ML20140H467
| ML20140H467 | |
| Person / Time | |
|---|---|
| Issue date: | 06/13/1997 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-3002, NUDOCS 9706180268 | |
| Download: ML20140H467 (143) | |
Text
.
Official Transcript cf Proceedings
.O NUCLEAR REGULATORY COMMISSION AcPS7 3o 02,
Title:
Advisory Committee on Reactor Safeguards 442nd Meeting TRO4 (ACRS)
RETURN ORIGINAL TO BJWHITE M/S T-2E26 Docket Number: (not applicable) Q}
- Location: Rockville, Maryland O
Date: Friday, June 13,1997 n nc q i 8 AulS u-.p c~e ipy - Re:w.
., ,m m f0FD8_Fe 01 ne u,o:trninae 1
Work Order No.: NRC-1135 Pages 425-533 61 8 970613 T-3OO2 PDR I
NEAL R. GROSS AND CO., INC. <
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
y ,__. .
7
( 95 Washington, D.C. 20005 g-' q l,!
(202) 234-4433 l . ;
i bJl ; ,,
Illl!!!O!I:llJi!Willl!11!!!I.Ill
i I
l f^)
..]
DIBCLAIMER l
PUBLIC NOTICE ;
BY THE !
UNITED STATES NUCLEAR REGULATORY COMMISSION'S l ADVISORY COMMITTEE ON REACTOR SAFEGUARDS j JUNE 13, 1997 l l
The contents of this transcript of the proceedings of the United States Nuclear Regulatory
[^T Commission's Advisory Committee on Reactor Safeguards on MAY
,O 13, 1997, as reported herein, is a record of the discussions recorded at the meeting held on the above date.
This transcript has not been reviewed, corrected and edited and it may contain inaccuracies.
l l
c NEAL R GROSS Cot'RT REPORTLRS AND TRANSCRIBERS 1323 R1IODE ISI AND AVENUE, NW (202)234-443; WASil!NGTON. D C. 2000$ (202)234-4433
425 l
l 1 UNITED STATES OF AMERICA
,- 2 NUCLEAR REGULATORY COMMISSION l
l V> 3 +++++
I i
l 4 442nd MEETING l i
l 5 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS) 1 6 +++++
7 FRIDAY 8 JUNE 13, 1997 I
9 +++++
i 10 ROCKVILLE, MARYLAND 11 12 The Advisory Committee met at the Nuclear 13 Regulatory Commission, Two White Flint North, Room T2B3, V 14 11545 Rockville Pike, at 8:30 a.m., Robert L. Seale, 15 Chairman, presiding.
16 COMMITTEE MEMBERS:
17 ROBERT L. SEALE CHAIRMAN 18 DANA A. POWERS VICE CHAIRMAN 19 GEORGE E. APOSTOLAKIS MEMBER 20 JOHN J. BARTON MEMBER 21 MARIO H. FONTANA MEMBER 22 THOMAS S. KRESS MEMBER 23 DON W. MILLER MEMBER 24
\ O\ 25
(_/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASH!NGTON, D C. 20005-3701 (202) 234-4433
426 1 ACRS STAFF PRESENT:
2 JOHN T. LARKINS Exec. Director
/,.s%
\. /
l 3 ROXANNE SUMMERS Tech. Secretary 4 SAM DURAISWAMY 5 CAROL A. HARRIS 6 RICHARD P. SAVIO 7 PAUL BOEHNERT 8 NOEL DUDLEY 9 MEDHAT M. EL-ZEFLAWY 10 MICHAEL T, MARKLEY 11 AMARJIT SINGH 12 j i
13 ACRS INVITED GUEST:
,r~
kJ 14 BOB UHRIG 15 16 ACRS FELLOW:
17 AUGUST W. CRONENBERG 18 19 ALSO PRESENT:
20 MIKE SNODDERLY 21 TOM KENYON 22 GARY HOLAHAN l
l 23 BRIAN McINTYRE 24 JOHN RIDGELY Ch
( ) 25 BOB PALLA NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS
[
1323 RHODE ISLAND AVE., N W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
427 l
1 A-G-E-N-D-A w 2 Acenda Item Pace I \
\"/ 428 3 Opening Remarks 4
5 Policy Issue Pertaining to AP600 Design 6 John J. Barton 430 7 August W. Cronenberg 431 8 Michael Snodderly 436 9 Brian McIntyre 507 10 11 12 13 f
w .,
i :
'k J 14 15 I
16 l 17 18 19 20 21 22 23 24 f M. .
't] 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C, 20005 3701 (202) 234-4433
428 1 P-R-O-C-E-E-D-I-N-G-S gS 2 (8:35 a.m.)
() 3 CHAIRMAN SEALE: The meeting will now come to 4 order. Congratulate those of us that were with us last j 5 night visiting with Commissioner Rogers, on your hale and l
6 hearty appearance this morning. On the other hand it l
7 appears that John Barton, who spent the night vorking on 8 some preparation for the first report this morning must 9 have -- he had a rougher night than the rest of us did.
10 MEMBER BARTON: Cheaper though, cheaper.
11 CHAIRMAN SEALE: Well, your dedication is 12 truly noted, or duly noted.
13 Our meeting will now come to order. This is 14 the third day of the 442nd meeting of the ACRS 15 Subcommittee on Reactor Safeguards. During today's 16 meeting the committee will consider the following:
)
l 17 1) Policy issue regarding containment spray )
, l l 18 system for the AP600 design; l l
l 19 2) Report of the planning and procedure )
i 20 subcommittee; and i
21 3) Proposed ACRS Reports.
22 Looking over there at the board, I see there's 23 still quite a few, and they are of mixed maturation. So 24 we've got our work cut out for us today.
(-
'\,_) 25 A portion of today's meeting will be closed to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBEh3 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON D C 20005 3701 (202) 234-4433
l 429 l
- 1 discuss organizational and personal mattars that relate 1
l l f- 3; 2 solely to the internal personnel rules and practices of 1
\J 3 this advisory committee, and information the release of 4 which, would constitute a clearly unwarranted invacion of 1
5 personal privacy.
6 This meeting is being conducted in accordance 7 with the provisions of the Federal Advisory Committee Act.
8 Mr. Sam Duraiswamy is the designated federal 9 official for the initial portion of the meeting.
10 We have received no written statements or 11 requests for time to make oral statements from members of 12 the public regarding today's sessions. A transcript of 13 portions of the meeting is being kept, and it is requested
/ ,,N
( 1
\ 14 that the speakers use one of the microphones, identify 15 themselves and speak with sufficient clarity and volume so 16 that they can readily be heard.
i 17 We do have a multi-faceted issue this morning I 18 to discuss. The first thing is the policy issue 19 pertaining to the AP600 design on sprays, and John, you're 20 the chairman of that subcommittee, and so I'll turn it 21 over to you.
22 VICE CRAIRMAN POWERS: John, before you start 23 -- and I think it's important for the committee to know --
l l
l 24 that I will recuse myself from this discussion. I have
()
(_,, 25 had some personal involvement in at least some of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l l
430 ;
I
! 1 ancillary issues associated with this particular question.
( fs 2 So I think it's best if I sit here and stay quiet.
I \
G 3 MEMBER BARTON: That's hard for you to do.
4 VICE CHAIRMAN POWERS: It's going to be a real 5 -- this one is going to be very tough to be quiet on. But 6 I'll do my best. Thank you, Mr. Chairman.
7 CHAIRMAN SEALE: We'll continually scan a 8 smile or frown on your face for meaning.
9 VICE CHAIRMAN POWERS: I would not search for 10 meaning there.
11 CHAIRMAN SEALE: Proceed.
12 MEMBER BARTON: Thank you, Mr. Chairman. The 13 purpose of this session is to review and comment on the n
--) 14 proposed staff position that a containment spray system or 15 equivalent, be incorporated in the AP600 design in order 16 to provide additional, accident management capabilities 17 following a severe accident.
18 In SECY-97-044, policy and key technical 19 issues pertaining to Westinghouse AP600 standardized 20 passive reactor design, the staff provided justification 21 for its position and requested Commission approval.
1 22 Westinghouse Electric Company responded to SECY-97-044 in 23 a letter to the Commission dated March 13th, 1997, and 24 stated that the AP600 design meets existing regulatory
(
(3 ) 25 prevention and mitigation criteria; that the requirement NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D.C 20005-3701 (202) 234-4433
l 431 t
1 for a sp:wy system is neither justified nor warranted.
l 7,
2 In order to summarize previous ACRS positions
( )
\' ,/
3 and statements on this issue, we will open up this morning 4 with a presentation from Senior ACRS Fellow, Gus 5 Cronenberg, followed by presentations from the staff and 6 in Westinghouse. Gus?
7 DR. CRONENBERG: As John said, there will be 8 other presentations by Westinghouse and the staff. I'm 9 just sort of setting the stage here this morning. As you 10 know, the Commission did look at this issue and they 11 requested the ACRS to revisit and be clearer in their 12 position.
13 So what I'd like to cover just briefly this
's
)
k/ 14 morning is the prior ACRS positions and statements on this 15 issue, relevant issues to consider, and then try to 16 emphasize that the Commission is really requesting a clear 17 statement of what ACRS views are, and then your technical 18 and your reasoning behind that position.
19 These are some of the prior positions. We had 20 some meetings last summer -- a subcommittee meeting and 21 then a full committee meeting -- where we looked at the 22 three policy issues: one, severe accident mitigation; 23 another one on onsite equipment for DBA support and 72-24 hour actions; and then the other to consider containment
,f3
( , ) 25 performance for vessel breach scenarios.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
432 1 On those three policy issues, this was the 7~3 2 beginning of the new format of the ACRS: where you had aa
( )
~
3 up-front conclusion and then discussions that followed.
4 This in a paraphrasing of the up-front conclusion.
5 We endorse the positions of the staff in addressing 6 the three policy issues on AP600. The outstanding issue 7 still here is this prevention and mitigation of severe 8 accidents. This is a paraphrase of your statement from 1
9 that letter. l 10 The staff is seeking approval of non-safety l l
11 rated systems -- non-safety systems -- to address 12 uncertainties associated with passive fission product i
13 removal for DBA analysis and balance between prevention j Y >i 14 and mitigation of severe accidents.
15 The applicant's submittals provide some 16 support for demonstrating fission product removal using j 17 only passive mechanisms. Nonetheless, we are persuaded by 18 the staff's position that systems beyond passive removal 19 mechanisms should be evaluated to provide greater 20 confidence in mitigating design basis and severe 21 accidents. We recommend Commission approval of the staff 22 position.
23 There was a disconnect here, I should point 1
l 24 out, between non-safety-rated systems for DBA support and r~
l (,T.) 25 for severe accidents. That has been since clarified. DBA 1
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C 20005-3701 (202) 234 4433
433 1 support is with safety-related systems only. So that has
,S 2 all been taken care of.
i ! :
i %.J 3 And then prior to that, in the summer of '95, 4 we looked at ten key policy issues with respect the AP600, 5 and the two germane to this discussion today are:
6 containment performance -- the staff intends to use 1
7 deterministic and probabilistic containment performance 8 goals to review the AP600. We believe that approach is 9 appropriate.
10 Then on long-term severe accident mitigation l
11 and assessment of consequences, it was stated that: the 1 12 post-accident pressure in the containment will remain j 13 positive longer than with active cooling sprays. Removal j
/ \,
\
\' 14 of radioactive species is expected to be less with passive 15 means than using active system sprays or filters.
16 The staff believes this situation calls for 17 additional means -- a non-safety spray. And it's a little 18 more. We believe the radioactive removals should be 19 considered with respect to risk and the safety goal. So 20 that's a little different than saying we support the staff 21 position on severe accident mitigation.
22 There was also a caveat in there. Arguments 23 on radionuclide concentrations would be unnecessary if a 24 performance-based criterion were derived. So I think what p)
(, 25 the Commission is looking for is a little more clarity NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
434 l
1 that those types of --
,r'3 2 MEMBER APOSTOLAKIS: Can you put it back on?
\ )
, xs l 3 DR. CRONENBERG: Yes. This are paraphrased.
4 MEMBER APOSTOLAKIS: Yes, I understand. When 5 -- in item 7, "the staff intends to use deterministic and 6 probabilistic contained performance", does this mean 7 numbers? I mean, is there a document someplace where 8 there is a probabilistic analysis of the containment 9 performance?
10 DR. CRONENBERG: There is the -- we haven't 11 reviewed the PRA that's coming up later this summer, but 12 there are release estimates and that sort of thing, in the p_
13 PRA.
i
{ b 14 MEMBER APOSTOLAKIS: No, but from the i
15 documents that you sent us -- I mean, the staff sent us in 16 preparation for this neeting, all I see is a discussion of 17 uncertainties. The uncertainties are larger, would like ;
1 18 to see these, and so on. But is there a quantitative 19 statement as to how large they are? I mean, what does it l 20 mean that the staff intends to use probabilistic 21 containment performance goals?
22 DR. CRONENBERG: At this time, in June of two 23 years ago, I think you were just telling the staff yes, 24 look at both PRA and deterministic analysis to assess r~N f
i
(
( j) 25 containment performance. And certainly, the various 1
NEAL R. GROSS COURT RFPORTERS AND TRANSCRIBERS 1323 RHODE IS!.AND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
435 1 revisions of containment failures modes have changed, even
,- 2 in that 2-year period. So, I don't know if I'm answering I i 3 your question.
4 MEMBER APOSTOLAKIS: Has the staff done a 5 probabilistic analysis of the containment performance?
6 DR. CRONENBERG: Tom Kenyon, or --
7 MR. SNODDERLY: Mike Snodderly, containment 8 systems branch. Are you asking, did the staff have a 9 quantitative goal or --
10 MEMBER APOSTOLAKIS: Quantitative analysis.
11 MR. SNODDERLY: No. This issue has been 12 approached, to this point, more from a qualitative 13 standpoint. What I'm going to do in my presentation is 7- ,
\s) 14 try to show some deterministic quantitative analyses using 15 the MELCOR code, to give some idea of the timing.
16 But as far as goals, in SECY-93-097 we defined 17 a deterministic containment performance goal which would 18 be that the containment would maintain itself as a leak- ,
l 19 type area for approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. And then there was l
20 the probabilistic containment performance goal of ten j 21 percent of all your accidents that would result --
22 MEMBER APOSTOLARIS: Yes, I know, that there 23 is a goal. The question is, is PRA playing any role in l l
24 this, or is it just qualitative argument and deterministic
['N
() 25 analysis?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBECS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
436 1 MR. SNODDERLY: In my presentation, I would 7- 2 say no, I don't see -- we realized that because of the low
~~'
3 core damage frequency of this plant, they're going to meet 4 the safety goals with all the preventative features, and I 5 think that's where we get into this problem of a balance 6 between presentation and mitigation. Because with just 7 the preventative features, they meet all of our existing 8 performance goals, so --
9 MEMBER APOSTOLAKIS: Including the containment 10 goal?
11 MR. SNODDERLY: Yes, yes. With a core damage 12 frequency of 10.
13 MEMBER APOSTOLAKIS: But you dispute that?
!\ 't 14 MR. SNODDERLY: No sir, I don't dispute that.
15 I think what I --
16 MEMBER APOSTOLAKIS: Okay, what --
17 MR. SNODDERLY: I think -- what my perspective 18 on this is that, one of the weaknesses -- in my mind, one 19 of the weaknesses of that goal is that you could meet your 20 10-6 goal with just preventative features. And so there 21 you'd have to say, then --
22 MEMBER APOSTOLAKIS: I understand.
23 MR. SNODDERLY: Okay. l 24 MEMBER APOSTOLAKIS: Yes, I don't want to get r'~N .
) 25 into your presentation now and interrupt Gus', but I'm I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
437 1 just trying to understand, when we say probabilistic I
,._ 2 analysis -- although it says goals there; it doesn't say e i
\j '
3 analysis -- is there such an analysis? Has somebody 4 calculated the condition of containment failure i 5 probability or --
6 MR. SNODDERLY: Yes.
7 MEMBER APOSTOLAKIS: -- probability of a large 1
1 8 release? j 9 MR. SNODDERLY: Yes, Westinghouse has.
1 10 MEMBER APOSTOLAKIS: But not the staff?
11 MR. SNODDERLY: Yes, and we're reviewing the l l
L2 Westinghouse -- l I
13 MEMBER APOSTOLAKIS: You have done it l
'q V) 14 yourself?
15 MR. SNODDERLY: No sir. We 16 MEMBER APOSTOLAKIS: Okay. That's okay --
17 DR. CRONENBERG: Some of those numbers are 18 coming up. i l
19 MEMBER APOSTOLAKIS: And you are saying that 20 the PRA will be given to us later?
21 DR. CRONENBERG: We'll be reviewing the PRA ,
1 22 and the design certification process which is, as I l
23 understand it, late summer /early fall. We have a lot of 24 work ahead of us this -- unfortunately, there a little rx 25 disconnect here, but --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
438 1 MEMBER APOSTOLAKIS: We'11 do it in August.
l 2 MR. SNODDERLY: I think I can present some of l f[n) 3 the Westinghouse PRA results to give you an idea, but I j 4 think we have to be careful with those results because if l 5 you look at Rev. 8 of the PRA, it showed most of the risk j 6 coming from containment bypass sequences, steam generator ,
i l
7 tube rupture.
8 And then what happened was, they went back and 9 did a re-analysis, and if you credit a decontamination i 10 factor of a hundred for the steam generator tubes, well, 11 steam containment bypass sequences drop out of the risk 12 and then risk gets dominated by early containment j l
13 failures.
/m I '\
V 14 So, I think that that's one reason I think --
15 at least in my opinion, I think we need to take this 16 argument away from maybe the probabilistic reference i 17 numbers, and I'm tending more towards some deterministic 18 analysis with MELCOR.
l 19 So I think when I present those analyses, I'm 20 hoping that's going to give the committee an idea of what I 21 impact we think this spray system would have and what we 22 think it will do. And let's see if -- and then if that 23 doesn't help the committee then yes, we can look at the 24 PRA numbers and see if that would help.
g l
V 25 MEMBER APOSTOLAKIS: Very good.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
439 i I 1 MEMBER BARTON: Go ahead, Gus. j l
I f3 2 DR. CRONENBERG: Okay. Then the Commission l
/ i i /
3 wrote a letter back to the staff, a staff requirements 4 document -- and I'll paraphrase what the Commission said 5 up to this point on the AP600 severe accident mitigation.
6 The Commission agrees it is important that the 1
1 7 AP600 design include adequate means for accident 8 management and long-term mitigation. However, the 9 Commission does nur support the staff's request for the 10 inclusion of additional systems for accident management 11 and long-term mitigation following a severe accident as i
12 presented to the Commission -- not because it is l 1
13 inappropriate, but because the basic design and
- \
k- 14 performance requirements have not been bounded or adequate 15 specified, and the requested additional systems not appear 1
16 to be consistent with the concept of a passive design.
17 The Commission is open to reconsideration of 18 this issue if the staff cac be more specific in terms of 19 what systems are contemplated -- the contemplated is the 20 spray system for accident management including those 21 design and performance regairements for that. And the 22 staff did come back with some more specifics on 23 performance requirements for a spray system.
24 The primary questions here for the ACRS to
(~h)
(, 25 consider is, I think could be framed in terms of arguments NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
440 1 of defense-in-depth versus risk-informed. And this is not
- ,s 2 the first time these types of issues have come before this
( )
\ '/
3 committee.
4 Is a spray system needed to provide defense-5 in-depth? We have defense-in-depth, we have containment, 6 we have ECC, we have borated water. Is really a spray 7 system -- are we requiring a spray system to meet defense-8 in-depth goals, to ensure adequate protection for severe 9 accidents?
10 A risk-informed view: would a spray system 11 provide sufficient reduction in risk for severe accidents 12 to justify its cost? I mean the AP600 has a lot of, as 13 Mike just said, its core damage frequency is low. Th're's
, ')
kJ 14 been a lot of work by this committee about thermal l l
15 hydraulics, and things look good. So you have a that l l
16 perspective on this argument.
17 What is the technical basis for ACRS' i l
l 18 recommendation? Where do you come down and -- ,
19 MEMBER APOSTOLAKIS: Before you -- I think the 20 language is important here because we are going to maybe 21 test some of the ideas we discussed during the development l 22 of the risk-informed regulatory guides.
1 23 So, I mean, the second bullet -- risk-informed 24 view. Would a spray system provide sufficient reduction p) s, 25 in risk? I'm not sure that risk-informed means only NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l 441 1 reductions in risk or the absolute value of risk. )
,, 2 Seems to me the uncertainty in the estimate of
' So I would 3 the risk is part of a risk-informed decision.
4 like to make that clear. In the future I think it's going 5 to make a big difference -- is going to be a major player 6 in our discussion.
7 DR. CRONENBERG: Okay, just some heads up. On 8 the deterministic perspective. Defense-in-depth, 9 prescriptive-type regulation view, this should be "less 10 than" -- sorry about that. In general we could say, 11 diffusiophoresis, what AP600 is relying on is good under 12 submicron size, particle size.
13 We also have a lot of data showing that spray n
(V 1 14 systems are good and very efficient on the one to tens of 15 micron size.
16 MEMBER KRESS: I thought diffusiophoresis was 17 independent of particle size. l 18 DR. CRONENBERG: Well, not large ones for 19 example. I'm just saying, the dominant mechanism. It is 20 --
l 21 MEMBER KRESS: Oh, you're saying, if they are i 22 less than one micron, then the dominant mechanism --
23 DR. CRONENBERG: Yes, the dominant mechanism.
24 MEMBER KRESS: I understand what you're saying
{) 25 there.
l l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i
1323 RHODE ISLAND AVE., N.W.
l
. (202) 2344433 WASHINGTON, D C 20005 3701 (202) 234 .433 l
442 1 DR. CRONENBERG: Okay, so diffusiophoresis is
,g, 2 dominant, then sprays are dominant in tens of micron sized
\ l 3 particles, and then you get the big guys and they fall out 4 by gravitational settling. So we need to know what we're 5 talking about here: the accident sequence; where we are on 6 the accident sequence; how much has got to the containment 7 by the time it's there; and all those sort of 8 deterministic, mechanistic assessments of what we're 9 talking about.
10 MEMBER KRESS: If sprays are present, you're 11 saying they're not clearly as effective for particles less i
12 than one micron, as diffusiophoresis? l There is something -- you l 13 DR. CRONENBERG:
14 have to catch the particles. When they're very small the 15 drop -- there's arguments that you don't catch the 16 particles for very sub-micron size. So -- this is just 17 order of magnitude -- my own view of how to sort out the 18 issues on aerosol depletion. And I sent a memo to this 19 committee about eight months ago where I did some 20 calculations and showed the regimes I thought were 21 dominant, without getting into all the agglomeration 22 calculations and that sort of thing.
23 But if we're going to do deterministic we have l
t 24 to understand our timing of our source time, what we're i 25 talking about, what the chemical species are, who's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 443 1 residing on what particles, that sort of thing -- if we i
g-) 2 really want to get, what is the ef;iciency of the spray. l
(_ / 3 If you're going to make those arguments, there's also I
I l
4 containment leakage and pressure history and bypass ,
1 1
5 scenarios.
I 6 And we really have, not so much of a ;
l 7 validation of diffusiophoresis, we do have quite a bit of l l
8 information on sprays, but what the effectiveness is for i
9 the AP600 design of aerosol with depletion mechanisms.
10 There's not much there in terms of testing lata.
l 11 From the risk-informed view -- and these l 12 numbers have changed over time as people have learned l l
13 things going through the AP600 design. On the Rev. 9 the
(~h. l
\ 'J 14 most recent one -- which I think will be the one that's l 15 reviewed by this committee in late summer -- early l l
16 containment failure is the dominant mode. l 17 Early containment failure is containment 18 failure before major melt relocation due to, most likely, 19 hydrogen coming off during the zirc oxidation, the 20 boildown phase getting in the containment, and then 21 assuming that hydrogen burn buckles the reactor 120-some 22 psi at 400 degrees F.
23 So if you assume that the containment or fail 24 or buckle during the early phases of a severe accident,
/~~T
( ,) 25 then your sprays may not do much for you anyway if you NEAL R. GROSS COURT REPORTERS Af:D TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
444 1 haven't turned them on. You're in the early phases, not when most of the major fission products are coming off.
fr~3 ..
\
~ I 3 Containment isolation failure -- if 4 containment is not secured you have some leakage out, some 5 access or whatever. Containment bypass, for example, 6 steam tube ruptures, popping during a heat-up phase and 7 getting out to the secondary side and leaking off the 8 secondary side and other.
9 So from a risk-informed view, why are we 10 looking at sprays anyway? Basically, we're bypassing the 11 containments, or the containment is failing early, not in 12 the late phase, severe accident mitigation. But then one 13 could say the containment did survive at TMI and we did
- \
\- 14 make use of it, and it was a good thing it was there.
15 So maybe it's overly pessimistic to say that 16 there's early containment failure. Maybe it will survive 17 a hydrogen burn and spray could. So there's uncertainties 18 here that you're talking about based upon what the 19 assumptions were and the analysis.
20 CHAIRMAN SEALE: Gus, did we have anywhere in 21 that analysis a devaluation of the incremental reduction 22 in early containment failure if we had sprays which would l 23 reduce the pressure in the containment?
24 DR. CRONENBERG: Maybe Westinghouse can talk
<s
! 's
\ _) 25 to that question, Bob. I can't answer that one.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
445 ;
l 1 CHAIRMAN SEALE: It may just change that 7s 2 profile profoundly. '
i \
3 DR. CRONENBERG: It would change that profile.
4 MEMBER APOSTOLAKIS: It will change the --
5 DR. CRONENBERG: The heat-up.
6 MEMBER APOSTOLAKIS: The first number, the 7 83.9 number?
8 DR. CRONENBERG: Well, this containment 9 failure is based upon shell buckling, I believe, at 1?0 10 psi and 400 F. So if you kept it cooler, seems like you )
11 could go to higher pressure. But I'm not prepared to 12 answer those questions.
13 Then there's higher pressures to pop the head, (3
t 4
\
4
's/ 14 and there are higher pressures to open ports, and it goes )
15 on. l 16 MEMBER MILLER: Are we going to get an idea of 17 the uncertainties in those numbers?
18 DR. CRONENBERG: Maybe from Westinghouse. But 19 we were going to go, unfortunately, again -- for 20 disconnect, we're going to look at PRA maintenance 21 summary.
22 MEMBER APOSTOLAKIS: When you say risk, what i
l 23 do you mean, individual risk?
24 DR. CRONENBERG: I mean population boundary
,a
( j)
~
25 risk.
l NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(
(202) 234-4433 WASHINGTON, D C. 23005-3701 (202) 234-4433
l 446 l
l 1 MEMBER APOSTOLAKIS: Yes, individual risk. l 2 DR. CRONENBERG: This is based upon footnotes l <7;
\',/
3 at the population boundary. j i
4 MEMBER MILLER: From a kind of a purist l
5 viewpoint, when you quote 83.9 that tells me the 6 uncertainty is 1 percent.
7 DR. CRONENBERG: I'm just taking the numbers 8 from --
9 CHAIRMAN SEALE: Oh, yes. How are you on 10 bridge ownership? I might have some to sell you here.
11 DR. CRONENBERG: I tried to put this in 12 perspective. This is not quite what I wanted. I wanted a 13 juggler -- you know, we've been juggling defense-in-depth n
- 14 -- but I couldn't get off a page major or juggler. But 15 anyway, I think that's a little too right, so I gave --
16 there will be a little more weight here, you know.
17 MEMBER KRESS: Does that clown have any 18 reference to the ACRS?
19 DR. CRONENBERG: No. That's why I weighted 20 the heavy scale of justice here.
21 ME.4BER APOSTOLAKIS : I'm not sure though --
22 CHAIRMAN SEALE: You missed the blindfold.
l l 23 MEMBER APOSTOLAKIS: -- that this is the 24 correct thing. I don't think defense-in-depth competes, 4
f3 25 or is separate, or should be weighed against risk-
\.a1 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
l
447 l 1 informed, because I think risk-informed again -- the way 2 you have it here, you meant that some sort of a 0.1, the ;
7s i
G/
3 mean --
4 DR. CRONENBERG: I meant a balance --
5 MEMBER APOSTOLAKIS: Yes, right. Obviously.
6 What I'm saying is -- and I think what you presented is 7 shared by most people. The idea that risk-informed 8 regulation means you have the mean value there and you 9 know, whatever dominates you try to do something about it.
10 As I said earlier, it seems to me that it's i
11 the whole uncertainty distribution that should be the 1
12 basis for a risk-informed decision, and the uncertainty in l l
13 that distribution, or the spread of that distribution I
['Ts
\ j
\ 14 plays a major role, and I think defense-in-depth is one 15 way of addressing that. So they're really one and the 1 16 same thing. Even though defense-in-depth comes from the 17 deterministic traditional way.
18 DR. CRONENBERG: And again, it would have been 19 nicer if we had really struggled with the final PRA before 20 the Commission was looking for an answer and some guidance 21 from this committee now.
22 MEMBER MILLER: It would seem to me if you 23 took the risk-informed perspective and your confidence, 24 your uncertainties were low, and you had a 10 core damage 7"X
) 25 frequency, then you could argue that you would not need a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
- l l
448 1 l
1 1 containment.
l <w 2 DR. CRONENBERG: Those are what this is all
- )
L./
3 about.
4 MEMBER MILLER: You could meet the safety goal b that way.
6 DR. CRONENBERG: Yes, yes. Okay, Mike 7 Snodderly will present next from the staff.
8 MR. HOLAHAN: This is Gary Holahan from the 9 staff. Can I make some comments on a few of the things 10 we've heard already, before Mike goes into his planned 11 presentation?
12 DR. CRONENBERG: Go ahead, Gary.
13 MR. HOLAHAN With respect to the discussion fn
\' -) 14 of containment failure modes, I think we have a little 15 different view as to how that fits into the thinking.
16 First of all, and we've discussed with the committee over 17 the last year or more, the difficulty of using conditional 18 containment failure probability versus large early release 19 frequency. l 20 Because the CCFP numbers can change very 21 radically rather easily with changing assumptions as to l
22 how conservative is the analysis of tube rupture versus 23 vessel breach versus treatment of hydrogen. And a lot of i
24 the things that go into determining containment failure
,/rh
() 25 are hard calculations to do, that have a reasonably large NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHOD 2 ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 i
l 1
l 449 l 1
1 amount of uncertainty. j j
73 2 And depending upon how much conservatism you l
> 1 i .,r 1 3 put in one calculation versus another, these numbers are 4 rather volatile.
S MEMBER MILLER: Over what range would those 6 numbers be?
7 MR. HOLAHAN: I think you can change these 8 numbers completely. You could change your perception of -
9 - and it has in fact, changed historically, as to wheti.er 10 this design -- whether the containment releases are 11 dominated by tube ruptures or early failures. I mean, the 12 numbers can switch from 10 percent to 90 percent depending 13 upon, you know, how much conservatism there is in one l\ i
'-) 14 calculation versus another.
15 And in some cases they have. So I think that 16 makes me relu ' it to give any given set of numbers on one l
17 day to say, well, let's decide what systems and i
18 performance characteristics are appropriate for this 19 design.
l 2C The other thing is, even if I look at the 2: numbers that were up there that said early containment 22 failure, say containment buckling, a dominant failure 23 mode, 83.9 percent. Well, to say that that means that the l
l 24 sprays are irrelevant or wouldn't play a big role, is not
("h
- ( ,) 25 my view of how I would see that sequence.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
450 1 A conservative analysis says, well, I get to a 2 temperature pressure point and all of a sudden the
, -s
\ ')
3 containment is completely failed, and you know, everything 4 is burped out into the environment. Well in reality, all 5 you've done is, you've gotten to the point where you may 6 get containment buckling or your ability to calculate 7 containment responsements, and so it would have gotten 8 beyond the state-of-the-art.
9 At that point, I don't imagine that you had 10 instantaneous, 100 percent release of containment 11 atmosphere. Even if this were a containment failure, it 12 could very well be a relatively small tear in the 13 containment, it could result in leaking over a number of r
i
\/ 14 hours, and you might say well, this is a good situation 15 under which you wish you had a spray.
16 Here you have a containment atmosphere will a 17 lot of aerosols. Gee, I wish I had a containment spray 18 just now, when I know it's going to leak for the next few 19 hours. So I think it's pretty hard to get from a list of 20 the dominant failure contributors to a sort of on/off 21 decision -- yes, there should be a spray; no there 22 shouldn't.
! 23 I think you have to make a lot of judgments 24 about how changeable these numbers are and what do they r~8
() 25 really represent? Do they represent a containment /no l
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON. D.C- 20005-3701 (202) 234-4433 l
l
451 1 containment, or do they represent the possibility of the i
,, 2 beginning of a long-term leak? And the value of a spray i
~
3 changes from one of those assumptions to the other.
4 CHAIRMAN SEALE: More than that, the spray 5 effectiveness is not just to knock down aerosols, it's to 6 depressurize the system, and if the nature of the leak is 1
7 that it's pressure-dependent, then you really change the 8 nature of the release?
l 9 MR. HOLAHAN: But I think, even that question '
l e l 10 is a little completed --
l l
11 CHAIRMAN SEALE: I appreciate that.
12 MR. HOLAHAN: -- because this isn't a 13 traditional spray system with a heat exchanger. At least, r~N s- 14 what we're envisioning. It isn't very good for 15 suppressing pressure over a long period of time, but it 16 might be helpful during a time of -- early time of 17 leakage.
18 CHAIRMAN SEALE: This is a steel shell, too, 19 instead of a concrete block, too?
20 MR. HOLAHAN: Yes, yes. And I think that 21 raises some questions about what the failure modes are.
22 But I think to say, well, the calculations today show l
l l 23 this, therefore the failure modes are that, therefore the 24 spray is needed or not needed. I think is a little too
(%
( ,) 25 simple for making this complicated a decision.
L NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433 l
l
l 452 l
1 CHAIRMAN SEALE- I think we're on the same l
p)
\
2 page.
3 MR. HOLAHAN: Mike?
4 MR. SNODDERLY: Thank you, Gary. Good 5 morning. My name is Mike Snodderly. I'm with the 6 containment systems and severe accident branch. As Gus i
7 showed you this issue, we've been wrestling with this 8 issue for over two years now.
9 And I think part of the problem -- or not 10 problem, but part of the challenge that's been presented 11 to us is that we in containment systems, in the past have 12 relied on containment sprays for temperature and pressure 13 control and aerosol removal due mostly to impaction, as f%.,
k/ 14 Gus showed. And we've been able to meet all of the goals 1
15 and regulations with that type of analysis. We have our l l
16 SRPs and we're very comfortable with how to analyze that ;
17 situation.
18 And as we all know, it's human nature to 19 resist change but we've been presented with the challenge 20 of reviewing the AP600, which now we're going to be using 21 the passive containment cooling system, this cooling on 22 the outside of the shell, to control pressure and 23 temperature, and also for aerosol removal.
l I 24 Chairman Seale, before we get started, I l /N 25 understand --
i l
!ss)
HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
453 1 CHAIRMAN SEALE: He's your chairman.
,- 2 MR. SNODDERLY: Oh , I'm sorry. Excuse me.
l( '~'
)
l 3 Subcommittee chairman Barton, before we get started, I i
4 understand that Dr. Powers has recused himself, but a big 5 part of my presentation is going to be talking about how 6 the staff is going to analyze natural removal mechanisms.
7 And the staff's best understanding, or our 8 ability to do that, is based on, in my opinion, three 9 works that were done by Dr. Powers. The first is 10 NUREG/CR-5966 which came out in June of 1993, where he 11 developed a simplified model for aerosol removal by 12 containment sprays.
13 And then he did a companion to that -- there's b
\_/ 14 four NUREGs I'm going to be referencing to. The other is 15 a simplified model of aerosol removal for natural 16 processes, and then finally, he applied those 17 methodologies to the AP600.
18 I'm going to do my best to present his work, 19 but I would like to ask that, if in this course of this 20 representation I misrepresent it or I think there's some 21 things that Dr. Powers can provide us with some 22 clarification or insight, I think it would be very 23 important to the committee in resolving this issue.
l l
l 24 MEMBER BARTON: I don't see any problem with 1
/
j i) s u
25 that, Dana, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
i 454 i I
1 VICE CHAIRMAN POWERS: A recusal does not gg 2 eliminate me from offering comments of clarification or ;
! 1
\ ,/
3 fact.
4 CHAIRMAN SEALE: Yes, that's correct. l 1
5 MR. SNODDERLY: I just wanted to let the 1 6 committee know --
7 CHAIRMAN SEALE: Thank you --
]
l 8 MR. SNODDERLY: -- before we got started.
9 CHAIRMAN SEALE: -- for putting him on notice.
I 10 VICE CHAIRMAN POWERS: Darn, now I've got to '
i 11 pay attention.
l 12 MR. SNODDERLY: As I alluded to before, we're 13 used to using sprays, and we have a lot of comfort and
( %
'w 14 confidence in our ability to look at sprays to control 15 pressure and temperature in aerosol fission product 16 removal. Now we're going to be looking at, instead of 17 spray impaction to remove aerosol fission products, we're 18 going to be looking at some different natural removal 19 mechanisms. What are those?
20 First, gravitational settling. That's fairly 21 easy -- just gravity, large particles falling out. And as 22 Gus showed, for the larger particles that's well I 23 understood.
24 The next dominant natural removal mechanism is
(%
(). 25 diffusiophoresis, and I think that can best be described i
i NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433 i
I
455 1 as, if you have steam condensing on a cold surface, as f3 2 that steam moves it's going to be pulling aerosol fission
( )
\' '/
3 products with it.
4 The next natural removal mechanism is 5 thermiophoresis. And again, as hot particles move towards 6 cold surfaces, you're going to be moving aerosol fission 7 products with it. The final natural removal mechanism 8 that I'm going to discuss today, and the most 9 controversial, is hygroscopicity.
10 And I really think that the bottom line on 11 this, based on some deterministic analysis that we've done l
12 with MELCOR, if you believe in hygroscopicity, than the 13 impact of a spray system, it's just not -- I don't see it l
/~N l l )
N/ 14 being there. But if you only credit diffusiophoresis and 15 thermiophoresis, then it looks like you do get an impact 16 of the spray system. I'm going to get to that later. l 17 So I think I've set the stage. And the first i 18 thing I'd like to discuss before I get to those analyses 19 is that, as I said, this issue's gone on for a long time 20 and we've argued it more I think, from a qualitative 21 standpoint, so to try to bring this issue to a head or to 22 resolution, I felt that there was a need for some l
23 deterministic analysis to give you, the committee, and 24 also the Commission, an idea of what do we think this g)
( 25 system is really going to buy us? What do we think it's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
! (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
l
l l
456 1 going to do for us?
,.-.s 2 I see the only policy issue as associated with
- )
3 this as, that we think a containment spray system is 4 needed for severe accident management for the AP600. I 5 don't think that NRR is saying that we can't license a 6 plant without a spray system. I don't think that we 7 should attempt to try to define a balance between 8 prevention and mitigation.
9 I think that the process that we have works in 10 the sense that, as we've done this overall review of the 11 AP600, I think you have to look at it from a design-12 specific 'nt of view. And I think as we reviewed this 13 plant we ..aticed that there seemed to be this -- when we n
i 1 k/ 14 compare this plant to existing operating plants, there l l
15 seems to be this -- I don't think weakness is the right j 16 word, but there's -- this ability to move aerosol fission 17 products is not as strong as in previous plants.
18 So with that in mind, as I said, one of the 19 reasons we're concerned about the ability of the AP600 to 20 remove aerosol fission products is that setting aside the 21 low core damage frequency, let's just say, given a failure 22 of the passive core cooling system, and given a source 23 term inside containment, what systems are available?
i 24 And really, the only system available is the
(N
( ,! 25 passive containment cooling system. We thought that the I
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
I 457 l
1 1 containment fan cooler system may be available to assist l
~~, 2 in the removal of aerosol fission products, but it turns 3 out because it's a non-safety related system, on a primary 4 containment isolation signal such as a LOCA, the fan 5 coolers are going to be isolated and they're going to 6 remain isolated until the containment temperature drops 7 below 228 degrees Fahrenheit.
8 And that's because there was a Generic Letter 9 that we issued recently, 96-06, which talked about a 10 problem identified at Haddam Neck where, because of high 11 containment temperatures, you could get the water in the 12 heat exhangers of the containment fan coolers would heat 13 up, and then when the system was unisolated you could get
/ 'N
)
xY 14 boiling, water hammer, and then a break outside l 15 containment.
16 So to address that concern, Westinghouse has 17 designed the containment fan cooler system to isolate on a 18 primary containment isolation signal and to remain 19 isolated until the temperature drops below 220 degrees F.
20 Now to give the committee an idea of what 228 21 degrees F means, I went to the AP600 -- this is from the 22 PRA where they did some analyses of the 3BE sequence, 23 which is a tailure of the direct vessel injection line.
1 24 .ind that's the dominant sequence; it's the most analyred l
( ) 25 sequence by Westinghouse and by the staff, and this is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
i 458 I considered as a failure of the IRWST to inject failure one
,_x 2 train of the CMT and a LOCA out of the direct vessel I
- ' ~/
i 3 injection line.
4 And what you see is, you don't go below 230 5 degrees F until about ten hours into the accident. So 6 right now, the only system that's going to be available to 7 us in those first ten hours for this particular sequence 8 is going to be the PCCS.
9 Now, what impact do we think the sprays are 10 going to have? Well, the main impact, it's going to give 11 the operators the ability to intervene and provide control 12 over the course of a severe accident. What I'm trying to 13 say there is, one of the biggest advantages of passive i
G
\~) 14 systems is, it doesn't require operator recognition of the 15 accident. The systems are there, they're going to do 16 their job.
17 I think the perspective we're coming from is, 18 what if we do recognize the sequence and we do recognize 19 what's going on? We see a great deal of benefit if we 20 have that ability to intervene with some type of a simple 21 non-safety-related spray system. And also, we think 22 aerosol's inside containment can be reduced in a short 23 time period with sprays.
24 Now, I'd ' cc to try and show you why I O,
25 believe this to be true, and I'd like to -- what I'd like i,~.s)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
459 1 to do :s to show you -- this is a work that was done by
,- ~s' 2 Dr. Powers for NRR, and this work is being done primarily
( ;
\/
3 to assist the staff in its design basis evaluation of 4 aerosol fission product removal.
5 And what Dr. Powers did for us, is for the 6 AP600, he took some thermal hydraulic boundary conditions 7 from MELCOR, plugged them into CONTAIN with a design 8 basis-type of source term to see what type of a 9 decontamination coefficient, or lambda, could be 10 calculated. And what he found was --
11 MEMBER APOSTOLAKIS: Excuse me, what is the 12 definition of the decontamination coefficient?
13 MR. SNODDERLY: That would be lambc't. So you t :
k/ '
14 have aerosol fission products. We calculate them being 15 removed from the atmosphere in an exponential function, E- l 16 lambda (T). So that would be the lambda. l 17 And as you can see, these bars represent the 18 50 percent median value. This black line is what CONTAIN 19 calculated lambda to be, and the top line would be if --
20 and help me out if I misrepresent here -- the top line 21 would be that 90 percent of the lambda's he calculated 22 with a Monte Carlo uncertainty analysis would fall below 23 that line, and 90 percent of the values calculated would l 24 fall &bove this line.
c3 (j 25 So from a design basis standpoint, the staff NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
460 l
l 1 has decided to use this ten percentile line as the basis 73 2 to credit nateral removal mechanisms.
V 3 Now Dana, this did not -- how does this 4 consider hygroscopicity? It doesn't, or does it?
5 VICE CHAIRMAN POWERS: For this particular 6 calculation, the containment atmosphere throughout most of 7 that calculation is superheated so that unless the 8 material has an extraordinarily high hygroscopicity -- and 9 that would be up to and beyond something like cesium 10 hydroxide -- will not absorb water. It does assume that
]
11 you get enough internal condensation from the particles to 12 spheridize; that is, that they are -- they have spherical 1
13 envelopes; they're not perfectly dense.
\ ')
(
14 We've seen that particles do tend to do that i
15 if you have a little bit of moisture in the atmosphere.
16 Still, this atmosphere is very, very dry for these 17 calculations. Those boundary conditions -- it does become 18 capable of condensing water up around 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br />, but most 19 of the time that Mike's interested in here is in the zero 20 to ten hours. That atmosphere is just very superheated.
21 MR. SNODDERLY: Okay.
22 MEMBER KRESS: Did you put a density factor in 23 the spheres?
l l 24 VICE CHAIRMAN POWERS: A density factor --
's_j/ 25 there's a shape factor that includes the effect of less i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
'v
461 l 1 than perfect density -- the size-dependent shape factor.
,-~ 2 And some of that uncertainty range that you t
3 see there does reflect the uncertainties in that shape 4 factor calculation effort because -- oh, it's an involved 5 thing, involving factual geometries and things that we 6 understand, perhaps qualitatively but not precisely 7 quantitatively. And they're reflected 'tn those 90 to 10 8 percentile bands.
9 MEMBER KRESS: Is this for a specific 10 sequence?
11 VICE CHAIRMAN POWERS: Exactly, a 3BE 12 sequence.
13 MEMBER KRESS: So the uncertainties --
t'~~%
\~- 14 VICE CHAIRMAN POWERS: Boundary conditions are l
15 --
16 MEMBER KRESS: -- those numbers do not reflect 17 the numbers in the sequences?
18 VICE CHAIRMAN POWERS: The boundary conditions 19 in geometry are fixed -- the zero uncertainty in those.
20 MEMBER APOSTOLAKIS: Is this saying that 21 lambda is a function of time?
22 MR. SNODDERLY: Yes, it is. And there's two 23 things I want the committee to take away from this slide 24 before I go on to the deterministic analysis we did. And t /-
( ,s) 25 that is, that the dark black line is what CONTAIN HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISt.AND AVE., N W.
(202) 234-4433 WASHINGTON, O C. 20005-3701 (202) 234-4433
! 462 1
l l 1 calculated lambda to be, and these models have been
! s 2 incorporated into MELCOR.
i
't x/ 3 So the next slide I'm going to show you which 4 shows what we think cesium iodide is going to do for the 5 3BE sequence, to give you some idea of what's -- in other 6 words, we're going to look at cesium iodide versus time 7 and we're going to look at it with the spray system and 8 without a spray system to try to give you some idea of 9 what we expect.
10 And lambda does vary over time, and you can 11 see that the average lambda is about -- for the 50 12 percentile mean, is about 0.4. You can see, you .know, in 13 less than five hours it drops down very rapidly to about tQ s/ 14 0.2, then rises up to about 0.9 in the first five hours. I 15 And that's the time period that we're going to be looking i 16 at.
17 But yes sir, you're right --
18 MEMBER KRESS: George, with time-varying 19 lambdas you have to view the definition as being DC with 20 respect that time is equal -lambda (c). You have to change 21 -- you don't use the exponential definition because you 22 only get an exponential if it's a constant, so you have to 23 view the definition of lambda a little differently.
24 MEMBER APOSTOLAKIS: Yes, so this is the
()
\ /s 25 instantaneous rate.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C, 20005-3701 (202) 234 4433
463 )
l 1 MEMBER KRESS: It's an instantaneous rate. I l
,s ~
2 MEMBER APOSTOLAKIS: Now, what kind of (s )
'" /
3 uncertainty are we displaying here? In other words --
4 MR. SNODDERLY: Well, I think --
5 MEMBER APOSTOLAKIS: Is there any randomness 6 here?
7 MR. SNODDERLY: I think the best way to l l
8 describe the uncertainty is that we assumed that there's 9 no thermal hydraulic uncertainty in the sense that we took 10 the results from MELCOR and used those as a boundary i
11 condition for -- because we -- it is very sensitive to l 12 containment temperature, containment pressure, and the 1
13 supersaturation ratio. l i
/% >
k/ 14 But we had to assume something, so we took 15 what MELCOR gave us for 3BE. So that part of the 4
1 16 uncertainty is not reflected. But I think, as I said, I i 17 think where the uncertainty is, is that say for ten hours, 18 we believe that the instantaneous lambda at that time is 19 mostly likely going to be whatever, 0.4, and there's a 90 20 percent, whatever, confidence that it will be less -- it's 21 going to be somewhere between whatever this is, 0.7, and -
22 - this is hard to read from --
l 23 MEMBER APOSTOLAKIS: Yes, I understand what 24 you're saying. I understand what --
r^s.
i ,) 25 MR. SNODDERLY: That's where, I guess, as far i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
464 1 as --
l
,w 2 MEMBER APOSTOLAKIS: Okay, is this --
I \
3 MR. SNODDERLY: -- consider uncertainty.
4 MEMBER APOSTOLAKIS: Are there any random 5 phenomenon occurring here, or is it purely a matter of us 6 not knowing --
7 VICE CHAIRtiAN POWERS: George, this is 8 strictly epistemic type of uncertainly. There's no 9 stochastic variability considered here.
10 MEMBER APOSTOLAKIS: So the thing has a timed 11 variability which is displayed primarily by the contained 12 calculation? Is the contained curve more or less 13 representative of the timed dependence?
\2 14 MR. SNODDERLY: No, I think all I wanted to 15 try to show you here was that, if you take a particular l l
16 point and say -- CONTAIN is going to calculate a certain 17 lambda versus time --
18 MEMBER APOSTOLAKIS: I understand that, yes.
19 MR. SNODDERLY: And it shows that it tracks 20 pretty well with the median. And now this is the model 21 that I'm going te use to try and give you some idea of 22 what impact a spray --
23 MEMBER APOSTOLAKIS: So basically, the time 24 dependence of the rate is really the continuous curve
- p
! ( ,) 25 there? Essentially? There is a dip, and then goes up, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
I 465 1 and then turns around. Is that what the medians do too? ;
1 3 2 Right. l
)
~j 3 MR. SNODDERLY: The median doesn't dip quite 1
4 as much --
5 MEMBER APOSTOLAKIS: It doesn't -- exactly.
6 MR. SNODDERLY: I'm just trying to --
7 MEMBER APOSTOLAKIS: I understand that.
8 MR. SNODDERLY: I think what I -- )
i 9 MEMBER APOSTOLAKIS: So then, if I wanted to 10 calculate then, the concentration of contaminants, l 1
l 11 whatever -- particles -- then I can't really use a pure '
12 explanation, right? I have to go and use the integral of 13 w: ever your rate --
14 MEMBER KRESS: Yes, you integrate it over the 15 time.
16 MR. SNODDERLY: And what I'm --
l 17 MEMBER KRESS: If you do it on a point, well, 18 you couldn't do it on the segments.
19 MR. SNODDERLY: And what I'm going to do is, 20 I'm just telling you that I'm going to use something 21 pretty close to this, in the analysis --
22 MR. HOLAHAN: Microphone.
23 MEMBER KRESS: The 0.4 value that you have is, 24 to me, slightly higher than I'm used to for just
() 25 gravitational settling, so there must be some NEAL R. GROSS COURT REPOhTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W (202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
466 1 thermiophoresis --
7- 2 MR. SNODDERLY: And diffusiophoresis.
3 MEMBER KRESS: Well, the diffusiophoresis are 4 probably those bumps at the front, I would guess. It's 5 probably super-heated during the rest of the time. You 6 don't get much diffusiophoresis.
7 MR. SNODDERLY: Again, I think my objective 8 here is to try and give you an idea -- because I think 9 what was missing from the committee is, what do we think 10 this spray system is going to buy us? So all I'm really 11 trying to show you here is that, I'm going to use this 12 removal, this lambda versus time, this modeling in the 13 next slide that I'm going to show you.
( '1
\2 14 Again, I'm trying to show you to give you some 15 confidence that the model I'm using is pretty good 16 compared to the median and the lower bound and the upper 17 bound as calculated by Dr. Powers for the AP600. So why 18 don't we get to those analyses and then we'll see if that 19 helps us or doesn't.
20 MEMBER APOSTOLAKIS: Why isn't it monotonic 21 though? I understand what you're trying to do, but I'm 22 trying to understand this picture.
23 MR. SNODDERLY: Okay, sure.
24 MEMBER APOSTOLAKIS: Why does it go up around O
(j 25 17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br />?
HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON D C. 20005-3701 (202) 234-4433 l
l
467 1 MR. SNODDERLY: Dr. Powers, could you help us
,3 2 out there?
i 3 GI 3 MEMBER APOSTOLAKIS: And then comes down 4 again. I mean, what are the physical phenomena?
5 VICE CHAIRMAN POWERS: The CONTAIN code itself 6 is an interesting model of saturation in the containment 7 and fog formation. And what you're seeing at 17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br /> is 8 a fog formation in the containment that starts sweeping 9 out aerosols.
10 MEMBER APOSTOLAKIS: Yes.
11 VICE CHAIRMAN POWERS: Okay. I personally 12 don't think -- I certainly haven't investigated it very 13 thoroughly but I -- most of the aerosol community doesn't I
kl 14 agree with CONTAIN's fog formation model and that's not 15 reflected in the uncertainty analysis, but that's what !
16 happens at 17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br />.
17 That's a fairly unimportant part of the 18 analysis. Mike will probably be concentrating on the 19 first ten hours.
20 MR. SNODDERLY: Exactly, Dana, I was -- thank 21 you, very much. Yes, what you're going to find is, my l 22 analyses only go out to ten hours. Again, all I'm just l 23 trying to do is give the committee for a feel of the 24 lambda that I'm -- or, the amount of credit that I'm
,rN
() ,
25 giving the natural removal mechanism inside the AP600, to l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RriODE ISLAND AVE, N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433 i
1 468 1 1 give you some idea of the reasonableness. I
,- 2 MEMBER MILLER: And you're going to do the
( )
'~'
3 same as if you put a core spray in, right?
4 MR. SNODDERLY: Exactly. And then --
5 MEMBER MILLER: Where you apply lambda for 6 core spray?
7 MR. SNODDERLY: Right. So why don't I put l
8 that --
9 MEMBER APOSTOLAKIS: This is model l 10 uncertainty? Because you just said that most of" l l
11 aerosol community doesn't believe in that. In oth r 12 words, you have incorporated chis -- l 13 MR. SNODDERLY: It doesn't believe in this
/
o \
U 14 flip here.
15 MEMBER APOSTOLAKIS: But how about the earlier i 1
16 one? Are these uncertainty bands parameter uncertainty, 17 primarily? Various parameters you have there, or you have 18 judgments regarding the models as well?
19 VICE CHAIRMAN POWERS: Yes. There are both 20 uncertainties in parametric values for models that are 21 accepted, and there are selections of different kinds of 22 models for the same phenomena in the uncertainty band, l 23 yes. We tried to cover all of the uncertainties aside 24 from boundary conditions in geometry.
(A)
% ./
25 MEMBER APOSTOLAKIS: Yes. And one last i
NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS
! 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
t 469 1 question -- I'm sure you'll be glad. Is tnis curve l
I 73 2 disputed by -- a figure disputed by Westinghouse.
b 3 MR. SNODDERLY: I think it would be best if l
4 Westinghouse --
5 MEMBER APOSTOLAKIS: After two years you don't 6 know?
7 MR. SNODDERLY: -- answers it. Well, I think, 8 with Westinghouse -- and of course, correct me if I'm 9 wrong, Brian, but they believe I think, that we should 10 have used a curve somewhere here for design basis analysis 11 -- in this range. Or in the .6 range, sorry.
12 MEMBER APOSTOLAKIS: Or a little higher.
13 MR. SNODDERLY: And they beli>'re that we're I
\*] 14 overly conservative in using that type of a curve. But --
15 MEMBER APOSTOLAKIS: Okay, so this --
16 MR. SNODDERLY: -- this is for design basis 17 accidents --
18 MEMBER APOSTOLAKIS: T'aey don' t dispute the 19 validity of the curve. This is a choice of the curve --
20 MR. SNODDERLY: We met with Westinghouse --
21 MEMBER APOSTOLAKIS: There is a dispute --
22 MR. SNODDERLY: -- in the January timeframe 23 with their consultant, EPRI, and we talked about in great l
24 depth, this curve and they talked about Dr. Powers' work.
ry
, () 25 I believe we came away with, I think, a mutual NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
470 I 1 understanding of their position -- l
,r 3 2 MEMBER APOSTOLAKIS: The lower the lambda, the i
3 more conservative you are? '
4 MR. SNODDERLY: The less you remove. And the 5 --
6 MEMBER APOSTOLAKIS: The lower the more 7 conservative --
8 MR. SNODDERLY: -- greater the lambda, the --
9 MEMBER APOSTOLAKIS: So you decided to go with 10 the 10 percentile, right? !
11 MR. SNODDERLY: Yes, for design basis.
l 12 MEMBER BARTON: This is design basis, now.
13 MR. SNODDERLY: But now we're in the severe
- p_ \
\~ 14 accident -- what I'm trying to say is, I'm going to give 15 them credit for something along in these lines.
16 MEMBER APOSTOLAKIS: The CONTAIN?
17 MR. SNODDERLY: Yes. Because that's the best 18 model I have right now available to me. Because as I 19 said, I think all along we've been arguing this thing from 20 a qualitative standpoint. I really want to try and give 21 the committee some idea of what sprays do and don't do for 22 you, and I thought --
23 MEMBER KRESS: This lambda does not include 24 fan coolers?
O
(,,,) 25 MR. SNODDERLY: No sir, it does not. Because NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
471 1 of what I tried to establish earlier; that they're not
,- 2 going to be available in the current -- as they are
)
3 currently designed. That's not to say that there's some 4 simple fixes that one could do to make them available, but 5 as of right now they are unavailable.
6 MEMBER APOSTOLAKIS: So for design basis you 7 use the 10 percentile --
8 MR. SNODDERLY: Right.
9 MEMBER APOSTOLAKIS: -- and for accident --
10 MR. SNODDERLY: And for severe accidents --
11 MEMBER APOSTOLAKIS: Okay.
12 MR. SNODDERLY: We probably would use 13 something like this, but I didn't have that model
-' 14 available to me. The thing I had available to me is this.
15 So that's what I'm going to use.
16 Okay. And here we go. Again, these analyses 17 are very preliminary -- we just did them a couple of weeks 18 ago and we just got the results early this week. But as I 19 said, when we were invited back to come in front of the 20 committee, I felt that we needed something more to tell 21 the committee. We .. ' some new information because 22 we've -- I didn't feel it was appropriate to come before
~
l 23 you and just say, we don't have anything new.
24 So what I'm trying to do here is to -- okay, q_) 25 what have we got? This is cesium iodide versus time.
NEAL R. GROSS l COURT REPORTERS AND TRANSCRlBERS i
1323 RHODE ISLAND AVE., N W.
! (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
I
472 1 This is using MELCOR calculation of the 3BE sequence, and 2 to give you an idea of the core inventory, at four hours i \
GI 3 this is about 84 percent of the total cesium iodide inside j l
4 containment. So it's a fairly large release of the total !
5 fission products. This --
6 MR. HOLAHAN: Excuse me. It's 80 percent of ,
l l
7 the core of the source. It's 100 percent of what's in the I l
8 containment, no? )
I 9 MR. SNODDERLY: I'm sorry, Gary -- 84 percent 10 of the total core inventory. So 84 percent of the cesium I
11 I iodide in the core we've released to the containment.
1 12 Okay, so it's a fairly upper bound --
13 MEMBER KRESS: Is it 84 percent of the total d 14 inventory?
l 15 MR. SNODDERLY: Total -- perhaps John Ridgely 1 16 can help me out, but I believe it's of the total ;
17 inventory. John Ridgely is from the accident evaluation 18 branch of the office of research who did the analysis for 19 us.
20 MR. RIDGELY: I'm John Ridgely, and it's 34 21 percent of the initial inventory at the inception of the 1
22 accident.
l 23 MEMBER KRESS: That's a pretty high release.
24 MR. SNODDERLY: Yes, it is, but for some
(
25 reason that's what MELCOR 1.84 calculated. To give you an NEAL R. GROSS COURT REPORTERS AND TRANSCR'BERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
i 473 l 1 idea of NUREG 1465 would calculate for a severe accident, 1
,s 2 77 percent of the total at the top. And for -- just the l l 1
! ( '/
3 early in vessel would be 40 percent of the total. So to 4 let the committee know, yes, I would consider this a 5 conservative upper bound. It's a lot of the cesium 6 iodide. But let's see what happens to it.
7 MEMBER KRESS: Well, actually, conservative is 8 in the eye of the beholder. The less you put in the less 9 effective removal mechanisms are.
10 MR. SNODDERLY: Thank you, Dr. Kress, that's 11 right.
12 MEMBER KRESS: I have a question on --
13 MR. SNODDERLY: Perhaps I -- let me withdraw 73
\ }
\/ 14 conservative or not conservative. It's 80 percent of the 15 total initial core inventory.
16 MEMBER KRESS: Does this also include some i 17 structural and non-radioactive aerosol calculations?
18 MR. SNODDERLY: It's just what MELCOR told --
19 what I wanted to do was to try and get a feel for what it 20 would do to a typical aerosol fission product that --
21 MEMBER KRESS: But MELCOR has some options on 22 fission product release. Did you use CORSORM or -- which 23 option --
f 24 MR. SNODDERLY: CORSORM.
(~%
(_,) 25 MEMBER KRESS: See, one of the options in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4433
474 1 MELCOR lets you turn off the structural aerosols and not
,-s
, s 2 use it.
. t i 1 \/
3 MR. RIDGELY: John Ridgely again. CORSORM.
4 MEMBER KRESS: CORSORM?
5 MR. RIDGELY: Yes.
6 MEMBER KRESS: This one does not include 7 structural aerosols, as best I remember.
8 MR. RIDGELY: I don't know.
9 MEMBER KRESS: I don't recall.
10 MR. RIDGELY: I know there was a study done of 11 the different release models some time ago at Oak Ridge, 12 and it indicated that CORSORM was the best model to be
- 13 used for this kind of a case.
('-
14 MEMBER KRESS: Well, we'll get representative 15 numbers, but my main interest is, of the natural 16 attenuation processes will be strongly dependent on how 17 much non-radioactive aerosols you put in, and you have 18 generally options as to whether you use that or not --
19 MR. SNODDERLY: So Dr. Kress are you saying, 20 so those non-radioactive aerosols would assist in 21 agglomeration and helping other things -- helping remove 22 these particles?
i 23 MEMBER KRESS: This curve on no sprays will be 24 strongly dependent on that assumption, and I don't know o
i
( )\ 25 what -- I was just trying to figure out what the NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202).234-4433 i
l 475 l 1 assumption was, i
l I
,-~. 2 MR. SNODDERLY: That's a very good point. I'm
> t a 1
'N / i 3 not sure how to address it. Yes, this is what I have at 4 this time --
5 MEMBER KRESS: Well, in a design basis space, 6 normally you don't put non-radioactive aerosols in, so I .
l 7 don't know what --
8 MR. SNODDERLY: Yes, this is not design basis, 9 so yes -- because I think we got --
10 MEMBER KRESS: Well, the other question is, 11 you know, how good are these cprays? Are those really 12 fine droplets, a high flow rate, or do you have some --
l l
~~.
13 MR. SNODDERLY: What we modeled is 1000 gpm l N- 14 spray drawing from a 200,000 gallon tank, and when I l
15 calculated the water flux, Q, it comes out to about .005.
16 So it's not --
17 MEMBER KRESS: Droplet size, you've got any 18 idea of what that was?
19 MR. SNODDERLY: I'm sorry, sir?
20 MEMBER KRESS: The droplet size, mean size?
21 MR. SNODDERLY: Only what I did from the water 22 -- no, only from the water flux. Which would, in a way, 23 consider droplet size.
24 MEMBER KRESS: But certainly, this spray looks I yw
(_,) 25 to me like one that really requires a prompt and an active NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
476 1 system, pretty good pump, high pressure, and --
f- 2 MR. SNODDERLY: Yes, that's -- we tried -- I'm
( !
~'
3 going to get to that later. The system that is available 1
4 in the current design that we would envision, would be 5 supplied by the fire protection system which has -- as 6 it's currently designed, has one electrically-driven and 7 one diesel-driven, fire pump capable of 2,000 gpm.
8 There's not a design of head for that pump, 9 but I'm assuming that such a pump could be either uprated 10 or losses or designs su h that you could provide 1000 gpm 11 flow.
12 To give the commi't.ee an idea, for typical
^
13 operating plants, water flux which would be flow rate
??v e
\- > 14 divided by the cross-sectional area of the containment, 15 varies from .01 to .06 for operating plants, which for a 16 10,000 square foot cross-sectional area of a containment, 17 would correspond to about 1500 gpm to 9000 gpm.
18 So this being .005, it's on the low -- it's 19 not a great spray. Like I said, the lowest on an 20 operating plant would be .01 and the one I modeled -- or, 21 the one being modeled here is .005.
22 MEMBER FONTANA: Do you have a feel for what 23 that spray curve would look like if you used the 90 24 percent value and 10 percent value of lambda? Just to get
,a l
( ,) 25 kind of a feel for what the range would look like on this l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 477 1 picture here?
f, 2 MR. SNODDERLY: Yes sir. What I tried to do, )
( '
) '
~'
3 or the reason I had the -- the answer to your question is 1
4 no. That would have been a very good sensitivity. What I 5 would have really like to use is this curve, but I didn't 6 -- that's not modeled. This is the one I had available to 7 me, so no sir, I don't have one for this or this.
l a
8 MEMBER FONTANA: I understand. l l
9 MR. SNODDERLY: But that would be a good )
10 sensitivity case.
i 11 VICE CHAIRMAN POWERS: Mike, you might 12 actually have a curve that you can show Dr. Fontana, l
13 There are a set of calculations in the analysis for the I
(~) l
'\ // 14 AP600 in which for instance, iodine is plotted --
I 15 calculated using the 90 percentile, the median, and the 10 16 percentile in there. He could probably show you those 17 curves.
18 I think you'd probably be surprised at how 19 little change there is in the curve.
20 MEMBER FONTANA: That's what I was wondering.
21 VICE CHAIRMAN POWERS: Within the scatter of 22 his xerox flaws in the points, you're probably about right 23 in this time interval. The big differences show up toward l
24 the end.
r~s
() 25 MEMBER MILLER: Would there be more change if NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
478 1 we assumed a -- this ascumed totally cesium iodide. Dr.
2 Kress indicated that in any case, you need to assume a lot 73 o s
~
3 of non-radioactive aerosol particles. Would that change 4 it more dramatically then? Like when you say --
5 VICE CHAIRMAN POWERS: The calculations --
6 MEMBER MILLER: -- particles --
7 VICE CHAIRMAN POWERS: -- done for the AP600 8 of course included everything that's coming into the 9 containment, as I suspect Mike's did. He's just shocring 10 you the cesium iodide component here.
11 MEMBER MILLER: I thought he only -- clarify 12 this, Mike.
13 MR. SNODDERLY: It did include the non-5- / 14 radioactive aerosols, I believe.
15 MEMBER MILLER: So this did, okay. !
1 16 MR. SNODDERLY: Yes.
17 MEMBER MILLER: The CONTAIN model does 18 automatically include --
l l
19 MR. SNODDERLY: Right. But it's less 20 conservative in the sense that, now that provides more I 21 particles that can agglomerate with the radioactive 22 particles and remove them. I believe that was the point 23 that Dr. Kress was trying to make. So this does include 24 the non-radioactive aerosols.
(G ,) 25 MEMBER MILLER: And that is approximately what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
l 479 l
1 Westinghouse is assuming, also? l
,3 2 MR. SNODDERLY: I'm not aware of an equivalent
( )
, ~~/
! 3 analysis, or one that I could compare to. I'm sure it's t
4 been done; I just -- I can't answer that. Sorry. l 5 Anyways, what you see here is that when I used i
6 the model that was in CONTAI:; -- and you can see, this is l
7 what happens to cesium iodide. And about nine hours 8 you've gone from 12/13 kilograms to one. And at ten hours I
9 you're at less than one. So then what we did was, we l 10 turned on the spray at one hour and like I said, it goes l 1
11 on for 4.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />, and that's why you see this tail to l l
12 start to come up again, because the spray stops. l l
13 But it's our belief that, even though it comes q
\
\-) 14 up it will stay down.
15 MR. RIDGELY: Excuse me -- John Ridgely again.
16 Since this graph was made we are continuing to run these 17 cases, and those curves for the 1-hour and the two-and-a-18 half hours and the hygroscopic, all hover at about half-a-19 kilogram, on out as far as we've got them out so far. And 20 some of them gc- out to five, six, and eight hours.
21 MEMBER KRESS: That just reflects --
22 MR. RIDGELY: So that upturn --
23 MEMBER KRESS: -- your continuing sources, 24 obviously.
I r~T q,/ 25 MR. RIDGELY: Right, and the upturn is not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-44 9 WASHINGTON, D.C. 20005-3701 (202) 2344433
480 1 going up very far. I just wanted to make that point.
r- 2 MEMBER KRESS: Because the source is probably
(
%/
3 practically all gone by then.
4 MEMBER BARTON: All gone by then. <
5 MR. SNODDERLY: And in this case we waited 6 until two-and-a-half hours after the initiation of the 7 break. And you can see that the spray system takes it 8 down in about four hours.
9 Now this curve here, what we -- there's an 10 option that's been put into MELCOR to model hygroscopic 11 effects, and when we turn that model on you can see that -
12 - this is with -- I'm sorry, let me tell you that all of 13 these include the PCCS on, so this really shouldn't be
(~T i 1 Y- 14 there. All of these have the PCCS on -- even the spray 1
15 bottles.
16 But if you give credit for the hygroscopic 17 model in MELCOR, you can see that you're even below the 1
18 sprays, and so that was a key finding. l l
19 I think the other thing that surprised me, or l 20 I learned from this was, I expected this curve but I think 21 I expected this curve to come down to less than a kilogram 22 at a farther time out -- say in the 20-hour timeframe. So 23 this was a bit of a surprise to me.
l 24 MEMBER FONTANA: Did I misunderstand a
,7 3 l (_) 25 statement made earlier that hygroscopic effects would not NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
481 1 be effective prior to several hours? Talking about the 73 2 system being super-heated? Did you say that or did Dr.
( )
3 Powers say that?
4 MR. SNODDERLY: I think Dr. Powers did.
5 MEMBER KRESS: The CONTAIN will supposedly 6 calculate thermal hydraulic conditions, and if it's super-7 heated it's built in, but I think what Dana says is, 8 there's some contention about this hygroscopic model. It 9 has super-heat things built into it, and it will reflect 10 that.
11 MR. SNODDERLY: And I think that's why you see 12 it being such a strong effect, because at this time in the ,
13 break you've got a supersaturated atmosphere, and so I j
's>i '
14 hygroscopic is going to show up as a big effect. l 15 So I guess I just wanted to let the committee 16 know that, that we did find that out. But anyways, this 17 does give you an idea that if you turn the sprays on in an 18 hour2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br />, you're reducing this amount of fission products that 19 would then be not available for release if something 20 should happen in the first -- I guess really in the first 21 -- the time of hours of concern is one hour to probably 22 five hours. After that, you know, there's really not much 23 in there.
24 But I think this does -- so now I'd like to t'mx
(,)
I 25 give you an idea of how effective this spray system was NEAL R. GROSS COUPT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
I l
482 j l
1 that we proposed. And we can of course, come back to this 2 slide. l p-l \ ) l
'~'
3 MEMBER MILLER: I'm sorry, I'm confused now.
i 4 The core spray you used in this previous model was not the i
5 core spray proposed for the Westinghouse plant?
6 MR. SNODDERLY: Well, first of all, there is 7 no formal spray proposed for the Westinghouse plant, so it 8 would just -- what I did was -- ,
9 MEMBER MILLER: It's not the one that the 10 staff proposes?
11 MR. SNODDERLY: Yes, this is 1000 gpm with 80 j 1
l 12 percent coverage and a water flux of .005.
13 What I want to just show you here is to give A
! \
K/ 14 you an idea of -- this is a cesium iodide in containment 15 atmosphere, so really what we're looking at is -- we're 16 going to look at this segment right here, so it's kind of l l
l' a blowup of this -- this curve right here.
18 So this is what MELCOR calculated, and what 19 Jason Shaperow of the accident evaluation branch did for 20 us was he said, if I just pick a constant lambda of 0.4 21 per hour and I plot -- that's that plot there -- and if I 22 use the simplified model developed by Dr. Powers of the 23 natural processes, it tracks it pretty well.
24 So I'm just trying to give you an idea of the
( ,) 25 reasonableness of the spray system we've proposed, or how l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
l t 483 l
1 w 've modeled this. So I guess what it shows you is that i
73 2 .the lambda calculated for the spray system -- this is just t
\"j 3 natural removal mechanisms was 0.4, and the spray at --
l 4 when you turn the spray on at one hour it resulted in a l
l lambda of 1.6, and when you turn this -- I'm sorry 5 -- a 6 lambda of about 0.8 when the sprays are turned on at one 7 hour, and a lambda of 1.6 for sprays at two-and-a-half 8 hours.
9 MEMBER APOSTOLAKIS: So the simplified model 10 uusumes a constant lambda, is what you mean?
11 MR. SNODDERLY: No, I'm sorry. The black dots l 12 assume a constant lambda; the open dots are Dr. Powers' 1
13 simplified model, which does vary over time.
, q t t
V 14 MEMBER FONTANA: Now that's -- the top two l l
15 lines are without a spray? l l
16 MR. SNODDERLY: All of these lines are without l 17 a spray.
18 MEMBER FONTANA: But you have, the bottom line 19 says, lambda of 0.4 per hour.
20 MR. SNODDERLY: If you just assumed -- I was 21 just trying to give you a feel for how effective the spray 22 is. The spray gave you a lambda of 1.6 and a lambda of 23 .8. And natural movement mechanisms give you a lambda of 24 0.4.
(v) 25 MEMBER APOSTOLAKIS: Okay, I see.
l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433 l
l ,
484 l
1 MR. SNODDERLY: So to get the effectiveness i f- 2 that I've shown you. you would need to supply a spray that t
V 3 would be about twice as effective as the natut removal 4 mechanisms. And you've got that for 1000 gpm spray with 5 80 percent coverage.
6 So I don't know if this helps the committee or 7 if this is -- but that's pretty much what I really wanted 8 to present today.
9 MEMBER KRESS: I really do need to know what 10 particle size, droplet size, you input for the spray also, 11 because your description of the flow and the coverage is 12 incomplete without that.
13 MR. SNODDERLY: Okay. So Dr. Kress, you were
[)
' 14 looking for a description of the --
15 MEMBER KRESS: Mean droplet size that you've 16 input in the spray.
17 MR. HOLAHAN- I think we should also, as a 18 matter of fact, clarify the treatment of non-radioactive 19 materials. And my view is that for matters of fact, you 20 ought to get them 100 percent correct. For matters of 21 opinion, you ought to be right at least half the time.
22 DR. CRONENBERG: This is Gus Cronenberg. I 23 would think the droplets were on about 100 microns.
I h
j 24 VICE CHAIRMAN POWERS: Not for a spray.
i (q,)
25 DR. CRONENBERG: That's normal input for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
485 1 CONTAIN, is that not correct, Dana? About --
,e w 2 VICE CHAIRMAN POWERS: No, that's -- for a i
V 3 spray? Typical monodisperse spray models -- and I'm 4 pretty sure CONTAIN is a monodisperse spray model; I'm not 5 absolutely positive about MELCOR -- that they use 1000 6 micron droplet.
7 The simplified model, we looked at the actual 8 distribution for the -- of spray -- the two types of spray 9 nozzles that are used in plants pretty universally. And 10 there are some interesting things that happen with sprays.
11 If you choose a monodisperse droplet size you end up with 12 a residual aerosol that just doesn't get trapped very 13 fast. And that's an artifact of the model.
(3 )
\- 14 If you actually use the distribution, that 15 residual aerosol that just defies trapping goes away 16 because it's actually a result of the fact that there's a 17 minimum in the ability of spray droplets to capture 18 particles as a function of particle size, but that minimum 19 moves with the spray droplet size. So if you get a 20 distribution you get rid of it. j l
21 But I believe 1000 microns which is the mass l l
22 mean diameter of one particular class of spray nozzles 23 that's widely used, is what is typically used for i
l 24 monodisperse models. I know the CONTAIN model is a l ,/ 3 k_- 25 monodisperse spray model. I suspect MELCOR is too, but I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
486 1 don't absolutely know for sure, g3 2 The user can adjust that, so the guys still I
J Lj 3 need to go find out what they put in, but if you take the l
4 default, I think you get 1000 micron droplet.
l 5 MR. SNODDERLY: I think to summarize, what I'd 6 like to say is, I'd like to make two points. Number one, 7 I don't intend for this one analysis to answer this 8 question. I was just trying to give the committee a feel.
9 I think it's important to remember that this is just one 10 data point; this is just one sequence; it's just one. But 11 I think it was something that was missing from this 12 argument, and I hope it will assist you in your decision -
13 -
['N MEMBER APOSTOLAKIS:
'-- 14 In risk --
15 MEMBER BARTON: Do we know what the 16 probability of this sequence is?
l 17 MR. SNODDERLY: I should know that right off -
18 - maybe Westinghouse can help me. Bob, do you know the j 19 frequency of the 3BE.
20 MR. PALLA: I don't -- Bob Palla of 21 containment systems branch. I don't know precisely. It's 22 one of the dominant sequences -- at least 30 percent --
23 MEMBER APOSTOLAKIS: Mr. McIntyre there has 24 something to say. ,
l
(^T 25 MR. McINTYRE: I have a slide that addresses
( ,/
NEAL R. GROS $
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
487 l
i 1 this.
l l 7 ~., 2 MEMBER BARTON: Okay, thank you.
! I
~
3 MR. SNODDERLY: And --
l 4 MEMBER APOSTOLAKIS: Mike, what does this tell 5 me about risk? I mean, you made the point that --
6 MR. SNODDERLY: Yes, well --
7 MEMBER APOSTOLAKIS: -- you used sprays, you 8 have --
9 MR. SNODDERLY: Well, I just --
10 MEMBER APOSTOLAKIS: -- smaller 11 concentrations, which is not a surprise to anybody.
12 MR. SNODDERLY: Well, yes. I guess I would 13 say that you can see that here the area under the curve is z3
( )
's_/ 14 at about half, and here it's a third, so then I would --
15 again, this is just conjecture, but I guess I would go on 16 to say that I would expect that the dose that could be 17 released from such an accident to be a third or a half.
18 MEMBER APOSTOLAKIS: Fine. But I thought the l 19 argument that the staff was making for the inclusion of ;
20 the spray system, is not related to a reduction of risk.
21 In risk you said that, you know, the risk may even be low 22 enough it meets the goals.
23 What makes you uncomfortable is the 24 uncertainty about that risk.
("T
( ,) 25 MR. SNODDERLY: I guess my --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
488 1 MEMBER APOSTOLAKIS: So I don't know how this
,. 2 helps me to reduce that uncertainty.
( )
3 MR. SNODDERLY: Well, what I tried to 4 establish in the early part of the presentation was that 5 the only system available is the PCCS, is this system.
6 And this is what this system can do, and thereby, what 7 would an additional feature buy us? And I think it buys 8 us the ability to reduce aerosol fission products inside 9 containment quickly and substantially. And so I --
10 MEMBER APOSTOLAKIS: Sure, and nobody disputes 11 that, right? I don't think anybody --
12 MR. SNODDERLY: Right, give you a feel for --
13 MEMBER APOSTOLAKIS: Without the calculation
,m
- 14 you can say, yes, if I have a spray system I will have a 15 lower release. ;
i 16 MEMBER MILLER: But it gives you a feel for 17 the difference in time. I think that's the key point I'm j l
18 addressing here.
19 MEMBER APOSTOLAKIS: But, but -- l 20 MEMBER MILLER: Without a core spray it takes 21 about ten hours; with a core spray it takes about four.
22 MEMBER BARTON: Containment. Containment 23 spray.
l 24 MEMBER MILLER: Containment sir, I'm sorry.
((~ ,)^. 25 Containment spray, yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
489 1 1 MEMBER APOSTOLAKIS: Yes, so what? I mean, rw 2 the way I understand the product, the goals are met even
\
('x./ /
3 without a sprav. Right? So even if I divide -- I mean, 4 if I had a spray, instead of ten hours it happens in three 5 our four hours. And if I do something else I may even do 6 it, you know, in half-an-hour.
7 I have to stop somewhere. So the main 8 argument that is advanced by the staff -- and I think it's 9 a legitimate argument -- is that yes, the risks may be 10 low, but the uncertainties are so large that they make us 11 very uncomfortable. And we would like to see the spray 12 system to reduce the uncertainties.
23 And this analysis, so far, has not addressed
[~ \ l
's s .4 that, has it? The issues -- l l
15 CHAIRMAN SEALE: Certainly not with a complete l l
16 --
17 MEMBER APOSTOLAKIS: -- should be the 18 uncertainties, not the actual value.
19 CHAIRMAN SEALE: Certainly not with 20 completeness.
21 MEMBER APOSTOLAKIS: Yes, I think I don't see 22 it. Maybe it's hidden there, but I don't see it. And if 23 the argument of defense-in-depth is based on the spread of 24 these things, of the distributions, I think we need to see
) 25 that. We need to see the argument that by putting the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
490 1 spray there, that uncertainty is reduced.
,y 2 MR. HOLAHAN. Unfortunately in this case, what
- i 3 dominates the uncertainty is what hasn't been analyzed, 4 remember.
S CHAIRMAN SEALE: Yes.
6 MR. HOLAHAN: And I think the uncertainty 7 we're talking about is the incompleteness.
8 MEMBER APOSTOLAKIS: Okay.
9 MR. HOLAHAN. For example, the AP600 has 10 treated seismic events with a seismic margins analysis, so 11 none of the numbers you have seen or will see, include 12 seismic risks.
13 MEMBER FONTANA: Of course, that spray may be s
rm i
\ 14 ineffective under those conditions anyway.
15 MR. HOLAHAN. Well that -- sometimes it would 16 and sometimes it wouldn't.
17 MEMBER APOSTOLAKIS: So what you're saying 18 Gary, is that this analysis here is not really central to 19 the argument?
20 MR. HOLAHAN- No, I think this analysis shows 21 that the source term is substantially affected by a spray, ,
22 and it gives you an idea of the timing and the magnitude l l
23 of the risk. Sprays will reduce the source term in the 24 containment atmosphere quickly and efficiently, whereas !
l
[)
(_) 25 natural mechanisms are effective but they take a ,
I NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISt.AND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
491 1 substantially longer time.
- s. 2 MEMBER MILLER: So another way to look at it,
(~) 3 it will reduce that now unknown uncertainty for a 6-hour 4 period.
5 MR. HOLAHAN: Well, I think it doesn't -- what 6 it does is, it reduces the risk because the risk is 7 proportional to the source -- for a 6-hour period, 8 approximately.
9 MEMBER APOSTOLAKIS: No, a risk is a risk, 10 period.
11 MEMBER FONTANA: But basically, it's the area 12 under the curve that --
13 MR. HOLAHAN: The area under the curve is
/~N k- 14 changed by a factor of 4 or 5 or 6, or whatever the --
15 MEMBER FONTANA: Or maybe it's -- compared 16 with all the --
l 17 MEMBER APOSTOLAKIS: Why is it the area? l 18 Because the risk is proportional to an integral of --
19 MR. SNODDERLY: Well, I think -- I'm sorry. I l 20 think we're focusing a little -- I think this -- you've 21 brought up a very good point, and I think this analysis 22 does not address uncertainty in general -- which would be 23 the low net driving forces, the thermal hydraulic 24 uncertainty -- because all these analysis are very
(~)s
(_ 25 sensitive to containment pressure, containment NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 48".,3 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 492 l
l l 1 temperature.
7\ L N And so you get to the uncertainty from the 1i 1 l
\' ~)
3 fact that we have the passive containment cooling system 4 is -- the majority of that database is based on the one-1 5 eighth large-scale test.
1 6 And although that database is going to allow l
7 us to come to a regulatory decision concerning design 8 basis accidents, there's some uncertainty there about how 9 much margin is available to do that system.
10 There is also that we don't have a lot of 11 operating experience with a -- we don't have any operating 12 experience. So I think that's where the uncertainty is 13 coming in that you're talking about, that no, I don't have f}
(m ,/ 14 a slide or something I can show you to --
15 MEMBER APOSTOLAKIS: I understand.
16 MR. SNODDERLY: That risk is inherent -- or 17 I'm sorry, not risk but uncertainty --
18 MEMBER APOSTOLAKIS: Go ahead.
19 MR. SNODDERLY: -- is inherent, and I think 20 that a spray system does -- it gets you to the point that, 21 given a failure of the passive core cooling system, given j 22 a source term inside containment, the only thing we have 23 to reduce the aerosol fission products is the PCCS.
l 24 And then an additional mitigative feature such (A) 25 as a non-safety-related, single train, simple spray system NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 2344433
493 1 that Westinghouse estimates would cost some $415,000, and gg 2 I think our contractor estimates it somewhere around
( )
3 $237,000.
l 4 The point I'm trying to make is a simple 1 5 system costing somewhere, you know, less than a half-6 million dollars but more than $200,000 -- but something in 7 that range -- a reasonable spray system is going to be 8 able to do a lot.
9 MEMBER KRESS: George, it would be a mistake )
10 to try to interpret these curves in terms of risk because 1
11 these are all for unfailed containments. I 12 CHAIRMAN SEALE: Yes.
13 MEMBER KRESS: And the risk is dominated by
,r~\
, < s i_/ 14 early containment failure, and you have to --
15 MEMBER APOSTOLAKIS: No, but -- ,
1 \
l 16 MEMBER MILLER: -- and some bypasses would be l I
17 in there and the sprays don't affect bypass accidents. So l 18 you have -- you don't want to look at the curve just for i l 19 risk --
20 MEMBER APOSTOLAKIS: Yes, that's my question.
l i 21 But I think there is somewhere in the documents here that i l
22 early containment failure probability itself would be 23 affected by the existence of a spray.
24 CHAIRMAN SEALE: That's right.
g, (w,) 25 CHAIRMAN SEALE: As a matter of fact --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
494 1 MEMBER KRESS: It would affect suppression,
,c3 2 yes.
( )
wJ 3 MEMBER APOSTOLAKIS: It would still be a high 4 contribution. Well, I -- 3BE is what? Remind me again 5 what is that sequence?
6 MR. SNODDERLY: That's a failure of the direct 7 vessel injection line. It's a 4-inch break, and there's 8 also a failure of the IRWST to inject, and failure of one 9 of the two core makeup tank trains.
10 MEMBER MILLER: What's the probability of that 11 occurring?
12 CHAIRMAN SEALE: We were going to find out --
13 MEMBER APOSTOLAKIS: We'll find out later.
,r'x 14 Now, Gary mentioned the issue of incompleteness.
1 1
15 MR. HOLAHAN: Yes, I think the problem is ;
1 16 later, we're not going to find out what t'ne probability of 17 these things are. We're going to get some information on 18 that --
19 MEMBER APOSTOLAKIS: Yes.
20 CHAIRMAN SEALE: Yes.
21 MR. HOLAHAN: Because if I thought that we 22 could assign probabilities and have a complete 23 understanding, then I think all you do is, you take these 24 curves and you integrate out and you would compare them to p\
(_) 25 the safety goals and you make a decision.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
495 1 MEMBER APOSTOLAKIS: So you are saying that ex 2 there is a serious issue of incompleteness here? Is that
( )
</
3 what you're saying, Gary?
4 MR. HOLAHAN: Yes. But I --
5 MEMBER APOSTOLAKIS: I don't understand. I 6 mean, what I see here is fairly complete, so can you tell 7 us what the incompleteness is that drives your uncertainty 8 --
9 MR. SNODDERLY: Oh, I think the incompleteness 10 is, is this is just one data point. This is one sequence 11 of which -- it was just a typical sequence. I merely was 12 trying to give the committee a feel, because I don't think 13 -- I think there was some -- I sensed an apprehension from
/ 1
\'v) 14 the Commission that, we don't know what you're asking for.
15 Are you asking for a spray system with heat 16 exchangers, with dual train, with safety-related -- and so 17 all I tried to do here was to say, here's what a simple 18 spray system can do for you. But -- I'm sorry, I' n, not 19 answering your question. But the uncertainty -- so this 20 is just a point. I mean, this is --
21 CHAIRMAN SEALE: It's a part of a point.
22 MEMBER APOSTOLAKIS: What I'm trying to 23 understand is, if we could have -- can we write there 24 somewhere -- if there is a curve that is at some time -- I (x,
( / 25 don't know, maybe that's the wrong measure, but we pick a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBER $
1323 RHODE ISLAND AVE., N W.
(202) 2344433 WASHINGTON, D C. 20005-3701 (202) 234-4403
496 1 time and we have an uncertainty of the aerosol mass in the 1n 2 containment. Is that a good measure? I mean, that's what 3
\' ~j 3 you're doing here. Cesium iodide, aerosol mass in the i 1
4 containment. I 5 So if I pick a time, say four hours, I will 6 have an uncertairty about that value, right?
7 MR. SNODDERLY: Yes.
8 MEMBER APOSTOLAKIS: And that ur> certainty 9 comes from many reasons. Some of them I have quantified 10 already -- now, we're talking about the uncertainty in our 11 minds, not the actual curve we've introduced.
12 So that comes from a number of sources. You 13 have addressed here one of them, okay, 3BE?
14 MR. SNODDERLY: Yes.
15 MEMBER APOSTOLAKIS: I'd like to understand 16 m.a t are the primary contributors to that uncertainty, and l
17 how a containment spray system will help me eliminate 18 them. Now, can we draw a curve to there --
l 19 MR. SNODDERLY: I think I can help you out 20 there.
1 21 MEMBER APOSTOLAKIS: That's the key argument i 1
22 here, is it not?
23 MEMBER KRESS: You want uncertainties in LERF I i
l i 24 or in CDF or -- .
i
,rK (v) 25 MEMBER APOSTOLAKIS: Well, whatever measure 1
NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
497 1 they select.
,- 3 2 MR. HOLAHAN: I think the uncertainty is --
I )
3 MEMBER APOSTOLAKIS: They selected the mass of 4 cesium iodide aerosol in the containment.
5 MEMBER KRESS: Why do you want the uncertainty 6 in that?
7 MEMBER APOSTOLAKIS: I don't want it -- I 8 don't want it anywhere. I just want to understand why I 9 need this spray.
10 CHAIRMAN SEALE: That's not the problem.
11 MR. SNODDERLY: I don't think that's the 12 issue.
13 \~ w MEMBER APOSTOLAKIS: That's not the issue.
g
! )
N/ 14 Peak --
15 MR. SNODDERLY: I think the problem is, 16 aerosol fission product removal inside containment, and 17 that is sensitive based on a number -- okay, it's the 18 removal of inside of aerosol fission products inside 19 containment, and that is sensitive to a number of thermal ,
1 20 hydraulic parameters, such as containment temperature, 21 pressure, mole fracture of the steam condensation rate --
l 22 all these things. l I
23 And there's a lot of uncertainty with that and i 24 -- I'm sorry, Gary.
(~
(-) 25 MR. HOLAHAN: I don't think that's the point.
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
l l (202) 234 4433 WASHINGTON, D.C. 20005 3701 (202) 234-4433
498 l
1 I don't think that is the uncertainty that's important for
~3 2 this decision. That you can draw those curves and you can 1
() 3 do all the analysis you like and put uncertainties on the 4 spray analysis and put uncertainties on the source term 5 and on the no spray and on the natural removal mechanisms.
6 And what you'll find out is, is all of those 7 analyses say that a spray is several times more effective 8 than the no spray. But what's important to the decision 9 is, what's the likelihood of the sequence? That's what 10 you want to know.
11 What you're saying is, the source term between 12 the two, you can do the uncertainty analysis but what 13 you'll find out is you're quite certain that the spray is
,/m kY 14 more effective than no spray and what you want to know is, 15 how important is this? I 16 MEMBER BARTON: Right.
17 MR. HGLAHAN: And that's determined, not by an 18 uncertainty analysis of this, it is, what is the 19 likelihood of these sequences with a containment failure?
20 MEMBER FONTANA: For example --
21 CHAIRMAN SEALE: That's right.
22 MEMBER FONTANA: For example, this would have 23 no effect on the containment --
l l 24 CHAIRMAN SEALE: But we -- I'm sorry.
,( -)s 25 MEMBER FONTANA: For example, this would have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
! 499 1 no effect on the containment bypass sequence because that ,
l gg 2 never sees a spray. )
(
%/ )
3 MR. HOLAHAN: Well, that's not true. l 4 MEMBER FONTANA: Sure it would. If you have a l i
5 straight shot from the core to the outside --
6 CHAIRMAN SEALE: But if the availability of I
7 the straight shot depends on the internal pressure that's l l
1 8 within the containment building -- '
9 MR. HOLAHAN: Not necessarily.
10 MEMBER FONTANA: Not really.
11 CHAIRMAN SEALE: Not necessarily, but it 12 could.
13 MEMBER FONTANA: Yes. But the point Gary's i s
'- ' 14 making is very important. That you've got to kind of look 15 at the big picture and all the other things --
16 MEMBER APOSTOLAKIS: That's what I want -- a 17 big picture.
18 MR. HOLAHAN: But the problem --
19 MEMBER APOSTOLAKIS: In a qualitative way.
20 MR. HOLAHAN. The problem with the big picture 21 is, if we look at a sequence like this particular 22 sequence, it's probably estimated to be a 10" sequence.
23 And you're saying, well how important is it to knock down 24 the source term by a factor of 2 or 3 or 4 or whatever the
, /~
(_)h 25 number, for a 10" sequence?
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
500 1 And I think if you do that analysis it will j
,f y 2 say, well, that's not very important and I wouldn't be L.) 3 surprised if, a little later, we're going to hear that 4 from Westinghouse. And what you have to say is, there is 5 an unquantified, but I think, real benefit to a spray, 6 because you know -- what you're certain about is that it 7 has a positive benefit. What you don't know is, what's 8 the likelihood of needing it?
9 The sequences you've analyzed show that it has 10 a benefit but it's probably small, because the sequences 11 are at low probability. You have to ask yourself, what 12 haven't I analyzed? How convinced am I that this is, in 13 fact, a 10" core damage frequency, or a 10-8, large early
/ \ ,
\ '/ 14 release design?
I 15 And I think to convince yourself of that you 1
l 16 need to address lots of -- you have a real serious ,
17 completeness question. And in the absence of a complete j l
18 seismic probabilistic risk assessment from a seismic point 19 of view, and I think some other things that are not in the 20 PRA, you can't convince yourself that we are in fact, at l 21 the 10" core damage frequency and we should, you know, 22 leave this issue.
l 23 And it's that uncertainty. What's the 24 likelihood that -- in fact, maybe there are other things p)
(, 25 that are in the 10-6 range.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
501 1 MEMBER APOSTOLAKIS: So the issues I see are
,s 2 two, from what you just said. First of all, we have some
( )
L.J 3 quantif.ication of -- well, the fundamental problem seems 4 to be that the frequency comes from the low value of the 5 frequency -- comes from the preventative measures --
6 MR. HOLAHAN: Yes.
7 MEMBER APOSTOLAKIS: -- not mitigation?
8 MR. HOLAHAN Yes.
9 MEMBER APOSTOLAKIS: That's one thing that 10 bothers me -- seems to bother me. The second is that the 11 uncertainty around that frequency has not really been 12 quantified, and there is a serious -- even if you quantify 13 the parameters that you put in there -- you may go, you O
'Es/ 14 know, between 10~6 and 10-', or 5 X 10-6, 10-'. But what 15 you're saying is, the most important part is that there 16 are unquantified, or maybe unquantifiable, contributors 17 that you have to take into account.
18 Now, they could do a seismic analysis; they 19 chose to do the margins things, but they could do it.
20 MR. HOLAHAN: Yes.
21 MEMBER APOSTOLAKIS: The margins showed that 22 the numbers are low --
23 MR. HOLAHAN: Well, the margins show that, I l
24 think, the safety systems in the design meet the seismic (h
( ,) 25 margin of 1.67 above the SSC.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
502 1
1 MEMBER APOSTOLAKIS: So it all comes down to -
2 -
r'sg '
8 )
3 MR. HOLAHAN: Which I think we would say is )
1 4 reasonable, but not extraordinarily --
l 5 MEMBER BARTON: Robust.
l l
6 MR. HOLAHAN: Not extraordinarily robust. i i
1 7 MEMBER APOSTOLAKIS: So it really comes down '
8 to a policy decision. It is not a technical decision.
l 9 MEMBER KRESS: It's really just what I said. j l
10 This is a purely defense-in-depth issue; that's a policy 11 issue.
12 MEMBER APOSTOLAKIS: It's a policy --
13 MEMBER KRESS: And you cannot resolve it
(~3)
K/ 14 altogether on technical grounds.
15 MEMBER APOSTOLhKIS: Now is, do you want all 16 your defense to come from preventative measures or do you 17 want some of it to be --
18 MEMBER KRESS: And how much defense-in-depth 19 do you want? And how do ycu define it?
20 MR. HOLAHAN: That's right; it is that kind of 21 judgment.
22 MEMBER MILLER: I think corollary to that is, l
t 23 then how much will the spray actually improve your 24 defense-in-depth?
,\
() 25 MR. HOLAHAN: And the importance of the NEAL R. GROSS l COURT REPO'4TERS AND TRANSCRIBERS i
1323 RHODE ISLAND AVE., N W.
I (202) 234 4433 WASHIN3 TON, D.C. 20005-3701 (202) 234-4433
I 503 l 1 analysis we've seen is, if the spray wasn't very l
73 2 efficient, then in fact, that makes the decision very i
'% )
3 easy.
4 MEMBER APOSTOLAKIS: Yes, but the thing is 5 this now. You mentioned the unquantifiable seismic 6 contribution. Would it be fair to say though, that if you 7 had one of those earthquakes, the spray also would --
8 MEMBER KRESS: You have to factor that into 9 the reliability of the sprays.
10 MR. HOLAHAN Sometimes.
11 MEMBER APOSTOLAKIS: But we're not doing any 12 calculations now. We're talking in the --
13 MR. HOLAHAN: But you can make judgments about 14 them.
15 MEMBER APOSTOLAKIS: -- rarified space of 16 unquantifiable contributions.
I 17 MEMBER KRESS: The margins analysis kind of 18 tells you that seismic -- if you can do your plant design l 19 right, the seismic contribution is low enough to be 20 acceptable in terms of frequency. That's kind of what you 21 interpret.
l 22 MEMBER APOSTOLAKIS: I think that's what the 23 margins -- the hydrology has --
24 MEMBER KRESS: Which tells you its effects on
\
77
( ,) 25 the sprays is probably low enough to be acceptable in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHlhGTON. D C. 20005-3701 (202) 234-4433
504 1 defense-in-depth thinking.
73 2 MEMBER APOSTOLAKIS: So they're shooting i )
%d 3 themselves in the foot? Because if that means that, i 4 because the seismic contribution is so low, the impact of l
5 the earthquakes on the sprays is fairly low, therefore go 6 ahead and install the spray. l l
7 MEMBER KRESS: That would be one ;
1 8 interpretation, yes.
i l
9 MEMBER APOSTOLAKIS: That's really something. i 10 CHAIRMAN SEALE: Where are we?
l 11 MEMBER BARTON: Do we have any more questions 12 to Michael?
13 MR. SNODDERLY: Thank you for this opportunity p.
(
N ') 14 to come before you --
15 MEMBER APOSTOLAKIS: Wait a minute. I --
16 MEMBER BARTON: Do you have a question?
17 MEMBER APOSTOLAKIS: I have a question, yes. l 18 I really think that -- I didn't see in the documents that 19 were sent to us, or in the letters that, you know, you 20 have sent to the Commission, these kinds of arguments. I 21 don't think that, in fact the Commission said that you may l
l 22 be right, but you have not made your case. Isn't that 23 what they said essentially, at some -- in one of the SRMs?
24 MR. HOLAHAN: Yes. It wasn't accepted as r'
k,%) 25 presented.
NEAL R. GROSS COURT REFORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE, N W.
(202) 234-4433 WASHINGTON, D C, 20005-3701 (202) 234-4433
l 505 1 MEMBER APOSTOLAKIS: As presented, yes. So
'q
, 2 you know.
i
'~
3 MR. SNODDERLY: But I think -- I can't speak 4 for the Commission, but my understanding of what their 5 concern was, was --
6 MR. HOLAHAN: That they didn't understand what 7 was being proposed.
8 MR. SNODDERLY: Right. Do you expect a system 9 with heat exchangers with -- a safety-related system 10 which is going to be much more expensive and a much 11 greater impact? And so, that's why we went back with the 12 response describing what type of system we envision which 13 is, non-safety-related, single train, and would use (n)
U 14 existing equipment already onsite.
15 MR. HOLAHAN: But on a factual basis, it's 16 clear that the Commission didn't jump at that offer 17 either, or that --
18 MR. SNODDERLY: Or we wouldn't be here.
19 MR. HOLAHAN: -- referred it to the committee, 20 MEMBER APOSTOLAKIS: But it seems to me that 21 if you want to make a case, you really have to address the 22 issue of uncertainties, and maybe instead of going to 23 these calculations, which nobody -- I mean, you can do i 24 that too, but I think it would be very, very helpful to n\
() 25 explain which point it is you're dealing with, first of i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
,l (202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
1 l 506 l l 1 all, and what -- even draw curves freehand and say look,
.- s 2 we don't know what this curve is. And the major reasons
('~' )
! 3 are. And if you put the spray there, this is what we 4 expect will happen to these curves. ;
I 5 See, it's not an issue of absolute values; .
l l
6 it's an issue of uncertainty here, by mainly from '
7 unquantifiable contributors. So freehand drawings to make l
8 a qualitative argument, in my opinion, would at least 9 clarify your position.
10 And then it comes down to this policy issue.
11 You know, defense-in-depth means this. Even if you have 1
12 10-9, but it all comes from preventing core damage, should l 13 you do something on the other side in the name of defense-
/~S x-)
, 4 14 in-depth?
15 MEMBER BARTON: George, can we continue this 16 discussion --
17 CHAIRMAN SEALE: After we've heard from 18 Westinghouse.
19 MEMBER APOSTOLAK.TS: It's done now.
20 CHAIRMAN SEALE: No, no, it's not done. But I 21 think we'll have a more complete picture after we've heard 22 from Westinghouse --
, 23 MR. HOLAHAN: We want to get the question.
24 CHAIRMAN SEALE: -- their side of the concept.
, (m) v 25 MR. SNODDERLY- Thank you, Chairman Barton.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
1 j
507 l
l l
l 1 MEMBER BARTON: Thank you, Mike. Do we have 1 i
,s 2 time for a little break before we --
i \
's J ,
3 CHAIRMAN SEALE: Yes. I was going to say, I l 4 know that Brian is jumping at the bit to punch all this l 5 stuff out, but we're going to let him collect his thoughts 6 and marshall his arguments for about 15 minutes. We'll be 7 back at 10:30.
I 8 (Whereupon, the forer 'g matter went off 9 the record at 10:18 a.m. and went back on 10 the record at 10:34 a.m.)
1 11 CHAIRMAN SEALE: We're back in session. l l
12 MEMBER BARTON: Thank you, Mr. Chairman. At 1
13 this time we'll hear from Westinghouse and Mr. Brian i
/N l l 14 McIntyre.
15 MR. McINTYRE: Thank you. I'm going to be 1
16 presenting the Westinghouse -- I guess I'd call it a l 17 viewpoint -- on where we think we are with respect to 18 SECY-97-044. What we see the staff position as being is, 19 they are looking for -- if you trace this back also, I 20 guess it was the 96-128, and the time periods before that 21 -- of what the staff is looking for is additional -- in 22 96-128 it was some sort of an additional system and now
! 23 it's been gone, it's been changed to a containment spray l
24 system or an equivalent.
f~% 25 And the three reasons that they had were to y _)
HEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
508 1 address the uncertainties -- and we talked about that this
- f. 2 morning. The second thing is part of severe accident s .
3 management. And I think one of the best ways to look at 4 that is, when we first started talking about this, the 5 director of NRR at the time said, well you know, what do I 6 tell the governor I'm doing? :
I 7 So they're looking for some sort of an action 8 that can be taken. They're looking for some sort of a --
9 something that you can actually do -- and I think that may 10 be one of the frustrating things about a passive plant, 11 because Westinghouse marches in and says, trust us, it 12 takes care of itself. And if I were the staff I'd be 13 looking for something that I could turn on or look for
,m
\/ )
I 14 some way to be involved in this. ;
15 And then the third thing which we've talked 16 about a lot is, is for this balance between prevention and 17 mitigation, that we're still puzzling over to some extent.
18 And what the staff was looking for was, what they describe 19 as a very simple system -- and I'll talk a little bit more 20 about this in a second -- the single train as a few 21 thousand galluns a minute, an external water supply.
22 In the letter it described either actively or t
l i 23 passively driven. I could see what Mike presented today 24 was based on an active system. The original design of the
.fm. And if you
( j 25 AP600 did have a passive spray system in it.
NEAL R. GROSS
! COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433 l
509 1 really thought about that and you made it totally passive, l
f- x 2 it takes a fair bit of intelligence to know when to turn
\
/
3 it on, even if you're thinking and being able to analyze 4 all of the conditions.
5 When we were designing it to the TID source 6 term it was quite simple. If you got any kind of a 7 radiation signal you turned that on, all the stuf. rauld 8 come up, you cleaned out the atmosphere and it was dandy.
l 9 You only had to spray it for about a half-an-hour. i 10 Then as we moved to looking at more realistic 11 source terms, the 1465, then it had to be a much smarter l l
12 spray system. So then we were looking for different ways 13 to extend it out over time, and it would phase itself, and
(' \
'I 14 it got so complicated that we didn't think we needed it.
15 And also, we got to the point in the design 16 and doing the accident calculations that we just said, we 17 really didn't need it because we were meeting the 18 requirements and the safety goals that were laid down 19 without the spray systems.
20 And finally, it does not have to be safety-21 related but as we talked this morning, it would clearly 22 have to be seismic 2, because you couldn't have 23 ringheaders and things up on the top of the containment 24 falling down, taking out other things. It would have to tO
(_,/ 25 live through some sort of a seismic event to make sure NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
510 1 that it was there.
2 Now, when Mike was talking about his example f
-)
G
3 of the acceptable spray system, he made the comment that 4 we do have fire pumps that are rated at 2000 gpm, and that 5 is a fact -- we do have two fire pumps.
6 You need to keep in mind that -- that is 2000 7 gpm is runout flow. If you go back to that ABWR 8 calculation where we looked at the SSAR for the ABWR, and 9 that pump that was delivering 950 gpm with no containment 10 back pressure was also a 2000 gpm pump. If you look at 11 the systems design people there's a lot more of them 12 saying, I need a pump this big. It's got to actually make 13 that much water.
i )
'~# 14 And if you look at our pumps at 2000 gpm, or 15 even take this GE pump and say all of a sudden it's going 16 to be pumping against a containment backpressure -- which 17 even if it stays below design and it may not for a severe' 18 accident -- could be 50 or 60 or even up to service level 19 C of over 100 psi -- that's going to be a little difficult 20 to pump against. Plus we would have to pump it up to the 21 top of the containment, which I believe is 135 or some 22 number like that, so you'd also have that head.
23 And that was the basis of the spray system 24 that we came in and talked to the staff about when we
(-)
(,) 25 first started discussing this back in 1995. And that was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
i 511 l
l 1 the system that made -- I think it was -- 334 gpm at the l
2 crane rail, j
i 7-t
- A_/
3 So what we would have to do, and as Mike 4 pointed out, we would have to make some changes to the 5 plant -- some upgrades to the pumps. We would have to 6 make them certainly larger, certainly a much higher head, 7 a much flatter set of pump curves, and we would have to be 8 very, very intelligent as to when we turned this one.
9 Because to some extent, you can't just 10 continue to pump water into the AP600; you can't --
11 obviously, you're not going to fill this thing up to the 12 top. Mike had a 2000 gallon tank; I'm not sure how much 13 that would increase the flood level, n
- 14 At this point we would have to go back, I'm l 1
15 sure, and that the staff would be asking us, well if you 16 have the spray system in how are you going to protect this 17 from flooding, how are you going to do that? It would 18 require a fair bit of work to go back and do that.
19 And if you can only have 200,000 gpm for the 20 system Mike was talking about, what was it, 3.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> of -
21 - 200 minutes it must be -- that you want to make sure ,
22 that you did that at the right time. And if you look at 23 all the different accident sequences that would require 24 potentially, a pretty clever, I guess algorithm or set of l /~S.
t
(_) 25 conditions that you would actually use this system under.
NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
i 512 ,
1 We went back -- what I'm going to be talking i
7s 2 about is certainly much more in line with making a policy
( )
3 decision. I'm not going to have curves about lambdas and I
4 all those sorts of things, because I think it's really the 5 question you guys are faced with. Is how good is good 6 enough?
7 We went back and tried to find out, when you 8 get into the severe accident -- particularly the 9 mitigation features of what the requirements are -- and 10 it's our belief that the requirements of NEPA that came 11 out and were in the SECY-91-229 staff requirements 12 memorandum, is that's where you need to go back and look 13 at severe accident mitigation design features that are in r8
- 4
\2 14 the plant. l l
15 And the Commission came down and said, you 16 need to do that as part of your design certification 17 rulemakings. And that's where you look at it, and it's a 18 rulemaking, and that's it. It's been really a fundamental !
19 goal of the AP600 to resolve and to have mitigation design 20 features. It was something that's been built in from day-21 1, 22 It was one of the objectives of the ALWR 23 utility requirements documents and we conform to that, so 24 we're -- you know, this is something that's been built in, f3 25 which is one reason that we come at this stage and say,
(/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
! 513 t
1 well you've got all this and now you need this too.
.s 2 We looked at -- if you looked through our l \
\ /
'~'
3 SAMDA evaluation, it's in Appendix 1B of the SAR -- we had l
4 a brainstorming session at Westinghouse. We came up with 5 a number of SAMDAs that we thought could possibly help 6 things. We looked at 15 of those in detail, and they were 7 things like: more battery power, taking the steam 8 generator safety valves and running it back into the IRWST 9 to help deal with things like the containment bypass 10 issues.
11 We looked at two things in there -- both a 12 safety and non-safety-related containment spray. Neither 13 of those -- and I'll get to this in a few minutes in the
,m I
)
'w / 14 slide on risk -- neither of those passed the muster of 15 being required to be added under the -- I think it's $1000 16 per averted man rem is the requirement that we looked at. )
l 17 But yes, the safety spray system was one of l
18 the really nice, big, 2000 gpm systems like you see in 19 current day plans and in System 80+. The non-safety 20 system was pretty much the system that we had discussed 21 with the staff and our estimate was $415,000 to build that 22 system.
23 Our approach at that point was to make the l
24 cost of the systems as cheap as possible, because if we l 25 made it higher than possible it looked like we were trying l
!ou-)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
514 l
1 to lead the witness and weren't passing muster.
,esg 2 I told Mike during the break that if in the l
' ; 1 1 1
%_J 1 3 end the Commission does require us to have a containment I 4 spray, then whoever his contractor is will get the bid to 5 build that system for $237,000 because --
6 MEMBER BARTON: Fixed price. i l
7 MR. McINTYRE: Absolutely. You can't build a 8 reasonable house in Pittsburgh for that, so if we can do 9 that, he's got it.
10 MEMBER FONTANA: This is more than a house.
11 MR. McINTYRE: This is a lot more than a I
12 house, i
13 CHAIRMAN SEALE: Just faced with a whole bunch p)
\
'- 14 of judgmental issues.
1 l
15 MR. McINTYRE: It is. This is -- this is l
16 almost one of the times it's better to be sitting up here l l
17 than to be sitting back there having to decide how to make l l
18 the decision.
19 I had mentioned on a previous slide where we 20 have put a lot of what we consider our severe accident 21 mitigation design features in the AP600. And I'm not going 22 to go through all of these, but there are three slides of l
l 23 things that we have done to deal with mitigation.
24 I think that's important to keep in mind
(( ,) 25 because if you look at the low core dataage f requency that l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE , N.W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433 l
l
515 1 we have, the reaction certainly is, well you guys have S 2 just adJressed the prevention; you haven't addressed the t I x_/
3 mitigation.
4 And the mitigation doesn't necessarily show up 5 quite as obviously. As the low core damage people look at 6 that they seize on whether or not there are uncertainties 7 in it, but that's a very popular number and a very visible 8 number.
9 We have done a large number of just design 10 things. We have the ADS. It has a much lower leakage 11 containment. It's a 0.12 percent per day containment 12 leakage. We have increased the design pressure of the 13 passive RHR system; we've put an extra valve in it. It's
/ 3
% /
14 been moved inside containment -- but these are the types 15 of things that we have done that don't -- where do I get 16 my credit for that because it's not obvious when you first 17 look at it.
18 We have to deal with the containment isolation 19 issues. We have reduced the number of penetrations down 20 to 45 in this. But the nice thing of moving all these 21 things inside is you don't need as many penetrations. We 22 have located them in one room of the auxiliary building so 23 it's a small area. And I guess the good news about that j 24 is, it's easier to take care of and you can get some
/'~T
, ( j/
25 credit for settling.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
516 i
i 1 If we were to take it in the auxiliary l
7\ 2 building, sort of the bad news is, that almost provides --
l ( '~~')
3 when we were looking at the containment spray design, the 4 only way we would get a benefit was by spraying in that
, 5 area. It's a room; it's basically -- it's not where you l
6 get as much credit for spraying in the upper compartment.
7 It's a large containment; it has a high design 8 pressure; it's a 45 psi design pressure. We have ways to 1
9 flood underneath the reactor vessel that we've put in. We 10 have ways -- when you get into accident management 11 strategies -- which I think Mike talked a little bit about
- 12 -- you could turn on the normal RHR and remove energy from i
13 the containment.
l (~h
' 14 That way it can be run and you're not taking 15 it directly out of the atmosphere is the problem. You're 16 taking it out of the water that's at the bottom, but you 17 are getting energy out. The important thing, if you're l
18 going to reduce the pressure is to get the energy out of 19 the containment.
20 With regard to the fan coolers which we looked 21 at earlier, the same time we were looking at sprays, and 22 our conclusion was that the fan coolers, the way we looked i 23 at them, were as effective as the sprays with the 334 gpm 24 design.
f~h
( ,) 25 Mike's comment was that the containment fan l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
517 1 coolers as configured -- and I think that's a fair !
l
~s 2 assessment -- end up getting isolated during a LOCA until
!' \
/
~'
3 the containment atmosphere drops below 228 degrees, and 4 that's in response to a Generic Letter, and of course we 5 did make that change.
6 If you were having a severe accident and you 7 were really trying to resolve things, I'm reasonably sure 8 you would find a way to isolate those and use them. I 9 mean, that would just make sense. You would do whatever ]
10 you could do to try to reduce the pressure and get the i
11 energy out.
l 12 MEMBER FONTANA: I'm just curious. What's the 13 problem as far as for the ignitors -- hydrogen ignitors?
I 1
's
' 'h 14 MR. McINTYRE: They are run from the -- l 15 actually, we're in the process of changing that. We're 16 having to put a tech spec on that they can run from 17 offsite power or either one of the diesels, or the non-1E 18 batteries. I think there are four different ways that 19 they can be run.
20 MEMBER FONTANA: Thanks.
21 MR. McINTYRE: Our PRA results. And this has 22 two sets of numbers. It has the baseline and the now 23 famous, focused PRA -- which is the baseline PRA only 24 taking credit for the safety-related systems. This is a n
( ,) 25 slide that I use for other purposes and I just keep that NEAL R. GROSS l COURT REPO9TERS AND TRANSCR!BERS 1323 RHUDE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
518 1
1 on there to keep things in perspective. I g- 2 And if you look at the total baseline PRA for 3 at-power and shutdown -- Gary points out, without seismic 4 events because we did a margin analysis -- we're at about 5 2.3 X 10-7 with all systems., With just the safety systems 6 we're about 8 minus 6. And we're comparing that to the 10-7 4, and that was our objective and that's what we're doing 8 for the RTNSS -- probabilistic part of RTNSS for the 9 determination of what systems need to be looked at.
10 For large release we're at 3.2 X 10'8 for the 11 baseline PRA, and 8E-7 compared to a goal of E-6. That's 12 where we are in our PRA results. Now the question was for 13 risk.
r~w i
(
\'
)
14 MEMBER APOSTOLAKIS: Now, let's look at these l
15 numbers a little bit. l I
16 MR. McINTYRE: Okay.
17 MEMBER APOSTOLAKIS: I take it you have not 18 done an uncertainly analysis, so these are point values?
19 MR. McINTYRE: That's correct. l l
20 MEMBER APOSTOLAKIS: So if we look at the l 21 baseline LERF, which is 3.2 X 10~8, do you have a 22 qualitative feeling as to what the upper percentile would 23 be? I mean, are we talking about one order of magnitude i 24 higher, or higher than that? What kind of uncertainty are r"T
!.s , ) 25 we talking about here?
NEAL R. GROSS COURT REPORTERS AND TRAN3CRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
519 1 MR. McIUTYRE: No, I do not, g~ 2 MEMBER APOSTOLAKIS: The staff probably does,
\_./
3 but maybe not in numerical terms.
4 MR. HOLAHAN: I don't recall and we didn't 5 bring our Level 1 --
6 MR. McINTYRE: I think at this point it's fair i I
7 to say -- where we are in the review process there's ;
1 I
8 probably not a plus or minus on any of these numbers. I 1
I 9 We're still discussing what the reliability of the check l l
10 values and squib values are, so --
11 MEMBER APOSTOLAKIS: But it seems though, to 12 be central to this argument, does it not?
13 MR. McINTYRE: Well, let's get through the
(~h l
( )
14 next slide, because that's part of it, and I think the 15 question was risk. And this is -- hopefully this matches 16 up with what was presented earlier, where if you look at, 17 on a percentage basis, you look at the early containment 18 failure and it's like 86 percent, and if I were you guys 19 sitting there I'd say Oh my goodness, what are we going to 20 do about this?
i 21 And it -- the little red asterisk I need to 22 explain first. We did not do -- and this is building on 23 something that Gary had said earlier -- that how would the 24 containment failure - does the top fail, does the top r')
( ,/ 25 blow off it and everything that's inside gush out into the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
520 1 atmosphere? Is there a small hole ripped in the side of l
i
,m 2 it?
( )
</
3 To simply our calculations, that what this 4 assumptica is, that if we get a vessel failure we get a 5 containment failure. There's been no mechanistic 6 calculation done whatsoever. It's just, you know, we're 7 vot sure how to do this, that would be hard to do, there 8 would be a lot of uncertainty, and nobody can argue that 9 this is not conservative. I think I can make this 10 statement.
11 So keep that in mind; that these numbers were 12 certainly -- I guess I would say, in a reality, for 13 whatever that means, in this particular area.
I
\
)
14 MEMBER KRESS: Does your vessel failure 15 incorporate the flooding mechanism -- )
I 16 MR. McINTYRE: Yes.
17 MEMBER KRESS: There might be some questions l
18 about --
19 MR. McINTYRE: And the staff is going through 20 reviewing that, yes. In this case, things that we've done l
21 through help vessel failure is, there are no bottom run 22 instrumentation tubes in this, so there's no -- I believe 23 there are no penetrations below the core in the AP600 l
! 24 vessel.
l
(_) 25 If you J ook at this on a -- the question is, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND owe , N W.
(202) 234 4433 WASHINGTON, Of 2AS-3701 (202) 234-4433
i 521 1 so what do sprays do for you? It depends on the question, 73 2 how effective do you think the spray is going to be? The
( ' ' )
3 total number down at the bottom is basically 80 - 3.
4 Spray is not going to help either the bypass or I think 5 the isolation -- well, I guess it could help the isolation 6 failure some.
7 For the intermediate failure and the late 8 failure, it's probably not going to help those at all 9 because if you looked at Mike's summary -- and it's going-10 to run for 3.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> depending on when you trigger it --
11 I'm not sure how that would look for other sequences, 12 given --
13 MR. SNODDERLY: Well, that's one proposed n
l \
V 14 methodology, but I think the other thing is, if you just 15 pulsed this very system for 15 minutes I think you could 16 see there that you would get some benefit.
17 So it could be either -- I would see it being 18 used more as a pulsing effect than in watching your 19 containment radiation monitors, as opposed to running it 20 for the complete 3.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. But there's both of those 21 options.
22 Also, Dr. Kress, for the droplet size 23 question, it's 1.25E'3 meters. It was based on the Surry 24 drop size.
l') 25 MR. HOLAHAN: Maybe we should do this later.
( , .
NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W. l (202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
\
522 I i l
l 1 MEMBER APOSTOLAKIS: So your argument is that l
,3 2 the spray would affect primarily the early failure?
! ( )
3 MR. McINTYRE: If activated, it can help the l 4 earlier failure which, as Gary points out, is a point 5 estimate of 7E-9. Trying to put that 86 pcrcent in some 6 sort of a perspective, I think that that's important. Is 1
7 yes, it would help. I think you can always make things 8 better.
9 There's a possibility -- each one of the 10 SAMDAs that we look at would help something, somewhere, 11 somehow. But somehow you need to decide whether or not ,
i 1
12 you're going to do it. And it was our looking at the 13 SAMDA evaluation and what it's actually going to help and rx )
b' '
14 what sort of frequencies that we have decided to not do 15 that.
16 MEMBER APOSTOLAKIS: So this is the principal l 17 contributor to risk, isn't it?
18 MR. McINTYRE: Yes. On a percentage basis, 19 yes.
20 MEMBER FONTANA: I forgot -- what's the time 21 for early failure, that you use?
22 MR. McINTYRE: Two hours, I believe.
23 DR. UHRIG: This is turned on manually?
l l
l 24 MR. HOLAIEN: Yes sir. By "this" you're I
(v) 25 talking about the system that the staff would envision?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
523 1 DR. UHRIG: The proposed system, yes.
s 2 MR. HOLAHAN: Yes.
?
\
)
'~/
3 DR. UHRIG: And it has to be turned on at a 4 specific time?
5 MR. SNODDERLY: I would envision that. High 6 containment radiation, radiation monitors greater than 7 probably, 10 4 rads / hour would be the decision.
8 CHAIRMAN SEALE: Are you having a problem over 9 there? Okay.
10 DR. CRONENBERG: Can I just make a comment? I 11 think I remember in TMI, the operators shut the 12 containment off at spray anyway, right? They didn't want 13 to muck it up with sodium hydroxide. So I thought it was
\~ / 14 only on for seven minutes. I don't know if that helps 15 this argument, but we have sprays but they don't like to 16 use them.
17 MR. McINTYRE: So this -- I think I was asked 18 to come up with what the risks are, and I hope that this 19 helps. It shows what we're trying to -- where it might 20 help, where it probably won't help. But I think the point 21 is and the point that Gus made earlier this morning was, 22 clearly early failure is 86 percent or whatever the number 23 was, of the total risk from these categories.
24 A lot has been said about uncertainties. I 25 know that the paper talks about uncertainties and check NEAL R. GROSS COURT REPORTERS AND1RANSCRIBERS 1323 RHODE ISLAND AVE., N W (232) 234-4433 WASHINGTON, D C. 20005-3701 (202) 334-4433
524 1 valye operation and thermal hydraulics. We have spent at l
~x 2 this point, about $40 million on AP600 testing, and the
! I v
3 thermal hydraulic subcommittee has -- I am going to say, 4 beat this to a pulp. We've spent a lot of time -- beat us 5 to a pulp.
6 We've spent a lot of time talking about this, 7 trying to move this along, and I think that they are now 8 convinced that yes, we have run the right testing program 9 and we're getting to the point of completing the 10 validation of the computer codes. And these are the lists 11 of the types of things that we have tested.
12 We made some changes to the design to be 13 responsive to concerns that the staff had. There was,
\-) 14 from day-1 there was a big question on, gee you guys are 15 holding check valves shut for 23 months and 29 days and 16 you're having an accident. You've got 2200 psi and you're 17 going to tell us that 12 psi is going to coen this thing?
18 And we changed the design. We put Squib 19 valves between the high pressure and the low pressure in 20 that cace, so that the check valve isn't sitting there 21 being, in their case, concerned about welded shut or 22 crapped up with boron leaking through it and 23 crystallizing, or just anything.
24 We've tried to do design features to address
/
k ,h) 25 some of these uncertainties because you know, quite l
t NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
525 1 honestly, they are really the right thing to do. If you i
(- x) 2 can design something out, quite simply, that's an J t ;
\_/ :
3 appropriate thing.
1 4 The bottom line being that we have run a 5 significant number of tests. Finally, I think really the 6 important ones are the integral systems tests, both in 7 SPES and Oregon State University, and the large and small 8 scale containment tests that were run in Pittsburgh.
9 Which gets us to really what the --
10 MEMBER BARTON: Did any of these simulate fan 11 coolers --
12 MR. McINTYRE: No. The question on fan 1
13 coolers has been -- actually, I guess it isn't a question, p_
l Y ') 14 it's been more, well so Westinghouse, we'll give you l
15 credit as long as you run a testing program on the fan l 16 coolers. Knowing I think, fairly well, that we're j 17 probably not going to spend a couple of million dollars i
l 18 doing experiments on fan coolers for non-design basis
\
l 19 accident consideration.
l 20 So that's -- we think they'll work, we think 21 that they would make an effective, severe accident 22 mitigation design alternative, severe accident management 23 strategy. We think that they would be a help. We think l l
24 in reality that they would be used in the plant.
( ,) 25 Now, can we get credit for them? And when you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
WASHINGTON, D.C. 20005-3701 (202; 234-4433 (202) 234-4433
526 1 start getting into spending millions of dollars to get
,._ 2 credit for something, to me, as fuzzy as achieving an
('~' /
3 appropriate balance of prevention and mitigation without a i
I 4 quantitative goal, it's hard for me to sell that to my 5 management.
6 So the question gets down to, what is the 7 requirement? And I think that that's where -- actually 8 you guys might use these for your deliberations because 9 it's -- you're being asked to say how good is good enough?
l 10 It's our viewpoint that the SAMDA requirements 11 are what we should -- we think that that's what the 12 Commission has actually, you know, stood up and been 13 counted for. The results of the severe accident policy (3
4
'w./ 14 statement, the safety goals.
15 If you go back and look at the SRM ch a9-102, 16 the comment was, gee, yes we do expect the advanced 17 designs to be safer. We applaud the commi tment on the 18 part of industry to make these designs safer; however, 19 that the NRC will not use these design objectives as the 20 basis to establish new requirements.
21 And that's -- the emphasis in there is not 22 mine; that is the way it's written in the SRM. So it's, l 23 go ahead and make them safer, and we think that that's the 24 right thing to do, but let's not also raise the bar while
,m (v ) 25 these guys are trying to get over it. This is -- have an i
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS j 1323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 - (202) 234-4433 l
1 l
527 1 industry objective. j t
J
~ 2 Also in there, in that same SRM, is that the
(
'~'
l 3 safety goals are silent on the issue of cost, because 1
4 you're not -- and the price of safety, you don't look at 5 cost. However, when looking at how safe is safe enough, 6 that the safety goals should be something that can be used 7 as a yardstick for that.
8 And I think that that is also consistent with 9 where the ACRS was in their letter on April lith, where 10 they talked about raising "the CFF subsidiary goal to the 1
11 status of a fundamental goal can be considered as a 12 defense-in-depth principle that provides the balance 13 between prevention and mitigation" 7-M 14 So I think that we're in line with that.
15 We're meeting the safety goals, we think that that's what 16 Westinghouse thinks, that that's what the Commission had 17 in mind when the policy statements were passed.
18 And I think the first, the top bullet there is 19 something that we talked about earlier, is the risk-20 informed regulation, and it's certainly our perception 21 that telling us we have to have a spray system is not 22 risk-informed regulation. I search for the words that I 23 would say on the record.
24 It just -- it's -- we're supposed to be the
, ~N i
( ) 25 designers. We think we've met the requirements. I guess
%J NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
528 1 is probably the thing that I should leave. And we keep
,, ~3 2 looking for the appropriate balance. And I think that's
! )
\ /
3 probably one of the things that people wish had never been 4 written down.
5 This appropriate balance, I don't think it was 6 ever appreciated that we would be searching for some 7 miracle value with the entire ACRS trying to figure out 8 how you decide what is appropriate and what is good 9 enough.
I 10 CHAIRMAN SEALE: Do I take it you'd like to 11 borrow Gus' slide? The one with the --
12 MR. McINTYRE: Yes, it's --
13 MEMBER BARTON: The juggler.
3
- 14 MR. McINTYRE: Well no, I like the scale 15 better. There was things I liked about the juggler, but--
16 CHAIRMAN SEALE: Okay, sorry.
17 MR. McINTYRE: It's a difficult question.
18 Then we also wonder how this gets regulated in an 19 operating plant, or how would you -- for the reasons that 20 it's put in, how would you answer the 50.59 questions of, 21 are the consequences are the accident increased? Geez, I 22 don't know. For a system that's being put in to achieve 23 an appropriate balance between -- and they get to be very, i
! 24 very hard questions.
, r'n, 25 To be quite honest, I don't know how I would
( )
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
529 1 defend this. Because I know if we're told to put a ey 2 containment spray system in, I'm not sure how -- because I (v) 3 know the staff's going to come and ask questions. Well, 4 how did you guys design this? Is this big enough? How 5 are you protecting this?
6 And we're also going to be called in hearing, 7 at design certification time, to defend why this is in the 8 plant, why this is good enough. And there's an 9 interesting, I think, implications of the deliberations 10 that you guys are going to have later today on this. It's 11 a hard question.
12 I can tell you that it will not -- make one 13 thing clear -- it will not make or break the AP600. If
/s
- 14 we're told to put in a containment spray we'll design some 15 sort of a containment spray and put it in the plant. l 16 So in conclusion, we think we have done -- we 17 meet the -- what the ALWR URD was set out to do. We think l
1 18 we meet the requirements that the NRC has written down for 19 us to meet -- at least the ones that are quantifiable --
20 now we're into the qualifiable types of things.
21 We've done a significant amount of severe 22 accident mitigation design features inccrporated in the 23 plant. Our SAMDA evaluation shows that we don't need 24 things beyond -- and I think the word " additional" here is O.
( ,) 25 important -- beyond what we already have on the plant.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
530l 1 And we still - I think as was printed out g3 2 earlier -- that we still don't think it's neither t ;
Q/'
i 3 justified nor appropriate based on, at least our 4 interpretation of the regulation. And that concludes what 5 I have to say.
6 MEMBER BARTON: Any questions?
7 CHAIRMAN SEALE: Well, at this point maybe we 8 ought to ask you if you have any questions. Are there any 9 questions that you'd like to resolve while we've got all 10 of these people here, John, so that -- to help you in 11 trying to encapsulate all of this?
12 MEMBER BARTON: I don't think so. While I 13 think we'll continue to have some discussion later, we
-' 14 probably should have a representative of the staff stay.
15 But at this point I don't have any other questions.
16 MR. McINTYRE: Can I talk come more? I'd 17 forgot tnat I borrowed this slide from Mike, and 18 understand, I'm the licensing guy and I promised I 19 wouldn't have any curves. But I've been looking at this 20 and having sat through the meeting in California for two 21 days where we talked about this at great length, and you 22 start talking about the effectiveness of the sprays and 23 obviously, if you turn them on at one hour you can get l 24 this type of a number, if you can get two-and-a-half hours l
r~T
(_,) 25 -- and it's my understanding that if you believe what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
531 1 MELCOR says, that you can get that kind of a number, and 2 this gets down to, how good is the MELCOR number.
Is t
3 Say it's only half as good. So therefore, 4 maybe it's going to be somewhere in between, which would 5 make it look like, even without sprays, if you believe in 6 natural effects -- and this is where obviously, the 7 passive nature of the plant which Westinghouse is keen to 8 not upset, or certainly at least, limit the upset of the 9 passive plant design -- if this is a zero probability, 10 and you assume that it's going to be somewhere much, much 11 higher for a -- should leave it down in this direction for 12 the natural processes, then it's an uncertainty, but 13 certainly puts it on a parallel with having a spray
,r h t
k/ )
14 system, which is our belief.
15 DR. UHRIG: Is there any implication here that 16 misoperating the system would make the situation worse 17 than not having it? Turning it on at the wrong time?
18 MR. McINTYRE: That would be a concern that I 19 wouldn't want to have to defend. Because that gets into -
20 - and that's why I asked Mike in the break -- well, how 21 did you turn this on? And he gave an answer that would 22 look at one sequence. And that's all -- admittedly, he 23 said, you've done one set of calculations and this is for 24 the one sequence where we did the calculations back there,
,f -
, ! ) 25 that it would work.
m/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
532 l 1 Yes, but you'd need to look at all sequences l
,_ . 2 and a lot more sequences, and you still wouldn't know e i i ('"
/
3 about all sequences. So there would always be some -- I 4 use the word uncertainty -- in this system, as to whether 5 or not it would be there if it had been this operated.
6 On the part of the utilities, their reaction 7 would be, so now we've got this system. I think as Gus 8 pointed out, they really don't like to run these, and it's 9 not our envision that we would have sodium hydroxide in 10 it, but you wouldn't want somebody to accidently turn it 11 on and crap up the containment.
12 DR. UHRIG: Well, my experience goes back 15 13 years or so, but the plant people would do almost anything ry i
's>/ 14 to avoid an accidental trip of the spray system.
15 MR, McINTYRE: Yes, they would. That's why 16 they were encouraging us to not have one.
17 DR 'HRIG: Thank you.
18 MR. McINTYRE: Thank you for your time.
19 MEMBER BARTON: Thank you.
20 CHAIRMAN SEALE: Thank you, Brian.
21 MEMBER BARTON: At this point, I'11 turn the 22 meeting back over to you, Dr. Seale.
23 CHAIRMAN SEALE: Well, we have another item on i
i 24 the agenda for this morning. And I guess we should go on
,c w
(%j ) 25 to that. I want to thank the staff, I want to thank NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
533 1 Westinghouse representatives, thank our Fellow for their
,s 2 presentations, and obviously this is an issue we're going l \
x ;
' ~'
3 to have to delve into.
4 I'll remind you also we have that famous --
5 or notorious, as the case may be -- list of questions that 6 we have to decide how we're going to dispose of as well.
7 So this thing is going to require a lot of mileage this 8 afternoon.
9 In the meantime, we should go ahead with our 10 next item which is the report of the Planning and 11 Procedures Subcommittee. I would point out that after our 12 meeting yesterday, the NEI people were concerned that 13 there might have been some misinterpretation or lack of
[^h
\j 14 understanding of the activities they've had recently in 15 the steam generator area.
16 So they're going to be here at one o' clock and 17 we will address the Letter on that issue at that time. So l I
18 whatever we do we want to adjourn early enough to get back 19 at one so we will not inconvenience thos people in their !
20 decision to come see us. And so I guess the next item is 21 Planning and Procedures. John, are you going to help me 22 out on this one?
23 DR. LARKINS: Sure.
24 (Whereupon, the 442nd Meeting of the Advisory 25 Committee was concluded at 11:12 a.m.)
%/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C, 20005-3701 (202) 234-4433
l O CERTIFICATE l
This is to certify that the attached proceedings before the United States Nuclear i
Regulatory Commission in the matter of:
i Name of Proceeding: 442* ACRS Docket Number: N/A ;
Place of Proceeding: ROCKVILLE, MARYLAND I l
i 1
were held as herein appears, and that this is the original l
transcript thereof for the file of the United States Nuclear l'
Regulatory Commission taken by me and, thereafter reduced to O
typewriting by me or under the direction of the court
, reporting company, and that the transcript is a true and accurate record of the foregoing proceedings. I d
[CCIBETT RIN$R a
official Reporter Neal R. Gross and Co., Inc.
i i
l-lO I
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 R! LODE ISLAND AVENUE, NW (202)234-4433 WASil!NGTON, D C. 20005 (202)234-4433 l
I. -
INTRODUCTORY STATEMENT BY THE ACRS CHAIRMAN 442ND ACRS MEETING, JUNE 11-14, 1997 O
l THE MEETING WILL NOW COME TO ORDER. THIS IS THE THIRD DAY OF THE 442ND MEETING OF THE ADVISORY COMMITTEE ON REACTOR l SAFEGUARDS. DURING TODAY'S MEETING, THE COMMITTEE WILL CONSIDER
! THE FOLLOWING:
(1) POLICY ISSUE REGARDING CONTAINMENT SPRAY SYSTEM FOR THE AP600 DESIGN (2) REPORT OF THE PLANNING AND PROCEDURES SUBCOMMITTEE (3) PROPOSED ACRS REPORTS A PORTION OF TODAY'S MEETING MAY BE CLOSED TO DISCUSS ORGANIZATIONAL AND PERSONNEL MATTERS THAT RELATE SOLELY TO THE INTERNAL PERSONNEL RULES AND PRACTICES OF THIS ADVISORY l
COMMITTEE, AND INFORMATION THE RELEASE OF WHICH WOULD CONSTITUTE A CLEARLY UNWARRANTED INVASION OF PERSONAL PRIVACY.
THIS MEETING IS BEING CONDUCTED IN ACCORDANCE WITH THE PROVISIONS OF THE FEDERAL ADVISORY COMMITTEE ACT.
MR. SAM DURAISWAMY IS THE DESIGNATED FEDERAL OFFICIAL FOR THE INITIAL PORTION OF THE MEETING.
WE HAVE RECEIVED NO WRITTEN STATEMENTS OR REQUESTS FOR TIME TO MAKE ORAL STATEMENTS FROM MEMBERS OF THE PUBLIC REGARDING TODAY'S SESSIONS. A TRANSCRIPT OF PORTIONS OF THE MEETING IS BEING KEPT, AND IT IS REQUESTED THAT THE SPEAKERS USE ONE OF THE MICROPHONES, IDENTIFY THEMSELVES AND SPEAK WITH SUFFICIENT CLARITY AND VOLUME SO THAT THEY CAN BE READILY HEARD.
1 i
O
O O O W e G
/s .gR RE *>
+++++ i i
AP600 CONTAINMENT SPRAY POLICY ISSUE August W. Cronenberg ACRS Fellow -
i 442nd Meeting of the ACRS !
June 13,1997 !
t i
i i
AP-600 CONTAINMENT SPRAY SYSTEM-?
- Prior ACRS Position / Statements t
- Relevant issues to Consider i i
i
- Commission Request:
ACRS Opinion and Basis for that Opinion 1
433 ACRS MEETING (LTR. AUG 15/96)
- Conclusion We endorse the positions of the staff in addressing the 3 policy issues on AP600-e Prevention & Mitiaation of Severe Accidents j The staff is seeking approval of non-safety systems to address uncertainties associated with passive fission product removal for DBA analysis and balance between prevention and mitigation of severe accidents.-The applicant's submittals provide some support for demonstrating fission product removal using only passive mechanisms. Nonetheless, we are persuaded by the staff ,
position that systems beyond passive removal mechanisms should be evaluated to provide greater confidence-in mitigating DBA and severe accidents. We recommend Commission approval (of the staff position). ,l e e2 9 -jy
O O O
! 422 ACRS MEETING (LTR: JUNE 15/95) i l
- Item-7: Containment Performance The staff intends to use deterministic and probabilistic containment performance goals to review AP600 We believe the staff position is appropriate
- Item-10: LonaTerm Severe Accident Consecuences Post accident pressure in containment will remain positive longer than for a plant with active cooling-removal of radioactive species is expected to be less with passive means than using active sprays or filters-The staff believes this situation calls for additional means,-a nonsafety spray.
-We believe-that radioactive removal should be considered with respect to risk and the safety goal.
- Caveat:
Arguments on radionuclide concentrations would be unnecessary if a performance-based criterion were derived 3
COMMISSION STATEMENT ON AP600 ACCIDENT MITIGATION SYSTEM
~
The Commission agrees it is important that the AP600 design include adequate means for accident management and long term mitigation. However, the Commission does not support the staff's request for the inclusion of additional system (s) for accident
! management and long term mitigation following a severe accident as presented, not because it may be inappropriate, but because the l basic design and performance requirements have not been l bounded or specified, and the requested additional system (s) do
- not appear to be consistent with the concept of a passive design.
The Commission is open to reconsideration of this issue if the staff can be more specific in terms of what additional system (s) are contemplated, including the design and performance requirements.
l o
e -
O* O
O O O t
'i i
PRIMARY QUESTIONS FOR ACRS CONSIDERATION ,
- Defense-in-Depth View:
Is a spray system needed to provide defense-in-depth and to ensure adequate protection for severe accidents ?
- RiskinformedView:
Would a spray system provide sufficient reduction in risk ,
for severe accidents to justify its costs?
What is the technical basis for ACRS recommendation to the commission question 5 i
_-----------j
DETERMINISTIC PERSPECTIVE (Defense-in-Depth / Prescriptive Regulation) 4 e Aerosol Depletion Estimates (Models)
= >1 diffusiophoresis dominant
= 1-100 sprays dominant
= > 100 gravitational settling dom.inant ,
i
- Aerosol source term / timing?
- Containment leakage / pressure history?
e Validation of aerosol depletion mechanisms for AP600 containment conditions?
~
O O O -
RISK INFORMED PERSPECTIVE AP-600 Release Mode Risk Contribution
- Coniainment isolation Failure 9.6-%
Early Containment Failure 83.9-%
Containment Bypass 5.8-%
Other 0.7-%
- population boundary dose risk -72 hr; PRA-Rev. 9 7
AP-600 CONTAINMENT SPRAY SYSTEM-?
3..-
c
).-
f-
~
< C- p,/
Q if i
\ i
\\ !
\ l I
\'
i p
, \1
\
(- .
.)
~
~s "'
Defense-in-Depth Risk Informed
~
e e' O -
!.. 3 4
)
l l
SUBJECT:
Need For Additional Mitigative Feature i
in AP600 l
l DATE: June 13,1997
! PRESENTER: Michael Snodderly 1
i O TirtE/ ore: Reactor Systems Engineer Containrnent Systems & Severe Accident Br Division of Systems Safety and Analysis Office of Nuclear Reactor Regulation
\
i l l TELEPHONE: 415-2047 i 1
l 1
i
- O l
~
O O O .
. Policy issue Before The Commission '
f, i
It is the staff's position that the AP600 include a containment spray system or equivalent for accident management following a severe accident Slide 1
O O D .!.
Need For Containment Spray i i
l The staff is concerned about the ability of the AP600 containment to remove aerosol fission products during a severe accident I
- l i Given a failure of the passive core cooling system, the containment i becomes the primary mitigation system ,
t e Containment Fan coolers are isolated during a LOCA and remain isolated until containment atmosphere drops below 228 F i
i Slide 2 i
!T ' ET- - . - - . .3
"~
' =E
.4 Sescre Accident Phenomena Treatment O
i CASE 3BE-1 3BE DOMINANT SEQUENCE DVI LINE BREAK ;
Containment Upper 'emportment Temperature l
- --- C o n t a i nme n t Lowe- -
oportment Temperature '
480 e. - 400 l
^ : '
m x 460 _
g v
- fg - 350 v 440
- ~ {IP.I
- 420 44 - 300 o
- if ' u
' 00 b %dm _
- 250' { $l K
D ?'
r _. 'e->
' n::- - 200 o et360 '
E a
- E o 340 w -
- 150
- n 320 -M ' ' ' '
0 6 10 15 20
'c j 25 30 Time (br) l k
l Slide 3 l
Figure 3411 l Case 3BE-I: Containment Temperatures j F.evision: 8 September 30,1996 m,\ap600pravev,Buc34 3 wpf.1MN 496 h ,_ Yj Westinghouse 34 78 .
O O O -
Impact of Sprays The ability to intervene and provide control over the course of a severe accident has benefit in terms of accident ~ management t
[
Aerosols inside containment can be reduced :n a short time period with sprays Slide 4 I
,a as se es se a 5 5 E33gd[ drR m 1. 5 . . . . , . . . .
~
l -
j - ._____ 90 PERCENTILE .
H z i __________ 10 PERCENTILE
~
a W t o 1. O _2 - A MEDIAN _
- 's ,- s
'g E
CONTAIN W -
g .
O s U - s -
[ s 2
o H
s________________________ _
I v" s o. 5 -
,'N -
Z I i '.'
r
,/ \, ,
x .
l 's # 1 =
=.
f -
j \~~. .,
2 -
~ ~ ~ ~ ._------~~~----.---~~~----....
O
- 9 -l __ . .
W -.
a . . . .... .... .,.. ....
5 10 15 20 25 TIME (HOURS)
Slide 5
, 14. Comparison of Instantaneous Decontamination Coefficients to Predictions Obtained With the CONTAIN Code
L O !
" Summary of Results for Csl Aerosol Depletion in AP600 for 3BE Sequence 14 , , , .i i .i .i ,
m 3 i .
. . . [
h . -. - - - - . - - - - - . . - - . ._-- . .
v: 13 . . . . , .
.f. . . . '
. : Cal Released : >
- 3 - * - -~ -- - --- - -- -.- ----- - - -- - -- -- - -
a a 3g -- -- - -.
j j to Contain%t j
- e { { !
C - . . .:- . -
t
.- - : . - . - 4 ._
z _
11 . . . . . ...a.
S . .
10 ..
- .. ., J:
. C.". .
f (U . _ . . -
. _ . . . . . . __-.--.-.--m . ...-.-
9 . . . . _ . .. .
., u .
i C . .
-r O _, . . .-.,-. - ., -- -
8 ,_. . . , , .
Q m . .
--j
--j- -. j- - - - - j - . ;- .
g g 7 .- : ,
f.r--],_,s., - -
t
/ .
.a
- _ J- - .- .J -- - .,
..L. - . ..
6 m
. ... ~,
m . . .
- .. - 4
__r-
.1 - .
5 .. . , ,
. . . . i
. , . m. % . . . '
's. .'
- . . . SputeF 9# ,- .. - .- .. {
4 1
3
- "0"" .
. s.
ui .
e ca .
. t .
s .
. t t
] O .. ...s. - ..J -
-[ - - - ,.,---- - - y - - - - l. '. - --
8,.
,L
..1 n .
e
. .ig. .
. . . ,. s. . .._- -. .-,.__. . ... . . .. - - -- - -
2 ... - . .. .
e .. . . . --...
. .,.. ..- q ..
s.;..s.
D ., ., s. . . .
e s . .. . .
- j . . _ . - - j . . - . -- g _ - . g - p. e g.- - J .a i w-4 % q-t O ,
I
"'T --
1 E I
E 1
1
'-- l l
- 0
~
-1 0 1 2 3 4 5 6 7 8 9 10 !
- Time - (hrs) s t
Slide 6 i
a
n
'G) (o~ .]
e AP-600 (3BE Sequence) a C with simplified model 27 O
C i MELCOR without hygrosecpic model '
86 E
E c 5 I O I E
.E 4- l
.5 C
O O 3 ,
o m
2-lambda of 0.4 per hour /
o CD x 1 NNg
= J.
0= - i i i i i i 0 1 2 3 4 5 6 7 8 9 Time (hr) ,
i Slide 7
O O O
.l Example of an Acceptable Spray System l
t A simple containment spray system capable of injecting into containment without dedicated pumps or heat exchangers e Non-safety-related e Single train For aerosol removal, a single actuation of the spray system will suffice because pH control measures in the cump should prevent revolatilization The sprays would be expected to mix the containment atmosphere, especially the boundary layer inside the containment shell i
t The heat capacity of the subcooled spray water will reduce containment pressure upon injection for a short time i
Slide 8 i
O O O -
Example of an Acceptable Spray System (Cont.)
Both evolutionary designs have ac-independent water addition systems The ABWR system consists of piping and manual valves connecting the fire protection system to the containment spray ring header The diesel driven firewater pump for the ABWR dehvers .
approximately 950 gpm flow with no containment backpressure j Both designs provide an external hookup outside containment for i connection of a fire truck pump to an alternate water source l
AP600 Fire protection system includes an electric motor-driven pump and a diesel engine-driven pump .,
Each pump is rated for 2,000 gpm Slide 9 n
Conclusion The staff believes that either the use of a simple containment spray system or the use of the fan coolers (backed by supporting data) could be found acceptable in addressing the uncertainties associated with mitigating severe accidents, and for providing operator intervention capability as part cf an accident management strategy l
Slide 10
_. a o o o -
J l !
PRESENTATION TO ADVISORY COMMITTEE ON REACTOR SAFEGUARDS AP600 TECHNICAL AND POLICY ISSUES .
SECY-97-044 JUNE 13,1997 WESTINGHOUSE ELECTRIC CORPORATION Brian McIntyre 1
s 0524BAM WPF/I
O O O PREVENTION AND MITIGATION OF SEVERE ACCIDENTS Lkn SECY-97-044 NRC POSITION The staff recommends that the Commission approve the AP600 require the use of an additional severe accident mitigation feature in the form of a containment spray system or equivalent for accident management following a severe accident.
4
- 1. To adatess the uncertainties associated with the passive systems in mitigating severe accidents
- 2. Operator intervention as a part of long term accident management i
- 3. For balance between prevention and mitigation of severe accioents Very simple system Single train A few thousand gallons per minute External water supply Actively or passively driven Nonsafety-related ;
os24a AM wrFt2
~:
O O O SEVERE ACCIDENT MITIGATION <,
SECY-91-229 Requires SAMDA Evaluation be Completed as a Part of 10 CFR 52 Design Certification Rulemakings .
- We Believe This is the Regulatory Requirement Severe Accident Mitigation is a Fundamental Goal of the AP600 Design
- Significant Severe Accident Features Built into AP600 Design Results of AP600 SAMDA Evaluation
- Fifteen SAMDAs Selected for Final Evaluation
- Including Safety and Nonsafety Containment Spray
- No Additional Features Justified i
i 0524H AM Wif/3 ll
~
O O
~
O
~
i SEVERE ACCIDENT MITIGATION DESIGN y FEATURES INCLUDED IN AP600 DESIGN km Containment Bypass Mitigation .
- Passive residual heat removal and automatic depressurization system (ADS) mitigate steam generator tube rupture without the need for operator action.
The ADS is a multi-redundant and diverse system .with control from both the protection and safety monitoring system and the diverse actuation system.
- Lew leakage containment !
l
- ADS mitigates high pressure and temperature induced tube ruptures i
- Passive residual heat removal system is inside containment :
- Normal pumped residual heat removal system has increased isolation with a third barrier, and has a higher design pressure such that it will not rupture at i reactor coolant system pressure. ,
a s
05248 AM WPF/4
O O O SEVERE ACCIDENT MITIGATION DESIGN m FEATURES INCLUDED IN AP600 DESIGN hm Containment isolation Failure Mitigation
- Reduced number of penetrations -
- Penetrations are into the auxiliary building where any radienuclide leakage can settle rather than being released to the atmosphere
. Open penetrations connected to the reactor cooiant system or containment are i fail closed ,
- Use of smaller valves with higher reliability and lower leakage
- Large dry containment with high design pressure
- Containment isolation controlled by both the safety-related protection and safety monitoring system (PMS) and the nonsafety-related diverse actuation system (DAS) i 0524H AM WPF/5 .
P
O O O
~
SEVERE ACCIDENT MITIGATION DESIGN q r FEATURES INCLUDED IN AP600 DESIGN kN Severe Accident Phenomena Mitigation Passive containment cooling is a fail safe system for decay heat and fission product removal.
Hydrogen control in a large dry containment via redundant safety-related passive autocatalytic recombiners (PARS) for slow release and using igniters for fast release
- ADS mitigates high pressure core damage accidents
- Reactor vessel flooding to prevent vessel failure and debris relocation to the containment. No bottom penetrations on reactor vessel.
. The normal residual heat removal system can provide heat removal from the containment.
=. - . _ . . . .
O O -
O PROBABILISTIC RISK ASSESSMENT RESULTS N Core Damage Frequency Large Release Frequency At-Power Shutdown Total At-Power Shutdown Total 1.7E-07 5.5E-08 2.3E-07 1.8E-08 1.4E-08 3.2E-08 f"[U"'
7.7E-06 4.1E-07 8.1E-06 5.5E-07 2.6E-07 8.1 E-07 E'{"5'd p
[sarety ,g,,4 ,g.,,
0524BAM.WPFD >
O O O AP600 RISK
SUMMARY
N Release Frequency 72 Hour Risk Population TEDE Dose (per year) (man-rem) (man-rem /yr) ,
Early 6.6E-9 1.0E6 6.8E-3 Failure Intermediate 1.3E-11 3.5E5 4.6E-6 Failure Late 1.5E-11 1.5E4 2.2E-7 Failure Isolation 3.6E-10 2.1E6 7.7E-4 Failure Bypass 1.1E-8 4.2E4 4.7E-4 Totals 8.06E-3 0524 B AM.WPF/8
).m
O O
~
O it 1' AP600 TESTING PROGRAM Component Testing .
- RCP Air and Water
- High inertia Rotor Single Parameter Effects Testing
- Air Flow Path Resistance Water Distribution ' '
- Wind Tunnel
- Heated Plate
- Core Makeup Tank t
- ADS Integral Systems Testing
- Small Scale Containment
- Large Scale Containment
- SPES 2
- Oregon State University ,
I I
t,524B AM WPF/9
o o O
WHAT IS THE REQUIREMENT ? k. i SAMDA Requirements Severe Accident Policy Statement .
Safety Goals SRM on SECY-89-102 (June 15,1990) ,
- "The Commission Applauds Such a Commitment. However, ttip NRC Will Not Use industry's Design Ob_iectives as the Basis to Establish New Requirements"
- "The Safety Goals are Silent on the issue of Cost, But dc Provide a Definition of i
'How Safe is Safo Enough' That Should be Seen as Guidance on How Far to go When Proposing Safety Enhancements ... "
Risk Based Regulatory Acceptance Criteria - ACRS letter of April 11,1997 t
- Safety Goals as Risk Assessment Criteria
- " Elevating the CDF subsidiary goal to the status of a fundamental goal can be l' considered as a defense-in-depth principle that provides balance between prevention 'and mitigation" l
0524BAM WPF/10 [
I
~~
o 0 0 -l WHAT IS THE REQUIREMENT ? $
Risk informed Regulation
- SECY-97-044 Request Seems Counter to this Concept t
What is the Appropriate Balance Between Prevention and Mitigation this Spray System is to Achieve?
How Would Such a Spray System be Regulated in Operating Plant?
= 10CFR50.59 l
i 0524BAM WIWil i
_ _ - _ _ _ _ . _ _ _ _ _ . - - . _ _ _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ . _ _ - - _ _ _ _ _ _ _ _ -?
O O -
O '
CONCLUSIONS N AP600 is the Most Tested Design Ever Reviewed by the NRC AP600 Meets the ALWR Utility Requirements Document Which was Reviewed and Approved by the NRC Significant Severe Accident Mitigation Features included in the Design SAMDA Evaluation Results Were That No ADDITIONAL Severe Accident Features Required Addition of a Containment Spray to the AP600 is Not Justified or Appropriate Based on Regulations 0524B AM WPF/12
_____________71