ML20246P476
| ML20246P476 | |
| Person / Time | |
|---|---|
| Issue date: | 07/12/1989 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1748, NUDOCS 8907200219 | |
| Download: ML20246P476 (183) | |
Text
r"'M f y"f L f W[
O DR a \ A' UNITED STATES NUCLEAR REGULATORY COMMISSION ADVISORY COMMITTEE ON REACTOR SAFEGU
)
In the Matter of: I
) l
) J SUBCOMMITTEE ON AUXILIARY )
and SF.40NDARY SYSTEMS )
)
\
Pages: 1 through 152 Place: Bethesda, Maryland Date: July 12, ~ --
19X2S 3fficb Cem_
r - /
Retair
_ _ _ _. . ,_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.; [ar ~~ ^2- '
's -
44 4.,, _ _ _ _ . .
nUtLIlv
'_%l^r.I I L V V i u l u V U g ; n i j j t (gj HERITAGE REPORTING CORPORATION O oeuaw-1220 L Street, N.W., Suite 600 l
Washington, D.C. 20005 (202) 628-4888 890/200219 090712 PDR ACRS PNU e
T-1748 l .-
p ,( 1 PUBLIC NOTICE BY-THE 2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S
- 3 AL"ISORY COMMITTEE ON REACTOR SAFEGUARDS 4- July 12, 1989 5=
6 7 The, contents;of this stenographic transcript of.
-8 the proceedings.of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards 10 (ACRS), as reported herein, is an uncorrected record of the r1 11 discussions recorded at the meeting held on the above date.
4 y 12 No member of the ACRS staff and no participant at 13 this meeting accepts any responsibility for errors or 14 inaccuracies of statement or data contained in this 15 transcript.-
16 17 18-19 20 21 22 23 24 25-Heritage Reporting Corporation (202) 628-4888 i
.___ _ __=_ -_ _ _ _____________ -
O s d
[?[C
- - $3
- 1. .
I '
UNITED STATES NUCLEAR REGULATORY COMMISSION.
. ADVISORY' COMMITTEE ON REACTOR SAFEGUARDS
~
~
.In-the Matter of:
- )-
= . )
SUBCOMMITTEE ON AUXILIARY. )
and. SECONDARY SYSTEMS )
)
- Wednesday, g July ~12, 1989 Room P-110, Phillips' Building' 7920 Norfolk Avenue Bethesda,; Maryland-
'The meeting' convened, pursuant to notice, at.1:30 p . m .-
'BEFORE: MR..CARLYLE MICHELSON, Subcommittee Chairman-
, Retired Principal Nuclear Engineer
~
' Tennessee-Valley Authority
' Knoxville,. Tennessee,'and-Retired Director, Office for Analysis n and: Evaluation of Operational Data U.S. Nuclear. Regulatory Commission-Washington, D.C..
ACRS MEMBERS PRESENT:
MR..'AMES J CARROLL' Retired Manager,. Nuclear Operations Support Pacific Gas & Electric Company San Francisco, California DR. IVAN CATTON Professor of Engineering
' Department of Mechanical, Aerospace & Nuclear Eng:.neering School of Engineering and Applied Science
,b University of California DR. WILLIAM KERR Professor of Nuclear 2ngineering
< Director, Office of Energy Research University of Michigan Beritage Reporting Corporation (202) 628-4888 a --__ __ _ _ __ _ -__ _ _ - - _ - .
p _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _
2
_ (N
'V.-
ACRS MEMBERS'PRESENT:
DR. CHESTER P. SIESS Chairman, Structural Engineering Professor' Emeritus of. Civil Engineering University of~ Illinois Urbana, Illinois MR. CHARLES J. WYLIE Retired Chief Engineer Electrical Division Duke Powe.r Company Chr.rlotte, North Carolina COGNIZANT ACRS STAFF ENGINEER:
.MR. SAM DURAISWAMY NRC STAFF PRESENTERS:
JOHN H. FLACK CONRAD McCRACKEN JOHN CHEN ALSO EEg,SENT:
'\- ] JOHN LAMBRIGHT JERRY WERMEIL Heritage Reporting Corporation (202) 628-4888 L ,
l- _____________________J
s ,
r 3 1 l
1 F_ B 2 C E E R .I H G A 2 MR. MICHELSON: The meeting will come;to order.
- 3. MR'. FLACK: Good afternoon. My name is. John l
4 Flack.- I.am from the Office of Research, Severe Accident 5~ ' Issues' Branch, and previously'the Project Manager of the 6 Fire Ris't Scoping-Study, which has since been complete'.
7 The purpose of my presentation is to present the 8 Staff's response to the recommendations reported. in the Fire 9 Risk Scoping Study.
10 .These recommendations and the Staff's response 11 were formerly presented to the Commission in the Commission 12 -paper. dated. June 7, 1989.
-13 Basically, my presentation will be broken down 14 into three areas: background, discussion and summary.
[
15 The background I will not spend a' lot of time on.
16 We have gone through the study before. But discussions, I 17 will present the recommendations and then what the staff's 18 response will be to those.
19 Okay. Basically, the Fire Risk Scoping Study was 20 to assess the risk significance and dominant sources of 21 ur.cortainty associated with fire and nuclear power plar.ts.
22 For important PRAs were used in this evaluation, 23 or assessment: the Indian Point II PRA, the Oconee, 24 Limerick and Seabrook.
25 The second purpose was to assess the risk of Heritage Reporting Corporation (202) 628-4888 l
4-( 1 potential fire-related issues.- These were issues raised by 2 the ACRS and other people in the. community.
- 3. There is a third purpose. And that is'to-assess 4' or evaluate the need for additional fire protection:research 5 based on 1 and 2.
'6 In summary, the study used state-of-the-art 7 methodology. It did not develop any new methodology. There.
8 were no funds available for that-at the time.
9 It used the most up to date data available. It 10' co'nsidered Appendix R modifications. And as mentioned 11 before, it considered all the potential issuer, 12 There were three main findings from the' study, 13 which were reported in the Commission paper..
('s 14 The first is that the study confirmed that fire is
. \.)
15 still an important contributor to core melt, at'some nuclear-16 power plants, even after the regulations have been met.
17 Number two, the state-of-the-art methodology has 18 modeling shortcomings which could result in large 19 uncertainties, depending on how the codes are used, 20 basically, the codes that are used.in the fire analysis for 21 modeling propagation.
22 And the tb4.rd finding was that there were se'reral 23 important areas that were not adequately addressed in PRAs.
24 .And these are listed here:
25 Fire induced alternate shutdown-control room panel O eritaee neeerei e cereeraese-(202) 628-4888
4.'
E 5 j l' interactions; 2 Smoke control and manual fire-fighting 3 effectiveness; 4 Adequacy of fire barriers; 5 Equipment survival in fire-induced environments;.
6 and 7 Seismic / fire interactions.
8 So those are the three main findings of the study.
9 DR. KERR: 'How much uncertainty is there in the 10 'first finding? You mentioned that there were large 11 uncertainties in the methodology and modeling.
12 MR. FLACK: Okay. I forgot to mention that John 13 Lambright is here'from Sandia who was. involved heavily in.
14 ~ the study. I would like to direct that question more to 15 John, who is more knowledgeable in that area.
16 MR. LAMBRIGHT: My name is John Lambright from 17 Sandia National Laboratories.
18 And as far as modeling uncertainties --
19 DR. KERR: My question was, how much uncertainty 20 is there in the first conclusion, which is that fire is an 21 important contributor?
22 MR. LAMERIGHT: As far as overall uncertainty 23 estimated in total core damage frequency for any of the four 24 PRAs, range factors vary from factors of 5 to 20.
25 DR. KERR: I'm not making my question clear. The 0 erie ee Regereine cergeraelen (202) 628-4888
6
- ( ') 1 conclusion, or the finding is that fire is an important 2 contributor -- this is a general statement -- at some NPPs.
3 And my question is how much uncertainty is there 4 in that conclusion?
5 MR. LAMBRIGHT: Well, I guess we don't have a 6 quantification for that uncertainty.
7 DR. KERR: Well, I'm willing to accept a 8 qualitative.
9 MR. LAMBRIGHT: Well, based on findings from other 10 internal events, and external events, it is not in the 11 noise. It is above the noise level. It comes in as a 12 rather 10 percent --
13 DR. KERR: Ten percent confidence?
14 MR. LAMBRIGHT: 200 percent confidence?
15 DR. KERR: No , 10 percent.
16 MR. LAMBRIGHT: Oh. Per the fire PRAs to date, 17 fire using similar methodologies to internal events and 18 seismic methodologies, has been a 7 to 50 percent 19 contributor to the overall total core damage frequency. In 20 the NUREG 1150 risk assessments that I was involved with for 21 both Surry and Peach Bottom, fire contributed a factor of 3 22 to a factor of 6 greater to core damage frequency than all 23 internal events combined.
24 DR. KERR: The fact that you identify, or I guess 25 identify is the right word, modeling shortcomings, and I am
( )y Heritage Reporting Corporation (202) 628-4888
L 7
l() ll not sure what'is. meant by "could result" in large i
g 2 . uncertainties, would lead me'to believe that this conclusion.
l
.3 that fira is an important contributor would have similar 4 uncertainties.
5 Is that a valid interpretation?
'6 MR. LAMBRIGHT: I think in' Item.2 there are 7 several codes out in the market right now.
8 DR. KERR: I am not talking bout the codes. I am 9 talking about the conclusion which is that the modeling h 10. could have large uncertainties. It would seem to me that if 11 the modeling has large uncertainties, which leads you to 12- believe-that the results have large uncertainties, and you 13 use the results to reach the forced conclusion which is that
, 14 fire is an'important contributor,-that that conclusion would 15 .have the same uncertainty.
16 Am I missing something's 17- MR. LAMBRIGHT: No, that's true. It could be.
18 That could be. However, in the case of the NUREG 19 1150, for Peach Bottom and Surry, the same models and same 20 entries were used as was in the case in internal events.
21 DR. KERR: This does not necessarily give me a lot 22 of confidence.
23 MR. LAMBRIGHT: Well, let me finish my :onclusion, 24 if I could.
25 DR. KERR: Okay.
Heritage Reporting Corporation (202) 628-4888
- _ _ - _ _ _ - _ - _ - - _ - _ _ _ _ _ _ - _ _ _ - _ _ - _ _ _ - - - _ _ - _ _ _ _ _ _ _ - - _ - _ _ _ _ _ _ _ _ _ _ - _ _ _ _ - - _ _ = - _ _ _ - _ -
8 l
f'% i
'(,) . 1 MR. LAMBRIGHT: The fact that uncertainties 2 calculated for the fire analysis were of similar order of 3 magnitude to the uncertainties, overall uncertainties of the 4 internal event analysis, and with using the same models as 5 in the internal event analysis, the fire was calculated to 6 be a factor of 3 to 6 greater in total overall core damage 7 frequency than all internal events combined.
8 DR. KERR: But with large uncertainties?
9 MR. LAMBRIGHT: With range factors of up to 8 or 9 10 which-is simi3ar to internal events uncertainties.
11 DR. KERRt What does a range factor of 8 or 9 12 mean?
13 MR. LAMBRIGHT: That is the difference --
(} 14 DR. KERR: I mean, does it mean a factor of 8 or 9 15 uncertainty?
16 MR. LAMBRIGHT: Well, it is a calculation for if 17 there is a normally distributed distribution of a variance 18 between the median and the 95th percentile.
19 So typically a variation in overall core damage 20 frequency of between the 5th and 95th percentile for fire 21 would be an order to an order and a half of magnitude.
22 DR. KERR: Okay. So there is then about the same 23 uncertainty in one's conclusion that fire is an important 24 contributor?
25 MR. LAMBRIGHT: As is internal events an important
) Heritage Reporting Corporation (202) 628-4888
9-
'1 contributor,lyes'.
f 2 MR. FLACK: But'I think I wanted to. point out here-3 that there are other uncertainties based on past PRAs that 4 the staff, at least John, has not-reviewed, which had used 5 different codes, which would also contribute to these 6 uncertainties.
7 Now, if they~use for example COMPBRN I code versus 8 a later revision, COMPBRN'III, that would introduce another 9 type of uncertainty.
10 DR. KERR: I am just trying to understand. There 11 seems to be in number 2 a conclusion that there probably
~
12 could be, I am not sure which uncertainties, in whatever one 13 predicts.
14 Number one does not seem to at least explicitly; 15 point to any uncertainties in the conclusion that fire is an 16 important contributor. And it would seem to me that one 17 would expect'about the same amount of uncertainty in those 18 two conclusions.
19 MR. LAMBRIGHT: As I said, if_one uses a similar 20 methodology in comparison of internal events for fire or 21 seismic, then there are going to be similar uncertainties, 22 and when you take a look at distribution there is a 23 potential that in fact one internal event could be a more 24 dominant contributor than fire or maybe even more so fire 25 than internal events.
() Heritage Reporting Corporation (202) 628-4888 l
b ,
10 1 DR. KERR: Okay. Now, you are saying that fire 2 turns out to be a much more significa1.t contributor than any 3 of the other contributors to core damage?
'4 MR. LAMBRIGHT: Except for seismic' initiators.
5 DR..KERR: Well, let's see. Is fire being treated 6 now as an internal or an external?
7 MR. LAMBRIGHT: As an external everst.
8 DR. KERR: Okay.
9 MR. LAMBRIGHT: However, in the case of the 1150 10 methodology that'I developed, fire is treated now in a
- 11. similar fashion as_an internal event because similar models, 12 and the same fault trees and event trees are used. In fact, 11 3 all the random failures not related to the fire itself, are 14 treated in a similar fashion as they are in internal events.
15 So now there is direct comparison between the way fire is 16 treated and the way internal event' initiatives are treated.
17' MR. MICHELSON: What do you define as an' internal-18 event?
19 MR. LAMBRIGHT: Something such as just a normal 20 loss of power.
21 MR. MICHELSON: Yes, I understand those as 22 internal events, but how do you rationalize that fire is an 23 internal event in that sense? It is rtainly external to 24 the equipment.
25 MR. LAMBRIGHT: Well, what I am saying is that it Heritage Reporting Corporation (202) 628-4888
i 11
[ -
'If is'. comparable with internal events. I'm not saying'it is an 2- . internal event.
3 MR. MICHELSON: Okay. I thought you were 4 indicating it was treated as an internal event, and I was y 5 confused.
6 MR. LAMBRIGHT: No.
~
7 MR. MICHELSON: Because I didn't think it fell 8 under the internal event category.
9 MR. LAMBRIGHT: No.
10 MR.-MICHELSON: Okay.
11' DR. CATTON: Do you model fire suppression?
12' MR. LAMBRIGHT: Yes. Fire suppression is modeled.
13 DR. CATTON: How?
14 MR. LAMBRIGHT: In the case of the 1150 plants, 15 fire suppression is modeled based on a data base that was 16 developed at Sandia National Laboratories where we have data 17 on roughly 70 fire events that give time to suppression.
18 And a distribution was developed for probability of non-19 suppression in any given time frame.
20 DR. CATTON: So you really don't model it, you 21 just have some data and you have correlated it one way or 22 another#?
23 MR. LAMBRIGHT: Aspects are addressed such as 24 taking a look at the fire brigades and their response times 25 to see if they are comparable with what we feel is a good
'O rit 9- a rorti-9 core r ti -
(202) 628-4888
i-12
- l. ) 1 fire brigade.
2 DR. CATTON: Are you familiar with the McGuire i I
3 incident that took place a month or two ago?
4 MR. LAMBRIGHT: No , I'm not. l 5 DR. CATTON: The fire suppression system didn't 6 work. !
I 7 MR. MICHELSON: It worked, but it wasn't 8 effective. ,
9 DR. CATTON: Didn't put the fire out.
10 MR. MICHELSON: Didn't put the fire out. Right.
11 DR. CATTON: Maybe I'm using the wrong word.
12 MR. MICHELSON: Yes.
13 MR. LAMBRIGHT: One point I would like to make
(~T 14 though is now --
G 15 DR. CATTON: Was it at Oconee?
16 MR. MICHELSON: Yes, it was Oconee.
17 DR. CATTON: No wonder I haven't been able to find 18 the LAR.
19 (Laughter) 20 MR. LAMBRIGHT: One other point I would like to 21 make is development on 1150 methodology. Now we are 22 treating plant specific fire suppression in this fashion in 23 which we have independent fire protection engineer to come 24 in and evaluate the plant for every specific critical fire l 25 area, look at brigade response times to those areas, assess
() Heritage Reporting Corporation (202) 628-4888
13 s i
(
q,i
) 1 factors such as what will be in the earth when the fire 2 brigade arrives to develop specific engineering judgment 3 distributions on all critical fire areas based on specifics 4 in any given plant's fire brigade.
5 DR. CATTON: Do you bring to bear any of the work 6 that has been done through the National Bureau of Standards 7 by. people like K.T. Yang at Notre Dame?
8 MR. LAMBRIGHT: In fact, a couple of the codes 9 from the National Bureau of Standards are used --
in DR. CATTON: I am talking about the work that is 11 published in the literature by Yang and by people like 12 Giluria and others. I don't hear these names or see these 13 names anywhere.
f~g 14 MR. LAMBRIGHT: K. T. Yang was consulted as part
%)
15 of the fire risk scoping study as one of our experts in our 16 expert solicitation process.
17 So we have tried to use the knowledge of these 18 experts as well, information from the National Bureau of 19 Standards and other expert groups around the country for the 20 development of our methodology.
21 DR. KERR: From what you have said, can one draw 22 the conclusion that the results in 1150 do not include these 23 modelings or the fire contribution or that the most recent 24 , edition I guess which we have not yet seen of 1150 does 25 include the fire results?
() Heritage Reporting Corporation (202) 628-4888
14
'l- MR. LAMBRIGHT: I'm sorry. Could you repeat the
, '2' question, please?
\
3 DR. KERR: Does the revised edition, if that.is 4 the right word,. of 1150, include the latest fire risk
.5 results.as part of the core damage total?
6 MR. LAMBRIGHT: 1250 fire methodology has taken a 7 look at all potential fire-induced initiators in a sense as-8 internal events would have. However, some of.the specific s 9 issues that were identified such as potential independence 10 of the remote shutdown panel were not specifically addressed 11 because of funding constraints.
12 DR KERR: So'what I am'trying to understand is 13 whether this most recent edition of 1150 shows,a significant.
e '14 increase in the core damage because.of the contribution of 15 fire which was not treated in this way earlier.
16 MR. LAMBRIGHT: In the case of both Peach Bottom-17 and Surry the estimated core damage. frequency for both 18 plants was 1 in 2 E to the minus 5 per year, where for the
-19 internal events core damage frequencies for those plants
.20 were factors of 6 to like 3 below that.
n 21 DR. KERR: So it does represent a significant 22 increase?
23 MR. LAMBRIGHT: That is correct.
24 DR. KERR: Yes. Thank you. l 25 MR. MICHELSON: In doing your analysis, the latest
.() Heritage Reporting Corporation (202) 628-4888
h 15
,m.
i
(,) 1 analysis under 1150, are you yet trying to model in the 2 effects of heat and smoke migration?
3 MR. LAMBRIGHT: At this point, we haven't dir'ectly 4 taken a look at that particular issue of heat or smoke 5 migration into adjacent areas.
6 However, in the case of Surry, we have done a 7 subsequent analysis of potential effect on core damage 8 frequency in inadvertent suppression, and we did take a look 9 at some issues there such as smoke migration.
10 MR. MICHELSON: Do you mean you looked at 11 inadvertent suppression caused by the migration of heat and 12 smoke?
13 MR. LAMBRIGHT: That is correct.
f) 14 MR. MICHELSON: What did you use to model the L/
15 migration of heat and smoke?
16 MR. LAMBRIGHT: What we did was we estimated 17 carrier failure rates based on generic failure rates for 18 different types of barriers.
19 MR. MICHELSON: The migration I am particularly 20 interested in of course is the open zone migration where we 21 are using physical separation of 20 feet for instance as the 22 barrier between two trains of redundant hot shutdown 23 equipment. Maybe Surry didn't have any such cases. I don't 1
I 24 know.
25 MR. LAMBRIGHT: That was the case for Surry. They l Heritage Reporting Corporation (202) 628-4888 l
l
16 I (,) 1 didn't have any such cases in Surry.
2 MR. MICHELSON: They did not. Everywhere that l 3 smoke and heat could go there was a three-hour, non-4 penetrable barrier for smoke and heat?
5 MR. LNMBRIGHT: The surry strategy, in the case of 1'
! 6 many of their fire areas, is filled out for cross connection '
7 from the opposite unit to mitigate any unintentional 8 effects.
9 MR. MICHELSON: Yes, but in order to do a fire 10 PRA, you have to consider the consequences of fire. What I 11 am trying to find out is how carefully you model what the 12 consequences might be.
13 For instance, did you model the migration of the 14 heat and smoke to the point of actuating other fire
(')T
's 15 protection systems in the adjacent areas? Because heat and l 16 smoke will do that. Then you didn't really model the 17 effects of heat and smoke migration.
18 MR. LAMBRIGHT: Not in the case of 1150.
19 MR. MICHELSON: In the case of Surry, did you? In 20 any case, has anyone that you know of modeled heat and smoke 1 21 migration to determine if there will be unwanted actuati on 22 of fire protection?
23 MR. LAMBRIGHT: Not to my knowledge, 24 MR. MINNERS: As part of Issue 57 at Sandia, did j 1
25 they -- I
() Heritage Reporting Corporation (202) 628-4888 l
~ .. _ _ _ _ _ - _ _ . _ _
I 17 l '. MR. MICHELSON: Did what?
2 MR. MINNERS: Which looked at migration.
3 MR. MICHELSON: Has that been documented yet, 4 where we can read about it? Which one is that? Oh, that is 5 that preliminary draft. Yes, I have a copy of that, and 6 .there you are trying to -- I didn't want to.get into that 7 because it hasn't come before the committee yet to be 8 discussed and so I was trying to find out.if in any other 9 forms,-and particularly any other PRAs, by industry, have 10 they ever attempted to determine whether or not adjacent 11- fire protection will be actuated because heat and smoke has 12 migrated from the fire area?
13 MR. LAMBRIGHT: Not to my knowledge.
r- .
14 MR. MICHELSON: In the fire PRAs which you have 15 .done, again relative to the effects of fire, have you 16' attempted to determine what effect heat and smoke will have 17 on adjacent equipment, not from the viewpoint of igniting it 18 now but simply from affecting its operability. How do you-19 determine if 20 feet away.there is a solid state cabinet,
- 20 whether or not it will function from the effects of the 21 fire, unless you somehow model the thermal change that is 22 occurring and whether or not that affects the equipment.
i 23 MR. LAMBRIGHT: I agree, this is a potentially l
24 important issue.
, 25. However, with limitations inherent in the fire Heritage Reporting Corporation (202) 628-4888
1 i
1: '
l 18
) II propagation code, it is just at this point not capable'of 2 performing --
l 3 MR. MICHELSON: To your knowledge, has anybody 4 attempted to model to that level of detail in their PRAs?
5 MR. LAMBRIGHT: In the case of Limerick PRA, a 6 model'was made for one area, I think the electrical switch 7 area, where there was some calculations made from a fire 8 that would go into a switch area and potentially fail in the 9 switch area.
10 MR. MICHELSON: It was sort of a direct radiant 11 heating model?
12 MR. LAMBRIGHT: Yes.
13 MR. MICHELSON: But no attempt to determine smoke-14 and heat migration effect as far as you know?
[
15 MR. LAMBRIGHT: Not to my knowledge.
16 MR. MICHELSON: Okay. Thank you.
17 MR. LAMBRIGHT: You're welcome.
18 MR. FLACK: The first topic I would like.to touch 19 upon is the state-of-the-art methodology with regard to 20 computer codes used to treat fire propagation.
21 Most fire PRAs use the COMPBRN code. There'are 22 three versions of COMPBRN, COMPERN I, II and III, and three 23 modified by Sandia, although modified COMPBRN III is not 24 available to the outside public at this point.
25 Basically the COMPBRN code is a zone model .and it
( Heritage Reporting Corporation (202) 628-4888 1
1
--_-__.m___ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . I
m 19.
.l' ..is tailored for nuclear power plant fires or cable tray-
~2 fires. Its simplicity-is highlighted,.which is'an economic 3- benefit'for running..the code. .The code, however, only 4 models thermal effects'and as I have mentioned,.there are 5 several versions. 'And all the versions lack adequate
+
6 validation.
7 It has been validated for' simple configurations,-
.. 8 but nothing'as complex as what we find in a nuclear power,
- 9. plant.
' 10.- DR. CATTON: Has anybody tried to establish how 11 good'it'has to be in order to support a-PRA? Because 12- clearly, the modeling capability.is out there.
13 MR. FLACK: You mean'the validation,-how-much 14' validation?
15 DR. .CATTON: I think the ability to model' fires is-
'16 pretty good.
But it costs you a lot of money.
17 MR. FLACK: That's right.
18 DR. CATTON: So has anybody.tried to establish 19 what the level of modeling is that is needed for a PRA? I 20 have heard arguments that COMPBRN, as weak as it is, is 21 ~ adequate for use with a PRA.
22 MR. LAMBRIGHT: And that is true in most simple 23 cases one might find. In fact as part of the fire risk 24 scoping study we attempted to determine what the uncertainty 25 would be in the overall core damage frequency estimates
'O rie se *i-9 cerrer tie-(202) 628-4888 L
20
]D m) .
1 given the fact that COMPBRN predicted almost literally no 2 time to damage or substantially shorter times, and the 3 effects on any given fire area might be changes in core 4- damage frequency from a factor of 3 to 20.
5 DR..CATTON: Oh, as high as 20?
6 MR. MICHELSON: But see, they are not using it to 7 determine the migration of heat and smoke beyond some 8 localized zone where they are trying to determine damage.
9 MR. LAMBRIGHT: That's right. COMPBRN 10 specifically is predicting damage --
11 MR. MICHELSON: Right at the fire.
12 MR. LAMBRIGHT: -- right where the fire is.
'13 MR. MICHELSON: Yes.
14 MR. LAMBRIGHT: And if it is 20 feet away from the 15 fire, COMPBRN III will not predict damage, unless the area 16 is of small enough size.
17 DR. CATTON: Or if the fire is big enough to 1
18 radiate, or do they just shut it off independent of the size 19 of the fire?
20 MR. LAMBRIGHT: Even with huge fire sizes. And by 21 that I mean fire sizes of up to 15 feet in diameter.
22 Because we have taken a look at let's just try and determine 23 uncertainties, and COMPERN III, one could have a cable tray 24 25 feet away, say, and not get damage.
l l 25 MR. MICHELSON: You are defining damage as not Heritage Reporting Corporation (202) 628-4888 l
t L__
21
(~')
v 1 igniting the cable though, I suspect.
2 MR. LAMBRIGHT: That's correct.
3 MR. MICHELSON: As opposed to elevating room 4 temperature enough to harm the operability of equipment 5 which is just as much damaging as burning it up. If you 6 heat it up too much and it fails, you've lost it whether you 7 ever ignited it or not.
8 MR. LAMBRIGHT: For cable you are looking at 9 ignition temperature as well as climate temperature.
10 MR. MICHELSON: Yes. Cable of course is fairly 11 resistant. You look at solid state instrument cabinets when 12 you do these, make these assumptions about no damage at 20, 13 25 feet, because some equipment is very sensitive to damage,
( 14 long before it ignites.
C]/
15 DR. CATTON: They usually paint it black, too, 16 don't they?
17 MR. MICHELSON: I don't know about that. But it 18 will get black from the smoke.
19 DR. CATTON: So what has to be done?
20 MR. FLACK: Well, I was going to go over that in 21 the next slide.
22 DR. CATTON: Okay.
23 MR. FLACK: Okay. The study had three 24 recommendations in this area.
l 25 One was to process and make available the
() Heritage Reporting Corporation (202) 628-4888
l 22
() 1 enclosure fire test data gathered under the old Fire 2 Protection Program. I understand the data is available, but 1 3 it just needs to be processed.
l 4 DR. CATTON: Are you going to include the NBS data 5 as well?
l 6 MR. FLACE: These are just the recommendations 7 that were made by the fire or scoping study.
8 DR. CATTON: Okay.
9 MR. FLACK: Two was to assemble and validats a 10 physically consistent fire simulation model. There had been 11 found inconsistencies with the COMPBRN III which had to be i 12 modified.
13 DR. CATTON: But what if you take one of the other (3
%)
14 off the she i kinds of programs that are available? Would 15 they be adequate?
16 It seems to me before you initiate another code 17 development program, which tend to be money sinks, you ought 18 to take a look around and see what else, what is available.
19 MR. LAMBRIGHT: That was done as part of the Fire 20 Risk Scoping Study and the conclusion was that while COMPBRN 1
1 21 III had limitations, it was currently the best code 22 available for modeling a nuclear power plant fire situation.
23 DR. CATTON: Why are nuclear power plant 24 situations different than buildings? l 25 MR. LAMBRIGHT: That, I am not really an expert on
) Heritage Reporting Corporation (202) 628-4888 l
1 l
E< 23 f( 'l. theLcode itself so I don't know if I can give you an 2' adequrate answer.
3 DR.-CATTON: I' don't think there is a difference.
4' MR. FLACK: Well, there is differences with 5 regards to, I understand, from furniture burning in a room 6 to fire propagation along a cable tray.
l 7- DR. CATTON: Yes,-but those are the source terms 8 for the program.
1 9- MR. FLACK: That's true.
10 DR. .CATTON: What you are talking about here is 11 the distribution of smoke and energy.
12 MR. FLACK: And the modeling of the generation of l
13 the smoke and energy, yes.
E14 DR. CATTON: That's the source term. You may-have
? 15 to change the source term. But the basic modeling seems to 1
16 me, I can see no physical reason for it to be any different.
27 MR. FLACK: I guess you're right. I guess you can 18 make other codes work, the same as COMPBRN could probably 19 work for other environments.
20 People use COMPERN since it has been around a long 21- time. But I suspect that they could transfer to another 22 code if one w.snted to. They are flow model codes. The s 23 Notre Dame code is one.
24 DR. CATTON: Notre Dame code is one of the ones I 25 was thinking of. Also the work that's been done by Giluria O erie 9e aegerei 9 cereer eie-(202) 628-4888 l
( l' at.Rutgers. .COMPBRN was done.for a purpose. It was put 2 together at'a level that they felt'was. appropriate for PRA.
3- And your conclusion is it is not' appropriate for PRA and I 4 think I agree with you.-
5 But I certainly would look at these other 6 possibilities before I would launch another code development.
p
'7 program.
8 MR. FLACK: I think we are going to the bottom 9 line where there is no action going to be taken at this 10 time.
11 DR. CATTON: I think that is not the way to go, 12 either.
13 MR. MICHELSON: That is the first hurdle to get
{} 14 15 over.
MR. FLACK: .8 is just what the recommendations 16 -were being made from the scop 3 ; study.
17 The third recom:nendation was to validate 18 correlations which were in the codes'themselves for 19 predicting cable tray flame spread rates, flame heights, 20 threshold teatperatures for ignition and damaging cables and 21 mass burn rates of cables.
22 And as I just mentioned, the staff action on these 23 recommendations is that there is none proposed at this time.
24 DR. CATTON: Did I understand you earlier when you j 25 said that the contribution to core melt was three to seven
() Heritage Reporting Corporation (202) 628-4888 ,
.-. .__- D
25 l -
(,m)
-1 times that of internal events?
2 MR. LAMBRIGHT: That is correct. Roughly about 3 two to six, I believe.
4 DR. CATTON: Two to six times, right?
5 MR. LAMBRIGHT: Times, that's right. Times total 6 from all internal event contributors to fire.
7 DR. CATTON: And the staff is not proposing to do 8 anything in this area?
9 MR. FLACK: Not at this time.
10 MR. MICHELSON: .And that was an oversimplified 11 model in terms of the real effects of fire and migrating 12 heat and smoke and setting off fire protection. None of 13 that was even in this. That's all uncertainty which I think
(} 14 tends to make the situation worse and not better. And those 15 uncertainties are not necessarily in here because you didn't 16 try to model the uncertainties in your estimates of 17 actuation of fir- protection outside the immediate fire zone 18 and'so forth.
19 MR. LAMBRIGHT: That is correct. In fact, any of 20 these areas could probably do nothing more than add to the 21 core damage frequency.
22 MR. MICHELSON. 'Yes. So you are looking at really 23 a conservative estimate of three to six times.
24 DR. CATTON: Not conservative. Conservative is )
i 25 when you go high. They went low.
) Heritage Reporting Corporation (202) 628-4888
26 (f l' 'MR. MICHELSON:- Conservatives meaning that you are
-2: worse off.than you were, than you-think you are.
3 MR. McCRACKEN: Conrad McCracken,.NRR.-
4 The assumption.that the Staff is going to do 5 nothing in fire is incorrect.
6 We have concluded we don't intend to do anything-7 additional on the code development.. That does not mean we 8 do not have other means of looking to see what the
'9 vulnerabilities are in improving areas where vulnerabilities 10 exist.
11 MR. MICHELSON: Are we going to hear about those 12 'later today?
13 MR. McCRACKEN: Yes.
14 MR. MICHELSON: Okay.
15 DR. CATTON: What about this business at Oconee 16 where the automatic suppression system didn't put out the 17 fire? Do you know why it did not put out the fire?
18 MR. McCRACKEN: An area , hat we have. concern'and 19 we have identified and are working with some plants on is 20 they have not done adequate full tests of some of their 21
~
suppression systems that have been installed.
22 DR. CATTON: I'm not talking about whether it j
23 turned on or didn't turn on. It turned on but didn't put 24 the fire out. The question is why. ,
1 25 MR. McCRACKEN: Right. i i
() Heritage Reporting (202) 628-4888 Corporation i
27
() 1 DR. CATTON: Unless you start a fire,.you are not 2 Lgoing to'be able to answer that question.
3 MR. McCRACKEN: An automatic suppression system is 4 designed for a certain basis. If that design is not tested
- 5 to verify that in fact it will' function and it will put in 6 .the-amount of suppression, suppressant'that you need,.~in the 7 . area that it is supposed to be, then you can come.up with a 8 problem.
9 We have found that in some plants and we have been 10 -having plants do testing of installed fire suppression 11 systems.
'12 DR. CATTON: How doLthey do that?
13 MR. MICHELSON: For the CO2 and Halon, they just 14 do a full test.
(
15' DR. CATTON: But if it'didn't put out the fire, 16- why didn't it put out the fire? Was it because it didn't 17 flow?
18 MR. McCRACKEN: I'm not sure --
19 MR. MICHELSON: It wasn't, it was a manual system.
-20 MR. McCRACKEN: at was a manual system.
21 MR. MICHELSON: They just brought in two carts of l
22 150 pounds each and that wouldn't do it and then 150 pounds 23 of chemicals and that didn't do it. And they finally turned 24 the water on in about an hour.
25 DR. CATTON: Well, that is kind of perplexing,
() Heritage Reporting Corporation (202) 628-4888 j
i
.j L---_--__- - _ _ _ . _ _ _ _
h y
b 28 1.- 't oo .
2" 1G1. MICHELSON: Yes. Took them about an hour to e .3 get that fire!down the road.
4 MR. McCRACKEN: The Oconee fire was a fire in a 5- controlled area where the fire brigade responded for a short' 6 time. Their' conclusion was.that they had time to suppress 7' the fire. They elected not to put water on it right away; 8 and they-went through a couple of other suppression 9 sequences as they tried to de-energize the buses.
i 11 0 When they finally de-energized the buses, they 11 went to then water suppression.
'12 Had that fire started to grow, they would have 13 gone to water suppression earlier.
[} 14
'15 DR. CATTON: But wouldn't it be helpful if you knew whether or not the suppression method would work 16 beforehand? Then you wouldn't have fooled around trying.
17 these other methods. You might have gone straight to the 18 water.
19 HR. McCRACKEN: You don't go to the water unless 20 you have to.
21 DR. CATTON: Yes, but you ought to know'whether 22 you have to before, not after.
23 DR. KERR: What would you suggest they do? Start 24 a fire in this thing --
25 DR. CATTON: No, but --
( Heritage Reporting Corporation (202) 628-4888
~
e v 29
'DR. KERR: -- so they:can put'it out?
'l-
-2 DR. CATTON: -
if their plans are.that they.re
- 3. . going to1 wheel in 150 pounds of,' what, CO2, or whatever.it 4 is, to;put the fire out,Jtheylought to.know whether it's 5 going to work or not.
l y 6 .And if that means that you mock up the fire, or do
/
7 some-kind of-analysis on it, you should do that.
l 8 It seems.to me you ought to.know before.
l-
-9 MR. MICHELSON: Yes, the effectiveness of CO2 as a
-:10 fire.mitigant is a. question --
' ll- DR. CATTON:- Whatever the fire mitigant is you
. 12 ~ ought-to know something about its' effectiveness under given 13- circumstances.
14 MR. McCRACKEN: They do. That's why they had a-
~
15 trained fire brigade there.
16- DR. CATTON: And that's why the trained Ifire 17 brigade used stuff that didn't work?
18 MR. McCRACKEN: No.
19 DR. CATTON: Didn't train them very well.
20- HMR. McCRACKEN: The trained fire brigade knew that 21 those suppressants were not as good as water to put out the 22 fire, but.they did not want to start putting the water on an 23 electrical fire. They knew that the fire was controlled 24 within a certain local area and not spreading. They knew 25 they had time to try the less drastic measure, which was to
'O rit 9- rorei a cerver ** -
(202) 628-4888
30 cm
(. ) 1 use the CO2.
2 They were not positive it would work, but they had 3 time to try it. If it didn't work, they were still 4 'available to use water at any time they saw that fire start 5 spreading.
6 DR. CATTON: ic seems to me that the best place to 7 try it would be out in my back yard, not during a fire.
8 MR. McCRACKEN: Every cabinet and electrical fire 9 is going to be different. Some you will put out with CO2.
10 Some you won't.
11 DR. CATTON: It just seems to me that you ought to 12 know a little bit about it.
13 MR. McCRACKEN: The easy conclusion is, just put
(} 14 water on everything. And that is not a good conclusion.
15 DR. CATTON: Not necessarily.
16 MR. McCRACKEN: You can't model every possible 17 cabinet configuration and cable configuration.
18 DR. CATTON: When I hear that two to six times 19 internal for the core melt probability, it seems to me that l 20 it deserves a little more attention.
l 21 MR. McCRACKEN: That number by itself doesn't tell 22 you.
23 DR. CATTON: I understand that. I understand.
24 MR. McCRACKEN: If the internal events number were 25 10 to the minus 7, I would expect fire to be 10 to 20 times (G,/ Heritage Reporting Corporation (202) 628-4888
31
() 1 what the internal events is.
2 DR. CATTON: Okay.
3 MR. MICHELSON: We're talking about pretty high 4 magnitudes. I thought it was around 1 to 2 times 10 to the 5 minus 5, is the kind of magnitudes we are talking about.
6 MR. LAMBRIGHT: That is correct, for the case of 7 Surry and Peach Bottom.
8 MR. MICHELSON: Yes. So we are not talking about 9 low probability events. We are talking about relatively high 10 probability events, and this is the one that sticks out 11 amongst the relatively high. Correct?
12 MR. McCRACKEN: I agree with that.
13 MR. MICHELSON: So it is a serious concern.
14 DR. KERR: We had a presentation on motor-operated
}
15 valves and it turned out that they also increased core melt 16 frequency about an order of magnitude.
17 I'm beginning to think that if you pick out almost 18 any issue and study it carefully that you get -hout an order 19 of magnitude increase in core melt frequency.
20 MR. MICHELSON: That's because, Bill, I think we 21 look, if you pick out na issue and look hard you will find 22 out what the situation really is. And until you look hard, 23 it is back over in the noise. And fire has been in the ,
i 24 noise until relatively recently.
25 DR. SIESS: I suggest that something we haven't 1
f~)
I q) Heritage Reporting Corporation I (202) 628-4888 )
i
__-______-__-_ a
k I
32 l( };
~
l 1T thought about or haven't looked at at all'is going to be'the 2' dominant contributor to the core melt problem.
3 MR. MICHELSON: It might very well be.
4 DR. SIESS: Why do we bother with PRAs?
5 MR. MICHELSON: Hopefully they are going to tell-
- 6. -you what is going on.
7 Let's proceed.
8 MR. FLACK: Now I would like to discuss the five 9- potential issues and the response, the staff's response to 19 those issues.
il The first issue, fire-induced control system
+ .12 interaction.
13 Basically a study reported three types of 14 interactions that could occur, given a control room fire.
(
15 First is loss of control power for an operated 16 device prior to transfer. And this.could be due to a blown 17 fuse, for examp'e.
18 The second kind of interaction included spurious 19 actuation prior to taking control from the remote shutdown 20 panel. And this could be for example the opening and 21 closing of valves or the running of pumps through the hot
.22 shorts.
23 And the third being the total loss of component 24 function due to redundant control equipment failure. The 25 third would not meet the criteria of Appendix R if that Heritage Reporting Corporation (202) 628-4888
)
(:
33 1
[D w) 1 equipment was needed for safe shutdown.
2 So those are the three types of interactions.
l l
3 DR. CATTON: Didn't you miss one, where you have 4 to de-energize and turn off the reactor coolant pumps?
5 MR. FLACK: I don't know if the reactor coolant l
l 6 pumps -- it depends on the scenario.
7 DR. CATTON: I mean Oconee.
8 MR. FLACK: Oh, Oconee. I'm not familiar with the 9 Oconee scenario.
10 DR. CATTON: Gee, I would think that being in 11 Research and dealing with fires, that one of the few fires ,
12 that has happened you would have read about it.
13 MR. FLACK: Well, I did read about it.
14 MR. MICHELSON: You didn't bring it to their 15 attention.
16 MR. FLACK: I did not analyze'it to that level.
17 DR. SIESS: All three of these things are the 18 result of fire, is that what you are saying?
19 MR. FLACK: These are interactions that could i
20 occur if there was a fire in the control room, that were j
(
21 identified by the fire and scoping study. !
22 The scoping study then proceeded to do sensitivity i
23 studies based on these interactions and they found under the 24 two cases that with no protection against any interaction, 25 or they would occur, core damage frequencies could be above 1
() Heritage Reporting (202) 628-4888 Corporation l l
I h _ _ _ _ _ _ _ _ _ . _ _ _ _ _ . . _ _ _ . _ _ _ _ _ _ _ _
1: ,,\ 34 1 ..
1 .10-to the minus'4.
2 And with one method of shutdown available,_
3 sequences'could beiabove 10 to the minus 5. And for-
- 4. example,'the Army. analysis of LaSalle did have a sequence 5 ~that.was reported at 8.6 times 10 to the minus 6.
6 MR. MICHELSON: .These numbers, I want to kind of l
7 keep-them in perspective, now.
l 8 These numbers are with the simplistic models that 9 were used'or are these results'with some uncertainties 10 applied to get them down to this low a number, or just what?
11 MR. LAMBRIGHT: In the case of the first two 12 . bullets, a look'was taken at the four PRAs, that have re-13 evaluated as part of Test 2 of the Fire Risk Scoping Study.
14 In other words, Seabrook, Indian Point II, Oconoee and' 15' Limerick. And bounding analyses were done on those 16 . sequences looking at the specific assets of point 1, 2 and 3 j 17 on the previous slide. Ana the conclusions were that if, 18 for 1 or 2, if there.was no protection against some 19 potential interactions, there were some sequences that could j
.20. raise core damage frequency --
I 21' MR. MICHELSON: So this is what you would result l 22 in if you considered just the three items 1on the previous 4
23 slide?
.24 MR. LAMBRIGHT: That is right. However, in the 25 case of LaSalle, detailed analysis was done on all relay Heritage Reporting Corporation
( (202) 628-4888
35
(~
(_)% 1 control circuitry -- and in one subsection of electrical 2 distribution cabinet, a number of interactions were 3 discovered which did potentially compromise come functions 4 of the shutdown panel through total interactions, and if 5 those interactions took place, it could have an ~verall 6 effect of roughly 9E to the minus 6 per year on c..e damage, 7 or something like that.
8 DR. SIESS: Now, what is the significance of that?
9 That if you do it right you get 9E to the minus 6, but if 10 you don't do it right the first two cases give you much !
11 higher values?
12 MR. LAMBRIGHT: That is correct. In the case of 13 the 1E to the minus 4 value that is assuming that there are 14 some interactions --
.( }
15 DR. SIESS: That is just assuming it?
16 MR. LAMBRIGHT: Yes. And a bounding analysis to 17 determine the potential effect if there is no protection 18 against certain interactions.
19 DR. SIESS: Did LaSalle have that protection, or 20 was it just that that assumption was too conservative?
l 21 MR. LAMBRIGHT: In the case of LaSalle there were 22 a number of subtle interactions that I mentioned.
23 In fact one was a spurious actuation of a breaker 24 caused by a relay in an electrical distribution panel >
25 failed, and they had to relay the trip automatically that Heritage Reporting Corporation (202) 628-4888
36 f(/'Y) 1 was controlled'from the shutdown panel.
2- MR. MICHELSON: Now what caused the spurious 3 actuation'of these devices?
4 MR. LAMBRIGHT: It was a. relay --
S. MR. MICHELSON: No, no, I mean what is the --
6 MR. LAMBRIGHT: A fire in a particular. subsection.
7 MR. MICHELSON: Is it the heat from the fire or
~
8 something or the actual fire involvement'of the device?
9 MR. LAMBRIGHT: Fire within one subsection.
10 MR. MICHELSON: Was it a cabinet containing that-11 device? Is that what you mean?
12 MR. LAMBRIGHT: That's right. But the assumption 13 was also made in the case of the LaSalle that a fire would 14 not spread to any other subsections of a given cabinet.
{
I 15 MR. MICHELSON: Now, how did you determine the 16 response of devices exposed to fire in terms of their 17 ability to produce certain kinds of electrical responses'of 18 shorts and opens and other kinds of maloperations?
19 MR. LAMBRIGHT: Spurious actuation and other kinds 20 of --
21 MR. MICHELSON: No, not spurious. These are 22 consequential actuation, not just random. These are 23 occurring because the fire in the cabinet is heating the 24 cabinet, producing smoke in it, water is coming in because 25 you are mitigating it, CO2 or whatever you are using is
() Heritage Reporting Corporation (202) 628-4888
37
() 'l coming in. That is what the device sees. That environment.
2 And now~you have to somehow in your PRA predict how that 3 device responds to that environment.
4 MR. LAMBRIGHT: In the case of LaSalle, they had 5 the most detailed models to date, where a lot of 6 ' considerations were taken as far as circuitry was concerned
,7 but I can't claim that any potential interaction was - ,
8 MR. MICHELSON: Now, do you think the LaSalle 9 .model accounts'for the consequences of the environment that 10 is created within the cabinet when determining what its 11 failure probabilities will be and what failure modes it will 12 have?
13 MR. LAMBRIGHT: Not totally. However, --
14 MR. MICHELSON: You say not totally. Well, tell 15 me how it does it even partially.
16 MR. LAMBRIGHT: It takes, for any given accident 17 sequence, what the worst state would be for a given contact 18 or relay with a given cabinet.
-19 MR. MICHELSON: What the worst state would be.
20 ~What do you mean by the worst state? I don't know. It 21 depends upon what is happening to me as to what the worst 22 state would be. What else is happening in the' plant at the 23 time determine what the worst thing &'-t this device could 24 do might be.
25 PRA has all this in the model, hopefully. But,
() Heritage Reporting (202) 628-4888 Corporation L 1 I
l
_ _ _ _ _ _ _ _ _ _ . - - - _ _ _ ___ -_ _ _ _ _ _ __ ___ _ _J
38 m)
(_ 1 when you get to the model and you go to-a particular 2 component, how detailed)',e 'his model? Has it modeled in 3; branches that show how the component could potentially 4 respond and what the probability of response in those ways 5 might be? Because this is what fire does. It gives you 6 'some unusual responses of devices.
7 MR. LAMBRIGHT: I agree. In the case of LaSalle, 8 since it was the most detailed model, modelled for control 9 circuitry or modelled down to the contact pair level, so it 10 was the best available model --
11- MR. MICHELSON: Okay. When they modelled it to 12 the pair level, did they now examine the consequences of a 13 pair either opening or closing? !
(} 14 MR. LAMBRIGHT:
}G1. MICHELSON:
Yes, that is correct.
15 And what the consequences of what 16 that opening or closing might be in terms of equipment that 17 turns on or turns off or changes some state?
18 MR. LAMBRIGHT: That~is correct.
19 MR. MICHELSON: Okay.
r 20 DR. SIESS: May I return to my question?
1.
21 MR. MICHELSON: Sure.
22 DR. SIESS: Item i seems to say, and correct me if
?3 I am wrong, that if I assume there is no protection against 24 interactions, which means that they would occur, I get 1 25 times 10 to the minus 4.
Heritage Reporting Corporation (202) 628-4888 1
39 MR. LAMBRIGHT: -That is correct.
'l 2- DR. SIESS:' LaSalle is two orders of magnitude 3 lower than that.
4L MR. LAMBRIGHT: Roughly an order of magnitude.
5 DR. SIESS: I'm'sorry. One order. Is that 6 because LaSalle was designed with protection against 7 interactions?.
8 MR. LAMBRIGHT: Yes, they do have a remote 9' shutdown panel.
'10 DR. SIESS: And is that unique to LaSalle?
11 MR. LAMBRIGHT: No.
12 DR. SIESS: So Item 1 does not apply to any plant 13 you know about?
14 MR. LAMBRIGHT: It would apply to any plant that 15 has not been adequately designed against potential 16 interactions.
17 DR. SIESS: But you don't know which plant that 18 is?
19 MR. LAMBRIGHT: No.
20 DR. SIESS: Okay. If we knew which one it was, we 21 could shut them down, I guess.
22 MR. LAMBRIGHT: The intent of this was to assess 23 what the potential effect would be, given no protection l
24 against interactions.
25 In the sense of the PRAs we had to work with, we Heritage Reporting Corporation (202) 628-4888 L
L - - --- __- _ - - _ _ - - _
b 40-
~
i '. 1L didn't have, for the four.PRAs, we didn't have.models.
2- -DR. SIESS: Yes, but I can do that on'anything,
~
3 can't I? I can go inJand say gee, if this plant- wasn't 4> designed for an earthquake, the' core melt probability would' 5 be X times 10.to the minus Y or if-it wasn't designed.for'a 6~ tornado. Is that a proper way to go about it, just to make 7 an assumption to see.what is'the worst thing you can think, 8 of?
9~ MR. FLACK: I think what the study is doing is it 10 is showing the importance of the remote shutdown panel.- It 11 is saying if you did not have one you would be looking at 12 that kind of number, and therefore it is an important 13 component in the fire analysis.
14 DR. SIESS: Now, this doesn't say with no remote 15 shutdown panel. It says with no protection against 16 interactions.
17 MR. FLACK: Which would be essentially the same 18 thing.
19 MR. LAMBRIGHT: That's correct.
20 DR. SIESS: Interactions then do not involve the 21 remote shutdown panel?
22 MR. FLACK: No. If they did, and they failed at 23 remote shutdown, what we are saying is it would be similar 24 to if you did not have the panel to begin with.
25 MR. MICHELSON: Well, you are really going to have O rie 9- verei 9 cerror tie-(202) 628-4888
t 41 :
() 1 to prove in the process of doing your PRA that there aren't 2 any interactions between the fire in a given area and the 3 capability of the remote shutdown panel to function 4 properly.
5 MR. FLACK: That's true.
6 MR. MICHELSON: That doesn't come automatically 7 because you got one. It is because you have done the 8 analysis, that shows that even though you have a fire in a 9 given area and it spreads and the smoke goes around and the 10 heat goes around, it still does not prevent the operability 11 of a certain group of safe cold shutdown equipment. That is 12 what you have to show.
13 MR. FLACK: That's correct.
~N 14 MR. MICEELSON: And you would show it because some (O
15 of the equipment gets damaged but the damage does not 16 propagate to this other thing you have put in. This remote 17 shutdown panel still is able to overcome or bypass or work 18 around the damage enough to keep a certain complement of 19 equipment functioning. That is all you are trying to do.
20 MR. FLACK: That's right.
21 bM. MICHELSON: And it's got nothing -- System 22 interaction has to be looked at just as carefully as LaSalle 23 with a shutdown panel as with any other plant without a 24 shutdown panel. You have to make sure that whatever you 25 have put in works.
Q
(,,/ Heritage Reporting Corporation (202) 628-4888 e _ _ _ _ -
42 m
() 1 DR. SIESS: I thought all of them had a shutdown 2 panel.
3 IGt. MICHELSON: Of some sort, some more effective 4 than others.
5 DR. SIESS: If you had gotten 10 to the minus 7 6 under one, would you just have stopped there?
7 MR. LAMBRIGHT: Well, if you would write that 8 using a bounding analysis of 10 to the minus 7.
9 DR. SIESS: 10 to the minus 8.
10 MR. LAMBRIGHT: Compared to other things that 11 fires have caused in and of themselves with a frequency of 12 10 to the minus 5 then you would think this issue would be 13 less significant than some of the other fire issues.
14 DR. SIESS: That is about as much of an answer as
(~)S
\.
15 I guess I can get. That's no answer, though.
16 MR. FLACK: Okay. The recommendations that were 17 made as part of the fire risk scoping study with regards to 18 fire-induced control systems interactions were:
19 One, to develop standardized guidelines for 20 analyzing fire-induced control system interactions; and 21 Two, to review the adequacy of remote shutdown 22 implementation practices using these guidelines.
23 DR. KERR: What is meant by standardized 24 guidelines?
25 MR. FLACK: I envisioned it as being sort of a
() Heritage Reporting Corporation (202) 628-4888 l
l
)
I L.
a<
- 43. I es .
.1? checklist,.to the best one.can conceive it to be.
(J 1
2 MR. MICHELSON: -There is a little' nomenclature 3 problem here. Control' systems to-son.e. people mean the'non-14- st 'ety related systems, because those.are commonly called- g 5 control as opposed to protection, q 6~ What-do you mean in.this case about, what do you 7 mean when you talk about control system interactions? Are 8 you talking about the interactions amongst the non-safety l'
9 related control arrangements?
10 MR. FLACK: If it had an impact on the safe 11 shutdown --
12 MR. MICHELSON: Well, let me ask, are you looking 13 at the.effect of safety-related control arrangements, which 14 are not called' control systems, they are called normally 15 protection systems.
16, MR. FLACK: Well, I have actually summarized that 17 title, by it should really read control room remote shutdown 18 system.
19 MR. MICHELSON: Okay. You really mean safety 20 grade stuff as well-as non-safety. Okay. But it is that 21 funny nomenclature. We struggled this morning through it, 22 because control and protection are kept separate, and they 23 have entirely different meanings.
24 MR. FLACK: I apologize for not using the entire 25 heading.
() Heritage Reporting (202) 628-4888 Corporation
p l
[ ,
t v 44.
l- Tha Staff action at this' point is to consider the-
)
.2 . issue with3n the IPEEE framework. And it has also been"
~
3- raised as a potential generic-issue. That is being'.
4- 'prioritized at'this moment.
'5 DR. SIESS: -John,.is the IPEEE going to be carried-6 out to that level of detail? l 7 MR. FLACK: At this. time,.we are still considering~
l 8 what is to be included within the IPEEE.
9 My personal opinion is no. But I don't know to 10 what detail we would be looking att it. This is still'being 11- discussed.
12 DR. SIESS: The kind of thing that we talked about.
13' on the remote shutdown at LaSalle, for e:: ample, where it was 14 a very detailed event tree or whatever, suggests that the -
15 IFEEE would have to be done with the PRA.
' 16 - MR. LAMBRIGHT : - To do a vital area analysis of'a 17 control' room cabinet or a vital area analysis of any 18 particular. area using the fault trees and event trees 19 developed by internal events is absolutely essential in 20 determining whatever the fire risk is at any given point.
21 DR. SIESS: So to include fire in the IPEEE would 22 mean PRA.- They would.have to have the PRA which they could 23 then use for the IPEEE.
24 MR. FLACK: Well, like I said, this is still being 25 discussed, which is the reason for raising it as a potential Heritage Reporting Corporation (202) 628-4888 j
45
~
i) 1 generic issue. It may address some aspects of'it,-but not 2 all ofLthem. ,
3 And you're right. It.will_probably likely be some 1
4 probabilistic studies.
5 MR. MICHELSON: Does the IPEEE require though a 6- PRA? Not talking about fire. The IPEEE process. I thought
- 7. part of that process was going to be a plant-specific PRA?
8 MR. FLACK: For the internal events. The internal 9 events at least at level one.
10 MR. MICHELSON: Yes, for' internal events. I 11 MR. FLACK: Yes.
l 12 MR. MICHELSON: It is clear everyone will have an '
13 internal event PRA. Is that right?
14 DR. SIESS: They still allowed the incore i 15' methodology, didn't they, as an option? Or has that been .j i
16 changed? i 17 MR. FLACK: It's been enhanced.
18 DR. KERR: The generic letter has not yet been 19 approved, so we don't know --
1 20 MR. MICHELSON: Well, I was just trying to find 21 out where they are at.
22 DR. SIESS: The last one we saw permitted either ;
I 23 method. ]
24 MR. MICHELSON: So if they decide not to use a 25 PRA, then they will have to come in and explain how they do 1
() Heritage Reporting (202) 628-4888 Corporation l
46
- h. 1 all.these things without a'PRA.
2- MR. McCRACKEN: Yes. Generic Letter- 89,. 88.20.
3- Generic Letter 88.20 allows them three options: a level one 4 PRA with certain additions, incore with certain additions,.
5 or another method as you determine and the' Staff reviews.
i l 6: DR. KERR: Now, 88.20 has not yet been released, 7- has it?
8 MR. FLACK: The generic letter has been released.
9 DR. KERR: It has?
10 MR. FLACK: Yes. It's out.
- 11. MR. MICHELSON: Now, when they do these under any 12 one of those options, do you have to~ include external ~ events -
13 in your consideration?
14~ MR. McCRACKEN: No. External events is to come 15 later.
16 MR. MICHELSON: That decision will be a separate l 17 instruction?
18 MR. McCRACKEN: Right.
19 DR. SIESS: What they are going to do for external 1:
20 events has not yet been settled. It does not mean that the 21 external event IPEEE will come after the internal event 22 IPEEE. It's just that the Staff is deciding sequentially.
23 MR. McCRACKEN: The licensees were told if they 24 want to do internal on their own, now, go ahead and we will 25 review them individually.
() Heritage Reporting (202) 628-4888 Corporation
--__________-----.__.__._____._____m_ _ _ . - _ _ _ _ _ _ _ _ _ . - - _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ - . _ _ _ - _ _ _ - -
2 __. ___________m.______
lJ
.4y.
, 1 MR. FLACK: The external events?
2 MR. McCRACKEN: The-external events. Yes.-
3 okay. The next issue is smoke-control and manual-4 fire-fighting. effectiveness.
5 The primary fire source in nuclear power plants is 6 ' lubricating oils and cable insulation, both of which 7 generate a lot of smoke.
l 8 The impact of smoke is pointed out in the fire-l l
9 risk scoping study, in the following areas:
10 Reduced manual fire-fighting effectiveness; 11 Cause misdirected suppression efforts; 12 Damage electronic components; 13 Cause evacuation of control centers; 14 It could initiate automatic suppression systems
)
15 away from the fire; 16 And it could potentially impact an adjacent unit.
17 The recommendations that came out of the study 18 were two:
19 First was to review further plant practices and 20 investigate the probable linkage between these plant 21 practices and brigade effectiveness.
22 And the second recommendation was to refine 23 methodologies for us in the evaluation of fire detection and 24 suppression systems adequacy.
25 The Staff action at this point is again, to Heritage Reporting Corporation (202) 628-4888
48 1 consider the issue within the IPEEE framework and it has 2 also been raised as'a potential generic issue.
>3 The next issue is adequacy of fire barriers.
~
4 The study pointed out that high barrier 5 reliability is essential in keeping the' risk from fire-6 propagation low. This is due to the number of barriers that 7 are needed to prevent going to core. melt should the barrier l 8 fail. It is in the neighborhood of 20, I believe, was 1
9 reported in the study.
10 The reliability was questioned, specifically with 11 respect to the penetration seals.
12 There was only one_ recommendation that-came out, 13 and that.was to assess the actual performance reliability of
{} 14 15 the 0.S. fire barrier elements. And this was, the concern there was the w6y the fire barrier elements were tested, 16 under pressure, or prescure differential across the barrier 17 versus non-differential pressure across the barrier.
l 18 MR. MICHELSON: Before you leave that slide, did 19 you think any more or determine anything more about the 20 question of the ability of fire barriers to withstand 21 elevated pressure differentials?
22 In other words, the fire heats up the room and
. 23 pressurizes it to some extent and that creates a 24 differential pressure across the barrier and that has to be 25 added to whatever the effects of fire are on the barrier to
() Heritage Reporting Corporation (202) 628-4888
j d
I 49
- 11 determine'how'long it'will last. And I am not sure this 1( )
.2 .typeLof pressurization was done during the barrier tests, 1
3 and the question is now what do we do'about it.
-4 MR. FLACK: That issue is still open. In fact,
~5 that is one of the --
6 MR. MICHELSON: Is that a part of your adequacy of 7- fire barriers, though?
8 MR. FLACK: Well,'actually, the potential generic 9 issue that came out has two parts to it.
10 .One exactly addresses that concern. That is, the 11 pressure differentials across the barrier, and how they are 12 tested.
13 The other aspect is the barrier integrity, and
('j 14 that addresses concerns regarding penetration seals were
- (_/
15 installed properly, et cetera.
16 IGR. MICHELSON: Of course this is all to dacide 17 whether we have a generic issue for consideration and 18 resolution three or four years from now. What do we do.
i9 about today's plants in the next generation which will 20- already be well into design and may even be certified before 21 that resolution ever comes out? What do we do about the 22 differential pressure question? It is just one of quite a-
- j. 23 few here-by the way but it is an important one, and 24 apparently at least one other country elected to do things a 1
25 little differently in this regard.
() Heritage Reporting Corporation (202) 628-4888
iL m
50
/ 1- MR. FLACK: That is' correct.
2 MR. MICHELSON: And now why are we doing and 3 continuing-to do what we think is acceptable and why,do we
- 4 think:it.is. acceptable?
5 I thought Research was going to churn something up-b 6 :o:try te'get a. resolution and not just pass it on to yet 7 :another_ generic-issue.
8 MR. FLACK: Well, there are.two aspects.to the-
- 9. issue. -The one issue, I do repeat, it has been raised as a 10 gener'o i issue, and it has been a controversial issue for 11 several years now.
'12 And at this moment, there is no research going on 13 in'that area.
- 14 MR. MICHELSON: In any area of fire protection, as 15 I understand it.
16 MR. FLACK: That's right.-
i 17 DR. CATTON: Is some sort of a standard test'used?
18 MR. FLACK: Yes, they do have standard tests for.
19 testing it.
20 DR. CATTON: Who sets the standards for the tests?
21 MR. FLACK: Do you know that offhand, John?
22 MR. LAMBRIGHT: I'm not aware what's done exactly.
23 MR. FLACK: There are qualification stat.dards that 24 are act. I believe it's by the National Fire Protection 25 Association?
Heritage Reporting Corporation (202) 628-4888
_ - - _ _ _ ___ _ __-___ _____-_______-_ D
l:)
I 51
() 1 MR.'McCRACKEN: They are in FPA standards on 2 testing fire barriers in the U.S. The barriers at nuclear 3 power plants are tested consistent with those st&ndards.
4 DR. CA " TON : And they allow --
5 MR. McCRACKEN: They allow you' testing without 6' excess presrure.
7 DR. CATTON: So the pressure is the same on both 8 sides?
9 MR. MICHELSON: No. Not quite. It's negative on 10 ~the side of the fire.
11 MR. McCRACKEN: Due to a fire you can either have 12 a negative or positive pressure on the side where you have 13 the fire.
14 MR. MICHELSON: The test, though, is done such 15 that they routinely end up with slightly negative pressure 16- on the fire side because of the exhausting of the heat and 17 smoke from their test apparatus.
18 DR. CATTON: They pump it out.
19 MR. MICHELSON: Yes, they pump it out, sure.
20 DR. CATTON: At a little faster rate than it ought 21 to be pumped out.
-22 MR. MICHELSON: Yes. One of the question is,
,3 2 doors is a fire barrier. To what extent is a door a smoke l 24 barrier? Do we allow cracks under the doors? I think you
-25 can have a crack under the door under the fire code as long O rie e- gerei e cereer ele-(202) 628-4888 l.
_ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ __a
52
() -1 as you pass the three-hour test.or whatever rating you want 2 for it.
3- But smoke certainly comes through cracks, as I 4 understand it.
15 DR. KERR: I thought from the recommendations that 6 you see moving around in a fire that the smoke tended to go-7 to the top of the room.
'8 MR. MICHELSON: It also tends to come:under.the 9 doors. That is why you put the towels on around the' door of 10- your hotel room, to keep it out of your room for a while.
11 DR. CATTON: That's when the hall is filled with 15 smoke.
13 MR. MICHELSON: This is the room, though,.where 14 the fire is, it is filled with smoke. I think we have seen 15 enough testing to prove that.
16 Well, that's another question on the_ adequacy of
~
17 barriers. What keeps the water in the suppression syst' ems 18 running? The water is running under the door.
19 DR. KERR: Is there some evidence that there is 20 likely to be a good bit of pressure buildup unless a room is 21 sealed?
22 MR. MICHELSON: Unless the room is sealed. That's 23- how, the better the fire area is sealed, of course, the 24 faster the pressure rises.
25 The only thing keeping the pressure down in the O erie ee aererei e cereer eie-(202) 628-4888
53 u
f) ll- fire area is the leakage out in all directions. And if you L :2 build a real tight room, then the pressure rises are very' 3 significant. This is why the British put the chimneys on 4 the size well and put the blowout panels to the chimney.
5 The British have, you know, you have read the size well bit,.
6 they have put the chimneys in the blowout panels to keep the 7 pressure down. None of the plants in this country,.to my 8 knowledge, have even done the analysis to determine whether -l l
9 a chimney would be needed. .j 10 DR. CATTON: But if you seal the room tightly and ;
11 the pressure goes up, don't you burn up all the oxygen? ,
I 12 MR. MICHELSON: If you don't blow out the i 13 penetrations first, yes, you eventually certainly would 14 -consume the oxygen, if there was no in-leakage. Sure. Yes. .
15 'But some of these things-supply their own oxidants. Some of ,
i 16 them supply their own oxidants.
17 Read the size well and you'll see how they came to j
~18 that conclusion.
l 19 MR. FLACK: The next issue is equipment survival ]
l 20 in fire-induced environments.
21 The fire risk scoping study actually did more of a 1
22 qualitative evaluation because of lack of fragility data and ]
23 analytic tools to treat equipment survival. j 24 They did report it was very plant specific. There j 25 was some interesting data they did accumulate during the 1
() Heritage Reporting Corporation (202) 628-4888 l
I L 1 1x _ _ _ _ - _ _ _ _ _ _ _ _J
, 54
(} 1 ' study.
2 The first thing is the frequency of inadvertent 3 actuation which they found to be high, .14 per' year,_ and 4 Lthat one'out of every four events either induced or resulted
- 5. from a plant t'ransient.
L 6 25 percent of the events resulted in degradation 7 of one or more safety systems.
8 DR. SIESS: Is that the reactor year?
9 MR. FLACK: The 1.4 a year? I believe that was.a 10 reactor year.
11 DR. SIESS: That's per plant per year.
12 MR. FLACK: Per reactor year. That's a generic 13 number.
14 MR. MICHELSON: Of course that number is kind of' 15 irrelevant to the real problem, which is the initiator-is 16 the fire and now the consequential actuation is what you are 17 worried about, not random inadvertent. That number there is 18 random, true random. That is when something sets off in the 19 plant all by itself. That is irrelevant to fire. It is 20 already big.
21 And now what the bottom line says, if you do, if 22 the fire indeed does cause an unwanted actuation of fire 23 protection, there is a high probability it will damage 24 equipment.
25 DR. SIESS: Is there a difference between whether Heritage Reporting Corporation (202) 628-4888
55-
^O 1 it's ene er mere 2 2 MR. FLACK: There certainly is.
3 DR. SIESS: That doesn't tell me very much. If it 1
4 is only one system, it is not likely to cause much of a 1
5 problem. If it is more than one system, it is. Right?
6 MR. LAMBRIGHT: More likely, from a core damage 7- frequency standpoint, the more safety systems that you L 8 compromise with an inadvertent suppression or a fire-induced 9 suppression event.
10 DR. SIESS: Yes. That 25 percent can't be broken-11 down any further?
12 MR. MICHELSON: My understanding is that none of 13 the events were fire-induced.
'14 Okay. Fire gets the first system clearly. One
[
15 actuation gets the --
16 DR. CATTOM: -- tied into the fire circumstance.
17 MR. MICHELSON: That's the one of real concern.
18 DR. SIESS: Maybe it would help if we were both 19 talking about the same thing.
20 I'm looking at a thing that says frequency of 21 inadvertent actuation. Inadvertent actuation, I think I 22 know what it means. And that is .14 per year. 25 percent 23 of those were either induced or resulted from plant 24 transients, so I'm now down to .025 per year and 25 percent 25 of those --
) Heritage Reporting Corporation (202) 628-4888
V '
, 56-L/~T
~
1 DR. CATTON: No , I don't think he means'that.
D
,2: DR. SIESS: It's' bad enough to change from. 14 to 3 1,out of 4 to 25 percent without confusing me further by
. 4 havingftwo things meaning the same thing. And~there's three-5 ' ways of representing ratios up there, and you got them all.
'6 (Laughter) 7 DR. SIESS: And those that result in one' safety 8 system degradation are well within the design -- if it is.
9 more than one, I'm beginning to get interested.in it. So I 11 0 just don't knOw what's up there. What does the 25' percent
~11 refer to? 15 percent of'the .14?
12 MR. FLACK: That's right. 25 percent of the .14.
13 DR. SIESS: So that is two separate items.
14' MR. FLACK: That's right.
15 MR. McCRACKEN: What he is telling you is that r
16 once in every 30 years you are going.to have an-inadvertent 17 suppression system initiation.that will result in a trip or 18 transient and loss of one or more systems.
19 MR. FLACK: Let me move on. There were some other 20 aspects of fire environment involving smoke and low level 21 thermal exposure. But there was lack of data to quantify 22 these contributors.
23 The recommendation stemming from the study was to 24 perform a more extensive review of the experience base in 25 order to better evaluate the issue.
Heritage Reporting Corporation (202) 628-4888
=
l 57
()
i 1 The Staff action at this time is we put this data 2 into Generic Issue 57 and it is being studied.
3 MR. MICHELSON: Now, what do you think Generic 4 Issue 57 is covering? What do you think it includes? By 5 title it says inadvertent actuation, I think, of fire 6 protection systems, something to that effect.
7 MR. FLACK: Yes, that's right.
8 MR. MICHELSON: It has nothing to do with fire-9 induced actuation. It is an entirely different issue. The 10 issue there in 57 is how safe is the plant from inadvertent 11 actuation caused on a random basis. And it works through 12 the exercise and finds out how important it is. And it 13 probably is not all that important.
f/3 14 What it does not deal with, at least as far as I U
15 know, ~it doesn't deal with fire as the initiating event 16 causing as a consequence the actuation of several different 17 fire protection systems, and their subsequent consequence.
18 DR. SIESS: Is the fire causing inadvertent --
19 MR. MICHELSON: It is not inadvertent anymore, at 20 least in my opinion, it is consequential.
21 DR. SIESS: As opposed to --
22 MR. MICHELSON: No, no, not outside the fire zone, 23 they better not.
24 DR. SIESS: That's what I'm getting at.
25 MR. MICHELSON: Yes, consequential action.
() Heritage Reporting Corporation (202) 628-4888
58
,( ) 1 DR. SIESS: It is inadvertent because it is 2 outside the --
3 MR. MICHELSON: Well, it is unwanted if you want 4 to use -- some people use inadvertent in the same way as 5 random.
6 DR. SIESS: Oh, no.
7 MR. MICHELSON: And it's not. Well, 57 is not 8 dealing with inadvertent. It is dealing with random, as I 9 understand it. Not as a part of a scenario.
10 MR. LAMBRIGHT: Those same analysis techniques 11 that are used for fire analysis are directly applicable to 12 looking at this issue as well.
13 MR. MICHELSON: Except they aren't directly
(~} 14 applicable because they don't include the migration of heat v
15 and smoke in the models.
16 MR. LAMBRIGHT: Well, in the sense that the fire 17 suppression code pumper can't handle that.
18 MR. MICHELSON: That's right.
19 MR. LAMBRIGHT: But he can take a look at data to 20 determine, since the generic probabilities of --
21 MR. MICHELSON: You can't look at data if you have 22 not had any experience. You don't have any data, with fires 23 and bow the heat and smoke migrates in a nuclear power 24 environment. So you don't have anything to look at, except 25 you can speculate.
n
(_) Heritage Reporting Corporation (202) 628-4888
59
-( ) 1 Now, we need something to look at.
2 DR. CATTON: That's right.
3 MR. MICHELSON: We need some calculational 4 techniques and we need some experiments. We are not getting 5 any of that. We are just getting lip service.
6 DR. CATTON: There might be some calculational 7 tools that you could use other than COMPBRN that might be 8 helpful.
9 But I think Carl is right. If you just do a 10 review of the experience base, and you stay within the 11 nuclear business, you are going to have not. ting when you are 12 done, or very little.
13 MR. LAMBRIGHT: I think it calls for looking at
{} 14 nuclear as well as non-nuclear data.
15 DR. CATTON: I am not sure -- what you need to 16 know is, if a fire is in that corner, how much is it heating 17 the equipment up at this side?
18 MR. LAMBRIGHT: That is correct.
19 DR. CATTON: When the temperatures are still 20 moderate.
21 I am not sure you are going to find that kind of 22 data.
23 MR. LAMBRIGHT: I am not sure exactly what data is 24 available at this point to address this issue.
25 However, certain groups do have --
) Heritage Reporting Corporation (202) 628-4888
.) .
E if - 60-
[ 11 DR. CATTON:' I think'you are' going to have'to.
2 marry;the data. search with a better computational.-tool'than
- s .
.3 .' COMPBRN to get~what you.need.
- . 4
- y. 5.
6 7..
.. 8 9
10
'11
- 12.
13" O 'i 15
.16 .
17
.18 -
19
'20 21 22 23 24 25:
l Heritage Reporting Corporation (202) 628-4888
I
'I 61:
q(_f ~
1 MR. FLACK: .The last'l* ** La seismic fire 2 . interactions.
3 'Again, the study did not quantify specifically the 4 contribution of fire seismic interactions of the core damage.
5 frequency. But it did point out that these interactions 6 could be eliminated by performing plant-specific-
-7 evaluations. And that the study did provide a list of 8 potential seismic. fire interactions.
9 The recommendations: what it defined in-document 10 potential vulnerability is concerned to nuclear power 11 plants. And the Fraff action is to work this' issue.out 12 within the IPEEE framework.
13' And then utilize the scope of study for
{} 14 information performing that study.
15 DR. SIESS: You don't really mean that the 16 interactions could be eliminated by. performing plant-17 specific evaluation.
18 ' MR . FLACK: No, it would be within the framework 19 of the evaluation 20 DR. SIESS: Evaluations will not eliminate 21 anything.
22 MR. FLACK: No, not at all.
23 DR. SIESS: It would have to change the hardware.
24 It would have to be plant modifications made.
25 MR. FLACK: That's right.
Heritage Reporting Corporation (202) 628-4888 I
- )
62-1 MR. MICHELSON: ' Whet islthe. approach that.one
]}
2 should'use in doing these evaluations when it comes to such 3 matters as the seismic fragility of the non-safety' fire
- 4. protection systems, since it'is a non-safety system.
5 Sometimes it's qualified at least to stay in place. It's-6 not qualified to operate.
7 In fact, because it may operate during and 8' following an9 earthquake. It may set itself off particularly 9 if it used mercury switches, for instance, which' fire 10- protection people seem to like. These are quite' fragile 1:L when it comes to even relatively modest earthquakes.
12 What do we do in using those kinds cf equipment in 13 places like diesel rooms where we may fill all four diesel 14 rooms of CO2 as a result of the modest earthquake?
15 MR. FLACK: That's a potential concern because of 16 the fact that it's going to the potential blackout that 17 could only.be mitigated by bringing offsite generators.
- 18. MR. MICHELSON: It just blows my mind as to where 19 we're heading in this or what are we doing in a tangible 20 fashion. We're doing a lot of report writing which is good.
21 We are understanding the problem better, which is good. But.
22 what are we doing? What's going to happen?
23 It isn't going to do any good unless it gets some 24 real information and guidance to work with.
25 DR. SIESS: You don't think they could look at
/ Heritage Reporting Corporation (202) 628-4888
( '1 their diesel generator halon --
2 MR. MICHELSON:. Do you think they can dream up'---
t 3 the problem is they can't dream up -- if they're. going to do:
4 a fire PRA to-determine their vulnerabilities which might'be U ,
5 one way. .They're going to have to dream.up the techniques,.
6 the data base andiso forth, which to'do a real fire PRA.'
7 The IPE per se doesn't do that.
8 DR. SIESS: That's what it is supposed to do.
9 MR. MICHELSON: No,[it's just.a program. The 10 . tools have to be provided for the program which research 11- could. provide or others could provide.
12 MR. FLACK: To summarize the presentation: first, 13 we have no proposed fire protection research at this time.
14 The.following will be considered in the IPEEE.-
(} ~ ^
15 Control room remote shutdown panel interactions Smoke 16 control and manual fire fighting effectiveness. Seismic.
17 : fire' interactions. And the adequacy of fire barriers..
.18 Three of these have been identified as potential 19 generic issues. And the information coming from equipment' 20- survival is-being integrated into the Generic Issue 57.
21 MR. MICHELSON: What do you do in IPE with 22 potential Generic Issues or Generic Issues now being worked 23 on? What does IPE do with that sort of thing since it 24 doesn't know how it's going to come out. It doesn't have 25 any guidance from it.
Heritage Reporting Corporation (202) 628-4888
r >
64 1- What does this mean?
}{~ f O MR. FLACK: Well, I can only speak for the IPE for 3 internal events.
4 1G1. MICHELSON: Well, this is external' events that 5 we're really interested in at this meeting.
6 MR. FLACK: Well, there are actually two' 7' independent tracks: one is the Generic Issue and one is the 8 IPE.
9 Now, what is done in the IPE may be adequate to
~
10 resolve'the Generic Issue; and therefore we don't have a 11 Generic Issue.
12 Another alternative is the Generic Issue is not 13 high priority that it need not be addressed and that would come out of the'prioritization.
O.1415 I see the Generic Issue as a formal way of 16 resolving the issue and being on the record that it is 17 resolved and that information is'not lost. That to be 18 independent of the IPE.
19 MR. MICHELSON: Is the IPE guidance going to 20 really try to define what you expect eventually the Generic 21 Issue resolution will be so that they can build it in ahead i
22 of time; is that what you're saying?
23 MR. FLACK: Well, I think if we ask it properly in 24 a submittal guidance it could go a long way to addressing 25 the issue. It's ear 1y to decide.
Heritage Reporting Corporation (202) 628-4888
65
(_,) 1 MR. MICHELSON: Yes, we haven't seen any of this 2 guidance yet. It might be that good. If you follow the 3 guidance whenever we get around to resolving the issue you 4 will be in good shape. If you've got that kind of guidance ]
5 now that would be commendable.
6 DR. SIESS: Is it the IPE or the IPEEE?
8 DR. SIESS: What kind of guidance do we have for 9 the IPE?
10 MR. FLACK: The IPE is a submittal guidance which 11 the utilities will follow in submitting their findings to 12 the Staff.
13 DR. SIESS: What's the form of that? Is that the
()
. (_./ -
14 Generic Letter or NUREG?
15 MR. FLACK: That's NUREG-1335, it should be issued 16 by tho end of the month.
17 DR. SIESS: Okay.
18 MR. FLACK: That concludes my presentation.
19 MR. MICHELSON: Thank you.
20 Any questions? Before you leave, John, you better 21 see if we've got any questions.
22 MR. FLACK: Sure.
23 MR. MICHELSON- I ou. ..er.c , ev 1 guess you're 24 free.
25 DR. SIESS: Let me ask one. I'm afraid I'm not p
(_) Heritage Reporting Corporation (202) 628-4888 1
1 i
( a 66
!" asking the right person. Do you see a difference between
(.( l' 2 the amount of information you would have to have, say, it j 3 was all to research or some other source of information, to 4 do a fire PRA and what you would have to know to make a 5 plant safe against fire?-
~
6 MR. FLACK: You mean identify some vulnerability?
7 I have none.
8 DR. SIESS: Anybody else want to answer that?
9 Because I've heard not so much here but in another 10 meeting, a lot of emphasis on what you need to know to do a 11 fire PRA, as if that were the end in itself. Obviously, 12 it's not. The objective is a plant thet's safe against fire 13 or safe enough against fire, I won't deal in absolutes. And f g. 14 .I was just' wondering if you thought about research'in terms
- G 15 of what you need to know for one thing-versus the other.
16 MR. MCCRACKEN: My hope is that by the end of the 17 next two hours I will have answered that question.
18 MR. MICHELSON: The next two hours.
19 This might be a good time'to.take a break then 20 before we get into our next two hours. Let's come back at 21 10 minutes after.
22 (Whereupon, a recess was taken.)
23 MR. MICHELSON: Go ahead.
24 (Slides being shown.)
25 MR. MCCRACKEN: The second presentation this
() Heritage Reporting Corporation (202) 628-4888
67
()
- ,m 1 afternoon is going over what the NRR and Research group are 2 doing primarily in the area of IPEEE or fire. How we intind 3 to search for vulnerabilities, the methodology, and the 4 basis for using the methodology that we intend to go to.
5 And part of that will require that I go through a little bit 6 of background on where we have been in fire and how we got 7 where we are and what the rules and regulations are that are 8 driving us to go the direction we are going.
9 I am providing this as a status teport because 10 we're involved in a lot of ongoing meetings. We have had 11 about a half a dozen meetings with NUMARC and EPRI in trying 12 to come to a resolution of how we're going to address the 13 issues for a fire vulnerability search at the individual 14 plants.
f]
\ ,/
15 The last of those meetings are on Monday this 16 week. The next one is scheduled sometime in the middle or 17 the third week in September.
18 I think the first thing you have to discuss when 19 we're looking at fire vulnerability and where we're going in 20 this area is where we are today with fire protection nuclear 21 power plants. And that is, status and compliance with the 22 regulations.
23 Appendix R inspections have been conducted at all 24 applicable plants. The findings at almost all plants were 25 that they were in compliance with minor exceptions. There rm
() Heritage Reporting Corporation (202) 628-4888
'68 I were exemptions written. But basically, they were worked
-- w f [--
2' on.
3 However, a few plants did such a poor job of-4 Appendix R that we're going back and doing a complete-5 reinspection, and these are in the process of being taken
-6 care of probably by early next calendar year we will have 7 those finished.
8 For the plants-that did not come under Appendix R,-
9 they had pre-operational inspections conducted and they were 10 required to comply.with basically the Appendix R 11 requirements, but-these became part of the.SRP. And we went 12 through those on each plant before they're permitted to 13 start.
14 Now we take all plants and they're'under'a
- 15. tri-annual. audit. program that requires us to.go back and-16 look at the status of their fire protection and what they 17 have done since the last time.we looked at them. They, 18 themselves, are required to do independent audits where they 19 bring in an independent organization to review their fire i 20 protection program to see what happened.
-l 21- In summary in this issue is, all plants currently l 22 meet the regulations or will do so shortly. And plants that 23 are found not to meet the regulations will be required to do 24 so.
25 DR. KERR: Is there evidence that the j
() Heritage Reporting Corporation (202) 628-4888 i
a_____________ _
69 1 implementation of Appendix R appreciably decrease core melt
([ }
2 frequency?
3 MR. MCCRACKEN: Yes.
4 That was one of the questions on the attachment to 5 your agenda that I was going to hit later. That was some of 6 the work that had been done by the Sandia study. But 7 basically, with Appendix R implemented they saw a reduction 8 of about, in order of magnitude in core melt frequency due 9 to those modifications.
10 MR. LAMBRIGHT: Sandia pointed to specifically, 11 there is an order of magnitude reduction in fire induced 12 core damage frequency. In the case of Lemrick it was 13 roughly a factor of three.
14 DR. KERR: And there is anothav order of magaitude
(-
U 15 contribution still -- no, wait a minute. Yes. And that 16 contribution of fire, though, is still about six times the 17 contribution of all other internal initiators?
18 MR. LAMBRIGHT: That's cccrect.
19 DR. KERR: Even with Appendix R.
20 MR. LAMBRIGHT: Even with Appendix R.
21 DR. KERR: So fire short of Appendix R contributed 22 about 60 times everything else. That seems odd to me.
23 MR. MCCRACKEN: I think the number could be odd.
24 The fact that fire prior to Appendix R was a very 25 significant contributor is not -- I got that a little later
() Heritage Reporting Corporation (202) 628-4888
l e 70
,(t ;1' lin the slides -- but basically, fire does not,make any new b 2 events. All. fire does is take out systems you were going to i
3 relyfon.. If you had very poor fire protection capabilities,
~
4 then'the reliability to all the' systems you thought were
- 5. redundant systems wasn't very good.
6 DR. KERR: Well, it just seems to me with the 7 operational history we had before Appendix A, if.the 8' contribution to core melt frequency was almost 100 times l
H 9' that or everything else, that we would have seen some core 10 melts caused by-fire.
11 MR. FLACK: That was the history; our first big-12 ' accident was a fire.
'13 DR. .KERR: Now, wait a minute. I'm talking about 14 core melt; I'm not talking about a big accident ~. Browns 15 Ferry in spite of what we may think, so far as I know, not a 16 core melt.
17 MR. MCCRACKEN: I think there's.a logical reason 18 for the discrepancy you think you're seeing. Fire 19 protection up until Browns Ferry was existence; it wasn't 20 nonexistent. But it was there primarily to protect the 21 economic interest.
22 DR. KERR: Now look, I asked what the core molt 23 frequency reduction was of Appendix R and I'm told about an 2t order of magnitude. And I'm told that'there still is a 25 contribution due to fire about six times, in summation of Heritage Reporting Corporation (202) 628-4888
r- ,
i F
71 r
( 1 all the other internal events. And I multiple the two and I 2 get about 60 or just roughly two orders of magnitude more 3 than the other internal events or whatever you want to 4 classify them.
5 Am I doing the wrong kind of arithmetic?
6 MR. CHEN: This is John Chen from Research.
7 I think the example what has been cited here is 8 for Peach Bottom itself. But the whole thing is, Peach 9 Bottom's internal event --
- .0 DR. KERR: Okay, give me different numbers. I li MR. CHEN: The number -- you cannot always use 12 that number because the Peach Bottom shows fire is higher in t 13 the risk, in term of the risk proportional. But the
.14 absolute value is still very low.
(~)')
15 DR. KERR: Tell me what number to use? Is it 16 order of magnitude reduction of Appendix R the wrong number? l 17 MR. MCCRACKEN: For some plants it's not the wrong 18 number. You're trying to generalize across with one number.
19 DR. KERR: I'm trying to get some idea of what )
20 Appendix R does generally in reducing fire risk. The number 21 I was given was in an order of magnitude.
22 MR. LAMBRIGHT: That's the case for an Indian 1
2? Point-2.
24 MR. MCCRACKEN: For a specific plant, Indian 25 Point-2.
7-(_)/ Heritage Reporting Corporation (202) 628-4888
i 72 1 DR. KERR: But the number would be. typical.
2 MR. FLACK: I think Indian Point-2 and Lemrick are 3 the only two cases where'one was able to go back and 4 calculate explicitly the contribution of Appendix R.
5 Modifications have been made on other PRAs, but it wasn't a 6 before and after comparison.
7 DR. KERR: So the order of magnitude number I 8 should disregard as being typical.
9 MR. CHEN: I don't think you can say that --
10 DR..KERR: Tell me what I.should say?
11 MR. CHEN: I don't think there is any typical 12 number you can compare. That's'only one set of numbers.
13 DR. KERR: Okay. So then I don't-know how much
/ 14 reduction Appendix R has produced.
15 MR. CHEN:- You cannot just say, because we impose 16 Appendix R we're going to have it 10 times less. But we can 17 say because it's more safer than what's previously without' 18 Appendix R.
19 DR. KERR: Well, if I describe safety in terms of 20 core melt frequency, what can I say about core melt 21- frequency reduction, nothing?
22 MR. MCCRACKEN:. In some plants it may well have 23 been close to.nothing depending on what they had to start 24 with. They were all designed and built differently. You're 25 not going to get one number.
O. erit ee aegerei-e (202) 628-4888 cerreratien
73
[() 1 DR. KERR: Apparently, I'm told that the status of 2 L compliance with the regulations is something, and that's 3 good because the regulation is the law of the land.
4 But I'm also interested in how much reduction in 5 core melt frequency I've got from Appendix R? I'm-asking:
6 this because I've heard some rather derogatory. comments
.7 about Appendix R.
8 What do we do if Appendix R is good, if it reduces 9 things by an order of magnitude then maybe there ought to be 10 another Appendix R which would reduce things by another 11 order of magnitude; I don't know. If it isn't any good, it 12 seems to me we ought to know it.
13 And I'm trying to discover what was accomplished 14 by Appendix R, not in terms of compliance with the (r] .
15 regulations but in terms of risk reduction.
16 MR. MCCRACKEN: I have not, number one, heard 17 anybody say that Appendix R wasn't any good.
18 DR. KERR: Well, I have.
19 MR. MCCRACKEN: Appendix R was not perfect.
20 DR. SIESS: You haven't heard the industry say 21 that.
22 MR. MCCRACKEN: I heard a few members of industry 23 say they didn't like it.
24 MR. FLACK: Could I make a comment.
25 What Appendix R did I think generally was, remove O 1e e- eerei 9 cereer eie-(202) 628-4888
t 74
'l single point; vulnerabilities where you could take out 2 ' multiple trains of equipment.
L 3 What we're left in the'PRA now is train of i 4- equipment being taken out by the fire and another failing 5 randomly or unavailable at the time. These are the 6 contributors that we see now in the fire PRAs.-
-7 If we were to do fire PRAs'before Appendix R I 8 think you would see single point vulnerabilities. Now these 9 have since.been removed, but it's hard to quantify that 10 unless one goes back to the PRA and sees what it was before 11 and then compares it to after. We have done that in two 12 cases and 10 and.3 were the values thht were estimated.' I 13 think that's all we can say at this time.
14 MR. MCCRACKEN: Looking at. efforts beyond the
[}
15 ' regulation which is where we're going now, we started with.
3 16 the plants currently meeting the regulations.
17 Severe accident policy statement basically summed 18 .down says that plants are expected to perform a limited 19 scope. analysis to discover particular vulnerabilities; and 20 that's where we' re going with IPEEE and trying to come up 21 with such a methodology.
22 DR. KERR: Did it say you were supposed to 25 discover particular vulnerabilities or outlier?
24 MR. MCCRACKEN: These are the exact words taken 25 out, "particular vulnerabilities."
() Heritage Reporting Corporation (202) 628-4888
75
( ,) 1 DR. KERR: Okay. I didn't remember what the 2 language was.
3 MR. MCCRACKEN: I have it highlighted.
4 DR. KERE: Good.
5 MR. MCCRACKEN: To accomplish this we have an 6 External Events Steering Committee formed to provide 7 directional implementation of policy. That goes to the 8 director of the Office of Research. And reporting to that 9 committee is the External Events Subcommittee which provides 10 recommendations to the Steering Committee. I'm the Chairman 11 of the Subcommittee on Fire.
12 Plant-specific vulnerabilities which are found 13 will be addressed then through the backfit process, because 14 they already meet the regulations; and therefore, if we want
}
15 to implement something we've got to go to the backfit 16 process to require implementation.
17 DR. KERR: And what is the particular 18 vulnerability? How do you know one when you see one?
19 MR. MCCRACKEN: That's the purpose of the ;
20 methodology, is to go through and see where you have 21 particular vulnerabilities such as those that have already 22 been addressed through the PRAs where you have a high 23 probability of fire because you have a system out of service 24 and a fire in another area would give you a higher frequency j 25 core melt. l 1
(h
() Heritage Reporting Corporation I (202) 628-4888 1
l 76
() 1 DR. KERR: So a vulnerability will be judged on 2 the basis of the quantitative core melt contribution?
3 MR. MCCRACKEN: No. The methodology is not going 4 to require a level 1 PRA. Some plants have done level 1 5 PRAs and they do have numbers.
6 DR. KERR: Well, I'm just trying to understand how 7 one knows when one has a particular vulnerability.
8 MR. MCCRACKEN: You will use a fault tree type 9 methodology to go through and look at the systems you have 10 to have to achieve safe shutdown.
11 DR. KERR: This will not be described 12 quantitatively.
13 MR. MCCRACKEN: The end result will not be
/~'T 14 quantitative. As you're going through the methodology you O
15 will use probability numbers to decide where to continue to 16 look. You will go through a screening process to find which 17 are the dominate sequences.
18 DR. KERR: And you will be looking at the 19 reliability of what?
20 MR. MCCRACKEN: You will be looking at, one, the 21 reliability of the systems tied into the internal events 22 analysis.
23 And then, two, the problem of a fire occurring, 24 the magnitude of the fire, location.
25 DR. KERR: Will you be looking at the contribution in
(,) Heritage Reporting Corporation (202) 628-4888
_ ____._.m .-___--___m_.._______m.m
1 77
[() 1 of unreliability of core melt?
2 MR. MCCRACKEN: Yes.
3 DR. KERR: Whether you have a vulnerability then 4 will be judged on the magnitude of the contribution to core l
5 melt frequency.
6 MR. MCCRACKEN: Yes, without using a'rpecific 7 number to accomplish that.
8 DR. KERR: How do you judge a contribution without 9 using a specific number?
10 MR. MCCRACKEN: You're-going to judge the 11 contribution by doing an engineering analysis once you get
~
12' down to the bottom line of, which of these contributors 13 appear to make sense as a type of problem that could lead 14 you to a loss of a system function.
15 DR. KERR: I'm sorry, but I don't understknd that
- 16. statement.
17 MR. MCCRACKEN: We're going to use numbers to 18 screen where we find --
19 DR. KERR: I mean, core melt frequency numbers?
20 MR. MCCRACKEN: Yes.
21 DR. KERR: Okay.
22 MR. MCCRACKEN: Where ve find that there are core 23 melt frequency numbers that are below a screening criteria 24 which we have not yet set. Those areas wil2 be further 25 investigated.
Heritage Reporting Corporation (202) 628-4888
i-78 f()
'1' DR.'KERR: What. sort of numbers are you. thinking; 2 about?'
3 MR. MCCRACKEN: I'm' thinking about something, I '4 perhaps, 10 to the minus 6 as'the' initiator.
5 DR. KERR: So anything'that's bigger than.10 to 6L the minus 6.
7 MR. MCCRACKEN: We investigate further, 8 DR.' KERR: ~ And once you investigate it further,
-9 then-what do you do?
10 MR. MCCRACKEN: You look at the potential' 11 sequences that~it's involved in. The'means you have to
.12 mitigate those' sequences or what becomes the limiting factor-13 mitigating those sequences, and make a determhsation as to-14 where.the vulnerability exists that would prevent you-from.
)
15 mitigating a sequence.
16 DR. KERF _: Sct you don't make any changes,Jyou 17 just look?
l 18 MR. MCCRACKEN: 'This is to identify l 19 vulnerabilities.
20 DR. KERR: Oh, okay.
21 And a vulnerability then is anything that 22 contributes nore than 10 the minua 6 to core melt frequency.
I 23 I mean, if you decide finally to use it.
24 MR. MCCRACKEN: If we decide finally to use that, 25 that would be something we would ccnsider vulnerability.
r Reporting Corporation Heritage (202) 628-4888 1
1 _
4
- 7 9 :.
(f 'l But again, that's a number that's going to.be 2 worked out and jointly agreed on.
- 3. MR. MICHELSON: What do you do now for the - .in 4 trying to determine your vulnerabilities you have to deal 5 with the real world situation which is an initiated fire 6 somewhere with multiple things as a consequence happening as 7- a consequence.
8 In determining whether you have a vulnerability.
9 are you going to.somehow determine what all these multiple 10 things are and then collectively examine their affect on the 11 . core melt that you - got from a PRA that did not even ' include 12 these in their branches,' necessarily, these events?
13 MR. MCCRACKEN: Those issues will be considered.
14 .I think-the simple answer to it is based on the majority of 15 work that has been done to date, for instance, the last PRA 16 that was done which was the Brunswick PRA, they have found
~
17 no sequences that go beyond a given fire area, i.e., the 18 -three hour rated boundary.
19 So when you come into sequences of events that 20 cause problems they are normally, as a consequence, of 21 problems within a fire area, different zones within the same 22 fire area.
23 MR. MICHELSON: Are you saying that there aren't 24 any things that could happen at Brunswick that would go 25 beyond a given fire zone?
() Beritage Reporting Corporation (202) 628-4888
90
() 1 MR. MCCRACKEN: No, I didn't say that.
2 I said, the vulnerabilities that were identified 3 showed that the sequences that were dominate were sequences 4 where you had a cross zone problem as opposed to through a 5 three hour rated fire area.
6 MR. MICHELSON: Now, in order to determine if you 7 have a vulnerability you have to consider all the things 8 that are happening as a consequence of the fire.
9 MR. MCCRACKEN: That's correct.
10 MR. MICHELSON: So you somehow stack these up.
11 And if you do a PRA these kind of come out to some extent, 12 if it's well modeled. But having not done the PRA you are 13 doing this somehow in your head or on flow sheets of some
(} 14 sort.
15 But you are stacking up e very single thing that 16 can happen as a consequence of the fire in determining 17 whether or not there is a vulnerability existing here that 18 needs to be fixed.
19 MR. MCCRACKEN: Yes.
20 MR. MICHELSON: Is that what you're doing?
21 MR. MCCRACKEN: Yes.
22 MR. MICHELSON: And I guess you can do this in 23 your head with lots of flow sheets and a lot of 24 understanding.
25 MR. MCCRACKEN: The intent is that the guidance Heritage Reporting Corp 0 ration (202) 628-4888
01
- () 1 will be rather specific.
2 MR. MICHELSON: What are you going to do about 3 things like the mitigation of heat and smoke beyond the 4 immediate fire, issues of that sort; ure you just ignoring 5 those in this vulnerability analysis?
6 MR. MCCRACKEN: No. You're going to look at the 7 fire area designation and whether there is a potential for 8 spread beyond that fire area.
9 MR. MICHELSON: By spread you mean the heat and 10 smoke can spread.
11 MR. MCCRACKEN: Heat, smoke and/or fire.
12 MR. MICHELSON: Well, I'm just worried about heat 13 and smoke for the moment, because I want to actuate some 14 fire protection in other areas and sprinkle them down. Is 15 that going to be part of this vulnerability analysis?
16 MR. MCCRACKEN: That is going to be considered in 17 the vulnerability analysis. We're going to look at what 18 significance we think it contributes and address --
19 MR. MICHELSON: To do that you will have to have 20 some kind of a calculational tool that tells you how hot and 21 how smokey a certain area is getting from a consequence of a 22 fire in another portion of the room or whatever. You will 23 have to have that kind of a tool. I don't know that we have 24 it, but you will have to have it if you're going to follow 25 heat and smoke effects.
(A) Heritage Reporting Corporation (202) 628-4888
t o -
q a
-82 s
h1 MR..MCCRACKEN: I will have to have an 2 understanding of how that plant-is designed and built with 1
3 the' ventilation. systems --
4 MR. MICHELSON: Well, you will havis to have a
! 5 calculational tool that tells-you if you have gotten to the 6 actuation temperature of the particular sprinP.lers, for 7 instance.
8 MR. MCCRACKEN: No, I don't agree with1that.
9 MR. .MICHELSON: You mean_you don't --
l 10 MR. MCCRACKEN: I don't think I do need that' tool.-
11 DR. CATTON: Are you going to guess?
12 MR. MCCRACKEN: No.
13 MR. MICHELSON: What are you going to do?
14 MR. MCCRACKEN: I am going to do a limited search
{
15 for vulnerabilities beyond the regulations. I am.not trying 16 to cross every "t" and dot every "i" and solve every problem-17 anybody can come up with.
18 MR. MICHELSON: ~Are you just going to assume they_
19 do actuate; is that what you mean then?
20 MR. MCCRACKEN: If systems are identified that are 21 close by, that if actuated could affect your other redundant 22 train that you have to have, then that has to be identified 23 as potential vulnerability.
24 MR. MICHELSON: So having decided that, yes, there 25 is a vulnerable area and I will actuate in my analysis, what i
() Heritage Reporting Corporation (202) 628-4888 i
83
,[ ) L1 is thei consequence of . actuation that you will use for your- ]
2- analy/is?
j 3 MR. MCCRACKEN: 'You then have to look_at the_
4 consequence of. actuation and make a detertaination as to 5 whether.that system of actuated is going to wipe:out a 6 ' component'you have to have.
7 MR. MICHELSON: Well let's say there is a solid.
8 ' state control cabinet that you are now spraying.with water, 9 what are you going to assume about what that cabinet does in,
'10 doing this vulnerability analysis?
11 MR. MCCRACKEN: The.first thing you have to assume 12 is, you're going to lose.
13 MR. MICHELSON: Well, losing it mi;ht be the nicest thing that could' happen to you.- That.might be'the
-( f 14 15 nicest thing. The thing that won't be so nice is when it 16 starts malfunctioning, give false signals to do things, 17 opens valves that you really didn't want to have open.
18 That's the thing that you start worrying about.
19 If it just quits and doesn't do anything, that 20 might be okay.
21 So what do we do-in our analysis of vulnerability 1
22 when we come to components and we say, okay, we have 23 actuated the fire suppression, we're going to spray them 24 down; what's the affect of spraying them down?
25 MR. MCCRACKEN: The affect of spraying them down Heritage Reporting Corporation (202) 628-4888
84 1 is.the fact that you have a vulnerable component. And-
, 2 ' spraying it down it can do things 1you don't.want it to do.
3 LYou'have identified it as-a vulnerability, which is the 4 purpose of the search.
5 MR.1MICHELSON: Then you'll go back and make sure-6 it doesn't' happen.
7 MR. MCCRACKEN: Then you make an assessment as to 8 whether you.need to fix it, so it won't happen, or you can 9 live with it.-
10 MR.-MICHELSON: Well, you can't determine you can-11 live'with.it until you know what it would do.
12 MR. MCCRACKEN: Well, you have to know what the 13 consequences are.
14 MR. MICHELSON: Yes.
)
15 MR. MCCRACKEN: First you have to determine 16 consequences. And then you may put splash shields on it.
17 MR. MICHELSON: Well, I.think we're together then.
18 MR. MCCRACKEN: Yes.
19 MR.-MICHELSON: This is a rather conservative 20 assumption that everything actuates, that's potentially
.21 within this heated zone.
22 MR. MCCRACKEN: But that's the easy way to do your 23 first cut to determine if you have a problem.
L 24 MR. MICHELSON: But determining whether or not 25 actuating, the fire protection is going to be a concern or
() Heritage Reporting Corporation (202) 628-4888 l
85-
)' l' not'is a little more complicated because now you have t'o 2- have detailed understanding of' responses and the responses
.3 just quits operating.
4 MR. MCCRACKEN: That's correcc.
5 MR. MICHELSON: That's perhaps the most-desirable 6 response. What you worry about are these other things, and
~
7 I guess you're going to do that.in your analysis.and inyour. ,
'8 guidelines you're going to tell people'how to do this, Juid 9 I guess you've got some data or something that tell you 10 typically what to' expect.
11 MR. MCCRACKEN: The intent is that we be able to-12 identify those components that are vulnerable. And'in.most 13 cases what the probable solution would be is to protect them
.14 as opposed to worrying about what they' re going to do.
{}
15 In other words, you'll see what they can do is 16 going to be over such a wide variety of sequences.
17 MR. MICHELSON: When we decide we will actuate the
- 18. fire protection because we can't seem to argue it won't 19 act'uate , do you chase the water on down the floors and 20 through the cracks and into the conduits and that sort of 21 thing, too, as a part of this process?
22 MR. MCCRACKEN: You have to look at, are there 23 components below in the next fire area that if you, in fact, 24 get leakage there, it's going to wipe out that fire area and 25 do you need it. If you don't need it, you don't worry about Heritage Reporting Corporation (202) 628-4888 m _ - _ _ - - _ - - - - _ - - - . - - _ - _ - _ - -
l 86
.. m -
() 1 it.
2 MR. MICHELSON: No, no, it's not that I need the 3 component down below, it's what that component down below 4 may do to equipment I do need that I worry about. I have-to 5 worry about both. It isn't just the availability of that 6 equipment; it's what that equipment down there could do to 7 other equipment.
8 It might be the heating and ventilating that now-9 gets lost as a consequence. Or it might be more fire 10 protection that now actuates and does something else. You 11 got to chase - you can't just assume that --
12 MR. MCCRACKEN: Then you will know.
13 MR. MICHELSON: You don't just look for your hot 14 shutdown equipment and that's all you ever worry about, in 15 other words, which has been the approach in the past.
16 Many times people have been looking just for the 17 hot shutdown equipment and make sure it's okay from the 18 viewpoint, well, the fire isn't there; the smoke isn't j 19 there; the heat isn't there. But the affects could be there 20 from the -- well, it's essentially the system interaction 21 problem.
22 MR. MCCRACKEN: You have to identify that there 23 are no system interactions that's going to cause that to 24 cause a problem.
l l 25 MR. MICHELSON: So if the guidance is written (O, Heritage Reporting Corporation (202) 628-4888 l
87
- ~s
(, )_
1 right and comprehensive enough it can do the job. I agree.
.2 So now we have to see the guidance.
3 DR. KERR: I go back to the vulnerability thing a 4 bit. If we use the table on the vulnerable fire FRAs and if 5 one accepted, say, 10 to the minus 6 as the final number, 6 and realize you have it; then every one of the 15 or so 7 plants that is listed in that table -- some in duplicate --
8 would fail to meet that screening criteria. Indeed, if the 9 screening criterion were 10 to the minus 5, most of them 10 would fail to meet it.
11 Now that says then that Appendix R has been 12 inadequate for every one of that, I guess, inadequate for 13 every one of that sample.
14 MR. MCCRACKEN: The 10 to the minus 6 criteria
{}
15 would be a criteria or screening initial cut sets for 16 vulnerability; that's not overall core melt frequency. Now 17 you' re talking total core melt frequency which is a 18 combination.
19 DR. KERR: Well, it says, fire core melt 20 frequency, I assume this means the contribution. And I'm 21 looking at that table. And even if one used 10 to the minus 22 5 almost every plant in there would fail to meet the 23 criteria.
24 MR. MCCRACKEN: For that initial cut set to decide l 25 whether to search further.
() Heritage Reporting Corporation (202) 628-4888
o
[ 4 88
- 11 ' Remember,Twe're trying to F.ecide whether to' search:
'2 further. The initial cut sets can be conservative.- How; I= '2' ' conservative, we haven't decided.
-4: DR. KERR: Oh, I see.- So I.shouldn't'take these 5 1 numbers too seriously at this point. .
6 MR. MCCRACKEN: I said, this is your initial cut 7 -set. 'To say, beyond this we're not going to look; it's 8 small enough'we're going to forget about11t.
9 ,DR. KERR: At what point will one be -- suppose 10 that one, decides that something needs to be'done,.how wills 11 you decide how much needs to be done?1 Will that~be a
'12 quantitative thing-or will you just look at the' obvious 13 things.that can be done and do them?
. 14 MR. MCCRACKEN: We're going to have to go through 15 the.backfit analysis and we're going to have toLdo the 16 normal cost benefit' analysis to do.it.
17 DR. KERR: =And at'what point -- so.it will-be a 18 -matter of -- now you've got to actually go throurih the risk and -- well, I guess'you will use a generic population
. :2 0 distribution in order to calculate.
21 MR. MCCRACKEN: My assumptions we would use 22- generic, I.would certainly hope so.
23 DR. KERR: That's going to be a complicated 24 process.
25' MR. MCCRACKEN: I agree.
Heritage Reporting Corporation (202) 628-4888
y L- 89 h ~1 MR. LAMBRIGHT: I would like to make a comment on 2 .the screening. criteria. I have'been. involved now in'doing 3 :four fire'PRAstin the last two years and using a screening-4 criteria of 10 to the minus 4 of probability and' assuming:
5: that'all' things within a fire zone'would fail tends to only 6 give you a' reduction:in fire area you look at, at any given 7 plant that I have looked at, by only-about half or 8 something.,
9 So you just get a limited reduction from doing 10- screening in this fashion.
11 DR.HKERR: A limited reduction in what?
12 MR. LAMBRIGHT: In fire 3reas under consideration i: 13- at any given plant.
14 'DR. KERR: What are you telling me then, that-15 Appendix R has really not produced an acceptable level of 16 protection in any case, probably?
17 MR. LAMBRIGHT: No. I'm saying if you use that 18 particular screening criteria, assuming that all things 19 within a particular fire area could fail and you set it at a 20- level of 10 to the minus 6 or even 10 to the minus 4, you're 21 'not going to reduce the amount of fire areas under 22 consideration to any great extent at any plant.
23 DR. KERR: I'm trying to find out what has been 24 accomplished by Appendix ' which was a very elaborate 25 procedure, to get some idea oi' what is going to have to be
() Heritage Reporting Corporation (202) 628-4888 l
I 1
p 90 c.,
( )~ 1 done to produce this further reduction. Presumably, you 2 have already taken care of those obvious things. Now you're 3 going to have to do stuff that's really tough. And I wonder 4 how much and what direction this is going to take and where 5 it is that you stop.
6 Now, presumably where you stop is going to be 7 determined by cost benefit analysis, which has many more 8 uncertainties even than PRAs And if you're really serious 9 about this approach it surely is going to be expensive in 10 time on both the Staff and a licensee and full of 11 uncertainties, but I suppose that's life.
12 MR. MCCRACKEN: This is the initial screening 13 criteria. Once you go through the initial screening 14 criteria, then you go down a little further which is, you go
{~-}
s 15 into the individual fire areas, you look at fire sources, 16 initiators which can give you further screening.
17 There are means of screening further as you go 18 down. When you get down to the very bottom of the analysis 19 you will limit it to a smaller number of areas you're going 20 to look at in great detail.
21 DR. KERR: See, if I were a licensee I would 22 realize that I didn't have to do :his just for fire, but I 23 have to do it for any other sequence that produces more 24 than, let's say, 10 to the minus 6.
25 MR. MCCRACKEN: This will be tied into other O Heritage Reporting Corporation -
(_f '
(202) 628-4888
)
l 1
l l
f
91 1 i
.( ).. 1E ' issues-that are being addressed. Like seismic willLbe' tied, 2 fire =and seismic are tied together and fire it tied to IPE; -
3 they all go together.
4 MR. LAMBRIGHT: I wou22 like to make an additional 5 point that.one of the plants at Sandia that we have 6 evaluated the fire risk that used strictly engineering 7 techniques and plant walkdown procedures to determine what 8 plant areas had potential. fire vulnerabilities. And in that 9 analysis two fires were identified totally at'the plant.
- 10. And the conclusion at that given plant was that 11 fire wasn't an important contributor to core damage 12 frequency. When vital area analysis techniques were 13 computed or used 16 plant areas that were never identified
( 14 during the plant walkdown were identified with core damage
.15 frequencies of 10 to the minus 7 per. year or greater. And 16 fire was identified as being the dominate contributor to 17 overall core damage frequency for this plant.
18 DR. KERR: Do you feel good about how much
~
19 credence to give to something that contributes 10 to the 20 minus 7 per year to core melt frequency?
21 MR. LAMBRIGHT: Well in the sense, if you're using i
L 22 similar models and event trees and fault trees and similar 23 assumptions as to failure rates for random failures as 24 internal events, you have already a comparison, I think, 25 between a fire analysis and an internal event analysis.
Heribage Reporting Corporation (202) 628-4888
- e. _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ -
e 92
'(,) - 1 What the relative numbers mean exactly, I think 2 you just have to make a comparison.
3 DR. KERR: I guess I personally don't give much 4 credence to any sort of number that -- PRA type number that 5 gives me a 10 to the minus 7 core melt frequency, but that's 6 a subjective thing.
7 MR. FLACK: We have to be careful though when we 8 define sequences between functional and systemic. You may 9 find many systemic sequences at 10 to the minus 7 which can 10 contribute to a substantial number at the functional level.
11 DR. KERR: If you depend on getting your total by 12 adding up these, whatever you call them, 10 to the minus 7s 13 per year --
r 14 MR. FLACK: 1150 at 10,000 sequences.
15 16 17 18 19 20 21 22 23 24 25 A
V Heritage Reporting Corporation (202) 628-4888 l
l l
.s 93
) 1 MR. MICHELSON: Are you. concluding that the' 2' walkdown process is not'ani. effective way of finding the real-3 fire vulnerabilities?
4 MR. FLACK: No. I-think the walkdown process.is 5 really at the heart-of the issue. I think that's is the 6 most important part.--
7 MR.-MICHELSON: Then I didn't understand the 8 earlier comments about as soon as you did the PRA you'found 9 all these other things.
(
10 MR. FLACK: ' Well, one is screening criteria that 11 is to be reported to the staff. I'm referring to the IPE 12 for internal-events. First is what one defines as a 13 vulnerability.
(} 14 15 MR. MICHELSON: Well, I think the whole point here, of course, is we axe trying to reduce the risk from 16 fire and if we try to avoid doing.a PRA and we do something 17 else in lieu of it, that's fine, provided it can lead to the 16 conclusion th.st you have an acceptable risk from fire.
l 19 Whether it is' quantitative.or qualitative.
20 MR. McCRACKEN: 'I think.the point here was a 21 s,' tem walkdawn without an organized prior search to tell 22' you where you need to be looking and what potential 23 vulnerabilities exist is not going to find a whole lot.
24 MR. MICHELSON: Okay.
25 MR. McCRACKEN: This is basically what happened
( Heritage Reporting Corporation !
(202) 628-4888 p
94
() 1 with Appendix R. Appendix R is a walkdown.
2 MR. MICHELSON: Yes.
3 MR. McCRACKEN: We did a walkdown. We said, 4 "Okay, everything that looks like this, make it look better 5 so it won't be a problem. Eliminate the single fire 6 vulnerability issue."
7 We did that. That was a big reduction in 8 potential problems at power plants.
9 Now, we're saying: Okay, there are systems.out 10 there --
11 DR. KERR: Now, wait a minute. You were telling 12 me earlier that you really only knew how much of a reduction 13 it was in about two plants.
14 MR. McCRACKEN: We gave you the numbers.
}
15 DR. KERR: You said an order of magnitude and when.
16 I pressed for details, you told me got an order of magnitude 17 on one plant. On another plant, .you got a factor of 3.
18 Now, I will accept those, but unless you have some other 19 elements, I don't know how much of a reduction was l
20 accomplished across the boards. Do you?
21 MR. McCRACKEN: I don't have a number for it.
22 DR. KERR: If you don't have a number, you don't 23 know.
24 MR. McCRACKEN: Yes, I do.
25 MR. MICHELSON: Well, then we will never do if we Heritage Reporting Corporation (202) 628-4888 L
l L
i h 1 N
h 95 1
"1 don't do'PRA. That's the problem.
.2 MR. McCRACKEN: I know that if_I have v.wo trains, 3 an A and a B train in this room and I wrapped one'and-4 protected it in a 3-hour barrier, that that plant is safer 5 than it was before I did that.
6 'DR. KERR: But you don't know how much safer.
7- MR. McCRACKEN: I can't quantify it.
! 8 MR. MICHELSON: You don't know if it is safe 9 enough, yet.
10 MR. McCRACKEN: I know it's safer.
11 MR. MICHELSON: . Safer.
12 MR. McCRACKEN: Appendix R made them safer. I 13 .have absolutely-zero doubt in my mind.
(} 14 15 MR. MICHELSON: Well, I.have doubts about that statement because I think we may have made them a little 16 less safe by too much automatic fire protection, for 17- instance. So, it's not out of hand clear that'everything 18 you put in is good. It may be in some cases not be good.
19 MR. McCRACKEN: In some cases, automatic 20 suppression may not be.the best choice.
21 Okay. The basic conclusions from the Fire 22 Subcommittee were that (1) each operating reactor should be 23 evaluated to determine plant specific vulnerabilities to 24 internal fires -- that was one of our charters -- should 25 they be examined. The reasons that we came to that Heritage Reporting Corporation (202) 628-4888
96 cm
() 1 conclusion was that industry performed PRAs to demonstrate 2 that fire can be a significant contributor to core melt. I l
3 And there are approximately 12 to 18 PRAs that have shown l t
4 that.
5 Some vulnerabilities identified through the PRA 6 process were corrected prior to reporting --
7 DR. CATTON: Do the industry PRAs yield the same 8 kind of numbers that the Sandia PRA, you know, this factor 9 of 10 for Appendix R and a factor of 6 times core melt for 10 fire?
11 MR. LAMBRIGHT: In fact, in the case of most 12 industry PRAs, and most of them, they were performed in the 13 early Eighties, core damage frequency in most of those PRAs 14 were due to fires higher than the Sandia predicted core
~}
15 damage frequency for either Peach Bottom or Surry.
16 DR. KERR: But these frequently did not consider 17 Appendix R. Now, I don't know how many -- because I have 18 asked that specific questions on a number of cases and --
19 MR. McCRACKEN: They didn't consider Appendix R 20 separately. They considered their plant as it existed.
21 DR. KERR: That's right. 1 22 MR. McCRACKEN: With Appendix implemented.
23 DR. KERR: Well, many of them did not consider 24 Appendix R implemented. I have asked in a number of cases, i
25 MR. McCRACKEN: Well, the ones that I have seen
) Heritage Reporting Corporation (202) 628-4888 I
97-h 'l considered'what had been implemented as implemented and if 2 something had been implemented --
3 DR. KERR: No, but in many cases the PRAs were 4 done before Appendix R had been implemented. That is the 5 point I am making. And those did not-consider that 6 Appendix R had been implemented.
7 MR. McCRACKEN: There are very few that are in 8 that category. .The majority of them had a lot of Appendix R 9- modifications implemented before they ever tried to do a 10 PRA.
11 MR. 1AMBRIGHT: In fact, in the case of the four 12 PRAs that were re-examined as part of the Fire Risk Scoping 13 Study, even with Appendix R modifications core damage
'{} 14 . frequencies were still in the range of 5, either the minus 5 15 to roughly 90 to the minus 6 per year. And then also for 16 Peach Bottom and Surry-with their Appendix R modifications, 17 core damage frequencies were greater than 10 to the minus 5 18 per year. That's six that I'm aware of.
19 MR. McCRACKEN: And the other reason for our 20 conclusion that we had to do a vulnerability search was the 21 Fire Risk Scoping Study results indicate the potential for 22 plant-specific vulnerabilities. And we felt that that had 23 to be part of the search.
24 The second conclusion we came to was that a Level 25 I PRA was an acceptable methodology to perform the
() Heritage Reporting Corporation (202) 628-4888 y
f' L
98 l:
p ()
n 1 evaluation for-fire vulnerabilities and that other-2- methodologies may be acceptable, but hadn't yet been 3 . developed. And that there are limitations =in the 4 application-of the fire spread codes, however, in the-fuel
-5 ' locations per plant, once you-have narrowed down your areas-6 of vulnerability where you have to use it, you can,do this 7 in combination with engineering evaluation and make a 8 determination as to whether in fact you do have a 9 significant vulnerability.
10 And looking at where fire is in relationship to 11 IPE, something I stated a little bit earlier: Fire doesn't 12 create a new accident; it increases the probability of 13 another accident by taking out systems you would have relied 14 upon. Therefore,-whatever is done to resolve the internal 15 ~ events has to be compatible with what we do on the IPEEE.
16 Generic Letter 88-20 permitted, and we. discussed this a 17 little bit this morning, a PRA with some modifications, 18' Level I, and enhanced IDCOR or other-systematic methods 19 found acceptable by the Staff.
20- MR. MICHELSON: I guess I don't understand your 21 first bullet. Could you tell me a little more? We talk 22 about increases the probability on existing ones. What do 23 you mean by that? Fire apparently isn't --
24 MR. McCRACKEN: Existing accident sequences.
25 MR. MICHELSON: Are you saying that fire is not a
() Heritage Reporting (202) 628-4888 Corporation l
- .__ _____=__ _ - ___ . . _ _ . _ _. _ _ _ - _ _ _ . . - _ ___ . _ _ _ _ _ _ - - _ _ _ - _ _ _ _ _ _ _ _ _ . _ - _ _
l; i
k 99-l ..
II :( [ 1- new accident but somehow i?. changes the' story on old 2 accidents. Is that.what you're saying?
3 MR. McCRACKEN: On the probability by eliminating L 4 equipment that would have been relied on. It decreases.
5 equipment reliability.
l l 6 MR. MICHELSON: Well, this puzzles me. I guess
'7- what you're saying then that you have already analyzed, for 8 instance, the-core melt contribution from, say, the. loss of.
9 all cooling water. And that a fire that might'cause a loss 10 of all cooling water doesn't change that any.
11 MR. McCRACKEN: It's going to be the same 12 accident. You are just getting to it from a different 13 means.
14 MR. MICHELSON: Yes. If ultimately that's the 15 worst . case of the consequence of the fire is to get all the 16 cooling water then, indeed, I shouldn't expect any 17 difference in the answer, but I expect a little difference 18 in the initiating event that led to it because now it fires 19 the initiating event with a certain probability.
20 MR. FLACK: Excuse me.
21 MR. MICHELSON: Sure.
22 MR. FLACK: The initiator is usually not the fire, 23 itself. It is an event that the fire causes.
24 MR. MICHELSON: Oh, no. I'm going to speculate 25 that I have a fire as the initiating event which there is
() Heritage Reporting (202) 628-4888 Corporation
_______________-____A
, 100
/-
.( j(L l' nothing wrong with that speculation. That's the name of the-2 ' game'.
3 MR. FLACK: Well, the initiating event may be a 4 plant' transient'which causes a trip. But it is something 5 that would affect the plant.
6 MR. MICHELSON: That might be, but another 7 initiating event will be a fire. A.short in an electrical 8 cable, it starts burning, and you get a' fire. And when you 9 get that' fire in certain circumstances, you might lose all 10 cooling water, depending on where those cooling pumps are 11 relative to the fire. In many cases,.they are in the 12 auxiliary building of the component cooling water and 20 or 13 30 feet apart and we are using physical separation of space
{} 14 15 to take care of it. That kind of a' fire, now, could lead to loss of all component cooling water, which could lead to the 16 loss in most plants of all capability to feed water to the 17 core.
18 MR. FLACK: I think we are saying the same. But 19 what will ultimately happen is the plant will trip without 20 component cooling water. See, that would be --
21 MR. MICHELSON: Oh, the plant will trip along the 22 way, sure.
23 MR. FLACK: That would be the initiating event.
24 MR. MICHELSON: No , no. That's not the -- the 25 fire is what starts all of this.
Heritage Roporting Corporation (202) 628-4888 i
L_____._____.__.__.______.______ __ _ _ _ _ . _ _ _ _ _ _ _ _______._i
101' I' .
'l - MR . FLACK:. I know, the way'it's treated in'a PRA, 2 -it is a little bit'different.
i 3 MR. MICHELSON: Well, fire, I hope the PRA 4 starts it --
n ;5 DR. KERR: What he is saying'is they are not going-6 to permit fire to be an initiating event. If they don't, it-7- Ecan't be an' initiating event.
8' MR. MICHELSON: Yes, I think that's what he is 9 saying, all right. And I don't understand the rationale on 10 approaching ~it that way.
11 DR. KERR: Well,-I mean he just doesn't want a 1
12 fire to be an initiating event. ,
13 MR. MICHELSON: Well, it's not logical, I guess.
i 14 MR. McCRACKEN: I think he is talking about he's 15 talking about.his cut set sequences when he is doing the IPE-16 that it is not aut initiating event. That is not saying that j 17 whatuyou're saying is incorrect.. A fire can start'that will 18 cause something, but when you get into their fault trees for 19 IPE, it isn't going to say, " Fire started." It will say, 20- " Systems started."
21 MR. MICHELSON: Yes. I think what you'll find is 22 when you go through those cuts, that's for other events that 23 you will find that the cut sets for other events that you 24 will find that the cut sets that you're going to get this 25 time are going to be a little bit different, coming from
() Heritage Reporting Corporation (202) 628-4888
e-102
~
-1 -different directions and whatever.
~
2 The ultimate' consequence is indeed loss of all L L3 cooling water, say, is one; but thereLare many other 4 possibilities, too. And I just don't follow the rationale 5 of the statement, but I guess it's all right. Clearly,.in l
6 'the end, fire can cause the loss of a core melt down.
7 MR. McCRACKEN: Well,-the rationale of the.
8 statement or what it's saying is if you.have -- you do an 9 PE, you come up with a core melt probability. Fire will.
10 always make it worse.
11 MR. MICHELSON: Yes.
12- MR. McCRACKEN: Because all fire does is take out 13 systems.you are going to rely on. It is not a new accident.
14 It simply takes out something you intend to rely upon.
-15 MR. MICHELSON: But see, what happens is you do 16- your PRA and you come up with all these other events as-17 being relatively low probability and then you say, "Well, 18 fire didn't even show up in them. It's a non-contributor."
19 And I am not sure that that's the case.
20 MR. McCRACKEN: No. You didn't do fire before.
21 And now you are superimposing fire on top of what you 22 already did and you may then look at your dominant sequences 23 and say, " Oops, a fire in this location is a very high 24 probability problem. We need to go fix it."
25 MR. LAMBRIGHT: In fact, I think you don't want to
() Heritage Reporting (202) 628-4888 Corporation
103 f(j ' 1 just look at your dominance sequences, but you want to look 2 at all potential fire induced sequences because something 3 that was dominant from internal events doesn't mean that it 4 is going to be dominant from a fire standpoint or a seismic 5 standpoint.
6 MR. MICHELSON: Precisely. You have to do a fire 7 PRA.
8 MR. LAMBRIGHT: You have to do a comprehensive 9 look at all potential fire induced initiating events.
10 MR. MICHELSON: Yes.
11 DR. KERR: Well, when you get beyond the screening 12 situation and decide to do a backfit which I believe you 13 said one had to do because of regulations. Who is going to 14 decide what it is that one does to reduce the frequency. Is
(/}
15 this going to be a responsibility of the licensee who picks 16 out one or two things and does a backfit analysis or is it 17 going to be the Staff's responsibility to decide on a fix 18 and do the analysis or is that beyond where you have gotten 19 in your decision process?
20 MR. McCRACKEN: I think where we are going to be 21 in the decision process is that we anticipate that the 22 majority of licensees for those items identified that are 23 relatively easy to fix, we'll go ahead and fix them. I ara 24 sure there will be a few who won't. And the ones who won't 25 we're then going to have to do the backfit analysis.
() Heritage Reporting Corporation (202) 628-4888 l
l - - - __._________ o
m-L
'104 '
- m Is,). 1 -We-don't. anticipate, but.there'may.be some high-2' probability events that are going to be significant in costs 3 'and we are going to get into a --
L 4 DR. KERR: Well, sir,'a: licensee might decide 5: that a 10 to the minus 6 contribution is not very~much.
6 MR. McCRACKEN: Correct.
7 DR. KERR: And you disagree with him presumably or 8 would you?
9 MR.-McCRACKEN: I would have to look at it and 10- what I'think it is and do an evaluation on that individual 11 .and make up my mind.
12- DR. KERR:. Well, you mean just the fact that 13 something is 10 to the minus 6, you would want to know what
() 14 kind of 10 to the minus 6, whether it was green or blue.
15 ' MR . McCRACKEN: Sure. I would want-to know 16 whether that 10 to the minus 6 was a number or whether it 17 was an engineering evaluation in support of that that I 18 should be worried about it.
19 Going'a little bit to a subject you'll get in on 20 as a requal about a month from now, what we're doing with 21 the'ALWRs in the area of fire is basically trying to 22 eliminate a lot of the questions we keep asking ourselves on 23 the current generation of plants.
24 The next plants will have to be capable of safe 25 shut down assuming total loss of any one fire area, assuming Heritage Reporting Corporation (202) 628-4888
1po- ,
I f 105 j L 1 no operator-re-entry. .And the exclusion for that is the 2- reactor containment building with specific reviews of the 3 locations there. The punch line is you eliminate the 4 20-foot separation criteria that all redundant trains will~
~
i- 5' be outside of 3-hour barriers. So, you will have a 3-hour-l
~
6 barrier between any redundant train, other than when you'get 7 into the reactor containment building, and there will be 8 specific guidelines on how you handle those.
1 9 MR. MICHELSON: Now,-I've got a couple of 10 questions on this slide.- Some people build 3-hour fire 11 barriers around, particularly around' oil-containing l 12 equipment because they worry about, particularly about fire 13 there, so, they put a concrete wall around it.- Now, it 14 isn't a wall'from floor to ceiling, it's a' wall that might
[
15 be 8-10 feet high. And it's a 3-hour rated wall.
L 16 However, all the heat and the smoke from the fire-17 can go on up as a chimney up and come on out and spread 18 around and so forth. Are you still going to allow that type 19 of fire barrier?
20 MR. McCRACKEN: The second bullet, we're saying-21 that you have to now consider smoke and heat, which is part 22 of that, the overall document --
23 MR. MICHELSON: So, if you want to use a wall that 24 does not have a ceiling on it, then you have to take account 25 of the plume effect --
Heritage Reporting Corporation (202) 628-4888 L -- - - . - _ - - - _ - - - _ _ _ - - - __ ---_-----. ------___- _________--- ____
106 L
L h 1' .MR.'McCRACKEN: Right. You are going to have.to
'2 'show that that oil burningLis not going to totally' wipe out 3 the whole rest-of the area due to smoke and heat.
4 MR. MICHELSON: From whatever consequence, now.
5 MR ~. McCRACKEN: Right..
6 MR. MICHELSON: Are you going to include the 7 consequence of act':.ating other fire protection equipment and 8 so forth in that area?
9 MR . McCRACKEN: They have to look at it.
10 MR. MICHELSON: Now; how do you define fire area 11 in this particular case?
12 MR. McCRACKEN: Three-hour rated' barrier.
13' MR. MICHELSON: Well, in the case of the wall
.%/
(') 14 that's 8-feet high and the smoke coming out'and so forth, 15 where is the fire area?
16 MR. McCRACKEN: The locations you are talking 17 about are locations that were found adequate in already 18 designed and buf.It plants. We don't anticipate having a1 19 whole lot of that type of area in the new plants. If they 20 come in with an area like that, they have to justify why 21 they can address smoke and heat control coming out of an 22 area where you simply have 3-hour walls, but not all the way 23 covered on top.
24 MR. MICHELSON: Now, in this case, we say we have 25 to take the total loss of equipment. Would that be only 1
() Heritage Reporting Corporation (202) 628-4888 I
4 l
107 I^')
V 1 within this 3-hour wall that we would have to take total 2 loss?
3 MR. McCRACKEN: Within that 3-hour area.
4 MR. MICHELSON: Okay. So, the area may not even 5 be enclosed as we are discussing and it is not yet even on 6 new plants required to be enclosed, as I understand what you 7 said. Is that correct?
8 MR. McCRACKEN: No. Anybody who would come'in and 9 wants to have a non-closed area is going to have to 10 demonstrate how they can control fire or smoke or heat from 11 that area to not affect any other area.
12 MR. MICHELSON: Okay.
13 DR. CATTON: That's a complicated problem. How (g 14 are you going to decide? You know, it's not like a big open V
15 area with a heat source sitting out in it which COMPBRN or 16 some of those codes will handle. This is difficult L;cause 17 you have recirculation between the top of the wall and the 18 roof. How are you going to decide?
19 MR. McCRACKEN: Well, number (1), I don't think 20 you are going to have a lot of areas like that located in a 21 large area where you have got a big overhead ceiling. You 22 aren't going to have large storage tanks put there because 23 it is a dumb place to locate them.
24 MR. MICHELSON: You mean in future plants.
25 MR. McCRACKEN: Yes. It is easier to locate them
() Heritage Reporting Corporation (202) 628-4888
i.
108
() 1 elsewhere than to address that issue.
2 DR. CATTON: I don't understand the answer, but 3 .that's okay.
4 DR. KERR: Is this going to be accomplished by 5 some new regulation?J 6 MR. McCRACKEN: We have told all of.the applicants 7 for construction permits that they need to meet this.
8' DR. KERR: So, it is not a regulation.
l 9 MR..McCRACKEN: And two.of'the three have said 10 yes.
11 DR. KERR: 2o,'it's not a regulation. It is just 12 something they have been told to do.
13 MR. McCRACKEN: Right. We have.gone'beyond the fg 14 SRP.
V
'15 DR. KERR: .But you have also gone beyond 16; regulations.
17 MR. McCRACKEN: Yes.
18 DR. KERR: Without regulations. How do you have 19 authority to do this?
20 MR. McCRACKEN: They volunteered to do it. They 21 thought it was a good idea.
22 DR. KERR: Oh, come on. Come on. "They
- 23. volunteered."
24 MR. MICHELSON: There is a policy statement,'also 25 that talks to this.
Heritage Reporting Corporation (202) 628-4888
n 109
) MR. McCRACKEN: The policy statement addresses the 1!
-2. issue that the new generation of plants should be safer.
3 DR. KERR: Well, why not have a regulation'that a 4' sdys what they should do.
5 MR. McCRACKEN: Because by the time I get a
'6 regulation finished, they will be done and their application 7 will be approved. So, as part of the review process, we 8- .said, "This is what we want to see modified. These are the 9 areas that we find the vulnerabilities in plants that exist
< 10 today.
11 DR. KERR: But, see, the problem with this is a 12 problem just now raised by Mr. Catton. Unless you have
'13 thought the thing through enough to formulate a regulation, 14 somebody is going to come in with something that will be
[}
15~ difficult to judge and one review team is likely to judge it 16 in one way and one another so that you have inconsistencies 17 and you may not accomplish what -- even with a good 18 regulation, you may not accomplish what you want to 19 accomplish. And if you just have sort of an understanding, 20 you may miss the target completely.
21 MR. McCRACKEN: I don't think so.
22 DR. KERR: I know you don't think so. I am trying 23 to convince you that you may be making a mistake with this 24 approach.
25 MR. McCRACKEN: I have a schedule to complete the Heritage Reportinpr Corporation (202) 6?8-4888
i:
110 c( ) :L ' review.- If I stop;the~ review to go come up with'a new 2 regulation, _ we're not going to get them finished. ~That's my 3 alternative.
1 4 DR. KERR: Well, finishing the review is a 5 worthwhile objective, but making the review accomplish the 6 risk reduction that you'want to accomplish is also a 7 worthwhile objection, it would seem to me. And it appears 8 to me that both ought to receive consideration.
9 MR. McCRACKEN: The changes we.are talking about 10 here are not difficult changes when you are designing a 1 '. brand new plant.
12' DR. KERR: I don't know what changes I'm talking 13 .about. I don't have a regulation. I just have something 14 that says -- and I hm not even sure how it is being 15 interpreted either by the review or by the licensee. And lC the licensee certainly doesn't know how it is going to be 17 interpreted until he brings something in.
18 MR. McCRACKEN: That's correct.
19 DR. KERR: And you who may. understand it, are you 20 going to be doing the reviewing or is another group going to 21 be?
22 MR. McCRACKEN: No. My group is doing that 23 reviewing.
24 DR. KERR: Well, suppose you leave and go to work 25 for somebody else, will in%y understand exactly what you'had O rie e- aereres-9 correr eie-(202) 628-4888
h 111 l() 'l- in mind when you formulated this?
2' MR. McCRACKEN: Everybody involved in doing it 3 knows exactly what we are trying to accomplish. It.is in
-4 writing.
5 MR. MICHELSON: I have a couple of.other questions 6 on this same thing. Could you tell rme what you mean there 7 by total loss?
8 MR. McCRACKEN: Total loss means all equipment.
9' MR. MICHELSON: In particular,. keep in mind the
~10 concern that in the process of degrading during a fire a
~
11 piece of equipment may do unwanted things. Is that to be 12 included or do.you --
13 MR. McCRACKEN: Yes.
14 MR..MICHELSON: Do you just assume it does nothing 15 unwanted and simply ceases to function?
16 MR. McCRACKEN: No, no. It can do what you're 17 talking about --
18 MR. MICHELSON: Okay. It would include the 19 unwanted actions that might result from the degradation of 20 the equipment which is quite a bit different than assuming 21 total loss.
22 MR. McCRACKEN: You can have partial loss in the 23 fire area.
2
.4 MR. MICHELSON: Okay. Another question: The 25 3-hour enclosure, that's the boundary that the roof to the
() Heritage Reporting (202) 628-4888 Corporation
._ _ - _____________ __ 2
112 j ,,
( ) 1 floor, the sidewalls, and so forth, that is rated for x.
2 three hours to enclose fire, but it doesn't necessarily i 3 enclose smoke migration or even heat, perhaps, of lower 4 grade heat, migration beyond this boundary because the doors 5 and so forth are not necessarily rated in terms of anything 6 other than the ignition temperatures on the other side of 7 the door which generally is to 53-50 or something. That is 8 way above what we might worry about in terms of degrading 9 equipment on the other side of the door.
10 So, to what extent are we asking new plants that 11 they consider such things?
12 MR. McCRACKEN: That's why we put in that they 13 have to address the smoke and heat.
("g 14 MR. MICHELSON: Okay, even beyond the 3 -hour
(/
15 barrier?
16 MR. McCRACKEN: There are -- well, there are 17 3-hour barrier designs that you could put in that would also 18 provide for smoke control.
19 MR. MICHELSON: It could, yes. l 20 MR. McCRACKEN: The current ones don't, but there 21 are designs available that can, in fact, give you smoke 22 control.
23 MR. MICHELSON: That's kind of what I'm getting 24 at. Either they have got to show that it's confined within 25 the zone or they have to show that they have analyzed the
() Heritage Reporting Corporation (202) 628-4888
4 113
(, ,,) 1 consequences of migration, ss-2 MR. McCRACKEN: Correct.
3 MR. MICHELSON: The same question kind of goes:
4 We have these fire barriers, if they are rated for three 5 hours for fire, they are not necessarily environmentally 6 rated for to keep bad environments from moving through. And 7 I have in mirG there water, for instance, from suppression 8 in the fire. This water is getting on the floor and so 9 forth. Have we told the utilities or the prospective 10 utilities what they would have to do about the environmental 11 migration of bad things like water out of this 3-hour zone?
12 MR. McCRACKEN: That is part of what they are 13 supposed to be addressing is what occurs -- the information 14 notices that have gone out, they need to address as part of 15 what they are submitting.
16 MR. MICHELSON: One final question. Within this 17 zone, now, it looks like it is going to be kind of a tight 18 zone and, in fact, some of the layouts of the ABWR are 19 fairly small rooms.
20 MR. McCRACKEN: Yes.
21 MR. MICHELSON: How are we going to address the 22 requirement to look at the pressure buildup? A problem with 23 pressure buildup, are we asking them to look at pressure
, 24 t'uildup in the process?
25 MR. McCRACKEN: Haven't asked ther to.
() Heritage Reporting Corporation (202) 628-4888
_ _. - - _ ~ _ . .
114
~
7() 1- MR. MICHELSON: Are we assuming it-is a 2 non problem without analysis or what will be the approach?~
3 MR. McCRACKEN: We are assuming it isn't that 4 -significant a problem because the 3-hour barriers are based 5 on. testing in a fully involved fire with a large fire source 6 and we den't have a lot of those areas.
7 The majority of rooms and areas that contain or.
8 ~are: adjacent to safe shutdown areas are not:large fire 9 source areas in the context of testing a fire barrier.
10 MR. .MICHELSON: Are you saying that we have tested 11- the fire barriers under differential pressure conditions?
12 MR. McCRACKEN: No. I'm saying that the fire 13 source that you put.when you do a fire test to the stendard
'14 is a very large fire source compared to what you normally
)
15 have.
16 MR. MICHELSON: But it doesn't necessarily 17 establish any differential pressure on the penetration being 18 tested. If I've got a big enough induction system, I don't 19 build up any pressure. So, it is not a test of the ability 20 of the penetration under fire to withstand differential 21 pressure.
22 MR. McCRACKEN: No.
23 MR. MICHELSON: Now, the question is then is that 24 an issue that needs to be even looked at? The English did 25 look at it and made special provisions for it.
Heritage Reporting Corporation (202) 628-4888
-=____- _ _ .-
115 l ) 1- MR. McCRACKEN: The English looked at it'and made 2 special provisions and --
3 MR. MICHELSON: I guess we are saying that we just 4 dcn't think it is necessary in our country. Is that what we 5 are saying?
6 MR. McCRACKEN: We looked at their. provisions and 7 it looks as though what they have installed in our opinion 8 may well not accomplish what they think they are going to 9 accomplish anyway. It is going to take a fairly significant 10 stack to relieve the pressure they think they want to 11 relieve and get rid of the amount of smoke they are trying 12 to get rid of.
13 MR. MICHELSON: So, what does that say then? Is 14 it that the pressure build-up is really an imagination to
}
15 begin with?
16 MR. McCRACKEN: Oh, nu. Pressure build-up occurs
]
17 but the question is: Is the pressure build-up going to 18 degrade that fire barrier to any significant extent faster 19 than if you didn't have the pressure build-up.
20 MR. MICHELSON: Yes. And on what basis do we know 22 the answer? Have we done tests or something to get that 22 answer? And.I think not. We have not tested under 23 differential pressure --
24 1Et . McCRACKEN: I am not aware of any tests that 25 have shown 1-for-1, side-by-side, the difference between the i
l rx ?
(,) Heritage Reporting Corporation (202) 628-4888
- p. ,
i.. '116
- ) . .
J l 1 .two.
2 MR. MICHELSON: I don't know of any tests'ia which~
3 we have applied significant differential-pressure-to the 4 . barrier because we purposely in fact try'to, for other 5 reasons, we put in' big induction systems to. keep the heat 6 and the smoke away from'everything and then as a result, we 7 don't apply any differential pressure.
8 MR. McCRACKEN: Right.
9 MR. MICHELSON: So, I guess it.still leaves me 11 0 . cold, though. We recognize that pressure is building up,
'll but-we don't think it's a problem, but we don't have any 12 good reason to think it is not a problem.
13 MR. McCRACKEN: In the opinion of the. fire q
-14 experts, the fire protection engineers, the people who are' L15 involved in these fires in all'other areas, they haven't 16' .seen it for this type of application, where we are talking 17 3-hour barriers, the types of barriers that we have 18 typically at nuclear power plants.
19 MR. MICHELSON: We haven't seen it because they 20 . purposely avoided it during the testing.
21 DR. CATTON: Have they made a written critique of
- 22 the English position?
23 MR. McCRACKEN: No.
24 DR. CATTON: Something that somebody could look 25 at?
Heritage Reporting Corporation (202) 628-4888
117
- 1 ' MR.I McCRACKEN: No . ' We don't have a written 2 ': csitique of it. We had one'of our consultants review it.
3 DR. CATTON: Did'he write a report?-'
4 MR. McCRACKEN: No, he hasn't provided al written 5 critique to the best of my knowledge.
6 DR. CATTON: Is he going to?
7 MR. FLACK: We will be receiving a letterLfrom fjj 8 Steve Nolan at-SandiaLwithin the next week or so on.the J
Y' l 9 English' approach to fire protection.
10 I wanted'to remind you that we did raise this as a 11 generic issue and it is being looked at.
12 DR. CATTON: It was in that' blue,'the.NUREG from 13 Sandia. I thought it was fairly clear.that it was a
'( 14 problem.
15 MR. FLACK: Well, .they have raised it as an issue' 16 and that is why it is being identified as a generic issue, 17 as a potential generic issue. And it will be resolved.
18 MR. MICHELSON: But for our next generation 19- plants, we are going to ignore it because we don't think it 20 is, I guess, a serious enough or'a significant enough 21 generic issue to make any design provisions for it. Is that 22 our position?
23 MR. McCRACKEN: The position is that we didn't 24 have sufficient information to require new plants to do it.
- 25. DR. CATTON: But you have whatever the rationale Heritage Reporting Corporation (202) 628-4888 i
118
's that led'the English to do it.
( 1 i
'2 MR.. FLACK: I guess it is the rationale.that is 3 being questioned at this point.
4 MR. McCRACKEN: Yes. I don't think we necessarily.
5 have their-rationale norLdo we think we even necessarily
.6 agree with their rationale, that they have come up with a 7 system that is'really going to accomplish what they want to 8 . accomplish.
9 MR. MICHELSON: That part'I don't want to debate
~
10 because I don't know the answer to that either. But I do' I
11 .want to-debate why we are ignoring even the issue. Have we 12 made any calculations and'so forth? Have we'done any 13 testing to verify the effects of pressure build-up? Have we 14 done - anything to start to get an answer 1x1 pressure 15 build-up? Or are we so confident that it is a non problem 16 that we don't have to?
17 MR. McCRACKEN: The answer from the NRR point of 18 view is they meet the current regulatory standards in the 19 United States,_the current fire protection standards.
20 That's how they are tested.
- 21. We have found those standards adequate. If we see 22 something that tells us they aren't adequate, then we can 23 address it.
24 DR. CATTON: But you are not going to know unless 25 you do the test correctly.
Beritage Reporting Corporation (202) 628-4888
119
- - f5 .MR. MICHELSON:
ty 1 Unless you-have a fire.
! 2 DR. CATTON: If the test is -- and-I thought I-3- -heard earlier today that'the. test was inadequate in the way 4 it was< performed. And if.that.is the_ case, how are_you-5 going to know?
6 MR. McCRACKEN: You heard conjecture'fron some 7 people that it may be inadequate in the way it is performed.-
8 You didn't see any evidence that that'wes true.
9 DR. CATTON: Well, I didn't see any evidence that-
_10 it was correct. And there is a lot of evidence around'that 11 this kind of-fire testing is very weak.
12 MR. McCRACKEN: No, I don't think there is a' lot 13 of evidence around.
(} 14 DR. CATTON: There is a very nice article th'at you
- 15. ought to read,.being in the fire business, that is in 16 Scientific American, written by Carrier maybe 10 years ago.-
17 It shows.a random number generator.can rank materials just 18 as well as the testing programs in the various countries.
l 19' That, to me, is a strong indication of weakness.
20 And how can you argue otherwise?
21 MR. McCRACKEN: The fire barrier testing is not 22 designed to come up with an exact 3-hour barrier. A 23 3-hour rating does not tell you that that 3-hour barrier i
24 will last for three hours for all fires you can postulate.
25 There are many fires that will last longer and there may be i_ Heritage Reporting Corporation L (202) 628-4888 1
120
()[
~
1 some types of fires that'will'last shorter. Three hours is
, 2 not a magic number. In any' accident sequence you~have or-3 scenario, the average response time to'get a fire brigade i
4 there is about:20 minutes.
l 5 MR. MICHELSON: The fundamental question is: What 6 do we know, if anything, about the behavior of the barrier 7 under' differential pressures? What kind of tests have we 8 ever done that tell us this at all? What'other kinds of 1'
- 9. work has been done that leads us to' conclude that it is'a 10 -non problem or sufficiently non-problem not to incorporate 11 in design. Because, to my knowledge, we just don't have any 12 test data on differential _ pressures applied to these 13 barriers under fire conditions. If we have,.I would like to-14 hear about it. I would even like to know'what the static 15 rating of the barrier is under differential. pressure with'no 16 fire on either side. How much differential pressure'will 17 the barrier take? Have you ever tested a barrier to see if 18 three inches of water blows it out? We are talking about 19 fire barriers, not the walls or the penetration barriers.
20 'There are lots of test on the concrete.
21 MR. McCRACKEN: The current plans with Haline 22 systems and CO2 systems, when they go to automatic 23 suppression testing, wind up pressurizing those 24 compartments. I have forgotten what the actual pressure 25 they get to is when they do it.
Heritage Reporting Corporation (202) 628-4888
121
(') l' MR. MICHELSON: You mean the rooms that they're 2 in?
3 MR. McCRACKEN: Yes.
4 MR. MICHELSON: And what pressures -- that would 5 be one potential interesting number.
6 MR. McCRACKEN: I forgot what the design pressure, 7 but in no cases where they.have done those testings and they 8 are required to test them, have we seen any evidence of I
9 losing seals as a consequence of that pressure test.
i 10 MR. MICHELSON: I would like to know what that 11 pressure.is-that it got'to and then it might be a 12- .potenti. ally useful' data, yes.
13 MR. LAMBRIGHT: I am aware of one test that -- a
-14 pre-operational test at Three Mile Island where they set off
.(~y}
15 their CO2 system and actually blew out fire ~ doors.
16 MR. MICHELSON: I know of one case when it blew
- 17 out'a fire door. It didn't take a whole lot of pressure to 18 do that, either. And they are worried now about the 19 electrical penetrations as well. How much can they stand 20 from whatever evidence that we've got.
21 I kind of suspect your CO2 releases and your 22 Haline releases are not giving you elevated conditions where 23 they have, in the case of TVA, blew out a set of fire doors 24 trat way. Yes, indeed, it only takes less than pound 25 internal. pressure.
1
()
Heritage Reporting Corporation (202) 628-4888
. _ - _ _ - _ _ _ _ _ _ _ _ . _ _ _ _ . . _ _ _ _ . _ _ _ = _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ - - _ _ _ _ _ - _ _ _ - _ _ _ - _ _ _ _ _ _ _ _ _
,t' I
122-L ?% .
- j. ,)!
, 1 1 Now, .the English pre' dict;in their calculations )
72' such pressures. easily from a fire. . And-have;we doneiany 3 ~ predictions of fire pressure build-ups of some sort that 41 :say, "Oh, it will never be more than'a few inches of water'
- 5 cpressure. "
6' MR. McCRACKEN:' No.
7 MR. MICHELSON: Or certainly well less than a
'8 pound.
9 If we haven't even done that,cthen-we haven't.even-10 thought'about the problem, have we?
11' MR. McCRACKEN: No , I don't' agree., We have not 12 seen a fire in roughly 1000 operating' years of. nuclear power 2 13 plant experience.in the U.S. thaa has given us'that problem.
L14
]p So, to say that we haven't thought about it is 15 ' incorrect.
16 17 18 19 20
'21 22 23 1 -
24 25 Heritage Reporting Corporation (202) 628-4888 !
m 123
, .1 MR. MCCRACKEN: We understand that all of the
'2 answers'are not in.' We understand that'there are l 3 uncertainties. But if we wait to get-all of the J
~
4 uncertainties resolved, we are never going to do the 1ook.
5 .And we think that there is a lot.more'to be gained by.
1 L' 6 looking with the knowledge that we currently have than to 7 wait around longer _trying to come up'with more answers on 8 how to-look better.
9 We believe~that by doing the examinations'and g 10~ doing a cross-check of the findings across all of the plants 11 that the really significant vulnerabilities that exist will ,
121 be. identified and we will be able to get them corrected l 13 where they need to be corrected. , i 1 14 The staff is working with NUMARC and EPRI to 15- Ldevelop an acceptable methodology. Our goal is that by the 16 iend of calendar 1989 to have that close to completed and to 17 be used on a' test BWR and PWR by 1990.
18 .The types of' methodologies that will be available 19 will be tile NUREG 1150 type Sandia methodology for plants:
20 that use PRA which is the'one that we have just gone 21 through, or an alternative methodology compatible with 22 enhanced IDCOR and the others that are going to be permitted 23 by the IPE generic letter.
24 We have agreed with industry that the applicable 25 fire risk scoping study results will be included in the Heritage Reporting- Corporation (202) 628-4888
.i l
124 !
l There will be means.there as to how'they
~
( ) 1 methodology.
]
2 should be addressed.
I 3 For the analytical codes, the opinion is'that they j l
4 are adequate to search. The intent is to build reference 5 tables that would show typical. types of compartments.' And i 6 these tables vill be built using a combination of probably 7 the'COMPBRN III and the. French MAGIC code combined with.
8 engineering judgment to tell you which components you need 9 to protect, where they are, what types of conditions that 10 you will get to. And the analysts can then simply go to a 11 table to look up where he is' going to.be.
12 Again we know that there are uncertainties in 13 here, but there would be uncertainties in the code if we
.f 14 developed it for an additional two years.
15 DR. CATTON: What kind of code is this MAGIC?
16 MR. MCCRACKEN: The MAGIC code is similar to what 17 COMPBRN III does.
18 DR. CATTON: Which means that it really cannot 19 handle compartments very well.
20 MR. MICHELSON: Not very well.
21 MR. CHEN: The MAGIC code itself is similar to a 22 zone model type of code, is similar to the Harvard code.and 23- the COMPBRN. Essentially it is one dimensional. But the 24 MAGIC code have some slightly greater, it does cover certain 25 things about migration from one room to the other room.
() Heritage Reporting Corporation (202) 628-4888 i
125 1 DR. CATTON: Do you have anything written on
[J)%
! 2 MAGIC?
3 MR. CHEN: The French ~has those written. And we 4 have a brochure that just stated what capability they can 5 do. 3 6 DR. CATTON: So you do not know really what its 7 internals are, you just have a description?
8 MR. CHEN: They have a description, and they also 9 use that to test the HDRs, the German reactor. They use 10 that and they use a small fire. And in a few cases they use 11 that to test that or verify how you can predict and what 12 sort of value you are going to get using that code itself.
13 They have this report indicator.
14 DR. CATTON: Did you send us a copy of that 15 report?
16 MR. CHEN: Pardon.
17 DR. CATTON: Could I have a copy of that report?
18 MR. CHEN: I can make a copy for you.
19 DR. CATTON: Good. Thank you.
20 MR. MCCRACKEN: Seismic fire interactions will be l l
1 21 handled by a procedurally directed walt.down. l 22 MR. MICHELSON: Before you get to that, that first i
23 bullet there on applicable analytical codes and so on, are 24 you going to end up with some kind of environmental map, is 25 that what you are saying, that the people walking down will
() Heritage Reporting Corporation (202) 628-4888 l
l l
t
Li .
() 1 have an environmental map that says this area is at such'and 2 such a temperature and humidity?
l l 3 MR. MCCRACKEN: It would take a room and say this 4 size room and say if you have a fire source of this size and 5 it is burning that these are temperatures that you could 6 ' reach over on that side.
7 HR. MICHELSON: It tells.you more than' temperature 8 I hope. It has to tell you part of the combustion content 9 so you know the effect of equipment. You know, certain 10 electronic devices do not like a whole lot of smoke 11 circulating'through them. So you-have to know the smoke 12 composition and you have to know the water vapor 13 composition. Because if you are fighting a fire in one 14 corner, you are going to transfer a. hell of a lot of steam
)
15 around the room. I mean that water evaporates the steam and 16 results in a very high water burden including water-17 droplets. That is traditional for a fire.
18 Are you going to include that sort of thing, the 19 environment around the room?
20 MR. MCCRACKEN: It is going to tell you only the 21- heat because that is what the available generation of codes i
22 tell you. l 23 MR. MICHELSON: I see.
24 MR. MCCRACKEN: You cannot go beyond what the 25 available generation of codes have for you.
() Heritage Reporting Corporation (202) 628-4888
127
(), 1. DR. CATTON: Are they not typically, I mean it is
- 2. a single stratified layer, is it not, it is hot,and below it 3 there is a step change to whatever the temperature was in 4 the room.before the fire started?
5 MR. FLACK: Well, COMPBRN breaks into.a hot gas l
6 layer and a plume and then you have the' cold layer below.it. .
7 DR. CATTON: Yes, but that will not deal with the ;
8 kinds of questions that Carl is raising where maybe a' piece !
9 of electronic gear where 250 degrees will wipe it out, j 10 because it'does not do that. You do not have 250 degrees.
11 You either have the' flame temperature or you have the room:
.12 temperature.
13 MR. MICHELSON: You have got to have the direct !
14 radiation component in there also.
15 DR..CATTON: Well, you need more.than that.
16' Because it is not a step change in temperature but it is a .i 17 gradient. And the plume feeds into a distribution but not !
18 into a step change like you find in COMPBRN. !
19 Is MAGIC built the same way?
20 MR. CHEN: ' MAGIC is more or less the same as 21 COMPBRN.
22 DR. CATTON: If it is the same as COMPBRN, you are 23 really not going to be able to deal with some of these ,
24 questions with that code. There are others that could do ;
I 25 it, l
() Heritage Reporting Corporation (202) 628-4888
p '
I 128
- e
'() l' MR. MCCRACKEN: We are not trying to' deal with.it 2 with'the code. .We are using the code plus engineering 3 judgment lon.when we are going.to get into. problems, and we.
4- are going to use available fragility data.that already
- 5 exists on apecific types of equipment or components to make 6- an assessment.with the code and an. engineering l analyst, the-7 equipment person plus the fire person to makeLan assessment 8- as to whether we think that component is in danger and put-9 out a code or a set of tables to tell us when you get into 10 trouble based on the type of room you have and the source 11 that you have.
12 DR. CATTON: I would be very. interested'in seeing 13 'this process as it moves along.
{) 14' 15 MR. MICHELSON: Are we somehow accounting also for the condensation of water vapor and so forth in the far 16 reaches of this room-from the fire and the fire-fighting in 17 the one portion?
- 18 MR. MCCRACKEN: No.
19 MR. MICHELSON: Do we think that water 20 condensation is not a problem for equipment?
21 MR. MCCRACKEN: Water condensation may be a 22 problem for equipment.
23 MR. MICHELSON: Somehow this environmental map 24 that I am going to give these walk-through people is going 25 to have to describe what the environment is. Including if Heritage Reporting Corporation (202) 628-4888
k_-
iI 129-
{): 1 it is-condensing moisture at the time, that has to be 2 included. -
3 MR. CHEN: Let me provide some clarification.
^
4 Really we are not going to get that. The condensation and 5 all of these things will not be there. I think essentially 6 what this would-develop is a calculation put in with a lot
'7 of input data which is supposed to be used for.COMPBRN. And 8 you can develop or maybe you consider certain factors like 9 how far is the distance away from a source and how large is 10' the source, and how far away is your cable and what is the 11 capability of the cable itself to see it whether it is 12- ignited or not. But this is just without the suppression 13 system getting too involved.
( 14 MR. MICHELSON: You realize of course that the.
15 ignition is not really the concern here. It is the 16 operability of equipment that is the concern. And 17 unfortunately, many times I read these articles and they 18 keep saying that we do not reach ignition temperature so 19 many feet away and that is immaterial.to the operability.
20 MR. MCCRACKEN: We are not looking for ignition 21 temperature. We are looking for degradation of the ability.
22 MR. MICHELSON: We are looking for whatever goes 23 on that destroys the environment in which the equipment 24 might have to function properly.
25 MR. MCCRACKEN: I think that I need to go back to
() Heritage Reporting Corporation (202) 628-4888 i
l
l 130
. (.\
i) I the initial number one priority which is to get the 2 examinations conducted to identify vulnerability. We know 3 that there are uncertainties. We know that all of the 1
4 "ts" are not crossed and the "is" dotted. But if we wait 5 around to do that we are never going to do the search.
l l 6 MR. MICHELSON: What is the search going to me .1 7 if we do not do it?
8 MR. MCCRACKEN: It is going to mean that we will 9 find that a large majority of vulnerabilities exist which is 10 better than waiting around forever.
11 MR. MICHELSON: I was hoping that was done when 12 you did the Appendix R walkdowns that you found those kind 13 of vulnerabilities.
(} 14 MR. MCCRACKEN: No. Appendix R did not consider 15 the issues of one system out of service, the issues that are i
16 being identified now through the PRA. It was not a PRA 17 analysis. It was just a deterministic of go protect them if 18 they are close together.
19 MR. MICHELSON: I thought this perhaps was not a 20 PRA analysis either depending on whether you use IDCOR or a 21 PRA.
22 MR. MCCRACKEN: This is a systematic system based 23 approach. It could be an IDCOR type, but it would still 24 have a methodology to it where you systematically go through 25 the systems and components that you have to have.
(m_) Heritage Reporting Corporation (202) 628-4888 >
1 131
, . 1 .MR. MICHELSON: I. thought'.that.was Appendix R.. I ;
i 2 have read these big volumes of Appendix R with all of this f.
3 systematic approach to finding what were the problems and 4 whether you had to put in fire barriers.and all of that.
5 That was looking at the equipment needed for safe shutdown 6 .and.how you handle it. I thought that caught the majority 7 of the problem and I thought that we were dealing hopefully 8 with more of the fringes of the problem here.
9 MR. MCCRACKEN: We are dealing here more with a PR 10 based methodology to find out where interactions are 11 1 occurring that Appendix R would never have picked up.
12 MR. MICHELSON: 'Yes, that is right. And I.am not 13 sure what'is being proposed so far is going to find much 14 more than we already=know, simply because it is ignoring
(}
15 most of the mechanisms.
16 MR. MCCRACKEN: No, I do not think so.
17 MR. FLACK: It is hard to think that generic 18 compartment temperature profiles will adequately address 19 potential risk points.
20 MR. MICHELSON: Certainly the starting point of 21 understanding what your problem is is to know what the 22 environment for the equipment is.
'23 MR. LAMBRIGHT: Specific equipment for cabling may 24 be located in a zone where the fire necessarily has to 25- probably be located.
Heritage Reporting Corporation (202) 628-4888
132 p~-
i j/. 1 MR. MICHELSON: I_ guess you must have missed the 2 point then too. The fire does not.have to be where the 3 equipment is, but the fire must create an environment 4 adverne to the equipment. It can be in the other corner of 5 the room.
6 MR. LAMBRIGHT: That is right. It can be in the 7 other corner of the room. But typically in a problem 8 situation the fire will have to be located.
9 MR. MICHELSON: That is because typically you have 10 been using COMPBRN and it does not work after you get out a 11 few feet. So you say okay, it is a non problem because I 12 have not reached the ignition temperatures any further out.
13 And that is what the Sandia study did as well in many
{} 14 respects when it did look at some of this. Because that is 15- all they can do, and they used ignition as the mechanism of 16 deciding whether or not the equipment was lost. And I am 17 saying that ignition is not the question, but it is the 18 operability and the environment that is created. And it 19 could be well below ignition temperatures or way below.
20 MR. LAMBRIGHT: Both ignition and time and 21 temperature to predict damage to the cabling. ,
l 22 MR. MICHELSON: Oh, in cabling. Forget cabling.
23 That is not our worry.
24 DR. CATTON: You only have got two temperatures.
25 You have the high temperature and you have the room rx
(_) Heritage Reporting Corporation (202) 628-4888
1 l
l 133
( ,) 1 temperature.
2 MR. MICHELSON: Yes.
3 DR. CATTON: I do not understand where you are 4 coming from really. In this kind of circumstance, the 5- problem with COMPBRN is because the person who put it 6 together made an assumption about the hot layer and the 7 plume feeds to hot layer and that is it. If you look at the 8 results of studies of these kinds of things, you find that 9 there is a temperature distribution. The temperature 10 distribution is kind of similar.
11 Once you know what it is, you can have a COMPBRN 12 type model that is meaningful to the type.of study that.you 13 want to do, and it does not take forever to develop it. So 14 I do not understand your statement that you are just not 15 going to do it.
16 MR. MCCRACKEN: We have a schedule to get done 17 with IFEEE. We started doing research in this area in 1975.
18 There are still answers that we do not have. I have no 19 doubt that in 1995 that we will not have some answers.
20 DR. CATTON: But that is just because of the way 21 that you carried out your research program. The physics are 22 understood. There are people around who know how to do 23 these things.
24 MR. MCCRACKEN: I am sure that there are people 25 around who may know how to do it. The codes that I have
'(O
,j Heritage Reporting Corporation (202) 628-4888
.134
(). ~1 seengthat look close to what you are discussing, .the lhst I 2 heard we are talking about perhaps a year to eighteen' months 3 to:get a code like'that up and available and not necessarily 4- verified and a couple of million dollars. . I am not aure 1
'5- that what I will gain from that would necessarily verify the
, 6 expenditure. There'are not'that many locations in the 7 plant, there are not that many areas that you havesto look
- at using that type of analycis.
9, 'In the majority of the areas, you are not going to 10 have to use'that. You do not have a hypothetical component 11' sitting over here twenty' feet away. It has already been 12 protected. It is already wrapped because weLwrapped it-13 under Appendix R.
14 MR. MICHELSON: You used twenty feet under 15 : Appendix R. That is all that you needed.. You did not have 16 to wrap it.
17 MR. MCCRACKEN: Nineteen feet.
18 MR..MICHELSON: I think it says over twenty feet.
19 I think the real question of course is whether we'have the 20 right kind of guidance or sufficient guidance to help the y 21 people really look to what we think that they have.to look 22 for, or are they just going to do a walkdown because the NRC o 23 says that you have to do a walkdown.
24 MR. MCCRACKEN: No.
25 MR. MICHELSON: So clearly the guidance document Heritage Reporting Corporation (202) 628-4888 L - _ _ .- -_ _ - - - _ _ - - _ _ _ - - _ _ _ _ _ _ _
135 n I will be the one that we ought to critically review, and we 3 )
2 r.re looking forward to receiving it at such time that it is 3 ready and we will focus c7 the guidance document instead of 4 worrying too much more about it here. Because we are really 5 going to look hard at'the guidance to see whether it is 6 going to result in anything meaningful. Because it should-7 not be done if it is not going to add any more to safety.
8 MR. MCCRACKEN: I agree. If it is not a useful 9 product that is going to in fact identify the majority of 10 vulnerabilities then we should not be doing it.
11 MR. MICHELSON: So we will just waAt for the 12 guidance to see what it really consists of and what we think 13 it might accomplish.
14 MR. MCCRACKEN: Going through the rest of these, I 15 can go through fairly quickly on them. For the fire barrier 16 qualifications, we are saying that individual plants have to 17 justify their assumptions. When they make an assumption on 18 whether their fire barrier is going to work, they have to 19 show in fact that they have implemented the types of 20 resolutions necessary to address the issues that we have had 21 with fire barrier qualification.
22 There have been some problems with seals, and l 23 there have been problems with dampers. We have put out a l
l 24 number of information notices in that area. And they need 25 to justify in fact that they procedurally are maintaining
() Heritage Reporting Corporation (202) 628-4888 l --. _ _ . __ - - - _ - - _ _
136 these in accordance with their design, that they have some
( )- 1 2 means.of showing that the barriers wi13 work.
3 Manual fire fighting. effectiveness. Again that is 4 individual plant specific. 'They have to justify.that.they
- 5. ~can reach and suppress a fire that-they need to in any area L
6- where they say that they have to suppress a fire. They have 7 to.show through training.
8 :DR. KERR:- Did you say earlier,.I thougnt that I-9L understood you to say that on the average that fire crews 10 get.to a-fire within twenty minutes?
11 'MR.- MCCRACKEN: An average number would be twenty 12 minutes, yes. That is slightly misleading. That is for the 13 fire brigade to get there. The majority of plants when a.
14- fire alarm sounds sends an operator who is trained in fire
-15 fly 7 and he' can do local fire-fighting himself which the 16 operator puts out.a lot of the fires. He'then says hey I 17 need the fire brigade and the total time from. initial 18 contact is twenty minutes.
19 DR. KERR: So if he has a garden hose available 12 0 and he gets there in five or ten minutes he could handle l 21 most fires?
22 MR. MCCRACKEN: Yes. The majority of them are 23 handled by the first peracn arriving.
24 An item that you addressed a little bit early, 25 total equipment survival. I think that the punch line here Heritage Reporting Corporation (202) 626-4888 e__ _ _ _ _ . -_ _ _ - - -_ __ ._ _. __ __- _ - __ _ _ __ _ _ -
b 137 h 1c is that for Generic-Iss.ue 57 that-this resolution is being 2- coordinated'with this particular search. The area where'we 1F intend to try to address it is that in the systems walkdown
'4 .that we.will identify all areas that are necessary to 5 achieve safe shutdown and will identify where' automatic.
6 suppression' systems exist and identify those as potential 7 vulnerabilities as part of the seismic.walkdown.
8 Once they are identified, then as part of the c 9 resolution of Generic Issue 57, if there is an issue.with-V 10 those, we know where they are in each individual plant.and l
11 whether a vulnerability exists as a consequence of something 12 identified in Generic Issue 57.
13 DR. KERR: What do the last two "EEs" stand for?
, 14 MR. MCCRACKEN: External events.
15 .IMR. KERR: All right. Thank you.
16 MR. MICHELSON: When you do this walkdown, you 17 said something about seismic vulnerability.
18 Is that going to be included in this I}SEE 19 process?
20 MR. MCCRACKEN: Part of the seismic walkdown will 21 include certain things like fire systems that could come 22 loose. In other words, it is not a fire.
23 MR. MICHELSON: It is a separate walkdown from the 24 dire walkdown?
25 MR. MCCRACKEN: The fire aspects will be
() Heritage Reporting (202) 628-4888 Corporation l
.__________a
138 1
r~s i j 1 incorporat*d in the seismic walkdown. We are trying to 2 eliminate a lot of excess walkdowns.
3 MR. MICHELSON: So you are really doing a. seismic 4 walkdown when you look at some of these fire 1
.5 vulnerabilities, is that what you saying?
6 MR. MCCRACKEN: No. We are in the seismic 7 walkdown procedures going to include certain arpacts of 8 issues that are identified under fire, such as initiation of 9 automatic suppression systems, because it is not a fire.
10 MR. MICHELSON: What is puzzling me is the crew 11 that does the walkdown for seismic the same crew that does 12 the walkdown for fire?
13 MR. MCCRACKEN: Yes.
14 MR. MICHELSON: It is kind of a different b(~N 15 expertise that is needed by quite a bit.
16 MR. MCCRACKEN: Right. The two walkdowns may 17 occur simultaneously. All I am saying that the procedure or 18 the one that identifies look at sprinkler systems to see if 19 they are going to come loose and cause a problem would be on 20 the seismic side. The fire side would be going 21 simultaneously. There would probably be one walkdown. The 22 walkdown will include several different types of discipline 23 to accomplish the walkdown.
24 MR. MICHELSON: Will the seismic walkdown include 25 looking at the fire protection from the viewpoint of seismic
() Heritage Reporting Corporation (202) 628-4888
139
( ). 1 actuation?
2 MR. MCCRACKEN: Yes. Seismic fire and reaction 3 would be part of the seismic walkdown procedure.
4 MR. MICHELSON: Interaction was not the word but 5 actuation.
6 Are they going to look into the system to see if 7 they have mercury switches and so forth?
8 MR. MCCRACKEN: They are going to look to see 9 where those vulnerabilities exist like that particular issue 10 identified, yes.
11 MR. MICHELSON: And that will be under the 12 seismic.
13 MR. MCCRACKEN: That needs to be in the specific 14 procedures of how they do their walkdown.
15 MR. MICHELSON: All right.
16 MR. MCCRACKEN: There is an attempt to coordinate 17 these walkdowns so that we have fewer walkdowns.
18 The last item on control systems interactions. We 19 intend to have individual licensees certify that basically 20 they meet the regulations. They were told to have separate 21 shutdown panels. The issue raised by Sandia of if that did 22 not occur you may be in trouble is really a separation 23 issue. It is do you have clear separation so that a fire in 24 the control room area is not going to wipe out the panel.
25 There have been a couple of information notices on i) Heritage Reporting CorporatAan (202) 628-4888
_____________D
140
) 1 that particular subject. And we are asking.each. licensee to 2 'n i fact' verify that is something is. going in the control-3: room that it is independent of the' control room fire area 4 before it comes to the alternate shutdown panel. So it=
5 truly is an alternate panel.
6 MR. MICHELSON: That was just a'particular problem 7 with the design. That doas not get into the questions of 8 what happens when you have a fire in the control room fran 9 .the systems interaction viewpoint.
10 MR. MCCRACKEN: No.
11- MR. MICHELSON: It does not get into that?
12' MR. MCCRACKEN: No.
13 MR. MICHEL.90N: There was a particular_ unique 14 question about how you loop your wiring and that was handled 15 there.
16 MR. UCCRACKEN Yas.
17 MR. MICHELSON: AJ1 right.
18 MR. MCCRACKEN: The ACRS agenda items which I 19 notice was not attached to the agenda today. The majority I 20 have gono over somewhat but I will go over them a little bit 21 more I think if you want.
22 Regarding the 6/23/88 letter to the External 23 Events Steering Group from the Fire Subcommittee, the 24 question of fire size, growth and frequency, I think that I 25 have addressed that which is the fe.ct that we are going to
() Heritage Reporting Corporation (202) 628-4888
_ --___ _ __ _ j
l 141 m
() 1 use the best currently available data base. Right now we 2 have a recent update on the LERs that is either in or will 3 be in shortly.
4 4 And in addition to that, industry is going to go i 1
5 to their data base and provide an update before we get into 6' this so that we have current or as recently as we can have 7 it a data base on all fires that are applicable. This 8 particular item may be critical path on that 9/90 date that 9 we had for getting this done in the first two plants, the 10 BWR and the PWR.
11 The codes we just talked about. We basically 12 going to develop. tables to give the analysts an idea of 13 whether or not we have a vulnerability. And again the same r~T 14 issue on fire detection suppression times which are part of
\_)
15 the same issue which the code tells you when you can 16 suppress. We intend to use a lot of engineering judgment on 17 when you can get a fire brigade to a fire or how good 18 automatic suppression will put lu out.
19 Component damage thresholds and behavior. Again 20 in methodology we are going to base this on available 21 information that we have on sonne components that have been 22 tested and there is only a sn.all number.
23 MR. MICHELSON: What will do you do for he rest 24 of them? It is a very small number that has actaally been 25 looked at.
() Heritage Reporting Corporation (202) fi28-4888
142 1 MR. MCCRACKEN: Right.
2 MR. MICHELSON: What will you do for all of the
'3 rest?
4 MR. MCCRACKEN: You are going to have to make an 5 engineering assessment as to whether they a.e or not that 6 significant, and how much uncertainty we have in it, and is 7 it a critical factor. One way of having some of this 8 disappear is you simply look at it and assume it fails. And 9 if it fails and it does not cause you a significant event, 10 then you can make it go away.
11 MR. MICHELSON: But to determine that it does not 12 cause a significant event, you have to know what its failure 13 modes are in the presence of this disturbed environment.
14 MR. MCCRACKEN: You ha"e to consider the potential 15 failure modes which could be half open, all the way closed 16 or whatever.
17 MR. MICHELSON: No. There is a lot more to it 18 than that. It is not quite that simplistic I do not 19 believe. But you do have to have some kind of experience 20 base or test base or something, some idea of knowing l
l 21 electronic equipment behaves under elevated temperature for 22 instance in order to evaluate what the fire in a room might 23 do to that ptrticular electronic equipment.
24 How does one do that in this case?
l 25 MR. MCCRACKEN: I thin that there is data
() Heritage Reporting Corporation (202) 628-4888 l-
143 I
() 1 available onLthat.
2 MR. MICHELSON: You think that you have data so i 3 you know what elevated temperature in a room will do? ;
I 4 MR. MCCRACKEN: I think there'is data available on 1 5 telephone company fires that show what happens to certain l4
{'
6 types of electronic switch gear when you get heat or fire 7 around it.
f 8 MF. MICHELSON: I think that I can cite you such 9 data as well. The question is that we are not talking about
.10 the equipment that is being engulfed in flames because that 11 is tough. We are talking about the equipment located in the 12 room but not engulfed in the flames but seeing elevated 13 temperatures.
14 What does it do, solid state control devices 15 opposed to modest increases in temperatures, what do we 16 expect?
I 17 MR. MCCRACKEN: That is part of what we'need to i 18 put in the methodology.
19 MR. MICHELSON: I will see that when I get the 20 guidelines then as to how you handle that, is that right?
21 MR. MCCHACKEN: Yes.
22 MR. MICHELSON: All right.
23 MR. MCCRACKEN: On the next @st of ACRS agenda 24 items on the back of that, I think that all of those have 25 been included in the previous discussions which were control
() Heritage Reporting Corporation (202) 628-4888 l
. 144 1 systems interaction, seismic fire, manual fire fighting
[v)
2 effectiveness, equipment-survival, and fire barr'..er 3 adequacy.
4 MR. MICHELSON: One of them that we did not ;.
)
5 discuss is the fire fighting effectiveness from the 6 viewpoint of the. people doing the manual mitigation and )
7 spraying the wrong equipment or spraying adjacent equipment 8 which turns out to be the wrong thing to do. How are we 9 evaluating that sort of thing or the potential for it? I 10 believe that it is required under Appendix R.
11 MR. MCCRACKEN: Right.
12 MR. MICHELSON: That when you are using twenty 13 feet of separation that you have to show that the fire 14 mitigation on one side is not affecting the other side 15 adversely.
16 MR. MCCRACKEN: That is already addressed under 17 Appendix R. And the intent here is that we will have them 18 verify when they find an area where that is a problem where 19 you have a given piece of equipment that fire brigades are 20 in fact trained to know that that is an area where they 21 cannot be spraying if they have a fire in another area.
1 22 MR. MICHELSON: Under that same requirement of i 23 Appendix R, you also have to show that the fire suppression 24 being set off at the twenty feet away location does not 25 adversely affect the ability for hot shutdown. .
I r~%
() Heritage Reporting Corporation (202) 628-4888
____-___A
c l ? -
- f. 145
( l How'is that.done presently'by utilities?
'2 MR. MCCRACKEN: The; current analysis ~showed it
'3 aimply-by the spray pattern you used, that your equipment 4 was outside of the spray pattern.
,; 5 MR. MICHELSON:- Well, no. My equipment is in the 6 spray pattern train two. The fire'is in train A'and I am 7 going to get smoke over into train B, set off the fire 8 protection system and spray train B like there were a fire 9- 'there but there'really is not.
10 How do we analyse the consequence of that?
11 MR. MCCRACKEN: That is Generic Issue 57.
12 MR. MICHELSON: No, no. That is an Appendix R' 13 requirement. That is not a generic issue. Appendix R says 14 that you show the effect of the mitigation equipment in that 15 area and its effect on hot shutdown, not in the area with 16 the fire. You go back and read it. It clearly states that.
.17 you have to consider that. It is one of the few places 18- where it was clear.
'19 MR. MCCRACKEN: It was clear, but remember that 20 Appendix R only applied to a certain subset of plants.
21 MR. MICHELSON: Well, yes. I am going to argue 22 which subset even.
23 How do the utilities handle that problem when they 24 come up to it?
25 MR. MCCRACKEN: I cannot tell you how they did the
() Heritage Reporting Corporation (202) 628-4888
146 m
) 1 analysis. I did not review any of those.
. (G 2 MR. MICHELSON: To that analysis, again I thcucht 3 that I would need some kind of a code that.shows either I 4 did not set off the fire protection from the fire on the 5 other side of the room or alternatively my equipment was all 6 well protected because of the spray.
7 MR. MCCRACKEN: I know in some cases they did put i
8 splash shields on equipment, but I cannot say that that was 9 the purpose for putting it on. I think that is why they did 10 it, but I cannot say for sure that that occurred.
11 Again the continued agenda items and uncertainty 12 in analytical tools. Each IPEEE submission to include'a 13 thorough discussion of uncertainties. Bottom line numbers 14 are used only to provide discipline in the process. There 15 where they make an assumption they have to tell us why they 16 used an assumption. We intend to limit the majority of 17 those by hopefully having the guidance clear enough that 18 they use only the assumptions that are within the guidance, 19 or if they deviate they have a very good reason for 20 deviating from that guidance.
21 The next question was you were looking for some 22 specific guidance from the consultant. Thir was on a 23 simplified methodology that we had had him work on. This 24 was simply a straw man to focus on the issues. We do not 25 intend to finalize it and it is not going any farther. It
() Heritage Reporting Corporation (202) 628-4888
E ..
147
(): l' is not part of where we intend-to be goins. ..It was to make
- 2. sure that we had a. discipline process to focus on all of the 3 issues that we needed to identify.
4 How will licensees consider the above ideas that 5 we have been discussing, and again that guidance will be 6 incorporated into the methodology'on how to address each 7~ individual item and make it plant specific.
8 MR. MICHELSON: Is that methodology the one 9- scheduled for about the end of the. year, is that correct?
10 MR. MCCRACKEN: Yes.
11 MR. MICHELSON: Do we expect to see this in 12 December?
13 MR. MCCRACKEN: We expect to see a reasonably 14 detailed' subset of the whole thing in the mid September to 15 end of September meeting, and then go through some fairly 16 extensive comments on it. And hopefully by the end of 17 December we will have it. At that time we probably will not 18 have the updated data base nor will we have the tables 19' finalized. We will have a pretty good idea of how we are 20 going with the tables at that time.
21 MR. MICHELSON: The September meeting, what.is 22 that? I 23 MR. MCCRACKEN: The September meeting is a m'seting 24 between the Fire Subcommittee and NUMARC EPRI.
25 And again Appendix R effects. I think that sas
() Heritage Reporting (202) 628-4888 Corporation i
l
148
[u.-)' 1 discussed quite a bit already. Another item was the 2 12/28 letter from Gillespie to Beckjord which basically said 3 that the items identified in the fire risk scoping study 4 would be handled through the IPEEE process. We have 5 discussed that, and I think that it is clear in our 6 previously stated positions that that is exactly what we are-7 doing.
8 Regarding the Siswell B report on fire protection, 9 as I said we have looked at it. The consultant report will 10 be within a week or two I guess.
11 MR. MICHELSON: And we will get a copy of that 12 when that comes in.
13 MR. FLACK: I will forward you a copy.
rs 14 MR. MICHELSON: You are not looking at just the U 15 question of pressure buildup but the consultant is looking 16 at the entire report?
17 MR. FLACK: From what I understand they will be 18 considering all aspects of the report.
19 MR. MICHELSON: All right.
20 MR. MCCRACKEN: Based on the verbal comments that 21 I have received, they have gone through the whole report.
22 MR. MICHELSON: Because I am particularly 23 interested in their views on the infrared detection versus 24 products of combustion and thermal.
25 MR. MCCRACKEN: Their comment on the infrared l
/~T 5
j Heritage Reporting Corporation (202) 628-4888
g 149'
- 1 -detector.was that'a'similar' French detector does not work in.
-[v~Y 2' radiation environments.
3 MR. MICHELSON: I am not sure that that is the
~
4 answer that we need to hear.- There was a reason why you
~
5 need something better than products of combustion'and 6: thermal. energy, and infrared happens to be the best
- 7. available.
8 MR. MCCRACKEN: I was just giving'you the comment.
9 MR. MICHELSON: Many of these areas are notla 10 radiation question. Inside containment'it clearly'would be, 11 1 but in a lot of these other areas there is not even a 12 consideration.
13 MR. MCCRACKEN: That is the end of the 14 presentation.
(g f-
.)
15 MR. CHEN: Could I provide some clarification on 16 this point?
17 MR. MICHELSON: Yes.
18 MR. CHEN: We have identified the methodology for 19 the fire as a fire PRA. Currently what you hear is another 20 alternative methodology which will be developed by NUMARC.
1' 21 MR. MICHELSON: In the case of the fire PRA you I
22 have defined the methodology, but where have you defined 23 where the data will come from and where the models will come l
24 from to go into the methodology?
25 MR. CHEN: For the fire PRA currently we have 1
() Heritage Reporting Corporation (202) 628-4888 4
i
,1 150 l ~1 NUREG 1150' asian approach.- We also:have NUREG CR-2300'which, 2 describe the details about how to perform a fire.PRA,' lots 3' of' details.
f4. MR. MICHELSON: 'We have~ heard the shortcomings'of-5 all of ? this today.
6' MR. CHEN: The shortcomings will be. incorporated r 7 'into our guidance.
8 MR.-MICHELSON: For the'PRA? i
~!
'9 MR. .CHEN: For conducting IPEEE which will be-10 comparable.to NUREG 1335. So what is identified in the fire 11 scoping study of.how we.want them to address and how we'are
.12 going.to have integrated walkdown procedures.will be.
13 described in the NUREG document.
14 ' . MR '. MICHELSON: And that is the one out in-15 December?
~~
16 MR.'CHEN: Hopefully. As far as the computer.
'17 tools, this is probably the best available currently la the.
18 market. We are trying to issue.a guidance for.. people as to 19 what can be'used. It is the best available, what is the 20 best available and yet can serve our purpose. We do know 21 that there is a field code out there which is a three
- 22. dimensional code, field code. That can provide much better 23 capabilities, but we do not see that for our purpose that it 24 is necessary that for us to identify and to specify that 25 those are things to be used for the PRAs.
Heritage Reporting Corporation (202) 628-4888 >
s m.
'151'
~
[ 1 And we'also identified certain other codes. For 2 instance,.NBS has several codes, the Harvard ~ code and there 3 'is also.the Notre' Dame code all available. And some have in 4 a sense-been validated by applying _it to different kinds of 5 power plants like b'uildings and'these things.
6 MR. MICHELSON: Are you intending that the PRA' 7 consider the migration of heat and smoke from the fire 1 area-8 and to other parts of the room and elevate temperatures that 9 might affect operability?
10 MR. CHEN: We point out'that those are the issues.
11 MR. MICHELSON: That is an issue but what are you 12 asking them to do about it in their PRA because that may 13 ' determine whether you have to go to more exotic codes to get 14 decent answers way out from the. fire. Some will not do'it (J}
15 and some of the other zonal codes will not do it, but there 16 are ways or some of these mechanisms or codes that are 17 available might do it. I do not know. It has to go to 18 three dimensional sorts of things.
19 How do they decide to do that unless you tell them
'20 what'they have to consider in terms of environmental changes 21 in the area and that sort of thing, because that will 22 determine what codes you need?
23 MR. MCCRACKEN: There is no intention of going 24 that far or having them go that far. We think that the 25 vulnerability search can be conducted with what is available
() Heritage Reporting (202) 628-4888 Corporation ]
1 1
l
l- .
E !
--152 j 1
3 J; 1= going to the tables we said.
2' MR.~MICHELSON: Are there any remaining questions?
L 3 DR. CATTON: The position is certainly clear.
l 4 MR. MICHELSON: I think that is right 1 5- DR. CATTON: There io no difficulty understanding 6 it . - .
j 7- MR.- MCCRACKEN: Then I accomplished my purpose.
8 MR. MICHELSON: I think that if we have no other '
9 questions _for Mr. McCracken then I believe that we are 10- finished.- And we thank you for your presentation ~. I think 11 that the subcommittee needs to have a short discussion which' 12 .is the only thing remaining on the agenda now.
13 MR. MCCRACKEN: I did have one question, do we 14 need to be back here tomorrow?
15 MR. MICHELSON: That will be determined by our 16- subcommittee discussion, so hang loose. We will now finish 17 the transcript since there are no more formal presentations.
18 (Whereupon,.at 4:56 p.m., the hearing was 19 concluded.)
'I 20 21 22 23 24 !
25
() Heritage Reporting Corporation (202) 628-4888
M J
() .1 CERTIFICATE I 2
.3- This is to certify.that the attached proceedings before-the 4 United States Nuclear Regulatory Commission in'the matter 5 of: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS !
6 Name:
SUBCOMMITTEE ON AUXILIARY l 7 AND SECONDARY SYSTEMS 8 Docket Number:
9 Place: Bethesda, Maryland 10 Date: July 12, 1989 11 were held as herein appear:", and that this is the original 12 transcript'thereof for the file of-the United States Nuclear 13 Regulatory Commission taken stenographically by me and, 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.
18 /s/ '2h 1 -
'A- /dk 19 (Signature typed) : IRWIN L.
Cy-FENBERRY& I 20 Official Reporter 21 Heritage Reporting Corporation i 22 23 .!
24 25 Heritage Reporting Corporation (202) 628-4888 l
i . . _ - - _ - - - - - - - _
H C
R A
E G N S E
N I
O S R Y
F E M TH R E SC O I
E E 8 YN T LA A R
B T TS 8
9 AR L NB U E T Y I
1 A
E E G
YU R O H H MS , T S C T LI S R I
2 M 1 B TE A I
A NL R O OY I K L EC A T CR Y C E DU L A RI C N E B A U L F &C F S U L I
J .K AO E S I HSE RE X
I R N R EI C H .VF S U O J SSO REF E R A R
I C Y B
F A D E
T N
E O S E
R P
~. .
.O.
PURPOSE: To present the Staff's response to the recommendations reported in the fire risk scoping study.
REFERENCE:
Commission paper (6/7/89) entitled: FIRE RISK SCOPING STUDY:
SUMMARY
OF RESULTS PROPOSED STAFF ACTIONS bESENTATION OUTLINE:
Background
Purpose Findings Discussion Recommendations Staff Actions .
Summary O
1- - - - -
2
O ,
BACKGROUND l
FIRE RISK SCOPING STUDY PURPOSE:
- 1) To assess the risk significance and O dominant sources of uncertainty associated with fire at NPPs.
- 2) Assess the risk of potential fire related issues.
state-of-the-art methodology up to date data Appendix R modifications potential issues g
O FINDINGS 1:1 Confirms that fire is an important contributor to core melt at some NPPs.
- 2) State-of-the-art methodology have ,
modeling shortcomings which could result in large uncertainties.
- 3) Several important areas were identified O that are not adequately addressed in fire PRAs: )
i Fire induced alternate shutdown-control room panel interactions. I Smoke control and manual fire fighting i effectiveness. 4 Adequacy of fire barriers. 3 Equipment survival in fire induced environments.
seismi / fire interactions.
O 4 3
0- P STATE-OF-THE-ART METHODOLOGY ,
COMPBRN CODE:
Zone Model Tailored for NPP Fires
- Simplicty Highlighted O
Only Models Thermal Effects Several Different Versions Lacks Adequate Validation t
O ANALYTICAL TOOLS RECOMMENDATIONS:
- 1) Process and make available enclosure fire test data gathered under the old l NRC Fire Protection Program.
1 g Assemble and validate a physically. 1 consistent fire simulation model. {
- 3) Validate correlations for predicting cable tray flame spread rates.
flame heights.
threshold temperatures for ignition and damaging cables.
mass burning rates of cables.
STAFF ACTION: None proposed at this time O
_ _ _ _ - - _ - _ t
O. ,
j POTENTIAL ISSUES )
- 1) Control System Interactions
- 2) Smoke Control and Manual Fire Fighting Effectiveness g Adequacy of Fire Barriers
- 4) Equipment Survival in Fire Induced Environments
- 5) Seismic / Fire O
-. .__ _x
10 FIRE INDUCED CONTROL SYSTEM INTERACTIONS Three general types of control room i
. remote shutdown panel interactions were identified. j l
- 1) Loss of control power for .1 O perated devices prior to transfer
- 2) Spurious actuation prior to taking !
control from the RSP.
- 3) Total loss of component function due to redundant control equipment failure.
l
.O l e
O SYSTEM INTERACTIONS (CONTINUED)
Sensitivity studies indicated:
- 1) With no protection against ;
interactions, (i.e. they would occur), core damage frequencies could be above 1.0 E-4 /'f r.
g
- 2) With one method of shutdown operable, sequences could be above 1.0 E-5/yr.
RMIEP analysis of LaSalle had a sequence 8.6 E-6/yr.
O
---m_______.____.____________________m_ ___ , _ _ _ _ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _
Q-FlRE INDUCED CONTROL SYSTEM INTERACTIONS RECOMMEND.ATIONS:
- 1. Develop standardized guidelines for i analysis. l
- 2. Review adequacy of remote shutdown O implementation practices.
STAFF ACTION:
I Considering the issue within the IPEEE ,
framework. I l
Raised as a potential generic issue.
1 o
+
O SMOKE CONTROL AND MANUAL FIRE FIGHTING 4 EFFECTIVENESS Primary fire source-lubricating oils and q cable insulation, g Impact of smoke:
Reduces manual fire-fighiing effectiveness.
Cause misdirected suppression efforts.
Damage electronic equipment.
Cause evacuation of control centers.
Initiate automatic suppression systems.
Potentially impact an adjacent unit.
O
\\
s O
SMOKE CONTROL AND MANUAL FIRE FIGHTING EFFECTIVENESS RECOMMENDATIONS:
l
- 1. Review further plant practices and investigate the probable linkage between O plant practices and brigade effectiveness.
- 2. Refine methodologies for use in the evaluation of fire detection and suppression system adequacy.
STAFF ACTION: 1 Considering the issue within the IPEEE framework.
O -
Raised as a potential generic issue.
u
1 O
ADEQUACY OF FIRE BARRIERS ,
High barrier reliability is essential in ,
keeping the risk from fire propagation low.
I 1
Reliability questioned, specifically with respect )
to penetration seals. I COMMENDATION:
- 1. To assess the actual performance reliability of U.S. fire barrier elements.
1
. STAFF ACTION: !
Consideration within the IPEEE frame' work.
Raised as a potential generic issue.
O a
O EQUIPMENT SURVIVAL IN FIRE INDUCED ENVIRONMENTS Qualitative Evaluation Lack of Fragilty Data and Analytic Tools. t Plant Specific.
A) Potential Damaging Effects of Fire O suppression systems.
Frequency of inadvertent actuation equal to .14/yr.
1 out of 4 events either induced or resulted from a plant transient.
25% of the events resulted in degradation of one or more safety systems.
O l
l p- ;
_ 7
- O EQUIPMENT SURVIVAL B) Other aspects of a Fire Environment smoke low level thermal exposure Lacks data.
COMMENDATION:
- 1. Perform a more extensive review of the experience base.
Staff Action:
Provide input into generic issue issue 57.
i m
O
O SEISMIC / FIRE INTERACTIONS Interactions could be eliminated by performing plant specific evaluations.
Study provided a list of potential seismic / fire interactions, COMMENDATIONS:
- 1. Define and document potential vulnerabilities ,
of concern to NPPs.
STAFF ACTION: .
Consideration within the IPEEE framework.
Utilize Scoping Study information in performing the IPEEE.
O I
b
SUMMARY
A.. No proposed fire protection research at this time.
B. Consideration in the IPEEE:
Control Room-Remote Shutdown Panel Interactions
.O- Smoke control and Manual Fire-Fighting Effectiveness Seismic / Fire Interactions Adequacy of Fire Barriers C. Three potential generic issues raised.
I D. Information utilized in resolving Generic issue 57.
O 1
~
,s
%. /
l ACRS SUBC0mITTEE EETING ON THE AUXILIARY AND SECONDARY SYSTEMS JULY 12, 1989 EXTERNAL EVENTS FIRE SUB-C0FMITEE O aA m R m RT i
CONRAD E, f1CCPACKEN, CHIEF PLANT SYSTEMS BRANCH OFFICE OF NUCLEAR REACTOR REGULATION l
I 4
v ,
( .~ ' , . .
O l
STATUS OF COMPLIANCE WITli REGULATIONS APPENDIX R INSPECTIONS CONDUCTED AT ALL APPLICABLE PLANTS o
FINDINGS IN SEVERAL CASES WERE THAT IMPLEKEATION WAS l INADEQUATE. .TRSE WILL PECEIVE A C0fPLETE RE-INSPECTION' PRE-0PEPATIONAL INSWCTIONS CONDUCTED AT ALL PLAT (TS NOT
- UNDER APPEt0lX R COMPLIANCE REQUIRED PPIOR TO STARTUP g *
,TRI-ANNUAL AUDITS NOW Pelf;G IIREFENTED AT ALL PLANTS IN Stit1ARY, ALL PUWTS CURREMLY MEET THE REGULATIONS OR WILI DO SO SHORRY ,
PLANTS FOUND NOT TO MEET REGULATIONS WILL BE REQUIRED TO DO SO O
m_--__ _ _ _ _ _ ._ _ _ _ . _ _ . _ _ _ ______.-..__m. _ _ _ _ . . - . _ . _ _ . _. - . _ _ . _ _ _ . _ _ d
r~
\,J EFFORTS BEYOND THE REGULATIONS SEVERE PEACTOR ACCIDENT POLICY STATB1ENT:
EXPECTED TO PERF0Pfi A LIMITED-SCOPE ANALYSIS TO DISCOVER PARTICULAR VULNERABILITIES 4
EXTERNAL EVENTS STEERING CCMMITTEE (EESG) F0PK D TO PROVIDE DIPECTION ON IMPLEMENTATION OF POLICY Q
v EXTERNAL EVENTS FIRE SUB-CO MITTEE PROVIDES REC 0F70 0ATIONS TO EESG PLANT SPECIFIC VULNERABILITIES TO BE ADDPESSED THROUGH BACKFIT PROCESS O
E
O FIPE SUB-CLYNnTEE FINDINGS EACH OPERATING PEACTOR SHOULD BE EVALUATED TO DETEPHINE PLANT SKCIFIC WLNERABILITIES TO INTERNAL FIPES INDUSTRY PERFORWD PRAs DEMONSTRATE THAT FIPE CAN BE A SIGNIFICANT CONTRIBlff0R TO CORE FELT SOE WLNEPABILITIES IDENTIFIED THROUGH THE PRA PROCESS hERE CORRECTED PRIOR TO REPORTING FIPE RISK SCOPING STUDY RESULTS IfDICATE THE POTENTIAL FOR PLANT SKCIFIC WLNERABILITIES A LEVEL 1 PRA IS AN ACCEPTABLE ETHODOLOGY TO KRF0PN THE EVALUATION FOR FIPE WLNEPABILITIES n -
OTHER METHODOLOGIES MAY BE ACCEPTABLE BlfT PE0VIRE RIPTHER U DEVELOPMENT LIMITATIONS EXIST IN APPLICATION OF FIRE SPREAD CODES.
H0kEVER, IN THE FEW LOCATIONS PER PLANT WHERE THIS IS AN ISSUE ENGINEERING EVALUATIONS CAN BE CONDUCTED O
_ -- _ - - -- i
I ..
l L ,
b i
l FIRE RELATIONSHIP TO IPE FIRE DOES NOT CREATE A ND! ACCIDENT IT INCREASES THE PROBABILITY ,
ON EXISTING ONES. THEPEFORE, FIRE RESOLifTION MUST BE CCFPATIBLE ;
WITH PLANT SPECIFIC INTEPA'AL EVENTS PESOLlffl0N I
GL 88-20 PERMITS:
PRA ENHANCED IDCOR 0THER SYSTEMATIC METHODS FOUND ACCEPTAT1LE BY STAFF n
V O
(^)
v FIRE RESOLifTION FOR ALWRs PLANTS MJST BE CAPABLE OF SAFE SHUTDOWN ASSLEING TOTAL LOSS OF ANY ONE FIRE AREA, ASSLNING NO OPERATOR RE-EfffRY (PEACTOR CONTAlff0(T FilILDING EXCLUDED WITH SPECIFIC .iVIEWS) 20 FOOT SEPARATION CRITERIA IS ELIMINATED i
SM0KE MUST BE CONSIDEPED O
u__ _- -_ _
o s i6 COURSE OF ACTION FOR IDENTIFICATION OF FIRE VULNERABILITIES THE NO.1 PRIORITY IS TO GET EXAMINATIONS CONDUCTED TO IDENflFY VULNERABILITIES
]
FINDINGS WILL BE CROSS CHECKED
{
l STAFF IS WORKING WITH NtFARC/EPRI TO DEVELOP AN ACCEPTABLE I l
METHODOLOGY BY THE END OF C Y.1989 TO BE USED ON TWO TEST PLANTS BY 9/90 NUREG-1150 SANDIA ETHODOLOGY COMPATIBLE WITH PRA PLAKTS ALTERNATIVE METHODOLOGY COMPATIBLE WITH ENHANCED IDCOR AND OTHERS APPLICABLE FIRE RISK SCOPING STUDY RESULTS TO BE INCLUDED IN ETHODOLOGY
- ANALYTICAL ;0 DES: ADEOUATE TO SEARCH, REFEPENCE TABLES BEING b3 PPEPARED USING CCFPBURN III, MAGIC, AND ENGINEERING JUDGEENT
- SEISMIC / FIRE INTERACTIONS; PROCEDURALLY DIRECTED WALKDOWN
- FIRE BARRIER QUALIFICATION; INDIVIDUAL PLANTS JUSTIFY ASSUMPTIONS PANUAL FIPE FIGHTING EFFECTIVENESS: INDIVIDUAL PLANTS JUSTIFY ASSUMPTIONS
- TOTAL ENVIRONMENT EQUIPERT SURVIVAL; 1) COMBUSTION PFODUCTS A10 SUPPRESSANT EFFECTS NOT ADEQUATELY DEFINED 2) FIPE SUPPPESSICN SYSTEM EFECTS ON SAFE SHUID0kN EQUIPMEKT TO BE INCLUDED IN PLANT WALKDOWN. 3) GI 57 RESOLUTION IS BEING COORDINATED WITH IPEEE. i
- CONTPOL SYSTEMS If!TERACTIONS; INDIVIDUAL LICENSEES CERTIFY THAT THEY EET REGULATIONS AND INF0PNATION NOTICES 1
1 q
)
O ACRS AGENDA ITFFS A. PEGARDING 6/23/88 LETTER TO EESG 1A: FIRE SIZE, GROKiH, AND FREQUENCY LICEFSEES HILL USE BEST CURRENT DATA BASE PLUS PLANT SPECIFIC DATA WERE APPLICABLE; DATA BASE UPDATE MAY BE CRITICAL PATH MODIFIED COMPBUPN III OR EQUIVALEfff PLUS ENGINEERING JUDGEE NT TO DEVELOP TABLES 18: FIRE DETECTION AND SUPPRESSION TIES COMPBURN III OR EQUIVALBff PLUS ENGINEERING EVALUATION Ic: COMPONENT DAMAGE THPESHOLDS AND BEHAVIOR
- TO BE ADDRESSED IN METHODOLOGY, BASED ON AVAILABLE INFORMATION ID: PURT STATE NOPFAL OPERATION O
l7 1 1
O ACPS AGENDA ITEMS (CONTINUED)
A.2A: CONTROL SYSTEM IfGERACTIONS INCLUDED IN PREVIOUS DISCUSSION A.2.B: SEISMIC / FIRE INTERACTIONS INCLUDED IN PREVIOUS DISCUSSION A,2.c: FIRE FIGfrING EFFECTIVENESS .I INCLUDED IN PREVIOUS DISCUSSION A.2.D: EQUIPE?R SURVIVAL n -
INCLUDED IN PREVIOUS DISCUSSION
%)
A.2.E: FIRE BARRIER ADEQUACY INCLUDED IN PREVIOUS DISCUSSION O
o
'\ ;
ACRS AGENDA ITEMS (CONTINUED)
A.3 UNCERTAINTY IN ANALYTICAL TOOLS EACH IPEEE SUBMISSION TO INCLUDE A THOROUGH DISCUSSION OF LFCERTAINTIES. BOTTOM LINE NLPBERS ARE USED ONLY TO PROVIDE DISCIPLINE IN PROCESS-l A.4 SPECIFIC GUIDANCE FROM CONSULTANT THIS WAS A STRAW PAN TO FOCUS ON ISSUES AND WILL NOT BE FINALIZED A.5 .HOW WILL LICENSEES CONSIDER PLAAT S WCIFIC APPLICABILITY OF A.1 THROUGH A.4 GUIDANCE WILL BE INCORPORATED IN MET 10D0 LOGY.
A.6 APPENDIX R tittds APPENDIX R AND OTHER FIRE PROTECTION GUIDANCE HAS PADE PLANTS SAFER. SANDIA ESTIPATES UP TO A FACTOR OF 10.
HOWEVER VULNERABILITIES STILL EXIST l
.O ,
l
]
%d ACRS AGENDA ITD S (CONTINUEDI B. REGARDING 12/28/88 LBTER FP&, GILLESPIE TO BECKJORD THIS LBTER IS CONSISTEfff WITH THE PREVIOLGLY STATED POSITIONS At0 SECY 89-170 MlICH PROVIDES THE PES ASSESSMENT TO TllE C0ft1ISSION C. PEGARDING SIStELL B REPORT ON FIRE PROTECTION
%.)
O