPORTING CORPORATION -- (202)628-4888
110 o r l
" (~ -- 1 you can and not the utility?
l 2 MR. CHOKSHI: I guess your question, in part there 3 are w number of studies going on and the type of, for example I
4 might replace certain relays if you find them critical. That 5 is already being recommended for some relays.
l 6 CHAIRMAN SIESS: I'm not sure I heard all of that. l 7 When they looked at relay chatter, aren't mercury relays just l l I 8 automatically--
9 MR. CHOKSHI: Supposed to be out. Supposed to l 10 exist.
11 CHAIRMAN SIESS: We ought to put a shaker on every 12 one of these plants. We could finally put the--
(~
I -
13 MR. MICHELSON: That's what a seismic event is is a l
l 14 shaker.
15 CHAIRMAN SIESS: Not everywhere.
16 MR. MICHELSON: How much--you have looked at heating 17 and ventilating systems. Walk-downs are going to have people l l l l
18 experienced in the damper and other kinds of arrangements, you j 19 often find some of which trip under seismic events.
20 MR. CHOKSHI: It depends on the, what the system l 21 people tell the people. If they tell air systems and the HSE i 22 is important from a systems point of view, then the walk-downs 23 will be, will include those systems dampers.
() 24 MR. MICHELSON: Seismic traps and things, cine of 25 them I saw at one time several years ago was on a diesel HERITAGE REPORTING CORPORATION -- (202)628-4888
'4, 111
. , ~ .
_\ '
1 engine air intake, little mousetrap arrangement, which-was
'2 with'the fire protection, which was tripped by fire protection 3 signal. It turned out it was quite obvious it was not a 4 seismically-qualified mousetrap.
5 MR. CHOKSHI: We look at those kind of things 6 -identified by system people that is necessary.
7 CHAIRMAN SIESS: 1,et me ask the general question. I 8 realize that there has been references to plant unique 9 vulnerabilities, but I'm not sure they really meant that 10 literally. One plant may be involved.
11 As the various licensees go through these studies, 12 one type or another, in the IPE or IPEEE, and find 13 vulnerabilities, what are you going to do or what is the 14 industry going to do to promulgate that information to the 15 other utilities?
16 Somebody may have a pretty good system to pick up a 17 vulnerability. The other plant is doing it a little bit 18- different way, doesn't pick it up. If they went back and 19 look, you would find it.
1 20 MR. BECKNER: We have identified the need for that, 21 and the staff would serve the clearing function for that type 22 of information.
23 CHAIRMAN SIESS: The staff would not? NUMARC or j i
() 24 EPRI?
25 MR. BECKNER: NUMARC as well would do that, too, but HERITAGE REPORTING CORPORATION -- (202)628-4888
l 1 l
112 {
[ . _
k/ 1 at this point the staff will be performing that function. ,
l 2 MR. BAGCHI: Information notices aro quite prolific.
3 We use that.
4 CHAIRMAN SIESS: It takes you too much to get one 5 out. I bet you EPRI can get them out a lot faster.
6 MR. BAGCHI: Information notice gets out very 7 quickly. Generic letters take more time.
8 CHAIRMAN SIESS: Okay. They don't have to go to 9 CRGR.
10 MR. BAGCHI: Right.
11 MR. MURPHY: Okay. Turn to the.last item on this 12 vugraph, and that's simply the identification and correction 13 of vulnerabilities is incumbent upon the utility, meaning they 14 will define what is a vulnerability, items that nominally will 15 fall below the review level that has been set by the staff.
16 The utility will decide whether or not that's a vulnerability.
17 In making that decision, they will notify us, and they will 18 also tell us whether they are going to take any correction, 19 corrective action about that.
20 If the utility doesn't carry the, out a correction 21 under their own initiative and the staff feels that it is a 22- significant enough safety issue, the staff then takes the 23 initiative under the backfit ru3e, and the burden then falls
() 24 on the staff to carry that out.
-25 CHAIRMAN SIESS: One thing, inevitably the different HERITAGE REPORTING CORPORATION -- (202)628-4888
i L l 11'3 l
. , ~ l t w'1 utilities are going to have different thresholds.
1
- 1. That is one j l 1 2 thing that is inevitable for either identifying and/or 3 correcting vulnerabilities.
4 MR. MURPHY: That is correct.
5 CHAIRMAN SIESS: The staff will recognize that 6 somebody is particularly sensitive and wants to fix things at
-7 some level. They won't impose that on everybody?
8 MR. MURPHY: That's the decision.
9 CHAIRMAN SIESS: You can't impose anything on 10 anybody. You really can't impose anything on anybody?
11 MR. BAGCHI: That is correct.
12 MR. MURPFY: That is correct.
13 CHAIRMAN SIESS: You are implementing a policy 14 statement by the regulation.
15 MR. BAGCHI: Exactly.
16 CHAIRMAN SIESS: I remember the SEP on one of the 17 plants, there was something the SEP group said that that is 18 not enough risk to be a real problem, the utility went ahead 19 and fixed it anyway.
20 MR. MURPHY: That is correct. I mean the case of 21 Main Yankee is a good one in point. It is a good utility.
22 They went through, and as Chokshi indicated, as they did the 23 review, they found items that were questionable. They went
() 24 ahead and fixed them, so that we ended up with a very few 25 items on the list when it finally came out at the final list.
HERITAGE REPORTING CORPORATION -- (202)628-4888
e - -
114
(,
'> 1 The water tank is a case. They selected to go ahead and fix 2 that.
3 CHAIRMAN SIESS: Most of-those~ items were probably 4 easy to fix, too, waren't they?
5 MR. MURPHY: Yes, sir. A lot of them were 6 inexpensive.
7 CHAIRMAN SIESS: Now the water tank.was not a cheap 8 one to fix.
9 MR. MURPHY: No, it was not, but'that utility made 10 the decision that it was better to have that extra 6 tenths, 11- 600ths of a G, so they went ahead and they fixed it.
12 CHAIRMAN SIESS: I think the water tank is easy to
/~ .
\- 13 fix, but I don't think it is cheap.
14 MR. MURPHY: A lot of manhours.
15 MR. BAGCHI: Was $200,000 is what we understood it.
16 It is not cheap.
17' CHAIRMAN SIESS: That's cheap.
18 MR. MICHELSON: In doing this beyond design basis 19 events, you may find that you have a problem that needs 20 correction that should have been fixed even within the design j 21 basis. It didn't even meet the design basis requirement.
22 CHAIRMAN SIESS: That's clear. They told you 23 earlier.
L 1
() 24 MR. MICHELSON: Wait a minute. Let me ask the 25 question. If you find that there is a vulnerability created j
b- HERITAGE REPORTING CORPORATION -- (202)628-4888 L - ___ _ - _ _ _ - - - -
115 e
's / 1 by a deficiency in the design that was within the original j 1
2 design basis, how is that handled? Is that handled as the 3 deficiency reports process? !
I 4 MR. BECKNER: That's an enforcement issue. l 5 MR. MICHELSON: It is processed by the normal. It i 6 is not a backfit? ,
i 7 MR. BECKNER: It is not a backfit.
8 MR. MICHELSON: If you find the vulnerability is i 9 created by a beyond the design basis requirement on the l 10 component, then it, you thought it needed fi: ting, that would l
11 become a backfit? Is that the way you look at it?
12 MR. BECKNER: That is correct.
<3 k 13 MR. MICHELSON: Okay.
14 MR. MURPHY: And to simplify or to confuse the l 15 issue, I think it probably is inappropriate to lock at those 16 enforcement issues as vulnerabilities. Those are within the 17 design basis and will be, definitely be handled that way.
18 MR. MICHELSON: They were just deficiencies not 19 recognized until the time you did the beyond design basis 20 examination.
21 CHAIRMAN SIESS: What complicates it is there are 22 going to be plants that meet their design basis but the design 23 basis has changed since they were built.
() 24 MR. MICHELSON: Then--
25 CHAIRMAN SIESS: That is still a backfit?
HERITAGE REPORTING CORPORATION -- (202)628-4888
l L 116 l '. . ,, -
k- 1 MR. MICHELSON: Yes. Some enforcement issues, it is 2 a backfit or isn't, but that, as that's a part of the 3 enforcement issue process as to whether it is a backfit or 4 not.
5 MR. DECKNER: I think if they meet the design basis
.6 that is on their license, anything beyond that is a backfit 7 CHAIRMAN SIESS: Yes. That's a backfit, unless you 8 can find that there is not adequate protection, in which case I
9 it is not a backfit. At least that's what the lawyers tell j 10 us.
'11 Have you got lawyers working for you?
12 MR. BECKNER: No.
13 MR. MURPHY: There's enough confusion without them!
14 The next vugraph lists a number of the issues that the 15 Subcommittee has decided to take a look at. These are related 16 to the seismic design margin program, which in the second 17 bullet, is the SDMP. That's what that stands for.
18 There are a number of issues. The first one that is 19 listed here is the question about the high frequency ground 20 motion that has been observed in the eastern United States, 21 and there are some question as to how this fed into the 22 standard specter that was used to generate the fragility 23 curves used for the seismic design margins. I
() 24 CHAIRMAN SIESS: What is the high frequency problem?
25 MR. MURPHY: For the--it has been observed in the 4
I HERITAGE REPORTING CORPORATION -- (202)628-4888
117
\- - 1 . eastern Units.d States that'the specter has a higher or is 2 enriched in high frequency ground motion as contrasted to that 3 specified by.the specter used in--
4 CHAIRMAN SIESS: We have discovered, we have 5 discovered in the past few years that the specter we have been p 6 using is wrong?
7 MR. MURPHY: It can be improved. Let's put it this 8 way, and that's what we are saying. We are looking into the 9 consequence of having enrichment in higher frequency ground 10 motion in the specter that was used to set up the fragility 11 tables used in the design margins methodology.
12 CHAIRMAN SIESS: Is this tied to relay chatter or to i
13 anything else?
14 MR. MURPHY: It is tied to the whole thing, not just 15 relay chatter.
16 CHAIRMAN SIESS: There are other component 17 structures of systems that are sensitive to high frequency.
18 MR. BAGCHI: I don't know if I can explain this any 19 better than this. It is the concern that you have a high 20 frequency motion that knocks the system down to a lower 21 frequency. Then you have power in the earthquake to cause 22 further damage.
23 CHAIRMAN SIESS: Okay. I remember that. How high
() 24 is high in this case?
25 MR. REITER: It depends on the earthquake level.
HERITAGE REPORTING CORPORATION -- (202)628-4888
L l
118 i.
~
'O
'\ / 1 There is--for the-large earthquakes, we are talking about w
2 'something greater than 5 hertz; small earthquakes, talking 3 about 5 hertz.
4 I think Andy hit upon the really important point 5' which was that the screening tables that we have been using, 6 assuming that the point, we have .3, .5G screening tables, it 7 has been implicitly assumed that those are tied to the NUREG 8 0098 specter, and one of the things that we are investigating 9 now that indeed in talking with people who put these specters 10 together is that really may not be a correct description. You 11 have some argument that those are really tied to a stronger 12 high frequency ground motion, and you heard some argument of 13 that presented by NUMARC and EPRI, and they are going to 14 present us with a report sometime in September looking at 15 that.
16 CHAIRMAN SIESS: The high frequency component 17 approacher tne zero period?
18 MR. REITER: Right, one of the things that we found.
19 CHAIRMAN SIESS: Do you think there should be more 20 amplification?
21 MR. REITER: What we find, for instance, is that if 22 you, typically if you look at reg guides specter, typically 23 the typical earthquake specter you see is usually lower at
() 24 frequencies, depending where you--let's say greater than, less 25 than 5 to 10 hertz, but it is higher in frequencies greater HERITAGE REPORTING CORPORATION -- (202)628-4888
L 119
,l -- r
- - 1 than 5 to 10' hertz, and associated with that-is where does the i
.2 constant, where do you anchor the spectrum? Once we get these 3 recording instruments out in many eastern earthquakes we are 4 -finding a spectrum greater than 33 hertz.
.5 ' CHAIRMAN SIESS: I'm trying to understand. I have 6 got--let me talk first in terms of the structure. It is up in l
7 the 15, 20 Hertz range, and the concern is that it wasn't-8 designed for a high enough G value, and that as it was 9 designed,.this G could not be exceeded? Right?
- 10. MR. BAGCHI: That is correct.
11 CHAIRMAN SIESS: Which will drop it down into a 12 lower G value and some other component.of the earthquake, that 13 the lower'G value would also be exceeded?
14 MR. BAGCHI: Exactly.
15 CHAIRMAN SIESS: Does that apply to equipment as 16 well as to--I think a structure is, concrete structure is.
17 where the damage will drop the frequency because it will crack 18 it or something, but I think I can't quite visualize machine 19 components.
20 MR. BAGCHI: I think machine components, too, not 21 for machine components, too. That's--Kennedy and R6ed have 22 proposed that if you limit the deformation to something like 23 20 mils, then--and you can absorb the earthquake within that
() 24 deformation limit, then you don't have much cf a damage, so to 25 look at the high frequency effect, they are going to look at HERITAGE REPORTING CORPORATION -- (202)628-4888
.I..
, 120
<~
kTl' 1 the, oh, the limitation of the damage to a certain amount, and 2 if it is less than that, then there is no vulnerability to 3 that.
'4 CHAIRMAN SIESS: To what? What are we talking about 5 damage?
6 MR. BAGCHI: Deformation all the way up to 20 mils; 7 this was for machine vibration, pumps, long shaft pumps and 8 things like that.
9 CHAIRMAN SIESS: Okay.
10 MR. MURPHY: Okay. The next item is the same sort 11 of thing. The fragility tables generated are used with the 12 design margins methodology. Implicitly I had a, a limited 13 mind of a, I believe it was 6.5, thereabouts. We were 14 concerned for the eastern United States, the possibility of 15 earthquakes larger than 6.5, 7, 7.52, 7.5, and we were 16 concerned whether or not there was a limitation on the 17 fragility tables given a little bit larger earthquake.
18 This is principia11y a question of the duration of 19 the ground shaking. It is one of the items at issue that we 20 have taken and looked at to see whether or not there is an i
21 impact upon the fragility tables themselves. !
22 The next item was of use to us in trying to come up 23 with criteria for selecting the bins. We were interested to
() 24 see if there was some generic relationship between the HCLPF 25 levels established for a plant, and the hazard or core damage l
1 HERITAGE REPORTING CORPORATION -- (202)628-4888
121
(~T
\I 1 frequencies. This would help us if there was a generic 12 relationship--
3 CHAIRMAN SIESS: You are speaking in the past tense.
-4 HCLPF, have there been HCLPF levels assigned to plants other 5 than the-one or two they did?
6 MR. MURPHY: There have been a few assigned to the 7 plants that we don't, and there have been--
R CHAIRMAN SIESS: Catarba, main Yankee. ;
9 MR. MURPHY: A few conversions made of PRA numbers ;
i 10 to HCLPF.
11 CHAIRMAN SIESS: So what you are saying is a plant 12 where you have had a PRA and a seismic margin study, is there 13 any correlation?
14 MR. MURPHY: Where you had the seismic margin study, i
)
15 if that PRA number was converted to a HCLPF, that was a, as is 16 one case. Where there was a PRA and a, and a margins done, it 17 is another case. Where there were margins done and looked to I 18 see whether or not we could get core damage numbers out of !
i 19 those, those were things that we were interested in doing to 20 see if there was a generic relationship. !
l 21 CHAIRMAN SIESS: How many plents have you got to do l 22 this on? t 23 MR. REITER: A study that we have, Livermore did for
() 24 us, I think we have something like eight or nine plants which ;
)
25 are examining this relationship, and again, Andy points out HERITAGE REPORTING CORPORATION -- (202)628-4888
122
- f) i.s 1 seismic margin'is a ECLPF-related approach, and the idea is we
.2 want to see if, what if any insights would we get in terms of 3 risk or core damage frequency from that, and cne of the things (
4 we are looking at is supposing.we assume ve know what the 5 chance of exceeding HCLPF is? Is there a plant HCLPF, is 6 there-some relationship, generic relationship to what the 7 associated core damage frequency is?
8 CHAIRMAN SIESS: The plant HCLPF is just the point 9 at which the first component will fail, right?
10 MR. REITER: That's right. Yes, success path.
11 CHAIRMAN SIESS: Takes multiple system or component I 12 failures to get you up into a core melt, so--
13 MR. REITER: Again, what we are dealing with is a 14 seismic design margins program that was not specifically put 15 together for severe accident, that it had another purpose. We 16 have a lot of--
17 CHAIRMAN SIESS: It was put together for severe 18 accidents. It just wasn't put together for probability l
19 analysis.
20 MR. REITER: Originally seismic design raargin 21 program was thought up, the idea of supposing we seismologists 22 came up-and said hey, the design basis is wrong? Instead of 23 be .17G, it should be .22G. We are trying to expand it. We l
O 24 think it is a good vehicle, good way for people to learn with l
25 the plant, and one of the questions always raised is what does i
i l HERITAGE REPORTING CORPORATION -- (202)628-4888 !
I
p 123 ps
1 that'mean in terms of risk or core damage? And this, what we 2 are trying to do is explore that, get insights as to what that 3 means. We are getting some insights.
4 CHAIRMAN SIESS: Is there a relationship?
5 MR. REITER: There is a very general relationship, j
'6 very general, with a large scatter around it.
7 CHAIRMAN SIESS: Wouldn't expect it to be much 8 relation between the design basis earthquake, and the hazard.
9 I'm not sure the HCLPF--
10 MR. REITER: I'm not talking about design basis.
11' CHAIRMAN SIESS: I know it. I'm not sure the 12 HCLPF--
j .3 MR. REITER: There is a scatter relationship between 14 HCLPF and design basis, and that's, that's one of the lessons 15 we are learning, and there is a--but there seems to be a 16 generic relationship between HCLPF or chance of exceeding that X
[ 17 HCLPF and core damage. It is not as robust as we might like l
l , 18 it, but it is, there is a generalized relationship.
L l
19 MR. BAGCHI: Calculating--let us try. Given that 20 the plant HCLPF is exceeded, what is the chance that there 21 will be a core damage? I think that kind of--
22 CHAIRMAN SIESS: It will be less than the 9595, l'
- 23 because that's only one component.
() 24 MR, BAGCHI: They are finding that there is one in 25 perhaps ten chances or one in five, that kind of range.
l
[ HERITAGE REPORTING CORPORATION -- (202)628-4888 w________---_______-- - >
124
.f)/
\_ 1 CHAIRMAN SIESS: You see, if you fixed your 2 vulnerabilities, first component that failed., you would 3 probably fixed the second one up. Now you have got'it further 4 down the line.
5 MR. BAGCHI: That is correct.
6 CHAIRMAN SIESS: That's interesting.
7 MR. MURPHY: Okay.
8 CHAIRMAN SIESS: I'm not sure how valuable-it is,
.9 'but it's interesting.
10 .MR. MURPHY: That's the tact we took with it. It 11 was a thought process of coning up with a justification for 12 making selections on the binnings. Some of these things are
(
13 going to be good. Some of them aren't going to work.
14 The next item that we looked at with the seismic 15 design margin methodology was the question about containment, 16 containment performance.
17 Quite frankly, it is not considered within the 18 design margins methodology. It's something that we definitely 19 want to take a look at within the severe accidents, so we 20 looked at it to see whether or not it was possible using a PRA 21 insight to extend it.
22 CHAIRMAN SIESS: Hold up a minute. This is 23 containment performance in an earthquake or containment
() 24 performance in a severe accident? Containment performance in 25 earthquake is included in the ESP? Fragilities for I
HERITAGE REPORTING CORPORATION -- (202)628-4888
_ _ _ _ _ - _ - - . - _ _ i
y m.- <
4 ;
1 .i 125 p,) ; W o '
M(- .
1 containments?
il MR. CHOKSHI: Screening table in the margin rule out fd 3 containment. figures--strong enough.
4 CHAIRMP.N SIESS: It was structurally strong. That 5 .ain't the way containments are going, a!
6 MR. CHOKSHI: Functions are always ruled out because
'7- they are, in effect won't affect core damages frequency. What 4
8 we are doing is the containment as a. mitigating' system, but 9 earthquake in use is severe accident.
10 CHAIRMAN SIESS: The containment isn't going to fail 11 structurally.
12 MR. CHOKSHI: No. It is a function.
/
('
13 CHAIRMAN SIESS: It is a function, is going to be 14 the penetration doesn't close or something else.
15 MR. CHOKSHI: That's--we have a presentation for 16 you.
17 CHAIRMAN SIESS: So containment here is the whole 18 concept of what you do to keep the--
19 MR. CHOKSHI: Mitigation.
1 l
20 MR. IfURPHY: Right, to keep it inside.
21 CHAIRMAN SIESS: Why does the MARK I ice condenser i
122 need special attention? They lower pressure, I mean lower l l
23 pressure. ;
8
() 24 MR. BAGCHI: We simply don't have the PRA types for 25 those types of containment.
\
HERITAGE REPORTING CORPORATION -- (202)628-4888
= x_ - __.
l 126. l J '7 l 35
/ 1 (Slide)
S 2 MR. MURPHY: The first item is a question, not a 4 3. statement, and that is do we need core damage frequencies? Do 4 we.need specific. frequency numbers in order to carry out an
- 5. IPEEE? Do we have to have the numbers in effect generated by l ) ,'
6 a seismic design margin methodology which they do not at this 7 time? And.the guidance that we have been getting on so far is 8' that we do not need specific core damage frequencies in order j 9 to successfully carry.out a vulnerability search in the 30 seismic area.
l 11. -CHATRMAN SIESS: That is saying you don't 12 necessarily have to define a vulnerability in terms of core i-LJ 33 damage frequency?
14 MR. BAGCHI: Right.
15 MR. MURPHY: That is correct.
16 MR. SHAO: The uncertainty is still large.
17 CHAIRMAN SIESS: I don't know why the uncertainties 18 margin is everything. I'll admit it is worse than seismic, 19 but--
20 HR. MURPHY: We want to carry that thought a step 1
21 further in that we didn't feel that it was appropriate to 22 generate these numbers and then use them in an absolute sense
\.
23 to make bottom line comparison across the board with the other 24 vulnerabilities and fragilities that are discovered by this
( )-
l 25 program.
,- HERITAGE REPORTING CORPORATION -- (202)628-4888
127 p/
\- 1 CHAIRMAN SIESS: That sounds reasonable, but 2 insights, you know, this is titled insights, and insights to 3 me means more than core' damage frequency calculations.
4 MR. MURPHY That's fine. That we begin to pick up.
5 You.want to say in the next bullet that we are interested in 6 carrying these things out, making comparisons to gain the 7 insights that are available to us.
8 HR. CHOKSHI: I think the plan demonstrates 9 difference in insights, containment and that sort of questions 10 in qualitative sense.
11 MR. MURPHY: Okay. One of the important things that 12 the staff has to realize is that the, looking at the third 13 bullet here, is that if we do carry out a seismic design 14 margins methodology, we will not get the core damage frequency 15 numbers. We will not have that kind of information available 16 to us if it is necessary to go into a backfit scenario with a 17 particular utility that doesn't want to fix an important or a l
18 safety significant vulnerability.
19 Tne next, the next item simply tells us that because 20 of the way the EPRI design margin methodology is carried out, 21 it is more difficult to get core damage frequencies to get 22 insights into the risk significance.
23 CHAIRMAN SIESS: You know, I don't know that we ever l
() 24 heard or certainly not recently heard a discussion of the EPRI
( 25 seismic margins versus the staff seismic margins. I thought i
HERITAGE REPORTING CORPORATION -- (202)62F-4888 )
l' h 128 l { {k}' 1 they differed only'in the hazards analysis that was done by t.
2 different expert opinicn groups.
L 3 Are there basic differences?
L 4 MR. MURPHY: There art other differences, and if we 5 can wait a while, that will be presented in two discussions 6 from now. Next we will have--
7 CHAIRMAN SIESS: We will be glad to wait.
l L 8 MR. MURPHY: At this stage, unless there are further 9 questions, that concludes my overview, and'next we will have 10 Dr. Chokshi talking about the issues involved with the seismic 11 program.
12 CHAIRMAN SIESS: Okay. You covered your overview
_7
/ 13 and you went into some on the seismic margins stuff. Chokshi 14 is going to talk about the PRA approach and then Kenneally J15 about the seismic design margins approach? Thore are the two 16 parallel approaches to the seismic?
17 MR. MURPHY: That is correct, ano under the margins
- n. , ' 18 appror.ch, we will take a look at both the EPRI and the NRC, 19 Livermore approach.
20 CHAIRMAN SIE3S: Okay, and then the containment h 21 issues?
27 MR. MURPHY: With Kenneally again..
23 CHAIRMAN SIESS: Okay. If it hasn't been clear E
) 24 before, I gue.as I should mention that we are not going to 25 finish up at 12:30 today. I don't know who conspired with I
O HERITAGE REPORTING CORPORATION -- (202)628-4888
- w -9,
- i , 129-k/. l' that~ schedule. I suspect it was somewhere between Mr. Bauchi 2 I and Mr. Igne. The thing is to show the proper degree of 3 pessimism for this business, so the question is what would be 4- a good time to take lunch?
5 MR. SHAO: The staff is willing to stay until around 6 4:30.
7 STAFF: Not for lunch!
8 DR. SIESS: I don't think we can finish in another 9 hour, so we might as well consider lunch.
10 MR. SHAO: Okay.
11 CHAIRMAN SIESS: And my question right.now is should 12 it be before or after Chokshi's presentation?
O
(> 13 MR. MURPHY: Chokshi's presentation was scheduled 14 for 10 minutes.
15 CHAIRMAN SIESS: Okay. Let's--
16 DR. SHEWMON: So was Larry Shao's.
17 MR. MURPHY: Fo was mine.
18 MR. REITER: Put the challenge to the ACRS 19 Committee!
20 CHAIRMAN SIESS: Let's hear Chokshi then, and we 21 will get lunch in between the PRA and seismic design margin.
22 MR. CHOKSHI: I think what I will talk about is the 23 Subcommittee's thinking of making some recommendations when
() 24 the utility decides to use a PRA approach in the context of 25 IPEEE, and PRA approach we are structuring around the family HERITAGE REPORTING CORPORATION -- (202)628-4888
h , I i+ ; ~ ; e x . F 1
130
- Fr~ :..
at i h0 ll h)' ' i 1.i . objective, and that is the objective is the plant-specific l i
%T ..
~'
i l 2 examination to R and D vulnerabilities and understand manned j g ;3 -response, de-emphasizing bottor.: line numbers, and we have some L , 4 assumptions and some solutions and we would like your comments 5 .and reactions to what, what our recommendation, this tentative 6 recommendations.
~7 (Slide)
/
8 HR. CHOKSHI: Or they are not recommendations at 1 1
( , 9 this. P.ime ,
10 There are two approaches which have been used i 11 widely, and neither one is acceptable, the one being used by 12 the industry, and other used by the NRC-sponsored PRAs.
(-
13 Example of that is the NUREG 1150.
.14 CHAIRMAN SIEGS: You call NUREG 1150 simplified.
- 15. MR. CHOKSHI: I put in the quote marks. That's how 16 it was characterized. It does not necessarily mean that.
17 MR. MICHELSON: That's what you mean by simplified? ;
18 MR. CHOKSHI: Because that's what it was called in 19 -1150, and I think in the same range I think we also are 20 talking about, a little later about the use of existing PRAs.
21 I think the next two bullets are the one, I think the 22 statements we are trying to--I think that when utility doing a 23 new PRA, they may choose to do a full uncertainty analysis and I) 24 full quantification, but at least from the Subcommittee point 25 of view, that's really not necessary, but you can do a point HERITAGE REPORTING CORPORATION -- (202)628-4888
331 l l
('s
\
,1 estimate type of calculations for the type of information 2 . which is pertinent in terms of the policy, and the reason, in 3 the next slides I will go and say why that we feel that that's 4 an important point, and some of the things, we have already l
5 discussed the walk-down, and we think that the seismic
'6 walk-down in PRA also has to be the similar to the one 7 prescribeed in the margin methodologies, and in the past 8 number of periods did not include relay chatter, and we think 9 that they have raised that issue. Given these enhancements, 10 the last enhancements, I think that those even enhancements 11- are needed in a PRA approach.
12 (Slide) 13 MR. CHOKSHI: I guess the question is now about the 14 point estimates and why we think that it will depend on a l
15 point estimate--it will depend on point estimate type of 16 calculation, and I think this is where we probably like your 17 comments.
18 There are some we think that the point estimate 19 will--still does the, you know, the familiar items., the 20 objective of identifying the seismic areas, and it is going to 21 de-emphasize on the bottom line numbers. It can still allott 22 us to rank, categorize. the dominant sequences, which' margin 23 approach does not let you do that, and I'll show you NUREG/CR
() '24 1150, I mean the NUREG 1150 analysis tahich shows what this, 25 when you have a dif f er ent hazards, what it does do in terms of HERITAGE R', PORTING CORPORATION -- (202)628-4888
132 A
k- 1 insights other than numbers, j 2 This will also highlight the problem of or 1
3 temptation of comparing the core damage frequencies from other 4 initiators and then claiming that seismic is high, low, and s
5 you know, instead of finding them--arguments of that nature 6 .rather than looking at what you are learning from the ;
I I
7 analysis.
8 It also will hopefully award on the selec'ing hazard li 9 curve because it is not really that crucial. The bottom line )
10 numbers are no longer of that much importance as the, as 11 before, and I think benn is going to talk about the type of 'l 12 hazard one night select for a point estimation type of
( 13 analysis.
14 CHAIRMAN SIESS: It seems to me that getting 15 yourself all tied up with uncertainties when you start off 16 with a hazard curve that we have, which embodies the sum total 17 of ignorance of a seismologist, what else?
18 MR. REITER: Some people look at a glass and say it 19 is half full, and some people say it is empty.
20 CHAIRMAN SIESS: Now the uncertainties and the 21 fragilities do get introduced fairly specifically in the i
22 seitemic design margins thing. They also introduce the PRA 23 approach.
.( ) 24 MR. CHOKSHT: Right.
25 CHAIP. MAN SIESS: But again, they represent another HERITAGE REPORTING CORPORATION -- (202)628-4888
t 133 (m_
t'
'l level of ignorance which is fairly broad, and I think, you 1
1 2 know, if you can't find your vulnerabilities using point 3 estimates, I don't know how you are, how you are going to do a 4 better job.
5 MR. CHOKSHI: Yes--just'doing full uncertainty 6 doesn't gets you better information.
, 7 ' CHAIRMAN SIESS: Because I tried to point out at the 8 beginning, it is not the objective to show that earthquakes 9 are or are not a dominant contributor to risk. Some people 10 think they are. In some places, they surely are; in some, 11 not.
12 The important thing is what we are trying to do here !
L O 13 is to find plant-specific vulnerabilities, preferably those 1
14 that are easily fixed.
1 l 15 MR. CHOKSHI: Yes, and that's the thrust of the type l l \
l 16 of recommendation we are making is to avoid precisely those l l
l 17 kind of complaints, argument, about which hazard curve is 18 right and which is wrong, and I think we get, we still get a J l
19 lot of benefits which you get from the PRA analysis like ]
i 20 ranking the sequences, and being able to evaluate the 21 importance in terms of risk selections and other measures, and l 22 I think the last bull bullet was the point which you made, Dr.
23 Siess, that unc' ainty hazard goverr.s the final numbers so
() 24 much that making small changes in a plant component does not 25 reflect, you know, does not reflect in any appreciable sense, HERITAGE REPORTING CORPORATION -- (202)628-4888 )
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___ J
134 v 1 and--
I 2 MR. MICHELSON: Would you clear up a problem that I 3 have right now with PRA?
4 In the case of, in the case of what--we have a 5 certain probability that a type of a given size is going to 6 fail. You start that out as an initiating event. We go 7 through and say what the probabi3ity of that resulting in core 8 melt might be.
9 What do you do in the case of a seismic event?
10 First of all, you know what probability experience in a 11 seismic event, some idea. That's the initiating event.
12 The consequence of a seismic event, however, might 13 he the failure--let's say you have six pipes in a plant. I 14 These probably are all non-seismically designed, perhaps, 15 varying sizes, but probably smaller pipes, not bigger pipes, 16 but six of them fail.
17 Does the PRA, have you have seen any PRAs in which 18 those consequential effects of the earthquake are included as i
19 a part of initiating event for the trees and the flooding I l
20 effects and the water effects, whatever, all those chase down 21 through the tree?
l 22 MR. CHOKSHI: There is varying degree of PRA. Some l 23 FRAs have greater and in more detail, have treated--
l
() 24 MR. MICHELSON: Have you ever seen one in which all l 25 consequential failures of pipes due to an earthquake are HERITAGE LEPORTING CORPORATION -- (202)628-4888 )
l
135 (Q s. . 1 included concurrently?
2 MR. CHOKSHI: Generally the pipe failures in 3 earthquake do not show up as a major contributor.
4 MR. MICHELSON: Don't know what it.is. I don't 5 know.
6 CHAIRMAN SIESS: Pipes don't fail in 9arthquakes.
7 Pipes don't fail in earthquakes.
8 MR. MICHELSON: Wait a minute now. Whose basis, 9 what basis is there to believe that a pipe doesn't fail? j 10 CHAIRMAN SIESS: It doesn't--lots of earthquakes and l 11 tots of pipes.
12 MR. MICHELSON: We are talking about all sizes now, ')
13 Chet, not just big 20 inch. I won't argue 20 inch. I will :
14 argue 1 inch, 2 inch, 3 inch.
l 15 CHAIRMAN SIESS: We have gone out and looked at le plant after an earthquake, and they don't find any pipes that 17 fail.
10 MR. MICHELSON: You have walked down and found 19 equipment that fails and falls on pipes.
20 CHAIRMAN SIESS: Might make it, j 21 MR. MICHELSON: You are backing orf from your j 23 statement already.
23 DR. SHEWMON: He is saying when you go look, you !
() 24 don't find that earthquakes fail pines. That you
(
25 failures--what you can hypothesize he can't defend hirself ;
HZRITAGE REPORTING CORPORATION -- (202)628-4888
. . l
1 l
1 136 t
. (-
f^) - 1 against, but his statement is still valid.
2 fMR . KELLY: Dan Kelly from the staff--in the PRAs i 3 snd the margin studies we have looked at the earthquake 1
1 4 experience data, and most welded pipe, you know, hadn't filed.
5 It is only a case where you have an anchorage failure of a !
6 tank, slides, breaks it.
7 MR. MICHELSON: This is irrespective of size, not 8 seismically designed?
9 MR. GUZY: Yes.
10 MR. MICHELSON: That's a pretty interesting 11 statement.
12 MR. GUZY: What you are concerned about is threaded
( 13 piping, fire from piping that has to be--and cases where a, 14 you know, a like a tank or something breaks and fails or 15 interacts on a pipe, but there has been no inertial cause 16 failure also of piping. The same set of plants experience 17 base used for A-16--they just don't fail is our experience.
18 MR. MICHELSON: There really is no reason to 19 seismically design piping then, is there?
20 MR. GUZY: Some people make that argument.
21 CHAIRMAN SIESS: Fome of us have been trying to say 22 that for 20 years.
23 MR. MICHELSON: How about the system interaction?
() 24 How about the A-46 interaction effects? In it assured that 25 all those are fixed under this program? Where you have a HERITAGE REPORTING CORPORATION -- (202)628-4888
137
./
~ f' 1 component which you know the anchorage is weak and it. fails, 7, c 2' and it impacts on say a 2 inch pipe, that can cause 2 inch 3 pipe to fail, are those included?
4 MR. GUZY: I am not sure I quite understand your 5 question.
6 CHAIRMAN SIESS: Falls over.
7 MR. GUZY: Interactions effects are considered in 8 the margins approach. They are not, piping is not considered 9 in A-46, but that's one of the differences between A-46 and 10 the margin.
11 MR. MICHELSON: You are saying free standing piping, 12 that has no impact potential, will not-fail even if it is 13 non-seismic, irrespective of size?
14 MR. GUZY: Irrespective of size, welded piping above 15 ground.
16 MR. MICHELSON: I have read the flood work, followed 17 it pretty closely. I never realized they included that, 18 included any size. I thought they specifically excluded 19 certain sizes. In their analysis they were looking at the big 20 stuff, 6 inches and over.
21 MR. GUZY: Small pipe is welded.
22 MR. MICHELSON: Is that documented in some place?
23 MR. GUZY: Yes. We have--
() 24 MR. MICHELSON: Where do I read it?
25 MR. GUZY: Well, NRC has a, we have an addendum to HERITAGE REPORTING CORPORATION -- (202)628-4888
_ _ _ _ _ _ _ _ _ . . _ f
138 r}~
k/ 1 1061, the addendum, Volume 2 of that report. There is a, )
i 2 there has been an EPRI EQE study that I don't know the number 3 on it, but there is a summary of piping experience data, and-- )
i 4
4 CHAIRMAN SIESS: Where don't you provide him a list 5 of those in writing, and it might be worthwhile to include the i
6 E tech stuff on how pipes do' fail, j 7 MR. BAGCHI: Is it.possible that Mr. Michelson is 8 perhaps thinking of something else? A-46, there was a basic 9 assumption that large LOCA is not considered simultaneously l 10 with the equipment failure, that large LOCA comes from damage 11 in piping. Small LOCA is considered for equipment selection.
12 MR. MICHELSON: I realize that.
13 MR. CHOKSHI: The same thing with margin.
14 MR. BAGCHI: So in A-46, although they didn't look 15 at the, don't look at piping, but they do look at interaction.
16 MR. JAMALI: I would like to make a comment. My 17 name is Kamar Jamali. I am a PRA specialist if you will. I 18 work for NUS Corporation.
19 Now we have done seismic PRAs. On a number of 20 studies, namely, for limiters and so forth, that the, you 21 have, Mr. Michelson, system interaction type of things, are 22 considered in the PRA approach. This is the type of thing you 23 try to deal with when constructing entry, so system
() 24 interaction in pipe failures, what not, of all kinds, will be, 25 are attempted to be looked at in a PRA approach.
l HERITAGE REPORTING CORPORATION -- (202)628-4888 L___-__-___-_-_____________-
139
<s
(' J- 1 Now the question of whether every potential 2 interaction will be properly identified or not is a question 3 of completeness, which to the best of capability of the 4 analyst performing the analysis should be, all be covered.
5 MR. MICHELSON: That wasn't specifically the 6 question under discussion at the moment. However, that 7 did--the question at the moment was if you have a non-seismic 8 piping, irrespective of size, is there an earthquake level 9 above which it fails in PRAs, or does fail at all?
10 What do we look at beyond the design basis? When we 11 look beyond the design basis, do we look to see whether we 12 start to find that non-seismic type of failure, irrespective L '13 of size?
14 MR. JAMALI: In most situations, it of course allows 15 for a screening out of any such failures. For instance, if it-16 does not, does not make any difference whether all the balance 17 of plant systems have failed because your ceramic insulators 18 are very high powered--
19 MR. MICHELSON: You are missing the point. The 1
20 point is that--
21 CFAIRMAN SIESS: Carl, I think we are going to have 1 i l 22 another tutorial on PRAs, and I hope we will get some people 23 in there for seismic, and that's the place I think it is more j
() 24 generic than this.
i 25 MR. JAMALI: I have one question, Mr. Chairman, '
l HERITAGE REPORTING CORPORATION -- (202)628-4888
____ _ _ _ _ _ _ _ _ _ _ _ _ . - - - - - - _ _ _ _ _ _ _ _ - - - - - _ . -- .-- J
1 IL I L
140 1' also. I have one question with regard to point estimates.
2 I would like to know what you would choose as a 3 point estimate because the problem in performing the 4 quantification of seismic is the fact that you have got sets 5 of hazard curves that might be considered, and then you have 6- correlations. Let's say you have several relay chatters that 7 will come into your core melt equation. It will be 8 correlation of one of those failures, so uncertainties play a 9 major role in the point estimates that you would come up with 10 and the correlated random as a result of it. I'm not quite 11 sure what your point estimates mean in that regard.
12 MR. CHOKSHI: I think you will hear about the hazard 13 core which one might, would be using for point estimation, and i
14 the correlation and the hazard I'm quite clear what is 1 i
15 the--you can't get correlation irrespective of hazard. ;
16' MR. JAMALI: Well, if you determine that you have i 17 let's say five relays that will be significant contributors to 18 a seismic core damage sev,ence, and the information that you 19 have on the failure of the relays given the seismic event 20 hazard curve have all come from the same data, they basically 21 are the same in value so to speak. There is correlation among 22 the probabilities.
23 MR. CHOKSHI: But that has to do with the hazard?
() 24 MR. JAMALI: What will happen is that when you 25 quantify the frequency of the--
i i _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _
141 f'.
CHAIRMAN SIESS:
- 11 I think we are getting in detail 2 that we are not interested in now. Thank you.
3 There are five relays that are important. I don't 4 know why we shouldn't replace them.
5 (Slide) 6 CHAIRMAN SIESS: I just can't understand anybody 7 even starting to make an analysis with seismically essentially
! 8 relays in the plant. There are 10,000 of them. I think any 9 responsible plant owner would replace them before he made the 10 analysis and save money on his PRA.
11 MR. MICHELSON: It would be easier.
12 DR. CATTON: Fewer meetings with NRC.
13 CHAIRMAN SIESS: Okay.
14 MR, CHOKSHI: This is the, on the NUREG 1130, and 15 here is the comparison for the Peach Bottom using the both 16 EPRI hazard codes, and I think the point is there is a big 17 difference in the core melt probability, order nf about f3ctor 18 of 25.
19 CHAIRMAN SIESS: Now these were, these are PRA 20 results?
21 MR. CHOKSHI: Yes.
22 CHAIRMAN SIESS: Made by the same people but using 23 two different hazard curves?
I) 24 MR. CHOKSHI: Correct.
25 CHAIRMAN SISSS: Who made them?
HERITAGE REPORTING CORPORATION -- (202)628-4888
L .
142
.m
- 1 MR. CHOKSHI: Sandia, Mike Bohn, but I think the 2 point is that--
3 CHAIRMAN SIESS: What you are trying to show is the 4 relative contribution, right?
5 MR. CHOKSHI: Right. And the order that--it clearly 6 is, the bottom line, effect on the bottom line numbers is 7 'quite large, but as far as getting insight as to the dominant 8 components or sequences--
9 CHAIRMAN SIESS: What is the first sequence? Is 10 that station Blackout?
11 MR. CHOKSHI: That's the transient, yes, the 12 seismically-induced loss of off-site power.
13 CHAIRMAN SIESS: Okay. What is the next one?
14 MR. CHOKSHI: Next one is the large LOCA, excessive 15 LOCA, and then this one is medium and small LOCA.
16 I think even with the two hazard curves which are 17 different at least in terms of magnitude, there is a 1
18 robustness with respect to the sequences and the dominant i
19 components, and that's what I think we are using, one of the ;
20 bases for saying that the point estimate type of calculation 21 can still provide us with the valuable insights. l 22 (Slide) l 23 CHAIRMAN SIESS: The hazard curves differ by about a
() 24 factor of 10 in probability?
25 MR. REITER: Depending where; depends on what G l
HERITAGE REPORTING CORPORATION -- (202)628-4888 l
L 143 L 1 value.
2 CHAIRMAN SIESS: Okay. Sorne day we will see them?
3 MR. CHOKSHI:- I have them.
4 CHAIRMAN SIESS: I get your point.
5 MR. CHOKSHI: T1.e next one is the use of existing l
l 6 PRA, because that option is also included with the plant to 7 each have a PRA, and we think that, that PRAs which exist that l 8 can be used might leave few more things, and this is I think 9 the general IPE requirement they should reflect the current 10 status.
11 CHAIRMAN SIESS: Let me ask you a couple of things.
12 I think you answered one of them. It says even if old.
13 Didn't quite know what you meant by old.
14 MR. CHOKSHI: Ten years back; that was--
15 CHAIRMAN SIESS: If it was ten years back, it 1 16 wouldn't reflect the current plant status?
17 hR. CHOKSHI: Right.
18 CHAIRMAN SIESS: So old isn't really--
19 MR. CHOKSHI: I think that's the one which is 20 included.
l 21 CHAIRMA'4 SIESS: I wouldn't accept a PRA that was 2a done ten years ago, because I don't think there is a plant out 23 there that hasn't changed in ten years. They would have to
() 24 justify the PRA was still valid.
25 MR. CHOKSHI: That is correct.
HERITAGE REPORTING CORPORATION -- (202)628-4888
i 144 l je
\ ~- - 1 CHAIRMAN SIESS: How well performed,1[ hate to ask, l 2 but have we got a definition of that?
3 MR. CHOKSHI: I think some of the things, that later 4 part of the bullet-- l 5 CHAIRMAN SIESS: This is well performed?
6 MR. CHOKSHI: This is sort of deficiency in the i
7 PRAs, that they are not-- j 8 CHAIRMAN SIESS: Now how many plants have got PRAs, 9 seismic PRA?
10 MR.'CHOKSHI: I would say about 20 to 25.
11 CHAIRMAN SIESS: Twenty to 25 plants that you know 12 of?
13 MR. CHOKSHI: Yes, but not submitted to the NRC--
14 few.
15 MR. BAGCHI: Eight or nine submitted to the NRC.
16 CHAIRMAN SIESS: Counting all the ones that have 17 been done but haven't been submitted--I remember Bob Kennedy 18 gave a number once. Now those 20, 25 plants probably are, 19 pretty good proportion of those are in a high seismic area?
20 MR. BAGCHI: No, not necessarily. Many of those--
21 CHAIRMAN SIESS: Don't have Turkey Point, South 22 Texas, in that list, do you?
23 MR. CHOKSHI: South Texas cleaned it out, but I
() 24' think the point here is that the PRAs can be used to do some 25 additional work, primarily in the walk-down, the relay HERITAGE REPORTING CORPORATION -- (202)628-4888 L________'_____.____
w f ,
145 1 .- m.
1 chatter, and we will also require that from the old PRA we 2 also get a' margin type of information, plant HCLPF, and--
3 CHAIRMAN SIESS: It says estimation of the HCLPFs.
4 You mean use of the PRA to estimate them?
l 5 MR. BAGCHI: Using the fragility curves 95, go to 6 the curve.
7 CHAIRMAN SIESS: You already have it for the PRA?
8 MR. BAGCHI: That is correct.
9 MR. CHOKSHI: Generally you have the component 10 level. You do not have the plant level. It is not reported.
11 The PRA, they can go into the PRA and get that out.
12 CHAIRMAN SIESS: For the PRA, they have something 13 very similar to the HCLPF.
14 MR. SHAO: You can get it.
15 MR. BAGCHI: You can calculate it.
16 MR. SHAO: You can get it.
17 CHAIRMAN SIESS: This isn't something that will 18 discourage anybody from using a PRA?
19 MR. BAGCHI: It shouldn't.
20- (Slide) j 21 MR. MICHELSON: Before you leave that one, don't i
22 these supplementary plant walk-downs--what kind of assumption j 1
i 23 do you make concerning the potential failure of piping during j
() 24 seismic events? Without physically going back, free standing 25 piping, physical impact interaction, do you just assume none HERITAGE REPORTING CORPORATION -- (202)628-4888
146 1 of that falls during seismic event in terms of walk-down?
2 MR. BAGCHI: They developed the fragility curve for 3 piping, for passive components. That's the piping, and they 4 screened out the piping. Fragility comes out, it will be 5 pretty high. It depends on where it comes out. It depends on 6 supports that particular plant has.
7 MR. MICHELSON: So you do assume that there is a 8 potential high fragility of the pipe, is that right?
9 MR. GUZY: I think in the walk-downs, like I think-10 it was pointed out before, you look for any pipes that do fail i
11 to have an interaction like fire protection pipe. If you look i 12 at that piping, their piping, you consider it as failing.
13 Threaded piping has a history of failing. If it was welded I
14 piping, you would assume it wouldn't fail. l 15 MR. MICHELSON: Even though you think this piping ;
i 16 doesn't fail, you do assume it fails?
17 MR. GUZY: No. If you think it, if--you look at 18 piping by physical use characteristics. If it is welded l l
)
19 piping, and it is well anchored at the ends, and if it fits 20 within our experience data base, say no, that it is highly L 21 unlikely it is going to fail. We have high confidence it l 22 won't fail.
I 23 If there is something unusual about it like it is
() 24 threaded piping or some other kind of mechanical connections 25 that could fail or next to nome unanchored tank or something 1
o_ _ _ _ -- - - "" ^^ " " -- <202>628-4888
F.
147 m
u 1 that could pull it out of anchor points, then you do assume it 2 is going to fail.
3 CHAIRMAN SIESS: What if it is attached to a piece 4 of equipment that is poorly anchored?
5 MR. BAGCHI: That would be failure initiated by 6 anchor movement, and there is plenty of example of that in the 7' seismic data base.
.8 CHAIRMAN SIESS: There are examples of failure in 9 the anchor movement?
10 MR. BAGCHI: That's right.
11 CHAIRMAN SIESS: Could I look at the stuff? That 12 took an awful lots of cycles, awful lot of movement. They 13 were moving the pipe, not the equipment. That would be the 14 same thing, and they never did make them fail.
15 MR. BAGCHI: By land or motion, if you consider 16 unanchored equipment, that is an unusual amount of motion. I 17 don't think any kind of piping would withstand that.
18 MR. CHOKSHI: I think the last vugraph on the PRA is 19 the type of reporting, the requirements in parallel to what 20 IPE generic letter requiren and the type of information one 21 would, the PRA puts out.
22 CHAIRMAN SIESS: These are possible reporting--
23 MR. CHOKSHI: Yes. They aren't final list.
() 24 CHAIRMAN SIESS: If somebody does a PRA as a part of 25 IPEEE, these are possible reporting requirements?
HERITAGE REPORTING CORPORATION -- (202)628-4888
148
,- m .
t
)
1 MR. CHOKSHI: No. Some of them will say in our 2 final guidance you should report, and these are the type of 3 items we are looking at, you know, including in those 4 reporting requirements. These are the possible items which 5 would be in the final letter or final guidance.
6 CHAIRMAN SIESS: This would apply whether they are 7 using an old PLA or new one?
8 MR, CFOKSHI: It would apply to both, but we held--I 9 think one of the, we haven't clearly decided on specific 10 criteria. We may not even--that you have to report sequences 11 at 10 to the minus 5, minus 6.
12 CHAIRMAN SIESS: The fourth bullet says dominant 13 components, human actions, non-seismic random failure.
14 MR. CHOKSHI: Right.
15 CHAIRMAN SIESS: Now I am trying to fit that into l
16 the IPE, IPEEE or whatever.
17 MR. CHOKSHI: Coupled with other seismic, seismic, 18 but the non-seismic failures, seismic initiating event. For l
l 19 example, if you have a seismic-induced loss of off-site 20 power-- 1 21 CHAIRMAN SIESS: Is the seismic PRA something 22 different than a PRA including seismic?
I
! 23 MR. CHOKSHI: Seismic PRA typically also includes l
() 24 non-seismic failures.
25 CHAIRMAN SIESS: There are two types, those with t
HERITAGE REPORTING CORPORATION -- (202)628-4888 L_________
149 p 'k- 1 seismic and those without?
l L '2 MR. CHOKSHI: Yes.
l l 3 CHAIRMAN SIESS: There is no PRAs that includes 4 seismic and nothing else?
5 MR. CHOKSHI: We are talking about the seismic 6 initiator accidents.
7 CHAIRMAN SIESS: Why are you interested in the 8 fourth bullet?
9 MR. CHOKSHI: The reason for that, for example, if 10 you have a seismic-induced loss of off-site power, then you 11 demand it to start, but the diesel fails not from the seismic, 12 but just because from failure to start, and that becomes 13 important if that failure rate is very high.
14 CHAIRMAN SIESS: I have heard that is the dominant 15 contribution from seismic, seismically-induced loss of 16 off-site power and non-seismic random failures of components.
17 MR. CHOKSHI: The fix is not a seismic..
18 CHAIRMAN SIESS: What has that got to do with the
- 19. IPE?
20 MR. CHOKSHI: That also from the seismic-induced 21 accident, it is not a seismic vulnerability. It is something 22 else, and that you are not fixing the wrong thing. I think 23 that's important, that fixing something else, seismic supports
() 24 or something, will not result in that problem.
25 I think if there is no more questions, the HERITAGE REPORTING CORPORATION -- (202)628-4888
7-,_--____-_________
150 l
- ["N ,
i ' 1 next--that's all I would.say about the PRA approach we are 2 thinking about.
i 3 CHAIRMAN SIESS: You are thinking then two possible l
'4- approaches. One is'a seismic PRA, old or new, meeting some 5 reasonable criteria--appropriateness, completeness, and the 6 other would be some form of seismic design margin program 7- where you'still got a lot to work out on the--I still can't 'j 8 remember the name of that earthquake.
9 MR. CHOKSHI: Review earthquake.
10 CHAIRMAN SIESS: Of the review earthquake; and you 11 think they will be equally good at showing up the 12 vulnerabilities, even though you don't know what a 13 vulnerability is?
14 MR. SHAO: Right.
15 MR. MURPHY: Acceptable.
16 CHAIRMAN SIESS: Okay. I think that's clear. Let's 17 take time cut for lunch now and be back at 1:20.
18 (Whereunon, at 12:20 p.m., the hearing was recessed, 19 to reconvene at 1:20 p.m. the same day.)
20 21 22 23 1
() 24 25 HERITAGE REPORTING CORPORATION -- (202)628-4888
m.9 <
,r ;
1 i 'y' r
.h jc l:
i ^ +
l.__ t t
(.
i-- i
.hkT ,
..u,!\_l=
l 151 j
l' .
AEb&BE'QQH EEEELQH
- h 2 , '(1
- 50 p.m.)
,3 CRAIRMAN SIESS: We will reconvene.
4 'I: guess Roger Kenneally is on the program.
5 (Slides being shown.)
6 MR. KENNEALLY: One of the methods for satisfying
!7 the' intent of IPEEE has been the seismic PRA. The other m" ~ '
8 method that is being considered by the subcommittee is the 9 seismic design margin methodology.
10 The initiative objective of this.was spurred by 11 ACRS comments back in the 1983 time period of developing a 12 method to assess the inherent capability of a plant to
() '13 withstand earthquakes above the design level. -There was
- 14 quite a bit.of correspondence back and forth between the 15 - ACRS and the Staff.
16 And the concerns at that point was, what happens 17 if there.was slight changes in seismic input, for example?
18 Is there a method that we could come up with that would give 19 us an intent of what is the margins of safety if the CPA 20 were to go from 0.15G to 0.17G. The current critarion would 21 call for a complete new analysis and the ACRS was concerned 22 for these slight differences; could we come up with methods 23 of assessing the margin.
24 There are two methodologies that are currently
- 25 available to do this. One of them having been developed by O 3eritage Reporting Corporation (202) 628-4888
L.
i <
m;
(
p A s,h ^
152 1 the NRC. It's the fault tree / event tree type approach.
2L .It's documented in NUREG/CR-4334. That NUREG is dated in 3 the '84 '85 time period.
4 The principal features'of the NRC methodology was 5 to guarantee a shutdown of the reactor and provide cooling.
6 The Electric Power Research Institute, EPRI, also developed 7 a methodology subsequent to the NRC's and it uses a success 8 path or paths, plural. It is documented in a report.EPRI 9 NP-6041.
F 10 They've got a little different motivation. They 11 want to provide either a hot or cold shutdown provision and 12 maintain it there for 72-hours.
'13 CHAIRMAN SIESS': Roger, what's the basis -- after 14 72-hours they assume that they could have restored whatever 15 was lost.
16 MR. KENNEALLY: Yes.
17 CHAIRMAN SIESS: Especially offsite power.
18 MR. KENNEALLY: Right.
19 CHAIRMAN SIESS: Okay.
20 MR. KENNEALLY: Both the methodologies derive from 21 the early seismic PRA insights. -These were some that were 22 already published. Some that the consultants that worked on 23 then had knowledge of or whatever. And these were the ones 24 that were available in the ' 83 ' 84 time period.
25 The attractiveness of the seismic design Heritage Reporting Corporation (202) 628-4888
( )
v 153 1 methodology is that it does reduce the scope of the systems 2 and components to be examined. We're not going at with the 3 robustness of the full seismic PRA.
4 However, this reduction has not sacrificed two of 5 the most important features that we have learned from the 6 seismic's PRA, and that is the integrated plant response, 7 spacial interactions, the impacting of buildings, the 8 concern of walls and piping and the like.
9 Also, the plant walkdown is maintained in the 10 seismic margins method.
11 The strong feature of it at the time of its 12 inception was the fact that the seismic hazard curves were
,a
() _
13 not utilized and all the uncertainties associated with them 14 were thereby eliminated.
15 The method is based on a screening criteria where 16 one would choose a review level earthquake or have one 17 assigned to them. And the utility would screen various 18 components based on that one. There are screening tables 19 that are provided in those documents referenced earlier.
20 And if the component is shown to have sufficient capacity at 21 the review level earthquake it is eliminated from further 22 consideration.
23 The screening tables are set up for basically a 24 0.3G PGA; 0.5G PGA. If you get into the area where you were 25 assigned the screening of higher than 0.5G on the PGA it's V
Heritage Reporting Corporation (202) 628-4888
l;
(_)
154
'l really recommended by all concern that you would use the 2 seismic PRA. -You're just not going to be able to eliminate 3 enough'to make it worthwhile.
4 The two different methods, like I said earlier, 5 are a little bit different in the way they go about 6 achieving the high confidence, low probability of failure.
7 The EPRI one uses a success path. And as I said earlier, 8 the NRC one is the event tree / fault tree one.
9 In this example it would show what the success 10 path logic diagram would be. The operability of components 11 of systems A and B in this case are necessary. And then 12 after one would get by there you would have a couple of
'() 13 alternatives to select on being able to attain a long-term 14 safe shutdown.
15 One option that could be chosen as a success path 16 would be to go from A, B, through F, and you could obtain 17 your shutdown. Other options would be A, B, up into C; and 18 then you would have a D or an E, and one of those succeeding 19 would get you to your long-term shutdown.
20 And so it is possible on the EPRI method to go 21 through one path, A, B, and F, examine it in great detail, 22 come up with a high confidence, low probability of failure 23 value, either the screening value or if it were lower than 24 the screening determine what that value would be and still 25 conceivably maybe have something go wrong at the plant in an Heritage Reporting Corporation (202) 628-4888 ,
1 I
)
_ _ _ _ _ _ _ _ - - . - - - - - - )
t r p
[ '-
n
'O 155 l' earthquake level less than what you would screen it on.
l Mainly, if you were in this chain that involved C, D, and E.
2.
3' So it isn't all capturing; you will guarantee a 4 success and you will be able to meet the intended functions 1 5- you had. And that's one:of the things that the subcommittee
~
6 is evaluating in the context of IPEEE is, how many success 7 paths would one want? Obviously, one wouldn't be 8- acceptable. In the work that the EPRI has done they have 9 chosen at least two and sometimes more of them. But that is 10 something that we are keeping in the back of our mind is to 11 make sure that the generic letter or anything would 12 guarantee that there would be enough success paths to meet
?~)
\_s 13 the intent of what we are trying to accomplish.
14 CHAIRMAN SIESS: Excuse me. The criteria for 15 success paths for a straight seismic margin review and the 16 criteria for a search for vulnerabilities; do they have to 17 be the same?
18 MR. KENNEALLY: That's what we're fighting with 19 right now. The margins approach, obviously, one could 20 obtain it if you vent through that initial selection, either 21 one of them. For the margins, we've got to make sure we 22 understand that and recommend that to management properly. j 23 Would we want a more inclusive criteria?
24 I think what I"m reading in your question is --
25 CHAIRMAN SIESS: Let me -- I didn't phrase it very Heritage Reporting Corporation (202) 628-4888 4
>f i i
~' '
O,
.y
H v l 156 H
l 1 clearly, I'll admit.
l 2 Suppose C had a very low probability of 3 succeeding. In other words, it's an outlier; it's a l 4 vulnerability. But F is rnal good.
5 Now, this would not necessarily find that 6 vulnerability in C.
7 MR. KENNEALLY: If your success path you would 8 pick was A, B, F, you would not have identified C.
9 CHAIRMAN SIESS: Now, would you consider that a 10 real vulnerability if there is an alternate path?
11 In otner words, if I were doing a PRA how much 12 would that weakness in C increase my core melt probability r'
(,s)'
a 13 if F waJ highly reliable?
14 MR. CHOKSHI: Let me answer little bit differently 15 that question. If I were to compute plant level HCLPF, high 16 confidence low probability failure, that will be gone by F 17 rather than C. The stronger component will go on the plant 18 capacity.
19 CHAIRMAN SIESS: Yes.
20 MR. CHOKSHI: So in terms of the core damage it 21 depends on the -- the contribution from C should not be as 22 important because F is available.
23 CHAIRMAN SIESS: So would we miss a vulnerability 24 if we use this?
. 25 MR. CHOKSHI: No, I think --
Heritage Reporting Corporation (202) 628-4888
e
- (~
T_)X r T 157-1 CHAIRMAN SIESS: Because it wouldn't really be a.
2 vulnerability of-the core melt; it would just be a 3 vulnerability of the failure of that --
4 MR. CHOKSHI: Component, right.
5 CHAIRMAN SIESS: -- component of the system.
6 'MR. CHOKSHI: The only thing you are relying on 7 one path. That's the reason for looking at another path.
8 The things you may not be looking at other kind of failures; 9 them failures are human actions and other things. That's 10- why to select an alternate path.
11 CHAIRMAN SIESS: But now if I did a PRA and the 12 weakness of C didn't have much effect, I wouldn't discover 13 it either. In other words, I wouldn't call it a 14 vulnerability because it wouldn't have much effect on core 15 melt.
16' MR. CHOKSHI: It may be dominant component, you 17 are correct.
18 CHAIRMAN SIESS: So you would know about it in 19 this case, but it's not dominant. In the other case it 20 wouldn't be dominant but you wouldn't know about it.
21 MR. KENNEALLY: That's right.
22 CHAIRMAN SIESS: And in terms of the bottom line 23 they're the same.
24 MR. SHAO: They're the same. The seismic 25 methodology essentially is derived from seismic PRA.
Heritage Reporting Corporation (202) 628-4888
j'^ ;-
L}
158 1, CHAIRMAN SIEGS: Yes.
2 MR. CHOKSHI: I think what you mean, Dr. Siess, 3 is, for example, if C is a simple fix which could 4 tremendously improve, okay, those kind of things, you know, 5 which is --
6 CHAIRMAN SIESS: That's right. If C was easy to 7 fix then we would have two good paths.
8 MR. SHAO: Good paths.
9 CHAIRMAN SIESS: You would probably do it. But if 10 you didn't do it, it wouldn't have much affect on your core 11 melt probability.
12 Okay.
r^S U 13 MR. KENNEALLY: The remaining part of the figure 14 here is showing the approach that would be suggested by the 15 NRC method which is the spin-off directly from the PRAs, the 16 event trees and the fault trees. Unless the subcommittee 17 would want more of the detailed presentation on that, I 18 don't intend to go through that right now. The main point I 19 was trying to make with this particular figure was that 20 there are some inherent differences between the two 21 approaches.
22 CHAIRMAN SIESS: I can see differences. I guess 23 I'm not sure that they' re inherent.
24 MR. KENNEALLY: Well, the success path may be only 25 not catching everything.
/,_'T
(_/
Heritage Reporting Corporation (202) 628-4888
(-
%)
159 1 CHAIRMAN SIESS: Yes. That's what we were just 2 talking about.
3 MR. KENNEALLY: Right.
4 MR. CHOKSHI: I think one important point -- if 5 you can put the slide down. The logic -- success path logic 6 is just operational sequences. And the reentry you would 7 assign probabilities to each branch. So in the NRC method 8 when you go through the reentry you quantify the HCLPF from 9 the ecquence.
10 CHAIRMAN SIESS: Yes. One place you're working 11 with HCLPFs and the other with probabilities.
12 MR. CHOKSHI: And the other one is looking at the
,A
(_f 13 component within the block.
14 CHAIRMAN SIESS: Okay, they are different. But 15 which is the easiest?
16 MR. KENNEALLY Some of the plant operators think 17 that the EPRI one is because it's working more on the space 18 that they're familiar with.
19 CHAIRMEN SIESS: If we look at step ahead of the 20 IPE to the accident management phase, which of these 21 approaches would be most useful in accident management 22 planning?
23 MR. CHOKSHI: My opinion is NRC type approach, l
24 And the reason is that you are looking in the space, so you jm 25 can deal with respond. Success path, I think is an ultimate
(_
Heritage Reporting Corporation (202) 628-4888
160 1- accident management, there is no failure in one sense, you 2 know, it does not look at the failure.
o 3 HR. KENNEALLY: We've discussed earlier that the 4 two methodologies have been applied to three plants: the 5 Maine Yankee'being the candidate for the NRC program. That 6 was a plant designed with an SSE of .lg. The initial high 7 confidence low probability of failure, HCLPF, plant capacity 8 was .219 And then with some more modifications was at the 9 .279 10 EPRI's initial benchmarking of the methodology was 11 against Catawba._ The SSE of .15g and the HCLPF was .3g.
12 CHAIRMAN SIESS: What was the review earthquake on 13 each of these, could you tell me?
14 MR. KENNEALLY: .3g.
15 CHAIRMAN SIESS: .3 on all of them?
' 16 MR. KENNEALLY: Yes.
17 And the Hatch approach is a combined NRC, EPRI 18 evaluation of a plant. And the initial estimate has been 19 .289 20 CHAIRMAN SIESS: Now, Maine Yankee was NRC and 21 Catawba was EPRI.
22 MR. KENNEALLY: Maine Yankee was NRC; Catawba was 23 EPRI.
24 CHAIRMAN SIESS: How do you combine them?
f_ , 25 MR. KENNEALLY: The Hatch is being done both ways.
Heritage Reporting Corporation (202) 628-4888 1
L
. e.
161
- 1. CHAIRMAN SIESS: Oh, both, p 2 MR. KENNEALLY: Yes.
t' 3 CHAIRMAN'SIESS: Okay.
4 MR. KENNEALLY: And the hard thing of determining 5 the HCLPF is probably none of the plants were actually 6- evaluated in their as found state. As they went through 7 some of their initial walkdowns there was such obvious 8 outlier, vulnerabilities, that they were corrected.
9 So even the .21, the .28 or'whatever reflects some 10 modifications; and that's why the Staff feels so strongly 11 that the walkdowns are so critical. And we've used the term 12 -. this morning of a seismic design margins walkdown as a 13 classification, shall we say. Put it in that context as 14 opposed to one of an A-46 which might have different intent 15 of inside the design basis or whatever.
16 So that's why the term seismic design margins 17 walkdown was used.
18 CHAIRMAN SIESS: So on Maine Yankee, for example, 19 if you had -luut done the walkdown and didn't -- well, the 20 walkdown might have showed you the tank was a problem.
21 MR. KENNEALLY: It did.
22 But there might have been things like an anchorage 23 on a cabinet that a bolt was missing or something. The 24 holes were there. Some of them were in, not all of them.
25 And so by putting in the remaining bolts or anything you I Heritage Reporting Corporation (202) 628-4888
l^
1 162 1 certainly increased the seismic capacity of that component..
'2 It's not to reflect the walkdown determined at'.21g 3 initially. There were some modifications to get to that 4 .21.
.5 Basically, this has just been a quick introduction
-6 as to where we ara with the current seismic design margins 7 method. It's not directly applicable to the IPEEE for a 8 couple of reasona. You know, there's some things like relay 9 chatter, containment performance, and the like that still 10 have to be folded into it; and that's what we're being 11 evaluated on. And also, the selection of the review level
- 12. earthquake, which is another critical feature is, what do 13 you start this ball rolling at for different places? And 14 those discussions will be made following mine.
15 CHAIRMAN SIESS: Yes.
16 But I guess I have a problem nowadays thinking 17 about relay chatter as a vulnerability that has to be dug 18 out by some sort of a procedure. I mean, I keep hearing 19 that there's only a dozen relays that could chatter or seven 20 or 20. And isn't it almost obvious now that those are 21 vulnerabilities without going in and doing anything else?
22 MR. KENNEALLY: Yes.
23 CHAIRMAN SIESS: I mean, are we beating a dead 24 horse.
25 MR. SHAO: We do find that the relay chatter can Heritage Reporting Corporation (202) 628-4883
l' Li- 1 f gr%
im>) ,
61 163 1
i i
L 1 be important, even in'the PRA, too, the major risk. J i
2- CHAIRMAN SIESS: But I mean, why don't they just l l
3 fix them and get along with it.
1 4 DR. SIIEWMON: They probably'use the wrong ;]
5 statistics when they find that's true. ]
6 CHAIRMAN SIESS: Are there any plants were relay 7 chatter is not going to present some kind of a problera when i I
8 you do a seismic analysis, evsn if it's just an uncertainty 9 problem?
10 MR. CHOKSHI: I think a lot of effort goes into' 11 finding those relays, which could cause problems in the 12 chatter; and then, you know, you can decide whether you want
- ,3 kJ 13 to fix them or look at the fragility from the chatter. But 14 a lot of effort is beina spent in identifying those.
, 15 CHAIRMAN SIESE: You have to identify the ones li that are isaportant.
l 17 MR. CHOKSXI: The ones that are important.
18 CHAIRMAN SIESS: Eeoause it*s assumed that any of 19 them are going to chatter.
20 MR. CHOKSHI: Yes.
21 CHAIRMAN SIESS: Okay. And that can be done 22 either with the PRA or one of these margins.
23 MR. CHOKSHI: Yes, it's a lot of trystems r6 view, 24 but you have to do that.
,, 25 CHAIRMAN SIESS: Okay.
L) i Heritage Reporting Corporation ,
(202) 628-4888
<m i
.] i 164 l 1
1 1 Now, Mr. Chokshi will get a chance to talk.
2 (Slides being shown.) 1 l
3 MR. CHOKSHI: What I'm going to talk about is some l 4 ways of handling containment performance issues in seismic 5 margin methodology context.
6 I think as Roger mentioned that the whole EPRI was 7 developed with respect to core danage. Whether the focus B was looking at a systems and components to prevent core 9 damage or as to the way you can get to the core damage. l 10 The severe accident policy talks about also the 11 features of containment, which could affect the performance 12 in both mitigation sense and in seismic is containment --
O-
\_j 13 could also initial accident sequence.
14 The first question is: to what extent containment 15 performance evaluation is necessary for seismic IPE7 16 and then the second -- tnother issue to answer is:
17 how to extend margin approach to improve containment 18 performance and what are the pe::formance aspects from the 19 seismic point of view. You ceuld have a containment bypass; 20 there could be a structural problem, which include 21 penetration as a part of that; heat removal functions of 22 containment; and then you have containment isolation 23 failures.
24 CHAIRMAN SIESS: What makes bypass different from 25 isolation failure?
~N f (V
Heritage Reporting Corporation (202) 628-4888
- S 165 1 MR. CHOKSHI
- Traditionally the bypass term is 2 used where you have a high pressure / low pressure system. So 3 in that sense I kept ueparate. Well, the system was it's a 4 little different.
5 CHAIRMAN SIESS: A bypass is an isolation failure.
6 MR. CHOKSHI: Yes, isolation will affect the 7 bypass containment, also.
8 We have a technical assistance contract with 9 Lawrence Livermore and Paul Amico is the principal 10 investigator. A1.d what he is doing is going back to the 11 PRAs which we looked at in our original development of 12 margin methodology and to get some insights for looking at k' 13 the high consequence sequences.
14 When we did the original margin method the core 15 damage -- most of tne sequences contributing to the core 16 damage were basically failure of containment, and basically 17 a fatalities type of risk.
18 The high segaance only risk -- high sequences do 19 not contribute much to the core damage frequencies, and so 20 we went back and looked at that.
21 And I want to just point out a couple of 22 preliminary conclusions from his examination -- and I'll 23 show some of the data in later slides.
24 For PWRs by looking at about four, five plants,
,s 25 all the consequence sequences involve gross structural ,
(_) I Heritage Reporting Corporation (202) 628-4888 l
?
L L 166 1 failures.
I 2 CHAIRMAN SIESS: Gross structural failure due to 3 overpressure or due to earthquake?
4 MR. CHOKSHI: Due to earthquake, at very high G.
.5 And a lot of that might be because it was not important in 6 the PRA context and all kind of things. In one case -- an'd 7- I think that's also -- where there was a case of isolation 8 failure which also related to an early fatality type of 9 event.
10 For BWRs --
11 CHAIRMAN SIESS: Now, here he was icoking at 7(J 13 MR. CHOKSHI: Seismic PRAs, right. For FARs where 14 we, I think -- he's still looking at some dr.ta base and 1 15 will cautivn that this is preliminary, particularly for 16 boilers, va don't hatre much plans.
17 He found sin.ilar things that either veu have a 18 gross internal failure, and in this case it was a failure of 19 reactor pressurized supports and that sort of thing. And
. 20 also, it was sequences, which also result in a high 21 consequences sequences. And I'll show the data from some of 22 the plants.
- 23 CHAIRMAN SIESS: The ATWS sequence is now 24 seismically induced?
25 MR. CHOKSHI: Yes, that's all seismically induced.
Heritage Reporting Corporation (202) 628-4888
i, 167 l' And as pointed out here it was the figure of 2 internals and then the random failure of the SLCs, standby 3' liquid control system. And the combination of that in a 4 couple of plants has shown up.
5 Again, the MARK I and ice condenser, we still.have 6 not really much data on it and we haven't come_up with how 7 to -- what type of guidance we can view -- different 8 guidance for the utilities.
l 9 CRAIRMAN SIESS: You mean nobody has done a PWR on 10 a MARK I.
11 MR. CHOKSHI: No, Peach Bottom is just finished 12 for 1150.
w i
13 CHAIRMAN SIESS: No seismic PRA on a MARE I?
1 34 MR. CHOKSHI: No, there are, recently.
15 MR. SHAO: Peach Bottom is the MARK I.
l 76 CHAIRMAN SIESS: Well,_then why do we know less 17 about the MARK I?
18 MR. CHOKSHI: Now we are looking at those things.
19 CHAIRMAN SIESS: Oh, they just haven't been looked 20 at.
21 MR. CHOKSHI: Just haven't been looked at.
22 CHAIRMAN SIESS: Oh, okay.
23 As I recall, Peach Bottom had a' core melt 24 frequency of two or three times 10 to the minus 67 25 MR. CHOKSHI: Peach Bottom is 7.5 and 10 to the Heritage Reporting Corporation (202) 628-4888
7 h*
E Q 168 1- minus 5, if you use Lawrence Livermore.
2- CHAIRMAN SIESS: That's in 1150?
3 MR. CHOKSHI: 1150.
4 CHAIRMAN SIESS: Okay.
l 5 MR. CHOKSHI: And if you use:EPRI it's 25 less.
6 Those two show -- these last two columns are 7 .important. This is the type of failure they found for which 8 you should take to the high consequence sequences. And I'll-9 get into the. structural figure of containment. In Indian 10 Point 2 it wp3 a shear failure because of a backfill.
11 At Millstone it was a crane wall collapse. And at
.,,, 12 Seabrook there is, because of a chatter, there is a
% } 13 possibility that a containment purge valve will be left 14 open.
15 CHAIRMAN SIESS: l'm not sure I understand the two 16 sets of columns there.
17 MR. CHOKSHI: Which one was that?
18 CHAIRMAN SIESS: The no containment bypass and 19 containment bypass; whr.c are we --
20 MR. CHOKSHI: In this table containment bypass is 21 the same as a gross containment failure.
22 CHAIRMAN SIESS: But if this is a PRA no 23 containment bypass is the core melt frequency assuming the 24 containment is infinitely strong?
- 25 MR. CHOKSHI: No. That just -- the terminology
. O-Heritage Reporting Corporation (202) 628-4888 1
169-1 used where the contain- -- the technology used for defining 2 all low consequence sequences. Reaches very low sequences
'3 because of accident. It includes late containment failure; 4 no containment failure; all cases. And this containment 5 bypass includes a containment failure.
6 CHAIRMAN SIESS: The left column then is the j 7 bottom line for the PRA including everything.
8 MR. CHOKSHI: Right.
9 CHAIRMAN SIESS: Including seismic. i 10 MR. CHOKSHI: This is just the core damage
-11 frequencies. Seismic core damage frequency.
12 CHAIRMAN SIESS: This is a seismic core damage
) 13 frequency, and the right hand column is the proportion of 14 that that's contributed by containment bypass or is not 15 proportion --
1 16 MR. CHOKSHI: This one is the seismic induced high l
-17 sequence, only fatalities. One is a core damage, other one 18 is a risk number.
19 CHAIRMAN SIESS: I'm sorry, I still don't 1
20 understand.
I 21 MR. CHOKSHI: This number represents --
22 CHAIRMAN SIESS: These aren't -- the frequency -- ;
1 l 23 MR. CHOKSHI: Right.
24 CHAIRMAN SIESS: That's not probability -- that's
-25 not core melt probability.
O.
Heritage Reporting Corporation (202) 628-4888 1
L.)
170 1 MR. CHOKSHI: This frequency in this column is a 2 core damage from seismic.
3 CHAIRMAN SIESS: It's a core damage frequency.
4 MR. CHOKSHI: Right.
5 CHAIRMAN SIESS: From a seismic PRA.
6 MR. CHOKSHI: From seismic PRA.
7 CHAIRMAN SIESS: Now, the second column is what?
8 MR. CHOKSHI: This is the HCLPF --
9 CHAIRMAN SIESS: No.
10 MR. CHOKSHI: This one?
11 CHAIRMAN SIESS: Yes.
12 MR. CHOKSHI: This one is the probability of early
,r's V 13 fatalities from seismic PRA.
14 CHAIRMAN SIES?: Oh, this is early fatalities.
15 MR. CHOKS3I: Right. And that'e how the 10 containment performance is related.
17 CHAIRMAN SIESS: Ch, boy.
1B Now I understand. But those two headings you 19 got --
20 MR. CHOKSHI: I know.
21 CHAIRMAN SIESS: -- if I take out no containment 22 bypass and put core damage frequency there, and on the other 23 side I put down early fatalities.
24 MR. CHOKSHI: I should have changed them. I just f-w 25 took them from the report.
Heritage Reporting Corporation (202) 628-4888
171 1 But I think the point was that all of this numbers 2 all are low. They are going by the type of failures which 3 are containment structural failures at very high -- in the 4 case of Indian Point the capacity do point greater than 2g.
5 DR. SHEWMON: HCLPF always has the units of G on 6 it. That's always --
7 CHAIRMAN SIESS: Yes.
8 DR. SHEWMON: Whereas frequency can change all 9 kinds of ways. We just learned HCLPF is always Gs.
10 CHAIRMAN SIESS: Now, when it says containment 11 bypass dominates, do you really mean containment bypass.
12 Now that I have erased it in the two columns --
rms k,) 13 MR. CHOKSHI: Right. What -- I think what he's 14 saying is, the containment bypass -- in this case the bypass 15 is caused by the failure of containment. Containment causes 16 both -- fe.ilure of containment causes accident as well as 17 containment failure.
10 CHAIRMAN SIESS: But now in the last column under 19 comments, containment bypass dominates, that means if 20 containment structural --
21 HR. CHOKSHI: Structural failure, right. The 22 second line --
23 CHAIRMAN SIESS: But before structural failure was 24 one alternative and bypass was another one; and here you're 25 using them as --
Heritage Reporting Corporation (202) 628-4888
l ,-~
-V 172 1 HR. CHOKSHI: That's because I'm using two l
2 different -- I was trying to use traditional terminology 3 before. But the way --
4 CHAIRMAN SIESS: If ycu blow a hole in the 5 containment you're not bypassing it; the thing is failing.
6 MR. CHOKSHI: Yes. But in this table the term was 7 used bypass, but it synonymous with containment failure.
l 8 Yes, I think the last one is more of a case of a bypass 9 where there's isolation failure.
10 CHAIRMAN SIESS: The last one is an isolation 11 failure, it says it.
12 MR. CHOKSHI: The other ones are structural fs 13 failure.
14 CHAIRMAN SIESS: Crane wall collapse. Now, how 15 does crane wall collapse release up to the environment?
16 MR. CHOKSHI: It is simply as unit - you know, 17 it's crane and when it collapse it's going to take the --
18 pull out some of the steam lines and the containment is 19 affected, is breached.
20 CHAIRMAN SIESS: That's an assumption.
21 MR. CHOKSHI: That's an assumption, right.
22 CHAIRMAN SIESS: That's like assuming that when 23 the roof failed at Zion that all seven pumps went out.
24 MR. CHOKSHI: Yes. I think a lot of this is based
,m 25 on assumptions.
\ )
</
Heritage Reporting Corporation (202) 628-4888 i
_ ___ __ _j
'; (~y J,
173
.:;- 1 CHAIRMAN SIESS: And where it says direct 2 ' containment failure.that is due to seismic action.
3 MR. CHOKSHI: Yes. And in the Indian Point case
- 4. it.was the seismic --
1' 5 CHAIRMAN SIESS: At .82G.
6 MR. CHOKSHI:--That's the HCLPF, yes. HCLPF, .82 7 nillion was more than 2.
8 CHAIRMAN SIF.SS: At .82G there's 95 percent
- 9. confidence that.--
10 MR. CHOKSHI: -Only five percent probability.of
-11. failure.
12 CHAIRMAN SIESS: That it won't-fail.
13 MR. CHOKSHI: Right.
- 14. CHAIRMAN SIESS: And that's why it's four E to the 15- minus 8 over there.
16 IE. CHOKSHI: Yes.
17 CHAIRMAN SIESS: I guess that's clear.
18 Go on.
19 /
's 20 21 22 23 24 LO Heritage Reporting Corporation (202) 628-4888 1
l k
.p k
174 l 1 MR. CHOESHI: The similar things for the BTW, but l-
$ 2 here the ATSW sequences are showing up more in the, 1 think l:
3 'in the second and third plant or second'and first plants is 4 the failure of reactor nodes, and also the failure of 5 standby liquid control system. In this case, it's an ATWS 6 sequence, but because of the high pressure generators from 7 ATWS nequence, even early containment failure.
8 Bypass is used as just to classify.high 9 consequence sequences, okay.
10 CHAIRMAN SIESS: Now again, I'm trying to, first, 11 why is it Plan A, Bc C and D? Did EPRI do this?
12 MR. CHOKSHI: Because some of them has not been h'
\ 13 submitted to NRC.
14 CHAIRMAN SIESS: Okay.
15 Now, on Plant A, reactor pressuro vessel support 16 failure. Now what is that? That's supposed to -- how does 17 that lead to a release to the atmosphere?
18 MR. CHOKSHI: Well, once you see the failure of 19 support and the reactor vessel moves, all the things 20 attached to the vessels, all the piping and things --
21 CHAIRMAN SIESS: It's all still inside 22 containment?
23 MR. CHOKSHI: Well, it is breached, because you've 24 got a large movement of the reactor pressure vessel.
25 CHAIRMAN SIESS: Now if this is a PWR, those pipes
.O
Heritage Reporting Corporation (202) 628-4888
\g ij 175 1 lead to steam generators?
2 MR. CHOKSHI: No, this is B. These are boilers.
3 CHAIRMAN SIESS: Oh, I'm sorry. These are all Bs.
4 MR. CHOKSHI: Bs.
5 CHAIRMAN SIESS: I'm sorry, I missed that.
6 So now those pipes lead outside of containment?_
7 MR. CHOKSHI: Outside of containment, right.
8 CHAIRMAN SIESS: I get it.
9 MR. CHOKSHI: And then another type of failure 10 which shows-up --
11 CHAIRMAN SIESS: And the second one is heat 12 removal failure, right?
- 13 MR. CHOKSHI: Second one is? I didn't hear.
14 CHAIRMAN SIESS: Service water system?
15 MR. CHOKSHIt. Yes. That's a service water system.
16 CHAIPMAN SIESS: Or the liquid poiscr:.
17 MR. CHOKSHI: The standby liquid control system.
18 CHAIRMAN SIESS: Okay.
19 MR. CHOKSHI: This is just a summary of the 20 results from the PRA. ,
21 The next slide I just follow more for information. l 22 CHAIRMAN SIESS: No, don't go quite so fast. Go 23 on back to Plant B.
24 MR. CHOKSHI: Okay.
25 CHAIRNAN SIESS: Suppose I did a non-seismic PRA.
I-)l l
Heritage Reporting Corporation !
(202) 628-4888 1
i l
O 176 L
C 1 MR. CHOKSHI: Non-seismic?
2 CHAIRMAN SIESS: Yes.
3 MR. CHOKSHI: That SLC is not likely to fail, or 1
! 4 its probability of failure is much, much lower for non-l; 5 seismic case than it is for a seismic?
6 MR. CHOKSHI: This is the -- this is not a seismic
- 7. failure. It's a random failure.
8 CHAIRMAN SIESS: It's a random failure. It's a 9 random failure.
10 CHAIRMAN SIESS: Well, wouldn't we have the same 11 random failure if it wasn't a seismic PRA?
12 MR. CHOKSHI: Yes, same. Right.
13 So if there was corresponding high, you know, 14 frequency of the reactor not going --
15 CHAIRMAN SIESS: So it has to be -- I'm sorry, 16 there is an "and" in there, right?
17 CHAIRMAN SIESS: It's a seismically-induced 18 internal failure or soil liquefaction and the random 19 failure.
12 0 MR. CHOKSHI: And random failure, right.
21 CHAIRMAN SIESS: I got lost in the words there.
22 MR. MICHELSON: I guess you are saying that the 23 probability of failure of a component during a seismic event 24 is no greater than normal.
25 MR. CHOKSHI: No, that's --
Heritage Reporting Corporation (202) 628-4888
l
,e l (_)
177 1 MR. MICHELSON: In other words, the random -- the 2 probability of a random failure during a seismic event is 3 the same as the probability of a random event any other 4 time.
5 MR. CHOKSHI: Most of the time that's the case.
6 In some companies, it might differ.
7 MR. MICHELSON: I was leading to the question, do 8 we correct at all for the effect of seismic disturbances on 9 components?
10 MR. CHOKSHI: We have a couple of things, but not 11 in a generic sense, no.
12 MR. MICHELSON: Now, do we have real good reason 13 to believe that that's a valid assumption, that seismic 14 events do not increase the probability of failure?
15 MR. CHOKSHI: There may be -- some of the valves 16 and things like that, you know, a question has been raised.
17 MR. MICHELSON: Oh, yes.
18 MR. CHOKSHI: And they are looking at it.
19 MR. MICHELSON: Motor-operated valves, for 20 instance, one ought to ask that question very seriously.
21 MR. CHOKSHI: Right, right.
22 MR. MICHELSON: You don't correct, as far as I can 23 tell. In the PRAs I've looked at, you don't seem to be 24 correcting for --
- 25 MR. CHOKSHI: I don't believe that it's been v
Heritage Reporting Corporation (202) 628-4888
L) 178 1 systematically addressed anywhere.
2 CHAIRMAN SIESS: Well, now, if you take Plant B 3 where you have got a random failure of the standby liquid 4 control, is that what SLC is?
5 MR. CHOKSHI: Yes, standby liquid controls.
6 CHAIRMAN SIESS: How does that -- what is that 7 random failure? Is it a value or a pipe break?
8 MR. CHOKSHI: I don't know which component it is.
9 CHAIRMAN SIESS: Or human error?
10 MR. CHOKSHI: But standby liquid system has 11 failed. I don't know exactly which component leads to that 12 failure.
es
(-) 13 CHAIRMAN SIESS: But Mr. Michelson's question is, 14 is that probability the same with and without the 15 earthquake? And it seems to me you can't answer that until 16 you know the actual mode of failure.
17 MR. CHOKSHI. But I believe they probably use the 18 same failure rates. Very few PRA change.
19 CHAIRMAN SIESS: I suspect the answer is you don't 20 know the failure --
21 MR. CHOKSHI: Yes, I don't know the --
22 CHAIRMAN SIESS: -- that good in either case, and 23 you make them the same, because we've been consistently told 24 that PRA analysts use the same fragilities of reliabilities
! 25 for QA Class 1 systems versus non-QA'ed systems.
(
Heritage Reporting Corporation (202) 628-4888
(
L_-__-____ _ _ _ _ _ _ _ _
pA 179 1 MR. MICHELSON: See, there is other affects. Of 2 course, even pipe loading effects on valves or on pumps and 3 so forth would affect the probability if failure of the 4 devise and we don't include any of that.
5 CHAIRMAN SIESS: This failure may not be a value.
6 It might have been a human error, and that might be higher 7 with earthquakes. I don't know.
8 MR. CHOKSHI: Yes, I don't know. You are right.
9 It may be human action involved.
10 MR. MICHELSON: There is even a more serious 11 limitation on PRA, and that is, we use the data base 12 generally comes from nominally-loaded equipment. And when
(\
13 in reality during the accident, the equipment is under its 14 accident-loading conditions.
15 CHAIRMAN S1ESS: Yes.
16 't. MICHELSON: And that can affect significantly 17 the operability of some types of components. And we don't 18 include any correction for that either. So we are using !
19 very optimistic numbers from that viewpoint, at least.
20 MR. CHOKSHI: Those two slides were to basically 21 explain what has been high consequence sequences in wall 22 structural failures at very high accelerations, and --
23 CHAIRMAN SIESS: What are these slides taken from?
24 MR. CHOKSHI: These are taken from a Sargent and l
g 25 Limney study. I don't think it's been published yet.
\J Heritage Reporting Corporation (202) 628-4888 l
I
l p.-
180 1 CHAIRMAN SIESS: Okay.
t.
2 And the previous one also?
i "'
3 MR. CHOKSHI: The previous one were later from 4 Paul Amaico to us on contract, and that will be published 5 very soon, his report.
6 MR. BAGCHI: We have made comments on these 7 contractors reports, and they haven't come back.
8 MR. CHOKSHI: Yes.
9 MR. BAGCHI: As soon as they are in some kind of 10 concrete to look at, we will send it out.
11 MR. CHOKSHI: And this is just to show that.from a 12 structural point of view, and this is a conservative 13 analysis, you still see a very high structural capacity and 14 for looking at some other containments.
15 And the reason for showing this is that I want to 16 make some preliminary reports in sites and get some 17 comments.
18 - MR. MICHELSON: Now, you are looking here at the 19 containment structure per see. You're not looking at the 20' isolation valves --
21 MR. .CHOKSHI: No, we are.
22 MR. MICHELSON: -- which are part of it.
'23 Oh, you are including the valves?
24 MR. CHOKSHI: Yes. That's what I want to talk 25 about.
Heritage Reporting Corporation (202) 628-4888
m 181 1 CHAIRMAN SIESS: Table 9.1 is strictly structural, 2 and go back to that table just a minute, will you?
3 MR. CHOKSHI: The 9.1?
4 CHAIRMAN SIESS: Yes. There is a statement of --
5 MR. CHOKSHI: This is strictly structural. You 6 are right.
7 CHAIRMAN SIESS: -- a limit state.
8 What's the difference between Task 3 and Task 57 9 Let's see, seismic plus dead.
10 MR. CHOKSHI: Yes, they are on a little bit 11 different load combinations and different failure 12 mechanisms.
l (G/ 13 CHAIRMAN SIESS: Okay.
14 But now when it says a limit stato is a failure of 15 biological shield wall, again it was assume at some point 16 that that failure meant a release to the atmosphere.
17 MR. CHOKSHI: Yes, this was not done in the 18 context of a PRA. It was just done ss a structural 19 evaluation.
I 20 CHAIRMAN SIESS: And somebody decides if the l
21 biological shield wall fails, that there is going to be an 22 opening to the outside.
23 MR. CHOKSHI: There could, yes.
l CHAIRMAN SIESS:
24 I mean, but they are not putting
- 25 -- the probability might be only one in 10, but they don't
\_/
Heritage Reporting Corporation (202) 628-4888 l
l 1
L ___ - __
L l'
\ ).
182 1 look at that.
2 MR. CHOKSHI: They don't, right.
3 CHAIRMAN SIESS: They stop the analysis at some 4 point.
5 MR. CHOKSHI: Well, they think that's very hard 6 to --
7 CHAIRMAN SIESS: Who helps the PRA analyst decide, 8 you know, if there was ever a systems problem, this is it.
9 The structural engineer can tell you when that wall is going 10 to fail, although I don't know what failure means, whether 11 that means it's going to fall over or just means it's going 12 to be stressed beyond the yield.
(D N/ 13 MR. BAGCHI: I think we should not take this to 14 raean too much. Only to the extent that this is such a high 15 resistance level. That's the only message we wanted to 16 leave with you.
17 CHAIRMAN SIESS: Okay, You don't want to go any 18 higher.
19 MR. BAGCHI: That's correct.
)
20 CHAIRMAN SIESS: In other words, you look at .45g 21 that cause a shear failure of the basemat, however what i 22 shear failure is. And you said that's high enough. I'm not 23 going to worry about it.
24 MR. CHOKSHI: It's pretty high.
25 CHAIRMAN SIESS: No need to go any further.
Heritage Reporting Corporation (202) 628-4888 i
l L
i
(^3 i
V 183 l l
1 Okay, I buy that approach.
1 l 2 MR. CHOKSHI: And the point is that, at least from l l l 3 what we have seen since all these calculations are linear j 4 type calculations, that the high consequence sequences which 1
5 results from a gross structural failure, it is calculates at J l
6 a very high level of outputs. And the first point is that l
i 7 the margin approach really cannot address those type of !
d 8 failure modes because they really don't look beyond .45g,
- 9 and the question is do we want to look at it.
10 CHAIRMAN SIESS: Yes.
11 MR. CHOKSHI: Is that a high enough g? And I 12 think the current thinking, and I think it's this one, that
/N
-) 13 maybe we don't have to look at structural type of failures I
14 for at least some type of containments. And maybe we shculd '
15 be fectising on something else more prescriptive area 7 16 requirements than sam:4 other areas like the things in I 17 isolation, and the heat removal function failures.
18 CHAIRMAN SIESS: Yee, 19 MR. CHOKSHI: Which may be more pertinent from the 20 accident point of view.
21 CHAIRMAN SIESS: Now that third item on reactor 22 supports.
23 MR. CHOKSHI: Right. We think that's --
24 CHAIRMAN SIESS: That's the thing we've been g es 25 looking at?
b Heritage Reporting Corporation (202) 628-4888
i dfm -
184 1 MR. CHOKSHI: Yes. 3 1
2 CHAIRMAN SIESS: Okay.
3 MR. CHOKSHI: And we've talked to a number of !
4 people, including Bob Kennedy, and what has happened that in 5 one. case some estistate were made. It's very hard to get )
6 information from the vendors to accurately characterize the 7' fragility of --
8 CHAIRMAN SIESS: I'm thinking about the recent 9 instance of the Trojan reactor vessel supports, for example, 10 with the embrittlement problem.
11 Now, is this just a broader --
-12 MR. BAGCHI: No , it's just a broader treatment.
13 CHAIRMAN SIESS: Okay.
14 MR. BAGCHI: In BWI MARC I's, for example, the 15 reactor vessel support may have less resistanco to
'16 earthquake. Those kinds of things.
17' 'MR. MICHELSON: Now, do you look at then, from the 18 viewpoint of the condition of the material after prolonged o ,
19 radiation?
20 MR. BAGCHI: If it's a significant concern, we 21 will consider that.
22 MR. MICHELSON: It's a little difficult to 23 calculate the affects of radiation embrittle:aent, you know, 24 how the metal will respond during an earthquake?
25 CHAIRMAN SIESS: But some BWR supports don't get Heritage Reporting Corporation (202) 628-4888 E_______________ _ _ _ _
185 !
I any radiation.
2 MR. MICHELSON: No , I'm thinking of PWR.
3 CHAIRMAN SIESS: Oh, okay.
4 MR. MICHELSON: But just trying to understand how 5 it works.
6 CHAIRMAN SIESS: Well, we've been looking at that o 7 in the context of recent problems. I mentioned the Trojan.
8 MR. MICHELSON: But it8s harder to do a seismic --
9 MR. BAGCHI: We really haven't determined that 10 radiational effect is contributor to initiate seismic 11 failure. We haven't determined that that's causing the 12 seismic failure.
,- y
(_) 13 MR. CHOKSHI: The bottom line is that we are still 14 looking at this issue, and I guess we are leaning toward 15 coming up with some very prescriptive, rather than doing a 16 full system and fragility evaluaticn, look at the things in 17 isolation and type of failure physically in the control 18 actoation systems, and control systems which are the type of 19 problems which can cause isolation failures.
20 MR. MICHELSON: In looking at the boiler water 21 reactor, how did you view the vent pipe header and the other 22 components that keep the suppression pool separated from the 23 drywell?
24 If those were to fail, of course, then the q 25 containment very quickly pressurizes and certainly it's a
(_/ j Heritage Reporting Corporation ])
(202) 628-4888 i
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ]
- p/
s_
186 1 9 od likelihood of fracture and failure.
2- But did you analyze then the vent pipe headers in 3 this process, for instance?
4 MR. CHOKSHI: In the PRAs, the --
5- MR. MICHELSON: Well, PRA doesn't do anything for 6 you. You've got to go in and do the analysis to tell the 7 PRA --
l l
L 8 MR. CHOKSHI: The fragilities were low depth.
9 MR. MICHELSON: Yes, you've got to do some kind of 10 calculation, don't you?
11 MR. CHOKSHI: Yes.
12 MR. SHAO: But in the PRA, the fragility analysis 13 that go in, the head analysis, the analyses go in the 14 fragility analysis.
15 MR. MICHELSON: Do you model the vent pipe 16 failure?
l 17 15L BAGCHI: It was not significant for Peach 18 Botton 2.
L 19 FDt. MICHELSON: Do you actually model it in there?
l l 20 MR. BAGCHI: Yes.
21 HR. MICHELSON: And for the event, once it failed, 22 -it was not significant?
23 MR. SEAO: Once it failed, important.
24 MR. MICHELSON: The probability of failure of the 25 containment is virtually --
L O
!' Heritage Reporting Corporation (202) 628-4888 1
?
J 187 n
, '1 MR. SHAO: Once it fails, the containment goes.
2 MR. MICHELSON: No, it's not a part of the 3' . containment. .You know what'the vent pipe header is.
4 MR. SHAO: . Yes, I know, but if that one fails, the-5 containment get a higher pressure.
6- MR. MICHELSON: It would be very quickly, in a 7 matter.of minutes, yes. It bypasses suppression.
8 CHAIRMAN SIESS: What the probability that the.
.9. . header will fail? Somebody must compute that?.
10' MR. CHOKSHI: We don't calculate for events or 11- venting.the --
12 MR. SHAO: I think the fragility of the head is
'13 very --
14 CHAIRMAN SIESS: The question is on a PRA.
15 Did somebody calculate the pressures of -- the 16 icads on the header?
17 MR. BAGCHI: We don't know the fragility nuniber.
18 for the headers. We can give you thosis. 7.'m qui.te sure 19 that --
20 CHAIRMAN SIESS: But there would be one, 21 MR. MICHELSON: But there would be one is any 22' question. I want to know if you codeled it.
-23 MR. BAGCHI: Well, it depends. To go back and --
24 for Peach Bottom, we'll give you an answer.
.o. 25 CHAIRMAN SIESS: Okay, that's good enough.
Heritage Reporting Corporation (202) 628-4888 i
L N 188 1 CHAIRMAN SIESS: Suppose things had come out a 2 little differently on those Trojan reactor vessel supports.
3 Would that have been a vulnerability?
4 MR. BAGCHI: We would have to look at the age of 1
5 the plant and it would have to be --
6 CHAIRMAN SIESS: I'm saying for Trojan.
7 MR. BAGCHI: Yes, for Trojan it would have been --
8 CHAIRMAN SIESS: . Plant unique.
9 MR. BAGCHI: Yes, it would have been a 10 vulnerability.
.11 MR. SHAO: For the ' trojan, the radius for the big 12 loading is not the seismic. The big loading is the pipe O.
13 breaks.
14 CHAIRMAN SIESS: It doesn't make any difference.
15 .I'm just talking about vulnerability now.
16 MR. SHAO: The pipe break is --
17 CHAIRMAN SIESS: You can't get out of this by 18 saying that's internal vulnerability. I'ta just still 19 talking about whtt's a vulnerab!13ty. Big loading a pipo 20 break, I'll admit.
22 But again, if either a pipe break -- wc>uld the 22 fact that it took a pipe break to fail it make it not a R3 vulnerability because it's such a low probability now?
24 MR. BAGCHI: If it had weakness, the seismic load l 25 is fairly tgh on that support system. If we were to assume Heritage Reporting Corporation (202) 628-4888
189 1 1 the seismic load was determining the resistance of the 2 rector vessel support, then it would show up as a 3 vulnerability.
4 CHAIRMAN SIESS: Yes. Now, that's the case where 5 it would meet the design basis, wouldn't it?
6 MR. BAGCHI: Right, it would meet the design 7 basis, and yet have vulnerability.
8 MR. MICHELSON: How do you handle tbe large 9 penetrations like say the personnel airlocks? Is that done 10 as a separate analysis? Is that later?
- 11. CHAIRMAN SIESS: I think we'll get to that.
12 MR. BAGCHI: I think that Table 9.1 that you have 13 in the presentation of the material has looked into those 14 penetrations.
- 15. MR. MICHELSON: Now, do you look at them at the 16 viewpoint of just the structural integrity, or do you look 17 at such things as pncurr.atic semie en them and the loss of' 13 the pn<ntnatic system tant keeps the seals --
19 MR. BAGCHI: If it's triggered by high 20 torporature, year that gets looked into it. But it's lt il primarily atnctursl integrity.
22 MR. CHOKSHI: But the:.e was some other work done 23 on --
l 24 MR. MICHELSON: But you worry about the seals if 25 'you get high temperature, or you worry about the seals if Heritage Reporting Corporation (202) 628-4888
l %-d l' 190 l
1 you get a serious seismic event that knocks out the 2 nonqualified air system.
3 CHAIRMAN SIESS: Now, you mentioned a lot of 4 things here, including structural-integrity, heat removal 5 functional function and isolation failures. And you've 6- talked about structural integrity. You had something on 7 bypass, although not as specific. Heat removal, you have a 8 service water system failure, right?
9 MR. CHOKSHI: On one of the --
l 10 CHAIRMAN SIESS: On one of them. But you didn't 11 say anything about isolation failures, did you?
I 12 MR. CHOKSHI: Yes, the last --
13 CHAIRMAN SIESS: You had one with relay chatter 14 and a valve.
15 HR. CHOKSHI: Yes, that's the one area I thinK we L 16 need to develop some requirements so people can look at type l
L 17 of what can cause the isolation failurs in the seismic, and l=
L 1B it appears that things you want to leok at is the actuation 19 systems and some of the control systems where the signals 20 initiate isolating the containment.
1' l 21 CHAIRMAN SIESS: Ihen an awful lot of valves have I
22 to close.
23 Mh. CHOKSHI: Yes, I think it's not going to be 24 more hardware. It's more of a system response type of 25 issue.
, Heritage Reporting Corporation L
(202) 628-4888
i <
l
(\
j 191 i 1 CRAIRMAN SIESS: I don't think you need an ]
l 2 earthquake to have a hardware problem. We've got valves 3 that aren't closing sven when there are not earthquakes.
4 HR. CHOKSHI: Yes. l 5 CHAIRMAN SIESS. Including swing check valves. I ;
6 don't know how it could be much worse.,
l 7 But you n.entioned penetrations back under 8 structural.
9 MR. CHOKSHI: Right.
I 10 CRAIRMAN SIASS: This would be deformations that 11 would atfect penetrations.
12 MR. CHOKSHI: Yes. And at least from the 13 penetrations which have been looked in the past does not 14 appear to be any problem at low g's. We have to go very 15 high g's when you start seeing the penetration problems.
16 CHAIRMAN SIESS: Mr. Michelson was asking about ,
17 door seals, right?
18 MR. MICHELSON: Yes.
19 NB . CHOKSH4: Again, I think that's -- we go to 20 high g's when the problems we' re looking, you know, g
/
21 NR. MICHELSON: Well, I s'uspect the seal itself 22 will go to high g, but I'Ia wondering if the air system 23 that's keeping it pumped up is going to go to a high g. And i
L 24 if you lose the air system, the seal is gone. It deflates.
25 And in the case of equipment penetrations, there is no I /_Y l (~/ l
} Heritage Reporting Corporation j
! ;202) 628--2888 l
l i
/~
,i-) 1 192 l 1 double doors cr anything. It gets to be pretty leaky.
2 CHAIRMAN SIESS: Some of them are -- all of them i e i 3 are air seals? d 1
4 MR. MICHELSON: No, no. Some vendors -- only sOme l l
1 5 people have used air sea.ts.
6 HR. BAGCHI: In some plants the air systems are 7 .not safety related and you would have to look at that.
8 CHAIRMAN SIESS: Wouldn't the air system inflated W 9 seals be safety related?
10 MR. CHOKSHI: There is no backup system. ,
I 11 MR. MICHELSON; Unfortunately not.
12 MR. CHOKSHI: No, no, not in older plants, f%
q) 13 MR. MICHELSON: Although it's performing a safety 14 function, it's keeping the containment seal.
15 CHAIRMAN SIESS: I think it's safety related, but 16 they don't call it that. Okay.
l 17 It seems to me that you have ten 6ad to eliminate 1 18 major structural failures as being a pretty low probability, 19 but it still leaves some other thinge --
20 MR. BAGCHI: Some air systems have accumulators.
21 You know that. But --
l l
22 MR. MICHELSON: Unfortunately, not on these door j 23 seals.
l
.. 24 MR. BAGCHI: Not on the door seals. I
/
25 MR. MICHELSON: They put them on valves and so j I
Heritage Reporting Corporation i (202) 628-4888 l l
l l
, A l ao'-
1 J
J
.,x L,b U 193 1 forth, but they never recognize the door seal as being kind 2 of essential, and I think we need to go back and I think IPE 3 ought to be picking this sort of thing up.
4 MR. BAGCHI: That's a good point.
5 CHAIRMAN SIESS: The probability of failure to 6 isolate I think is fairly high even without an earthquake.
7 I've seen figures in 5 percent up and down. And I guess the 8 important question to me would be how much is it changed by 9 the earthquake.
10 MR. MICHELSDN: Yes. Some of the big Persian vent 11 valves on containments are still open, although most plants 12 have tried to get those closed up and substitute smaller 13 valves. But some of the larger ones even are cocked open at 14 certain failure, 15-degree angles and --
15 MR. BAGCHI: But even if you had mechanical doubt 16 of reliability, you have a problem of the relay chattering 17 and keeping it forced open. That's something that was found 18 in Seabrook.
19 (Slide presentation.)
20 MR. REITER: I've gotten a not so subtle message 21 from my colleagues that they are waiting to talk after me, 22 and I shouldn't take up too much time.
23 I'm going to talk about hazard, particularly as it 24 relates to the review level earthquake in seismic margins 25 and a little bit about the use in seismic PRA.
O Heritage Reporting Corporation (202) 628-4888
l L- i em l
l Ib 194 1 There is.no doubt that this has been a knotty 2 issue from day one, and our thoughts have been evolving on 3 this. And in the spirit of what other people have told you 4 beforehand, we would like to expose you as to where our 5 thoughts are going and elicit your comments, because it is a 6 tricky issue. It's one that we've had a lot of interchange 7 with outside consultants, people within NRC, NUMARC and the 8 thought have evolved.
9 And the first topic I would like to talk about is 10 review level earthquake, and I thought I would outline some 11 of our basic concepts and run you through a few sensitivity 12 tests, looking at EPRI in Livermore has the calculations,
(
x/ 13 and then talk to you a little bit about seismic PRA.
14 Generally, an approach to the review level 15 earthquake, again that is the earthquake which we tied to 16 the specific screening level that's inherent in the seismic 17 margins approach. We've had several basic approaches or 18 basic aspects of our approach.
19 One, we would like to use both the Livermore and 20 EPRI hazard results to develop consistent selection of 21 review level earthquakes. Both these programs are 22 essentially state-of-the-art programs. I don't think we 23 have any equivalent anywhere for a seismically inactive 24 areas. Anywhere in the world nobody has done programs like
_ 25 this.
V Heritage Reporting Corporation (202) 628-4888
_.-n.
\_
195 1 The second thing is we feel that most plants would 2 use .3g. And originally, in fact, to give you an idea of the 3 evolution, we had originally thought that for the eastern 4 United States, east of the Rocky Mountains, all plants would 5 use . 3g, and that was sort of based on various assumptions.
6 One of them is that we thought we had a good --
7 could tie in the .3g review level earthquake with the 8 Charleston earthquake. But, unfortunately, recordings of 9 earthquakes that have actually occurred, and changing ideas 10 of ground motion have indicated that .3g is not an 11 appropriate representation of the Charleston earthquake.
12 Similarly, there has been a lot of pulling on
\ ,, 13 various sides. Namely, it was felt that there may be some 14 plants out there in the eastern United States which .3g may 15 be an inadequate approach for seismic hazards for screening.
16 We have a plant design base is .3g in a very high seismic 17 area, and essentially just by looking at .39 you wouldn't be 18 extending that search for the so-called vulnerabilities.
19 CHAIRMAN SIESS: Would you use that .3g to look at 20 liquefaction?
21 MR. REITER: No, I think the liquefaction would be 22 assoc -- the g level we pick is also a g level for 23 liquefaction, whatever was picked.
24 CHAIRMAN SIESS: Because there are some sites that 25 I don't think have a chance of taking .3g.
ts Heritage Reporting Corporation (202) 628-4888
y, ,
pq l' 'i ,.
Vi
$[ '
196
[ ,
~
, l' MR. REITER: Well, we've been looking at that.
1 2 By the way, liquefaction -- we haven't mentioned h 3 . liquefaction, and that really is important concerning --
l-l, 4 LIAIRMAN SIESS: Do you recall that --
I -
H 5 MR. REITER: Hatch is the first plant --
6 CHAIRMAN SIESS: Yes.
7 MR. REITER: -- that we are really looking at in P 8- seismic design margins is looking at liquefaction, and we've.
9 had some outside reviewers on this, and I think we are going 10 .to be learning a lot from that review, and want to 11 incorporate those lessons into our prescription of how to 12' deal with' liquefaction.
I) 13 CHAIRMAN SIESS: Now, this was pointed out a long 14 time ago, back when the proposal was made and actually being-15 considered to put a .2g floor on the SSE. Jim Devine 16 pointed out'that that was. going to give us a problem at 17 otherwise acceptable plants in terms of liquefaction.
18 HR. REITER: Well, one of the things I think that 19 we learned from the Surrey evaluation was, just because you 20 have liquefaction does not necessarily mean you have core 21 damage or plant failure. Essentially you can have 22 liquefaction essentially which has relatively little impact 23 on the plant. It depends where that liquefaction is and how 24 much and what's actually happening to the buildings.
25 CHAIRMAN SIESS: Yes, which is an area that can Heritage Reporting Corporation (202) 628-4888 m
_________.___._________.___.m._ _ _ _ . _ . _ _ _ _ . _ _ _ , _ _ _ _ _ . _ _ _ _ _ _ _ _ _ . _ . _ . _ _ . _ _ _ _ _ _ _ _ = _. _ _ _ . _ . _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ . _ _ . _ _ . -_ . _ . _ _ _ . _ _ _ . _ _ . _ _ _ _ . _ _ _ _ . _ _ . . _ _ _
0: 7.,
197 H
b :1 lead to endless argument.
\~
'2 MR. REITER: Yes.
3 So we Owlt that there might be some plants out 4 there that'would require a higher g level than .3 to .5.
.5 -And we felt that if the hazards at those planta were 6 consistently higher, that would be a criteria for looking at 7 that.
8 Similarly, it was felt that there may be some 9 plants out there which had such low hazard that even .3g
'10 - criteria may not be appropriate. You may not get your 11' money's worth in terms of that-you would if you would have 12 had really a very thorough walk-down.
You can talk about 13 perhaps some plants in Florida and Texas where everybody 14 agrees a seismic hazard is extremely low.
15 The other concept we have aside from these " bins" 16 was that we wanted to try and place this hazard in some sort 17 of context of core damage. And that's been a very difficult 18 . thing for us to do. And we have been wrestling with that, 19 and we think we have come up with some limited insights.
20 The report we talked about in the past was there 21 seems to be some generic relationship -- some, and that's a 22 very great parenthesis -- generic relationship between the 23 likelihood of exceeding the HCLPF and the core damage
'24 associated with that particular plant.
25 And people talk about a factor of 10 if one is Heritage Reporting Corporation (202) 628-4888
l t )
J 198 1 looking at the mean, and an even more vague relationship is 2 about the same if one is looking at the median. But again, 3 those relationships -- so the idea is to try and use this 4 hazard screening, put it in some sort of loose context of 5 core damage, but not overly rigorous because we don't think 6 it can be done.
7 The issues in trying to carry out these particular 8 ideas are several.
9 First of all, which parameters should we use, and 10 I'll show that right away. Which measure of seismic hazard ,
11 should I use?
12 The other thing which frequency. By frequency, I m
k-) 13 mean like which frequency of motion, not the likelihood of 14 occurrence.
15 And the third is how do I treat the Livermore and 16 EPRI. Do I treat them separately? Do I combine them?
17 Now, before I go on, I would just like to point 18 out that the idea of a .3g or the use as appeared as an 19 efficient way to go, we want to put in most of our plants 20 appears in several write ups in the seismic margins program.
21 And here is one we found in the EPRI write up, and it says 22 things that if too low a tri-level is set, then -- well, I 23 guess the bottom line, the important thing is that for most 24 plants in a low seismic regions, it appears most reasonable
- 25 to form the SMA -- that's the seismic margin analysis -- for
'J Heritage Reporting Corporation (202) 628-4888
i
'\ i
~.)
199 1 trial SME, which is a view level of either a .25 or .3g. It 2 can still be conducted cost effectively. Most elements will 3 pass the review and a small number of elements are not 4 likely to pass.
5 So there is a general feeling based upon past 6 experience efficiency that .3g is -- if we can find a 7 justification for it, is a good way, an efficient way to 8 look at plants, to carry out an efficient way to search for 9 the vulnerabilities.
10 CHAIRMAN SIESS: Now, that's a definition of 11 vulnerability of something that would have a HCLPF of less 12 than .3g, maybe.
(
's
- 13 MR. REITER: No, that's not -- I want to point 14 that out. We clearly make a differentiation between review 15 level earthquake and vulnerabilities.
16 CHAIRMAN SIESS: Okay.
17 MR. REITER: The review level earthquake is a 18 target for which one examines the plant. What one uncovers 19 after that may or may not be a vulnerability.
20 CHAIRMAN SIESS: Okay. Now, if you set the review 21 level too high, you will find that an awful lot of things 22 will fail, right?
23 MR. REITER: Yes, you might find more things that 24 will fail. You decrease the efficiency of the methodology.
25 CHAIRMAN SIESS: Yes, okay.
%s Heritage Reporting Corporation (202) 628-4888
g-"%
' .J 200 1 MR. REITER: In that you have screened out less 2 material.
3 CHAIRMAN SIESS: So if you go down to .25 or .3, 4 you are going to reduce the number of failures.
5 MR. REITER: Right.
6 CHAIRMAN SIESS: Now, you've got a terrain here 7 with various peaks and valleys, and what you're trying to do 8 is lower your level down to where there is not too many 9 peaks sticking up, right?
10 MR. REITER: I think one of the previous slides, 11 and maybe it was Neal or Andy had it, we have this -- there 12 is two things. We want to be able to uncover the
\-) 13 vulnerabilities and obtain efficiency. We have to find out 14 where that particular place is. And the feeling is that for 15 most plants in the United States is .3g.
16 There may be some plants which may require higher 17 and some plants that might be lower.
18 CHAIRMAN SIESS: But if there is one peak that 19 sticks way up --
20 MR. REITER: Yes.
21 CHAIRMAN SIESS: -- you would pick it up anywhere.
22 MR. REITER: By sticking high up you mean - -
l 23 CHAIRMAN SIESS: It might stick up to 55 or so, so l
24 you would pick it up at .5. But you want to get low enough 3 25 to pick up that one and some that might be sticking up Heritage Reporting Corporation (202) 628-4688
..v(}.
201 1- around it.
2' MR. REITER: Right. If the fragility is -- if the 3 fragility is greater than .3 9, if the HCLPF is greater than 4 .3g, then we won't pick it up in the .3g screening.
5 CHAIRMAN SIESS: That's right.
6 DR. SHEWMON: Leon, why don't you use two times 7 the SSE for all plants, assuming that the.SSE is what they L 8' were designed to and thus, two times is something -- you 9 .know, the same uniform probability of something untoward 10 occurring?
11 MR. KEITER: That's a very good thought, and 1:2 initially, in fact, we paid Livermore, one of our contracts, 13 to go out and to look at all the plants in the United 14 States, compare the .3 and .5g levels to multiplies of the l
l 15 SSE.
16 The SSE, one and a quarter, up to two times the 17 SSE, and we -- I think that's still there and we might use 18 it. The problem is that the SSE is not necessarily a very 19 robust indicator of hazard or either of plant capacity. And 20 the idea was that, as somebody once said, if you have two 21 plants in the same area, one with a .lg and one with a .25g, 22 why punish the plant, the .25g by forcing them to go to a i 23 much higher evaluation just because they have a higher SSE.
l l 24 What you are saying is a logical thought which we 25 considered, and we have the information, and we may look at Heritage Reporting Corporation (202) 628-4888
202 l
1 it. But I think at this point we're not ready to embrace 2 that as the single way to proceed.
? CHAIRMAN SIESS: The SSE, as you said, isn't'that 4 well related to the hazard.
5 MR. REITER: It's not that related.
6 CHAIRMAN SIESS: Even to the hazard as we 7 envisioned it at the time the SSE was set.
8 MR. REITER: Well, it's not only that, but 9 remember, the way we defined the hazard according to 10 Appendix A is to take the largest earthquake and float it.
11 CHAIRMAN SIESS: Yes.
12 MR. REITER: And you can float it for a thousand
,e
(> 13 miles. You can float it for 10 miles, and probabilistically 14 they may have very large differences. But in the context of 15 Appendix A, there is no difference. That's always one of 16 the problems we have, and that's why the SEP program, for 17 instance, we didn't rely on Appendix A evaluations. We went 18 towards using more probability to make sure that we don't 19 unduly punish some plants and not give other plants enough 20 of a ground motion level to look at.
21 CHAIRMAN SIESS: Good.
22 MR. REITER: Well, just quickly, the problem about 23 parameters, and this is a stick issue and the one I get 24 yelled at the most, and there are some people here, I'm 7,3 25 sure, who are ready.
O Heritage Reporting Corporation (202) 628-4888
h i
[
g l'
V O, .
203 i l
1 By the way, we have a very, let's call it a j 1 2 vibrant committee, who people don't hesitate to tell each
'3' other when they think they are off the wall or wrong, and 4 it's really probably one of the best committees I've ever 5 workred with in terms of the people on there and what they 6 bring to it. So'I think it's really good.
7 One of the parameters we can look at in the 8 hazards, okay, first there is the mean, and the mean is'very 9 attractive because --
10 CHAIRMAN SIESS: You're talking about the hazards 11 right now?
12 MR. REITER: Right. Remember, we're talking 13 about -- we have these various three bins, okay, and 14 essentially will be driven by the hazard, and how are we 15 going to put these plants in the various bins.
16 Well, the first thing is the mean, and the mean is 17 very attractive because it's risk related from an actuarial 18 sense. You know, insurance companies use'it. If I go to 19 PRAs, if I want to know the mean core melt, all you have to 20 do is take the mean hazard curve, co-valve it with the mean 21 fragility, and I get the mean core melt.
22 On the other hand, the problem with the mean is 23 that it's dominated by the treatment of outliers or 24 extremes. Very often what completely controls the mean is 25 whether or not I include, even at a low likelihood, some Heritage Reporting Corporation (202) 628-4888
!q.
Q '
204
(
1 particular thought about what the hazard may be.
2 So'the treatment of the extreme or the outliers 3 can completely control what that mean will be. As a result, 4 because there is some difference of treatment between 5 Livermore and EPRI, there are very large differences between 6 the EPRI means and the Livermore means. So that's the good 7 and the bad.
8 On the other hand, we have a median, which the bad 9 side is not related to risk. It really does not reflect the 10 uncertainties. But on the other hand, it's really the least 11 affected by the outliers.
12 For someone like myself, and I'll say is a (m
(./ 13 personal view, if someone asks me what do I think is the 14 best estimate of the central tendency of the hazard, I would 15 gravitate toward the median,-and I've put a question on 16 that.
17 CHAIRMAN SIESS: Did you use --
18 MR. REITER: And what's really good, there is --
19 from my point of view, there is the least difference between 20 EPRI and Livermore.
21 CHAIRMAN SIESS: Did you use best estimate in a 22 statistical sense there?
23 MR. REITER: No. I'm just saying I, as a 24 seismologist, what do I think -- if I had to look at five 25 plants and I wanted to rank them.
Heritage Reporting Corporation (202) 628-4888
205 1 CHAIRMAN SIESS: I know. I know what you mean.
2 But the best estimate has a meaning to statisticians.
3 MR. REITER: Right, and I'll talk about that right 4 away.
5 CHAIRMAN SIESS: But that's not what you --
6 MR REITER: No, no.
7 CHAIRMAN SIESS: Okay, just be careful when you're 8 talking to the full committee.
9 MR. REITER: That's with little b and little e.
10 And by the way, and some people have been saying 11 we have these hazard studies out there and they are 12 terrible. You know, they are all over the place. And the 13 answer is that's not true.
14 The answer is that compared to previous hazard 15 studies, we're getting a lot more agreement between EPRI and 16 Livermore than we have seen in the past. And I will show i
17 you some comparisons of that.
18 The trouble is, though, if we concentrate on my 19 ability to define the extremes of distributions at very low 20 probabilities or very high ground motions, that's where I l
21 have the greatest difficult. And, unfortunately, that's the l 22 area which dominates the mean, and for many of the cases the 23 mean core melt in PRAs.
24 CHAIRMAN SIESS: With the spread you've got on 25 those curves, if you keep spreading them, they've got to get Heritage Reporting Corporation (202) 628-4888
- i l/}
d-e . !
206 I 1 closer together.
2~ MR. REITER: No, no. I don't --'I think we have 3 really come a:long way. And some day if you want, I'll show 4 you some before and after comparisons.
5 CHAIRMAN SIESS: I know.
6 MR. REITER: But in the past, we never really 7 looked at means of distribution. We concentrated'on single 8 estimates. .Those single estimates have come.a lot closer 9 -now than they have in the past. i-10 CHAIRMAN SIESS: If you and EPRI both assumed a
- 11. uniform distribution, you could still argue about the 12 extremes.
75 k )'
m 13 MR. REITER: Well, it's.not a uniform -- well, 14 it's not a uniform distribution. The idea is I'have a 15 conglomerate of views out there, and how do I treat that 16 conglomerate of review. And depending on how I treat it can 17 essentially -- essentially how I treat that dominates what 18 my mean hazard, and therefore mean core melt.
19 20 21 22 23 24 O
Heritage Reporting Corporation (202) 628-4888
l 207
.1 MR. REITER: Another parameter or 85th percentile, 2 just,the 85th percentile distribution, has characteristics 3 which are the sar.e where between, namely which are somewhat 4 related to risk and the central tendency. And it seems to 5' me that this is sort of an intermediate difference between 6 the Livermore and EPRI results.
7 CHAIRMAN SIESS: Where would that lie in relation 8 to the mean?
9 MR. REITER: Well, I will give you an example.
10 For the Livermore curves, the mean in most cases is about 11 the 85th percentile level, in a few cases it is less, and in 12 many cases it is above. In the EPRI case it is always
{ss) 13 below. The mean is always below the 85th percentile. It 14 all depends again on the distribution.
15 Now something called the best estimate which is 16 like a best estimate in the context that you thought about 17 and which is calculated by Livermore, it is not rigorously a 18 best estimate. But each one of the experts was asked to 19 pick what they thought was the single, if they had to pick 20 one choice, what was the single choice that they would pick 21 in terms of zonation and maximum earthquake. We asked the 22 same thing of the ground motion experts, the singlo best 23 choice. And this best estimate is just an average of all of 24 those.
~,1 25 So what wt like to say is that if you force each Heritage Reporting Corporation (202) 628-4888
/'
208 1 expert to say rid yourself of your own uncertainty and give 2 me one choice and we average those choices, that is what we 3 would get.
4 CHAIRMAN SIESS: They were asked for one choice 5 for regulatory use?
6 MR. REITER: Let's say that you are an expert, and 7 I say give me one choice of the zones, one choice of what 8 the maximum earthquake in each one of the zones were. And 9 say I asked a ground motion expert and I say you have lots 10 of ground motion models, pick the one ground motion which 11 you think is the best. And then we have a number of experts 12 and we average that. Now the problem is that we do not have
,m
(_; 13 such an estimate available from EPRI. These three we do.
14 And another issue is frequency. By the way, what 15 we have available is information at these frequencies, peak 16 ground acceleration, I hertz, 2.5 hertz, 5 hertz, 10 hertz, 17 25 hertz, and a weighted combination which gives equal 18 weight to 2.5, 5, and 10 hertz, and half weight to the 19 PGA. That is the information available.
20 CHAIRMAN SIESS: These are available?
21 MR. REITER: You can get that from EPRI.
22 CHAIRMAN SIESS: These are available from your l
1 23 expert panels?
l 24 MR. REITER: We ran the calculations such that we
_s 25 got this information. We do not limit ourselves only to V
Heritage Reporting Corporation j (202) 628-4888 )
{
_ _ _ - - - - - - - - _ - _ _ _ - - - - - - - - - _ - - - - - - - - - - - - - - - - - - - .- - - - - - - - - - - - - - - - - - - - - - - - - - _ _--- J
l-F di-209 1 PGA.
2 CHAIRMAN SIESS: Oh, okay.
l 3 MR. REITER: Now the advantage to using PGA is 4 that it is most familiar, and most of the PRAs are governed 5 by PGAs. So it is most simply related to PRA. The 6 advantage of using some other element like 5 hertz or a 7 weighted combination is that they are more related to 8 damage. And in fact some of you may know that in the 9 Diablo Canyon PRA that the hazard is being scaled not by the 10 PGA but by an average over 2 to 10 hertz depending en what 11 you want. So this is a more damage related way to go.
12 Now by combining EPRI and Livermore, the first and p
j A,) 13 most important question is should it be done at all, and 14 there are very strong arguments why it should not be done.
15 If you can overcome those arguments, then the question is 16 how do I combine them, and there are various ways to do 17 that. I can combine them arithmetical 11y and I can combine 18 them geometrically.
19 CHAIRMAN SIESS: What is the alternative to 20 combining?
21 MR. REITER: Well, the alternative to combining is 22 having criteria such that, and I will show you right away 23 that I used both Livermore and EPRI results and depending on 24 where they fall in the Livermore resulto and the EPRI
,r.~ 25 results I do to one decision, and depending on another one I b,
Heritage Repcrting Corporation (202) 628-4888
- _ - _ - _ _ - _ i
J#%
a 210 1 go to another decision.
2 CHAIRMAN SIESS: You mean you take the worse case?
3 MR. REITER: No. Really the emphasis here is we 4 are not worse case oriented. If that is the case, if 5 Livermore is higher than the others, and since the mean is 6 the highest, we could just say take the mean of Livermore 7 and you have the worse case. I think that the idea of where 8 we are going and somebody brought up the word before is 9 robustness.
10 What we are looking for are what are the robust 11 conclusions that we get from both sets of hazard studies.
12 In other words, those conclusions that are consistent with r
(,) 13 both sets we are the most comfortable with. Let's try and 14 build around that and recognize the fact that it is not only 15 answers a lot of uncertainty that is out there and let's try 16 and proceed. And the strategies that we are doing are 17 trying to go ahead and do that.
18 Now I want to show you some comparisons that you 19 might find interesting between Livermore and EPRI. And 20 these comparisons were done, visually as we do not have the 21 printed results yet so there may be errors, were done in 22 57 plants. These sites are not included because they have 23 not submitted their calculations to the seismicity owners 24 group.
,, 25 Two of them, St. Lucie and Turkey Point, had done Heritage Reporting Corporation (202) 628-4888 L_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _ _
6 L ()-
211 e n. -
~
3
-:L calculations but they preferred to do the calculations l
2 themselves. Grand Gulf has now asked EPRI to do the I 3 calculations. So we will have these. But Shoreham, 4 . Cooper, Cook, Fort Calhoun, Palisades, Duane Arnold,
-5 Callaway, and Big Rock Point do not have EPRI calculations.
6 So what we have done is limited our results to the 57 sites 7 for which we have both sets of calculations.
8- CHAIRMAN SIESS: Now what are the two sets of 9 calculations?
10 MR. REITER: Excuse me.
11 CHAIRMAN SIESS: What are the calculations that 12 they are submitting?
k_s) 13 NR. REITER: These are hazard calculations for a 14 particular site.
15 CHAIRMAN SIESS: How does that relate to the 16 multi-volume report on seismic hazard?
17 MR. REITER: Livermore has everything, but EPRI 18 does not.
19 CHAIRMAN SIESS: Oh, okay.
20 MR. REITER: So I am saying here are the plants 21 that are not included in the EPRI study. And what I want to 22- do is compare results available from both Livermore and EPRI 23 to have a uniform set.
24 CHAIRNAN SIESS: Now the utilities were doing the 25 EPRI studies?
Heritage Reporting Corporation (202) 628-4800
- _ _ _ - _ _ _ - _ _ _ __ i
m , .. . . . . . .
~
,a .-
e a 212 i ? '
HR. REITER: Well, the utilities remember had
') 2- -formed-a. seismicity owners group, and they. asked EPRI to do 3; the study, and as a result they also produced calculations.
4 Now these plants were not in the seismicity owners group.
p 5- Grand Gulf was and then recalled but now they.are doing them 6 again. St. Lucio and Turkey point were involved, but they 7 want to do their own calculations. So it is a little 1
8 different.
9 .And what I am going to be showing you is-10 comparisons of Livermore and EPRI at the chance of exceeding 11 0.3g or the chance of exceeding the NUREG 0098 spectrum 12 associated with that. So it is a special ground level. And 13 the reason that I am interested in that is because again 14 .
0.3g seems to be the review level or screening level that we 15- are most interested in.
16 CHAIRMAN SIESS: Now you gave us'two likelies, 17 likely of exceeding 0.3 9.
18 HR. REITER: Yes. There is one slide that I am 19 going to show you. One is a PGA. I am showing you most of 20 the stuff in PGA, and one slide that shows you the 21 difference between PGA and some other measure, this weighted 22 frequency.
23 CHAIRMAN SIESS: Okay.
24 MR. REITER: So I would just like to show you to 25 give you an idea of where we are going. Here is the first Heritage Reporting Corporation (202) 628-4888 ]
l
{
i
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ )
l -f s l\ )
213 1 plct and let me just show you what we are looking at. What 2 you have here are probabilities of exceeding 0.3g. And 3 these are EPRI and Livermore. In other words, this is 4 10 to the minus 2 per year, and this is 10 to the minus 7 5 for Livermore, 10 to the minus 2 per year, and 10 to the 6 minus 7 for Livermore.
7 If all 57 plants were the same, they would all 8 fall along the line which says likelihood of exceeding is 9 the same for both. Each one of these squares represent a 10 half an order of magnitude. So for instance this is 1.1 10 to the minus 4, and this is three times 10 to the 12 minus 4, and that is 10 to the minus 3, three times s
f"
( 13 10 to the minus 3, and 10 to the minus 2.
14 So if you look at this, you see the 57 plants and 15 this is what is most encouraging, that although only say six 16 of them fall exactly on this line that most of the plants 17 fall pretty close to this line and form a trend. In other 18 words, there seems to be very good consistency between the 19 Livermore and EPRI results both in the relative distribution 20 and the absolute values.
21 Now the problem breaks down at the very low 22 probabilities. In other words, there are some plants which 23 Livermore says may fall between three times 10 to the minus 24 6 and 10 to the minus 5 while EPRI says between 10 to the
,-s 25 minus 7 and three times 10 to the minus 7.
Heritage Reporting Corporation (202) 628-4888
. (p
\_/
214 1 And the reasons these nre at very low 2 probabilities is because the methodology that appears with 3 this very low seismicity is very sensitive to what is called 4 lower bound magnitude cutoff and the way that you look at 5 the coismicity.
6 CHAIRMAN SIESS: What are the numbers?
7 MR. REITER: These numbers are the number of 8 plants.
l 9 CHAIRMAN SIESS: Number of plants at those points?
10 MR. REITER: That fall into these little bins.
11 And this again is extremely encouraging. Past plots would 12 have shown this thing much over and a lot less agreement.
,~
(,;/ 13 So again one could construct a criteria and we 14 have constructed such mode's criteria which would say that if 15 plants fall within this particular area that I will give it 16 a 0.3g, a reduced program or a 0.5. There are various ways l
17 to do that.
18 The next plot which is not as encouraging is the 19 mean. Remember that that was the median. And here we see 20 that indeed the plants seem to fall within a group where 21 there seems to be relatively little, which is offset from 22 the Livermore versus EPRI line and also is not systematic.
23 There seems to be a blob down here or less systematic.
24 The problem that we face with the means with even 25 0.39 is there seems to be disagreement in these particular Heritage Reporting Corporation (202) 628-4888
_ (m, t) 215 1 means. For instance you can look and say that there is one 2 plant over here to give you an example which for Livermore 3 is somewhere between three times 10 to the minus 4 and 4 10 to the minus 3, while for EPRI it is between three times 5 10 to the minus 7 and 10 to the minus 6. That is like the 6 worse case.
7 This is the problem that we have with the mean.
8 It is very attractive from the point if you are a PRA, but 9 from t'ne point of view of agreement because of what 10 dominates this mean it is not a very attractive choice.
11 On the other hand if we look at the 85th 12 percentile, we see sort of an intermediate picture. And
(~3 13 again maybe our statistician will challenge me on this, but t>
14 to me this seems to be somewhere in between. It seems to be 15 more systematic in that the trend is parallel to the line 16 but it is more offset.
17 CHAIRMAN SIESS: Is EPRI using a cutoff?
18 MR. REITER: No, there is no cutoff. But remember 19 that all I am looking at is 0.3g in this case, so a cutoff 20 would not affect it. This is the likelihood of exceeding 21 0.3g. One at the median level, at the mean level, or the 22 85th percentile level.
4 23 CHAIRMAN SIESS: So it does not make any 24 difference?
l
,.. 25 MR. REITER: The cutoff would not affect it unless 1 Heritage Reporting Corporation (202) 628-4888 ,
1 l
L 216 1 you cut off low at 0.3g, and I think that that would i 2 probably be hard to justify.
3 I think that this next slide might also be 4 interesting, and just bear with me a minute or two while I 5 explain it. This looks at-the consistency between the 6 results. And what I have done is looked at the individual 1
7 plants, looking at the individual plants, and have assigned l-8 them a number. I have 57 plants, and the number that I have l
9 assigned them is their ranking within EPRI with the chance 10 of exceeding the 0.3g median.
11 Plant number one is an unnamed plant somewhere I ,
12 which according to EPRI had the highest chance of the mean
( 13 exceeding 0.3g, and Plant 57 had the lowest chance. And 14 what I have done is looked at the first seven plants. So 15 what I am looking for is consistency between these numbers.
16 And I looked at the Livermore and I see that the highest i
17 plant in Livermore is number 36 in EPRI. And the difference 18 between this plant and that plant is a factor of 11.5. In 19 other words, the mean likelihood of exceeding 0.3g for plant L
20 number 36 is 11.5 times higher than the mean likely of 21 exceeding plant number one.
22 So we look at it and we see some plants which are l
L 23 in the top seven, and they would be four and seven; or they 1
24 would be four, seven, and one. But there are other plants 25 which are much lower down on the EPRI scale.
Heritage Reporting Corporation (202) 628-4888
J"
sa-217 1 And if you go down to the bottom, again the lowest 2 plant here, there is some agreement and some disagreement.
3 To give you an idea of the chance of exceeding the SSE for
'4 the lowest plant in the Livermore case is a factor of 108 5 greater than the lowest plant in the EPRI case.
6 When I look at the median, I get really a 7 different picture. Again I mm using the same identification 8 that I did in the mean. Plant number one is the same plant 9 that came out first in the EPRI mean listing. And what do I 10 see when I compare the Livermore and EPRI is almost the 11 exact same plant. Plant number one is first, plant number 12 three is second, plant number four is third, and plant O)
\_ 13 number seven is fourth, and number two there is a slight 14 difference over here, and then we go plant number fourteen.
15 And the difference in the medians between plant 16 number one is only a factor of 1.4. Here is a factor of 17 11.5. And when we go down to the bottom again there is a 18 lot greater agreement, and the difference between the bottom l 19 number is a factor of fifteen versus a factor of 108.
l 20 And again this is the kind of information that has l
21 been the most encouraging for us. If I look at the 22 85th percentile without going through all kinds of boring 23 details, it is somewhere in between. There is a little more 1
l 24 consistency and the differences are less.
,_ 25 I think that comparisons like these are going to V
Heritage Reporting Corporation (202) 628-4888 i
L__________________
(~ I 218 1 get in every which way possible from Livermore are going to 2 be very helpful for us in helping us choose what are the 3 appropriate plants. We are also looking not only to ranking 4 but what we are really looking.for are how the plants group.
5 What are the significant differences between the plante, is 6 there a top group or is there a lower group.
7 One final plot which I wanted to show you was how 8 would the use of a frequency or weighted frequency rather 9 than a PGA affect the plant. And here I do not have these 10 results for EPRI yet. We are developing it now and reading 11 the EPRI tape.
12 So all I have done is compared Livenmore and the 13 chance of exceeding the PGA versus the chance of exceeding 14 the spectrum level associated with that PGA. Now what I 15 find is a systematic down shift. In other words, there is a 16 lower likelihood of exceeding the weighted frequency than 17 there is of a PGA. So that tells me that if somehow I was 18 looking at the 0.3g 0098 spectrum if I use the PGA that'I 19 would overestimate what the hazard is.
20 Now on the other hand, again one of the other l
j 21 proposals that we had was to use these weighted frequencies 22 and combine it with the EPRI results.
23 CHAIRMAN SIESS: I do not see the downshift that 24 you are talking about.
- 25 HR. REITER
- Well, for example --
.(
i Heritage Reperting Corporation L (202) 628-4888 L
L _ _ _ _ _ _ - _ _ _ - _ _ _ _ _ - -
, j g ,
Q- -r ,
, , q 3 ,
sp%W.51;p&
v < m , 3,3 , b >J '
~i < >
l: { ' I ' ~
( ( l) . . .. n}W'.3']lI7',
i + -e ,
. -e , 0 , _ ,,i:t < >
' ' .,,' h s li' t ,;,: , s ~t u
- ' ~
uj. ,' ',
219 i
g ,.
L
,l' t ,
')',
p..,> 'l .;
CIGIRMAN SIESS: ~ Marking that 45 degree'line.;.
se .,, ,
l2 S [ >( i
, The ' ;45 degree' line .' . There areltwo
.t
"* ,c..
e 2 MR. REITER W '
.' . ,4 .
EZ, :3 plants that. fall.between 10 to the minus>4-and three. times 4 y, . , s -
. . , 9 g- l' 10.to the.minus 4. i If 'I: use PGA,L there -skre ao plants 'iq? I Y, b. as
.t , . .
5' Q, ' '
-use,the weighted frequency. .T.f I.go dotin to between thiree
..e ,>
j 6- biruss 10 to the minus 5 andL10 td the minus 4, there. a.re ' '
u 4 o ,
j,
. ~7 /tWnty plants. '
[ >
- 8. .
CHAIRMAN SIESS:i Oh,-okay. There is no . ,
s
.; 9 s$gnificance to where the twehty is?
l l
l < - 10' MR. REITER:--They.are all within this box,conly i 11 thi? represents the number of! plants whose PGA falls in that m n
'l~2 box.and this represents the number of plants whose weighted 13 frequency falls in that box. .
t 14 CHAIRMAN SIESS: I am lost, but I do not think (
8 l
l 15 that it'is worth it. l 16 MR. REITER: This box represents the ch:ance of 17' exceedance that is between three., times 10 to-the minus 5 and .]
18' 10 to the minus-4, okay?
19 CHA.IRMAN SIESS: ' Yes 20 MR. REITER If I look at PGA results, there are 21 twenty plants in Livermore that fall within that box. If I '
i 22 look at the weighted frequency, 2.5 hertz, 10 hertz, o~ ,
23 5 hertz, and a little bit of PGA, there are only three 24 plants that fall in that box.
25 CHAIRMAN SIESS: Okay. '
O l 1
Heritage Reporting Corporation 1 (202) 628-4888 1 i
'/ ,
I
\'
n j
' ' ' a. --_-___a- - . - - - . _ , - _ _ _ _ - - _ - - - . -- --w-_-- - - - - . - . - - _ _ -- - - . - - - - - - _ _ .
c; ; -- 7, --- ; - y . , ,
-h
, <te '4 's, p b - . .
M;.--
- yp -
.. 7 lt ; .,
k.y' i '
, ~i h~ y y f llgA{ [.
~
t, ?
..J
.w, '
220
[4.[ I!h 1- MR. REITER:t And what you see is that the; plants- l N,Uh n cm
.;ry c. >-
(
n ' ,7 c2 ;. have]been shifted downward in the weighted frequency, whichl J
O, , F - -tells me that if I use PGA alone that it is going to give me;
- p. . ,
4 a conservative or maybe'is going to overestimate the hazard
[.," :5 if 1' leak;at the chance'of exceeding the 0098 spectrum.=
1 ,
I ', % 6
~
SoLone of.the other strategies that we are looking iJ e 'a.
j" '
~
-1 . at is taking these kinds of numbers, combining them with
~
3 . EPRI to.get.somel cort of an average and using that. In 9L - other words,,. you can look at the two extremes. One look:s at-n ' b' L 10 - the snean and 'uses EPRI separate and Livermore separate, and
>3;
- 11' the other extreme looks at the median and looks at the
[i,8 g 12 frequencies and combines EPRI and Livermore.
- 13 And we are not at this point saying that those are
$' 14 the only two. methods but we can look at various methods and c 15 look.at those results, and using those results and again
- 16 this very' generalized relationship we have with core damage, 11 and try to arrive at what we consider is a good way to bin J18 these plants.
,fi' 19 CHAIT24AN SIESS: Is there a good reason in your f
f 20 mind for believing that the two methods are equally valid, 21 that you should not just use one and throw the other one
22 out?
23 MR. REITER: I do not think that you can do that.
24 CHAIRimN SIESS: You think that both have been 25 done by good people in good faith?
H3ritage Reporting Corporation (202) 628-4888 u
mW ma_am___-m.____m _. _Aik _ _ _ _ ____._*.-__._J 1
7y
't (1 221 i-l' MR. REITER: They bcth have been done by good Jn 2 people. They both have their shortcomings and they both 3 have their advantages.
4 CEAIRMAN SIESS: So you are looking for some way ;
i 5 to bring thera together?
6 MR. REITER: I think tha.t there is an brgument.
7 On the one hand people say, and say back when we convened 8 the panel with Bob Budnitz, with some statisticians, with 9 some decision people, with some seismologists and carthquake
, 10 engineers, and they said for the purpose that we are looking 11 at that we should keep them eparate. Sc 1 think that there 12 is a strong argument that you do not want to combine them.
/
$c% ) . 13 On the other hand if we limit the scope of what we
]
14 are doing,.you know we may want to combine them, that we may i
15 want to see whare that result leads to. So we are not 16 ruling tnat out. What I am saying is that there are strong (
17 arguments on either side.
18 CHAIRMAN SIESS: If our only purpose were to look 19 for vulnerabilities, is there reason to think that one is 20 better than the other?
21 MR. REITER: Well, again the only trap that you 22 want to avoid. falling into is because if I want to got the 23 most " vulnerabilities" I should therefore use the highest
~
,i 24 harard curve out there. And the highest hazard curve out j
{
,, 25 there is the Livermore mean. I do not believe that that is '
/%
V Heritage Reporting Corporation (202) 628-4888 i l
L
.r. ,
- 3. .
M .
- ,s f% '. .V l, t L_ I ( );
u 3.
o L i
.the way to go,.
n' ' . If '
"c lW 2 CRAIRMAN SIESS: Would'they uncover different Ji i , ! '
9 , ,j j j 3 types of. vulnerabilities?
l; if
'ju , ,
-4 MR. REITER: No, I think that was pointed'out, if 1 : t ;
- 5 CHAIRMAN SIESS: That is what I thought too. l A
. i .
s i i6' MR. REITER: The studies that we have looked for I. 4 7 Surry and Peachbotton, they do not.
?.. 8 CHAIRMAN'S7ESS: So from what I would expect from k I 9 .a few years of looking at these things the NRC sponsored 10 . method tends to give. higher earthquakes than.the industry
.0 t
, 11 .-sponsored method, and you think that both -)obs were done 12 ' pretty carefully and was actually an attempt'to be closer to I. ,
' 13 real~ life and to conservatism.
n -
i o .- ].
f,- 14. .MR. REITER: One way is to say let's try to
- 11 5 combine thean. Again I want to emphasize that I
- think there
- ' 16 ' is a ,1.ot of agreement and people tond to focus on the areas' l_ _
Q[y 17' of dit, agreement, namely tha mean hazard driving the mean
, s: ,
d.(
f, 4
' 18 core melt. And quite honestly the massage that'we are l; 19 getting is that there is uncertainty out there and we have
} ;, ,
?
20 ,to' learn to live with that uncertainty'and-learn to make 21 decisions in the face of that uncertainty which'is not
. 922' ~necessarily driven by enveloping.
23 . MR. MURPHY: Just one point. The NRC commissioned
-t 24 the National Research Council to carry out an in depth study
.25 cf the use of probabilistic hazatrd assessment within the
'O .
Heritage Reporting Corporation (202) 628-4888 i 7 } 'I w ,
__i -
N) 223 1 context of regulation. And that committee specifically 2 looked at these two methodologies and came back with a set 3 of criteria of what they called acceptable ways of carrying 4 out seismic hazard assessments. And both of these
'5 methodologies fit well into those categories.
6 CHAIRMAN SIESS: Is there a report on that?
.7 MR. MURPHY: Yes, there is.
8 CHAIRMAN SIESS: Recent?
9 MR. MURPHY: Six months.
10 MR. REITER: In fact it says that the EPRI and 11 Livermore were the mudels for the ways it carried out.
12 C;1AIPMAN SIESS: Could you get a copy of that?
/'~T
()m 13 MR. MURPHY: We can get-you a copy I think.
14 CHAIRMAN SIESS: Who did they have on that 15 committee that was not already involved in this thing?
16 MR. MURPEY: They were able to find some 17 individuals.
18 MR. REITER: The chairman was Professor Kaiki, 19 probably one of the world's most renowned seismologist who 20 has not been involved in hazard calculations.
21 CHAIRMAN SIESS: There must be something wrong 22 with him.
23 MR. REITER: Remember that you are looking st 24 estimating hazard in the eastern United States. And many 4 25 seismologists are really more concerned with what is Heritage Reporting Corporation (202) 628-4888
<k:
1 1
lt 224 1 happening with earthquakes in California or Japan. So he 2 spent his whole life in developing insights to what is 3 happening in the highly seismically active areas.
4 OHAIRMAN SIESS: Okay.
5 MR. REITER: And there are other people involved 6 also. But it included people like Kennedy, and it included 7 people like McGuire.
8 CHAIRMAN SIESS: So there were people on it who 9 knew this area?
10 MR. MURPHY: Yes, 11 CI'7 'RMAN SIESS : But it was tempered by 12 independence.
13 MR. MURPHY: Yes, sir.
14 CHAIRMAN SIESS: I would not have thought that 15 they existed.
l- 16 MR. REITER: There were both, people who were not i
17 and people who were independent. There were people in GS 18 who had z.ot been involved. Now the chief of the office of l
l 19 eartnquake studies at USGS. Rob Weston is a member, and 20 some excellent people.
21 CHAIRMAN SIESS: Get those names. We may need 22 them for consultants.
23 MR. MURPHY: We are get you a copy of the report, L ,
24 and they are all included in there.
l 25 MR. REITER: Professor Rocky is our consultant on Heritage Reporting Corporation (202) 628-4888 L
.n l
225 1 .Diablo Canyon atd ground motion. Rob Weston was our 4 2 consultant on the earthquakes.
3 Let me just sort of wrap up the seismic. So a 4 basic consideration that we have is we do not think that 5 there is a simple theoretical'colution. We have to look at i
6 the conclusions based on different criteria and different 7 parameters. As somebody said,.I think that it was Joe I
8 Murphy, we are slicing that distribution in various ways and 9 looking at it. There is no simple way to characterize it.
10 We have actually asked Livermore in the context of
! 11 the Charleston program to do all kinds of comparisons I 12 between the EPRI and Livermore studies that utilize both 13 EPRI and Livermore, and we are utilizing those calculations 14 based on different hazard measures, different frequencies, 15 and different statistics, and we will be getting those very 16 shortly, and using that as a firm base rather than some of 17 the visual kinds of readings that I gave you beforehand.
18 For the purpose of binning relationships, and this b
l 19 is an important assumption, the relation between hazard and i
l 20 core damage need only to be very general. We are not basing 21 it on a very rigorous connection between the two, and that l-22 is a very important assumption to us.
23 We want to utilize the kind of work that NUMARC 24 and EPRI is doing. NUMARC has started out on a different 25 approach. They are taking the hazard curves and usrag some Heritage Reporting Corporation
!. (202) 628-4888 L-_________-_--_-_
226 R 1- generic plant fragilities and coming up with some generic l
! 2 core damage and binning in that.way. And we want to look at 3 those results when they come in and try to utilize that 4 information also.
5 CHAIRMAN SIESS: What are-they doing that for?
6 MR. REITER: They are doing that for us, 7 essentially for them and they are going to submit to us.
8 CHAIRMAN SIESS: If it is generic then it cannot 9 be related to IPEs.
10 MR ~. REITER: No. In other words, we are 11 essentially binning on hazard. They are coming up with a 12 way that they want to try to use the hazard curve but use 13 some generalized fragility and come up with some answer that 14 way. And hopefully we will be getting that in a couple of 15 weeks. I see that Jim Whiteraft from NUMARC is saying yes, 16 so we can write that down in blood.
17 We want to utilize as necessary deterministic 18 information. And we always reserve the right for ourselves 19 to look at that and say well, there may be a particular 20 plant which may have something which may not be reflected in 1 21 the hazard curve which we may want to override. I 22 And finally we wanted to find plants which fall 23 into 0.5g and 0.3g, reduced program bins, based on 24 consistent patterns of different approaches. It is not
,, 25 going to be 100 percent consistent, bu+ 7e are looking for U
Heritage Reporting Corporation (202) 628-4888 i
fs V
227 1 consistency.
2 I have not mentioned the western plants. The 3 western plants, remember that we have two plants out there 4 in coastal California, Diablo and San Onofre, which the 5 margin is not going to work for them, and they are going to 6 have to use PRA. The other plants out there will probably 7 be 0.5g or 0.3g. But unfortunately San Onofre only had a 8 partial one for Unit 1.
9 CHAIRMAN SIESS: Before you take that slide off, 10 Item 5, suppose you never heard of PRA?
11 MR. REITER: Excuse me.
12 CHAIRMAN SIESS: Suppose that you had never heard 13 of PRA. Go back about twenty years, you probably had not.
14 How good of a job could you do of finding 15 vulnerabilities strictly on a deterministic basis?
16 MR. REITER: That is not a hazard question.
17 CHAIRMAN SIESS: No, but it is an open question.
18 MR. REITER: There is a question about that, yes.
19 CHAIRMAN SIESS: And I am going to get a PRA man 20 to answer it.
21 MR. CHOKSHI: I think that it is basically a 22 deterministic type of approach. It definea the plant 23 operational sequences, and the way that you get high 24 confidence and low probabilities is a deterministic type of 25 matter and not really probabilistic. There are no l Heritage Reporting Corporation l (202) 628-4888 l
Lm__.-_ _ _ _ _ _ _ _ .
l "v.,f g , . , , ./
, w.
Q
?i +
gyp _ ' 2' g-,- ~ 251--
q.'<>!'i (1 s.. / . But unless'they thou'ht g about a flood and some gy iways'of protecting against.-it, they weren't going to b'e that 12
=3. well off. <
L p '. i g gcib-l;e-
.:r; l . 4 MR. WESTCOTT: Okay. ' I' m' not sure what.'. the -
g ._
l e. ..
P 5 probability there was-for' Quad. Cities lfor the Mississippi-
+j , 6- flood.
- ;? c_ ,
7' It-could have been significantly, you know, or-8 fairly high.. .
- 9. CHAIRMAN SIESS: Well, this was a long' time-ago..
f:'y >
Q. , ' O _10 MR. WESTCOTT: Right.'
11 CHAIRMAN SIESS:- But they did have. warning.:
- . 12 MR. WESTCOTT: Right. . Well, that's the case r 15 certainly'for most coastal plantt:, where you are talking 14 about hurricanes.
<,..y . y - >
15 CHAIRMAN SIESS: .But unless they've'got a plan, L 16- sarning doesn't,do them any good.
1 17 MR. WESTCOTT: That's right. And you also have,to 18 have~ time. 'If you' re talking about a dam failure or a very l 19 small basin, you don't have any warning time, in many cases.
20 CHAIRMAN-SIESS: The damn failures got to be 21 pretty low probabilities, too.
22 MR. WESTCOTT: That's true.
23 CHAIRMAN SIESS: Okay.
24 MR WYLIE: Let me ask you, does the examination 25 apply to the facilities for additional generation as part of Heritage Reporting Corporation
, (202) 628-4888
_ _ _ _ _ _ - - - - _ _ _ _ _ _ _ _ _ _ _ _ - - _ - - - J
- c f3 ,
b 1 ,
B ;%7 - ,,
, ;i ( -f' I 252 P, -1 the blackout rule?
rv, ,
-2 . MR . s JENG:. I beg your pardon?
3 MR. WYLIE: I say, does this. apply to the 4 additional onsite' generation that is added as part of a 5 blackout rule,.to meet'the blackout rule?
6 MR. BECKNER: Do you mean are those facilities 7 required to protect from the flood?
8 'MR. WYLIE: Yes.
'9 MR. BECKNER: 1 don't know.
10 MR. WYLIE: I don't'think they are.. Br.t I'm just
,11 curious.,
- 12 MR. CHEN: For station blackout, I. don't think
' ,_ 13 flood is in the picture.
14 MR. WYLIE: Ttdditional facilities you add for the 15 . station blackout rule, are they required to be protected
~ 16 from the flood?
17 MR..CEEN: From flood or from wind or from 18 extern &l event? No. They don't require anything other than 19 --
l 20 MR. WYLIE: It's not even safety-related? i 21 MR. CHEN: No. !
22 MR. WESTCOTT: Reg. Guide 159, very specifically }
~.
I 23 requires flood protection for safety-related structures and ]
24 components only, That is for the design basis flood, or 25 hurricane.
O I Heritage Reporting Corporation (202) 628-4888 i
-_ .___--____-_A
i
\,) .
, ,. 253 u
E 1 CHAIRMAN SIESS: I think Lacrosse was an other one 2 that couldn't take the flood. But their first reaction, the 3 first thing they would do would be get a crane and take the-4 insulation off the outside of the vessel, outside-the 5 containment.
6 And they thought of a lot of ways to handle these 7 low probability cases without spending a lot of money. But 8 they have to think about them a little bit.
9 But how many plants have you got left that haven't 10 been looked at for flood? Not PHP, but real flood.
11 MR. WESTCOTT: Well, I think the way we looked at 12 it was, we expect to have potential problems with the pre-
}"
(.)k 13 1973 plants that haven't gone through SEP. And I don't know 14 how many --
15 MR. JENG: About 35.
16 CHAIRMAN SIESS: 35 pre-1973 plants?
17 MR. JENG: Yes. And take off 10 SEP plants, so 18 about 25 plants.
19 CEAIRMAN SIESS: Palisades was certainly a pre-73 20 plant, the CP anyway. And we worried the heck out of that 21 plant en saiches and surges and --
22 MR. WESTCOTT: You know, sometimes just by the way 23 the plant is sited, I think you mentioned before, if a plant 24 is sited up on a slope --
25 CHAIRMAN SIESS: Palisades is on Lake Michigan, Heritage Reporting Corporation (202) 628-4888
,t.
[ <U A ,e 254 1 right down'on'he t water. You can walk out there and stick 2 your toe in it.
'3 We had experts on saiches and surges and all of 4' that stuff worrying the heck out of them. And every plant l' 5L on Lake Ontario 1[ think we went through the same cockeyed 6 thing.- So-what happened in '737 What-is the significance 4
7 of '73? What happened there?
8 MR. WESTCOTT: That was when we had the Reg. Guide 9 159. That's kind of what we decided --
10 MR. SHAO: So maybe even before '73 people were 11 already using this reg. guide.
12 CHAIRMAN SIESS: Yes. I'm sure they were.
() 13' HR. BECKNER: I think the point for all these 14 hazards is that for the SRP plants, we know what'the 15 licensing basis was. Prior to that,-there was variability 16 in these areas.
17 CHAIRMAN SIESS: Palisades was an SRP. That's my 18 point. You see, in terms of date r they may have been an SRP 19 because of the FTL, but that would have put it even later, 20 you see.
21 MR. BECKNER: Right.
22 CHAIRMAN SIESS: But I know when Palisades was 23 licensed. And I'm sure the CP was long before '73.
24 MR. SHAO: Before the SCP, standard review plan 25 came out in 1973.
k.
Heritage Reporting Corporation (202) 628-4888
n I (
255 1 CHAIRMAN SIESS: Let's face it. We do have
-2 licensed plants before we had a standard review plan.
3 MR. .H;NG: 25 plants we are saying we suspect. .It 4 doesn't mean necessarily 25 plants are in trouble. Maybe 5 some of them would be.
6 CHAIRMAN SIESS: I don't think you are going to
'7 have that many plants that have a flood problem, 8 Okay. Go ahead.
9 10
~11 12 13 14 15-16 17 18 19 20 21 22 23
-24 25
.O Heritage Reporting Corporation (202) 628-4888
___ _ __ __ -- _-- _____-_ __ - _ ____-__-_--_ _ ___ -____-_ _ _ -_- _ __ _ = _ _ _ _ -
' ,)
256 1 MR. JENG: This morning an important comment about 2 even those minor hazards like sandstone lightening, we 3 should not prior screen out. We should push the licensee to 4 justify if they think it should be screened out without any 5 further consideration. And that comment has not been 6 included in this slide. We would certainly take care of 7 your comments and include it in the slides.
8 CHAIRMAN SIESS: The licensee knows his plant 9 better than you do.
10 What about tornado missile design? That went 11 through God knows how many changes, and it probably came 12 along later than flood design. There wasn't a single plant
()
\s 13 in the SEP that had had any tornado missile design.
14 MR. JENG: Yes, but I think come in around '72-73.
15 CHAIRMAN SIESS: And that's been relaxed since.
16 DR. SHEWMON: Do you guys ever accept the AEC 17 telephone pole at 100 mile an hour or whatever the darn i
18 thing was?
19 CHAIRMAN SIESS: Well, we used to have a 20 Volkswagen.
21 MR. JENG: Also Volkswagen coming at --
22 DR. SHEWMON: I asked about the telephone pole.
I 23 MR. JENG: Telephone pole is still a part of our 24 missile sector.
,, 25 DR. SHEWMON: You probably heard about the guy who ,
1
(%.))
Heritage Reporting Corporation (202) 628-4888 i l
_____________ a
g.
i;
-r kf 257 L 1 tried to buy one of those things. I don't know whether it 2 was EPRI or somebody else. You never could find one as big 3 as the AEC swore was going to be planned, or the NRC.
4 CHAIRMAN SIESS: Okay, let's go on to floods.
5 MR. JENG: As I mentioned, our current criteria is 6 very conservative in the wind floor and the other areas.
7 Therefore, I would like to sort of summarize what our 8 current criteria in the area of wind.
9 In the high wind, we have been designing to 100 10 recurrence wind, which ranges from 130 miles per hour. One
'll hundred thirty miles per hour is normally associated with-12 the coastal hurricanes in the New England area. I remember
.( ) 13 one plant was as high as 130 miles per hour. But normally 14 100 miles per hour is the norm for design consideration.
15 In the SRP sections covering the design of high 16 winds, it's SRP 331, which is the wind loads, and 353 which
., is the barrier design, and ANCI Code A.58.1 tells one how to 18 determine the intensity of the loads from the wind speed 19 given and how to consider a configuration of the structures.
20 It's how to determine pressure and loads for structures.
21 CHAIRMAN SIESS: If a ternado wind is 360 --
22 MR. JENG: That's tornado.
23 CHAIRMAN SIESS: -- what's the significance of 24 also designing for 100? Or are you designing for different 25 allowable stresses?
O Heritage Reporting Corporation (202) 628-4888
. 3
\m/
258 1 MR. JENG: Different allowable stresses.
2 CHAIRMAN SIESS: Different combinations of loads.
3 Which governs?
4 MR. SHAO: The low direction may be different.
5 CHAIRMAN SIESS: Which governs?
6 MR. SHAO: The loading condition is.
7 CHAIRMAN SIESS: It's like an SSE and an OBE.
8 MR. SHAO: Load combinations.
9 CHAIRMAN SIESS: But which governs?
10 MR. JENG: The tornado governs.
11 CHAIRMAN SIESS: Okay. So the combination with 12 tornado in it always governs?
y~.
( 13 MR. JENG: Yes.
14 CHAIRMAN SIESS: I would think so.
15 MR. JENG: Yes.
16 CHAIRMAN SIESS: As the pressure goes up, it's the 17 square root of the --
18 MR. JENG: Of the velocity.
19 CHAIRNAN SIESS: Of the velocity?
20 MR. JENG: Yes.
21 In the tornado design areas, I think it has been 22 very conservative, and the 360 mile per hour tornado has 23 been believed by the staff to be tied to the 10 minus seven 24 ball park number per year, and this number has been quoted 25 in the past evaluations quite often.
('_3
(/
Heritage Reporting Corporation (202) 628-4888
n.
259
.1 Again, in the SRP section covering tornado designs 2 the 3.3.2, and 3.5.1.4. 3.3.2 is the tornado loading.
3 CHAIRMAN SIESS: Why don't you skip that part 4 because that's not very helpful to most of us.- The SRP 5 chapters aren't of much interest to us.
6 MR. JENG: Okay.
7 CHAIRMAN SIESS: Just the loading is important, I 8 think.
9 MR. JENG: Okay.
10 CHAIRMAN SIESS: And some of us don't even know 11 what the SRP is.
12 MR. JENG: Let me go to the current criteria for
(~T
%,/ 13 flood design basis. There are two options. Option 1 is 14 based on the maintenance of safe shutdown capabilities for 15 all the sites, without one in time.
16 Now, if you are able to show sites where either 17 one in time can be provided, then that's different 18 equipment. Given that one in time variability, we are only 19 asking those items needed for safe shutdown to be designed 20 to be much more flood, but those items needed for safe 21 shutdown will be designed to the standard project fact, and 22 those items needed for maintenance of the safe shutdown 23 would have to be designed to the design basis flood.
24 So there is a variation in cases where one times 25 the variable, but either one of the option is acceptable.
Heritage Reporting Corporation (202) 628-4888 i
73-
%.]I 260 1 That's the current approach. And the basis criteria is l-L 2 1.59, which was again mentioned incorporated in 1973-74 time 3 frame.
4 CHAIRMAN SIESS: I'm confused a little bit.
5 You've got the reference to the probable max.fmum 6 flood, which I know what it is. And down in the bottom 7 you've got the standard project flood, and I know what that 8 is. . And then you've got a design basis flood.
9 MR. WESTCOTT: Yes. The design basis flood is E10 really the probable maximum flood with wind run up. It was 11 with wind rave run up.
12 CHAIRMAN SIESS: Okay. That's the distinction you 13 make in that.
14 MR. WESTCOTT: Yes, that's right. It's a NRC term 15 as opposed --
16 CHAIRMAN SIESS: The PMF just being a stage, and 17 the design basis being that plus all the other things 18 together.
19 MR. WESTCOTT: That's correct. It was a 20 distinction we made to separate our design from the Corps of 21 Engineers PMF.
22 CHAIRMAN SIESS: Thank you.
23 MR. JENG: In the accident evaluation area for 24 others like missile explosions, aircraft impact, we have the 25 following criteria.
O Heritage Reporting Corporation (202) 628-4888
.t 3
- LJ 261 1 .Among the initiators are the hazardous material or 2 activities in the vicinity of plant, and the result of that 3 initiator is the toxic cases releases, explosions, fires, 4 missiles. And the consequences, of course, releases of 5 significant quantities of radioactive materials.
6 And the criteria being used is the following.
7 Basically, they are based on the 10 CFR 100. That means the 8 two hour' exposure to single individual at the exclusion site 9 boundary should not exceed, or exceeding 25 REM whole body 10 or 300 REM thyroid.
11 There is a minor subtlety in the way we consider 12 acceptability. If the accident frequencies can be shown to I 13 be less than 10 minus seven based on the data alone, then 14 that will be acceptable. However, you want to make some 15 argument that although we cannot prove to you, but the 16 indication or our belief that the risk will be much smaller 17 than 10 minus six, then we will accept 10 minus six as 18 references.
19 CHAIRMAN SIESS: Now I'm confused, because that's 20 labeled " screening criteria", but what you' re putting up 21 there are design criteria under the existing standard of 22 EPRI, aren't they?
23 MR. JENG: Yes, right.
24 CHAIRMAN SIESS: But why is it labeled " screening 25 criteria", because that's not going to be your screening Heritage Reporting Corporation (202) 628-4888
-,3----_--------
?, M .
262 1 1
1 criteria for --
2 MR. JENG: We are proposing this to be the screen 3 ' criteria for --
4 CHAIRMAN SIESS: Well, I can understand the two 5 ' accident frequency estimates, but the two hour exposure to f an individual at the site boundary, et cetera, for something 7 that's tied into the severs accident policy statement sounds 8 strange.
9 MR. JENG: Well, our belief is based on the 10 frequency of such occurrence is very, very small, and 11 certainly in the middle of the ---
12 CHAIRMAN SIESS: But you see, what you've almost 13 got there is the proposed EPRI requirement for a definition 14- of a large release: 25 REM at the site boundary.
15 And it's hard for anybody to look at that without 16 relating it to the safety goal. And I don't think that was 17 the intention.
18 MR. JENG: This is not intended to be tied to 19 safety goals.
20 CHAIRMAN SIESS: It says if you can find an 21 initiator that doesn't -- that causes an accident, even a 22 severe accident that doesn't exceed 25 REM at the site
-23 boundary, you can ignore it, right?
24 MR. JENG: I'm talking about the current criteria 25 for design basis.
Heritage Reporting Corporation (202) 628-4888
'4
- ,.- g h .
' "' Q).
L 263 t'
1 CHAIRMAN SIESS: I know you are talking about
, 2 : current criteria, but you have got a slide up there labeled i 3: " screening criteria".
4 MR. JENG: Right.
l 5 We believe from the domain of the severe accident 6 consideration, if you are to be able to convince people'that 7 the hazards is 10 minus seven, I think it's consistent with
- 8. our quality goals that should be --
9 CHAIRMAN SIESS: But that's not what it says.
10 MR. JENG: You mean this one?
11 MR. BECKNER: It's not our intent to use that 12 first one.
(G,/! 13 CHAIRMAN SIESS: That's the one I'm looking at.
14' That says you're screening on the basis of whether it's a 15 large release.
-16 MR. JENG: That's not intended.
17 CHAIRMAN SIESS: But that's what it says.
18 MR. JENG: Yes, maybe we didn't make it clear. I 19 CHAIRMAN SIESS: You had better straighten it out, 20 because these are the design criteria.
4 21 MR. JENG: The objective of this slide is try to 22 justify why we feel 10 minus seven, 10 minus six type number 23 is good for screening out for IPEEE purposes.
24 CHAIRMAN SIESS: It's that second bullet -- that
- 25. first bullet up there that bothers me. Get your act Heritage Reporting Corporation (202) 628-4388 l 1
i
m '
)
R 1
. ,3 O
E Uh 3 (~N !
3,/
264 -
l1' straight on that. J 2 MR. BECKNER: That should not be on the slide.
3 It's not our intent to use that.
l4 CHAIRMAN SIESS: I think it's an existing design 5 criteria which we would barely find acceptable -- the safety 6 goal would say that's acceptable at 10 to the minus six 7 level.
8 MR. BECKNER: It's on the slide, I think, because 9 it's part of the current SRP, but it's not our intent to use 10 that.
11 CHAIRMAN SIESS: Okay.
12 MR. JENG: We are trying to provide basis for why lO
(_/-- 13 saying if a plan has been shown to meet current criteria in
< -14 the area of an accident, then we consider that to be 15 adequate basis of screening out, and this is nothing but 16 just to show the current criteria is in the order of this 17 number.
18 And if you were to meet this one, this would form 19 the basis for being screened out.
20 Earlier we mentioned if you were to meet the 21 current criteria, that by itself should be adequate basis to 22 be screened out. That is our premise. And I'm trying to 23 provide a basis for why such is the case.
24 CHAIRMAN SIESS: I don't believe that the 24-hour
_, 25 exposure of 25 REM is acceptable unless it's tied into a V
Heritage Reporting Corporation (202) 628-4888
2*%
() .
265 1 . frequency. The safety goal statement by the Commission --
2 MR. SHAO: Take this out of the screening
- 3. criteria.
4 HR.'JENG: Take this out.
5 MR. SHAO: Yes.
6 CHAIRMAN SIESS: At least somebody has decided 7- that's a good measure of a large release. And the 8 Commission says a large release ought to be 10 to the minus 9 six.
' 10 MR. JENG: I see, okay.
11 CHAIRMAN SIESS: So without the 10 to the minus 12 -
six on it, you are violating another Commission policy
' s_)
f3 13 statement which you are supposed to.be integrating, if you 14 will excuse the expression.
- 15. MR. JENG: Okay, we'll takeLcare of your' comments,
- 16. sir.
17- MR. BECKNER: We have no intent to use that at 18 all.
19 CHAIRMAN SIESS: I hope not.
20 MR. JENG: I would like to talk a bit about in the 21 walkdown what are we going to look at. This is basically to 22 look at the weaknesses associated with the structures and
- " 23 the plant tanks and so on. The retal sidings,. which tend to 24 'run away by some strong winds; structures with thin concrete j_
25 walls and roofs. These are primarily consideration on
(-)
L Heritage Eeporting Corporation l' (202) 628-4888
L.
266 1 missile impact. And the tall stacks, again they are very 2 vulnerable to large variation, and may fall down to hit the 3 safety components. And open tanks, again they are subject 4 to damages given the tornado, missile and so on.
5 In a flood area, of course, we have to cover all 6 the river, coastal and lake site plants, and their potential 7 flooding.
8 In the transportation areas, we look at any 9 changes in the industrial / military facilities, their 10 intensity, their potential hazards, their uses, new 11 structures coming up nearby, new transportation, new 12 potential airport coming up and so on.
,s
(.) 13 CHAIRMAN SIESS: I would think that the IPE 14 walkdown would include local flooding as well as the river, 15 coastal and lake site plants just to see if there was 16 anything odd about it.
17 MR. JENG: Okay. This is consistent with earlier 18 comments, local flooding. We'll take care of that.
19 CHAIRMAN SIESS: I mean, people ought to walk 20 around the outside of the plant, and at least consider that.
- 21. That's again a reference to that generic letter that went 22 out, but that ought to be really addressed as a part of the 23 IPE.
24 MR. JENG: Okay, we'll take care of that, i . 25 CHAIRMAN SIESS: It would make sense for them to 1 t
l \~]'
Heritage Reporting Corporation (202) 628-4888 l
j:
+.~
267 1 look at.it in connection with the IPEEE and address it 2 there, not 3 in -- well, that's what we said, I believe.
4 MR. JENG: Next one I would like to talk about 5 what are going to make use of the summary by the licensees 6 after they complete their review and report to use.
7 There are three possible outcomes. I.think this 8 question'was earlier raised by Mr. Michelson in the morning.
9 If the review --
10 CHAIRMAN SIESS: Dave, I think we covered this 11 with Larry this morning?
12 MR. JENG: Okay. So I think it's clear.
13 CHAIRMAN SIESS: You either backfit them or twist 14 their arm or not.
15 Incidentally, on the SEP plant, not one change had 16 to be invoked by the backfit. I don't know how many times 17 they threatened, but there were a lot of changes, but not 18 one of them had to be forced.
19 MR. BAGCHI: Times have changed, though.
20 CHAIRMAN SIESS: Pardon?
21 MR. GAGCHI: Times have changed, though.
22 CHAIRMAN SIESS: Well, I think the SEP people may 23 be threatened, but they never had to resort to it, and they 24 got an awful lot of changes made to this.
25 MR. JENG: Next, I would like to talk about our Heritage Reporting Corporation (202) 628-4888
n- l N; o .
fj4Q i
, l w]I I'2 268 l 1 past interaction with NUMARC, which represent industry 2 l group.-
y 3 We had three meetings and we did pass to them our
'4- basic approach, our thoughts, and we have received some-
't 5' informal ~ comments from NUMARC staff.
6 They promised to take a more thorough look at our 7- approaches, and should come back to us with more official 8 feedback maybe in two weeks. And we are continuing to work l
9- with NUMARC staff in the hope we can come up with an 10 approach'which is adequate for IPEEE purposes and also'not 11 cause undue-hardship on the part of industry unnecessarily.
12 CHAIRMAN SIESS: Let me get -- I sat in on a I
.- l13' meeting you had with NUMARC on June 21st,,but that was'with !
14 their high-level. people. It mentions three meetings here, 15 none of them were the one I attended.
i 16 NR.- JENG: I think you start only July'14th?
17 CHAIRMAN SIESS: My copy says June 21.
18 MR. JENG: June 21?
19' CHAIRMAN SIESS: That was for other.
20 MR. JENG: For others?
- 21 CHAIRMAN SIESS: High wind floods and others.
22 MR, JENG: How about August lith?
23 HR. SHAO: June something.
24 MR. JENG: June something.
25 CHAIRMAN SIESS: But I don't think that was NUMARC Hs 'itage Reporting Corporation (202) 628-4888 1
p.-f pf N( s fw w a 1 o
?'
E ,
269 I :1' - technical staff. I guess they were technical people.
... 2' Let's.see, who was there. Dave Modine from J3 NUMARC,' Ray Eng from NUMARC.
14 MR. JENG: Maybe I'm mistaken.
5- In that case, the --
6 CHAIRMAN'SIESS: Is that technical staff?
7 MR. BAGCHI: That's the technical staff that are
-8 , directing the program rather.than the committee experts.
9 CHAIRMAN SIESS: Okay, soithere were three 10- meetings after th'at'one at a working level. Okay, that's 11 . just what'I was trying to get straight.
12- 'MR. JENG: I see.
- 11 3 CHAIRMAN SIESS: Because at that meeting you just 14' told what you were doing, and you've had three neetings L 15 since the1.
- 16- MR. JENG: Okay. The last slide is talking about 17- ' the schedule and~end products. We expect to provide some-18 recommendations to the EESG in our areas to provide adequate 19 ' input for the IPEEE generic letter, and to develop some-20 information needed for the staff to review the outcome of 21 the licensee IPEEE results.,
22 The schedule we are talking about of the 9 23' procedures by end of October '89, and hopefully the final
- 24 result will be available at the end of the year.
25 This concludes my presentation.
Heritage Reporting Corporation (202) 628-4888 1
n v 270 1 CHAIRMAN SIESS: Thank you, Dave.
2 Does that conclude the staff's presentations,
~3 Larry?
4 MR. SHAO: Yes.
5 CHAIRMAN SIESS: I would like to raise an issue 6 'with the subcommittee members. The original thought that we 7 would have the staff come in tomorrow for the full committee 8 meeting and present something on this. We have an hour and 9 a half scheduled for it. And as I've sat here today over a 10 period of several hours, I have reached the conclusion that I
11 'I don't think we should recommend having them come into the 12 full committee.
1 13 First, I don't think it would be sufficient 14 benefit to the full committee at this stage of the game to 15 hear even an hour and a half of interrupted presentation on 16 this subject. And, of course, that in itself is an 17 impossibility.
18 Second, well, considering the fact that what the 19 full committee could get out of that presentation, I think 20 it's very unlikely that they could get back to the staff any 21 significant consensus report that would be useful to the 22 staff.
23 So I have suggested to the staff that we hold off 24 on the full committee until this thing is a lot further 25 along. I think the subcommittee or some portion of it, whet Heritage Reporting Corporation (202) 628-4888
r ,
s
)
271 !
I 1 her it's a combined subcommittees or maybe just the F.EP 2 subcommittee, could meet at least once more, maybe twice i 3 more with the staff and/or NUMARC, when they get down to the 4 final documents; whatever is going to be in the' generic-1 5 letter plus whatever supporting information. And some 1 6 committee will meet with them enough to understand the 7 situation fully, and then we will schedule an appropriate 8 length of~ time for the full committee, to get a full 9 committee comment on it.
10 Anybody disagree with the decision not to have the 11 report tomorrow. That will give us an hour and a half we 12 can use for something else.
) 13 MR. SHAO: The staff can come back some time at 14 your approval.
15 CHAIRMAN SIESS: But you keep us posted. We will 16 meet with you whenever we can get our teeth into something, 17 when you are far snough along. But if you are concerned 18 about a problem, don't wait until the last minute. But we 19 could probably at one meeting decide whether we've got a 20 problem. I think it's easy enough in this case to bring 21 something to the full committee that will be acceptable.
22 MR. SHAO: The staff would like to send to the 23 committee. We have very good comments today.
24 CHAIRMAN SIESS: Yes, you've heard a lot of 25 comments today.
Heritage Reporting Corporation (202) 628-4888
(Y 272 h' 'l MR. SHAO: We-have a lot of comments.
2 CHAIRMAN SIESS: You've,been receptive to some of, 3 them, and maybe some you shouldn't be. I' don't know. But I-
~4 think you are on the right track.
5 MR.-SHAO: Thank you, J 6 CHAIRMAN SIESS: But I~think you have got to bc 7 sure that track leads to finding vulnerabilities and not 8 .doing PRAs and bottom lines and so forth. I think wo all 9 should remember, whether we believe it or not, that the 10 Commission has stated in its policy statement on severe
' 11' accidents that the existing plants are safe enough.
12 But they ought to get the~USIs and GIs cleaned up.
13 That includes A-17 and A-46 and so forth in that category, 14 and because experience.has shown when we look at plants, we 15 usually find an outlier somewhere that isn't too difficult' 16 to fix. All of the plants should have an indiridual plant 17 examination to look for those vulnerabilities. And that's 18 what we are daing. We're not trying to solve all the 19 problems in the woriri. We look at the individual plant j.- 20 vulnerabilities.
21 MR. WYLIE: Will you provide e report to the full L 22 committee?
f1 23 CHAIRMAN SIESS: I will make some kind of report 24 to the committee which I suspect -- I suspect Mr. Michelson l
l 25 will want to supplement, but that's his privilege. And you Heritage Reporting Corporation j (202) 628-4888 1 _ ._ . _ - _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ ___ ___ _ -
~~ ~~~ ~
f Q ~~ %p 4: ..
~ ~]
- _ .g a.=.
n.~ j ; ,g. , , e .
$ p' . s w, ; ~
t
- r. :', . . .
(
i . i. : - '
.Ws.;>. . : ; n.
273
..j.> , i i ., .- .
1 "' can tell Ray that'I will take 15 mirutes for.a report.. Tell'
' '* 2' -
him'a. half. hour if he.wants the slot, but we won't have the
<n' e-3- ' staff there
~ n .: '
i
.;c, >4 The meeting is adjourned.. Thank you. ,
M, 'v
.5 J (Whoreupon, at 4:30 p.m., the meeting was
, (' ' . .
- L. 6' adjourned.)
4 1
.P i,t.,
- 8 g
q
.I
-,n 10
.c
.(
11, l12 s
g v- 13.
14' 15:
16.
u 17-18 19 20 21 22 23 24 25 O
Heritage Reporting Corporation (202) 628-4898
!, ^#-
.y- . -
-l' CERTIFICATE L' 2 Li 3 This is to certify that the-atte.ched proceedings before the 4 United States Nuclear Regulatory Commission in the matter
'S of:
NN NNNNRNON 1
L b 6' Name: NBCOMMITIEE ON DUTOE EXTLTNAL PIMMENA 7
'3 Docket Number:
9 Place: Be&asda, Maryland 10 Date: September 6, 1989
, 11 were held as herein appears, and that this is the original 12 transcript thereof for the file of.the United States Nuclear 13 Regulatory Commission taken stenographically by he and,
- v. 14. thereafter reduced to typewriting by me or under the 15~ direction of the court reporting company, and.that the 16 transcript is a true and accurate record of the' foregoing 17 proceedings. # l 18 /s/ ._ ~ D*%
19 -(Signature typed) : NN 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 Heritage Rerporting Corporation (202) 628-4888 r
f 1- CERTIFICATE' 2
3L This is to certify.that'the attached proceedings before the 4 United States Nuclebr Regulatory Commission in the matter of:
5 Name Advisory Cammittee on Reactor Safeguards,. Subcommittee on 6
Extreme External Phenomena 7 Docket Number:
8 Places ~Bethesda, Maryland 9 Date. September 6, 1989
.10 - were held as herein appears, and that this is the original 11 transcript thereof for the file of the United States Nuclear-12 Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewriting-by me or under the direction e-ms 14 of the court reporting company, and that the transcript is a.
V 15 true and' accurate record of the foregoing proce dings.
16 /S/[ --
m 17 (Signature typed): Catherine S. Boyd 18 Official Reporter 19 Heritage Reporting Corporation 20 21 l 22 23 24 25 O
Heritago Reporting Corporation (202) 628-4888
TREATMENT OF EXTERNAL EVENTS IN THE SEVERE ACCIDENT PROGRAM Q
LAWRENCE C. SHA0 -
SEPTEMBER 6, 1989 O _3_
l.: -
. %)..
o y..-
SEVERE ACCIDElli POLICY STATEMENTS NO DIFFERENTIATION BETWEEN INTERNAL AND EXTERNAL EVENTS O INDIVIDUAL PLANT EXAMINATION GENERIC LETTER MENTIONS EXTERNAL EVENTS O - - - -- --- - _ -
i NPC EXTERNAL EVENTS STEERING GROUP (EESG)
MISSION MAKE RECOMMENDATIONS TO NRC SENIOR MANAGEMENT REGARDING:
ROLE OF EXTERNAL EVENTS WITHIN HRC SEVERE ACCIDENTS POLICY sVIDANCES FOR IMPLEMENTATION OF EXTERNAL EVENTS INDIVIDUAL PLANT EXAMINATION (IPEEEst INTEGRATION OF ALL NRC EXTERNAL EVENT PROGRAMS NEEDED RESEARCH OR TECHNICAL ASSISTANCE
.O _3_
.. lO2 KEY EXTERNAL EVENTS EARTH 0tJAKES INTERNAL FIRES EXTERNAL FLOODS
- O-
- WIND AND TORNADOS ,,
TRANSPORTATION ACCIDENTS AND OTHERS 1^
LO -=-
9
l .
fl rs .
.O NRC EXTERNALfVEtlTS STEERit!G GROUP MEMBERSHIP CHAIRMAf': L. C. SHA0 NRR MEMBERS: G. ARLOTTO RES T. NOVAK AE0D R. W. HOUSTON RES EXECUTIVE: G. BAGCHI NRR SECRETARY:
O SEISMIC SUBCOMMITTEE:
CO-CHAIRMEN: L. REITER NRR A. F.URPHY - RES FIRE SUBCOMMITTEE:
CHAIRMAN: C. MCCRACKEN NRR HIGH WIND, FLOOD AND OTHERS SUBCOMMITTEE:
CO-CHAIRMEN: D. JENG NRR W. BECKNER RES O .
r ,
^ >
.g L.;OD . .
!F NUCLEAR INDUSTRY'S COUNTERPART ORGANIZATION 1
i NUCLEAR UTILITY MANAGEMENT AND RESOURCES COUNCll (NUMARC) 4 SEISMIC ISSUES' SEVERE ACCIDENT
- Qt WORKING GROUP
-WORKING. GROUP CHAIRMAN: 'W. LINDBLAD. CHAIRMAN: -CORDELL REED RESPONSIBLE FOR RESOLUTION OF RESPONSIBLE FOR RESOLUTION ALL SEISMIC ISSUES OF OTHER EXTERNAL EVENTS ISSUES AND ACCIDENT MANAGEMENT O. - _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ _ _
i i
l EXTERNAL EVENTS CONSIDERATIONS MANY POSSIBLE SOURCES OF HAZARDS LARGE UNCERTAINTIES ON FREQUENCY OF INITIATING EVENTS PLANTS DESIGNED TO VARIOUS CRITERIA EXTENT OF PROTECTION WITH REGARD TO EXTERNAL EVENTS BEYOND DESIGN BASES ARE UNKNOWN FOR MANY PLANTS
(])- (MAY NOT BE CONSISTENT)
PRAs INDICATE RISKS DUE TO CERTAIN EXTERNAL EVENTS CAN BE HIGH NEED APPROACHES FOR EVALUATING VARIOUS EXTERNAL EVENTS '
BEYOND DESIGN BASES EXTERNAL EVENTS PROGRAMS NEED TO BE INTEGRATED O _ _ _ _ _ _ _ ._ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ _ - _ _ _ _ _ _ - _ _ _ _ _ _ _ - - _ _ - - _ -
n
\~/
SElSMIC DESIGN VINTAGE PRE 1962 PLANTS: NO SPECIFIC CONSIDERATION FOR SEISMIC DESIGN.
1963 TO 1969 PLANTS: llSED UNIFORM BUILDING CODE OR EQUIVALENT STATIC METHOD OF ANALYSIS.
1970 TO 1975 PLANTS: USED HOUSNER'S RESPONSE SPECTRUM CURVES FOR SEISMIC INPUT. CRITERI A USED FOR EQUIPMENT QUALIFICATION VARIES 1975 TO CURRENT: R.G. 1.60 SPECTRilM OR EQUIVALENT USED FOR SEISMIC INPUT. EQUIPMENT QUAllF1 CATION IN ACCORDANCE WITH kq.
SEISMIC REEVALUATION EFFORT:
10 EARLIEST PLANTS UNDER SYSTEMATIC EVALUATION PROGRAM (BIG ROCK P0 lilt, DRESDEN I & 2, HADDAM NECK, GINNA, LA CROSSE, PAllSADES, OYSTER CREEK, SAN ONOFRE 1, .
MILLSTONE 1, YANKEE R0WE) WERE EVALUATED TO MEET A SET f
0F SEISMIC CRITERIA (RELAXED FROM SRP CRITERIA) WHICH WILL ENSURE SAFE SHUTDOWN CAPABILITY AND PRESSURE BOUNDARY INTEGRITY O 8
J 4 %
h, y:h. .
i
. SEISMIC IPEEE-AVAILABLE METHODOLOGY' SEISMIC'PRA ---
.SE1SMIC RISKS SEISMIC MARGINS -
PLANT VULNERABILITIES TO SEISMIC EVENTS 4
O _g.
t 4
.-x .. .
9 l'
L INDIVIDUAL PLANT EXAMINATION FOR FIE MfY IS INDIVIDUAL PLAfd EXAMINATICU FOR FIE E&SSARY?
INDLETRY PRA'S HAVE SHOWN FIE mlTRIBUTES TD UP TD O %60% OF CDRE ELT FECLDICY EXISTING CRITERIA D0 POT ADEQJA7B.Y ADDRESS ERTAIN FIE ISSUES O . _ - _ _ _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _ _ _ _ - _ - _ _ - _ - _ _ _ - - - - _ _ _ _ - - _ _ _ - - - _ _ _ _ - _ _
m -
]
l rrPF DAMGE FREOLENCIES FROM PRAs
[- (]
1 i T#l. CORE DANGE FREQLBICY/RX-YR FIRES, STATION FOR FIRES FOR ALL lillTIATORS FRACTION 2101 1-2 1.8 E-6 5.7 E-5 3% {
ltalAN PT. 2 1.4 E-4 4.7 E-4 30%
INDIAN PT 3 9.6 E-5 2.3 E-4 40%
BIG ROCK POINT 2.3 E-4 9.8 E-4 23%
L1ERICK 2.3 E-5 4.4 E-5 55%
2.3 E-4 SEABROOK 2.5 E-5 12%
OCONEE-3 L.0 E-5 2.5 E-4 4%
MILLSTDtE-3 4.8 E-6 7.0 E-5 7%
PEACH BOTTOM 2.0 E-5 1.0 E-4 (LLNL) 19%
2.7 E-5 (EPRI) 72%
SURRY 1.1 E-5 1.7 E-4 (LLNL) 7%
7.7 E-5 (EPRI) 15%
bG - II -
\
/
,a l
l FIRE IPEEE o FIRE PRA o ALTERNATE FIRE IPEEE METHODOLOGY (BEING DEVELOPED BY NUMARC)
O
- , = ~ = - ,
s:
F
/" 4
, . %.s l
HIGH WIND, FLOOD AND OTilER EXTERNAL HAZADPR MAJOR-SOURCES:
HIGH WIND, TORNADO, TORNADO MISSILES, FLOODS, EXPLOSIONS, T0XIC GASES, FIRES, AIRCRAFT AND SHIPPING ACCIDENTS WIND:
LOSS OF 0FFSITE POWER METAL SIDING STRUCTURES STRUCTURES WITH THIN CONCRETE WALLS AllD ROOFS TALL STACKS
(-).
OPEN TANKS FLOOD:
RIVER, C0ASTAL AND LAKE SITE PLANTS INTENSE LOCAL PRECIPITATION OTHERS: INDUSTRIAL / MILITARY FACILITIES TRANSPORTATION (AIRCRAFT, TRUCKS, TRAINS, SHIPS, PIPELINES) l O L
P SR R SSSR R S U E R R E E E R R E R NNR R R NNR O
R E
G E G T N G Y T I Y I N I L
H A I
R H L M H A E P C H S E M E R R G C K Y N O T U E A GN OG L L N
O C F S MT B N E HN E E U HCE J KK I
B U OST WREM C E A Y- H NR S RN T NNS D N E I
N D O U UH E VE G L
C E N E OSOI AL O I V AL GTJ NDGR M E NN S
I L AA E A MM S N RR I I R AA E HH T CC- -
O X OO E CC
b i SEISMIC SUBCOMMITTEE PRESENTATION ^ ,
OVERVIEW -' MURPHY PRA-ISSUES -~CH0KSHI SEISMIC MARGINS METHODOLOGY - KENNEALLY
< CONTAINMENT ISSUES - CH0KSHI' L
REVIEW' LEVEL EARTHOUAKE - REITER
- o 0 1
9 l
- o; I
o ) E lja 3
~~
> E M
I d3 h.
O <t O
S
...-....s
!a i ! ;I l 1
O t
s n
a E Y N
K A H A C U F A O I H ,
A O S ) T S I S
R T I R R E R E YN A I
E I P FE P T T T E E I P I TI D L
I I
R I
L A NC R L E B C N I
B EC O V A :
A C DA E R TW R I I C L E F.HO FGD g g E S OD R N N A T C E N W L
U AUK 3 5 L B
(
E V TO LA O0 U SU I SH V D E D M A
V E T E C TWTT AA F TI N I
O E S
I CE R
G R
A N
A L
P O
RTY NNTWW O
O C EI EE L N O A P K P .
YR BR A VV I I R S O P A A P O R UEE I
U T E U T O C E T QRR C R Q S N E EV Z E C E CHT I
P I V KO M S - SSS R I
EDIGGG I
E SA : R A UN E S A L QU I
BE RRR CL E E M I
M E P I E
P SO D
O R H
TO O L L O AAA MMM DU T
I S T H N F RT T I L N T G P A WEVCCC AN C TS E AS I
EH H I I I O
I N AI E M RE L
C LU G G S L ED MMM NP M U SSS OU E PD L T H EO E I I I I S PL I
V N O VGEE E T AT I
EB I
L BI CC E E E N ENSSS E AMM I
N L E L I E CN E TA I I S AR TSS M H FB WB ABC W G W I UE PI E EE I
R E EI O E L TM L N I I NU AL C SS T V H L VI L EC VU C E E ET DN EV A - - D R - - RS I I O o o o o O o
D E P E Z D I
M R C R D A S A N S D Z A T E A U
N N H M H A I R G S T I
E O O S S S Z T I I F N I N R S F R W L O P E A g
E F E L M O
K C P P D V M A H K E T L
EA U G L A
L R Q T N P N BT H N E M I
W OC T A S W R P E R L M D U N E P A P E N S I I
S E I D V Y FG A N E M E CN F OA T I
D E M L R NO O N D N B E T I
P A E E S
I H D O T O UO N D
U X R Q SE C E M- L E P O NR RD C OO I
T E S FN I N B S A TC I E
H U T I A T N C G O I R M L D E N O A C I
HG L RA N C A g o o o o o o
{ .
h _
O 3
Y M T
I S A L
I E
B C I N E S E E B A U L E QB Y F E I A
E S S M P N EA M O F P M O S D S S C N M S I D
T E T
L S H G :
AE G
N I NOT I
A T CB I
I R
o I
S T WN S GY P O OI E OA E N N HS E L M I
ST G O M G A DI T
, I E H A M O K S R E G AFE E
S E I MA R F
T I
P D S I
A D N R D S L UI N E E I
T VB T U R T E N HC O SG R A I
T T/ GF I
CP T D NI I O ND E I O GA - -
TC N
I E
N OO QF RM o o o o o
i; ..s
.O m
Ib 3
tn O e 9
9 o
T
(- o d
L 2 1
O
b , ..
- O PRA APPROACH FOR-SEISMIC IFE I
- o. A PRA APPROACH CONSISTENT WITH AND-TO EMPHASIZE LTHE PRIMARY OBJECTIVE OF THE IPE:
PLANT-SPECIFIC EXAMINATION TO IDENTIFY VULNERABILITIES AND UNDERSTAND INTEGRATED PLANT RESPONSE TO A SEISMIC EVENT
~
O
<O
U($) POSSIBLE RECOMMENDATIONS 0N PRA APPROACH o TWO PRA APPROACHES
- INDUSTRY AND " SIMPLIFIED" (NUREG-1150) o FOR IPEEE, ABSOLUTE PRECISE CDF ESTIMATES ACCOUNTING FOR FULL UNCERTAINTIES PROBABLY NOT NEEDED.
POINT ESTIMATE TYPE CALCULATIONS SUFFICIENT TO PROVIDE NEEDED INSIGHTS.
o FULL DETAILED UNCERTAINTY QUANTIFICATION ONLY.UNDER SPECIAL CIRCUMSTANCES WHEN RISK ESTIMATES NEEDED IN ABSOLUTE. SENSE.
o A SEISMIC PLANT WALK-DOWN IS THE MAJOR FOCUS OF THE IPEEE, SHOULD BE SIMILAR TO THE ONE DESCRIBED IN THE SEISMIC MARGIN PROCEDURES, l
0 RELAY CHATTER MAY NEED TO BE INCLUDED, 4.
{
u O a
h WHY A POINT ESTIMATE IS SUFFICIENT o ALLOWS UTILITY TO IDENTIFY POSSIBLE WEAK SEISMIC AREAS IN PLANT W/0 FOCUS SHIFTING TO BOTTOM LINE NUMBERS.
o EMPHASIZES IDENTIFICATION, RANKING, AND CATEGORIZATION OF DOMINANT SEQUENCES RATHER THAN ABSOLUTE NUMBERS, (NUREG/CR-1150 ANALYSIS USING TWO HAZARD CURVES INDICATED SAME DOMINANT SEQUENCES)
- o. NO DIRECT COMPARISON.WITH OTHER INITIATOR CDFS, o CONTROVERSY ON SELECTION OF HAZARD CURVE IS MINIMIZED.
o IDENTIFIES AND RANKS DOMINANT SEISMIC FAILURES, NON-SEISMIC FAILURES, AND HUMAN ACTIONS.
o ALLOWS UTILITY TO EVALUATE IMPORTANCE MEASURES SUCH AS RISK REDUCTION RATIOS, o ABSOLUTE NUMPERS (PARTICULARLY, MEAN) POOR INDICATOR 0F NEED FOR PLANT MODIFICATIONS, AS THESE NUMBERS ARE DOMINATED BY UNCERTAINTY IN THE HAZARD CURVES.
hC
V . 'q) -
~ Table C.11.4 Dominant Sequences - Peach Bottom .
Total Mean P = 7.66E-5 (LLNL), 3.09E-6 (EPRI)
Dominant Sequences LLNL EPRI T3 -33 = 3.69e-5 (48%) 1.61e-6 (52%)
ALOCA-30 = 1.84e-5 (24%) 6.70e-7 (21%)
RVR-1 = 8.92e-6 (11%) 3.27e-7 (11%)
5 3LOCA-70 = 6.67e-6 (9%) 1.85e-7 (6%)
RWT-1 = 2.76e-6 (4%) 1.75e-7 (6%)
52LOCA-42 = 1.20e .6 (2%) 4.90e-8 (2%)
By Sequence Type LLNL EPRI Transients (LOSP) = 3.69e-5 (48%) 1.61e-6 (52%)
<- LLCAs = 2.59e-5 (34%) 9.04e-7 (29%)
Vessal Rupture = 8.92e-6 (11%) 3.27e-7 (11%)
RWT Bldg Failure = 2.76e-6 (4%) 1.75e-7 (6%)
O C.11-19
" ~
j, l (ij USE OF EXISTING PRA OR PRA BEING CURRENTLY PERFORMED 9
e o PROBABLY, MOST OF A WELL PERFORMED PRA, EVEN IF OLD
^
SHOULD PROVIDE INFORMATION NECESSARY FOR IPEEE o SHOULD REFLECT CURRENT AS-BUILT PLANT STATUS, 1: 1 i
o ADDITIONAL EFFORT MAY BE NEEDED IN THE FOLLOWING ilREAS:
I SUPPLEMENTARY PLANT WALK-DOWN
~
0- -
RELAY CHATTER l
- ESTIMATION OF PLANT LEVEL, SEQUENCE LEVEL, AND L
l- DOMINANT COMPONENT HCLPFS.
ADDITION OF NON-SEISMIC (RANDOM) FAILURES.
l l
l hO C
y .
w .
L ,h 'POSSIBLE REPORTING REQUIREMENTS FROM A PRA SEISMIC ANALYSIS o CDF AND PLANT DAMAGE STATE ESTIMATES o HAZARD RELATED INFORMATION (SPECTRAL CURVE, HAZARD CURVE, ETC.)
0- DOMINANT ACCIDENT SEQUENCES AND SCENARIOS (USE OF A NUMERICAL CRITERION YET TO BE DECIDED) o DOMINANT COMPONENTS, HUMAN ACTIONS, NON-SEISMIC (RANDOM) FAlLURES o PLANT AND DOMINANT COMPONENT FRAGILITIES o HCLPFS FOR PLANT (WITH AND WITHOUT NON-SEISMIC FAILURES),
SEQUENCE AND COMP 0NENTS o PLANT MODIFICATIONS (HARDWARE, PROCEDURAL, ETC.) MADE, PLANNED, OR UNDER CONSIDERATION BY THE UTILITY o CONTAINMENT PERFORMANCE INSIGHTS 4
O
p-~--__
e O
J 9
A D
e in k
J
% J 2 T O u
- 'n Q 2 t 1.9 g hl o .
~ b E
I.D
'O O
L
) 1 R E 4 K R
AV 3 Y E L O 3
B 4 G C A -
O U R L NS C O E / )
AK G 1 D F A E 4 U R 0 O O Q U 6 N S H N
- WR T SH ( P OU E T R N DO E I R E O M TA ET I TH I
RU L E I RCG P H2 S BD TAN E S7 E LI (
N A
P N T R ) DR I
A N O O G AT E NO SL O
(
OF R CS H C A T H SE V WC T N E/OE NI T A R I M I G EEDD P A E
R W L I
OT N O L R TV S TI N G E OE T UO HR S
OA H
N TVE O DTSP E
I S E I TL OL C HM E V EN C - -
N HU - -
U T A I
D THA EF S CT LP GI C E S M I J S R E - -
E S E D E M OESWE L S BC I I
L SOH AR R E AAP T L N P S I I E
T A I
V O N I A
l 1llJ l
L
) m u
t n n
o
(
c S _
Y E G R _
O U L S T
D A O T N E D H F G A O I S S A D H N R E T I M PE S E E C S U
A T N M R S I MO E P Y R SD SP O S C I
S A N I E E EE 3 I
M HI N SR t E
G S TM TT V A R I NNN R A
E FA R S OX A AW U P M Y E TL O C E
L E E RPD O D A S P B N R P D K I
R R U G A O L A
I E CO MET A I Z E
T A S ST TR AW A I AA G E M O ES SGT H R GGP C
D HT PP R
TN OE N C G 3 5 O 5
F E
C N P N OO>
I D S E N SI S I
I N
M E CO N E E S V I U P I
A S E - - -
I R DM T R E E EO E O C 0 S D RC R N S ll l ll l
,P I j C
p#
SEIShSC - - -
SAFE MARGIN ESNDOWM
- E -
EARTHOUAKE -
- A - 5 -
1
+ F la) SUCCESS PATH LOGIC DMORAK A B C D E F l S EE r r r y ,r r FALLifE / ,
I , crag y
I M3TOK torOK
Is3T OK
@) EVEin TEE.
- CT
~
O 9 I
. PAAALLEL
'PA1NS F All h = W GATE (V)
O.*AN N E W ,
HEN 54AL CUTSETS:
=a PARALLEL ,
U T PATH F AILS i s
EAT Q VATAT I DMDE FAE.
(c) FAULT TREE.
ftpure 3-1. Comparisons of Success Path Logie Deepres.
Event Tree, and Fault inre 4 ,
s
3 K
R T F N P
- A L g g L 7 g 8 P C 2 3 2 H .
L A
N I
F N
G S I
R W T A E I N F MV AP L L g E P 1 - -
N R C -
L H 2 T o G T I A I
E Y
SN T .
D E A I N E DL I T
P E C g g L P
I L E g 5 5 MA S 1 1 1 M
O SI R S . .
I C E T S T E O E N K
N S I
A A Y B W E W H E I
N A C V I T T E o A M
A C
A H
R
O b
o a
M O _
2 3 tu -
O k T 2-0 O
d O
L O CONTA1NMeNT eeRe0RMANCe Issues e0R seismic iPe .
o TO WHAT EXTENT CONTAINMENT PERFORMANCE EVALUATION IS NECESSARY FOR' SEISMIC IPE?
o 110W TO EXTEND MARGIN APPROACH TO INCLUDE CONTAINMENT.
PERFORMANCE? WHAT ARE IMPORTANT PERFORMANCE ASPECTS (BYPASS, STRUCTURAL INTEGRITY, HEAT REMOVAL FUNCTIONS, AND ISOLATION FAILURES)?
~
O O
l 0- AreR0ACH UNDER INVESTIGATION o PAUL AMIC0 (LLNL) REVISITING PRAS CONSIDERED IN THE MARGIN DEVELOPMENT (NUREG/CR-4334) TO GET INSIGHTS BEYOND CORE DAMAGE WITn' TiPHASIS ON HIGH CONSEQUENCE SEQUENCES o FOR PWRS - HIGH CONSEQUENCE SEQUENCES INVOLVE GROSS STRUCTURAL FAILURES OF CONTAINMENT WITH ONE CASE OF ISOLATION FAILURE o FOR BWRS - FLIGH CONSEQUENCE SEQUENCES INVOLVE GROSS O. STRUCTURAL FAILURES-AND ATWS SEQUENCES RESULTING FROM SEISMIC FAILURE OF INTERNALS WITH RAND 0M FAILURE OF SLC o MARK I AND ICE CONDENSER PRAS NOT EXAMINED YET.
O e
io r mynn nfi m yi( teag mb ,i o r ot o( .d otao cyt . dirChr d ) . dinc l ne l oAtu s klse li p s ssli r samt sros sacp sat s t ul y a sia u stoc ae a aabatt p arrrl aaDe p tl pf i t i n y p eoi pf r ycnl y rile B yyvta y i b yt vi m by.d biao t
s t
btouf b i l) tsi mnc ln troVai bn n n tldit trls niml nac0ba e e nanrn nali eea.rr. eeol ee t1 rn. 2ot m m etate msDa m s' n m ma nm m nceoe n . w ncn pon Ai o nf con iivtr if ii af cep Ra Pt C i aym ici a amouu ao.e amnoseeo tsi n ts bl t kn Ro tl stt nrinn ni2ii n% sa nimemn o2ir oeoriis WC P ' oaeao Cesnc oe#ra Cm( tf C7rc CsDutli h
- dt 'J ei 2' tW ~ a c e" +
e^
ee l s n -.
7
. r eo ) f ~
r, 1 /.* .' ^
3 Sh T .
m o 's sP F
3 2 6
10 2( g d' rv f
sL 5 8 . .P s aC .
0 0 -
O.
pH 0 ss y ra op B 1
ty t uB n .
l b ey b
a it rn mc T te nn ie 7 3 8 1*
nm au -
E E
E on t q E 4 7 6 Ci 1
a ne 7 .
ct oi Ci in '
mo sC i
et Su s
.o sF 8 6 fh aP 5 2 3 ot pL 3 3 . .
i yC 0 0 nW LH 0 0 o -
ss t ie n rt aa e pt m mS n
iy o ac 6 5 5 Ceg t n 4 ne -
E E E a ou E 2 2 m Cq 2 4 8 a e D or
'e g NF t
a n P a l t t P n n y e i
i o e mh e
r m
a P o P n o
k o
o u N n3 t3 n2 a s r M t a lt b it it a
. n di li a di nn in e C. l nn IU MU S P IU
.R f '
oo p oo rt cSefS t dir m rt u pu.a uoP st n a eo pu.s bkeRq R sa sfdrd s bk sisV sisrOie lrf sc s nun s sriP sri uo ai aeala pr ir a
p atrR atres pn rell pf yasa pn yi .
y yoys yoyariie uoar bnk b sa lfd n
B bct islsf u gs s bctii l tii trnRa t t ttlr ttlri nsr norO s n n nnao nnaoaomi eattftoa e ete e
m e
m eatt mcau df mao n
m u t gl nbnna .
sn m mcau.
nifbe nifbsen il uam ie t
iiiinC Ai o ir if abr ar drL Ra C if ai yr u ai yr ad r do t o tteleS Pt tnltl tnltn ntt nnrit n ngrni ngrnredn oou ooounf Ro oiaoa oiaoerna Cnb cccbio WC B Csecf Csectuar h
dt) ei S
' tW c W r T o leeAs ) r )
eo s She r ) 5 r
o
- 1. ,,
T d o
- 2.
- E l e 1 0 . .
m u F 830 1( t os l sP 3 a 5. 3. (P ;,
rv cn 2 2. (P .P sL 2 .
l 00 /
f aC .
00 0
.s pH 0 '
ss I y ra B 2 opss t tya c)(_ l e uBp b y n
ey r it ) mc e b
a rn B ( nn ie 7 7 '
9
- t a
T tet ,
au -
E
.E l nm on en /v tq B
E 8 l
Cim ne an 1 or ct i 2 CF in a '
mot / )
sCn ~
)
r 3 i o 4 r 5 o
- E- r et C Su oR mssF o*E 778 8 2 e
t 3
3 fh W aP 10( 21( a .
oti B / , pL . .P .P l 0 h yC 00 nW(
o
( BH 00 ss t ie n rt e aa m r
pt n e mS iy 7 o ac 6 6 - t a
,., r Ceg tn - E ne E E 6 l a ou Cq 2 5 m
9 a e D or e NF g t a n P a l
P
) e y( y a
r m
a A B e c t C t D
t r N u
M t t t
n Jna S
,n
-. a n n a ,l ,l
. a a l i P / P l
l P C. P P R
4 '
O ~
l Ta bl e 9.1 Governing Limit States and Their Associated Capacities II) a g esme,+oea0 + ForW Containment Task 3(2)
Limit State Capacity (AH ) Limit State f (2)
Task 5
~
Cap (AH )
r e, j (Design SSE)
Fermi Ea:1.l ure cof, Fail ure .of . 0.399' hydt5 biological 0 3.99 biological ggg (0.159,I3)
Housner shield wall shield wall s pectra )
Failure of 0.45g Failure of 0.459 basemat in shear basemat in shear and bending and bending Clinton Liquefaction 0.83g Liquefaction 0.83g (0.25g,RG of soil under of soil under 1.60 spectra) basemat basemat Failure of wall 1. 0g reinforcing bars
~
(^)
C and liner Zion Failure due to 0.75g Failure of wall 0.34g (0.179 , interference reinforcing bars Housnar between contain-spectra) ment and auxili- Failure of wall in 0. 399 ary buildings transverse shear Failure due to 0.75g interference between containment and auxiliary buildings Shear failure at 0.759 buttress plates Sequoyah Failure of . 0.30g Failure of 0.30g ,
(0.189, shield building shield building l Housner
- spect ra) Failure of 0.529 I4) Failure of Greater basemat in basemat in transverse shear transverse shear than(4)
- 1. 0g
,^ Notes: (1) Capacities stated are in terms of peak horizontal ground
't) a ccel era tion, A, H and time histories consistent with Regulatory Guide 1.60 spectra.
l (2) Task 3 is load combination with dead, prestress, and seismic loading; Task 5 is the load combination including 0 9-4
klh PRELIMINARY INSIGHTS o FOR PWRS - HIGH CONSEQUENCE SEQUENCES RESULTS FROM SEISMIC LEVEL GREATER THAN 0.5G, MARGIN APPROACH NOT ADEQUATE TO ADDRESS THESE SEQUENCES-
- o. 'FOR BWRS - SOME HIGH CONSEQUENCE SEQUENCES CAN BE CAPTURED BY DOING MARGIN AT REVIEW LEVEL OF 0.5G o FOR BWRS - MORE DETAILED INVESTIGATIONS OF CAPACITY OF REACTOR INTERNALS AND REACTOR SUPPORTS NEEDED o CONTAINMENT STRUCTURAL AND BYPASS FAILURES NEED
. O-- NOT BE INVESTIGATED? EMPHASIS ON ISOLATION AND HEAT-REMOVAL FUNCTIONAL FAILURES?
O.
e
o l
w r u_
e e
(t D
o g g
b b .
J ID d.
3
.j b
In W
O
Y) .
AN APPROACH T0 THE SELECTION OF REVIEW LEVEL
'EARTHOUAKE (RLE)
USE BOTH LLNL AND EPRI HAZARD TO DEVELOP CONSISTENT SELECTION OF RLE u -
MOST PLANTS WOULD USE 0.3G IF HAZARD'IS CONSISTENTLY HIGHER USE 0.5G IF HAZARD IS CONSISTENTLY LOWER USE REDUCED PROGRAM BUILT AR0UND WALKDOWN O- TRY TO PLACE HAZA'RD SCREENING IN GENERALIZED RELATIONSHIP TO CORE DAMAGE ISSUES WHICH PARAMETERS?
WHICH FREQUENCY?
LLNL AND EPRI - COMBINED OR SEPARA1E O
3
ch
'WHICH PARAMETER?
MEAN - RISK RELATED, DOMINATED BY TREATMENT OF OUTLIERS, BIGGEST DIFFERENCES BETWEEN LLNL AND EPRI MEDIAN - NOT RELATED TO RISK, LEAST AFFECTED BY OUTLIERS BEST ESTIMATE OF CENTRAL TENDENCY (?), LEAST DIFFERENCE-BETWEEN EPRI AND LLNL 85% - SOMEWHAT RELATED TO RISK, AND CENTRAL TENDENCY OF HAZARD. INTERMEDIATE DIFFERENCE BETWEEN LLNL AND EPRI 10 BEST ESTIMATE -
LIMITED INCLUSION OF UNCERTAINTY (N0 INTRA-EXPERT VARIATIONS) - PROBLEM: ONLY AVAILABLE FROM LLNL O
z
_ ____-_- _ _____-___________ _ ____ _________--_ ___ ____-___- ___ _ _- __-______ _ _ _ _ = _ ___
Vv, t
N_
WHICH FREQUENCY?
AVAILABLE - PGA, 1,2.5,5,10 AND 25 HZ PLUS WEIGHTED COMBINATION PGA - MOST FAMILI AR, SIMPLY RELATED TO PRA 5 HZ OR WEIGHTED - MOST RELATED TO DAMAGE COMBINING EPRI AND LLNL L)
SHOULD IT BE DONE?
IF S0, HOW?
O l
1 -
i e4 L_---,_..s.-_._- _ _ _ . _ . - - _ _ . - - _ _ _ _ . _ _ _ . _ _ . _ _ _ _ . - - - - ._ _ _ - - _ _ _ _ - _ - - _ _ _ _ - - _ . _ . _ _ _ _ -
k PLANTS NOT INCLUDED-IN EPRI STUDY SHOREHAM COOPER COOK FORT CALHOUN PALISADES DUANE ARNOLD CALLAWAY BIG ROCK POINT GRAND GULF (CALCULATIONS UNDERWAY)
Si LUCIE (CALCULATIONS DONE BY UTILITY?)
TURKEY POINT (CALCULATIONS DONE BY UTILITY?)
,- \
V "
FOLLOWING COMPARISONS BETWEEN LLNL AND EPRI STUDIES USES 57 PLANT SITES FOR WHICH BOTH LLNL AND EPRI ESTIMATES ARE AVAILABLE - AB0VE PLANTS ARE NOT INCLUDED ALTHOUGH LLNL CALCULATIONS ARE AVAILABLE ALL ESTIMATES ARE OF LIKELIHOOD OF EXCEEDIllG 0.3s OR NUREG 0098 SPECTRUM ANCHORED AT 0.3G
,n v
. . . 4 i . .
ie ,
- s. . i i ,
, l . . ,
t e i t i..
.61 6_
1 . j t i .,e e ,
. w. . . . , I i !
--M---
- 4 . . . . . . ,s . , . .. 6 ,
. . . 4 i ei , i e ! q. . . 3
. 4 . , i
- - - ~
. , , , , ! i ; 6 i . . . . ,
---.--_j
, . i4 . 1
- ' LJ _p+A - M L, f /1 t / -. .
5 e i i . I i .6 # 3 w a gv r-
-- m - - --:
. . i i e ! . . . ! - ,,,.l--- .. ... ......i
, , ,; i, ; , , , . ,
i i 4 i . , -....--!
--- - - "_._. . . .]
. a 6 e . ,
, , , , 8
_ m . . _ . . . , , . j
. . . _ . , , f
- 8 _...-.. _.l i M
g p .. .
, . _ y'. .
r . . ... - .
- t. i
- a. . . _ . .
! . f m.>. - .t. .. .
. 1. , ,s , . . . i f ..p _ . ~ . . . s
.a. j[ *d
,n, e\
. ' m. .
,. e ,,,. ,u. _ .af
- - ,e- m ' ' - _ t'_.- _ _ _ ... .
4 i . t i \ ! i t . .
i t i i 1 r X t . 4e i 6e i i . A e i * . . i i . . ,i. .r i e i _
, i l e _ , T, a i . .
. j . 6ef i . 4 .
. .i. . 5 i
\ i . 3
. . Ii t , I i ei .
. i . 4 ,\ . * .
. I8 i . 4 6 i . i t N i . 1 . . . . . . . , . . , . 6 , .
i I .i
_ _ J1 su s 1 . t i 6 t .
! . ,\ . . , 4i a e e i 6 I t . i 4 .
~~ \ 3
! . . I 6 6 3 . i . 1 . i . . .6 i i 1 . I I
4
\
. ; _ 7 .. ! _ _
t e
. . 1 6 . ,' . . . . i , .
i . . >
a w . . . 6 , l . 6 i i . . .
E N . 4
,A .. 4 i. , i . . . , i i .
' ' N . . i s . . . . t , # 6 . .a
. . ., i . < . ,
.A r . ,
7'"
. s -
- * ~ ~
- u. L . i e i 3 ,,,A . . . 6 . . *y
-=. , , , . . . .
.. t . .I
- 'A
_._y i .
~
. b . ...) . n .
-- \ . . . . . .
I ___.'
.. . . _f z
s gw , e\ _-..*
+.7 A ew l_
.p . .- - .
g.
..._ ._. _ _ _ _ _ p_. _
, . . ... . u--_..
\~. . .. _,._. _w . . .
_ _ \_ _ __ _ _ __
. _ 5.. '
__)...
P m \ _.
. M & . - ; . .,- .e
._. 3 _. - . . . - - ._ . . .
-.65 _.
. i i i . p .
\ -.
t-
. i.
___ ._ _ _. q._ .
i 6
. i . . . - . . _ _ _._ . . ..
i .t e t .. _.. .
A i l ..
A
. t66 i , ,
m .y
. l m .
' ' ~-~
t . , . . . . . _ ,
'a............_
-w _.._ .___. s=> , ....._.....-
L. __ . .
m ._.M4 - _,.
---=_.. _. .. ..
. ,. . . . .< }
.. .. . ~ ..
..3 6
t s..a ... -_.._a
- . .. )
- - . . . . . j. __ . .
- . . .p , . . _ .. - . . . - .g
. . . . . ._1...a.... _. L_ J_. e
.. ..____a. . . . . . . . . _ . . _ . - . _ . _ ' . .. . _ . ...' .
...i - _ . _ . .
I
'L
_. .L _
.. . }. .. 2.
i l . .. .
e..
e Y
._ -. .. . _ . .. . . - . _.. . . .j . . . .. .l._....._ g. . .
l
.4 _ _ .
- 9. .l .
. _ .s g..._-._.-. .
1
\
4 . I '_ c a- _ _ . - ... ..... . . g.
,I
. . I t i .. . . .. . . . . . _ . . . _ ._. ..
. . 4 . ...__ .---. ... - _ . .
p._
.J . ..... . _ . . _ - - . . . _ . . . . . . . .
t t
I e . .-
135703 KEuFFEL a ESSER CD t
20 sees e o.a.a.
D . 0 so to vses os css - _ _ . _ - - - _ - . - . _ _ - _ . _ _ _ _ - _ _ _ _ _ _ _ __
i . . . i , ,
.e i . t . #6 iF . . e . ; )
, , I . ' . t I - "-!
. l g j e ,., t . 3 . 1 i e , ,
i i, . . . . , =~.
4 _ _ . . _- - y i , , , , i , ,
-- ~ __ ..
. . , i i e i . . ,
1
, , , ,i . . ,,, , i _
i i * .
,j- t
. i
. i , ! . . .
l a , , -, - ~ . ---_..
. i _ j i ,
- l . . . a /TV"- 19 --- ~ * ---- - ;( - - . - .j
. . 4
~~'7~~ , .# 11 ,' 6 i i eii _! LJ / L j
- ' ' ~ ~
it i I it i i t .t4 O id IL i -/
. , l
. i ,
- . a , s ,
, -r , -
, - . . _ _ _ .l ..
l, i , .i
-- - --. _-s... . . ..
- p. .
. . . 3
_ _. _.. i l 1. ' a. -
p..._._
l
-l
- i. .
l
,i J.-
w . Q rl .- --
l
+-. - .
i ,
i .
t_. ,
- n
/
t....__... -. ..) ,
g.
, i i y , .,, , p f, t -- 7 f _
.y- , , r r r. .' r ,
- ~ ' - -
i
.s. . . E ! ,
I i p
..** t - - es , _ ,W
- r. ,n, .\
n'" -eQ, g .
~
5, .i i
A
- ,. , , i g . , , . . j
. i i . .x i. '
3 . . . i
. ! A i. . . , i . s ,s .
i . . i_ T i . . . . 1 i i 6 .
! . _ , aJ. < -
! .. 1 . i i St . !
- I i , , , . s . .- , .
6
, , . . . i . me . i i . . i # , i
. i e I . i, i N' . i 1 . . , . ,t t . . . . ,
. 1 _ _ .1 >
. . ! ! , 6 i EU s , . , . i 1.- .g. ,- I .!
__ ~ e t i ! * , .
JVA . i i . . -
16 6 . t. t i i i t i i . . .
e . e l I i 3 . . . s i . . t . . , ai i i . i 3 , 6 , f . , i . i , i
+ . ] . t , , 1 x _ . . . '
! , ,, . , , , . i . . , , , , ;
, 3 .. A a - i . i i . , e . . . , i e i .
(
. i s i i . . . . . ,
f . i1 . { . . N i i , I . . t
- j . __
i . ( E. N . i . . i . . .. i . , ,
. ' ' A. s . .i 1 4 . . . , . . ,
. . A A i ,
i
. 4 . , .K . . . . . . i w M K 1 A . . t
. i ,,e
.. .A . . , e s i A i A.2. .
I
. ... . 'A i .
_. t
__n-
. t, .
n
_._ - 4 v7 m .
~
g . ,
.---~).- t m _ _
M- . _ _ . . . _ . . . . _ - ._ ,
9 Bi s= _._
\
.__ ___ .._ _ u.____..,
. ._. - . ..w _ . .e
- u. . . _ _
A
- x. _.__.
g,. .
_; ... . . . . .l E i ~
'\ '
.__. 3 ._. .
.Q . .
&__- --e i S ..
g m- . _ . . .
. i e i i,
. I 6 i.
.t . ,
. t i
- . 3 ..
s . .__ .
4 .
_i . . .
. _ . . . __. . _ . - _ . \ . .
, 4 . .
s . . ,
_. _ g .._ -
. . p. g
, ,,3, - _ ._
_. g _....._._ g t._.._
.:-...._._ . . . . .g. -
_= .
. ..i
. . = .. ..
, --..q
- - - - . " - . * . j ._ _6 h__ ----N . % . _ . _ _ .WW 6M - 6 N*-8
.-..t..__- . . _. t_...-- . - - __- . .
.. . _. s . . . _ _ . . - . _ . - _ . .. ._ _ . .. ..
t ...._...i.____..-
i ' --_ _ .
.)__._.____) _._._.p__.-.____.
i ... __..____..--
. _ _ . . _ .. p_- _.
__p : - .- . _ . . _ _ _ .
. 4 i ....j... -
.p, ,
_..l.___
l , i *
.. . 4. _ . . .
s L _ ._ .. . -- . -
l .-*
j, ,
q.__._..... a i. . .. ,
p-- --t
- a. _. .. _. .. . . . . _ -,
--_: i
. . . . l . s , 1 J
128708 KEUFFEL 3 (55tR CO E se 3 se To tses anscas ,
ones e.s 4.
we I
F i
._6 s
/%
3 - - -
._ _ 1 - w* . ..
e -
_-_._,,,,, j 3
we
. m. -. . . _ . . - -...o. ,.. O. . , , .
. , , , , ,.,4 y 1 - -
L_ _ ' ' ' ' ' -. y : ._ . ._ .-
- _ g -
b I b ._ _ . . . . _ , _ . _ _ . 6 f . .
i , i . .
i .4 - _- . . _ . . _ .;
g h 4 4 .f E f l e i e f s ) 4 l i 8
-._.. ..I .,I6 . ., ..
- . >I n - __
1 4-
.. _ . .. _ . . . _.. . .. e
. 1 p.__._- . ..
_,, . f_. __. . . ,
} __ _ _ . .
L .. .
.p 1,.......3 _
- g. __. .. . .._ 4
., .N . . '
- a
,. t.
l__ _
. _ - .____L_._.______...p
. . _ . ._. _ _ p . .._ . ..;
j j-g
.. } . .._. . . .__;... . .._. ,. . _ ._ ._ .
1
_.p _
.{ . - _.. L "L ...__.._......_..____....
.[._ _
=; ._.____3_. .
._2
,2
- _t - _ ,4. 4_. N _ _ _ _ _ .._. ' _..._. ...
.4 p ,. & _p g.ei _ . gA__ ___ _ . .. . _ ._ _.._-
.r .
. - . _ .i
_ . . . ..qa .
_ .L -. . .-
_p k .
f , . _ , -
4 4
\ .
N T i
i-l
_4 9. _ _ . , _ s
.. I
, N. _
4 i . . a
,,,__g- - g .
=
- x - -__ ._
f .
. . ~~T'
- a
_ .. .B . L - : *
- g ..
x . _.
.p e. e go. e e pe 4.. ge' -
N em._
l'
._ 3 -17~D '
1
. J l
, . _ . . _ _E '
E\ f f i ia . .._ _ . _ x & i
. . . ._. _ .. . -. w . . _ _-. . _.. ... . __.
_ ._ .. .. . . . . _ _ . ._ _ _.i e.. .~T _. N... _ _ _
_h _ . - W 6._ _
.N-
- _ .3 SD-- . N _ _. . A m . __. _ P' .
, - _ . I . .
._ . _ .. . . . m.. ._. - . 3
.. i.._. . ..ns, _
.e_ . . v . . _ . . . e f .d.
_ . a ..
...og.. ..g.. . . . , .
hb . .. sg_ .6 h w- h.... ... .h._. __..c. --
.- . _M -
. . . _ . . _ . . . .=_N ._he .h 6. .m ss .e .. .. .e._ati._ im e.. .e ^
._ _. . e .e. .
.w g.d
... .e.- .
. _ _ _ __ _a ... . _ _ _. __ _ _ . _ _ _. . ._ _ .
. _ . . . . E ,_. -
! ' I ' a
. . . . . . - - . . _ .L...__.....*
i _. ,
h.
.4 i I
_. E_ . . _ _ _
.N e..e e
,e. .
.s_ e-i...-.__.-__. '
e e g... .p i
g_
.y_..__._.._.
_. __ y _
_ . . q __ ...
e _. =>. .
.. e.... = _ _ . .w. . . . ..
I
.a...=e +-e= .==* =. **t
. .x 4 . . ee 4 g ,
A. ...g
- . .- t
,=.__ . . , . - . .
s
- 1. . .
..[..._... 1._ ..._.___[_._. .. . _ a _ _ . _ . . . 4__.___
. _.. _ . ._7 &-_
.l.
_. .._l
.__.....I
..[ . _. . _;_ ; , . . _ .
e
.6m. -
9 = *. v.+- . . * . . . . = . _~ . ,. .. .. .s
. . = . * . = *
- ,.. . _.4 . _ . . . . .
.,d '
g
- i. .. .,._. _ .
.._ .._ . . . . ., .6 . . . .
...h.-
_.q. l _-_ . .
/ r o
.._ . l . -.
. . " . .._,&. +s.e.a aim pg..g.mn .g . .e,g.,a, I
_.4 _ ___ . _
7.._._.._..._. . . ._. .. e
. . . . . . _ _ , j . .. .. ..
.e... 4
._ l. -
. e.. .% -
. e..
. w _. .
_p. ..- _. _ _. _ . _ . ;
g, . . . . . . .. .. _ . - . . .-+ _a 1 *
-g s i t
12 5705 .
< * -%,eD ' a to to taes sescos KEUFFEL S E55ER CO.
mass an e s 4
C u
1 l
h: ' RANKINGS OF 57 PLANTS (EXCEEDING 0.3G PGA)'
l MEAN MEDIAN 85%
-EPRI LLNL' EPRI LLNL EPRI LLNL ;
1 36(11.5) L1 1(1.4) 1 3(3.4) 2 .27 3 3 2 27
'3 4; 4 4 4 4
}
4 7 7 7 3 36
-5 18 2 2 7 7 6 38 8 27 5 1 7 1(26)' 14 14(2.2) 6 14(8.2)
.,:. g I
'51 19 45 52 45 55 j 52 26 53 48 53 20 q 53 39 52 54 52 53 )
54 42 54 57 54 32 55 32 56 53 55 57 56 ~ 20 55 55 56 54 ;
57 56(108) 57 56(15) 57 56(44) i d
NOTE: NUMBER IS KEYED TO RANK USING MEAN PROBABILITY 0F EXCEEDING 0.3G PGA IN EPRI STUDY
( ) = RATIO OF-LLNL PROBABILITY TO EPRI PROBABILITY LO:
. __.______.___________________________._____________U
, ,< t s ;
q j
l I _ i i I . * ! 6 i i i i i 9 , e i I , t . , 6 e i i , , . . , i iI . . , y 1 i t I ' ! e , 8 e>6 ie i 1
- f a t e i ! ! . . e 6 i . l .
- e e t ,
,, . . # , +
i . 6 6 16 e i 6 . . t I i e , i . .- ,
1 e ) , i i t i . . * ! I i t ! t 4 . * *e l l 5 6 6 t i e i . t ! I e i . 4 & . . . i , . , 4 e
e e . y u- ( i e e 6 l a . e eJ . e t I i e i i } i i _L ,.~ i . .
g_
i ' i 6e e tt '
Amt t _-, . -- - .
i ,- --
i i e i f IT t e i . } i
- l 1 6 , , , , [ <
g i ,
, y i e i i i i t . *f i e ! i a I , , I . , - * * - -
1 t i . 6 e l , I i i 4 I ( t . i a i . .
t . I ! !1
- I i 6 i ! . , a . i i , .
t t t . t 4 I t I f f I f e 6 j , , 4 . I . g g , , a t1 1I i i t , i i l1 6 3 .
6
{ . .
t j i 6 g l l , t i i 6 , , . $ { { . g
- ' i
. , . . i . 6 + 6 i i( , s .
- - - -*
- j
. i t t i i n i i 6 . v . . I 6 . a.
,A .
A
,7 l ----r
. , , , . ! 6 . , + . . . .si gg g --
. e i I . l i i i 6 i ,
i i
- 4 i ' l I i .
t i
i
- 4 t . ,
husul
,a
- ', ~
i r
i t t g.
. 3 ie e , I i ' 1 . 6 i i e t )6 t i !
- ~ ~ '
i . e ! ! ! i i e ! i
- i 6
- i i e i , f 6 , 4
. . , 6 i 4 , , a # i 6 , , t . . . ! 4f i 9 .
t . i i i i 1 e , i 4 , . . i i 1 i t i . I _I t 1 i 4 . i 1 e i t . i e i ., ;f . 9 i i g . .sg ,
- 1 i e i ..M i ' i 1 - M3- t t i 1 !m Jr i * .n=V + . . e i s an T " i * ' FD . t 4 i , 3 ; i yu < J.a - i su , 4 ss s- ,- . . ,
e i i t . .
9 *il ' t t % . t t I ! i . I i i , i +! t t 6 . . l . . . . 41 1 .e , t i 1 i 4 i e i ia ! , ' i i 6 i e i i i i ! t t ! 6 i
- , i t .
i e . . i a i .
see . e . , "i i t ! . a e . . I 6 t . . . e , e 1 6 (6 , i , 9 i i i . . ; . , . ; ,
~~"
a ii& 1 I t i 6 e
- i 3 4 a I . e i 6 5 i i i i,ie t 1 E.T -
r
', . 3
. 4 , . .
I t I i e i en i l i . I i
- 4 f i i . 4 . . t +f , f i i i e i 4 6 i ; , 4 6 e i e i ! t i f i * ! i ' 6 ' ! ! i
- 3 . I e i A i , . . 6 ,
1 i i e i 6 i . I , l 9 6 4 i . 6 % .
e i i 1 4 e ! i ! i e i 4 e i , e( i a 6 i , ! l i i 6 , 1 ! }
. 21 t , e e i . e 't 6 6 (U . . , , a 1 6 - .
. * +t t . 6 , t e .
I l 8 i e ,! ~
t i . i ! ii t i . i i 8
! . I6 e . i i t ie e 4 f 4 6 4 1 1e a ti 9 t M . t 6 665 ie *
- i I . i ,
3
. . I , 1 6 , I i 4 . t *
- l t t I t I ea ! 3 i i i i . i , e . I i i i e $e . . e ; e i e i
) .
_e i n 6 i i 1 . IN . i i . ae e i 4 i . i i i ; j e 6 4 e f: ! ' i ak i 1 , , e i i i 1 3 . e i e . . . ,
. i ! ! ! * , t i _X_ L.J i i a 1
+ i i t . . 6 I , . e te , i i . e . . a 1i i .
6 i , i 6 .e i i . 6 6 i et . i , , _ . i i j i . i . . i .
. . , 6
! e i . &
8h C f 4 a , . . * , i e i i t 6 e I I t 6 i ee . , e .
- M qug . 6 i i . # i i u _ , t . > . , . . , # .
6 ' i i !. . i n . e 1 6 e ! ! t . . . t # . 4 i i t i , t s e t e !
- i '
e i t t . . . 8 i s
- . . . i e . . i i . ; . i a e .
i kwm' g i 1 . t t i a i _a ttI !
- i e a e ' # .
. . i *
. i i i . . i , a i i 4 ,.I A_ . 6 i _
- * .. 6 1 . 6 i i 4 6
. e' e i . . . 4^ 4 . . g ~4,4p t i . . . , , , a . .
f* ! . ! . . , ! i . . . I N. 4 * . i * . s . . .
a . t. . e . ! . e . , . . .
- Q o 'k, ,= .
'~~
q vw -
Q -
t t e*
'J
~
3 %._ __.
BT _
l f
~.:
. L _ _ _ -
-._e
.,,_: _ m . , l
. i . i e . i e 4 i ,; - . . . . ie , . .
I
,6 . e i ' i_i _ :
. . .
- i 1 . i .
4 i '**
i . ,
k e i
_J U - > - , # 4
- t e l 4 6 i 4 , i
! l . 4 Y , a i 6 e . i .
1 ,
. . } e ( o I i i I . { ! 4 6 "a e . i . 4 i e t 6 . . ,
- f T"~ -
' i i ! 6 t I i4 6 e * ' 6 , i g L. g . e . . i . i i . . . e I . i 6
i 4
I e t i g
s e i
. i . . .
' i e' e f IJ l ( e a e e . i e e '
. . I e te _e i ' .
e i . 3 . . .
g I a i a 4 4 e _ e , . e .g . . ! e i 1
. . . t ,
i - q g
. . 1 .- . . .
_ 1.. . - ,, -
,e r . .
e .
19 7 i - , ,
~
! a I l m.7 } . l . - . 1 i l t .
- I , ; I '
I f
(' 9 1i t 8
,y * ._,,,g' t . 4 .
_ ,,, . e a .
I s , -
. gi e e i . . i . ?
i , s . .
i . 4 l . . . 4 ' .
_: . 4 ..-,. ~
h* .
. m' . + .
- L gt w. ;',_ ' ._
._7_.._L._.__...
- _ ,__ y _-'
hDTi,hhMI-hMO
~
l
~ -
. t-- ,
' I 6>
e e ,
I i-6%_. TA, . .
t ...
u
'14WOiR.. al%... ):. f_IWhidWM4~#~~
I_--
i ier i 1 , . e t i i . > , . a _._ ,_1. i
.g___,.
. . . _ .. . . . . J
, . 1 ,,. . . i , . . . . .
1 . . . . _ .._s 3 3 3 . e i . e i i } . . ' ' -
t -
- t, ""'"4
_. i .i ! , i +,, . .
- 1. I
- i i
- i IL_ ' . l. J_ _a * -
. i 1 --
.k) ,
SOME CONSIDERATIONS
+ . ;,
- 1. NO SIMPLE THEORETICAL SOLUTION - NEED TO LOOK AT CONCLUSIONS BASED ON DIFFERENT CRITERIA AND PARAMETERS
-2. UTILIZE OHG0 LNG LLNL CALCULATIONS WHICH GROUP AND RANK PLANTS BASED ON DIFFERENT HAZARD MEASURES FOR BOTH EPRI AN LLNL HAZARD
- 3. FOR PURPOSE OF BINNING PLANTS RELATIONSHIP BETWEEN HAZARD AND CORE DAMAGE NEED ONLY BE VERY GENERAL
- 4. UTILIZE NUMARC/EPRI CALCULATIONS WHICH USE PARAMETERIZED MEAN HAZARD AND GENERIC PLANT FRAGILITY
~ {)
- 5. UTILIZE, AS NECESSARY, DETERMINISTIC INFORMATION
- 6. DEFINE PLANTS WHICH FALL INTO 0.5G, 0.3G AND REDUCED PROGRAM BINS BASED ON CONSISTENT PATTERNS WITHIN DIFFERENT APPROACHES O
r
_m _ _ - _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ - _ _ _ . _ _ _ _ _ _ - _ _ _ _ _ . _ _ _ _ . - _ - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ _ _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ . _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ _ . _ - _ _ _
b ,
, SEISMIC HAZARD CONSIDERATIONS IPEEE PRA' SINGLE HAZARD CURVE IS VIABLE AS LONG AS USES OF PRA ARE NOT
" BOTTOM LINE" ORIENTED CURVE SHOULD REFLECT SOME UNCERTAINTY CANDIDATES - MEAN, 85TH%, BEST ESTIMATE IF POSSIBLE, REFLECT BOTH LLNL AND EPRI IF BOTTOM LINE NUMBERS ARE REQUIRED FOR SOME SUBSEQUENT r ANALYSIS, FULL UNCERTAINTY NEEDS TO BE TAKEN INTO ACCOUNT l
l l
l O
1 m ._ _ _ _ _ _ _ _ - _ _ - _ _ _ _ _
L ) ,
SEISMIC HAZARD CUT-0FF FOR PRA PRESENT CURVES TO 1.0G FOR LLNL AND EPRI SOME~PRA APPLICATIONS HAVE EXTENDED TO GREATER L VALUES (0NE STUDY TO 2.4G) WITH NO CUT-0FF FOR HIGH CURVES CUT-0FF CAN HAVE SIGNIFICANT IMPACT ON CORE DAMAGE OR RISK (?)
DIFFICULT TO DEFINE RIGOROUS PHYSICAL BASIS FOR CUT-0FF r3 PGA' HORIZONT-AL RECORDED AT 1.3G PGA VERTICAL RECORDED AT 2.0G +
INTENSELY SPECULATIVE WHAT PROBABILITIES ARE OFACCELERATIONSjb1.0G SOLUTION ASSUME ARBITRARY CUT-OFF BASED ON SPECULATIVE NATURE OF ESTIMATES?
AVOID DECISION MAKING BASED ON LIKELIHOOD OF SUCH HIGH GROUND MOTIONS?
(l' l
WQ-
'4 l
l INDIVIDUAL PLANT EXAMINATION FOR EXTERNAL EVENTS (IPEEE)
HIGH WINDS, FLOODS, AND OTHER HAZARDS O
PREPARED BY SUBCOMMITTEE ON HIGH WINDS, FLOODS, AND OTHERS EXTERNAL EVENTS STEERING GROUP 1
0 '
s ACRS 9/6/89 1 ;
-_---------_- ----_-----------_--_--- _ _ - _ _ _ _ _ _ a
.(
HIGH WINDS, FLOODS, AND OTHERS SUBCOMMITTEE -
COCHAIRMAN: D. JENG NRR COCHAIRMAN: W.BECKNER RES MEMBERS: T. CHENG OSP l J. STROSNIDER RI J.CHEN RES N. CHOKSHI RES R. WESCOTT NRR K. CAMPE NRR O
\
1 j i
l II i
0 10 1)
ACRS 9/6/83 2
+ >
l
-~ -- __ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _
. I O
SUBCOMMITTEE MISSION
- MAKE RECOMMENDATIONS TO EESG ON iPEEE FOR HIGH WINDS, FLOODS, AND OTHER HAZARDS l
- PROVIDE INPUT TO GENERIC LETTER FOR IPEEE
-- PROVIDE GUIDANCE FOR IMPLEMENTATION OF IPEEE FOR HIGH WINDS, FLOODS, AND OTHER HAZARDS i
O ii i,
h a %
4 l
.i l
O' l
- l. ACRS 9/6/89 3 i
i I l - _ _ _ _ _ _ _ _ _ _ _ _ . _____________L
- P .,
HIGH WINDS, FLOODS, AND OTHER HAZARDS MAJOR SOURCES:
HIGH WINDS, TORNADOES, TORNADO MISSILES, FLOODS, EXPLOSIONS, TOXIC GASES, FIRES, AIRCRAFT AND SHIPPING ACCIDENTS WINDS:
LOSS OF OFFSITE POWER (INCLUDED IN IPE)
. METAL SIDING STRUCTURES
. STRUCTURES WITH THIN CONCRETE WALLS AND ROOFS
. TALL STACKS
. OPEN TANKS
FLOODS:
RIVER, COASTAL, AND LAKE SITE PLANTS b.l .
L .
INTENSE LOCAL PRECIPITATION (Gl 103) .
l l
OTHERS:
l
. INDUSTRIAL / MILITARY FACILITIES
. TRANSPORTATION (AIRCRAFT, TRUCKS, TRAINS, SHIPS, PlPELINES) 1 1
d O..
ACRS 9/6/89 4 l I J
a,. ,
BASES FOR EXAMINATION HIGH WINDS
- PRAs SHOW THAT HIGH WINDS / TORNADOES CONTRIBUTE SIGNIFlOANT RISK TO CORE DAMAGE AT SOME PLANTS (e.g.
ARKANSAS, POINT BEACH, TURKEY POINT, AND INDIAN POINT)
- OLDER PLANTS AND PLANTS WITH FACILITIES NOT DESIGNED '
AGAINST NRC CURRENT CRITERIA NEED PLANT-SPECIFIC EXAMINATION TO DETERMINE THE RISK DUE TO WIND / TORNADO HAZARDS FLOODS
- PRAs SHOW THAT FLOODS CONTRIBUTE SOME RISK TO CORE DAMAGE AT OCONEE AND TURKEY POINT
- OLDER PLANTS, USUALLY DO NOT MEET CURRENT NRC O CRITERIA, NEED TO DETERMINE THE RISK DUE TO FLOODS TRANSPORTATION AND NEARBY FACILITY ACCIDENTS
- OLDER PLANTS WITH OL REVIEWS PRIOR TO THE IMPLEMENTATION OF THE SRP '
MAY NOT MEET CURRENT CRITERIA
- CHANGES IN LAND USE AT SITE VICINITY CHANGES IN FREQUENCY / SEVERITY OR PREVIOUSLY l EVALUATED HAZARDS INTRODUCTION OF NEW HAZARDS O
5 ACRS 9/6/89
b L U RECOMMENDED IPEEE APPROACH FOR WINDS, FLOODS, AND OTHERS REVIEW PLANT SPEC!FIC HAZARD DATA AND LICENSING BASES (FSAR) .
IDENTIFY SIGNIFICANT CHANGES, IF ANY, SINCE OL ISSUANCE DOES PLANT / FACILITIES DESIGN MEET CURRENT LICENSING CRITERIA YES (QUICK SCREENING & WALKDOWN)
NO IS DESIGN HAZARD ACCEPTABLY YES LOW?*
NO BOUNDING ANALYSIS * -SES (RESPONSE / CONSEQUENCE)
NO PRA*
DOCUMENTATION (INCL. IDENTIFIED REPORTABLE ITEMS AND PROPOSED IMPROVEMENTS)
- Thesse are optional steps.
O 6
ACRS 9/6/89
'()
e%
CURRENI' NRC SRP DESIGN CRITERIA FOR HIGH WINDS,'IDRNADOES, AND 'IURNADO MISSILES HIGH WIND DESIGN CRITERIA 100 YR. RECURRENCE WIND (100-130 MILES /HR.)
- SRP 33.1 AND 353
- ANSI CODE A 58.1 TORNADO AND TORNADO MISSILE DESIGN CRITERIA
- R.G.1.76 AND 1.117 (MAX. SPEED = 360 MPH, PRESSURE DROP =
3 PSI FOR ZONE I)
SRP 33.2,3.5.1.4, AND 3.5.2 p}
L
- ANSI CODE A 58.1 O
ACRS 9/6/09
,, ~ .
SRP FLOOD CRITERIA R.G.1.59 POSITION 1 RIVER PLANTS - SAFE SHUTDOWN SYSTEMS ABOVE PROBABLE
~
MAXIMUM FLOOL)
LAKE AND COASTAL PLANTS - SAFE SHUTDOWN- SYSTEMS ABOVE MAXIMUM HIGH WATER LEVEL DESIGN BASIS FLOOD PROBABILnY ESTIMATED TO BE LESS THAN 1.0E-5 PER YEAR R.G.1.59 POSITION 2-ALL PLANTS - ADEQUATE WARNING FOR SAFE SHUTDOWN ALL PLANTS - SYSTEMS TO ACHIEVE SAFE SHUTDOWN Q PROTECTED FROM STANDARD PROJECT FLOOD (ABOUT 2.0E-3)
ALL PLANTS - SYSTEMS TO MAINTAIN SAFE SHUTDOWN PROTECTED FROM DESIGN BASIS FLOOD ACRS 9/6/89
c.
)
SRP ACCIDENT EVALUATION CRITERIA ACCIDENT SCOPE
- INITIATORS HAZARDOUS MATERIALS OR ACTIVITIES IN VICINITY OF PLANT
- PLANT INSULT TOXIC GASES, EXPLOSIONS, FIRES, MISSILES
- CONSEQUENCES i
RELEASES OF SIGNIFICANT QUANTITIES OF RADIOACTIVE MATERIALS 1 SCREENING CRITERIA (SRP 2.2.3)
- OFFSITE EXPOSURES IN EXCESS 0 10 CFR PART 100 2-HOUR EXPOSURE TO AN INDIVIDUAL AT SITE EXCLUSION AREA BOUNDARY EXCEEDING 25 REM WHOLE BODY OR 300 REM THYROID
- ACCIDENT FREQUENCY LES$ THAN 1.OE-7 PER YEAR ACCEPTABLEWHEN ESTIMATED ON BASIS OF AVAILABLE DATA
- ACCIDENT FREQUENCY LESS THAN 1.0E-6 PER YEAR ACCEPTARLE IF REASONABLE QUALITATIVE ARGUMENT CAN BE USED TO SHOW THAT REALISTIC PROBABILITY IS LOWER 0
ACRS 9/6/89
s IPEEE PLANT WALKDOWN HIGH WINDS:
METAL SIDING STRUCTURES STRUCTURES WITH THIN CONCRETE WALLS AND ROOFS TALL STACKS OPEN TANKS FLOODS:
RIVER, COASTAL, AND LAKE SITE PLANTS TRANSPORTATION AND NEARBY FACILITY ACCIDENTS:
INDUSTRIAL / MILITARY FACILITIES TRANSPORTATION (AIRCRAFT, TRUCKS, TRAINS, SHIPS, PIPELINES)
SEISMIC AND FIRES WALKDOWNS PROVIDED ELSEWHERE 1
ACRS 9/6/89
h I I
NRC USE OF IPEEE RESULTS ;
The NRC will evaluate licensee iPEEE submittal to obtain reasonable assurance that the licensee has adequately enalyzed the plant design and operations to discover instances of particular vulnerability to core damage. or unusually ' poor containment performance given a core damage ~ accident. The NRC assessment may lead to one of the following:
- 1. If NRC assessment indicates that the plant design or operation should be changed to meet NRC regulations, appropriate enhancements will be required.
- 2. If NRC assessment indicates that plant design or operation could be enhanced by substantial additional protection beyond NRC regulations, appropriate enhancement will be recommended and supported with backfit analysis in accordance with 10 CFR 50.109.
- 3. Otherwise, enhancements would not be suggested unless significant new safety information becomes available.
ACRS 9/6/89
HFOS END PRODUCTS AND SCHEDULE
.END PRODUCTS:
- GUIDANCE AND CRITERIA FOR IMPLEMENTING IPEEE FOR HIGH WINDS, FLOODS, AND OTHER EXTERNAL EVENTS.
INPUT TO IPEEE GENERIC LETTER ADDITIONAL 'INFORMATION NEEDED FOR THE STAFF 'TO REVIEW AND AUDIT THE SUBMITTED IPEEE RESULTS SCHEDULE: CONSISTENT WITH ALL EXTERNAL EVENTS DRAFT BY END OF OCTOBER,1989 FINAL BY END OF 1989 Q
I r
'0 ACRS 9/6/89 e - - - _ _ - _ - - _ _
o .
L .- -
f -
1 NRC/ INDUSTRY INTERACTION
. MET WITH NUMARC TECHNICAL STAFF IN THREE MEETINGS l (7/14,8/11, AND 9/1/89) i
. PRESENTED STAFF IPEEE PHILOSOPHY AND SPECIFIC APPROACH TO NUMARC STAFF
. RECEIVED NUMARC COMMENTS STAFF /NUMARC INTERACTION CONTINUING ACRS 9/6/89
-__ ___-- -- ~ . - - _______ _ _
A,M. .y z ,,,_ ,
. grll' I!
s,lr"
'r +
)
m
- _ Floods, High Winds & Other Events l' Bases for Examination Basic Approach-Progressive Screening i-SRP Criteria and their General Conservatism Search for Outliers / Plant Walkdown
$' NRC/ Industry Interaction use of_IPEEE Results End Products and Schedule Subtotal = 3/4 hr.
^ '
_.______._______._m.__ ___ _ . - _ _m____.__m_m_____.________.mm____m_.____._