ML20210S342
| ML20210S342 | |
| Person / Time | |
|---|---|
| Issue date: | 09/23/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1548, NUDOCS 8610080098 | |
| Download: ML20210S342 (290) | |
Text
i ORIGINAL' WAW /5$T O UN.11ED STATES NUCLEAR REGULATORY COMMISSION IN THE MATTER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEES ON CONTAINMENT. REQUIREMENTS AND REACTOR SAFEGUARDS O
~
LOCATION: WASHINGTON, D. C. PAGES: 1- 245 DATE: TUESDAY, SEPTEMBER 23, 1986 5 L.l l [hh l
~
00 $0} Remoy9 f;9 00 ICO Ae-FEDERAL REPORTERS, INC.
OfficialReporters i
/h) g 444 North Capitol Street Washington, D.C. 20001 (202)347-3700 NADONWIDE COVERAGE 8610390098 860923 PDR ACRS PDR
[
T-1549
PUBLIC NOTICE BY THE
[a'T UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS TUESDAY, SEPTEMBER 23, 1986 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards _
(ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
No member of the ACRS Staff and no participant at
() this meeting accepts any responsibility for errors or inaccuracies of statement or data contained in this transcript.
l l
e O
. - - . - , - . . - . - - - , - - - - , - - - . . - - - - - - - - - - , - - - - - - - - - - - ~ + - - - - - - - - - -
r i CR28229 0 1 DAV/dnw
-- UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION l AVISORY COMMITTEE ON REACTOR SAFEGUARDS 4 ,
5 . SUBCOMMITTEES ON CONTAINMENT REQUIREMENTS I -
6 AND 7 REACTOR SAFEGUARDS I
8 9 Nuclear Regulatory Commission 1717 H Street, N.W.
10 Room 1046 Washington, D. C.
11 Tuesday, September 23, 1986
( Th.e subcommittees on Containment Requirements and Reactor Safeguards convened at 9: 00 a.m. Dr. Carson Mark, 14 presiding.
15 l 16 17 ACRS MEMBERS PRESENT:
18 DR. CARSON MARK DR. DAVID OKRENT 19 DR. CHESTER P. SIESS
! DR. WILLIAM KERR 20 l MR. JESSE C. EBERSOLE l
MR. CHARLES J. WYLIE 21 l ACRS CONSULTANT.PRESENT:
22 ,
! DR. PLESSET 23 24
_.'> 25 i l
ACE-FEDERAL REPORTERS, INC.
~ 2,02-347-3700 Nationwide Coverage 800 336-6646 l
2290 01 01 2 bAV/bc 1g PROCEEDINGS 2 (9:00 a.m.)
3 DR. MARK: The meeting will now come to order.
4 This is a combined meeting of the subcommittees of the ACRS 5 on Containment Recuirements and Safeguards.
6 I am Carson Mark, subcommittee chairman. Other 7 members of the ACRS in attendance are Jesse Ebersole on my 8 left, David Okrent, Chester Siess and Charlie Wylie.
9 William Kerr, we expect to arrive later this morning.
10 We have a consultant, Milton Plesset.
11 The purpose is to discuss first a draft of a 12 position paper on containment requirements design 13 objectives; CPDO, I think it's known as now, which is being 14 worked on by the Research Branch of NRC and I believe will
, 15 come into a final form late this year or early next year.
16 And there is also an NRR generic letter having to 17 do, of all things, with BWR Mach I containments.
18 Dean Houston, on my right, is the cognizant ACRS 19 staff member for this meeting. The rules for participation 20 have been announced in the notice of this meeting, 21 previously published in the Federal Register on September 22 lith.
l 23 A transcript of the meeting is being kept and 24 will be made available, as stated in the Federal Register 25 notice. It is reauested that each speaker identify himself l ACE-FEDERAL REPORTERS, INC.
! 202 347 3700 Nationwide Coverage 80 4 336 4 646
2290 01 02 3 1 AV/bc 1I or herself and speak with sufficient clarity and volume so 2 that he or she can readily be heard.
3 We have received no written comments from members 4 of the public. We've received no requests for time to make 5 oral statements from members of the public. So the meeting 6 will now proceed.
7 I have rather little to say before we go on. I 8 should say that I do have some concerns in mind about the 9 general subject of Containment Recuirements being reexamined 10 at this time this year.
11 One is that we ought to be very careful that we
,s 12 are not thrown off the track by Chernobyl. The only reason U 13 for mentioning that is that so much has been made, so 14 ridiculously, so fatuously about the fact that, because of 15 our containments, Chernobyl cculdn't happen there; and, 16 therefore, we must rush in and examine our containments.
l 17 I think there's just no connection between those 18 events, and we should be careful that we don't invent one or l 19 fall into it.
20 The other concern I have is that we have fairly l
21 recently subscribed to or had imoosed a safety goal policy l 22 by the Commission. There has to be, and there ought to be
! 23 some connection between the safety goal policy and what is
^ 24 or isn't to be worried about in connection with containment.
d(~
25 If we have a gas-cooled reactor, then we don't 1
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
2290 01 03 4 1 DAV/bc 1 need a three-foot containment if it is true that it can't 2 possibly have a core melt.
3 If we have a past generation light water reactor, 4 then it's true that we may want to worry about that. We 5 need to face the need. We have to sort these things out 6 and in some way make sure that we don't get focused on 7 containment as a thing in itself.
8 It really is a part of a fairly complicated 9 complex, and we need to know where the emphasis then 10 belongs. We shouldn't be coming out with the idea that 11 containment must have a probability of 10 to the minus X of 12 failing in the event og a core melt. And if the core melt 13 has a small enough probability that we shouldn't be worried 14 about it, we need to keep this clear, particularly with 15 respect to any rules, regulations that may emerge on this 16 subject. And there's a threat of there being rules and 17 regulations.
l 18 We've heard on quite high authority that the Mach 19 I containment has a 90 percent chance of failing in the l
20 event of a severe accident. This is complete garbage. The i 21 person who made the statement doesn't know from nothing what 22 the probability is, nor does he define what a severe
, 23 accident means.
l 24 We've got to be a little careful that we don't 25 buy such a statement and do something on that account.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nuionwide Coverage 80 4 336-6646
i 2290 01 04 5 l'%
kJDAV/bc 1 Do other members of the subcommittee have 2 comments?
3 MR. EBERSOLE: I just wrote down a few topics 4 here which comes out of somewhat ancient history, and I 5 would like, if we don' t have the present speaker amplify 6 these matters, they might be considered later.
7 The first item is I'd like to have discussed the 8 matter of the reliability of valve closures which ensure the 9 reliability of the containment. These are notably the purge 10 and vent valves, but also include steam generator isolation 11 valves, where you have large steam generators and the s 12' potentiality of coupling the discharge from more than one 13 vessel.
14 I'd like to have discussed another ancient topic, 15 the electrical protection, the short-circuit protection of 16 the electrical penetrations, since, in my memory, they 17 turned out to be the fast fuses on the circuits, including 18 those for the pumps.
19 And I recall that they were tripped with non-1 E 20 breakers and there was a puny little matter of having to 21 isolate the seismic event from the pressure relation of s
22 containment by invokino that these would not be 23 coincidental.
I thought it was interesting that the NRC staff
{} _
24 said that we would never expect a LCOA to generate an 25 ACE-FEDERAL REPORTERS, INC. .
\ h. 202-347 3700 Nationwide Coverage 800-33M64,
,a _ _ . - - - - -
1 2290 01 05 6
('T iIDAV/bc 1 earthquake and, thus, cause a mechanical upset.
2 Of course, the inverse is what one would think 3 about, that a seismic event would cause a depressurization 4 and then we'd have to deal with hostile conditions in the 5 containmen t in the face of degradation of the electrical 6 circuitry, which would not qualify.
7 One of the worst things about the containment 8 problem to me is like the old Army posts that gave you a 9 week when the General visited so you could pick up all the 10 cigarette butts, polish up the place. He could come in with 11 his white gloves and everything was okay.
12 Such is the way we test containments. We work on 13 them a while, fix them up. And when they're. tested, they 14 work.
15 I'm wondering, if we went in and said today, 16 "We're testing the containment fast with big compressors,"
17 and management didn't know that, what the results might be.
18 And, finally, I have a question on the matter of 19 venting. In general, the main thrust of the venting 20 question is:
21 What's the venting logic to prevent core melt 22 when you're trying to hold the radiation in, but you can't 23 get the heat out?
24 Containment is really a peculiar strainer. It (J3 25 lets the heat out or it better, but it keeps the radiation ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
2290 01 06 7
4 /DAV/bc 1 in, and that requires eculpment.
2 In the case of one kind of reactor, you know, 3 they had direct open cycle cooling. But, anyway, how do we 4 currently contemplate venting?
5 What pressure and temperature logic do we have to 6 prevent core damage? And to guarantee both the relief 7 opening as well as the subsequent need to close?
8 When do we require filters in this process, and 9 when don't we require filters?
10 And what might be the nature of those?
11 I've often thought that a suppression pool can be 12 built like a small lake next to the plant to do some 13 stripping if it would be needful. It would be very cheap 14 and simple. Well, nothing is simple; I withdraw that 15 statement.
16 Finally, I'd like to here on a more basic theme 17 what is the logic of releasing small amounts of radiation 18 from the containment to prevent a severe release which might 19 occur if you did not?
20 And where is the documentation on that?
21 That's all.
22 DR. MARK: Thank you, Jessie.
23 David...
2d DR. OKRENT: I'd be interested in understanding
}
25 the staf f's thinking on containment performance design ACE-FEDERAL REPORTERS, INC..
202-347-3700 Nationwide Coverage 800-336 6646
2290 01 07 8 O
l_jDAV/bc 1 objectives as they apply to current plans and future plans.
2 Is it the same objective or do they have to find 3 different objectives?
4 That's one point. I'd be curious to know if they 5 are including external events. I would assume they are.
6 But, as of now, NUREG 1150, for example, the IDCOR work, at 7 least in part, doesn't.get into this, which I think is a 8 deficiency that has to be remedied soon.
9 And just a minor point for our staff, I got a 1.0 partial document, which I appreciate, since I didn't carry 11 it down here. That was the report from Vermont Yankee, 12 which I think had 180 pages or something.
J 13 I would like to get the full document.
14 DR. MARK: Thank you. Chet...
15 DR. SIESS: Nothing now.
16 DR. MARK: Charlie?
17 MR. WYLIE: I'm fine.
18 DR. MARK: Our consultant, Mr. Plesset?
. 19 MR. PLESSET: Nothing.
20 DR. MARK: With that guidance, Mr. Burdick, could 21 you tell us about the thoughts that are building up on the 22 subject?
23 MR. BURDICK: Thank you, Mr. Chairman. Good 24 morning, members of the subcommittee.
! (
25 (Slide) i I
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336 6646
2290 01 08 9 (3
a_/DAV/bc 1 Some of your concerns I can address. Some are, I 2 think, under the heading of further implementation of what I 3 am going to discuss this morning, and would necessarily
'4 would have to be addressed in more detail under other 5 existing programs.
6 Venting. You're going to hear about that, at 7 least for the boilers, I think, from Bob Bernero this 8 afternoon. But I think, in the main, you will find 9 satisfaction in the performance objectives a little higher 10 level perhaps than some of you might desire at this point.
11 But I think you'll find that we have discussed it 12 reasonably.
}
13 ( S' lid e )
14 Let me give you some background to start off 15 with. Actually, there's a couple of items missing here.
16 The ACRS actually did recommend, I think, in a couple i 17 memoranda previous to this EDO, memorandum to the staff on 18 July 29, 1985, recommending that the staff go ahead and
- 19 develop in conjunction with work on the safety goals the l
L 20 safety goal implementation period; a containment performance 21 design objective.
22 We did draft a program plan. This was sent out i 23 for review in October. There was a briefing to the ACRS 24 subcommittee on safety philosophy technology and criteria.
- O, 25 This was October 1985.
l ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage A00-336 6646
l 2290 01 09 10 1 bAV/bc 1 We selected a group of experts and solicited 2 their views on containment performance design objectives 3 during the period November '85 to January '86. We developed 4 initial CPDO options for review by this group of experts.
5 The intent here was to have it reviewed prior to 6 the CPDO workshop at Harper's Ferry, May 12th and 13th, 7 1986.
8 Another important item bearing on the contaiment 9 performance design objective is of course you realize is the 10 containment for the Commission policy statement on safety 11 goals, which was just approved in June of '86.
12 Something bearing directly on our further O 13 activities is a remedial memo to us in response to the 14 safety goal policy statement, telling us that we would have j 15 to conform to that statement.
16 MR. EBERSOLE: Before you take that down, can I 17 ask you one cuestion? I have a particular reason for asking 18 this, since I have to say something about it at some point.
19 Is there a compendium of some sort? Or what has i 20 been our experience in finding faults with containment that 21 we have thought were built the way they should have been?
! 22 I'm talking about the experience in testing 23 containments and analyzing the containment design?
24 Is there some sort of collection or history of 25 defects? I've listed a few, you will recall. Is there an ACE-FEDERAL REPORTERS, INC.
202 3.t7 3700 Nationwide Coverage 800-336-6646
2290 01 10 11 (D
a /DAV/bc 1 integrated package like that that we can look at and 2 extrapolate perhaps to where we're missing other things?
3 MR. BURDICK: I know where such information 4 exists in pieces. I don't know if any one compendium --
5 Jim, do you know if any single document exists?
6 MR. COSTELLO: Yes. I'm sorry. I pointed in the 7 wrong direction at this.
8 Ji'm Costello, NRC staff. It is an Appendix J 9 kind of testing?
10 MR. EBERSOLE: With containment integrity?
11 MR. COSTELLO: There is a long history of 12 Appendix J type tests.
7-V I'm not talking about tests, I'm 13 MR. EBERSOLE:
14 talking about findings.
15 MR. COSTELLO: Of the tests and their outcomes.
16 There has been in the last five years or so a summary by 17 American Nuclear Insurers of at least a tabulation of data 18 of passing and failing of tests.
f l
l 19 MR. EBERSOLE: Oh, there is?
20 MR. COSTELLO: If you'd like some of that, I'm 21 sure we could get it for you.
22 DR. SIESS: They don't tell you how big the leaks l
t 23 were.
l 24 MR. COSTELLO: No, it's just either pass or not i s 25 pass.
I ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
2290 01 11 12 Il L /DAV/bc 1 DR. SIESS: Nobody knows the answer to the other 2 one. There are many type B tests and penetrations where all 3 they could record was the leakage was in excess of some 4 amount; because they couldn't measure it, they don't know 5 whether it was leaking inside or outside.
6 DR. MARK: There is an attempt to get a look at e
7 this on the part of Bill Brown of the staff here from 8 LER's. So what leaks were there when they weren't supposed 9 to be there, and to what extent, in general, I think you 10 must have a copy of that.
J 11 DR. SIESS: The Appendix J test is not intended
- 12 to provide that kind of information.
~
13 MR. EBERSOLE: No. Of course. It's a standing 14 leak test. It's like the General at the Army post after two 15 weeks of preparation.
16 DR. SIESS: Not always. It varies, depending on 17 how runs it and how. You've got to know the relationship 18 between Type A and Type B and Type C tests. The integrated 19 leak rate test is a type A. And it never finds any leaks ;
20 that you couldn't have found in a type B and type C because 21 it doesn't leak through solid steel.
22 And as I recall, recently, Clinton ran an 23 integrated leak rate test and passed it and had a great big i
i 24 hole in the steel liner.
25 MR. COSTELLO: I don't know about that one, ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6M6 l _ _ _ _ . , _ _ _ . _ _ _ _ _ . , _
2290 01 12 13 f3
'JDAV/bc k 1 Chet.
2 DR. SIESS: Somebody had burned a hole in the 3 steel liner and they still passed the integrated leak rate 4 test.
5 MR. EBERSOLE: Do you remember they used to cut 6 the pipe, slid down the middle and welded it across all the 7 containment as well.
8 What's the difference?
9 DR. SIESS: They still have leak channels on most 10 containment liners. Not the steel ones.
11 O 13 14 15 16
(
17 18 19 20 l 21 22 23 r 24 25 ACE-FEDERAL REPORTERS, INC.
- 22-347 3700 Nationwide Coverage 800-336-6M6
2290 02 01 14 DAVbw 1 DR. OKRENT: Do we understand why they passed?
2 DR. SIESS: They probably just didn't leak out to 3 the concrete.
4 DR. OKRENT: There were no pads, you're saying?
5 DR. SIESS: I haven' t got the details on how big 6 the hole was.
7 DR. OKRENT: But it's not ordinarily designed so 8 it would seal everywhere through the concrete and steel.
9 DR. SIESS,: I don't think we'll ever know.
10 MR. EBERSOLE: I didn't get a very straight 11 answer as to where I could lay my hands on this. I'll have 12 to try to lay my hands on whatever summary sort of data I 4
(1) 13 can.
14 DR. OKRENT: Could I ask, does there exist in the 15 Research Office or somewhere else a study of containment 16 performance, including all of what I'll call inadvertent 17 losses of containment integrity, whether it was the valves 18 that were enclosed, and so forth and so on?
19 DR. BURDICK: This information exists, for 20 example, in LERs.
21 DR. OKRENT: It's not been taken together, as far 22 as you know?
23 DR. BURDICK: I don't know if there's any one, as 24 I said before, single document residing in research or 25 elsewhere. There are going to be many problems with ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
2290 02 02 15 O There's going to be problems b 'DAVbw 1 implementing the safety goal.
2 with our measurement of safety. There are going to be 3 questions of what do we do with plants to make sure that 4 they adequately implement the goal.
5 These questions, I don't have plans to present to 6 you answers to them at this point. I would like to get out 7 to you what the goal is. Then we can see more clearly what 8 problems there are in implementation.
9 (Slide.)
10 Let me move ahead then with the Containment 11 Performance Design Objective Workshop..
12 We did reach some conclusions out of that 13 workshop. Although in many cases, there was no consensus 14 among the members, there were clear conclusions that we, the 15 Staff, could draw, that is, that contaimment is a necessary 16 element of a defense-in-depth philosophy. We need 17 containments. A minimum containment performance level 18 should be specified, and this was preferable, if it was i
19 quantitative, and to the extent feasible, it should be 20 mechanistic. And it was repeatedly made very clear that the l
23 CPDO and aspects pertaining to it had to be very well 22 defined and unambiguous.
23 And out of the meeting at Harpers Ferry, we did r- 24 think that one of the proposed containment performance
( }e 25 options was promising and was the basis for our further ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
2290 02-03 16 m
bYDAVbw 1 work.
2 For those of you who were there, it was option 5, 3 I believe.
4 DR. MARK: Is this all clearly in the context of 5 the present style or even advanced style lightwater reactors 6 with enormous stored industry in the system, or does it 7 distinguish between that and a type of reactor of which we
,8 don' t have any, which doesn' t have those characteristics?
9 Is it clearly --
10 DR. BURDICK: We do make a distinction between 11 existing and future reactors.
,- 12 DR. MARK: The performance objective needs to be U 13 specified. Well, it's need is a function, I hope, of the 14 thing we are dealing with, and if we are dealing with the 15 present style of LWR, that is one thing. If we are dealing 16 with an HTGR, that surely is something else.
17 DR. OKRENT: By the say, Carson, I think, 18 historically, that the follow on larger HTGRs, one was 19 called Summer and Fulton, perhaps, the other, were designed 20 with containments that were not built. There was a change, 21 and you might want to call it the containment design l
22 philosophy.
23 DR. MARK: Well, that is history that you are rs 24 more aware of than I, for sure. It seems to me we ought to b be careful that we aren't talking of containment per se,
- 25 4
ACE-FEDERAL REPORTERS, INC.
3rl-347-3700 Nationwide Coverage 80433&6M6
2290 02 04 17 1 DAVbw 1 independent of the thing we are talking about containing.
2 DR. OKRENT: I absolutely agree.
3 MR. EBERSOLE: May I ask a cuestion? Chernobyl 4 has got everybody all excited and the media are making hay 5 out of the sensational part of the accident in attempting to 6 disseminate the notion that we have some potential for a 7 Chernobyl like accident. Now it is a greater seller of 8 paper and TV and books, which is the prime product of the 9 media, the prime product to be sold.
10 Can you say anything about the integral aspects 11 of containment philosophy en the cold water reactors? I am 12 spepking of something-like Hyper, the MTR, the OMR, any q( /
13 reactor where you don't have all this accumulated energy and 14 the enormous problem of heat removal. Reactors which tend 15 to shut down, practically cool themselves under cooler 16 water, but which if one the other hand. undergo uncontrolled 17 excursions, this cooling can becone real sticky.
18 DR. BURDICK: Well, no, I can't address that 19 question here. Some of those reactors have design features 20 that wot.'.d make their implementation of goals perhaps easier 21 than others. Again, it is an implementation type problem.
22 MR. EBERSOLE: But are their requirements in any 23 integral form for these, I'll call it, cold water reactors?
24 DR. BURDICK: Not that I know of.
{
25 MR. EBERSOLE: You don't have any thermal energy, ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 804336%m
)
2290 02 05 18 bAVbw 1 except in the core?
2 DR. BURDICK: Matt, do you have any response to l
3 that? ;
4 MR. TAYLOR: Matt Taylor. I've been here quite a 5 while, and I am familiar with some of the recuirements, but 6 I would guess it would be sort of ad hoc. For example, the 7 CPPR, with respect to attentuation factors, it was ad hoc'd 8 to the particular reactor.
9 DR. MARK: I think what I am concerned about is 10 not different from what Jesse was just saying. The 11 statement on which you are working should make it as clear 12' as one can to what it relates. It relates to reactors of b'~
13 some type, these types, and if we had a reactor with very 14 different characteristics, we would have to go back and do 15 it again.
16 MR. EBERSOLE: These types and differences 17 between PWRs and BWRs, they are refined differences.
18 DR. MARK: There is more difference than that.
19 The modular HTGR is a totally different item. It shouldn't 20 be stuck with a 10 to the minus 3 containment failure, if it 21 doesn't need a containment at all.
22 MR. EBERSOLE: That is a more drastic one. I am 23 getting further down. For instance, the activity level of a 24 PWR is enormously, I think, ordinarily higher than that of a i
73 U
25 running boiler, because it doesn't go to the turbine.
ACE-FEDERAL REPORTERS, INC.
202 Nationwide Coverage 804336 4 46
.- . _ -347-3700 . - ,__. . _ , - - , -- . - , . .
2290 02 06 19 DAVbw 1 DR. OKRENT: If I can go back historically, when 2 the general design criteria were being developed,'it was 3 known then that there was active work on liquid 4 metal-cooled fast breeder reactors and other concepts. A 5 conscious judgment was made that one should develop general 6 design criteria, primarily, if you will, applicable to 7 lightwater power reactors and that other reactor types might 8 or might not fall under the design criteria or that there 9 would need to be modification -- deletions, additions, and 10 so forth, rather than try to prepare one set of criteria 11 that one was confident was all encompassing.
12 DR. MARK: Fine'. I believe that line of thought
() -
13 should be consciously in mind.
14 DR. BURDICK: I have got it written down.
15 Again, we must conform to the Commission policy 16 statement on safety goals, and we will refresh you here.
17 There are two qualitative safety goals having to do with the 18 risk from competing sources. These have quantitative 19 statements. These are on individual and societal risks and 20 were given by the Commission guidelines for regulatory 21 implementation of these safety goals. And simply stated, 22 the overall mean frequency of a large release of radioactive 23 materials to the environment from a reactor accident should 24 be less than one in a million years of operation.
r3 LJ -
25 DR. MARK: That is per reactor?
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 804336-6M6
2290 02 07 20
, bAVbw 1' DR. BURDICK: Yes.
2 MR. EBERSOLE: A point of clarification. Reactor 3 accidents, historically, have been the classical one, the 4 standard one, the large LOCA.
5 Does that mean, and should it not say, including 6 severe accidents?
7 DR. BURDICK: Well, you've come right to the 8 heart of the problem.
9 MR. EBERSOLE: I know.
10 DR. BURDICK: We have to define what is meant by 11 a large release; yes.
12 MR. EBERSOLE: You can have a large release with O. 13 a LOCA. Well, I don't know. I will withdraw that. If you 14 cool the core well, you don' t need to have one, but if you 15 do a part way job, you might.
16 So I am just trying to eliminate the notion that 17 the people would draw a conclusion, "Oh, that is the old 18 classic LOCA."
i 19 DR. BURDICK: I think you will see clearly that 20 that is not the case.
21 DR. MARK: I don' t remember it exactly. I guess, 22 in retrospect, I had an uncertainty whether the one in a 23 million was per country or per reactor.
24 DR. BURDICK: Fer reactor year.
25 DR. MARK: It wasn't clearly stated.
l ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336 6646
l 2290 02 08 21 1 bAVbw 1 DR. SIESS: No accident for the next 20 years was 2 in the U.S, but the large release was just 10 to the minus 6 3 per reactor year.
4 DR. MARK: That is what I thought was intended.
5 My question was whether it was clearly spelled out that 6 way.
7 DR. SIESS: One that said no accident in the 8 U.S. It wasn't in 20 years either. They forgot to put the 9 time on it. That was the zero probability of having a core 10 melt accident in the U.S.
11 DR. MARK: That's where k got lost a little.
g-- ,
12 DR. BURDICK: It is per year of reactor b
13 operation, very clearly stated.
14 (Slide.)
15 There are other important issues that we had to 16 wrestle with besides the definition of large release. There
- 17 was the overall consistency, of course, with the safety 18 goals and the CPDO. We also feel that the main driving 19 force here for the containment performance design objective 20 is defense in depth.
21 Also there has been a lot of consideration given 22 to venting in the case of accident situations. Also we 23 wanted to come out with something that was reasonable and j r^g 24 achievable. We would have wasted our time, of course, had
\m) 25 we come out with something that was totally not achievable l
l ACE-FEDERAL REPORTERS, INC.
- 202 Nationwide Coverage 80 4 336-6646
- --, _ _ . . -- . -347-3700_
2290 02 09 22 (O,
ldDAVbw 1 at all.
2 DR. OKRENT: Excuse me. I am curious that you 3 pulled out what I'll call one subtopic, venting, when there 4 are a multiplicity of subtopics related_to containment.
5 DR. BURDICK: I think it was the feeling among 6 some experts at the Harpers Ferry workshop that this should 7 be addressed. It certainly seems reasonable to me,. and I 8 think it is reasonable to our staff, that since this is a 9 very topical item, we should consider it and include it. If
, 10 you object --
11 DR. OKRENT: Well, my point is that if you're 12 going to get to what I'll call " subtopics," there are (O
another half dozen or so that are at least of equal 13 14 importance.
15 DR. BURDICK: This is not, yes, all-inclusive.
16 DR. OKRENT: That I would class es important 17 issues. But go on.
18 MR. EBERSOLE: Just a minute. I just want to ask 19 you one example. Let's say that I have this hypothetical 20 thing, which for so many years stood in front of us, the 21 large LOCO, and lo and behold, the ECCS system worked pretty 22 well with one train, but there was a bit hole in the 23 containment and the valves that were providing ventilation 24 didn't close.
25 Would that be called a large release?
ACE-FEDERAL REPORTERS, INC.
.. 3 w-~ - necv_ m~ _
2290 02 10 23 DAVbw 1 DR. BURDICK: There is a large hole in 2 containment, and you have a large LOCA?
3 MR. EBERSOLE: I have a large LOCA. The ECCS 4 worked, but the containment didn't work. You lost all the 5 primary coolant. I wouldn't think that would be called a 6 large release.
7 DR. OKRENT: I can't seem my way to the end of 8 the event, Jesse, because if I lose all my primary coolant, 9 and I lost my makeup water, you have to tell me something.
10 MR. EBERSOLE: Well, Safety Guide No. 1, I think 11 it was about 1967 or thereabouts, was because the NPSH would 12 suck water out of -- in this case, it would be the
-)
u) 13 doughnuts, would require that you have a preclosed 14 containment and guarantee that you have the presence of 15 noncondensable atmosphere for that increment of suction 16 pressure.
17 Do you follow me? That was the No. 1 Safety 18 Guide, which 's now -- what do they call it now? The " Blue 19 Ribbon Regulati.a" we have. And Dave makes the point that 20 we are going to lose a lot of steam and water through this 21 big hole. And at least at first, we are going to be getting 22 our water through the cooling water storage tank, and 23 whether that will simmer down and contain itself at a decent
{} 24 temperature. I think it will.
The first boiler would have had a problem.
25 ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 80 4 336-6646
-2290_02 11- 24 O
L_/DAVbw 1 Clearly lost in the first burst would be the 2 noncondensables. Fortunately, it was sealed and inerted.
3 later, unfortunately, it was not.
4 So if we are talking about large release, I get 5 the impression here that unless we degrade to the state 6 where we lose the coolant, we would not be having a large 7 release; correct?
- 8 DR. BURDICK: That appears to be the case. I 9 have a rationale here, which I think leads us to a 10 reasonable definition of a large release.
11 (Slide.)
12 Let's look'at the prompt fata'lity safety goal.
13 It appears that if we meet this prompt fatality safety goal, 14 we're probably going to meet the latent safety goal.
15 Let's look at mortality tables. Out of these 16 tables, we can see that this translates into the average 17 individual risk within a mile from the site boundary, being 18 less than or equal to about 5 times 10 to the minus 7 early 19 fatalities per reactor year.
20 Now if we let R be the large release low and mid 21 risk to individuals in the downwind sector, to be in 22 conformance here, then we have to have R over 16 less than 23 or equal to at 5 times 10 to the minus 7 early fatalities 24 per reactor year.
O 25 i
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33H646
2290 03 01 25
(~h b 'bAVbur 1 DR. MARK: What is that number 16?
2 DR. BURDICK: That comes from assuming a uniform 3 population of 16 sectors in the circle. You have 1/16th 4 chance of being in that 1/16th sector downwind.
5 MR. EBERSOLE: Do they stay there and breathe 6 deeply?
7 DR. BURDICK: I am just trying to offer a 8 rationale.
9 MR. EBERSOLE: What I am saying is, is evacuation 10 part of this?
11 DR. BURDICK: I am just working down from the 7y 12 safety goal. This is not telling me what individuals -- how
(_J 13 they are breathing.
14 DR. MARK: But it must be on the basis of the 15 people staying in place?
16 DR. SIESS: There is a fixed population 17 unifo rmly?
18 MR. MARTIN: Jim Martin, from the Research 19 staff.
20 You have to look at history on how these goals 21 were set up and what was going on when all of this was being 22 considered, and in all of the risk studies of evacuation and 23 sheltering it had been considered, but for the most part
{} 24 there have been very kind of delayed kind of evacuation 25 assumptions.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336 6646
2290 03 02 26
(?
k -DAVbur 1 So for the most part the people in close to the 2 plant got exposed,to most or all of the plume and a whole 3 bunch of contamination besides, but evacuation indeed has 4 been considered in all the PRAs. It is very delayed 5 evacuation.
6 DR. BURDICK: This deals with the average 7 individual, Jim, is that right? It does not consider --
8 MR. MARTIN: That is something the Commission did 9 not explicitly address in coming up with a policy 10 statement, exactly how do you make this calculation, but 11 historically, yes, ir. deed , that is part of the risk
- 12 analysis.
13 In fact, even today we are considering in 14 NUREG-ll50 calculations -- we are considering what we know 15 about some fraction of people who just won' t move fast, 16 people who just simply will not move. And so we assume 17 today -- actually, they are the ones that are at high risk.
18 We assume everyone else is at low.
19 DR. MARK: Now, the ground contamination term is 20 zero if the people move even tomorrow?
21 MR. MARTIN: Not if you have a large release.
22 That is a very, very major part of that pathway.
23 DR. MARK: But a major part of that pathway is in i 24 the next week, not in the next 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of the ground 25 contamination?
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
2290 03 03 27
,o b-DAVbur 1 MR. MARTIN: As used in NUREG-0001, and I guess I 2 can send you a graph which would show you that the dose rate 3 is a function of time and varying distances for a large 4 release. We need to be talking wind to 100 R per hour, 5 depending on where you are and how the release is.
6 DR. MARK: Look, I understand, vaguely at least, 7 how those terms would go, but a great deal depends upon your 8 assumption as to whether you evacuate when you evacuate. If 9 you can't get out in the first hour, then you get so much.
10 If you can only get out tomorrow, you skip most of the 11 ground release.
12 MR. MARTIN: Going back tg my bottomline response 13 to your question, historically in risk analyses, indeed, 14 protective action by the public are part of the assumptions 15 in coming up with a risk assessment. The Commission didn't 16 say whether we should do that or not.
17 MR. EBERSOLE: Let me see, I was going to 18 introduce the problem. Sometimes a release can be a release 19 not to the atmosphere but to the confinement around 20 containment to catch the leakage. Unfortunately, that 21 confinement might be a building. In that case you have got 22 to have some kind of basis for control of the heating, 23 electrical apparatus and whatever.
24 Do you in the course of thinking about this, 25 including the degraded cascade state where you leak into the ACE-FEDERAL REPORTERS, INC.
M-347 3700 Nationwide Coverage 80 4 336-6646
2290 03 04 28
(~
b DAVbur 1 auxiliary buildings and then expose the operators -- and we 2 are talking about operator types now.
3 DR. BURDICK: _ I am dealing here with the attempt 4 to define the large release.
5 MR. EBERSOLE: Straight to the atmosphere? You 6 are not going to talk about it coming where the penetrations 7 are and going into the machinery?
8 DR. BURDICK: Not at this point, no.
.; 9 DR. MARK: And you are talking, I presume, about 10 people offsite?
11 DR. BURDICK: Yes.
12 DR. OKRENT: It is my recollection that one or
( two of the Commissioners suggested that measures of a large 13 14 release which would lead to what you call values of R, I 15 suppose, or which would lead to quantities of radioactivity 16 released would be substantially smaller because I think they 17 were talking about 5 R at the site boundary if someone 18 didn't move two hours, this sort of thing.
19 I am not arguing at the moment for any particular 20 measure of large release. I am interested in hearing 21 rationales for one or another definition.
22 So therefore, why yours and not Commissioner 23 Bernthal's?
24 DR. SIESS: Gary, I had rather refer to 25 Commissioner Asselstine for the moment defining a large ACE-FEDERAL REPORTERS, INC.
Nationwide Coserage 80 4 336 202 _.,_3700 347
.,______________.___._______.__-6646, ,__ _ _ - -___
2290 03 05 29
{3 o /DAVbur 1 release.
2 DR. BURDICK: I will address this, Dave, this 3 question.
4 DR. SIESS: Exceeding 5 rem at the site 5 boundary.
6 Now, what you have got is no early fatality at 7 one mile from the site boundary.
8 Is there any way you can translate this into rem 9 either one mile out or at the site boundary?
10 DR. BURDICK: I guess we have moved ahead to the 11 next slide.
12 DR. MARK: Before you move ahead to the next or O 13 perhaps overlying it, when you talk about early fatalities, 14 is there or is there not a threshold -- I mean, there is no 15 early fatalities at less than 100 rem at all?
16 DR. BURDICK: At 100 rem that appears to be well 17 below the cutoff point.
18 DR. MARK: So in estimating early fatalities --
19 DR. BURDICK: That is red bone marrow dose, I 20 believe.
21 DR. MARK: You don't use the extrapolation down 22 to 10 to the minus 18th?
23 MR. MARTIN: Absolutely not. We use 24 approximately 200 rem applied to bone marrow.
25 DR. MARK: This is 200 rem, no early fatalities, ACE-FEDERAL REPORTERS, INC.
202 Nationwide Coverage 800 336-6646
- - . -- _ _ - - - ,-347 3700. ._ ,_. .,. _ . . _ _ , _ _ __ . . _ _ , _ _ _ - _ _ _ _ _
- 2290 03 06 30 D.
h l AVbur D 1 or something like that?
2 MR. MARTIN: We considered three insults. You 3 can't look at just one organ dose. We look at long 4 exposure, GI tract, red bone marrow, but, yes, effectively 5 in round numbers for any of these things, less than 200 rem 6 to any organ, no, you won' t see an early fatality.
7 DR. MARK: I just wanted to make sure you didn't 8 do this lousy integral.
9 DR. BURDICK: All right. I am building up some 10 IOU's on questions here.
11 (Slide.)
- 12 I think I can address them very shortly here.
13 Let me continue with the rationale.
14 If we, instead of taking that lower limit as 8 15 times 10 to the minus 6, let's add a little defense in-depth 1
16 and then for CPDO purposes let's take R equal to 10 to the 17 minus 6th early fatalities per reactor-year.
18 DR. SIESS: Is that one mile from the site 19 boundary?
20 DR. BURDICK: Yes.
21 Then we can define a large release as any release 22 of radionuclides to the environment which would likely cause 23 one or more early fatalities, and the large release thus 24 defined satisfies the prompt fatality safety goal. We have 25 dropped it by a factor of 8, and it complies with the safety i
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336 6646
_ . __. . , _ __. - - _ _ _ _ _ _ . _ _ _ - _ _ __.. , _ _ _ _ _ _ - - . . ~ _ _ _ . _ . _ _ _ . _ .
2290 03 07 31
(
b~3jDAVbur 1 goal implementation guideline on frequency of large release, 2 which says that a large release should not occur any more 3 often than one in a million reactor-years.
4 DR. SIESS: Well, that is true, except that the 5 safety guide, the safety goal policy statement itself did 6 not define a large release.
7 DR. BURDICK: That is why I have to for the 8 purpose of implementing it.
9 DR. SIESS: You still have not defined the large 10 release?
11 DR. BURDICK: I am offering a definition here.
12 DR. SIESS: Can you relate that definition to the 7-)3
\-
- 13 definition that was proposed by either Commissioner 14 Asselstine or Commissioner Bernthal?
15 Commissioner Asselstine proposed 5 rem at the
- 16 plant boundary and one in a million. Commissioner Bernthal 17 proposed 5 rem at the plant boundary with 90 percent 18 confidence limits at 10 to the minus 3 for all plants, and 19 10 to the minus 3 for 100 plants would be 10 to the minus 5, i
20 and I don't know what 90 percent confidence limit would do 21 with probably a factor of 10, and then we have something 22 comparable.
23 DR. BURDICK: I would suspect you would have 24 something comparable, yes.
25 DR. SIESS: It would be 10 to the minus 6th at ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800 3366M6
_ , _ . . _ _ _202
_ 347 3 700 _ - _ _ _ _ _ . - - . - - _ _ _ _ _ _ _ - .
2290 03 08 32 DAVbur 1 the plant boundary for 5 rem.
2 Now, can you relate that at all to this?
3 DR. BURDICK: I think, as we both just agreed, 4 the frequency considerations here are comparable. The 5 5 rem, however, appears to be significantly below what we are 6 talking about here. I am trying to distinguish between 7 venial sin and mortal sin.
8 DR. SIESS: You are talking about 200 rem at a 9 mile out versus their 5 rem at the boundary?
10 DR. BURDICK: I am talking about implementing 11 this definition in terms of a CCDF here per individual 12 plant. We want the lower limit to be 10 to the minus 6 per O,
13 reactor-year for having one 'cn: more early fatalities.
14 Anything higher than that correspondingly, of course, the 15 frequency should be much less.
16 Now, what that translates into man-rem or 17 person-remwise to individuals at a specific plant is going 18 to depend on various factors, one of which would be the 19 population in the vicinity of that particular plant.
20 DR. MARK: Would it or would it not?
21 As we just heard, there is no early fatalities 22 unless there is 200 rem at the point you choose, a mile from 23 the boundary.
24 DR. OKRENT: This would permit 200 rem, in fact.
r-)
(./
25 DR. SIESS: It is a uniform population.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6646
2290 03 09 33 (3,
M DAVbur 1 DR. MARK: No, that is not the point.
,2 DR. BURDICK: That is the cutoff.
- *3 DR. SIESS: That is population density.
4 DR. BURDICK: Jim Martin, what is the 1 percent 5 chance?
6 MR. MARTIN: Of the early fatality?
7 DR. BURDICK: Within 60 days, yes.
8 MR. MARTIN: Do you want it exactly? I can look
, 9 it up.
10 DR. BURDICK: Roughly.
i 11 MR. MARTIN: Roughly, a 1 percent chance, like 1
~ 12 220 rem red bone marrow. I will take a risk of providing an b 13 answer to questions about the 5 rem and the 25 rem.
14 Gary mentioned the difference between venial sin 15 and mortal sin. If I had 25 rems vis-a-vis the safety 16 goals, they are below the early fatality threshold. That 17 is in the dose world rather than the early fatality world.
18 At 5 to 25 rems you cannot have an early fatality. That is 19 holier than the church, to use an old Irish phrase.
20 If that is the way they want to do it, fine.
f 21 That is their prerogative. But they have got to realize l
22 that they are much more conservative than the safety goal l 23 because that is zero.
24 DR. OKRENT: This is all within some cloud of
}
25 uncertainty. So even if your mean, or whatever it is, were l
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6M6
2290 03 10 34
/~T L.jDAVbur 1 that you would get 5 rem at the site boundary, there would 2 be a distribution that would get you up to 500 rem at the 3 site boundary with some lowered probability.
4 So in no case can you eliminate higher doses.
5 MR. MARTIN: Yes. A lot of people want zero 6 risk. They simply can' t have it.
7 DR. OKRENT: But this is not a zero risk 8 statement.
9 DR. BURDICK: It is not a zero risk statement.
10 We are defining here --
11 DR. OKRENT: Yours nor theirs is my point. What 12 I am trying to understand is why the difference. You are 7-s '
R-) proposing something -- I will speculate -- 50 times larger, 13 14 roughly, than what Commissioner Asselstine proposed as a 15 measure of a large release, ballpark.
16 I mean, after all, in the end this is a question 17 of judgment. But I would like to understand the rationale
~
18 for the staff's recommending this rather than what the 19 Commissioners have recommended. That is all.
- 20 DR. BURDICK
- Fine. The rationale is that if you 21 look at the prompt fatality safety goal, if the plant meets
- 22 that, it is our experience that the latent cancer risk is 23 probably satisfied, also, and following the same argument, I 24 got to a definition of a large release which was below the 25 statement of the safety goal. Granted, there are l
1 i ACE-FEDERAL REPORTERS, INC.
! 202-347 3700 Nationwide Cove se 80 4 336-6646
2290 03 11 35 1 -DAVbur 1 uncertainties in the yardstick by which you measure plant 2 safety. How this is implemented, we will have to, I am 3 sure, take into account those uncertainties.
4 But as a goal here now, we are conforming to the 5 Commission's safety goal. It happens to be a statement of 6 release which is larger than the two Commissioners in 7 question have voiced.
8 Do you want the Commission's safety goal to be 9 smaller? Is that what we are saying here?
10 DR. OKRENT: I am trying to understand your 11 rationale.
12 DR. BURDICK: My rationale is we have got
' O 13 something identified, a release which satisfies the safety 14 goal, and we are still saying we do not want that release to 15 happen with 10 to the minus 6th or less frequency. It has 16 got to be at a likelihood, a frequency of -- again from the 17 safety goal implementation guidance, it has got to be less 18 than 10 to the minus 6 per reactor-year.
l 19 DR. OKRENT: Let me say it is my impression that 20 the safety goal, while it is some numerical high level risk 21 guidelines, like 5 times 10 to the minus 7 early fatalities, j 22 it doesn' t tell us how the Commissioners want uncertainties 23 managed in applying that goal.
l 24 DR. BURDICK: That is right. There is a lot of 25 implementation work to be done.
ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800-336-6646
2290 03 12 36 (3
t -DAVbur 1 DR. OKRENT: And depending on how one factors in 2 these uncertainties, one can move in one direction o.r 3 another, and I would say the staff in this proposal is 4 pressing up against the safety goal sort of from one 5 direction, whereas Commissioners Bernthal and Asselstine 6 were suggesting perhaps, for whatever reason, coming, if you 7 will, from the other side with regard to uncertainties.
8 Let me add one more thought, a think, if I recall 9 correctly, not explicitly and perhaps not implicitly pulled 10 out of the safety goal related to longtime loss of natural 11 resources.
12 In other words, suppose in fact you managed to O 13 have a release where offsite nobody got more than 150 R but 14 there was a substantial area.
15 Do you want me to invent a number? How about one 16 18 miles in diameter or in radius as an example?
17 It is not clear, I would say, whether the safety 18 goal is complete without giving thought to what I will call 19 loss of natural resources.
20 There are some countries which I understand in 21 fact have some very stringent proposed or adopted high level 22 goals of that sort, and I think it is clear that there are 23 many countries to which -- whether explicit or implicit --
24 that that is a very important aspect of a safety goal policy 25 and shows in some of the implementation requirements that ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6646
- _ _ - ~ , - .
2290 03 13 37
(~T R DAVbur 1 you use for mitigation, and so forth.
2 So I want to note that although it is not called 3 out in the Commissioners' safety goal policy, I think they 4 did a monumental job in getting five goals for something 5 under the pressures they did, and this issue shouldn't be 6 left out by the staff. It is clear it is important.
7 8
9 10 11 12 Lq ,
13 14 15 16 17 18 l 19 l
20 21 22 23 O 25
. ACE-FEDERAL REPORTERS, INC.
M-347-3700 Nationwide Coverage 800-336-6646
2290 04 01 38 l'
i-)/DAV/bc 1 DR. BURDICK: I'll write that down and we will 2 certainly consider it. Jim...
3 MR. MARTIN: Just two responses. In the long 4 deliberations on the safety goal, there were two things 5 considered but never adopted. One was the core melt 6 frequency number. And the other one, there was 7 consideration in some parts for inclusion in an economic 8 type of safety goal. .
9 But they didn' t do it. So we could only work 10 with what they give us.
11 DR. OKRENT: There's a dif ference between .what 12 I'll call an economic safety goal -- it has a certain
() '
13 different ring -- and the loss of societal resources.
14 There's a relationship, but I don't think you should equate 15 them. There are other effects besides economic.
16 MR. EBERSOLE: Really, I guess I'm asking Dave 17 the question. I recall a case in point where it went beyond 18 natural resources and including human-built resources. In 19 particular, SK 25, which had an importance in several 20 respects -- National Defense, and so forth.
21 So I don' t know whether we should just stick to 22 natural resources, thinking beyond economics.
23 DR. BURDICK: I will admit that there are other 24 considerations.
25 DR. MARK: I must say, I rather like the main ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6666
2290 04 02 39 O aspects of what's reflected here. There's the matter of 1-/DAV/bc 1 2 definition that has to be built in -- 10 to the minus 6 3 early fatalities per year. Maybe that's a good number.
4 Maybe it should be some other number. It doesn't matter.
5 When do we decree an early fatality? That has to 6 be spelled out. And what allowance do we make for 7 variability of the event? It has to be spelled out.
8 I think that what you're reflecting there, or 9 what I read from what's there, sounds good to me, with open 10 questions as a way of approaching things.
11 Chet...
12 DR. SIESS: Gary, your last statement says that 13 this satisfies the prompt f atality safety goal by about an 14 order of magnitude. And that's based on the fact that, on 15 the previous slide, you've got 8 times 10 to the minus 6th.
16 Now, the way you got 8 times 10 to the minus 6 was to divide 17 5 times 10 to the minus 7 by 16.
18 DR. BURDICK: Multiply it.
19 DR. SIESS: Multiply it by 16. I'm sorry.
20 Suppose I had multiplied it by 32? Or by 160? or by 2? It 21 seems to me that this statement that you have here is 22 dependent on what you define as a sector.
23 DR. BURDICK: I am saying implicitly here that in 24 a release, about 1/16th of the annulus is probably going to 25 be affected.
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336 6M6
2290'04 03 40
(')
1 /DAV/bc 1 Jim, do you have a comment on that?
2 DR. SIESS: That's related to the plume?
3 MR. MARTIN: You can go down as low as 1/20th.
4 The plume takes -- on an average day, a plume takes up a 5 substantial portion of the 22-1/2 degree circle, your 6 cardinal compass.
7 Actually, it takes up about 1/20th of a circle.
8 DR. SIESS: Okay.
9 MR. MARTIN: So you can' t go lower. You 10 shouldn' t go lower than that.
11 DR. SIESS: Now, if I tried to apply this design 12 objective to a plpnt that's a standard plant that's being 13 designed for a site that's not been selected, this assumes a 14 uniform distribution of popula' tion around the site.
15 DR. BURDICK: That's right.
16 DR. SIESS: And it doesn't make any difference 17 how big the population is as long a there is one person in 18 each of those 16 sectors. You have to keep the dose down to 19 meet this. Right?
20 What happens if I come up to a site that has all 21 the population in one quadrant? Does that make a difference 22 now?
2 23 DR. BURDICK: We've gone now from this rationale 24 to the CCDF interpretation.
I 25 DR. SIESS: Why is anything on that CCDF ACE-FEDERAL REPORTERS, INC.
3)2-347-3700 Nationwide Coverage 80 4 336-6646
2290 04 04 41 g'DAV/bc ks 1 important except for the value of one?
2 DR. BURDICK: I'm sorry. Would you repeat the 3 question?
4 DR. SIESS: As long as that curve decreases as 5 you move away from 1 on the horizontal axis, does it make 6 any difference what the shape of it is? Your criterion is 7 at one mile.
8 MR. MARTIN: At one person.
9 DR. SIESS: One or more persons.
10 DR. BURDICK: I'm sure the shape.is going to be 11 monotonically decreasing.
- 12 DR. SIESS: Does it make any difference what the
\~J -
13 shape is?
14 DR. BURDICK: You certainly could not --
15 DR. SIESS: As long as it decreases.
16 MR. MARTIN: The shape doesn't matter.
17 MR. TAYLOR: Excuse me. The shape does not 18 matter, but there would be a different shape in terms of the 19 person rems. In other words, if one was to look at the cost 20 of improvements -- more people, more rem.
21 DR. SIESS: Once this safety goal is satisfied, 22 then you need to look at something else, you need to do 23 this. But this is just dependent on the meteorology out to 24 one mile beyond the plant boundary.
25 MR. TAYLOR: The total ensemble of severe ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M646
2290 04 05 -
42 1 'DAV/bc 1 accidents.
2 MR. EBERSOLE: Let ne ask a question. The recent 3 event has now put an international perspective on this whole 4 business. And if it's an order, and appropriate for you to 5 say anything -- it may or may not be -- if I picked it, if I 6 picked the case of Great Britain in its consideration of, 7 say, the Westinghouse PWR, or Germany or France, and I'll 8 stop there, do they have similar constraints to these?
9 Have you looked at these? In short, earlier on, 10 you said you wanted to be practical about it so that there 11 is an industry. And that's an interesting thing, since 7- 12 there,is none at the moment.
\~)3 13 DR. BURDICK: Let me protect myself by saying 14 that's not what I said. I said I wanted it to be 15 achievable.
16 MR. EBERSOLE: Something like that. It's related 17 to that.
18 DR. BURDICK: A four-minute mile for the industry 19 versus a two-minute mile, if you will.
20 MR. EBERSOLE: I just wondered, is there a view 21 in the larger sense about what the countries do with packed 22 populations, such as Japan and Germany?
23 In general, they have lots more cooling systems.
24 They pad up the two-track system. And that is a case in 25 point. You know what I mean.
l I
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
2290 04 06 43 l JDAV/bc' 1 Are they aiming for goals higher than, or do they 2 have lower ones than these?
3 DR. BURDICK: Matt, you had your hand up. Do you 4 want to respond to that?
5 MR. TAYLOR: I can only report what Jim Sniezek 6 found in discussing safety goals with some of the 7 Europeans. There's no doubt about it that they do place a 8 high degree of emphasis on redundancy, N plus or N minus 2 9 design of their systems.
10 But my understanding is that, generally, they are 11 taking a CCDF approach, in gauging the overall risk 12 somewhere above the CCDF, so that you could draw a straight 13 line on that particular curve that you see on the screen.
14 So that's one approach that I've seen discussed.
15 At this point in time, I'm unaware that there are health and 16 safety goals in any of the European countries; they're still 17 debating this.
18 DR. OKRENT: I wonder if you would mind going 19 back to your definition of "large release". I am trying to 20 understand the logic. As I read it through, R is the large 21 release, lower limit risk for individuals in the downwind 22 sector.
23 And it says that this R may be as large as 8 r~s 24 times 10 to the minus 6 per reactor year.
d The safety goal says, I believe, that a large 25 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
2290 04 07 44
(';DAV/bc is/ 1 release itself should be no more frequent than 10 to the 2 minus 6 per reactor year.
3 Therefore, I would assume, if I were consistent 4 with that part of the safety goal and I assumed in fact this 5 factor of 16, that the largest release, lower limit risk to 6 individuals would be 1/16th of 10 to the minus 6.
7 DR. BURDICK: No. The safety goal speaks in 8 terms of risk to the average individual in the vicinity of 9 the plant. Now, if you have people in the downwind sector 10 as the only ones that are at risk, then the risk to the 11 average individual has to be 1/16th of the risk to people in 12 that sector. It's that simple.
7_
13 DR. SIESS: He's deriving this from a health 14 effects goal, not from that other 10 to the minus 6 15 release.
16 DR. OKRENT: I'm sorry. I realize I'm taking the 17 10 to the minus 6 risk of a large releace, and if we assume l 18 arbitrarily, which may not be correct, that such a release, 19 if it were to occur, would occur as pieces of pie, 1/16th of 20 the total, then given a large release, the chance of early 21 fatality should be 10 to the minus 6 divided by 16 if this 22 large release could produce 300, 400 or 500 more R at the
! 23 site population boundary; in which case, one would arrive at 3 24 some number, something times 10 to the minus 8, I believe, l ~)
25 for the individual risk that you derive backwards from the ACE-FEDERAL REPORTERS, INC.
3)2 347-3700 Nationwide Coverage 800-336-6646
2290 04 08 45 k /DAV/bc 1 large release -- assuming it was that large. -
2 If you took the Asselstine, Bernthal large 3 release of only 5 R at the site boundary, it would be some 4 less frequent release that led to higher doses and to still 5 lower risks.
6 So, unless I'm missing something, you will not 7 get a 1 to 1 compatability between the 5 times 7 10 to the 8 minus 6, which is in the high level safety goal for early 9 fatalities, and the 10 to the minus per reactor year for a 10 large release.
11 And depending which way you were to get a rather 12 different number, and you have chosen, it seems to me, not 3
13 to start with the 10 to the minus 6 per reactor year to 14 estimate the prompt fatality that might result -- you've 15 gone the other way.
16 DR. BURDICK: I started'with the 10 to the minus 17 6th. If you want to write this as .5 times 10 to the minus 18 6, then we're talking about a 50/50 chance of an early 19 fatality, given a large release at a frequency of 10 to the 20 minus 6.
21 DR. SIESS: You started with the health 22 measures. You did not start with 10 to the minus 6. Right?
23 That's what that slide says.
24 DR. BURDICK: But the 10 to the minus 6th is in 25 there. I don' t see --
ACE-FEDERAL REPORTERS, INC.
3)2-347-3700 Nationwide Coverage 800-336-6646
2290 04 09 46
(~')
1 /DAV/bc 1 DR. SIESS: There are two separate criteria.
2 DR. BURDICK: I could have started right with the 3 10 to the minus 6 and said the mission of the agency is to 4 protect the health and safety of the public, and we'll 5 define right away a large release up front as one which has 6 the chance of causing an early fatality at that frequency.
7 DR. SIESS: But you didn't do that.
8 DR. BURDICK: I thought it would be a little more 9 illustrative to tie it into the safety goal because its 10 conformance with the safety goal is something we're going to 11 have to address anyway.
12 DR. SIESS: As I understand it from looking at O 13 the two slides now, if you limit the large release as you've 14 defined it, the 10 to the minus 6, you are a f actor of 8 15 below the safety goal.
16 DR. BURDICK: Roughly speaking, yes. We are a 17 factor of 8. But, of course, getting back to the 18 uncertainties of how an individual plant measures up, that's 19 the question.
20 MR. MARTIN: Maybe I can help. I understand 21 Dr. Okrent's problem. I keep getting back to when I figure 22 out how to implement this, one could get around this factor 23 of 16 and get to Gary's next slide, the CCDF, and talk about 24 the conditional probability. Associate a large release with C 25 a conditional probability of the person who happens that ACE-FEDERAL REPORTERS, INC.
N2 347 3700 Nationwide Coverage 800 336-6646
2290 04 10 47 DAV/bc 1 happens to be right downwind, which is a factor of 16. It's 2 a new condition.
3 Since we're trying to figure out what it means, 4 we can say, well, that's what it means. And we get around 5 this problem of 16 and actual risk versus relative risk of 6 the poor person who would happen to be downwind from this 7 large release. But we'd still be going to the early 8 fatality'. So we'd still be at a much higher dose.
9 DR. SIESS: That's a significant difference. The 10 Commissioners, as opposed to the Commission, Asselstine and 11 Bernthal proposed 10 to the minus 6, was not based on early 12 fatalities. It was based, if you look behind it actually at 13 some of the words, it was based on obviating the need to 14 evacuate.
15 In fact, that's exactly what Asselstine said -- 5 16 rem is the emergency planning guideline from FEMA.
17 Bernthal took the 5 rem as part of the definition 18 of an accident that requires insurance payments. I forget 19 what it's called -- NRC's finding of a major accident.
20 MR. MARTIN: The extraordinary nuclear 21 occurrence.
l 22 DR. SIESS: So they would base their 10 to the 23 minus 6 large release, and defined "large release" as either 24 not requiring evacuation or not being an extraordinary 25 nuclear occurrence.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 04 11 48 rh b DAV/bc 1 You've tried to come up with a large release 2 definition in terms of fatalities, which is part of the 3 safety goal. And this is the difference.
4 Now, you are satisfying what the Commission told 5 you by relating your definition of "large release" to 6 something else, the Commission said.
7 DR. BURDICK: To something the Commission 8 approved, yes.
9 DR. OKRENT: I'm still interested in pursuing 10 this a little, so I understand what is being said. I don' t 11 really understand why. But if I think back to Chernobyl, 12 from what I read, there were no doses off site which were P) 13 large enough to fall into the significant likelihood of 14 early fatality. That's what I read..
15 Was that a large release, or was the meteorology 16 playing a role to keep it from producing doses that might 17 have some probability of early fatality?
18 Do you have any opinion? ,
19 20 21 22 23 24 25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
2290 05 01 49 (D I have spent l--DAVbur 1 MR. TAYLOR: I have an opinion.
2 quite a long time in discussing safety goals.
3 I believe we did debate the issue of repeated 4 release. I believe we now have some data on which the 5 repeated release will, if you will, jump over in place 6 perhaps of a dose further out.
7 DR. SIESS: Does that mean 1t might be better to 8 blow the top off containment than to just have a nice ~ sized 9 leak?
10 MR. TAYLOR: Yes. This was a prime release.
11 DR. OKRENT: Let me continue, if I may, and try 12 to pursue something. I had the impression. In fact, I 13 recall a statement by one or more members of the staff that 14 you could have an SST-1 at Indian Point and it would predict 15 less than one early fatality.
16 MR. TAYLOR: I am not aware of that. I do 17 believe that in the discussions of what constitutes a large 18 release it didn' t call for intervention, and if you did look 19 at an SST-1 it was around 10 to the minus 5.
20 Well, most sites in the country you could show 21 meeting the .1 standard.
22 DR. MARK: You say .1 percent?
23 MR. TAYLOR: .1 percent, which is the 24 safety goal policy.
25 MR. MARTIN: That goes back to the fact that ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
2290 05 02 50
^
()
1 /DAVbur 1 Tony was talking about. If he wants to specify the release 2 in terms of curies, like Chernobyl, a release that could 3 cause an early fatality but would not if it were lofted high 4 enough up in the air, it is lots of curies.
5 But you have really got to remember you would 6 meet the early fatality goal because the chance of an 7 arbitrarily located individual of even intercepting the 8 plume of any size is only 5 percent. So you go from 10 to 9 the minus 5 down to 5 times 10 to the minus 6th.
10 DR. SIESS: You said at a mile. That is the 11 boundary, isn't it?
12 MR. MARTIN: The safety goal says calculgte the O 13 risk for zero to one mile, and if there is nobody within a 14 mile then put a person at one mile.
15 DR. OKRENT: Well, I guess I am left with the 16 feeling that we don' t know why the staff seems to have 17 proposed that a large release be defined as being.a 18 release, what I will call, maybe mistakenly, substantially
! 19 larger than the release that Commissioners Bernthal and 20 Asselstine seem to be talking about.
l 21 MR. TAYLOR: I am not sure I can close that 22 question at this point in time. I think we do have to look
, 23 at this Commission directive. But it seems as though 1
i 24 Commissioner Bernthal and Commissioner Asselstine are 25 approaching it from one end of the spectrum and we are ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Covers:ge 800-336-6646
. , - ~
2290 05 03 51
(~)DAVbur b_/ 1 approaching it from the total spectrum, where we have got 2 both intermediate releases and intermediate frequencies 3 built into thi s approach as well as the largest release of 4 the type that the ACRS mentioned in their recent letter, a 5 very large release. I believe the ACRS themselves specified 6 a very large release.
7 Now, I am not sure exactly what you had in mind, 8 but I think we want the spectrum of releases reflected 9 here.
10 DR. OKRENT: It is my impression that the three 11 Commissioners were silent on the -definition of a large 12 release, and you implied that is their personal
.(:) 13 interpretation, am I correct.? -
14 MR. TAYLOR: There were two Commissioners that i 15 provided that personal interpretation.
16 As I said, I interpret this as coming from one 17 end of the spectrum instead of the whole spectrum.
18 DR. SIESS: I don' t think you need to talk about 19 spectrum. What the staff has done is try to define large 20 release in terms of an open-end safety goal, a quantitative 21 health and safety goal. I think both Commissioners were 22 trying to define large release in terms of some public or 23 lay understanding of what a large release might be, a 24 release that would require evacuation, a release that would
- 25 qualify as an extraordinary nuclear occurrence. They were ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage M336-6646
2290 05 04 52
('/
L_ DAVbur 1 looking for some other definition, which makes a certain 2 amount of sense.
3 To me, I think Bernthal in effect said in 4 connection with one of his criteria, not necessarily that 5 one, that this would be easily understood by a newspaper 6 editor, assuming the newspaper editor reads the newspaper, 7 which I doubt.
8 MR. TAYLOR: The problem, Dr. Siess, is that I 9 believe that Commissioner Bernthal in his comments on the 10 policy position suggested that perhaps a strategy of talking 11 to a standard of background would be a more appropriate 12 criterion.
) 13 DR. SIESS: But what comes out instead of these 14 numbers is that there is a large difference in the amount of 15 curies released and the amount of rem between a one in a 16 million and what would be required to meet the quantitative 17 fatality rate. Which one is right I don' t know.
18 MR. TAYLOR: The safety goal is stated in terms 19 of rem.
20 DR. OKRENT: If I can say one more thing -- and 21 then I will drop this point -- if I follow the logic on this.
22 graph, I arrive, it seems to me, at a permissible range for 23 larger leaks, which is 8 times 10 to the minus 6 per 24 reactor-year.
25 Is that correct?
ACE-FEDERAL REPORTERS, INC.
E-347 3700 Nationwide Coverage 800-336-6646
2290 05 05 53
("j\
L_ DAVbur 1 DR. BURDICK: That is not correct. This is not a 2 speed limit. This does not define something that is 3 acceptable. Again, this is a goal.
4 DR. OKRENT: Nevertheless, taking the staff's 5 logic, I would arrive at the result that R can be equal or 6 less than 8 times 10 to the minus 6, and I say this is 7 incompatible with that part of the safety goal policy which 8 says it should be less than 10 to the minus 6.
9 So I think there is a logical incompatibility.
'10 DR. SIESS: I have a slightly different problem.
11 If I take your approach, starting with the health effects,
_ 12 , coming down to the definition of large release and the O 13 median level until I get down to where you put another ,
14 factor of 8 on, I don,' t have a containment performance 15 objective. I simply have an overall safety goal objective.
16 DR. MARK: I think that is great.
17 DR. SIESS: I am simply quantifying the safety 18 goal objective in terms of the probability of a release.
19 Now, if you put another factor of 8 on it and
- 20 say, well, that is defense in-depth -- I think that is what 21 you said and I have used the same expression -- have we got 22 a containment design objective or have we now got just 23 simply another way of expressing the quantitative health 24 effects?
25 DR. BURDICK: As soon as the Commission approved ACE-FEDERAL REPORTERS, INC.
202 347 Nationwide Coverage 80 4 336-6646
_ _ _ . -3700 -. __ - . _ . . _. , _ . _ _ _ _ _ . _ _ _ _ - _
2290 05 06 54
/"'t the safety goal statement, they made a statement about b_/DAVbur 1 2 plants. That included -- plants includes the containment.
3 We are trying to refine this statement a little 4 for implementation, further guidance perhaps to the 5 industry, and we are adding a little more defense in-depth 6 than there was there to the safety goal, and you say does it 7 deal with containments?
8 of course, it deals with containments. When you 9 are talking about limiting releases to the atmosphere, you 10 have to be talking about- containments.
11 DR. SIESS: What is the probability of it because 12 I could presumably design a plant that would release, but if 13 the probability was low enough, I wouldn't need a 14 containment?
15 DR. BURDICK: That brings me to the next part of 16 the statement. We have not finished with the statement of 17 the CPDO yet.
18 DR. SIESS: That is what I said. At this point l
l 19 you do not have a containment performance design objective.
20 You only have a quantification of the health goals with an 21 extra factor for defense in-depth which you think will be 22 provided.
23 DR. BURDICK: You are absolutely right. At this 24 point our job isn't through.
25 DR. SIESS: Let's go on and tie this.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336 4 86
r .
J 2290 05 07 55 DAVbur 1 DR. MARK: Gary, would this be a good place to 2 give you a slight respite?
3 DR. SIESS: Not to mention us.
4 DR. BURDICK: I could use another glass of water, 5 I think.
6 DR. MARK: ' How about resuming at a quarter till 7 11:00?
8 (Recess.)
9 DR. MARK: The meeting will continue.
10 Gary.
11 DR. BURDICK: Thank you, Mr. Chairman.
, 12 (Slide.)
i O' 13 Now that.we have disposed of the definition of 14 large release, there were some other considerations.
15 DR. MARK: Look, there is a real need for 16 convergence on what is meant by large release. I think you 17 are approaching it in a way which gives promise of being 4
18 able to write something specific. Whether it should be 19 this, clearly it is very different from what Commissioner 20 Asselstine thinks, and I believe it should be different from 21 what he says.
4 22 DR. BURDICK: Well, I was offering a little 23 levity there. I certainly do agree with you that there is 24 some further look at the situation. This is just what we 25 are laying on the table at this time.
l-ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
2290 05 08 56 DAVbur 1 Some of these other considerations, however, have 2 to do with venting. As I said, we did feel in the workshop I
3 that this should somehow be addressed. There are other 4 countries which have implemented filtered venting systems, 5 at least one that I don't know of.
6 This venting should prevent high pressure 7 buildup. The vents ideally should be closed with high 8 reliability before core damage occurs. I think someone 9 brought this point up.
10 We do have to have reliable venting, open and 11 closure.
7,3 12 MR. EBERSOLE: I have a little problem with that iuj !
13 statement because i,t is almost implicit in it that if you 14 vent then comes core damage. That is not true. As a matter 15 of fact, you will be venting to stop core damage.
16 DR. BURDICK: Well, th'at is true. It is not 17 meant to imply that at all.
18 MR. EBERSOLE: But you see, it reads that way.
l 19 DR. BURDICK: If core damage is going to occur, l
l 20 it certainly should be closed before that.
l 21 MR. EBERSOLE: If it occurs after that for any l
22 cause whatsoever, then you might want to vent in a different l
23 way with filters.
l 24 DR. OKRENT: I really don't understand the second (v']
25 bullet. It seems to me it deals with a segment of the ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
4 2290 05 09 57
(")
1 /DAVbur 1 issue; namely, you are trying to prevent before core damage 2 occurs as a way of removing containment heat or something 3 like this. That is not the only situation.
4 DR. MARK: It might be, David, that you might 5 very well want to vent even after core damage in order to 6 control the amount of release from being gross.
7 DR. OKRENT: Such has been proposed and design 8 features have been included in some other countries, and I 9 believe such is being proposed by some of the current 10 licensees.
i 11 DR. MARK: That is, you vent to avoid super high 12 pressure and you vent to control the' amount of stuff that
( ,
13 goes out even after there is stuff there?
14 DR. OKRENT: There is some question about how 15 well you are controlling what comes out. That depends on 16 the conditions.
17 DR. BURDICK: Well, what the slide is intended to 18 illustrate is that there were a lot of other things we 19 thought about. Some of these maybe are not as accurate as 20 they should be. Necessarily, some of them are cryptic to 21 get them on the transparency here.
3 22 i ,
But there are a lot of issues involved when you 23 think about containment venting. I think there are a lot of 24 other questions that we haven't even addressed here.
25 In a venting situation, I think there would be ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 202-347-3700
_.. _ 800 336
_ _-6646. ._,_, . , --
2290 05 10 58
[D political, legal, and a lot of other things that you would lDAVbur 1 2 have to consider before you did make a decision to go ahead 3 and vent, of course, depending on what kind of system you 4 had.
5 There are a lot of open questions here.
6 MR. EBERSOLE: In that context, I think earlier 7 on I mentioned we do not now have any kind of a standing 8 principle or concept of how much radiation to release to 9 prevent a larger radiation release. That is a real 10 problem.
11 DR. BURDICK: That is a problem.
" 12 Jim, you had a comment?
t) Just so it doesn' t get lost, on that 13 MR. HARTIN:
14 second bullet I think a key phrase in that thing is 15 regardless of when you would be beginning or how much you 16 would tolerate in a release, you do want to be able to close 17 that valve.
j 18 MR. EBERSOLE: And reopen if you want.
l 19 MR. MARTIN: The reliability is, by George, you 20 want to be able to close it once you open it.
21 DR. OKRENT: And in fact there are some issues 22 concerning what happens to the venting channels downstream l
l 23 of the valves. So that is an incomplete slide in many I
I
{} 24 25 ways.
Let me give an opinion that this slide calls to ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6M6
2'290 05 11 59 l'iDAVbur 1 mind I more than once have read in what the staff has 2 written. One thing is, following some words, the conclusion 3 is drawn, prevention is more important than mitigation.
4 That is one kind of a statement you see.
5 Another kind of a statement that I see is, now it 6 is about time that we started looking at prevention; 7 mitigation has gotten the bulk of the attention in the 8 past.
9 If you insist, I will find the staff documents 10 that have those statements.
11 I have never understood the particular statement
, 12 that in the past mitigation has gotten more attention than 13 prevention. It has been my impression that since 10 CFR 14 Part 100, which is 1962, except for a little bit on 15 hydrogen and an odd and end here and there, the staff has I
16 hardly looked at mitigation.
17 So as I say, I am just giving you a reaction to 18 some of the stuff. In fact, it is starting to -- if I 19 understand correctly, almost for the first time -- to look 20 at it vis-a-vis the Mark I's, so that at least it seems to 21 me one should be careful about some of the generalizations 22 which may really not have anything to do with the history.
23 DR. BURDICK: I agree with you. The distinction between prevention and mitigation for some boiler sequences
{} 24 25 is certainly fuzzy. I can' t extend a lot of the statements ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800-336 6646
_ . __ _202
,-347-3700
~
2290 05 12 60 O
1"DAVbur 1 of other members of the staff. I am not sure who you are 2 referring to. -
3 4
5 6
7 8
9 10 11 12
~
13 14 15 16 i
'l 17 18 19 i 20 i
21 22 23 4
lO 25 l
! ACE-FEDERAL REPORTERS, INC.
_ . . , , _ _ _ . , - -~_.__ .202 347_.3700_ _, _ _ __ _ , Nationwide Coverage _
, _ _ _ _8043Q _
2290 06 01 61
(~'t lDAV/bc 1 MR. EBERSOLE: Would you call precore-venting for 2 a boiler prevention, or mitigation?
3 DR. BURDICK: That's what I say. There are some 4 sequences where it's pretty fuzzy. Obviously, if you are in 5 one of those scenarios and are successful in the cooling and 6 venting and preventing of core melt --
7 MR. EBERSOLE: I would, in my own mind, call it 8 prevention.
9 DR. BURDICK: It's prevention.
10 DR. OKRENT: So would I, Jessie. It seems to me 11 mitigation we might uniformly define as something you do 12 after the core is damaged, even though you would say you 13 are mitigating a LOCA by applying water.
14 DR. MARK: David, would you buy the idea that 15 bullet number two should read that you have venting 16 available 'to avoid pressure at some level, damaging i 17 pressure? And you use it providing that the venting is l 18 filtered and providing you have the ability to close it off?
l 19 That is, you would always use venting to avoid l
20 rupture, no matter what the state of the core, providing you 21 could turn it off when you decide that the containment was 22 no longer threatened.
23 DR. OKRENT: Then you have a filter?
24 DR. MARK: Yes, a filter.
{
25 DR. OKRENT: Measured to the amount of ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6686
2290 06 02 62
(
. P)DAV/bc 1 radioactivy involved and, of course, look downstream to make 2 sure you are not venting a lot of hydrogen into a building 3 that's then going to blow up. And that it's a properly done 4 job.
5 DR. MARK: Independent of the state of the core, 6 ,you would vent if it were filtered and you could stop it 7 when you desired.
8 DR. OKRENT: That's a personal opinion. But, let 9 me put it this way. If I were either the NRC, the licensee 10 or the Governor and somebody said we've got a plant whose 11 pressure is building up and the prediction, whatever that 12 means, is that it's going to go up to 133 psi and plateau 13 ou t -- four years ago, when the PRA, Sargent and Lundy, 14 estimated 140 plus or minus 4, so there's really no 15 problem, I must say that I would hope that the plant had a 16 system that would let me relieve the pressure far below 130 17 in a way that nothing much beyond noble gases got out.
18 DR. MARK: Now, that statement as written doesn't 19 cover that line of thoufra. It seems to me it should.
20 DR. BUPUI 1: -ell, as was pointed out, the list 21 There should be qualifiers in isnotcompletebyangmeans.
4 22 s.ome areas, again. We try to point out here that we did 23 consider a lot of things when we were thinking about the 24 venting portion of the containment performance design 25 objective.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336-6646
.i 2290 06 03 63 4
& DAV/bc 1 Another consideration was further defense 2 indepth. This was pointed out by Chet Siess, that unless 3 you had some degree to which you limited the probability to 4 containment, you might have this situation where you could 5 meet the 10 to the minus 6 by a core melt frequency. 'And 6 then who would meet containment?
7 That's not an acceptable situation, so we had to
! 8 address that question also.
9 (Slide) 10 Now I lay before you where my staff and I are at 11 the current time. I emphasize this is where my staff and I 12 are at. Of course, we've been keeping NRR involved but 13 there's certainly no office concurrence or anything else in 14 this statement as it stands.
15 This is what we have under consideration right 16 now.
17 MR. EBERSOLE: Are you talking about new or old,
! 18 or what plants, containments?
19 DR. BURDICK: This is specified in item 3.
20 MR. EBERSOLE: I'm sorry. I didn' t read down 21 there.
22 DR. BURDICK: What is parceled out for new and i 23 old plants. The other two statements, however, apply to all 24 the plants. We have this three-part statement then that i
25 containments should survive all but highly unlikely severe ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336 6646
. , . . _ _ , _ . _ _ _ , ~ . . . _ . _ _ _ ___ . . _ _ . - . . - _ . _ _ . _ . . _
2290 06 04 64
(~
h-)'DAV/bc 1 accidents without occurrence of a large release.
2 The aggregate frequency of a large release of 3 radionuclides to the atmosphere for all severe core damage 4 accident sequences should not exceed one in a million 5 reactor years. We define here again what we mean by a large 6 release.
7 A large release is defined as release of 8 radionuclides to the environment that would likely result in 9 early fatalities or equivalently we can take large release i 10 to mean life-threatening release.
11 As far as controlled venting is concerned, where 12 we are at now is that we should limit such activities as 0 13 venting of radionuclude species to the atmosphere, to those 14 limits specified in 10 CFR 100.
15 DR. MARK: Now, 10 CFR 100 would be --
16 DR. BURDICK: 25 rem whole body and 300 thyroid.
17 DR. MARK: Am I out from the boundary?
18 DR. BURDICK: At the site boundary.
i 19 DR. MARK: 25 rem in two hours.
20 MR. EBERSOLE: We've still got a problem that is 21 implicit in that. That is that venting carries new release 22 with it. It says control venting of containment. And says 23 release should not exceed this.
24 Implicit in that is the thought that you're going (a-}
25 to only vent after you've had some core damage. And I don't
- ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646
2290 06 05 65 1 bAV/bc 1 think that's appropriate.
2 DR. MARK: No, no. You might vent just steam.
3 MR. EBERSOLE: But I'm still, in the absence of 4 any guidance, you know, about the mechanism.
5 DR. BURDICK: What I'm saying, whatever hardware 6 and operations for its use that you design and install for 7 venting, it should conform to 10 CFR, Part 100.
8 MR. EBERSOLE: Say I'm a BWR operator and I don' t 9 know what my valves would do if the pressure is high and I'm 10 today at the board, and I have a problem with off site and 11 on site power in the containment.
12 I'm going to release a little radiation and I'll 73 g
13 never get any suppression pool cooling. What am I going to 14 do? What would you do?
15 I'm just a BWR operator. Where is it?
16 DR. BURDICK: If I were an operator, I would
[
17 probably follow my operating procedures.
l 18 MR. EBERSOLE: You won't find anything.
19 DR. BURDICK: Systems which are designed and 20 installed for emergency venting would, I hope, have well-21 defined operating procedures.
22 MR. EBERSOLE: There are none. Emergency 23 venting. As a matter of fact, there used to be a joking l
24 remark at Browns Ferry:
25 " Fellows, you should have out in the barn some ACE-FEDERAL REPORTERS, INC.
2C2-347-3700 Nationwide Coverage M336 6646
2290 06 06 66 1 AV/bc 1 ' place a howitzer because NRC doesn't have the notion in mind 2 that you're going to let this big doughnut blow up. You 3 know, just explode under pressure."
4 It was a casual sort of remark about that you 5 need to get rid of the pressure if you don' t have any core 6 damage. Yet, there is not in place today, after 20 odd 7 years, something to guide the operator in a semi-regulatory 8 context.
9 DR. BURDICK: In the event that these objectives 10 were adopted and approved, I would think again that, in 11 order to have some reasonable assurance that you can meet 12 this, you would for sure have to have a set of operating Oa 13 procedures.
14 MR. EBERSOLE: You haven' t got them.
15 DR. MARK: But he's proposing, as I understand 16 it --
17 DR. BURDICK: This is a goal.
18 DR. MARK: -- a set of things which would call 19 for a procedure: When it gets above 130 psi, you vent.
20 MR. EBERSOLE: Wait. I'm not sure that my valves 21 will work at even 50 psi.
22 DR. MARK: You now ge- and inquire of that.
23 They're not there now.
24 DR. OKRENT: Your item number two may in the end 25 be okay, but I'm not sure you've given us the rationale for ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6 9 6
2290 06 07 67
(~)
lw/DAV/bc 1 why under controlled venting, you would permit 10 CFR, Part 2 100. I assume it's the old 10 CFR, Part 100, since we're 3 talking about changing it, why you would permit releases up 4 to that value.
5 What is the reason that that's an adequate 6 measure of the performance of combined filter and vent?
7 DR. BURDICK: After taking all of these 8 considerations, looking at them, we have no regulatory basis 9 for any other statement.
10 DR. OKRENT: Sorry. I don't know that you have 11 any regulatory basis for that statement, because 10 CFR, 12 Part 100 is not in terms of some accident in being, it's 13 part of the siting criteria. ,
14 MR. MARTIN: More important though, it's part of 15 the design criteria. These are postulated as part of design 16 basis accidents. This would come into a design basis 17 accident.
18 DR. OKRENT: Part 100, when it was first adopted,
! 19 in fact took what it thought was a very conservative amount i
20 of radioactivity into the containment.
21 MR. MARTIN: You're talking about the footnote.
l 22 DR. OKRENT: It neglected to think about whether 23 the containment would maintain its integrity at its leak 24 rate. And then went through a recipe for computing some off 25 site doses.
ACE-FEDERAL REPORTERS, INC.
202 -347 3700 Nationwide Coverage 800-336-6M6
~ . - . _ _ _ . _ _ _
e ,
4
'2290 06 08 68 0- I'm reluctant to assume that it provides any lDAV/bc 1 2 philosophic basis for establishing what is or is not the j 3 right dose limit that some venting systems should be
'4 designed for.
5 MR. TAYLOR: Excuse me, Gary. I think you're 6 entirely correct that Part 100 has been the licensing siting 7 basis for design basis accidents. It does not necessarily
- 8 form a framework for risk management strategies, as these i
9 are.
10 Now, I understand that what Gary has put out r
11 there is a trial balloon. I don't think we're in a position 12 right now to say what would be the governing dose criteria.
13 I do think one has to do a little more looking at the risk 14 and risk tradeoffs between a controlled leakage and an 15 uncontrolled leakage.
i
~
16 MR. EBERSOLE: It seems to me what's needed here-
! 17 is what I think is sometimes called a rule curve, like the 18 flooding of reservoirs in Tennessee. It deoends on where 19 they are and what the likelihood of later rains are.
I
! 20 So it's a matrix. It says I'll vent now
! 21 according to a certain set of conditions, and what might 22 happen if I don' t vent, including a release.
23 Do you follow me?
DR. BURDICK: I understand what you're saying.
{}
! 24
! 25 MR. EBERSOLE: There may be more than two i
4 ACE-FEDERAL REPORTERS, INC.
- - - . _ - . ._ 3n-e _ m c~. - . .w- --
2290 06 09 69 1 bAV/bc 1 dimensions. But it is a procedure prethought out, not 2 churned up by an operator in stark terror.
3 DR. MARK: We have a comment here.
4 MR. HOLMAN: My name is Jerry Holman, from NRR.
5 Speaking for only one of the three operating divisions, and 6 only from my branch, the Plant Systems Branch, I don' t think 7 we would agree with number two, Gary.
8 Furthermore, the emergency procedure guidelines 9 for boilers now incorporate venting without any criteria 10 with respect to doses. It's there to save the containment.
11 MR. EBERSOLE: How long has that been in place?
gg 12 MR. HOLMAN: ,About a year and a half.
V If I can offer a comment, those.were 13 DR. OKRENT:
14 put into, I suppose, some kind of acceptance state, 15 prematurely, ill-considered and deeply require further 16 study.
17 So I think.they provide less rationale than item 18 two. Maybe you would find yourself having trouble with item 19 two, but I must say I have trouble with what the staff put 20 through a year and a half or so ago.
21 MR. EBERSOLE: Could I add something to this?
22 It looks to me like that's a wild swipe at the 23 solution of a problem. The operator I doubt knows when 24 these valves work or don' t work.
25 They haven't been designed to work or not work on ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336
. - , _ _ . _ - . _ - .. -_ _.- -6646 - .
2290 06 10 70 (h
l' #DAV/bc 1 pressures and temperatures. These valves are supposed to be 2 closed.
3 Now we have the other case. We've got an 4 instruction to the operator, but we haven't defined his 5 equipment. So what good is it going to do him?
6 DR. MARK: At least, the operator cannot be asked 7 to gauge his action against 10 CFR 100. He can be asked to 8 gauge his action against psi or something that's 9 recordable. He doesn' t know where CFR 100 cuts in or cuts 10 out.
11 MR. EBERSOLE: He doesn't know the psi either.
12 DR. MERK: He knows the psi, or at least he has 13 the chance to know it.
14 MR. EBERSOLE: He doesn' t know the psi, which his 15 valves won't open.
16 DR. MARK: Look, I understand your question about 17 the valves. They've got to be looked at and wondered 18 about. But, the psi, he can, at least in principle, have a
! 19 feeling for.
I 20 MR. EBERSOLE: Yes, that's true. He knows. It's 21 not quite the howitzer case.
22 DR. MARK: So, when you say the regulation is l
l 23 designed, correctly or incorrectly, to safe containment from
(} 24 rupture --
25 DR. OKRENT: Without thought to the doses that ACE-FEDERAL REPORTERS, INC.
202 347-3700 _ Nationwide Coverage _
800 336 6646
2290 06 11 71 DAV/bc 1 might occur without really proper thought to should they be 2 filtered, are the existing filter systems adequate?
3 DR. MARK: I don't debate that.
4 DR. OKRENT: A variety of things under a. host of 5 questions.
6 DR. MARK: At least it points at something that 7 could be defined and acted on without covering the 8 implications that you mentioned, or that Jessie mentioned.
9 DR. OKRENT: It would have been better to not 10 have been adopted.
11 DR. MARK: It shouldn' t have been adopted without g 12 thought.
U 13 MR. EBERSOLE: In this fragmentary way, Limerick 14 has come out of the ground with presumably equipment that 15 will vent at some point. Who knows where they should do 16 it.
17 I believe Grand Gulf has, and the other new 18 family of these big Mach III's, but not the others.
19 DR. BURDICK: Well, I appreciate these comments 20 and I'm succeeding in getting some feedback. Remember, this 21 is preliminary and I appreciate comments from all quarters.
22 I want to make that certainly clear.
23 This is where the research staff is at this
{} 24 point. Yes?
Can I ask a question on number 25 DR. OKRENT:
ACE-FEDERAL REPORTERS, INC.
D3#~3# Nationwide Coverage 80433MM6
2290 06 12 72 I~')
lDAV/bc 1 three?
2 It resembles, I realize, the ACRS suggested that 3 the trial balloon, NUREG 0739 -- I'm not sure. I can't tell 4 whether it's compatible with the safety goal policy, as this 5 part of the policy, the Commission has quite a full policy.
6 It says, pretty strongly urges the staff to look 7 at this 10 to the minus 6 per year, which the staff sort of 8 is running with.
9 It's picking up the ball and running. It's a 10 little bit hard to tell how that .01 for existing plants, 11 for some existing plants will line up with your frequency of 12 large release.
13 DR. BURDICK: Where the .01 would?
14 DR. OKRENT: The .l. I'm sorry.
15 MR. TAYLOR: I think I agree with Dr. Okrent.
16 Clearly, what one envisions is setting up a conditional 17 probability of containment. You have a set of sequences and 18 the set of sequences are a unity problem.
19 I think, in this context, you have to consider 20 the core melt cause. So you're talking about one sequence 21 where core melt is the initiator and the challenge to 22 containment.
23 Obviously, where you have, for example, loss of 24 coolant accident that has loss of containment heat removal, p) 25 that indirectly may challenge core melt or cause core melt.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-33MM6
2290 06 13 73
/^
4 DAV/bc 'l Theose are different conditional probabilities.
2 So I do think you have to look at the kind of sequence.
3 4
5 6
7 i
8 9
10 11 1
0 13 ,,
14
, 15 16 17 18 19 20 21 22 23 O .
25 1
1 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800
... . . . . . . - _ _. . _ - - - . . - . . . . - _ , . - . . _ - - - _ . - ._ .- 33 6 . . . .
2290 07 01 ,
74 rm b-DAVbw 1 In terms of .1, that is consistent with what has 2 been done in the past. I guess whether or not you need a 3 conditional probability or you need an overall large release 4 frequency is still a matter to be decided.
5 DR. OKRENT: I think there's been certainly, with 6 ACRS, urging that there be some kind of containment 7 performance criteria, aside from high levels goals, as part 8 of defense in' depth.
9 MR. TAYLOR: That is correct. It may be a 10 conditional probability.
11 DR. OKRENT: There are dif ferent ways of doing 12 it.
~
13 I am just thinking aloud that there may be some 14 reactors with, I'll say, a core melt frequency over -- in 15 between 5 times 10 to the 4 and 5 times 10 to the 5, having 16 to put in external events. If I take 20 percent of these, 17 having some potential for large release or even more, .1 18 doesn't necessarily get me to 10 to the minus,6.
19 MR. TAYLOR: That is correct.
20 DR. OKRENT: And that may be all right. We may 21 say, well, these may be all old plants, or whatever, but it 22 is going to take some thought.
23 MR. EBERSOLE: I have a momentary mental blank as 24 to what the accident that releases steam and water into the 25 valves and what is the likelihood of a pipe break or a ACE-FEDERAL REPORTERS, INC.
202-347 370v Nationwide Coverage 800-336-6646
2290 07 02 75
('8 b JDAVbw 1 machinery break. You and I well know. The one I think of 2 is the 10-inch main steam line, which degrades the 3 equipment, and it cascades to a total meltout right into the 4 outer world.
5 Do you follow me?
6 MR. TAYLOR: I think so. I believe you are 7 referring to the kind of accident that is developed in 8 WASH-1400. It's bypassing containment.
I 9 MR. EBERSOLE: You see, they released it, but i 10 they did not permit that release to degrade it. This is a 11 cascading -- how shall I call it -- well, it catalyzes or it 12 causes equipment which would prevent a problem. It causes 13 that equipment to die.
14 MR. TAYLOR: I think I understand what you are 15 saying.
16 MR. EBERSOLE: Considering the real liability of 17 valves to function under all conditions.
18 MR. TAYLOR: I think it is an indication that the 19 Staff is working on that is directed toward the BWR design.
20 MR. EBERSOLE: That number is still being chewed 21 up.
22 MR. TAYLOR: I've seen numbers that have come 23 from contractors.
- 24 MR. EBERSOLE
- What is the error band on the 25 uncertainty, particularly about valves.
l ACE-FEDERAL REPORTERS, INC.
I 202 347-3700 Nationwide Coserage 804 336 4646
2290 07 03 76 (D
h /DAVbw 1 MR. TAYLOR: It is large. That is 10 to the 2 minus 7 and 10 to the minus 5 to an 1/8.
3 MR. EBERSOLE: I am talking about the valves to 4 steam-driven diesel systems.
5 MR. TAYLOR: I understand. That is being worked 6 on now in Variation No. 5, and I can't tell you exactly.
7 MR. EBERSOLE: We need a progress report on 8 that.
9 MR. TAYLOR: This is true.'
10 (Slide.)
11 DR. BURDICK: To achieve such a CPDO, the 12 containment design may have to contain some of the 13 following. We have discussed already some of these. You 14 have to have low probability of containment bypass, the 15 sequences for which containment fails before core melt. We 16 would have to have some provision for venting before core 17 melt and high reliability of the system for reclosing the 18 vent.
19 It might be possible to install high stacks for 20 the venting of noble gases, for Ps, automatic 21 depressurization of reactor vessel to eliminate the 22 possibility of direct heating of containment atmosphere.
23 That might be something to consider.
- 24 I would certainly want to consider highly 25 reliable containment cooling systems.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 33MM6
2290 07 04 77 DAVbw 1 DR. OKRENT: Is that a PWR and BWR Vugraph?
2 DR. BURDICK: The considerations. Some of them 3 apply to both.
4 MR. EBERSOLE: At that point, I think you should 5 put these two plants in separate boxes and treat them for 6 what they are and what they are not. That would include 7 consideration which is being carried on now to vent the 8 primary system for the PWR. Glenn Reed is pushing that.
9 DR. BURDICK: Let me understand what you are 10 saying.
11 Are you saying we should have a separate 12 containment performance design objective for each one, or f-)
x) 13 are you saying separate the implementation?
14 MR. EBERSOLE: Separate the implementation, 15 because they are different cats, and you can' t keep stirring 16 them together on the basis that they are all the same. That 17 is just pure hogwash.
18 DR. OKRENT: It seems to me, you have an 19 interesting set of points, but it is really incomplete in 20 several ways.
21 What do you have in mind, when we know that, for 22 example, the BWR has suggestions for several modifications 23 which don't automatically fit, for example, in the items you 24 mentioned? There may be similar things or different things
}
25 for MK2s. MK2s, ice condensors or something. Is this then l
ACE-FEDERAL REPORTERS, INC.
_ Nationwide Coyerage 800-33H646
, _ _ Xc 347 3700 _ _
2290 07 05 78 (3
's # DAVbw I supposed to be, for example, to be further developed or 2 what?
3 DR. BURDICK: This is just a "for example" list.
4 It is not meant to be all-inclusive, by any means. There 5 are things that could be thought of that I and my staff 6 probably wouldn't even dream of also.
7 DR. OKRENT: Let me mention one that somehow 8 isn't on this list. I am not sure whether it is thought to 9 be important or not.
10 Is there a role that steam explosions can play in 11 leading to a loss of containment integrity? I don't 12 necessarily mean the alpha mode. I mean steam explosions.
g-V) 13 DR. BURDICK: That is something that has to be i
14 taken into consideration when you apply your yardstick to a 15 particular plant. Are you suggesting that we deal with 16 implementation measurement problems explicitly in the CPDO?
17 DR. OKRENT: Not the CPDO. This is an 18 implementation that you are talking about here. I mean, 19 you are saying, do the following things.
20 DR. BURDICK: In measuring your specific plant as 21 to whether or not it complies with or meets the safety goal, 22 there are a number of things you would have to take into 23 consideration, and steam explosion would be, I would think,
, 24 one of the things that would enter into your consideration
}
25 of the uncertainties, as well as direct containment ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800
-336 6646 - . . , .
2290 07 06 79
- !D Iw-DAVbw 1 heating, as well as a few other things.
2 DR. SIESS: You've got a conditional probability 3 of .1 or 01.
4 How would you meant that for a steam explosion?
5 The only thing I have seen is that there are steam 6 explosions, but the probability is, one, that the 7 containment is going to fail. That is a large release.
8 That is an early release.
9 DR. BURDICK: You have to remember that this is a 10 goal. There are contributions to uncertainty in our -
11 measurement of PRAs that may require additional 12 phenomenological research to give us some idea perhaps of O 13- what the uncertainty has been or could it be reduced if you
, 14 postulate that you are going to get something that would 15 violate this goal, then what can I say, if you make me 16 accept that as a postulate? s 17 DR. SIESS: If you didn't have the conditional 18 probability, you could say, all right, I'll do whatever I 4
19 need to do to reduce the probability of steam explosions 20 below 10 to the minus 6, 10 to the minus 7, or whatever.
21 Then you would meet your requirement of a large release, 10 22 to the minus 6. As long as you have a conditional 23 probability in there, then you say I've got to take a steam l
24 explosion somehow. You can't through it into the 25 probability of the steam explosion. You have assumed a 1
ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 202-347 3700
- . ., .__- - _ - ___ - 80_-6646 4 336
2290 07 07 80 1 AVbw 1 steam explosion or hydrogen detonation where direct heating 2 would cause an early containment failure.
3 You say now I've got to have a 1 in 10 that the 4 containment won't fail, given this extreme event. It seems 5 to me that puts you on a real spot. Now you are trying to 6 eliminate events on some kind of a basis that has to be 7 deterministic. It can't be probabilistic.
8 DR. MARK: I wante.d to raise a question which I 9 think is very closely parallel, Chet. This conditional 10 probability of 0.1, going back to Item 3 on the previous 11 slide, you've got to have along with that some statement as
'y e 12 to what it is that happened against which you have"this .1
?
~
13 protection. And if you say it is a steam explosion, as Chet 14 suggests, that is quite different than saying it is a .1, in 15 the event that so-and-so happens, but you really need that 16 underpinning.
17 What kind of thing am I going to use as my base 18 against which the .1 is to apply? And I don't think you can 19 allow a total steam explosion.
20 DR. BURDICK: I would appreciate any suggestions 21 you come up with.
22 DR. MARK: But you need something that is the 23 worst possible thing you are going to consider, and then
( 24 there are things which might be invented, which are not C]/
25 going to be considered.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
2290.07 08 81 1 DAVbw 1 DR. BURDICK: I understand your problem.,
, 2 MR. EBERSOLE: Gary, on the third bullet, high 3 stacks for venting of noble gases.
4 I think you mean'anything that comes out with 5 krypton and whatever, don't you? You mean emissions at 6 large. But let me go on. There is a sociopolitical problem 7 there. I was very interested a few days ago, in fact,-to 8 see the media do its usual thing. This time they took the 9 Chernobyl high lofting, which was advantageous, to avoid
, 10 local severe conditions, and they took the inverse of that 11 and said look how far this goop can go.
12 The local politicians and the papers, again, they
( sell trash to sell papers. They never mentioned the 13 i 14 relative advantage of having the high stack or the 1
15 distribution versus an inversion.
I 16 DR. BURDICK: I think noble gases was intended i
17 here.
! 18 MR. EBERSOLE: Was it just noble gases?
19 DR. BURDICK: If you have a severe core damage 20 accident, the noble gases are probably --
21 MR. EBERSOLE: Are you talking about TMI 2?
22 DR. BURDICK: They are going to lay other things 23 out.
24 MR. EBERSOLE: Okay. I understand.
25 DR. MARK: What that should say is really ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80433H646
2290 07 09 82 l' ) something different. It is a high stack, which will still 1-sDAVbw 1 2 releae noble gas, but you don't have it for that purpose.
3 You filter out the iodine and the cesium, and you don' t ask 4 it to contain the noble gas, but you don't have it there in 5 order to let the noble gas out.
6 DR. BURDICK: Just a representative list here of 7 things that you might do.
8 DR. OKRENT: I should know, it would help if you 9 said, you should have considered a number of things, 10 including the following.
11 Right now, it just says "the following."
12 ,
That leaves one possibly to draw the conclusion ss/
13 that these are all that you intend. The first point.
14 The second point, I still don' t get clear in my 15 mind whether the Staff has in mind, should one get into a 16 situation where you have lots of things under containment 17 besides noble gases like cesium, which is, as you say, a 18 likely candidate, whether you think it will be important 19 that the filter divisions are designed to remove some 20 significant fraction of the cesium, I don't know, .9 or 21 v s better, whatever number the Staff arrives at. I think the k
22 ) trench use .9 or better. Or whatever happens to have been 23 the filter system in the plant, which was really not
- 24 designed for a large accident, happens to go up in smoke,
[}
25 tough luck.
ACE-FEDERAL REPORTERS, INC.
E 347 3700 Nationwide Coverage 80 4 336-6646
2290 07 10 83 (3
l'JDAVbw 1 Do I make the distinction clearly 2 DR. BURDICK: Yes, I think so.
3 DR. OKRENT: It is not clear what your thought 4 is. And if I can come back to the steam explosion issue for 5 a moment, it is my impression, and I would be interested in 6 learning more, so I can see where the thrust of the 7 information is, it is my impression that, for example, for a 8 MK2 BWR, over and under, that the steam explosion might 9 occur in the pool, threaten the pedestal, the reactor vessel 10 head and the containment or maybe directly the containment.
11 So that is not the so-called alpha mode.
12' Similarly, there may be other steam explosions, V .
13 again, not the alpha mode, which sorts of disintegrates the 14 vessel, which, in fact, have a nontrivial effect, either 15 directly on contaimment integrity in the course of events, 16 which may then affect containment integrity, and I have not 17 seen this examined in detail in the MK2 PRAs that were i
18 looked at.
19 I can't tell whether the Staff had some separate 20 effort which is going to provide all the necessary 21 information soon, or what. But it is sort of in my own 22 mind, a question someone raised, to which I don't have an 23 answer.
{} 24 25 MR. EBERSOLE: Along that same line, but not steam explosions necessarily, if you put in some sort of ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336-6646
2290 07 11 84 b- DAVbw 1 suppression system, whether it be water or ice, and there 2 are problems with it, and that is, if you find a mechanism 3 that will bypass it, you are in trouble, and there are 4 mechanisms.
5 One of the notable and interesting things on that 6 MK2, we have those long pipes that hang down in the water.
7 The Germans put two pipes on such a path when they get into 8 the water. The American system is to put one. You can 9 argue that it is great, you know, and put braces on it, et 10 cetera.
11 The ice condensor system has a blow-back problem, 12 but fortunately -- no, I'm'not sure. No, it doesn't. I fS '
(/ .
13 can't remember whether it works from the top down. It .
14 certainly still works, in that you melt the ice, but it is 15 not near so effective when you sweep through it. I don't 16 recall the rate.
17 But I am just bringing up a point. The 18 suppression systems are special cases, and you'd better not 19 bypass them.
20 21 .
22 23
{} 24 ,
25 l
l ACE-FEDERAL REPORTERS, INC.
202 347 3700 Naticnwide Coverage 800 336-6686
2290 08 01 85 p/DAVbur
- s. 1 DR. BURDICK: Both your points are well-taken, 2 and I have a note here to look further into those issues in 3 considering the CPDO further.
4 Jim Martin?
5 MR. MARTIN: Just to make sure we don't lose 6 sight of the forest for the trees, implied in this slide, 7 how you get into this is you look at the suggestion that the 8 probability of a large release should be one in a million 9 reactor-years or you look at the conditional probability of 10 .1 for a large release, even a core melt.
11 If you really want to do that, if you want to 12 meet those goal s -- there is a lot of plants out there that O 13 can' t meet L ose goals -- they would require these kinds of 14 actions.
15 DR. SIESS: I am not sure you can ever meet the 16 conditional probability.
17 MR. MARTIN: I am not sure you can either, but if 18 you want to try, these are the kinds of things you have to 19 meet.
20 MR. EBERSOLE: Bypassing a boiler can occur with 4
21 the actuation of a normal release on turbine trip because 22 the water is burdened under containment.
23 DR. BURDICK: We will have probably some more l
24 insight into theso questions when we look at the reference (a-)
25 plant studies that come along.
l ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwule Coverage 8 6 33(H W 6
2290 08 02 86 (3
h -DAVbur 1 MR. EBERSOLE: I think it was Brookhaven did a 2 study that water doesn' t act as a condenser when you lay 3 down a hot film on top of it. You have got to go through 4 it. Ice worked pretty good.
5 (Slide.)
6 DR. BURDICK: For new plants again, a 7 representative list of options that might be considered, the 8 same caveats apply here.
9 DR. SIESS: That first item up there, when I look 10 at a Mark III I see a containment that looks like the large 11 dry containment structurally. It is designed for about 15 12 psi, and I know I can design it for 50 or 60 psi. There are g-)
(/
13 two dozen out there of the same general structure that are 14 designed for that.
15 How much good does that do me?
16 DR. BURDICK: How much does that do you with 17 respect to meeting this offering here?
18 DR. SIESS: Well, if I design Mark III 19 containments for 60 psi, how much do I reduce the 20 probability of a large release with core melt?
21 I don' t know what the challenges are.
22 DR. BURDICK: There is a lot of variation from i
23 plant to plant.
24 DR. SIESS: It can't be that much. There's only
(~}
v 25 four Mark III's. I mean, they have different containments ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80433MM6
2290 08 03 87 (9
l 'DAVbur 1 but the systems are busy.
2 DR. BURDICK: I guess I don' t know how to answer 3 that off the top of my head.
4 DR. SIESS: The problem is you worry about 5 bypassing the suppression pool. If I do that in the Mark I 6 or Mark II, of course, I am just out into a containment 7 building with some standby gas treatment systems in order to 8 handle so much, but if I do that in the Mark III I am into 9 that secondary volume, that so-called wet well, which is 10 very large. It could be designed for 60 psi.
11 I haven't seen those sequences analyzed to know 12 what I gain there because if you increase the pressure f-)
'wd .
13 capacity that is another thing. That is the logical place 14 to do it.
15 We talk about the future, how much you can do to 16 increase the pressure capacity on a plant that is already 17 built.
18 DR. BURDICK: Yes. We are talking about new 19 plants here and what an owner would buy into. There could 20 be a variety of things that an owner could do to meet the 21 containment performance design objective. He would have to 22 make analyses, choices, options that were more cost 23 effective than others.
24 I can't right here say what he would buy at a
}
25 particular plant for a particular option.
l ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800 33MM6
__ _ _ __ 202 M .3J00 . _ _ _ _. _ _ _
I 2290 08 04 88 gm, i jDAVbur 1 DR. SIESS: I am not sure what you can do on the 2 dry containment. I know what you can do when you look at 3 the difference between Sequoyah and Watts Bar. Sequoyah 4 reduced the frequency as it went up and Watts Bar didn't.
5 So you wind up with about twice the density on Watts Bar 6 than we struggled with on Sequoyah.
7 MR. EBERSOLE: I can add to that. Sequoyah even 8 had to warm the water that you spray into the containment so 9 it wouldn't implode.
10 DR. BURDICK: That concludes my presentation, 11 Mr. Chairman.
12 DR. SIESS: Can you make a containment strong 73 V 13 enough to survive a steam explosion? ,
14 DR. MARK: Look, I think to a greater extent than 15 is reflected in the slides some of the things here are both 16 partial lists and they are lists of things to consider 17 rather than decisions as to what to do.
18 Are there other questions within the context of 19 the status therein?
20 I believe we have raised a fair number, at least 21 for the next couple of years.
22 You have a plan to bring out a final discussion 23 of containment performance options or proposals.
24 What is your idea as to when you will converge on l 25 something final?
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
- I 2290 08 05 89 I'T i iDAVbur 1 DR. BURDICK: About, I would say, mid-February of 2 next year.
3' DR. MARK: Boy, you are ambitious. That is
. 4 great.
~ 5- MR. EBERSOLE: I will say.
6 DR. MARK: Anyway, you are thinking of something 7 early next year, the first half of next year?
s
- 8 DR. BURDICK: The first quarter of the calendar 9 year.
10 DR. MARK: I wasn't going to hold you to that.
11 That is why I said half.
12 In tnat context, I think we will want to look at bv 13 this again at such a time as that before you have really put
, '14 the last jot and tittle on it because there seem to be lots 15 of things open, both in our minds and I judge also in yours, 16 about which it would be good to hear more.
, 17 DR. OKRENT: Let me ask a kind of general 18 question, Gary.
In order for you to arrivo at a recommended 19 l 20 containment performance design objective consideration, 21 whoever is the higher authority, the first question is:
22 Do you feel that you do need to have these, what 23 I will call, implementation licts for the existing and 24 future plans as part of what it is you present? Do you
(
25 think that is rather necessary?
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
2290 08 06 90 DAVbur 1 I am not trying to argue either way. I am just 2 trying to understand.
3 DR. BURDICK: That is a very good point.
4 Originally, we did not plan to go that deeply 5 into the implementation. We were looking at our effort in 6 providing an additional statement to be added to the safety 7 goals.
8 The approved safety goal policy statement 9 apparently views everything known afterwards as 10 implementation of that policy. We now have to view the CPDO 11 as implementation, and I am expecting further guidance from 12 the EDO office in regard to this.
()
fs 13 If we do get deeper into implementation, of 14 course, the schedule is geing to be lengthened the first 15 quarter of next fiscal year or next calendar year. Whether 16 that would make it hopeless, I don't know.
17 Matt, do you have anything to offer on that?
18 MR. TAYLOR: In terms of actual schedule, I can' t 19 offer much at that point in time. In terms of Dr. Okrent's 20 question, I would definitely say the implementation question 21 was one in which we would perhaps address risk management 22 strategies suitable to existing plants and perhaps goals 23 that might be of aid more to future plants.
/~ 24 Now, I am talking out loud at this point in time, G;'
25 but I would see those as certainly being divided in terms of I
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage mn33MM6
I I
2290 08 07 91 (S
l DAVbur 1 their implementation.
2I DR. OKRENT: So we suggest then that it may be 3 valuable or even necessary for the staff to propose risk 4 management strategies at the same time it proposes 5 containment performance design objectives for existing 6 plants.
7 Is that your guess?
8 MR. TAYLOR: That is my guess. I think what we 9 have talked about here today is a mixture of both.
10 DR. OKRENT: A second question. It is my 11 impression that I certainly have less than adequate 12 information on the behavior of the several -- and I mean 13 more than five -- types of containments that exist out there 14 today and their failure modes under various conditions, and 15 so forth.
16 So if I were going to try to think about 17 management strategies for these, it seems to me I would need 18 better information than I now have on what is thought to be 19 the potential, likely or probable or possible failure modes i 20 under a wide range of scenarios, and in fact I would include 21 external events, which complicates life.
22 Do you feel that you have or will have adequate l
23 information of this sort by the time you mention the winter 24 or spring of the next calendat year, enough to come out at 25 least with the trial balloon risk management approach plus ACE-FEDERAL REPORTERS, INC.
__, 202-347-3700 _ Nationwide Coverage. ,_
8%336696 _ _ _
2290 08 08 92
- / DAVbur 1 the performance objective for existing plants?
2 DR. BURDICK: Well, I think there is at least one 3 area that we are going to have to look into with respect to 4 certainly the reference plant information as soon as that 5 becomes available. That is scenarios that can get you to 6 what you are facing with this conditional probability. You 7 have to look at that.
8 There are probably a few other things we have 9 discussed today they are looking into the reference studies 10 for, also.
11 I think as far as stating the performance 12 objective, we are probably pretty cloge.
13 Now, if you are talking about achievability and 14 what a plant may have to do implementationwise, that gets 15 pretty complex pretty fact because there are a lot of 16 differences in these 100 or so operating plants, and 17 implementation strategy for one would probably not be 18 acceptable for another, 19 So there are a lot of problems.
20 MR. TAYLOR: I just wanted to add a thought, 21 again my personal opinion, in response to Dr. Okrent.
22 If you think by early spring of calendar '87 we 23 will have answers to all of the technical questions 24 regarding containment failure mode, that is not correct
}
25 obviously.
We won't have those answers.
ACE-FEDERAL REPORTERS, INC. -
y.~. m a - a c - ,. g
2290 08 09 93 I
1 ')DAVbur 1 I think what one can do is to come up with some 2 general guidance and provide some sound principles to follow 3 and sound principles to follow with respect to risk 4 management in existing plants.
5 Certainly, that will not cover all contingencies, 6 but I do know WASH-1400, for example, got into this issue of 7 the beta mode failure, the steam explosion in the Mark I 8 containment .
9 I don' t think we have fully resolved any of the 10 issues at this point in time, even though this has been 11 around for a long time. I don't expect we will have all the
~
12 issues resolved by the spring of '87.
O 13 DR. OKRENT: Could I ask, you are NRR7 14 MR. TAYLOR: No, I am on Roger's staff. I view 15 my job right now as to try to get the implementation 16 underway.
17 DR. OKRENT: EDO?
18 MR. TAYLOR: Right, the EDO's staff.
19 DR. OKRENT: Let me suggest, at least in my 20 opinion, for a long time there has been insufficient 21 attention given to the question of just how do many of these 22 containments fail and what are their failure modes, and I 23 think even now there is not a sufficient priority given to 24 trying to get a comprehensive assessment even for what I 25 will call the internal events for the relatively large ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
2290 08 10 94 (b
1DAVbur 1 family of containments and how you are going to bring in 2 external events.
3 You may be able to do it generically by certain 4 varieties, guidelines, but I really do think it is going to 5 take more effort, more priority among the national labs and 6 contractors and with them feeling that we have got to get.
7 the information needed for this. I don't sense that is the 8 case.
9 MR. TAYLOR: Your observation is well-taken. I 10 would agree that for some period of time we have not put a 11 priority of resolving all of the questions. I think this 12 has turned around in the past few years.
13 DR. OKRENT: Or at least since April? _
14 MR. TAYLOR: I would think so.
15 DR. MARK: Che't, you have something?
16 DR. SIESS: Is it appropriate for me to make some 17 philosophical remarks?
18 DR. MARK: Oh, always.
19 DR. SIESS: First, I would like to suggest that 20 any containment performance objective or containment 21 performance design objective we come up w3th must be both
- 22 achievable and demonstrable. If it isn't achievable, it is 23 not going to do much to protect the health and safety of the
{} 24 public, and if you can't demonstrate it, it is not going to 25 be very much help in reassuring the public that you are ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 08 11 95
[D lDAVbur 1 regulating the plants to protect the health and safety of 2 the public.
3 The other point is that I don't think anybody I 4 have read or anybody I have heard has stated clearly why we 5 want a containment performance objective. I have read 6 everything the Commission wrote as a majority Commission. I 7 have read everything the ACRS wrote, and I have heard a lot 8 of discussion, and it is still not completely clear why we 9 want it.
10 I hear references to defense in-depth, to 11 mitigation versus prevention. We have not been all that 12 clear. We have many references to uncertainties, lack of 13 completeness in predicting the course of an accident.
14 I think it is going to be difficult to come up 15 with an acceptable answer unless we start out by defining 16 why we want it.
17 If I look at what Asselstine and Bernthal have 18 said, their thinking is the public perception of risk, l 19 avoidance of psychological trauma of evacuation, or some way 20 of expressing safety in terms that people can understand, 21 not in terms of health and safety of the public.
22 That is somewhat in the same category as the 10 23 to the minus 4 or 10 to the minus 5 on core melt. That
{} 24 doesn't tell you anything about the health risk. It is 25 simply one step.
1 ACE-FEDERAL REPORTERS, INC.
L d 202-347-3700 Nationwide Coverage 800-330 6 46
4 l
2290 09 01 96
[D I don't think we've got a clear picture. If you l >DAV/bc -1 2 have it, I'd like to hear it. But I don't think the 3 Commission has given you a good basis for selecting a 4 containment performance objective. You've dug one out.
5 DR. MARK: Milt?
6 DR. PLESSET: I don't think my remarks are 7 particularly valuable. I'd like to compliment the staff, 8 and even the Commissioners on their attempt -- but not their 9 details. And predict that they'll never get satisfaction on 10 details, at least not from this group.
11 The other remark I want to make is about 12 Dr. Siess' question about a 60 psi Mach III. Would that O. .
And my answer is I think we gained quite a. bit.
13 gain much?
14 I'd like to ask sometime today how much that 15 would cost.
16 DR. SIESS: Not much.
17 DR. PLESSET: Not much. That's a good answer.
18 DR. MARK: That's for new Mach III's.
l 19 DR. SIESS: That's an easy one to estimate l
20 because you can take any PWR.
21 DR. PLESSET: Well, that's a very useful thing to l 22 know. He deserves a lot of credit for bringing it up.
23 I also have remarks on the steam explosions.
l I've got to understand those a little better. I think I
{} 24 And you can make a 25 understand a little bit about them.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800
-- - . . . -33M446
2290 09 02 97 DAV/bc 1 steam explosion you can contain very easily pin glass 2 vessels. I've done that. It's not very hard. It's what 3 happens after the spike that counts. I think we ought to 4 keep that in mind. The spike is very dramatic.
5 DR. MARK: It's the same as a hydrogen 6 explosion.
7 DR. PLESSET: Exactly. A lot of times, people 8 forget that. And that's all I have to say, Mr. Chairman.
9 DR. MARK: I want to ask, if not the very next 10 thing, before we stop, what comments NRR might wish to throw 11 in here, if any.
f3 12 MR. HOLMAN: My name is Jerry Holman, from NRR.
V 13 - I think it would be a little premature for me to 14 offer any comments other than our intent is to work with 15 research to try and develop something that's useful for both 16 existing plants and for future plants.
17 And I want to note that we expect an application 18 for an advanced BWR next fiscal year, and intend to start 19 work on it. It is within our game plan for the next year.
20 So this is timely.
21 MR. EBERSOLE: Did you say it was a PWR? ,
22 MR. HOLMAN: An advanced BWR. I think it's the 1
23 Winatchie G.E. design.
(~} 24 MR. EBERSOLE: No pipes.
N._/
25 MR. HOLMAN: Inside.
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 33M686
2290 09 03 98
^( 3 l' DAV/bc 1 MR. EBERSOLE: This leads almost into my next 2 question. Thirty odd years ago, I think it was critical 3 that we throw this industry out to the American ingenuity, 4 entrepreneurship, et cetera, et cetera. And we invented all 5 kinds of funny reactors and converged to the LWR's.
6 I have a simple but difficult question. If we 7 had 10 reactors today and you attempted to slot them all on 8 one plateau of requirements, you'd have some problems 9 because the ones at the top could be a hell of a lot better 10 and you wouldn't need to do cheaply what they could really 11 do to improve their performance, because you had a common' D 12 passing grade.
13 But, today, we've got two. The question is 14 this. I'm not going to say which because you can figure 15 that out for yourself.
16 In your development of the requirements, I think 17 you have to look at these systems, these plants in different 18 boxes and argue the following:
19 The number one plant can vastly improve its 20 performance by doing a few simple and relatively inexpensive 21 things. The other one cannot.
22 I must impose some sort of logic in the interests 23 of the country in putting different requirements on the
{} 24 plants according to their ability to achieve them, as Chet mentioned, because there are different abilities to achieve.
25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M646
2290 09 04 99 DAV/bc 1 And I don't think you can any longer escape 2 focusing on these differences.
3 Are you with me?
4 DR. BURDICK: I'm with you. I understand what 5 you're saying.
6 MR. EBERSOLE: That's all.
7 DR. MARK: Dave?
8 DR. OKRENT: Just a few thoughts arising out of 9 what Chet says. I'm not sure myself that we'll be able to 10 have either high level safety goals or containment 11 performance design objectives achievable and demonstrable, 7, 12 although I'd like that to be the case. -
N_/
13 But it may be a tough thing because I think there 14 may be pretty broad uncertainty bands. And then I don't 15 know what it means to say "it's demonstrable". So I'll 16 leave that one caveat on the back.
17 It seems to me that the idea of defense indepth 18 provides an adequate basis for one's trying to develop any 19 performance design objectives for serious accidents.
20 I think this is a subject that should have been 21 dealt with in the late sixties, and we just, for a variety 22 of reasons, delayed it even after Three-Mile Island. But 23 l it's time.
24 Even if we have something imperfect and not very
(^)
'a j 25 demonstrable, there are many things that are not ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-66 4
2290 09 05 100 DAV/bc 1 demonstrable'about the nuclear power plants. I hope it's 2 not bad, or have very bad side effects.
3 The only last comment. I've heard more than one 4 representative of the nuclear industry, high level 5 representative of the NRC, people in similar positions 6 suggest to the public on nationally transmitted programs 7 that the American reactors have. containments that will 8 contain severe accidents.
9 DR. SIESS: They also say they are reinforced 10 concrete.
11 DR. OKRENT: With little or no qualification on 12 these statements, leaving, in my opinion, an incorrect 13 perception of reality. _
14 DR. MARK: That's not the first time.
15 DR. OKRENT: In fairness, there have been some 16 people pointing out that this is not the case.. And they're 17 even pointing it out to utilities. But, nevertheless, I 18 think the Commission needs to address this forthrightly.
19 DR. MARK: Look, I think probably we should call 20 this to an end. I'd like to thank you very much, Gary. I 21 have a personal feeling that you're approaching this problem 22 in a very, very nice way.
23 MR. EBERSOLE: I agree with that.
/~3 24 DR. BURDICK: Thank you, Mr. Chairman.
V 25 DR. MARK: Listen, it says here we should resume ACE-FEDERAL REPORTERS, INC.
E -347-3700 Nationwide Coverage 800-336-6646
2290 09 06 101 1 DAV/bc 1 at 1. How about resuming at 1:15?
2 (Whereupon, at 12:05, the hearing recessed, to 3 reconvene at 1:15 p.m., this same day.)
4 5
6 7
8 9
10 11 12 v
13 14 15 l 16 -
17 18 19 l 20 21 22 23 rn 24 C'
25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M686
2290 10 01 102 I I) l DAV/bc 1 AFTERNOON SESS ION 2 (1:15 p.m.)
3 DR. MARK: The meeting will resume. I'm supposed 4 to talk for 10 minutes, but since we're five minutes late, I 5 will not talk at all and turn things over to Mr. Bernero.
6 MR. BERNERO: I have some slides that I passed 7 out some copies of the slides for today. Let me begin by 8 some remarks. The subcommittee's objective is a containment 9 performance design objective consideration, at least so 10 Carson Mark translated that acronym for me a little 11 earlier.
12 ,
What I'm talking about today is enhancement of 13 containment performance with only the most general objective 14 in mind, and a' quantitative assessment of containment 15 efficacy to be undertaken after the fact, if at all.
16 And I think it's useful as a perspective of j 17 achieving containment safety without necessarily knowing i
18 exactly how safe that safety is.
19 (Slide) l 20 Let me start out. We're talking about a generic i 21 letter, one of the regulatory instruments available in the l
l 22 panoply of tools we have to make life miserable for 23 licensees.
(} 24 As you know, with generic letters, they have 25 regulatory force. It's a way to tell a certain subset or ACE-FEDERAL REPORTERS, INC.
_ __ 202-347-3700 Nationwide Coverage 800-33M646
2290 10 02 103 DAV/bc 1 group of reactors that they have to do something.
2 Now, the setting for this, we're talking about 3 closure of severe accident issues. That is, remember, the 4 Severe Accident Policy Statement said that the present 5 generation of reactors are safe enough for the time, but 6 that what we call the IDCOR process, the systematic 7 appraisal of these designs for outlyers, for severe accident 8 outlyers, is clearly an order and should be. continued. So 9 that the assumption is th't there are individual plant 10 evaluations being done under that policy.
11 And what we're talking about here is taking a 7- 12 snapshot of the backend, containment performance only -- not L.) Not system reliability. And reaching a-13 the front end.
14 generic conclusion on that and using a generic letter as the 15 regulatory instrument.
16 The 'ey k regulations of interest, and I'll touch 17 on them briefly to refresh your memory, are general design 1 18 criteria 16 and 15 of the book, both applying to 19 containment, of course.
20 Now, the subject plants, we have 37 boiling water 21 reactors with pressure suppression containments. There's a 22 little table in the back. Later on, I'll touch on it.
23 Twenty-four of them are Mach I containments, nine are Mach II's, four are Mach III's.
(-]
\_/
24 25 The focus of this discussion and the focus of the ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 10 03 104 h#DAV/bc 1 activity is first and foremost on Mach I's. The same 2 . principles and approach to a solution apply to Mach II's and 3 Mach III's. But to avoid confusion, we're going to segment 4 the action and segment the discussion, because whenever we 5 try to shift from one type to the other, it becomes very 6 confusing.
7 DR. MARK: Could you help me? Everybody else 8 knows this. Roughly, what does GDC 16 and GDC 50 say?
9 MR. BERNERO: I'm going to put it on the board.
10 (Slide) 11 First, I want to recall for you what the Severe 12 Accident Policy Statement said. It said the most cost-13 effective options for reducing vulnerability to severe 14 accidents shall be identified and a decision reached 15 consistent with the backfit policy, the cost-effectiveness 16 criteria.
17 Now, in determining whether any change is 18 warranted to any containment out there, it is appropriate to
. 19 consider it as a backfit, whether or not the owner 20 voluntarily goes to do it; because that, in principle, is 21 merely determining is it worth fixing? Is it worth doing 22 something about?
23 So, first and foremost, we're talking about'
{} 24 options for reducing vulnerability to large release in this case and looking for justification thereof.
25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
2290 10 04 105 DAV/bc 1 And then this is a key part of the Severe 2 Accident Policy Statement, it says:
3 Where the technical issue goes beyond current 4 regulatory requirements, generic rulemaking is preferred, or 5 in other cases, you can dispose of it through bulletins, 6 orders or generic letters.
7 In this particular case, being subsets of 8 reactors which are themselves subsets of all reactors, 9 namely, one type of containment for boiling water reactors, 10 we identify the generic letter as the appropriate vehicle.
11 (Slide) 12 But it does go beyond the regulations. Here is
( .
1 13 general design criteria in 16, or at least the key phrase in 14 it. General design criterion 16 on containment design calls 15 for an essentially leak-tight barrier against uncontrolled 16 release as long as postulated accident conditions require.
17 Now,.I cannot say, and I have yet to find anyone 18 who can say that the authors of this general design criteria 19 thought along these lines. But it certainly is not a 20 contradiction of logic to say that essentially leak-tight as 21 long as postulated conditions require can be fairly 22 interpreted as the containment within the design basis is 23 leak-tight it contains.
And when you go beyond the design basis for
(} 24 25 whatever reason, one might seek at least to have a ACE-FEDERAL REPORTERS, INC.
Nationwide Coverqq 80 4 336-6646
_. ._ 3!$.237@7Q _ _
2,290 10 05 106 1/ DAV/bc 1 controlled release of radioactivity, or an at least partial 2 salvage of mitigation.
3 But that's clearly beyond the regulation as it's 4 presently written.
5 Now, if you turn to the other regulation of 6 interest, general design criterion 50 --
7 (Slide) 8 -- this is the one that calls for designing under 9 degraded core cooling. The containment design basis, and 10 note it calls for the 5044, 10 CFR 5044 metal water energy, 11 and the key words here, the degradation but not the total 3 12 failure of emergency core cooling functions.
13 So I think it's fair to say here that the design 14 basis includes degradation of cooling, but not loss of core 15 cooling, such as is typically associated with a core melt.
16 So, once again, we are beyond the regulations.
17 DR. KERR: Mr. Bernero, I'm puzzled that we are 18 using criteria developed so long ago, since a completely 19 different approach for dealing with severe accidents is used l
l 20 as a justification for what's being done today.
l 21 MR. BERNERO: I'm not using that as a l 22 justification. I'm just warning you that the regulations 23 don't go that far.
- ("T 24 DR. KERR
- You're showing us that what you're 1 \_)
25 doing is illegal, is it?
l ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Courage 30 4 336 6646
2290 10 06 107 DAV/bc 1 (Laughter.)
2 MR. BERNERO: No, no. I think the Commission's 3 severe accident policy statement recognizes that we could go 4 beyond. And a point I would make is that if you go beyond 5 the regulations, you're encouraged to do rulemaking. And if 6 rulemaking there be, I have a recommendation for what that 7 rule would be.
8 It would be two simple severe accident design 9 criteria.
10 (Slide) 11 This is just a list for your convenience -- 24 7-) 12 reactors of product line, BWR 2, 3 or 4, with a Mach I LJ 13 containment. They all have licenses. There are nine BWR 4 14 or 5 series, with Mach.II containments. The two which are 15 not licensed out of this set are Nine-Mile Point Unit 2 and 16 Limerick 2.
17 There are four BWR 6 with Mach III containments.
18 And 30 minutes ago, I talked to the owner of Clinton, which 19 is not yet licensed, and it should be ready Thursday for a 20 license.
21 So the Mach III's should all be licensed before 22 very long.
23 (Slide) l /~S 24 The institutional process we're in is closure of
(_)
25 the severe accident analysis for containment alone, so that i
ACE-FEDERAL REPORTERS, INC.
Walk 37-37TO Nationwide CoGEm 800-336-6646
i i
2290 10 07 108 DAV/bc 1 individual plant evaluations on boiling water reactors would 2 be confined to system analysis and not include the 3 containment. This would be dispositive for containment. No 4 further analysis, unless some particular plant discovered 5 circumstances within that particular plant so unique that it 6 couldn't implement it as a practical matter, or some such 7 thing.
8 Yes, Dave?
9 DR. OKRENT: Do you feel you now know enough for 10 making some recommendations which are both necessary and 11 sufficient to take the position of closure?
g-) 12 MR. BERNERO: Yes.
V 13 DR. OKRENT: And when was this information -
14 developed that put you in this position?
15 MR. BERNERO: I would say that the information 16 was developed in the last four or five years, and has been 17 before us without grasp, without being grasped, for at least 18 two years. And if you go into all of the extant severe 1
19 l accident analysis, I think you will see it's a common 20 thread.
21 DR. OKRENT: What convinces you that you have 22 looked at enough of what I'll call the phenomenological 23 picture to know that what you're proposing in fact is I
24 adequate for the purpose?
(U~)
25 I would suggest that if we go through the ACE-FEDERAL REPORTERS, INC.
202-347-3700 N81onwide Coverage 800-33MA6
_. _ . , -_ .. _ _ . = _ _ _ - ~_ - .. ._
2290 10 08 103
[D representation and revisit the subject, then it would be far 1"'DAV/bc 1 2 easier for me to answer the question.
3 DR. OKRENT: Is there in your opinion.a written 4 document to which you can refer me?
5 MR. BERNERO: No.
6 DR. OKRENT: There is no written document to 7 which you can refer me which will systematically go through 8 the various potential phenomena that might need to be 9 considered in dealing with the matter under bullet number 10 one, and which with due reason says that something should be 11 included and something else may be neglected?
12 MR. BERNERO: No, there is not such a document.
13 The generic letter, of course, has a substantial regulatory 14 analysis and technical analysis associated with it. We have 15 an internal draft of that working comments right now within 16 the staff.
17 I'm not even sure that that is adequate for what 18 you specify. I'm trying to make it nearly adequate, if not 19 adequate. It has to be the basis. That conclusion has to 20 be supported.
21 DR. OKRENT: Yes. But is it that you're hoping 22 you'll be able to propose?
23 MR. BERNERO: No. I argue that the facts are
{} 24 there, Dave. If we walk through it, and I think you, 25 yourself, can look at it and be hard put to identify ACE-FEDERAL REPORTERS, INC.
1 Jt 202-347-3700 Nationwide Coverage 800-336 6646
2290-10 09 110 (l
lDAV/bc 1 phenomenological circumstances or failure modes that are 2 both considerable and uniquely outside the pale of this 3 framework of solution.
4 I think you know if you look at it first.
5 DR. OKRENT: Well, you may be right, but I must 6 say, before I would feel comfortable agreeing with your 7 position or having suggestions for its modification, I, 8 myself, would need to have a rather thoughtful, detailed, 9 written document to be able to read and reflect on, although 10 I hope to learn a lot today.
11 MR. BERNERO: Okay, let me proceed.
. 12 I mentioned earlier the spectrum of 13 administrative options ruling a 5054 F letter followed-by an 14 order. That's something where we would write to each owner 15 and say give me the following information so that I may 16 determine whether I should amend, alter or cancel your 17 license.
18 That's rather cumbersome. We think this is the 19 most direct way and it can come out of the director of 20 Nuclear Reactor Regulation or come out of Boiling Water 21 Reactors alone, and the precedent for plant-specific, or 22 class-specific, I should say, class-specific resolution 23 would apply to ice condensers or subatmospheric containments
(} 24 or anything else as well as BWR's.
25 What I'm trying to say here in the process ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage, _ _ _ _
800-336-6646 ,_ _ __
~ . - . .- - - ._.
2290 10 10 111 .
l' 'DAV/bc 1 characteristics, there is an enormous library of severe 2 accident analysis. Unfortunately -- at least, in my view, 3 it's unfortunate -- it is shot through -- whether done by 4 the industry or done by the NRC -- it is shot through with 5 the character of an assessment.
6 It's an assessment to find out what is the risk 7 to quantify accident sequence probability, containment 8 failure probability, risk or release probability, and all 9 that. -
10 It does not lend itself to the regulatory thought 11 by what means can one wrap this up and draw a conclusion
-m 12 with respect to containment performance.
13 Now, the process of doing this clearly has to be 14 open.
I 15 DR. KERR: Mr. Bernero, I need some additional l
16 elaboration on that statement in order that I can understand 17 it. My superficial hearing tells me that you were telling 18 me that regulatory people don' t need to understand phenomena 19 before they regulate them.
20 And since I am sure that's not what you meant --
21 MR. BERNERO: Yes, they do, indeed. But the 22 literature has discussions of an assessment of a phenomenon, 23 you know. My favorite example is core debris.
{} 24 If you go into the modelers of core debris and 25 its transport, there are two schools of thought, virtually.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage __ , 800-336-6646
2290 10 11 112
[)
lDAV/bc 1 There's the hot consumme school which says that core debris 2 comes down with all the lack of viscosity that a bowl of hot 3 consumme has.
4 It goes right over what's the wall and you get a 5 zipper failure of the whole drywell.
6 The other school of thought describes the debris 7 more like Hawaiian volcano lava. It's slow. It's 8 slugg ish *. It's got that pretty dark crust on it. It's 9 principal thermal or corrosive attack is downward on the 10 concrete and not outward on the drywell.
11 Now, those phenomena are not well-understood 12 today. We have some understanding of debris travel. We 13 have some understanding of the expected viscosity that one 14 might find or the expulsion force that one might have, 15 depending on reactor pressure, and we are not going to have 16 a vastly better understanding five years from now.
17 The point is what one needs to do is look at the 18 available literature and say by what strategy or by what 19 approach might we make the difference, the uncertainty about 20 the core debris less significant to the outcome? Less 21 significant to containment performance?
22 That's what's absent with this abundant 23 literature that we have.
{} 24 DR. KERR: So you don' t need any more f 25 information. '
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M686
2290 10 12 113
[d ~)
l DAV/bc 1 MR. BERNERO: I say not.
2 DR. KERR: You just need an approach which will 3 ignore the existing information.
4 MR. BERNERO: No. Analyze and understand it.
5 Stop trying to find the right answer. I think it's absurd 6 after all these years to keep trying to find the right 7 answer. Tha.t's the viscosity for core debris, and that's 8 the amount of hydrogen generation in a severe accident, or 9 that is the direct heating quotient for a PWR with large dry 10 containment.
11 We have analyzed this thing to death. We've 12 spent millions and millions of dollars. It's time to look 13 at that information and ask some questions.
14 (Slide) 15 Now, I talked to the chairman of the 16 subcommittee. There are two approaches to saying why is a 17 thing worth doing. And this is my favorite safety goal or 18 safety objectives.
19 These'saw the benefit of public light at the 20 Commission's testimony to Congressman Markey earlier this 21 year in a discussion of containment performance and its 22 likely acceptability or unacceptability.
23 And it really, if you look at it, it constitutes
{) 24 what I would consider a useful rule if we have general 25 design criterion 191 and 192, a reworded version of these ACE-FEDERAL REPORTERS, INC.
!, . . _ _ _ . 202-347-3700 _ _ _ Nationwide Coverage 800-336 6646
2290 10 13 114 f s, idDAV/bc 1 two that said:
2 With respect to severe accidents realistically 3 appraised, the likelihood of a severe accident should be 4 very low and -- and I would put a caveat here -- it's an 5 "and" only -- if you have distributable vulnerability if you 6 suppressed all your accidents down to only a cataclysm can 7 cause core melt, then you won't have containment 8 performance. But, given typical distribution of core melt 9 vulnerability, in addition, if a severe accident occurs, 10 there should be substantial assurance that the containment 11 will mitigate its consequences.
12 13 _
14 15 16 17 18 19 20 21 22 23 24 25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4646
2290 11 01 115 DAVbur 1 In other words, that is defense in depth, and 2 that says that for the distribution of accident secuences 3 that any plant may well be vulnerable to you should have 4 substantial assurance that the containment will work. This 5 is not to say that the Boeing 747 crash won't 6 cataclysmically cause the core melt and cause the 7 containment failure, but it says for blackouts and LOCAs and 8 transients and whatever, ATWS, whatever sequence, you can 9 have both.
10 So if you look at a containment -- and let us say 11 WASH-140 0 was right. Postulate for the moment that 12 WASH-1400's appraisal of Peach Bottom was accurate and
- it
(:) 13 said, boy, there is not much chance of melting that core 14 but, good lord, if you do melt it that containment doesn't l
i 15 do anything. It fails virtually every time, emitting a 16 fairly large release.
17 You could argue in that case that one doesn' t l
18 mean one of the evident safety objectives and therefore you 19 must do something about that containment, that that is an l
20 argument of justification that bypasses the cost / benefit 21 process.
22 MR. EBERSOLE: Bob, may I ask a question?
23 In these considerations, do you still mix the two 24 kinds of LWRs in one pot and stir it?
l O 25 MR. BERNERO: No. I would say this one applied ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
2,290 11 02 116 1" DAVbur 1 to any LWR.
2 MR. EBERSOLE: Is that consistent with what we 3 were told this morning, that if you have had a severe 4 accident in the old plants the likelihood of containment 5 failure, called a large release here, should not exceed .1 6 in the old plants or 1/100 for the new plants?
7 MR. BERNERO: I have a remark later about the .1 8 or 1/100ths. But if you are willing to stop at substantial 9 assurance, I will be happy with it.
10 But to quantify it, you kriow, the fellow next to 11 you is a terror for doing that.
12 (Laughter.)
13 That is one way to approach. Should something or 14 need something be done on containment? But the severe 15 accident policy statement says, oh, but do it by 16 cost / benefit as well.
17 (Slide.)
18 Let me postulate some numbers to play with, some 19 numbers to think about as the basis for a cost / benefit.
20 Usually we identify here is a change, here is the 21 risk averted by the change, therefore this is our 22 cost / benefit algorithm.
23 I would rather approach it from the other 24 direction and say how much justification can you dig out of 25 a modern BWR with Mark I containment? How :auch risk or loss ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage _ _ _ 80433M646
2290 11 03 117 g
hlDAVbur 1 can you avert to justify change?
2 Let's start out with right now the IDCOR process, 3 and the NRC NUREG-1150 and other people are bandying around 4 numbers on the order of 10 to the minus 5 per year as the 5 cumulative or total frequency of , core melt.
6 I have a real hard time digesting those numbers.
7 I communicate nightly with the Angel Gabriel, and he has 8 assured me time and again that 10 to the minus 4 per year is 9 a good value.
10 So let's say that we will quantify 10 to the 11 minus 4 per year for the sake of argument.
m 12 DR. KERR: What is core melt? '
13 MR. BERNERO: Severe core damage or core melt.
14 In the PRA we make them synonymous.
15 DR. KERR: So this is synonymous with all of the 16 core out.of the vessel on the floor?
17 MR. BERNERO: Yes, large scale core melt out of 18 the vessel, through the vessel.
19 DR. KERR: What is the difference between that 20 frequency and the frequency of 10 percent core damage?
21 MR. BERNERO: I don't know. We don't know, and I 22 don't think anybody really knows that. I know the direction 23 of that. There is likely to be a substantial fraction that 24 don't go that far, and remember what I am trying to do here 25 is what risk aversion might I find, what might I justify.
ACE-FEDERAL REPORTERS, INC.
202-347-37N Nationwide Coverage 800-336-6646
2290 11 04 118 DAVbur 1 DR. KERR: I am just trying to find out what it 2 is, what the risk is describing.
3 You are saying 10 to the minus 4 is a good number 4 for the core on the floor?
5 MR. BERNERO: Core on the floor. I am also 6 saying that it is a full spectrum of sequences, including 7 blackouts.
8 We have constant debate about is the distribution 9 for this kind of reactor 98 percent ATWS and 2 percent
'10 blackout or 98 percent blackout and 2 percent ATWs, and you 11 don't know that. And if you come up to the 10-to the minus
(-
V) 12 e 4 with me, I want you to distribute. You have got an 13 appreciable fraction of blackout and you have got'an 14 appreciable fraction of small LOCA and an appreciable 15 fraction of ATWS and an appreciable fraction of whatever 16 else, so that you have full spectrum capability in your
- 17. consideration.
18 MR. EBERSOLE: Do you include the capability, l
19 which everybody thinks is there but which is not there, to l 20 relieve the containment in a BWR before the core is hurt and 21 thus keep it from melting -- any BWR?
22 MR. BERNERO: We get to that. I am not counting 23 on that right now. That could be appreciable.
I
(} 24 And remember, in WASH-1400 two-thirds I think was the fraction of all the core melts where overpressurization 25 l
I ACE-FEDERAL REPORTERS, INC.
l M 347-3700 Nationwide Coverage 800 336-6616
2290 11 05 -
119 DAVbur 1 caused the core melt. Wasn't it something on that order --
2 a half or two-thirds -- where the containment failure led to 3 the core melt rather than the other way around?
4 Now, what is the present containment capability 5 out there in the field? 24 Mark I's.
6 Later on there is a table of them in the back if 7 you want to look them up.
8 It is uncertain, and it is variable. I have gone 9 in almost every Mark I containment, and I have talked to the 10 plant staff. I have looked at the emergency procedures, 11 looksd at the ducts and piping, and so forth, and I am going g-) 12 to say that one out of two core melts will give a fairly
\' -
13 large release. It is like 50/50.
- 14 With the plants as they stand today, with 15 incomplete implementation of this emergency procedure or 16 that one, you have got a 50/50 chance of averting a large 17 release, given a large scale core melt.
18 DR. KERR: You are talking now about almost all 19 the core being on the floor of the contaimment. That is 20 what you mean by a core melt?
21 MR. BERNERO: Yes. Large scale core on the floor
, 22 and that you have only got about a 50/50 chance of coping
- 23 with that and substantially mitigating the consequences.
{} 24 Now, after -- and this remains to be seen whether this could be achieved, but I think it is reasonable.
25 ACE-FEDERAL REPORTERS, INC.
__ . __ _ . ._.. _ . 202-347-3700_ _ _ _ Nationwide Coverage _ _ 800-336-6646
2290 11 06 120
(~) Remember, we are talking only about the containment. So k 'DAVbur 1 2 don't touch the core melt frequency. The IPE may take care 3 of that.
4 DR. SIESS: Could you change that to per year?
5 MR. BERNERO: Oh, yes.
6 Boy.
7 (Laughter.)
8 That is Browns Ferry.
9 (Laughter.)
10 The IPE may reduce, but here there would be no 11 credit or no compensation. So once again that would 12 maximize the value of'whatever containment change there was, g-) ,
13 and for containment capability I will assert for sake of 14 argument here that you can vary from plant to plant, but 15 something on the order of one chance in 50 is achievable.
16 And this, Dave, when you sit down and say go through the 17 uncertainties, well, you have confidence in that.
l 18 DR. SIESS: You have written one out of 50 core j 19 melts, and you said one chance in 50. Let me read it the l 20 way you have got it written.
l l 21 That says if one of those 50 core melts involves 22 a steam explosion there is no chance of the containment
! 23 working.
l 24 Is that acceptable?
25 MR. BERNERO: No. If you define steam ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6686
2290 11 07 121 DAVbur 1 explosion as a sufficiently energetic steam explosion to 2 eject the head of the reactor such as to cause early 3 containment failure, then that is the one out of 50, the 2 4 percent right here.
5 The 2 percent failure of containment to give some 6 substantial mitigation has to embrace the likelihood of such 7 energetic steam explosion, any other mechanism that would 8 give an early failure.
9 DR. SIESS: So it is not for any core melt 10 scenario to have a 2 percent chance, but overall?
11 MR. BERNERO: Yes, overall, because you don't g- 12 know the exact distribution. This full spectrum is very b) 13 important. You don't hang your hat on I will have no 14 blackouts.
15 Now, given that, if you take this as face value, 16 you know, this is kind of looking at it from the front end, 17 what risks might be averted and therefore what kind of a 18 money kitty can we develop for backfits?
19 Just recall from the many analyses that are in 20 print, on the order of 10 to the 7th person-rem is a large 21 release. You know, an SST-1 type release. That is a 22 population exposure of on the order of 10 to the 7th 23 person-rems.
/" 24 So if you taka in to the minun 4 per year times U]
25 10 to the 7th person-rem times .5, you don't have very i
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 33MM6
2290 11 08 122 (D
A /DAVbur 1 many person-rem per year. It is 500.
2 Even if Harold Denton's statement is still 3 true -- you know, 90 percent, you have got 1000 person-rem.
4 That is a million dollars a year, roughly, as the quantity 5 of money, because this is going to just vanish, you know.
6 The 2 percent is going to be a negligible subtraction from 7 it -- what it will say is that if you use this value or 8 this set of values, you can' t justify more than about a 9 million dollars a year or $500,000 a year equivalent in 10 present value cost.
11 If on the other hand IDCOR and NUREG-1150 are
- 12 correct, and this is an order of magnitude lower, then you x) 13 are talking S50,000 a year at its present worth. It is.not 14 very much money.
15 And the message is then I assert that from a 16 cost / benefit point of view you can clearly justify 17 expenditures that are at plant overhead levels, that one 18 should not haggle about a million here or a million there, 19 that 1- or S2 million is the kind of backfit cost. That 20 backfit cost is not appreciable in this context and can be 21 justified by this kind of an argument. But certainly an 22 order of magnitude above that cannot. You don't have the 23 case.
24 Yes.
25 DR. SIESS: Bob, we just spent a few hours this ACE-FEDERAL REPORTERS, INC.
202-347 3700 NationwMe Coverage 800-336-6646
2290 11 09 123 O
1 iDAVbur 1 morning discussing the definition of a large release.
2 Now, you have a fairly large release.
3 MR. BERNERO: 10 to the 7th person-rem.
4 DR. SIESS: I am using the words up there, the 5 bottom line. In two places you say a fairly large release.
6 Is that your 10 to the 7th?
7 MR. BERNERO: Again for the cost / benefit 8 argument, I am saying 10 to the 7th. That is an SST-1.
9 Now, if you say fairly large release and go in 10 and make a BWR-3 release, that is not 10 to the 7th. That 11 is more like 10 to the 6th person-rem, and there goes your 12 war chest. You can' t justify a big change.
13 DR. KERR: Bob, what I seem to be hearing is this 14 is done in the noise level so far that you really can't 15 calculate it. Let's do something. A million dollars is not 16 much, and it won't matter if it 13 correct or not.
17 MR. BERNERO: No. The answer is that the 18 approach should have a fundamental character, and that 19 fundamental character should be don't redesign the plant 20 with a silver bullet like the filter system mounted on the 21 ground nextdoor, but exploit the system you have.
22 DR. KERR: You just told me that we don't 23 understand the phenomena that involved core melt interaction p 24 with the flow. Yet what we are trying to do, I think, is to
'u) 25 design a system that will mitigate against that. So we have ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
2290 11 10 124 1%-DAVbur 1 got to design it without understanding it.
2 MR. BERNERO: We have to cope with our 3 uncertainties about it.
4 DR. KERR: That is another way of saying that we 5 have to design it without understanding it.
6 MR. BERNERO: But this is clearly telling you 7 that the preser.t expected level of risk is such that it is 8 not possible to justify a massive or major redesign. That 9 is why the severe accident policy statement said what it 10 said and that a strategy which exploits the capability of i 11 existing systems, albeit with a couple of bucks here or
~
12 there, is the only acceptable strategy.
- C) 13 MR. EBERSOLE: You mean on a deterministic basis?
14 MR. BERNERO: Yes.
15 DR. OKRENT: I would like to disagree with your 16 fundamental conclusions. In the first place, I think the 17 ACRS has questioned on more than one occasion whether 18 cost / benefit considerations should play the dominant role in 19 the situation.
20 The second thing is the American public has been 21 led, at least in large part, to think that American reactors 22 in fact will contain core melt accidents with a very high 23 probability. By many discussions since Chernobyl, including 24 some I have heard from high representatives of the tTRC, not 25 all discussions give this impression, but many -- and in ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
2290 11 11 125 O
lODAVbur 1 particular, many of the earlier ones, which are probably the 2 only ones the public listened to.
3 Thirdly, I don't believe for a minute that if you 4 were to go up to the Congress and say we have a class of 5 reactors that we are not quite sure what the containment 6 capability is, it may be that 10 or 20 percent of the time 7 it will give and give a big release. That is what the 8 industry is likely to say. Or it may be that on the order 9 of 50 percent of the time you will get a large release or 10 maybe even larger. And in fact we haven' t included external 11 events in our picture aside from all the other things.
12 But let's assume this is the range, and when we g ,
13 do cost / benefit analysis we come out with -- at most w'e can 14 find a million dollars a year to spend on a plant whose 15 placement costs, were it to burn down and the insurance 16 i company were to give you money to rebuild it, would be 17 on the order of 3- to $4 billion.
18 I think -- in fact, I am quite confident -- that 19 the Congress would say what is wrong with your overall 20 safety philosophy when you think that containment is of so 21 little value that you can't find more justification than of 22 the order of a million dollars a year? Where are you wrong I
23 because you are completely out of step in the first place 24 with the perception given to the public, and furthermore you 25 are out of step with what the bulk of the public wishes of 1
l ACE-FEDERAL REPORTERS, INC.
l l 202 347-3700 Nationwide Coverage 800 336-6646
2290 11 12 126 (D the containment?
lDAVbur 1 2 MR. BERNERO: May I speak to that now?
3 There is an implication in what you said that the 4 .present state of events, which I think you fairly 5 characterized; namely, that the containment has maybe a 10 6 percent chance of failure or a 2Q peicent chance of failure
.?'
7 'or a 50 percent chance of failure or more leading to a large 8 release, that that is acceptable.
9 It is not. Ap ;J assert not only that it is not,
- ~
10 but that either of these two approaches -- remember the 11 first one I did. I said the safety -objectives, I can look
.g.r 12 at that and say that is not substantial assurance, and on 13 those philosophic grounds alone somethinp4need be done.
14 But if I can identify a soluklon that also fits 15 this averted loss thing and provides substantial assurance 16 for the million dollars a year present worth, then I do have 17 a happy solution. That is what I assert we have in front of 18 us.
19 20 21 22 23 24
{~)
25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 12 01 127 p
1( )AVbw 1 DR. OKRENT: You are addressing a different 2 question. I am attacking your conclusion that $1 million a 3 year is the sum-available. Now you are saying, well, with 4 $1 million a year, I am going to be able to get one out of 5 50, 6 MR. BERNERO: No, this has no limit on it. If I 7 seriously believed that substantial assurance is necessary, 8 if the backfit policy clearly says this, if you need it for 9 safety, you don't need cost-benefit. And if you assert you 10 need substantial assurance of containment performance for 11 safety and certainly 90 percent likelihood of a large 12 release is not acceptable, not at the 10 minus 4 regime. If O 13 you have core melt frequencies of 10 minus 7 per year, you .
14 might argue it, but not 10 minus 4. ~But there are the two 15 solutions, and I think they are congruent.
16 What it boils down to is, if spending $1 million 17 a year gives you clearly substantial assurance, why, if at 18 all, would you spend another SS million a year or $10 19 million a year? I know they can afford it.
20 DR. OKRENT: If you can do it for $1 million a 21 year, fine, but the way it was presented was, we can justify 1
22 of the order of $1 million, and I say, I think that were you 23 to present that to the Congress, I don't know whether they 24 would vote for 50 commissioners or one, but they would l
25 certainly change this.
l ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 804336-6646 202 py00
~2290 12 02 128 1 AVbw 1 MR. BERNERO: I know of at least one state 2 official who has argued that that is the wrong basis, that 3 the right basis is how much can they afford. And his number 4 starts at $10 million a year and goes up from there, because 5 they can afford it, but if it is not achieving 6 substantial enhancement beyond the $1 million, I think-it is 7 arguable. Yes.
8 MR. EBERSOLE: You are talking about an 9 investment per year and trying to integrate.that. I take it 10 you mean that is over the operational life of the plant.
11 MR. BERNERO: It is averted loss per year. If g 12 you are going to make one backfit to avert that, you take V
13 the present worth and, you know, multiply it by 5. .
14 MR. EBERSOLE: I think most big plants like 12 or 15 13 or even 1000 megawatts, I think they tend to reflect one 16 trip.
17 MR. BERNERO: Certainly.
18 MR. EBERSOLE: It is $1 million a day.
19 MR. BERNERO: That is why I call it in the level 20 of plant overhead. You don't even notice expenditures like 21 this.
22 MR. EBERSOLE: And you don't have to use this 23 line of thinking, I don't think.
24 MR. BERNERO: Tell the Commission that. The 25 backfit policy is clear. It is enunciated. It is a ACE-FEDERAL REPORTERS, INC.
l 202-347 3700 Nationwide Coverage 80433(H56s6
2290 12 03 129
('M lx 3AVbw 1 majority vote.
2 (Slide.)
3 Let me carry on. If we get to the meat of it --
4 what I have now is a strawman. It represents the sort of 5 requirements we are thinking of, and these are not frozen or 6 sacrasanct. I gave you a briefing two months ago or 7 something like that about the five elements, where you look i 8 at hydrogen, you look at drywell sprays, containment 9 pressure, core debris and the training and procedures that 10 glue the whole thing together.
11 So let me walk through those elements, at least 12 the first four, and talk about strawman criteria, which O 13 could or would be part of this generic requirements letter.
14 Now all MK1 containments are inerted, you know 15 that, and they have these requirements about the purge and 16 ventilation systems, and so forth. Really about the only 17 issue left is, there is a tech spec that allows 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> at 18 the beginning and at the end of operation for inspection and 19 check out, 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> of inert and nonoperation.
20 Is it possible to cut the risk by a requirement 21 that says, change that tech spec to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and make it at 22 reduced power, and then, of course, this is a matter of 23 clarification. There is a window. If you have an 24 identifiable leak in the middle of a cycle, people ask 25 whether that tech spec permits de-inerting long enough to ACE-FEDERAL REPORTERS, INC.
I Nationwide Coverage 800-336 i
,_ _- - . 202.347-3700 -
,6646 .
2290 12 04 130 4
bAVbw . 1 go in see if it is valve packing that you could tighten in a 2 moment and stop.
3 So this would clarify, that at any time you need
- 4. the access, a 12-hour tech spec at reduced power might be an 5 appropriate thing. This is merely a refinement. There is 6 very little risk change in this.
7 In here in the strawman, we went on the 1
hssumption that a plant's purge and ventilation rate would 8
9 support de-inerting in a period of four to eight hours, 10 with a little bit of a lean toward the four-hour end of it.
11 Leakage inspection and minor repair might 12 reasonably be done in another four-to-eight-hour period.
' O 13 Again, a little more toward the four-hour end of it, and if 14 you did it at reduced power, 33 percent or less, which also 15 helps occupational exposure, then the combination of risk is 16 worth getting, and it is a minor administrative change.
17 It turns out, we got some data recently. Vermont 18 Yankee went through their log books, and in their operating history with inerting, only 1 percent of the time have they 19 l 20 been de-inerted using the old tech spec. So this one here 21 may be a worthwhile refinement, but it is not sure that 22 there is going to be any risk at the end of this rainbow.
23 DR. OKRENT: Is there a risk of transient going 24 from full power to reduced power, in order to de-inert?
O 25 MR. BERNERO Yes, there is, but that is not ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
2290 12 05 131 b
A sDAVbw 1 that big a transient, to go down to one-third power.
2 DR. KERR: Reduced to a third?
3 MR. BERNERO: Yes, reduced to one-third power.
4 It is not that big, but it is there.
5 You see, in theory, what you can say is, you can 6 change this to zero. You can say no de-inerted operation 7 and no de-inerted operation will eliminate 1 percent of the 8 vulnerabilikty, if you take Vermont Yankee as a 9 representative sample.
10 DR. KERR: And if it makes maintenance more 11 difficult, it may make the risk greater?
12 MR. BERNERO: Yes. There is a countervailing
'~'
)
13 thing.
14 DR. KERR: This is the reason that I am a little 15 uneasy about only looking at containment.
16 MR. BERNERO: Now wait a minute. To be 17 practical, I don't think anybciy is going to go in an inert 18 containment to do minor repair or leakage identification as ,
19 a practical matter.
20 DR. KERR: I don't know what will happen. I am l
l 21 simply saying, if you reduce the time you have to do 22 maintenance from 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, this puts pressure on l 23 people who are doing the maintenance.
l I 24 MR. BERNERO: Certainly it does.
25 DR. KERR: I don't know what the risk will be, ACE-FEDERAL REPORTERS, INC.
( 202 347 3700 Nationwide Cowrage 804336-6646
2290 12 06 132 DAVbw 1 but there is no guarantee that it may not increase risks.
2 MR. BERNERO: ,They are competing risks, but you 3 are not working with a very big risk change anyway.
4 MR. EBERSOLE: Bob, the first four pages of your 5 list here are generic in character. Then you switch to 6 U.S. boiling water reactors. Now you are on just MKls and 7 that is going to be the end of the road for you?
8 MR. BERNERO: No. I said that is the first one, 9 because it is the dominant set of the population.
10 MR. EBERSOLE: When we talk about hydrogen 11 control ---
12 MR. BERNERO: This same thing applies to MK2s.
C 13 MR. EBERSOLE: What about PWRs? Where are you 14 going to put them?
15 MR. BERNERO: They don't have the same issues.
16 The ice condensors and the MK3s have a variation on this 17 issue with the igniters, but if you believe the analyses to 18 date, the PWR with large dry, doesn't have a hydrogen i 19 problem.
20 That is the minimum significance one.
21 Here is the maximum significance one. Drywell 22 spray.
23 (Slide.)
l 24 This is perhaps the most important safety feature (v~}
25 of the MK1 containment for the full spectrum of events of ACE-FEDERAL REPORTERS, LNC.
%)2-347 3700 Nationwide Coverage 800 33MM6
2290 12 07 133
(~)
1s ilAVbw 1 core melt.
2 All of the MK1 containments have drywell spray.
3 Oddly enough, they have designed the drywell spray to be fed 4 by the RHR system, with a couple of early plants being the 5 exception.
6 The bulk of the MK1 containments use those very 7 large RHR pumps. Now RHR pumps range in size from 5000 GPM 8 to 10,000 GPM, and as a result, the sprays are enormous.
9 They are sized to fit the pump, and there are two things 10 that you get out of. .
11 One, you either RHR the plant or you spray the 12 plant. There is no dual choice. You transfer flow. You O 13 don't split flow. ,
14 Secondly, every plant has got precautions. Don't 15 collapse the containment. This humongous rate of water 16 coming in there, you don't think about that. That is a 17 gasoline tank truck a minute coming out of each pump, and it 18 is an enormous flow rate.
19 So if you look at the possibility of backup 20 spray, the only thing that backs up an RHR pump is another 21 RHR pump, but do you really need that spray. The first l 22 requirement here is, cut that in half, just change the 23 nozzles.
24 Now what you would get out of that is the ability 25 to satisfy this requirement. This, by the way, I left off.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-33& 6646
2290 12 08- 134 O
-b- DAVbw 1 That is proposed requirements'. Provide an AC-powered 2 back-up water supply for the spray, other than the RHR 3 system. The RHR system is already piped to it.
4 Provide another system that can spray. Now if
- 5 you reduce the design spray rate, there is a whole host of 6 pumps now that can produce the spray in there and provide a 7 second AC independent water supply that can also produce a 8 spray and the availability of these two backups, one behind 9 the other is by remote manual operation. That is the
- 10 operator in the control room pushing a button or turning a 11 switch or by simple reliable manual procedure.
12 MR. EBERSOLE: Bob, if my' memory is correct, the O 13 RHR system is not two, but three functions, as follows:.
14 One, it has to cool the core, core pressure, and 15 it had to spray, as you said, and then it had to provide 16 enough water flow to cool the suppression pool. This 17 resulted in a design I am familiar with, which did have a 18 split flow, because that is-the only way you can get a mass I 19 flow rate large enough to cool it, if you have a small 20 break.
21 So the flow is split and the spray is part of the 22 system.
23 Are you talking about a different system?
24 MR. BERNERO: No, I am just saying, change the 25 nozzles. It will frustrate the design purpose of the ACE-FEDERAL REPORTERS, INC.
.- . ww meemwaie_ -
2290 12 09 135
(~)
LJDAVbw 1 system, but will enable other systems to produce a spray, 2 not a dribble.
3 If you've got a smaller capacity system, you get 4 a certain benefit out of dribble, but you get more benefit 5 out of spray.
6 MR. EBERSOLE: As I recall the process, though, 7 it was essentially a bypass where you have a blowdown of the 8 MK2s. I think it is claimed that the spray would get up and 9 carry you home. I am not sure that is the current way in 10 the NUREGS you propose or where the possibilities are.
11 Do you follow me?
12 MR. BERNERO: Oh, yes. I know what you are 13 talking about. It is spray condensation instead of 14 suppression pool condensation. I don't know the answer to 15 that. But what I am distinguishing here. This is the 16 proposed requirements strawman. This is what would be 17 desirable features but not required.
18 One, if you've got alternate water sources to go 19 the spray header. It is nice if they can go to the core.
20 I will give you an example. Vermont Yankee has a 21 whole bunch of AC-powered back-up water supplies. There are 22 all different kinds of service watet pumps that are already 23 piped to that. Also piped to that, they have a fire main 24 that has two 2500-gallon vented pumps out on the Connecticut
\
25 River Pump House. One is electric and one is diesel-driven.
ACE-FEDERAL REPORTERS, INC.
347"I # Nationwide Coverage 800 336 6646
2290 12 10 136 1 DAVbw 1 It is AC independent. They would pump the river right in, 2 but they could cool the core with that thing. So why bother 3 spraying the containment, if you can prevent the core melt, 4 see?
5 MR. EBERSOLE: Let me ask about this too. I 6 recall this -- as a matter of fact, this system in 1966 was 7 called the standby cooling system.
8 MR. BERNERO: Jess, what.you are really 9 eliminating is, you can really go into eny MK1 containment 10 in this country and you find pumps all over the place that 11 can do this function. And I assert that if you have this
, 12 reduced spray value and you seek the 90-10 mode of RHR, 13 where the bulk of it cools the core and the balance of it 14 sprays the containment, what you get is, you have many' 15 available water supplies for the flows, and you have a very 16 high level of assurance that in any accident sequence, 17 including blackouts, there is a curtain of water coming down 18 the inside of that drywell. It wets the walls. It wets the 19 reactor vessel pedestal. It wets the floor, so the core 20 debris gets quenched to some degree uncertain.
21 It keeps the pressure down for the condensables.
22 It cools the noncondensables to some extent, depending on 23 the water source. And basically what you have, you ha a 24 aerosol scrubbing possibilities out of it. You get many 25 benefits just from having that spray.
l ACE-FEDERAL REPORTERS, INC.
202 347 3700-Nationwide Coverage _ y3366M6 _
.2290 12 11 137 O
i_/DAVbw 1 Yes,' David.
2 DR. OKRENT: If you had no spray, and you had the 3 core melting through the vessel, it would be in a relatively 4 dry area.
5 MR. BERNERO: No. You'd have the water that is 6 in the vessel, but you know, there is a great deal of 7 debate, you know. Does the steam blanketing or does the 8 water fall on the core debris?
9 DR. OKRENT: If you have the spray, you may have 10 substantially deeper water.
11 MR. BERNERO: No, you are limited to about a foot 1 12 of water. Remember, in the MK1 containment, you look O '13 around, and you've got eight vents. They are typically.
14 about a foot off the floor. So that if you get to a foot of 15 water, it flows down into this suppression pool.
16 DR. OKRENT: And you have a foot of water, 17 whether you have 10 percent or full flow of the spray?
18 MR. BERNERO: I don't know. I think if you 19 started early enough, you'd have the foot of water.
20 DR. OKRENT: Suppose you went to 10 percent of 21 the spray. Is it possible that you have a different 22 condition that the molten fuel sees with regard to the 23 water under the vessel?
24 MR. BERNERO: I think not, because the spray rate 25 at 10 percent of the flow, which is hundreds of gallons a ACE-FEDERAL REPORTERS, INC.
202 347 370) Nationwide Coverage ') % 336-6686
-2290 12 12 138
./~T l b '1DAVbw 1 minute, should be able to fill up the floor.
2 Terry, do you remember the volume of Monticello 3 there?
4 MR. PICKENS: How much water?
5 MR. BERNERO: Yes. What is the volume of the 6 free area up to the vents? Vermont Yankee had it in their 7 report. They calculated, if you melt the whole core in the 8 lower head and the core support structure, would it even get 9 up to the vent level. I think Vermont Yankee came out seven 10 inches deep.
11 MR. PICKENS: I think it is about 1000 cubic 12 feet.
~) .
13 MR. BERNERO: 1000 cubic feet is a fair number.
14 So that is 7000 gallons. Now 10 percent. That
- 15 takes 10 to 20 minutes to fill that up.
16 DR. OKRENT: I am just trying to see how 17 different that was.
18 Now the question I was leading to is, in your 19 thinking, how have you made the decision as to whether there 20 will or will not be an ex-vessel steam explosion?
21 MR. BERNERO: I have concluded that an ex-vessel 22 steam explosion is not a likely failure mode of the 23 containment, but whatever finite probability it has has to be demonstrably variable in that 2 percent that I talked
{} 24 about earlier, and the residual likelihood of early 25 i
ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800-336-6646
- - - - - - _ _ _ _ . . - _ - _ _ _ _ . _ 202 347 3700 _ _
2290 12 13 139 4
DAVbw I containment failure.
2 DR. OKRENT: What is the basis for the judgment 3 that an ex-vessel steam explosion is not a likely source of 4 containment failure?
5 MR. BERNERO: The extant literature up to and 6 including -- I can't remember the number of the report, the 7 steam explosion report that Speiss shepherded out 18 months 8 'ago.
9 DR. OKRENT: I thought they mostly looked --
10 correct me if I am wrong -- at the question of in-vessel.
11 MR. BERNERO: They must have looked at in-vessel, 12 yes, but the arguments remain the same. In order to get a 13 high yield ex-vessel, you know, to get the coherence of.the 14 melt and all that.
15 16 17 18 19 20 21 22 23
- O 25 ACE-FEDERAL REPORTERS, INC.
-_ _ _ - - . e m- m m e s e .t _ _ _ ~ n - .
2290 13 01 140 1 DAV/bc 1 DR. OKRENT: In the first place, there are people 2 apparently using different arguments to arrive at the same 3 conclusion concerning in vessel. But I don't know that they 4 apply ex vessel.
5 What do you do to arrive at your position of 6 closure concerning MK I's for the moment? I'll come back to 7 MK II steam explosion questions again.
8 How are you going to pr6 vide the backup for 9 whatever position you have on ex vessel steam explosions?
10 MR. BERNERO: I'm certainly not planning to do 11 any more analysis of steam explosions. I'm going to assert 12 that the record is sufficient now.
) ,
13 DR. OKRENT: I think, if you're going to assert 14 the record, you'd better find a record that you can quote 15 and that is quite robust.
16 MR. BERNERO: Duly noted. I would point out 17 the last bullet on there which I think is significant. If 18 you have high confidence in a spray functioning in the 19 drywell during the core on the floor event, it reduces the 20 probability of drywell failure by cooling the electrical 21 penetrations, by cooling the wall, by whatever potential it 22 has for quenching the debris and retarding its ability to 23 directly attack the containment wall.
/~N 24 And one thing I would point out to you, all the L) 25 source term modeling in the world that I know of fails to ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nanonwide Coverage 800 33 4 646
l 2290 13 02 141 O model the effect of the biological shield behind the lk JAV/bc 1 2 drywell. Every time you see someone modeling, there was a 3 direct drywell failure, there is a calculation which assumes 4 the release is directly into the reactor building.
5 Fortunately, there's an awful lot of reinforced 6 concrete back there and it would significantly retard the 7 progress of any such failure. But, nonetheless, che spray 8 is significant in reducing the likelihood of drywell 9 failure. And also suppression pool bypass, remember, later 10 on, when we talk about suppression pool venting, you're 11 trying to bubble the stuff through the pool to get 12 scrubbing.
O 13 If a vacuum breaker sticks open, at least the 14 spray has had a shot at the aerosols that go down that 15 vacuum breaker. So you get an added benefit. It's like two 16 filters in series. The spray curtain is a filter followed 17 by the bubble pool.
18 DR. OKRENT: Might the spray, coupled with the 19 vacuum breaker sticking open, lead you to a combustible 20 situation in the drywell?
21 MR. BERNERO: You mean in a blackout?
22 DR. OKRENT: Well, the spray would take out any 23 steam. And if the core was on the floor, you're making 24 hydrogen.
25 MR. BERNERO: You're making hydrogen out of the ACE-FEDERAL REPORTERS, INC.
I 202-347 3700 Nationwide Coverage 800-336-6M6
2290 13 03 142
.1 bAV/bc 1 concrete.
2 DR. OKRENT: And you have a vacuum breaker open..
3 If you can bring in oxygen that way -- I'm just asking. I 4 don't know whether you think that's just not -- if you don't 5 see a problem.
6 MR. BERNERO: I don't think so. It's the pool 7 bypass scenario because the mass flow rate is out and the 8 flow is going out, bypassing the pool, going directly to the 9 surface of it.
10 Later on, we'll talk about the vent.
11 Yes, Chet?
12 DR. SIESS: Bob, would you explain that second O 13 bullet, after desirable? ,
14 MR. BERNERO: This one here? Yes. As part of 15 this reduction, now some plants do have an ability to split 16 flow, but most don't.
1 17 The procedure says do RHR or spray. What I'm 18 looking for is the flexibility that you would get with 19 90/10. You get an effective spray while you're getting 20 essentially complete RHR.
21 DR. OKRENT: Vacuum breakers between the wet 22 well?
23 MR. BERNERO: Yes. You've got a vacuum breaker 24 to the outside as well, but that's to the wet well. I don't l 25 know of any plant that has it to the dry well.
1 ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
2290 13 04 143 bAV/bc 1 DR. OKRENT: But, between the wet well and 2 drywell, you have a vacuum breaker.
3 MR. BERNERO: Yes.
4 DR. OKRENT: I'll stay with my question. There's 5 oxygen in the wet well.
6 MR. BERNERO: No, it's inerted.
7 DR. OKRENT: Ah, my error.
8 MR. EBERSOLE: Bob, are you making a particular 9 point of spray versus o_her means to flood the core and go 10 part way up into it?
11 MR. BERNERO: I prefer spray as against dribble
~
12 because dribble, a 600-gal 1on a minute dribble could fill up i \ .
13 the floor quickly and get a clench, but it doesn't spray the 14 aerosols. It doesn't spray the EPA's, it doesn't filter a 15 bypass and so forth.
16 MR. EBERSOLE: So you really would want a stand-17 by cooling system?
18 MR. BERNERO: Right. And what I'm saying is just 19 , get out there is they're there.
I 20 (Slide) 21 Now, pressure control, this gets into the vending 22 issue. The proposed requirements would be a reliable 23 capability to vent the wet well and EPG pressure level that 7s 24 has to be plant-specifically chosen from the competing L]
25 matters of the design capability or ultimate strength ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
2290 13 05 144 1 bAV/bc 1 capability of the containment itself. The structural 2 integrity confidence.
3 Secondly, from the vent valve itself, can you 4 open a vent valve with confidence at that level, you know?
5 Is it qualified to open?
6 Thirdly, from SRV differential pressure 7 considerations --
8 MR. EBERSOLE: You weren' t here this morning?
9 MR. BERNERO: No, I wasn't.
10 MR. EBERSOLE: Earlier on, we took up the matter 11 and I'd like to get your reaction to it. What we need is 12 what we call a rule curve or guideline that tells when
(
13 you're going to do this in context of vending core damage, 14 so you can really prevent the damge first.
15 MR. BERNERO: That's what we're talking about.
16 If nothing else, you've heard me say before, the vending of 17 containment was smuggled into existence by the regulatory 18 staff and the owners. And I think it warrants a much more 19 explicit and higher level political consideration. And this 20 I think is the appropriate forum.
i 21 That includes the integrated assessment of the f 22 public health risk of the noble gas release direct. It's so 23 hard to find that release even identified in the 24 literature. It's buried in with other calculations, yes.
l 25 DR. OKRENT: How do you assure that it's noble l
I ACE-FEDERAL REPORTERS, INC.
..._p2-343M Nationwide Coverage 800 336-6646
2290 13 06 145
$ DAV/bc 1 gas release?
2 ,
MR. BERNERO: I use that as a shorthand. It's 3 noble gas release plus the residual traces of whatever is in 4 there, because you've got a finite DF in a spray followed by 5 a suppression pool. And the early volatile activity only 6 went into the suppression pool and never saw the spray.
7 So, you know, it's noble gas plus delta.
8 DR. OKRENT: See, before, when we were opening 9 this up, you said, well, some people have this picture.
10 They have this dominant sequence. And others have this 11 picture.
12 But I think we shosuld use a melange. Here it 13 seems to me there's a picture where you might want to vent 14 because the pressure is too high, but you've got a lot of 15 radioactivity in the containment.
16 I don't see how you ruled that out.
17 MR. BERNERO: No. What I'm saying is the symptom 18 high pressure leads to the venting. It is not an integral 19 system high pressure with no radiation.
20 Some procedures in the plants now have that.
21 Vent if you have high pressure, but if you have more than 22 X-millirem per hour expected dose in the stack, don't do 23 it. And it kind of cancels it out.
24 DR. OKRENT: You're saying vent if there's high
- 25 pressure?
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336-6646
2290 13 07 146 DAV/bc 1 MR. BERNERO: Vent on the symptom.
2 DR. OKRENT: And I'm saying I can envision a 3 certain fraction of the scenarios where you have high 4 pressure and high radiation and it's not all noble gases.
5 MR. BERNERO: Even with the high confidence that 6 you have sprays spraying in the drywell and bubbles bubbling 7 in th; wet well?
8 DR. OKRENT: Oh, I don't know when and where I 9 should attribute high confidence to systems. For example, 10 we haven't seen what external events are going to do here at 11 all and whether your system is seismically designed. That's 12 just one aspect of it. Okay?
O 13 So I'm interested in knowing are the filters that 14 are there capable of taking any plausible fission product 15 load?
16 DR. KERR: Which filters?
17 MR. BERNERO: The filters are the water. We're 18 not talking about standby gas treatment system.
l 19 DR. OKRENT: Is that going to vent?
20 MR. BERNERO: We'll get to that in a moment. The l 21 standby gas treatment system has essentially no capability 22 to handle eititer high flow or substantially wet flow, you l
l 23 know, steam, water vapor. Just blow it out the paper s
24 filters and'everything.
l
%,)
25 Now what I'm saying is if you vent, and most ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
2290 13 08 147 AV/bc 1 plants do have some kind of path that goes from the vent 2 valve up to -- and by the way, when I refer to an 18-inch 3 diameter valve, those valves are 18-inch or greater in 4 nominal diameter, but they don't fully open.
5 They're blocked so that they go 80 percent open,
, 6 or something like that. The hydraulic diameter of that 7 valve is less than that.
8 But it is, in my mind, clearly more than decay 9 heat r<noval level, which would be the appropriate size of 10 the vent.
11 Now that path goes out at least if you just 12 chopped off the d'uct downstream of the valve. It's released
( 13 into the reactor building. It puts you in a desirable ,
14 situation rather than a required situation.
15 By venting, you have at least managed the 16 containment pressure to avoid catastrophic overpressure in 17 the large majority of sequences.
18 Now you have a question of what I might call 19 battle damage. And that is do you greatly threaten the i 20 reactor building or the workers or accident recovery, or 21 something like that? Because, now, you are venting a 22 mixture that is rich in hydrogen but inert into the reactor l 23 building atmosphere, which has oxygen. And it has noble gas 24 in it, which is not a trivial dose, you know, when you 25 really get close to a puff of it.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
- l. - -, _-
2290 13 09 148 bAV/bc 1 And burst resistance -- that's supposed to be a 2 "t" -- burst resistance ducting in the reactor building 3 would greatly minimize complications.
4 If you, for instance, vented straight to the 5 stack with that, you would accomplish two things. One, 6 you'd keep the hydrogen hazard away. You wouldn't have 7 hydrogen combustion. You would minimize the exposure to 8 operators who were trying to recover from the accident. And 9 the noble gas activity would be vented.
10 You know, a BWR stack is a really nice thing for 9
11 doing that.
12 DR. OKRENT: It's that repetition of the words, O 13 the noble gas activity, as if there's some law of physics 14 that that's all that will come out.
15 Let me get at it another way. The French are 16 using only large dry containments. So I would assume most, 17 if not all, U.S. utilities would argue you could only have 18 slow, if ever, failures of containment. And the staff i
19 sometimes comes in and says that.
20 But, nevertheless, as you know, they have chosen 21 to put in event which goes through a sand and gravel filter, 22 which is not, as I understand it, QA'd to death.
23 MR. BERNERO: And they look for a decade out of
.? 4 it. One decade.
25 DR. OKRENT: And then to a stack, I assume. If ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800
__ _ _ _ _ . 336-6646
2290 13 10 149
,m.
L )DAV/bc 1 you're going to have burst resistant ducting, why not to a 2 sand and gravel filter to cover your bets, and then to the 3 stack?
4 MR. BERNERO: I'll tell you why not. First of 5 all, if I owned a PWR, I would have a vent on my 6 containment. I wouldn't go through a factor of 10 sand 7 filter that I had to build. I would get a six-inch pipe ,
8 with a bunch of flanges on it, flexitallic gaskets and run 9 it from the spare penetration up to a quencher in the spent 10 fuel pool. And I'd get 25 feet of water filtration with 11 cleanup and I'd get gas filtration above that, with elevated 12 release.
O 13 DR. OKRENT:- I think it's a good idea as long as 14 you've checked that your spent fuel pool has a good seismic 15 capacity. Okay?
16 MR. BERNERO: Now you and I differ on the value 17 of the seismic capacity.
18 DR. OKRENT: Okay. It's just that some of the 19 pools may have -- well, I agree otherwise. You've got 20 something built in.
21 MR. BERNERO: But, now this, we're talking about 22 an abundance of water in the containment, both with respect 23 to the spray curtain and the suppression pool.
24 Now the data are all over the place. All the 25 debate in the source term documents, there are obviously l ACE-FEDERAL REPORTERS, INC. .
202-347-3700 Nationwide Coverage 800-336-6M6
l 2290 13 11 150 DAV/bc 1 residual activities in that noble gas. But they are 2 certainly greatly reduced from the unsprayed or unfiltered.
3 DR. OKP.ENT: You see, you've assumed a way then 4 any Frents that could lead to substantial activity other 5 than noble gases, either above the pool or above what was 6 the pool. Okay?
7 MR. BERNERO: Yes.
8 DR. OKRENT: It seems to me you haven't really 9 gone through the events that might lead you into a 10 situation.
11 MR. BERNERO: When you get into core debris, 12 you'll see one. But that's why the spray is so important.
O 13 During the core melt phase, keep in mind the sequence of 14 events. During the core melt phase, the core is still in 15 the reactor. The noble gas and the volatile activity comes 16 out essentially through the SRV lines down into the 17 suppression pool.
18 It can't bypass unless you break the SRV lines.
19 Once you have reached that stage where you have large-scale 20 core melting, you're now passed the volatile release stage.
21 You melt through the reactor vessel.
22 Now you get core concrete interaction with lots 23 more hydrogen, with the spalling of the concrete and 24 tellurium, or whatever else is going to come out of that i
f 25 mess. That's where the spray is so important.
i l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 NationwideCo erage 800-33(H%46
2290 13 12 151 I bAV/bc 1 You're trying to bubble that stuff out through 2 the suppression pool for scrubbing, but you're trying to 3 scrub it with a spray in the first place. That spray in the j
4 first place is what you rely on if you bypass the pool or if 5 the debris itself attacks the pool, as I'll touch on in a 6 moment, and changes the geometry.
7 DR. OKRENT: I repeat, you have assumed away the 8 sequences that get tellurium and so forth into the wet well 9 above tha water.
10 MR. BERNERO: I've assumed them away because of 11 the presence of the spray greatly reducing their
-12 concentration. Now you can argue that the aerosols are too
( 13 young for the spray to be effective.
14 DR. OKRENT: This is a magic spray that's always 15 there, but all these other things that they've engineered 16 redundantly somehow are not always present.
17 MR. BERNERO: I'm taking great pains to assure 18 that the spray is there.
19 MR. EBERSOLE: Bob, may I ask a question?
20 It's interesting to me that you're approaching 21 the gray area, the borderline, at least, of deterministic 22 design.
23 MR. BERNERO: It's not a gray area. I'm there.
24 MR. EBERSOLE: Let me go back to that first
<~)
NJ l
25 bullet; what you left me with was a feeling that there is ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
2290 13 13 152 DAV/bc 1 only two issues.
2 If I'm venting and I get radioactivity, you 3 didn't define how much, I'm going to have to close up.
4 -But what I need is a continuity, I believe, or I 5 call it a rule curve that says whatever that radiation is, 6 if it prevents a large radiation Y, which may be many orders 7 of magnitude greater, I may not be here because I have a 8 double end to the problem.
9 MR. BERNERO: There are two arguments I would 10 pose. One is the philosophy of not containing when an 11- accident in its early stages might very well be within the 12 design capability.
O 13 I think, as a philosophical step, I don' t like to 14 take, we have a logical philosophy that has been developed 15 and implemented. And that's the design basis of the plant.
16 When in doubt, isolate. How much, I would like on ATWS to 17 open the MSIV's because that makes ATWS go away.
18 But the containment philosophy, the general
(
! 19 design criteria I showed you are essentially leak-tight i
20 containment if you're within the postulated accident 21 envelope.
I 22 23 C) 25 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336 6646
2290 14 01 153 AVbur 1 MR. EB'2RSOLE : If I can hypothesize that I have 2 had a modest or even severe release, it is not yet a severe 3 accident.
4 MR. BERNERO: Now, we get into the second thing, 5 the second reason I have trepidation about it. It is like 6 emergency preparedness.
7 There are those who say that you can optimize, 8 minimize the pdblic exposure with a complex strategy of 9 shelter first, then run for three miles, shelter again and 10 do something else, you know. Intricate formulae. At the 11 time of an accident the signals are simple but confused.
12 I think any strategy that tries to see through O 13 and diagnose is doomed to failure. Pressure is pressure.
14 If you have got a ruptured disk, do it with a ruptured 15 disk. That is what the Swedes do. They have got a ruptured 16 disk in the pipe.
17 DR. KERR: There is an extension of this which 18 says don't close containment up at all, leave it open.
19 MR. BERNERO: On the contrary, the design basis 20 says the opposite.
21 DR. KERR: But the design basis is not designed 22 to take care of severe accidents. We agree that the only 23 accidents that really cause significant risk are severe
- 24 accidents.
25 MR. BERNERO: And don't call every accident a ACE-FEDERAL REPORTERS, INC.
202-347-3700, Nationwide Coverage 80 4 336-6646
2290 14 02 154 1 bAVbur 1 severe accident because they are within the design basi's and 2 you will never even open that vent valve.
3 DR. KERR: And you don't even need containment 4 for them.
5 MR. BERNERO: That is true, but you have got it.
6 DR. KERR: I am simply saying if you really are 7 trying to reduce t.he risk of severe accidents it may make 8 sense to leave the containment open.
9 MR. BERNERO: I think if you do the integral 10 strategy wo are talking about you will substantially 11 8 eliminate, if you have got the right procedures to cope with 12 the pressure variations -- you know, you have got to D-) -
13 remember you are taking suction on the suppression pool, you 14 have got considerations of pump cavitation, but with the 15 right pump discipline, water flow discipline, I think the TW 16 problem -- I call it -- or containment heat removal 17 degradation ultimately leads to core melt. That literally 18 goes away.
19 DR. KERR: Leave the containment open.
20 MR. BERNERO: Well, with the spray it virtually 21 goes away. Outside water supply for the spray.
22 DR. OKRENT: Are there any problems on 23 qualification of things that you are assuming?
rx 24 MR. BERNERO: No. I am assuming these are
'u) 25 internal events and do not need --
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6M6
2290 14 03 155 ls JAVbur 1 DR. OKRENT: I am talking about seismic 2 qualification now -- steam, heat, whateyer, high pressure.
4 3 MR. BERNERO: No, because the stuff is all 4 ready. If you look at the spray, the vent valves are a 5 qualification issue. The spray is not, you know. Just the 6 hardware, the nozzles, that in the vent valves those valves 7 are designed to close. They are .ot designed to reopen.
8 They are butterfly valves.
9 And some of the utilities are looking -- I think 10 you will hear more from the Cwners Group -- at times, again 11 in this framework of the dollar levels we are talking about 12 you can look at the thing and say I can close up this valve O 13 to work or you can say wait a minute, why don't I tee off 14 here and put the right valve in and go directly to the stat 15 or something on that order.
16 In neither case are you talking massive 17 expenditure. It is a matter of what does the plant look 18 like. That is a plant specific detail, but you do get into 19 that. Can you manually reopen the valve in a blackout?
20 You,go in some of the plants that are mostly air, 21 open, spring shut vent valves. Some plants have the ability 22 to put them on noninterruptible air supply, you know, the 23 emergency bus compressor. Some plants do not have that 24 capability.
25 An accumulator, an air accumulator, or even a 1
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 14 04 156 1 DAVbur 1 manual operator on it -- you know, that is an option on 2l those valves -- would enable you if you acted promptly to 3 manually actuate the valves.
4 DR. OKRENT: And all the sensors and actuators I 5 assume that you need are well-qualified for the conditions?
6 MR. BERNERO: Yes, because the EPG pressure I 7 referred to is typically design pressure plus some delta.
8 It has got to be design pressure plus 2 psi or 8 psi or 9 something like that, a minor variation. You are going to 10 run into one of those limits rather quickly above that.
11 MR. EBERSOLE: You quickly jumped on the PWR
,- 12 desirability of having a containment venting process, I ,
i,._)
13 presume through some suppression water or whatever.
14 Do you involve that, coincidentally, with 15 releases that also go directly through the core to get away 16 from the dependence on secondary?
17 MR. BERNERO: I don't. regulate pressurized water 18 reactors, so I won't go into that. But I have streng 19 personal beliefs in that regard, yes.
20 Let me go on. I would like to wrap up.
21 (Slide.)
22 This is on debris control, the top of the chart 23 there, core debris.
24 If you look at core debris, the only requirement
(^)
'w ]
25 I could see -- and this may be analytically satisfied -- a ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 14 05 157 1 DAVbur 1 typical BWR has the torus in a chamber called the torus ,
2 room. It is a box chamber. There are doorways or archways 3 in some plants going out into what are called the corner 4 rooms in the typical geometry. In the corner rooms you find 5 RHR pumps and things like that.
6 Now, the torus has 500,000 to a million gallons 7 of water in round numbers. If the core debris, even in 8 spite of the spray presence, is able to travel over to the 9 wall, it would hit a dry well shell, a steel shell, and 10 there are two possible backings. The biological shield is 11 fairly thick, reinforced concrete behind the steel sheel, 12 only by about maybe one or two inches, and there is'a 13 ' construction spacer in there. It is not a free-flowing 14 path. It is not a free exit for core debris even if it cuts 15 through the metal.
16 On th+ other hand, there are eight large vents --
17 you know, the LOCA vents. They are like big ducts, and they 18 have a mushroom cap on them, and they penetrate through the 19 biological shield, arcing down to enter the wet well and 20 penetrated at the top to go down with -- what do they call 21 i t --- the rams ' heads and stuff, you know, for bubbling the 22 steam into the water of the suppression pool.
23 Now, if the debris attacks simply the concrete 24 backed wall it is not likely to travel very far nor be 25 released into the reactor building, but if it attacks near ACE-FEDERAL REPORTERS, INC.
_ NationwideCoverage 800-336-6646
_ 202-347 3700 _ . , _ _ ,_
2290 14 06 158 1 AVbur 1 a vent and in some cases the cavity opening -- you know, the
. 2 lower head area of the reactor has a doorway in it, an 3 archway, and in some cases that is aimed right at a vent --
4 if it goes there, a tongue of debris could reach out and 5 flow right into the torus room. If it did, it is very 6 likely to fail the torus.
7 Now, one level of assurance that is worth getting 8 is -- and this number is arbitrarily drawn -- to retain the 9 water. What you would like to assure is that the water in 10 the. torus, if cut loose from the torus, doesn't go 11 anywhere. It stays in the torus room both as a quenching 12 pool so that any debris that goes out there isn't vigorously 13 reacting with dry concrete. It is wetted.or quenched. _And 14 secondly, the water in the torus at this time will have 15 already absorbed the large amount of volatile activity, the 16 iodides, and you don't want that water running off as part 17 of the accident management.
18 MR. EBERSOLE: Bob, there is a perturbation on 19 that. Some of the early old donuts had a suction header, 20 and that was a bad thing. It led to a higher probability of 21 loss of torus water, and that led to the finding that if you 22 do -- as you say -- you tighten up the room and protect the 23 corners and then you continue to pump water, you would.have i 24 eventually a small steamship. In other words, the donut 25 floats. It carries enormous forces up with it.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-33M646
2290 14 07 159 bAVbur 1 MR. BERNERO: Actually, if you,take this -- some 2 owners have been looking at it. I can't say that it is yes 3 or no in all the reactors, but I have the impression that 4
4 all plants satisfy that requirement or can verify they can 5 satisfy it using the corner rooms.
6 MR. EBERSOLE: No, that is not right. The three 7 unit Browns Ferry plant has open squares for the very reason 8 that a finding was made if you pump the water in like that 9 the torus is going to come up off the concrete.
10 MR. BERNERO: But I am saying it flows out of the 11 torus room into the corner room but goes no further.
12 , MR. EBERSOLE: That is true, but it gets the RHR
(_) 13 and other pumps when it does that.
14 MR. BERNERO: Not always. Many plants say, no, 15 they have got them on pedestals, so it won't.
16 MR. EBERSOLE: Well, they have a limited level, 17 but --
18 MR. BERNERO: You are getting to my desirable 19 thing. If you have got any reason to diddle with the flow 20 in the reactor dry well -- you know, like the concrete floor 21 that is there now -- a modification of it that put concrete 22 curves that would retard debris would be a desirable thing, 23 but it is pretty hard to justify making that change on its 24 own merit.
l 25 But if you can avoid loss of recovery systems ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 14 08 160 DAVbur 1 like RHR if that torus room water flows out, that is 2 desirable, too.
3 MR. EBERSOLE: What are you going to do with the 4 flotation problem?
5 MR. BERNERO: It isn't there. I just don't think 6 it is there.
7 MR. EBERSOLE: If you put in water from this 8 continuously available external source, it is a once-through 9 process, it goes on ON, and you have hypothesized that the 10 torus is going to deliver it.
11 MR. BERNERO: But if you lift the torus, what 12 could you do to the torus except rupture it?
13 MR. EBERSOLE: It is already ruptured. -
14 MR. BERNERO: Yes. The failure mode.
15 MR. EBERSOLE: You haven't yet hit the floor.
16 You are cooling away, and the water is running out on the 17 floor.
18 MR. BERNERO: I think this is a rather benign 19 accident when you are that far out.
20 MR. EBERSOLE: I can't recall what the evolution 21 of events was.
22 MR. BERNERO: In any event, this argument is no 23 mandatory curbs, but an assured knowledge of where the water
{} 24 is.
DR. OKRENT: So you are willing to accept the 25 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverg 800-336 6 86 ___ ___
2290 14 09 161
/~T lk dAVbur 1 possibility that the steel liner may go?
2 MR. BERNERO: That the debris, if it reaches the 3 steel liner, has those two paths, and again independent of 4 the debris attached there I have got spray coming down on 5 it, and I consider it very unlikely that the debris is going 6 to get directly into the reactor building except via the 7 torus room. That is the path, and so if we have a quenching 8 pool there -- and then of course you know that is not 9 airtight. The vapors can get out of there, but the vapors 10 can only come out of the quenched debris in the pool or from 11 the dry well perforation which has been sprayed on the way 12 out through the perforation.
(:)
13 MR. EBERSOLE: Point of clarification." I was in 14 error.
15 You have preassumeo that the core is molten. I 16 was in the front-end of this.
17 Do you follow me? You have already got the core?
18 MR. BERNERO: This is with the core in full 19 melt.
20 DR. OKRENT: And again we don't have to worry 21 about a steam explosion?
22 MR. BERNERO: In the torus room that is right.
I 23 DR. OKRENT: Why?
24 MR. BERNERO: The quantity and the rate.
25 (Slide.)
ACE-FEDERAL REPORTERS, INC.
. a>=->m __ m - c - _ _x'>~ ..-_ _
2290 14 10 162
(~]
lDAVbur 1 This is a list for your information.
2 DR. OKRENT: This is documented somewhere?
3 MR. BERNERO: We will refer you to what 4 literature is available, Dave. You know, I can sit and say, 5 yes, there is still an uncertainty about everything we do, 6 and I just consider it an extreme stretch to speak of a 7 steam explosion in the torus room from a tongue of debris 8 eating its way' through the vent. You know, a sufficiently 9 energetic steam explosion to be of interest is just not in 10 the cards.
11 DR. OKRENT: You don' t see any chance for molten i
12 material going down there?
13 MR. BERNERO: Yes, molten material, but the yield 14 would be some vigorous embolism of steam and not much more, 15 and it is open to the reactor vessel.
16 DR. KERR: It seems to me you will almost 17 certainly have a steam explosion.
18 MR. BERNERO: Yes. We have to distinouish 19 between the ohysical phenomenon of steam explosions and a 20 sufficiently energetic steam explosion so as to do something 21 really terrible.
22 DR. KERR: I don't think it ought to be dismissed i
! 23 yet.
{) 24 DR. OKRENT: I am sorry. At the moment I can't 25 tell on what basis you judge that steam explosions are
[
l ACE-FEDERAL REPORTERS, INC.
L An-347-3700 Nationwide Coverage 800-336 6 9 6
2290 14 11 163
('3 h--DAVbur 1 unimportant in the dry well or in'the wet well. They may 2 be.
3 My impression is if you have molten material 4 falling into the water the chance of an explosion is 5 significant.
6 MR. BERNERO: Oh, yes. A steam explosion of a 7 tolerable type.
8 DR. MARK: You mentioned a number. Is that a 9 fairly decent number? Do you want a ton of molten stuff in 10 a millisecond to explode?
11 MR. BERNERO: Did I mention a number?
12 DR. MARK: You said a ton.
13 MR. BERNERO: I didn't say that.
14 DR. KERR: I have seen explosions with less than 15 that.
16 DR. MARK: Well, I don't care. Why need I care?
17 I have got to have something in the neighborhood of a ton of 18 stuff.
19 DR. KERR: I remember earlier arguments that said 20 in order to get the in-vessel explosion that was needed you 21 would have to have about a quarter to a half of the total 22 core falling simultaneously into the bottom.
23 DR. MARK: In a period measured in milliseconds.
24 Whether that is true or not, I don't know.
l O 25 DR. OKRENT: You see, again what you need perhaps l
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
2290 14 12 164
(~)
b 'DAVbur 1 is a certain large amount reacting at once, but to react at 2 once it doesn't have to fall at once.
3 So anyway, I am just saying I think the staff 4 needs to show that these in fact --
5 DR. MARK: Without having the numbers, I agree 6 they need to be looked for and looked at.
7 MR. BERNERO: Let me try and wrap up.
8 This is for your information. It is a list of 9 all the Mark I containments.
10 One of them, Vermont Yankee, has undertaken a 11 special containment study that dovetails rather neatly with 12 this generic consideration. They are looking at their Cs 13 containments plant specifically. ,
14 DR. SIESS: Does it make any difference if they 15 vent into a concrete containment? The gap between the 16 lining and the concrete is a little smaller than it is with 17 the others.
18 MR. BERNERO: I don't think that is a significant l
l 19 difference, nor do I think that Nine Mile Point, Unit 1, and l 20 Oyster Creek, which don't have the RHR-spray combination but 21 use a containment spray -- they are a variation on the 22 theme, but the principle is the same -- I think debris 23 travel and the value of the spray is what counts and it is l 24 the dry well spray that counts.
25 But I wanted to point it out, Vermont Yankee has l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6646
2290 14 13 165
,O l /DAVbur 1 a containment study that was made available in the first 2 week of September. We had a public meeting on that study on 3 September lith, and it is rather interesting. They broke 4 the study into two parts.
5 One part was an assessment of what they think 6 their containment failure likelihood is today. In other 7 words, are they a 90 percenter or what?
8 Then the second part was an explicit address of 9 the elements of containment reliability or failure that I 10 just spoke of in this generic letter context.
11 (Slide.)
12 I would give you one table from that study. It
,7 -
V This is a table taken directly. All I 13 is in your handout.
14 did is have the source typed at the bottom.
15 This table illustrates the danger of very careful 16 accounting of containment failure probability. It'goes 17 without saying that there are not three significant figures 18 in every one of these calculations and every one of these l
l 19 estimates. Those are clearly for bookkeeping purposes l 20 only.
21 What Vermont Yankee did was a rather rational 22 exercise. They took Peach Bottom, which is a much more l
23 heavily analyzed plant, and they set up a set of accident 24 sequences extracted from Peach Bottom and then used j 25 engineering correction factors to adjust the frequency of ACE-FEDERAL REPORTERS, INC.
,, _ _ ._ _202-347 3700 Nationwide Coverage 800-33M686
2290 14 14 166 - 167 I
l / DAVbur 1 those accident sequences to what they could reasonably 2 expect in Vermont Yankee.
3 In other words, if you are looking for something 4 that is dependent on diesel generator power, Peach Bottom 5 has four diesel generators for two plants, Vermont Yankee 6 has two diesel generators for one plant.
7 Then they have a hydro station nextdoor, and they
'8 treated the hydro station rather pessimistically, and they 9 made adjustments or fudge factor corrections to get a core 10 melt frequency contribution from each one.
11 Then they have containment event trees to get 12 early release, which is E, or late release, which is L, and 13 then high, medium, or low with respect to type of release --
14 you know, the geometry of release and energy.
15 And they came up, counting all of these, with 7 16 percent containment failure probability, and that is the 17 reference number they used.
18 So what they say is for the containment as it 19 stands today we think it has a 7 percent chance of failing 20 in a core melt.
21 22 23 i
() 24 25 l
l l
ACE-FEDERAL REPORTERS, INC.
i 202-347 3700 Nationwide Coverage 800 336 4 86
2290 15 01 168 DAVbur 1 I think an equivalent number now for Philadelphia
- 2. Electric is about 15 percent by their own estimate, and in 3 neither case is the owner now saying that is good enough.
4 Vermont Yankee , in their containment study, has 5 identified a number of things they are still looking at, and 6 they have expressed what appears to me to be a clear intent 7 to sort out -- I will call them -- the fixes or adjustments 8 or procedures that are most suitable to their plant so that .
9 they can enhance the containment performance and cet some 10 number notably lower than this.
11 I prefer to think of containment failure in
~ 12 decades, that 90 percent, 10 percent, 1 percent, .1 of 1 13 percent, and so forth.
14 Now, in my mind it is obviously not acceptable to 15 have 90 percent probability of large release given core 16 melt. 10 percent is a little too close for comfort. 1 17 percent might be acceptable.
18 I prefer to approach it as substantial assurance 19 rather than a quantity because every time I upset the ratio 20 of one sequence to the other I have to reevaluate and 21 reassess what I have, and I think that is a chronic thing.
22 I think that one can approach this thing essentially from a 23 deterministic approach, get the high level of assurance of the spray, and so forth, and then assess later and assess
{} 24 for years if need be as information develops and avoid 25 ACE-FEDERAL REPORTERS, INC.
. =ne _ _ --* cem._ mm _ __ _
2290 15 02 169 hAVbur 1 trying to pin it down to a containment performance criterion 2 that is strictly numerical.
3 (Slide.)
4 There is a chronology attached here for your 5 interest. It has the June 16th owners group meeting, the 6 June 30th Vermont Yankee study commitment, a July 25th 7 commitment by Boston Edison, which is also doing a 8 containment enhancement study although they haven't produced 9 a report yet.
10 September lith, we had two meetings back to 11 back. One was with the BWR owners group essentially on the 12 same slides that we just reviewed with you here, and then U,n 13 that afternoon we went and met with Vermont Yankee on the 14 results of their study and where they are going.
15 Now, there is a meeting going on today in 16 Bethesda, NRC/IDCOR meeting. Then tomorrow with Dr. Kerr's 17 subcommittee is a follow-up on that meeting.
18 Then let's see, this is today's meeting, the September 23rd, this meeting right here. September 24th is 19 20 Dr. Kerr's meeting.
21 Now, October 23rd, we expect to write a letter to 22 Vermont Yankee saying here is what we think of your study 23 and here are some questions.
24 On November 13th, in the morning, we are l
25 tentatively scheduled to do a revisit with the owners group ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
I 2290 15 03 170 3-mDAVbur 1 on results of further analyses that they will talk about, I U 2 am sure, in the next presentation, and in the afternoon of 3 that same day --
4 (Slide.)
5 -- we are tentatively scheduled to meet with 6 Vermont Yankee for the answers to our cuestions and 7 follow-up of the analyses they have been doing.
8 Our schedules calls for November 19th we would 9 have internal -- that is the CRGR review of this draft 10 generic letter and its regulatory analysis in preparation 11 for it going out for comment. We have scheduled a month 12 between the CRGR meeting and the actual issuance of the
(') 13 draft generic letter for public comment.
14 And then lastly, December 24th, we would expect 15 to put another letter to Vermont Yankee which amounts to a 16 safety evaluation report along the lines that you have 17 studied your containment consistent with the proposed 18 requirements we are talking about, and if the proposed 19 requirements as published, if they stand as the final, then 20 here is our safety evaluation of your containment, and I 21 think it would be very useful to all the owners prior to 22 getting the comments and of course getting to a final.
23 DR. SIESS: What interaction have you had with 24 the Commissioners on this?
() 25 MR. BERNERO: I am talking to one Commissioner ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage S(A336-6646
._ _ _ _ 202 347 3700
l l
l 2290 15 04 171 j 1 at 4:00 o' clock today on this, and I will be talking to UTDAVbur 2 their technical assistants, but I have had no formal 3 dialogue with the Commission itself.
4 DR. SIESS: And no advice from them?
5 MR. BERNERO: No specific advice.
6 DR. OKRENT: If I can make a few comments.
7 First, let me say I am pleased to see the 8 initiative taken by the staff to improve the Mark I 9 containments. So if earlier and in the near future I make 10 comments which sound negative, overall I am happy to see the 11 initiative.
12 Okay, I must say I was disappointed, speaking for
() 13 myself, to learn about this via -- I don't know whether it 14 was Inside NRC or getting a piece of mail that Denton had 15 sent to somebody a couple weeks later or whatever -- but it 16 seems to me personally that it is a big enough issue, and 17 furthermore it is an issue in which the ACRS has lona 18 expressed an interest that you might have considered in fact i 19 discussions with the ACRS -- and I don't mean a briefing --
l 20 long before going so far down the road.
21 And in fact in your listing I don't even see 22 specifically called out ACRS comments either prior, during, 23 or af ter public comment.
24 I don't know how the committee will feel. In my
() 25 opinion, they should be providing comments before CRGR l ACE-FEDERAL REPORTERS, INC.
l 202-347 3700 Nationwide Coverage 800 336-6646
2290 15 05 172 If-DAvbur 1 review, but that is a personal opinion.
(_)
2 MR. BERNERO: Let me interject. I briefed the 3 full committee under the guise of future business.
4 DR. OKRENT: I was there, but that was after.
5 MR. BERNERO: Two months ago.
6 DR. OKRENT: But the ball was already --
7 MR. BERNERO: Let me assure you, I had just 8 pulled the ball out of the closet, Dave.
9 DR. OKRENT: Well, you pulled it out of the 10 closet, but you know I can go back and look at B&L reports 11 and draft reports, and so forth, that antedate that time by 12 quite a bit.
() 13 In any event, I am offering a personal opinion, 14 but I am going to raise the same question with the fulf 15 committee as to whether they think procedurally they should 16 have been brought in earlier, and so forth.
17 A second and very different question. I 18 mentioned seismic before. It may be that if one had the 19 benefit of a reasonable seismic look at some plants and a 20 reasonable seismic look at the things you are talking about, 21 one would judge that in fact it didn't change anything. It 22 is possible. .
23 At the moment I don't even know -- you may 24 know -- if the dry well spray is Safety Class 1, seismically
() 25 designed, and so forth. I just don't know.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336 6646
2290 15 06 173 3DAVbur 1 MR. BERNERO: It might not be.
G 2 DR. OKRENT: My recollection is it is not 3 seismically designed, but I could be wrong.
4 MR. BERNERO: I have gotten mixed answers on it.
5 DR. OKRENT: So let me just suggest that there 6 are a range of questions. You know, there is a seminar that 7 the staff is holding -- I guess it is two weeks from 8 today -- and in the opening statement on the seminar it says 9 it appears that the seismic contribution to risk may be 10 appreciable.
11 This is the staff letter, not something the ACRS 12 wrote, and it is a meeting to try to consider how to modify 13 Appendix A on seismic and geologic criteria.
f( )
14 But nevertheless, it seems to me as a minimu'm we.
15 had better get scme smart people going about seismic PRA and'
! 16 also what is going on here to see could there be some weak-17 spots.
18 I mentioned one possibility on the spray itself 19 that you at least need to think about, as well as are there 20 some scenarios aside from what you are thinking about that 21 change things. I am bothered by that term " closure" l
22 obviously. I mean, if I were a utility, I would want to 23 take it seriously.
l 24 MR. BERNERO: I hope they do. I want to take it l
() 25 seriously.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 15 07 174 1 DR. OKRENT: On the other hand, if you have J^3DAVbur G
2 really left out a family of scenarios that are important --
3 MR. BERNERO: Dave, I know this sounds like let's 4 change our mind and get it over with without doing a 5 thorough job, but we have badgered and bewildered ourselves 6 with severe accident analysis for years.
7 We published an advanced notice of proposed 8 rulemaking of a severe accident rule in October 1980. It is 9 now almost October 1986, and for six years we have analyzed 10 the subject to death and not made a single change in a 11 blessed plant out there, and this is ridiculous.
12 We are looking at 7 percent or 15 percent today, 13 and that is no better-than what we had in 1980.
(]) ,
14 DR. OKRENT: I am more frustrated than you '
15 because as far as I am concerned the question began in 16 1966.
17 MR. BERNERO: Yes, it is ancient.
18 DR. OKRENT: Or earlier.
19 So -- and I said originally, and ! meant it, that 20 I am much in favor of the initiative. I think, though, in 21 the end it needs to be on firm ground or have the necessary 22 caveats and/or both or whatever.
23 MR. BERNERO: Incidentally, the ACRS 24 participation -- we can't schedule the ACRS. We can only
() 25 solicit comment. So I am clearly doing that. I was clearly ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
2290 15 08 175 T^NDAVbur 1 doina that when I talked to the committee before in July or G
2 June or whatever.
3 DR. OKRENT: My last comment is one you are used 4 to. You and I both know there are supposedly some PRAs 5 sitting out there for BWRs. We are talking about BWRs and 6 the Mark I's, but I assume that Mark II's are not far away.
7 So let me issue a public request for you to
-8 transmit to the appropriate utilities. I understand Browns 9 Ferry, Hatch, Susquehanna maybe, among others, have one form 10 or another of the PRA, and I think even though these were 11 done --
12 MR. BERNERO: Independently, outside I would love
() 13 to have them. ,
14 DR. OKRENT: They are sources of information, it 15 seems to me, that utilities would be wise to take the point 16 of view we will all be better off if we were all smarter.
17 DR. KERR: I certainly agree with Dave, and I am 18 pleased to see the staff taking an initiative in this 19 important area. I am uneasy -- and that is as far as I l 20 would push it because I have been almost since the beginning i
21 of severe accident research -- that so much emphasis is put 22 on what I would say or what I would characterize as the 23 worst possible situation.
24 We don't really, it seems to me, have an
( 25 extremely serious problem unless we have almost all of the ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6 686
2290 15 09 176
'J'NDAVbur 1 molten core on the floor of the containment. The small O
2 amount of experience we have had up to date would suggest 3 that there are other possibilities that accompany severe 4 core damage. The core may not even get out of the vessel.
5 Yet we have had, it seems to me, very little 6 emphasis put on research or investigations, or whatever,
'7 that would try to intervene during the process that might 8 begin when the core damage starts. Inadvertently, we have 9 had at least one intervention which prevented things getting 10 worse.
11 In my own view, this is the more likely accident 12 than the one that gets all of the core on the floor. I am
-() 13 just a bit uneasy that all of the emphasis is going toward 14 that accident because I am not sure it really envelopes the 15 other less serious accidents, and I am not sure in the long 16 run it is the best thing to do.
17 Now, I don' t want to get caught in the trap of .
18 letting the best eliminate everything else. So in no sense 19 do I say stop what you are doing, don't do it anymore. It 20 leaves me a bit uneasy because it seems to me all the 21 emphasis is going to the worst possible accident, and I am 22 not sure that this is the one that deserves this much 23 attention.
24 For example, if one could be convinced that the ,
() 25 likelihood, given 10 percent core melt, that you would get ACE-FEDERAL REPORTERS, INC.
_ - - , . , , . _.___ _ . M-317_-3700_ __ _ _ _ _ Nationwide Coverage _ . _800 3_3&6686__
2290 15 10 177 7"'DAVbur 1 the core on the ficor is one in a 100, this at least makes V
2 cuantitative considerations quite different than what they 3 are now where one assumes that the probability is one.
4 Yet after all these years of research we don't 5 have, so far as I know, anywhere a good result, or even a 6 bad result, that would give us that probability, and yet it 7 is an extremely important consideration. What do you do?
8 MR. BERNERO: Bill, as an underpinning of any 9 strategy to deal with severe accidents, I think one should 10 be very careful about confusing lack of a good quantitative 11 estimate for lack of a good strategy. I think we have 12 excellent methodology to analyze, to understand the
() 13 potential contribution to core cooling of different systems, 14 different operating modes, different emergency proceduEes.
15 Secondly, when we get into containment here, such 16 as the issues we just discussed, the greater importance 17 attached to the dry well spray as against the containment 18 venting, for example, significantly reflects that the 19 partial accident is the one that first of all may be 20 prevented if not arrested by system reliability, by cooling 21 the core, and which can be mitigated to a sufficient degree l 22 by dry well spray in many instances so that you never have 23 to vent the containment, so that you somehow, at least for j
24 containment pressure envelopes, stay within the design
() 25 envelope.
l ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800 336-6646
l'
-2290 15 11 178
?SDAVbur 1 You do have and can have a strategy which gives O highest priority to preventing or arresting the accident, 2
3 second priority to dealing with the intermediate levels of 4 the accident, and final priority to the worst case accident 5 and still not know quantitatively what you have.
6 You know, if every utility in the country that 7 has a Mark I containment goes out and does the final product 8 that I am speaking of, I don't know what the containment 9 failure probability is best quantified to be, but I really 10 don't care. It is down there low enough that it is 11 acceptable. ,
12 There is substantial assurance of containment 13 initigation of a core melt, and if the front-end is done in a
(])
14 commensurate way, you know the reliability of systems. Then 15 I think anyone with a boiling water reactor -- you take 16 Vermont Yankee. You look at their study and their l
! 17 systems -- and I have been up there to walk it myself -- I 18 think any reasonable person looking at it, excluding 19 external events and Boeing 747s -- but take station 20 blackout. It is virtually impossible at that plant. There 21 are so many ways to cool that core, even in a blackout.
22 You can have the strategy and yet not have a good 23 number, and I am not troubled by the lack of a good risk 24 number when the risk number is down in the noise.
() 25 DR. KERR: It seems to me that you don't really ACE-FEDERAL REPORTERS, INC.
202-347-3700 ' Nationwide Coverage 800-336-6646
2290 15 12 179 7' DAVbur 1 get in serious trouble with a Mark I unless you have most of
(]/ That is all I am saying.
2 the core on the floor.
3 MR. BERNERO: That is true. That is the 4 vulnerability. It is a very adaptable system. Every Mark I 5 out there has two independent, diverse ways to cool the core 6 if it doesn' t have AC power.
7 DR. KERR: It seems to me it would be smart to 8 really know what the probability is getting all that core on 9 the floor as contrasted with having sericus core damage 10 because it may be too low.
11 12 13 14 15 16 17 18 19 20 21 22 23 i
24
. 25 I
I ace-FEDERAL REPORTERS, INC.
202@7 73?M _ , , , _ _ . _ Nationwide Coverage 800-33H646
'2290-16 01 180 1 MR. BERNERO: It seems to me that you still have fAV/bc 2 the two safety objectives. And it seems to me that it's a
- 3 good idea to have the defense indepth.
a.
! 4 DR. KERR: Of course.it is. But the amount of 5 defense depends on what you think the likelihood to be.
6 After all, we've talked about numbers for a large release.
7 We said a large release shouldn't be greater than. We don't 8 say it should be =ero.
9 DR. MARK: You said any reasonable person.would 4 10 be encouraged by what is really there. Does that include 11 Governor Kunin?
12 MR. BERNERO: I think so. In fact, I've talked 1
() 13 to Vermont state officials who are following this very
~
14 closely. At the recent Vermont Yankee meeting that was a 15 remark I made to one of the state officials. He didn't seem 16 to disagree with it because they are not only in possession i
17 of multiple, just one af ter the other dif ferent ways to i 18 supply cooling water and already piped directly to the core, i 19 but they are exploding it with procedures and even looking
! 20 to modify some of the valves to make it even easier for the 21 operator.to use it.
l 22 I don't know if Governor Kunin is going to get up i
23 her nose enough about system reliability to comment on it.
. 24 DR. MARK: She knows enough to raise questions I
() 25 though. [
1 ACE-FEDERAL REPORTERS, INC.
?~ -- nmcv a~- __ --
2290 16 02 181
'T'\DAV/bc 1 MR. BERNERO: She raised some good questions.
()
2 It's a Mach I containment and it's an older plant. How 3 confident are we that it's in good shape, and all that.
4 So...
5 DR. MARK: Any other questions for Mr. Bernero?
6 Because he's going to try to escape as soon as we let him.
7 MR. BERNERO: No, no, no. I can never pass up an 8 opportunity to jeer at the other groups' spokesmen.
9 DR. MARK: If that's it, then we have another 10 presentation ahead of us. I suggest we take a break until 11 20 after 3.
12 (Recess.)
A 13 DR. MARK: The meeting will resume. I believe,
(_j 14 representing -- how on earth do you say BWROG7 -- is 15 Mr. Pickens. Do you really represent anyone besides 16 yourself, sir?
17 MR. PICKENS: I work for Northern States Power.
18 My name is Terry Pickens. I'm current chairman of the BWROG 19 owners group.
20 DR. MARK: Are all the BWROG owners with you in 21 the sorts of things you'll be saying?
22 MR. PICKENS: All of the General Electric boiling
, I 23 water reactors, 23 utilities are represented in the BWROG l
24 owners group. And we've been developing this and getting 25 the participation of the BWROG executives.
ACE-FEDERAL REPORTERS, INC.
t 202-347 3700 Nationwide Coverage 800-336-6646
2290 16 03 182 DAV/bc 1 We've had two executive meetings to address this 2 subject. And as I'll discuss in my slides, NUMARC is now 3 ~ going to be getting involved in the subject also.
4 So we are going to be representing the industry.
5 DR. MARK: You're giving us a pretty 6 well-representative view, not signed and secled?
7 MR. PICKENS: Not signed and sealed yet, but 8 we're moving towards that.
9 (Slide.)
10 With that, I'll move into the presentation. Bob 11 showed you the slides which he had last week, on 12 September llth. I will also use the slides which we used as
() 13 a presentation on our severe accident containment policy 14 industry evaluations.
15 The objectives of this evaluation, Bob called in 16 the BWR industry, or called in IDCOR and owners group 17 representatives on June 16th and presented us with 18 restructure of a severe accident containment policy, and the 19 five elements which are contained in there.
20 At that meeting, a commitment was made by IDCOR 21 and the owners group to go off and evaluate those five 22 elements.
23 Today, we want to present the results of that, 24 along with the IDCOR consultants, the BWR consultants which j
1
() 25 were working on this. We also have some individual utility ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6f>86
2290 16 04 183 7'~NDAV/bc 1 evaluations which were performed. Those have all been
\~/
4 2 folded into one-large report. Bob's been presented with 3 that report last week.
4 That's what forms the basis for the slides I'll 5 be showing you here today, and discussing some conclusions 6 reached by utility.
7 (Slide.)
8 You'll see this slide twice. But it's good to 9 run through what the actual conclusions were. It sets the 10 tone somewhat of the rest of the presentation that we'll be 11 going through.
12 The August 19th meeting which we held in
-( ) 13 Philadelphia during the period of June 16th to August 19th, 14 we put this work together, which was a pretty significant 1
15 chunk of work.
16 . We had 20 or 23 BWR owners group utilities 17 represented at that meeting, 20 executives. At that 18 meetirc, we presented the results of the IDCOR owners group 19 analysis, what some of our cost estimates were, risk 20 reduction, all the different types of things we had taken a 21 look at. And these are the agreements, and this is what was 22 proposed by the executives to continue working on these 23 things.
24 The first was to continue a constructive dialogue 25 with the NRC. They appreciated the work, I think, that the ACE-FEDERAL REPORTERS, INC.
- _ , ~ _
_ _ 202 347-3700., , _ . Nationwide Coverage _ 800-336 4 46_ __ _ _ _
2290 16 05 184 F NDAV/bc 1 owners group and IDCOR had done with the NRC. I think we UI 2 appreciated Bob coming to us, asking us for input on this, 3 going back and forth and having a lot of good dialogue on 4 this.
5 The actual agreements or the motion which was 6 passed at that meeting on August 19th was:
7 Commit to implementation of Revision 4 to the-8 emergency procedure guidelines. This is the result of lA 9 elements which Bob had on procedures and training and 10 concern that the emergency procedure guidelines and their 11 underlying strategies were not being implemented uniformly 12 across the BWR industry.
() 13 There was a lot of discussion at the meeting on 14 August 19th and it was concluded that this is really ju'st a 15 reaffirmation of the post-TMI, where we said we'd implement 4
16 emergency procedure guidelines. And if Revision 4 is the 17 best technical quality, best product we've got put together, l
l 18 we should all step up and say we will commit to 19 implementation of that revision.
20 An that commitment as been made now by the 21 industry.
l 22 We contacted NUMARC to consider the severe i
23 accident containment issues as a generic indust:y issue.
l 24 That has been done now. That was done last week via the
() 25 NUMARC technical committee and the NUMARC steering l
l ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800-336-6646 I. - , _ ._. , _ _ _ ___ ._ _ .202-347 3700
2290 16 06 185 DAV/bc 1 committee . And it was unanimously approved to be taken on f-]/
2 by NUMARC as a working group, a technical issue for them to 3 work on.
4 One of the desires, one of the reasons why it was 5 taken to NUMARC is because of the implications which, 6 whatever the BWR's do, might have flowing over into the 7 PWR's, both the ice condensers and the large tray. There 8 might be elements that come out of this if you start 9 handling severe accidents which are beyond design basis 10 accidents; it will have an impact on what the pressurized 11 water reactors will.do.
12 So NUMARC has taken the time.
()
us 13 -
Continue working with the NRC to better define .
14 the issues for resolution. There's a lot of work that still 15 needs to be gone into, from a lot of the questions I heard 16 during Bob's presentation, still need steel into refining 17 the elements to really look at what the pinned down 18 requirements will be.
19 Then, finally, we discussed -- we did not commit 20 to doing it, but the executives endorsed and sent me back 21 and some of the others to go back and propose through the 22 normal BWR owners group mechanisms a severe accident insight 23 review of the emergency procedure guidelines, Rev. 4.
24 This type of severe accident in site was
() 25 performend on emergency procedure guidelines, Rev. 3.
ACE-FEDERAL REPORTERS, INC.
M-347-3700 Ntionwide Coverage 804336-6fA6
2290 16 07 186
~
DAV/bc 1 There have been significant changes between 2 Rev. 3 and Rev. 4, and we see it as a benefit to go back a 3 and go through this same type of process which we went 4 through before, to do it on Rev. 4. And make sure that, in 5 terms of severe accident in sites, there's no conflict.
6 This doesn't say that this is the end-all and 7 be-all set of procedures for how you manage a severe 8 accident. It just says that there's nothing in the current 9 emergency procedure guidelines which would conflict with the 10 way you would address a severe accident.
11 (Slide.)
12 As always, a few qualifiers or cautions as we 13 move into the actual elements, as we run through the
(])
14 evaluation. The elements, as you'll see. them presented 15 today, that we evaluated are the elements as they were 16 presented to us on June 16th by Bob. >
17 They cover not only the Mach I's, but the 18 hydrogen igniter issues on for Mach III's. And also, when
! 19 we did our evaluations, we considered the whole set of them 20 for covering the Mach II's.
21 The cost estimates, there was a wide variation in 22 the cost estimates. The cost estimates which were put 23 together for our report were based on nonseismic, non-EO, 24 nonsafety-related, the type of things where, if you're going
() 25 beyond a design basis accident, you.might reasonably be able i ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage ,
. _ _ _ _ . . F**3E-3M___ ._ , 8Q33M646
2290.16 08 187 1 to expect that you wouldn't have to put all of this extra qgfAV/bc 2 baggage on. And that's the way this thing was approached.
3 The evaluations were performed on a small number 4 of plants. I mentioned earlier that, besides the IDCOR work 5 and the owners group work, that there were some individual 6 utilities who contributed resources.
7 Dr. Okrent...
8 DR. MARK: Could I ask? You mentioned that the 9 studies underway include Mach III and igniters. Is there 10 thought being given in that connection to the power supply 11 for the igniters in case of blackout?
12 MR. PICKENS: That is the element on the Mach
() 13 -III's,'is an AC-independent backup power supply for the 14 handful of igniters.
15 DR. OKRENT: A moment ago, you mentioned that, in 16 making estimates of the cost, you did not include 17 considerations of environmental qualification or seismic 18 design, for example.
19 It seems to me that if one had a reasonably full 20 peropective of the sources of, say, core melt frequency and 21 radioactive material release, one might be able to look at a 22 set of modifications, whether they're the ones that Bernero 23 proposes or some other set. And one by one say, well, 24 here's one that doesn't seem to have a seismic connection,
() 25 for example. And there's nothing unusual about the i
ACE-FEDERAL REPORTERS, INC.
_ _ _ _ Nationwide Coverage _ 800-336-6646
. . _. 202-347 3700 _ _ _
2290 16 09 188 1 environment it's going to see.
jggDAV/bc 2 So, if we're going to do it, let's just do it in 3 whatever is a reasonably sufficiently reliable way.
4 The next one might look quite different. So I, 5 myself, would hope that people, in looking at this issue--
6 and I don't want to say holding these specific things 7 Bernero has proposed -- tried to bring with it this broader 8 perspective.
9 And I would hope that, instead of people sort of 10 being at loggerheads on things, that a logic evolves. In 11 some cases, it makes sense to do something. And it's just 12 sort of done by both parties, saying, Gee, that's the right n
( !
13 way to go,. .
14 In other cases, it makes the other point.
15 There's no point making this safety-grade.
16 So, let me at least urge that the BWROG and the 17 NUMARC people try to keep this in mind. And, in the end, 18 we're looking for, if we can, BWR containments that look 19 pretty good.
20 . MR. PICKENS: I agree.
21 As I pointed out, it was nonseismic, non-EO, and 22 that was kind of the basis that Bob gave us to go off and 23 work on. I think you'll see, as we get into the elements, 24 realistically speaking, that's very difficult not to put
(_) 25 that baggage on there. You're interfacing with systems that ACE-FEDERAL REPORTERS, INC.
202 A47-P00 Pationwide Covera ge 800-BW
2290 16.10 189 7
-DAV/bc 1 have seismic requirements, which have safety-related, which 2 have EO recuirements. And for the interface, if you're 3 going to use a system to fulfill two purposes, both a safety 4 function and something beyond design basis, you're going to 5 make it seismic, you're going to make it EO and you're going 6 to make it safety-related anyway.
7 One of the problems or one of the things which I 8 think drove the process up, which we were evaluating, was 9 the fact that you have those interfaces anywhere you go 10 inside the dry well. You're going to end up with seismic, 11 and you're going to end up with EO, because the systems in 12 there are necessary systems already.
(s.
( ,) 13 So you're going to end up doing that already. I 14 think, and Bob can say whether or not he agrees with this, 15 that, primarily, that kind of qualifier applied to things 16 like the fire pump that you might use to provide your backup 17 water source, do we want to put a seismic requirement on a 18 fire pump now as a backup water source or spray?
19 I think the answer was no. It's a backup. So 20 ' those are the areas where we're talking. You don't need to 21 make your fire pump seismic. You don' t need to make it EO.
22 But that's the kind of thing we're talking about.
23 ' MR. EBERSOLE: You're acquainted with, I. guess, 24 the details of Gessar II. There appeared in that design, to r~w
(_) 25 my personal satisfaction, a little system which was ACE-FEDERAL REPORTERS, INC.
REEM7D Nationwide Coverage 800-33MM6 .. l
2290 16 11 190 T^1DAV/bc 1 oriented toward one specific end, continuity of containment 2 fluid.
3 We called it the UPPS system. I saw in it the 4 beginning of what I thought was common sense. To reduce 5 these horrible requirements on all these other systems which 6 are thrown all over the place, and are independent and 7 create monumental problems with OA and everything, the 8 problem that stands in front of us would be different, I 9 think.
10 But there's a unique opportunity for your kind of 11 plant to do some simple things by the relatively simple 12 process of depressurizing but with guaranteed capability as
()
13 you do not now have.
14 And then providing through the pressurization 15 process the privilege of of course going to the core 16 suppression pool, which initially condenses the coolant on 17 the backside to the atmosphere and in the prevent mode.
18 You heard us earlier talk about the proposed 19 release. Then one looks at the source of supply of the 20 water, just let's say barely above the core. You have a 21 two-thirds height core.
22 Anyway, I'm going back to Gessar II. I've been 23 told that immediately af ter scram, even a PWR, not to 24 mention a boiler, if it's supplied with the equivalent of
('T (J 25 just water enough to cover the core, which may be in a two-l t
ACE-FEDERAL REPORTERS, INC.
I ..,m,- m . m c _ ,. mm-
2290 16 12 191
[ ' DAV/bc 1 phased configuration, if you're not going to have core 2 damage -- and this leaves you with the implicit 3 contemplation of the cartoon which I've mentioned frequently 4 that all you need in the supply context to make up what 5 you've evaporated to the atmosphere, including the hot 6 pours, is an old man with a jockey pump and a gauge glass.
7 I'm sure that's about a 500 kilowatt caterpillar diesel 8 pumping out of a well into a concrete building, which 9 probably gets its supply out of the ground.
10 That seems to me to be getting close to 11 simplification, which the chairman mentioned was not in our 12 recent letter.
() 13 -
I think it's important and .I think you have a 14- golden opportunity, which your competitors don' t have.
15 MR. PICKENS: I think there are some who have 16 drawn the parallel between the ultimate plant protection 17 system and having the capability also to pump into the 18 vessel.
19 MR. EBERSOLE: You could put all of the defensive 20 elements in there -- fire, earthquake, floods, even 21 sabotage. It can be done -- without checking the costs.
22 MR. PICKENS: Other elements of our evaluation 23 performed on a small number of plants, we had two plants or 24 two utilities -- Philadelphia Electric and Northern States
! (~'N, l (J 25 Power -- who both did Mach I plant-specific evaluations l
l ACE-FEDERAL REPORTERS, INC.
l l RiR DS7-17 G Natsonwide Coverage 800-336-6646
- 2290 16 13 192
~
1 against Bob's five elements. And Commonwealth Edison did fAV/bc 2 perform them on. LaSalle for a Mach II. So we have a small 3 number of plants who did plant-specific evaluations.
4 That's both good and bad. I wish we would have i 5 had a larger number of plants. But, in the time frame we 6 had, I thought we did pretty well. We had three plants who 7 put it together, so you have to realize these evaluations 8 were done on relatively few plants.
9 The objectives do appear generic. The 10 enhancements appear plant-specific. The capabilities which
~
11 .the plants have right now is all over the place -- in terms 12 of venting, in terms of dry well sprays. Back up to the dry
() 13 well sprays, the whole shot.
j 14 So you have to be' careful there in looking a't the 15 cost estimated. It's going to be somewhere in there.
! 16 17 18 19 20 l 21 22 23 24
() 25 ACE-FEDERAL REPORTERS, INC.
- -_ . . _ . . _ . . _ , ,_ _ _ _ . 202 347,- 3 700 _, , __ Nationwide Coverase _ _ _y336 6646
2290 17 01 193 7-~DAVbw 1 Qualitative assessments of benefits and negative V
, 2 impacts have been performed. By no means did we have time 3 to go in and do a qualitative production assessment on the 4 benefit side or a risk increase assessment on the negative 5 impact side, so that we could look at it, crank it through a 6 source term and say this is how many dollars.
1 7 We did purely a qualitative assessment that might 4
8 allow us to say we think there is benefit, and we don't know 9 of much negative impact. If you put that on a scale, it 10 would tilt toward its benefit, and it would warrant further 11 study.
12' So those are the kind of qualitative assessments
() 13 that have been done to date. ,
14 (Slide.)
15 The first element is on hydrogen control, as I 16 mentioned. The objectives and requirements are pulled right 17 from Bob's slide as he presented them to us on June 16th.
I 18 The objective was to present hydrogen combustion and cause 19 failure. The requirements, either oxygen control by 20 inerting. That is the MKls and MK2s. Or hydrogen control i
21 on MK3s and that is by igniters and by burning. The IDCOR j 22 Owners Group evaluations, the options control by nitrogen 23 inerting is adequate for the MKis and MK2s; however, we did, 24 through talking with Bob, decide that it would be worth
() 25 taking a look at limiting the time de-inerted, as you shut
, ACE-FEDERAL REPORTERS, INC.
l s.<>e.. . ce .< .. eoo>34-ee.e
- .~ - 20*>413700
2290 17 02 194 DAVbw 1 down.
2 We are going to take that under review right 3 now.
4 MR. EBERSOLE: Let me ask something about kind of 5 a philosophy. I am astounded at the big flap that occurs 6 when somebody suggests that the few, very few kilowatts you 7 need to make the igniters work seem to be enormously. In 8 fact, you could do it with a gasoline engine generator, 9 which you can buy for S300 or $400.
10 DR. KERR: Seismically qualified, OA'd?
11 MR. EBERSOLE: They are designed to ride on 12 trailers on streets. But no, we don't do all this. You can
() 13 do it if you want to.
14 DR. KERR: I am simply saying, the cost 15 estimates.
16 MR. EBERSOLE: We are talking about a power 17 supply, which normally is geared to shocks far in excess of 4
18 G value. For the life of me, I can't see why it raisos a 19 big flap to find power for these few kilowatts. Virtually 20 nothing.
21 Could you explain why that occurs?
22 MR. PICKENS: I will get to it in the last
. I 23 bullet. The MK3 hydrogen controls is being addressed by 24 HCOG. HCOG is not here today. I have the slides which they
() 25 used and presented on that day. What they are addressing ACE-FEDERAL REPORTERS, INC.
.. _ ME*3M _- __ Nationwide Coverage _
80 4 336-6646
2290 17 03 195 1 is a test, some of the costs associated with it.
ygpAVbw 2 Maybe I will just jump a couple of slides back on 3 this one.
4 (Slide.)
5 .Under their proposal, HCOG is to supply the 6 design criteria for a back-up power supply for the hydrogen 7 igniters. A back-up power supply need not be 8 safety-related.
9 One of the things that they will be undergoing is 10 a test program as part of their test program right now to 11 identify what the configuration and how many igniters you 12 need to put on that~back-up power supply are.
() 13 MR. EBERSOLE: Do you have an estimate of the 14 kilowatt demand?
15 MR. PICKENS: No, I don't. Bob?
16 MR. BERNERO: The present igniter system on MK3 17 has about 100 igniters, each of which is 100 watts and the 18 expectation is that the test demonstration might show four, 19 five, maybe six would be an abundant or sufficient.
20 MR. EBERSOLE: I am reminded of the time about 21 the argument as to how much wood and how much concrete. I 22 once worked on a power plant, and the owner didn't want to 23 make them too wide, even though he would save concrete. He 24 didn't want to be bound by regulatory constraints.
) 25 MR. PICKENS: I think if you talk to the HCOG ACE-FEDERAL REPORTERS, INC.
202-347 3700_ _ __ _ Nationwide, Coverage _ _ _ _ 804336&M _ __
2290 17 04 196 DAVbw 1 people, and I've done a little bit, where they think that 2 some of the costs come in is on the interfacing end with the 3 safety-related system, and you can' t just take a little 4 battery, and you can't just plug it in there next to the 5 igniter or something like that, but that you are going to 6 have to maintain your Class 1E, and you are going to have to 7 --
if you just wa::t to pull up a generator or something 8 outside, you've got to run cabling over there, and you're 9 going to have to qualify that cabling.
10 That is the kind of things they are getting into, 11 but you know they have not put together a cost estimate. As 12 you can see from this slide here, they are going to identify 13 what the design is and then, on an individual basi,s, will
- (])
14 take a look at that design criteria and id.entify the 15 associated costs.
16 What they will do for the back-up power supply, 17 and then they will meet with the NRC and go through the 18 cost-benefit analysis, but they are not to that stage yet.
19 MR. EBERSOLE: The cable and conductors are 20 already okay. All you need to do is get something to put 21 into them.
22 MR. PICKENS: You have to have something to put 23 ' into them, but where you interface or tie into those cables 24 and things that are already there, the interface problems
() 25 that you have to work on, it is not as easy.
ACE-FEDERAL REPORTERS, INC.
_ , - _ _ NM 3E _.. _ _ pjonwide Coverage 804336-646
- 1 _ _ _
.2290 17 05 197 1 MR. BERNERO: The existing system is non-Class gggDAVbw 2 lE. _The implication is that they would tie into a Class lE 3 battery. That is where you get the interface requirement.
4 MR. EBERSOLE: Well, I am not going to do a lot 5 of that.
6 MR. PICKENS: Maybe we can talk HCOG into coming 7 in next time.
8 So on the hydrogen, back at the slide I had 9 previously, we are evaluating time de-inerted. This is the 10 Hydrogen Control Owners Group in the response that they put 11 together to Bob.
12 MR. EBERSOLE: That inerting system is not
() 13 carried at any differential pressure.
14 MR. PICKENS: No.
15 MR. EBERSOLE: Just a little bit positive.
16 (Slide.)
17 MR. PICKENS: The second. element is drywell 18 sprays. The objectives under this one are to quench debris, 19 scrub aerosols, lower your pressure and cool vulnerable 20 equipment.
21 The first time Bob showed this to us, he did not 22 put the words primary, secondary, secondary. That was 23 something that was added on by us.
24 In reviewing it, I guess we feel that the primary 25 purpose of the drywell sprays and the reason that you want ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336-6646
2290 17 06 198 DAVbw 1 to get that back to power supply is for the core debris 2 coolability, quenching your debris. The fact that you 3 scrub aerosols, lower pressure or cool vulnerable equipment 4 are, in fact, secondary, in our opinion, in terms of what 5 you are trying to accomplish.
6 The requirements are having a spray in the 7 drywell and back-up water sources and pumps.
8 The suggested use is that they were either by use 9 of the fire hose connection or cross tying out to your fire 10 main, out to your screen house, something like that. The 11 IDCOR Owners Group evaluations Bob mentioned before.
12 The typical spray capacity of an individual
() 13 header is designed right now to handle 5000 to 10,000 gpm 14 per head, which is a pretty significant flow.
15 MR. EBERSOLE: Bob said fix the pump rather than 16 the requirement.
17 MR. PICKENS: That is right. And in fact, when 18 we went back and dug back into some of the design basis on 19 this thing, you cannot find a lot of clear definition as to 20 what the design requirements were. It's as though we had 21 the RHR pump sitting there. Therefore, they pump 6000.
22 Let's design a header. Take the flow from those pumps and 23 put it up there.
24 MR. EBERSOLE: Regarding the split flow logic, I c () 25 recall distinctly finding out that you had to split the l
l l
ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800-336 6646
__. ._ _M-34_7 3700 _ _
2290 17 07 199 7'NDAVbw 1 flow to get enough mass flow to containment cooling,
(_)
2 especially cooling while you provided whatever you needed 3 for the core. That required a curious arrangement, which I 4 must say, I always didn't like.
5 What other configurations are in existence in the 6 boilers today?
7 MR. PICKENS: How we would split it?
8 MR. EBERSOLE: You've got to cover the core, but 9 you've got to get mass flow.
10 MR. PICKENS: We were kind of kicking that I around, while Bob was up here talking, in the back of the 11 12 room.
() - 13. I think what you are looking at is probably you 14 would take one RHR pump and put it to the drywall spray'. s 15 You take the other RHR pump, and you put it to the 16 suppression pool cooling; however, if you only have one RHR 17 pump, I don' t think right now that we have the plant 18 capability to split the flow of that one RHR pump.
j 19 MR. EBERSOLE: You can do it at one plant.
l l 20 MR. PICKENS: We might at one plant. I was only l
21 speaking for the experience at our plant. We would not have 22 that kind of capability to put a 10-90 split or control it 23 to that degree.
l The concepts considered. We did consider 24
() 25 connection to a hose station in the reactor building. The ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80433MM6 l - - __ _ _ __
2290 17 08 200 DAVbw 1 approximate flow that would be provided by that hose 2 connection spray would be 200 gallons per minute.
3 I don't think 'to anybody's surprise, this does 4 not provide the spray out of your drywall spray hitter. It 5 promotes, at best, a dribble.
6 We had some great testing done. We happened to 7 have a spare nozzle sitting around at Monticello. NuTech 8 was doing some work for us and took it back to California 9 and hooked it up to a hose and put some calibrated flow f
10 rates through it. We had 100 drywell spray nozzles on each 11 header, so we ran two, four, eight to see what 200, 400, 800 12 would be and would look like through this thing.
13 I can't say that even approaching the 800 gpm
(])
14 ones, that we were getting anything that even closely 15 resembled a spray.
16 So by the flow rates that we are looking at here, 17 you would not be providing spray. And again, I jump back up 18 to the secondary purpose. The primary purpose. You don't i
19 need a spray, necessarily, for the quenching debris. You i
20 need it for scrubbing the aerosols and things like that. So 21 then it now becomes another question of, do you want to 22 change your nozzles out, plug some nozzles, do some things, j 23 so that you get a spray with these lower flow rates, so that 24 you can accomplish the secondary ones. And that is just
() 25 another one that gets plugged into your costs.
ACE-FEDERAL REPORTERS, INC.
- .__ a m _ _ ~ ~ ec - w J m . - .
.- -= . ..
'2290 17 09 201 DAVbw 1 (Slide.)
2 The next concept that was considered was a
. 3 cross tie from your diesel fire pump to your RHR out in 4 screen house. We picked an eight-inch line cross tie. The
- 5 approximate flow provided would be about half of your fire 6 pump rating. And in the case of Monticello, that turned out
]
7 to be where you might be getting about 750 gpm pumped 8 through. Again, that does not give you a spray. In order 4
9 to achieve a spray, even.with the higher flow rate, not the j 10 250, but the higher one with the eight-inch cross tie, you
- 11 would have to.close off approximately 70 percent of the
, 12 nozzles that you currently have on a drywell spray header i O 13 to achieve that spray.
14 DR. WYLIE: And you say that is adequate?
15 MR. PICKENS: The flow rate is identified in one i
! 16 of the things we talked about when he presented his i 17 presentation and said about 10 percent of the rate flow i
l 18 would be the number of something like that'. I think we are 19 at the point where we think it would be better to define 20 perhaps a decay heat removal capability some point into the 21 accident, some power level that you want to look at.
f I
22 The flow rates which we have ider.tified, 250 or ,
23 750, if you get that kind of water on the floor of the t
. 24 containment with the debris coming in, that appears adequate O 25 to hand 1e, whae, 1 percene decar heae remove 1 capabi11ev 1
1 ACE-FEDERAL REPORTERS, INC.
- _ _ - _ _ =m _.a-*cm _ _ m ___
I 2290 17 10 202 DAVbw 1 or something? I think 1 percent was the number we picked.
2 So all we are saying about that statement is that
! 3 that flow rate is adequate to pull off a 1 percent power 4, level decay heat removal.
DR. KERR: What were the numbers you gave? How
[ i 5 6 many gallons per minute?
7 MR. PICKENS: 250 on the fire hose connection and 8 750.
9 DR. KERR: As I remember -- let's see, one 10 minute. Those two, I don't think -- the ten minutes 11 reminded me of it.
12 MR. PICKENS: Some of that gets into, when do you
() 13 initiate it? How long do you have some water on the floor 14 before the core gets down? Those kind of things.
15 MR. EBERSOLE: Would that handle the hypothesis?
16 Are you talking about all BWRs? Is it MK3s, MK2s?
! 17 MR. PICKENS: The particular evaluation we did, i 18 our plant is a MK1, so we know this is good for the MKl.
19 MR. EBERSOLE: The MK1 has some potential, as 20 MK2s and MK3s don't have for suppression bypass. If you 21 merely stick a safety valve in the pipe and close it off, 22 rather a safety spray, or whatever, because if the pipe 23 comes lose, you bypass the suppression pool and lay down a 24 lot of hot water on top of the suppression pool. I think
() 25 there is a rather vague claim that the spray at its ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 33Ms646
2290 17 11 203 E mDAVbw 1 present rate will let out hot water; is that correct?
U 2 MR. PICKENS: I think Bob was talking about the ;
, 3 spraying of the aerosols. !
i r
- 4 MR. EBERSOLE
- I am talking about the cooling 5 pump to cope with the suppression bypass. -
6 What happens then if you build un pressure in the j 7 wet well, it runs to the drywell, and it sprays it a certain l 8 way. You've got a tuck -- I think they are eight inches. i r
9 What are the sizes, ten-inch? They are big. But i
j 10 if they stay stuck open, and they start rumbling at a low l
- 11 pressure, and they come down whatever problem, I think the i
12 parachute is the spray system. You are cutting into it. Or l
l
() 13 is it not worth much? -
14 MR. PICKENS: That is one of the things we ,
15 identified, that we have to go back and take a look at. You 16 are talking about the existing design requirements for the 17 drywell sprays, and it is 10 percent of the capacity.
18 Is that adequate to handle all the different 19 things we are using?
20 MR. EBERSOLE: That particular pump is a dark 21 corner.
22 MR. PICKENS: That is what we have been looking 23 for. It appears that whatever the design basis is that it 24 is stuck in a dark corner somewhere. Debris quenching does
() 25 not require a spray. I talked about that. j ACE-FEDERAL REPORTERS, INC.
I 202 Nationwide Coverage 800 33M646 .>
. _ _ , _ _ _ _ . . __. . _ _ . __ _ _ _ _ __ __ -347-3700_..._._ ._-~____ _... __-._ _ __________ _._ _ . _ _ _ _ _ . _ _ _ . .
2290 17 12 204 DAVbw 1, Potential benefit risk on a scale that I talked 2 about says that it warrants further study.
3 When we looked at this too, as we went through 4 our estimates, one of the things which we included, which 5 has changed a little bit is putting in DC power operated 6 valve capability rather than using a simple procedural 7 step to handle the station blanket capability, if you are in 8 a station blackout situation.
9 So some of the cost estimates are based on that, 10 putting in DC power operated, because that would cut down 11 the costs.
12 DR. KERR: I am sorry. Did you say that it would
(}
- 13 cut down the cost?
14 MR. PICKENS: That would make it much more j 15 expensive, if we are allowed to use the simple procedural 16 step to go in and open them. That would cut the cost down 17 quite a bit.
18 DR. KERR: Thank you.
19 MR. EBERSOLE: I guess I never really understand 20 why, but there was a logic way back that you would use the 21 DC valves in series with AC valves to get some diversity on 22 the pipes, and you had a practice, which I don' t know how 23 far it went, that you never put the DC valves inside a 24 hostile environment. You would always use three-phase AC l
() 25 motors.
i ACE-FEDERAL REPORTERS, INC. l 202-347-3700 Nationwide Coverage 80433&6M6
2290 17 13 205 1 AVbw 1 Is that still with you?
2 MR. PICKENS: I don't think so 3 Dick, do you know? Do we still keep the DC power 4 operated valves out of bad environments?
5 VOICE: In general, yes. Like the HPCI and RCIC 6 steam line valves.
7 (Slide.)
8 MR. PICKENS: Number 3. Pressure control.
9 The objectives were to avert an uncontrolled over 10 pressure failure and controlled over least path by scrubbing 11 through the suppression pool.
12 Requirements. Substantial capability to vent the Q 13 wet well, remote reliable control other than valve, and 14 again, we interpreted that as being DC power operated '
15 valves, not necessarily simply reliable steps.
16 And then the ability to reclose the vent was a 17 requirement which Bob had on there at that time.
18 The concepts considered, we considered 19 upgrading. Some plants had two psi ducting as part of their 20 vent path. We considered upgrading that ducting and the 21 standby gas treatment system to containment design pressure 22 capability.
23 It is possible to upgrade the ducting, when you 24 talk about the actual standby gas treatment system. There 25 you run into real problems, in terms of trying to upgrade ACE-FEbERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
2290 17 14 206
?-wDAVbw 1 the pressure capability of that system, and we decided that V
2 it wasn't feasible to try and upgrade the standby gas 3 treatment system. So we rather looked at a hard pipe bypass 4 around the standby gas treatment system.
5 MR. EBERSOLE: It is interesting to me that you 6 work in isolation on the vent without associating it with
, 7 the core cooling function.
8 It is kind of an integral problem, and I note 9 that at least in GESSAR 1, that vent, in fact, was a vent 10 path for the energy for the decay heat, low pressure 11 and moderate pressure.
12 And that it is an integral problem that you can 13 package up, not a bunch of loose things all over the place,
(]) ,
14 if patched together in typical style.
15 MR. PICKENS: I think Bob has stressed to us, and 16 I think we all believe, too, that when you talk about the 17 drywell sprays and venting, that they are integral things 18 that you need to consider.
19 Without both of those, you are not really taking 20 care of the containment problem.
21 l
22 i
23 24
() 25 l ACE-FEDERAL REPORTERS, INC.
202 -347-3700 Nationwide Coverage 800 336-6646 I _ _
2290 '3 01 207
- DAVbur 1 MR. EBERSOLE: I am talking about pressurized 2 core cooling.
3 MR. PICKENS: We look to the hard pipe bypass 4 around the standby gas treatment system using something like 5 a Schedule 20 stainless steel pipe. Some of the problems 6 that you get into in trying to put a bypass around is 7 actually tying in and interfacing with your existing vent 8 system and trying to get into some of the control logic for 9 which event path are you trying to use it this time, 10 switching over to the other vent path, 11 Again, we considered DC power operated valves in 12 putting the costs together for this one, so that if you had 13 the desire after you had gone through something like an ATWS
(})
14 steam vent or something where you are venting mostly st'eam 15 and you got down to some of the lower flow rate type venting 16 situations, you might want to put it through the standby gas 17 treatment system to pick up the benefit of that.
18 But some of the problems that you get into with 19 the interface between an existing system and just trying to 20 put a bypass around it drove some of the costs up and made 21 it fairly difficult.
22 So we looked at what would it also take to hard 23 piping dedicated emergency vents, if you will, that might 24 come off of the wet well air space, go right up the reactor
, O)
(_ 25 building and out the roof of the reactor building and ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
2290 18 02 208 1 AVbur 1 activate event there.
2 MR. EBERSOLE: Were you waiting for an elevated 3 release? You don't always have stacks, do you?
4 MR. PICKENS: The stacks vary. I was talking to 5 Dick Diderich back here, and I said how much higher is your 6 stack, the roof of your building? He said, it is up on a 7 hill; it is about three times higher than the reactor 8 building.
9 But at Monticello, out where it is nice and flat, 10 our stack is on the same elevation as the reactor building, 11 and my observation is our stack is not that much higher to 12 start with from the reactor building.
13 MR. EBERSOLE: It is interesting that nobody
(])
14 looks to the right or left.
15 MR. PICKENS: Looking at it out the reactor 16 building roof, they say it is not going to be as elevated, 17 and that is probably true for some plants.
18 MR. EBERSOLE: Would you call that satisfactory?
19 As I recall, the old plant I used to work on -- ,
20 MR. PICKENS: We did not get into assessing l 21 whether or not what was there was adequate in terms of the 22 stack.
23 Interestingly enough, because you drop out some 24 of the interfaces with the existing system, some of the l () 25 control logic and things like that, the cost estimates ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 4 686
2290 18 03 209 1r'SAVbur 1 which we came up with, even considering the pluses and V
2 minuses, it is not quite as high, but you don' t have the 3 interfaces and those kind of things. The cost estimate for 4 both hard pipe systems turned out very close to the same. I 5 think there is a lot of people who say that they would 6 rather just go to a very simple dedicated emergency vent 7 rather than getting into all the control and logic problems 8 and things like that. The costs turned out about the same 9 on this.
10 (Slide.)
11 In terms of sizing the vent or what is an 12 adequate vent size, that is something that is currently
() 13 under review. It has been under review, and there has been 14 a lot of discussion going back and forth on what we ought to 15 be doing on this. There are two considerations, sizing it 16 probably either for a worst case ATWS scenario and MSIV 17 closure, the whole shop, or sizing it for decay heat 18 removal.
19 Discussions about do you really want to vent for 20 an ATWS scenario, do you need to vent for an ATWS scenario.
21 There has been a lot of talk and work done by the New York 22 Power Authority about venting under an ATWS situation, l
23 arguments saying they don' t think you should vent, and I l 24 have recently heard some other discussions about, you know,
() 25 if you have one RHR heat exchanger and you can pull the heat ACE-FEDERAL REPORTERS, INC.
202-347-3700 Natiorrvide Coverage 800-336-6646
2290 18 04 210 13DAVbur 1 away, they think that you can by lowering the water level V 2 inside the core during an ATWS scenario, getting your power 3 level down, that you will be able to pull out an adequate 4 amount of decay heat via one heat exchanger and things like 5 that and you won' t get into the need for venting.
6 Then it becomes a juggling game as to how you are 7 going to size those. I would not purport to say that we are 8 ready to answer that question yet, but there is a lot of 9 work going on right now.
10 MR. EBERSOLE: I wonder if the work doesn't cost 11 50 times as much as the difference in cost -- or maybe that 12 is the wrong way to go. Maybe it is a problem.
13 MR. PICKENS: I don't know. I have heard numbers f')T x
14 tossed out for an ATWS sized vent all the way out to 3F 15 inches in diameter. It is going to cost more, but I think 16 probably we have talked about different ways to do it, 17 playing out the probabilities of ATWS scenarios and looking 18 at how many you can cover by a one-time decay heat removal 19 sized vent to a two-times decay heat removal sized vent, 20 I just doing something to provide some kind of basis. But 21 that is under review.
22 DR. KERR: What happens if you just open up the 23 main steam isolation valve and feed the condenser?
24 MR. PICKENS: Your problem would go away. We (O) 25 could avoid the MSIV closure and go to the condenser.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80433MM6
2290 18 05 211 DAVbur 1 MR. EBERSOLE: There is a small problem. You 2 will blow the condenser. Of course, more practical is a 3 bypass around the main steam isolation valve, which they 4 needed in the Browns Ferry fire very badly, and they got 5 one. It was put in there for equalization.
6 DR. KERR: I don't see why you need a bypass.
7 Why not just open it?
8 MR. EBERSOLE: They have pressure on them on the O.
9 front side.
10 MR. PICKENS: If they are shut they stay shut is 11 what you are saying.
I 12 MR. EBERSOLE: They lock up. That is what saved
() 13 the day.
14 DR. KERR: Once they are close'd, you can't o'en p 15 them.
16 MR. EBERSOLE: There is pressure on the front l 17 side, am I correct?
l l 18 MR. DIDERICH: You have to equalize the 19 pressure.
20 MR. EBERSOLE: That little line kept Browns Ferry 21 alive in the big fire.
r 22 DR. KERR: It doesn' t take much to equalize l
23 pressure.
I 24 MR. EBERSOLE: Of course, the fill bins and the
()
I 25 stop valves control and you just fill it up. It is an ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800 336
. . - . - , . 202.347-3700
._ . _ _ ___...__.._.___MM. . . _ . _ . . _ _ _ . . _ . . . . . . _
2290 18 06 212 1-sDAVbur 1 interesting opportunity. But I think you make a good U 2 point.
3 DR. OKRENT: But the condenser is only there for 4 similar scenarios.
5 MR. EBERSOLE: It is gone.
6 DR. KERR: No, you don't necessarily have core 7 damage when this thing starts.
8 DR. OKRENT: No, but the condenser goes.
9 DR. KERR: But not with all of them.
10 MR. EBERSOLE: That is the AC power failure 11 case.
12 DR. KERR: But you don't necessarily have to have
() 13 AC power failure just because you have ATWS.
14 MR. EBERSOLE: I think it doesn' t matter. You 15 just blow the relief diaphragm and there she goes. It is 16 not a clean way to do it.
17 MR. PICKENS: Negative impacts needing further 18 review.
19 Secondary containment contamination. Bob has put 20 in at one time -- he talked to us -- is it acceptable to 21 burst your duct inside secondary containment. Now he has 22 got it in his desirable category to have burst resistant 23 ducting on your standby gas treatment system and on your 24 vent path, i
i
() 25 I think that needs to be evaluated, especially l
I
- ACE-FEDERAL REPORTERS, INC.
3)2-347-3700 Nationwide Coverage 804336-6646
2290 18 07 213 1 1 from the utility point of view in terms of recovering from a fAVbur 2 severe accident, whether or not we want to have these 3 secondary containments all wrapped up.
4 MR. EBERSOLE: It isn't really that. It is 5 whether the machinery will live through the process and the 6 operators will. That gets into a design problem.
7 I recall the design engineers said what are we 8 going to do if I hypothesize all the leakage is going to go 9 into the machinery rooms next to the control rooms in the 10 context of defining the source term for shielding.
11 If I remember correctly, we used WASH-1400 12 estimates of containment activity, the total standard leak f^)
v 13 rate multiplied by something that came into the machinery 14 hall and what size of machinery could live through the ATWS 15 and the radiation. The operators had to be shielded.
16 When that drifts off to the community, you lock 17 it up, that makes a dif ferent picture.
18 MR. PICKENS: It makes it very difficult to get 19 in there and take any recovery type actions. So I am not 20 sure that if we are going to say we have venting capability l
l 21 that putting the burst resistant ducting in the desirable 22 category is the way to go.
l 23 MR. EBERSOLE: Incidentally, that brought up the 24 fact that you had four drains for cooling in those old (j'N
( 25 machines, the RHRs. In the new ones you have got two.
l l ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6 4 6
2290 18 08 214 DAVbur 1 I am talking about for the long duration haul.
2 You are not in a pretty picture.
3 MR. PICKENS: The other negative impact needing 4 further review would be deliberate release out the stack.
5 In terms of what we are doing, I don't know if that has been 6 clearly --
7 MR. EBERSOLE: How do you feel about Bob's 8 opinion that if there is an X release that you close and ,
9 never open?
10 were you here when we had the discussion? What 11 is the owners group opinion?
12 He has got quite a machine there. I recall using
() 13 the S3 billion number thlt Dave used just to get you to trip 14 the pumps in ATWS.
15 But anyway, 3 billion is not a bad number.
16 What is your view about an activity release 17 because you are going to prevent a catastrophic release that 18 might occur a day or two later?
19 MR. PICKENS: I guess I really can't say that I 20 have an owners group point of view or a NUMARC point of view 21 to give you, but in terms of looking at, you know, a little 22 bit of release to prevent catastrophic, it seems to make 23 sense to me.
24 MR. EBERSOLE: We don't have that balance in the l () 25 regulatory process yet, and I think again we need a rule ACE-FEDERAL REPORTERS, INC.
202-347-3700 _ , _ Nationwide Coverage 800 336 4 646
2290 18 09 215 h DAVbur 1 curve.
(_)
2 DR. SIESS: It seems to me you have two 3 possibilities there. One is a deliberate release versus no 4 release and a deliberate release versus subsequent 5 catastrophic release.
6 Do yu have any idea now when you are in Condition 7 A and when you are in Condition B or what you have to know?
8 MR. PICKENS: I know there's different scenarios 9' where you are now into the different releases. Venting for 10 an ATWS scenario e if you did get into it, you are in an ATWS 11 scenario, you are venting before you have core damage.
12 DR. SIE,SS: This implies that you have some
() 13 alternative. If the alternative is no release, then it 14 doesn't seem to me that is much of a decision.
15 Alternatively, if there is a catastrophic release, it 16 doesn't sound like that is inuch of a decision either.
17 Now, is there a third alternative I don't know .
18 of?
19 DR. OKRENT: Yes, there sure is. Of course, we l 20 may not know with certainty if you don't release that you l
21 are going to have a catastrophic release.
22 DR. SIESS: That is not the third one. That is 23 just saying you don' t know which of the two it is.
24 DR. OKRENT: You may be in a position where all
() 25 you have is some probability, and also you may be in a ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336 486
2290 18 10 216 P' DAVbur 1 position where what you release is not radically different
( 2 from what would be released given a burst release except you 3 would be doing it spread out over time.
4 So there are ranges of situations.
5 DR. SIESS: How can you change that?
6 DR. KERR: It seems to me if you release you are 7 always in a situation where you don' t know whether you would 8 have a release if you waited.
9 MR. EBERSOLE: This is why I say it takes a 10 multi-dimensional chart or matrix.
11 MR. PICKENS: It takes a lot to understand all of 12 the different venting scenarios.
() 13 MR. EBERSOLE: Put yourself -- you are an 14 operator-on the board and you face this thing. To whom'do 15 you turn for instruction?
16 At TMI the engineers never told the operators.
17 DR. OKRENT: Bernero said he turned to Gabriel 18 every night.
19 MR. EBERSOLE: I guess that is what the operators 20 in BWRs have to do today.
21 MR. PICKENS: And I guess I would contend I am 22 not sure that we could develop a procedure that would give 23 him the definitive answer every time as to what to do, and 24 what we are going to do in these kind of sequences and the
() 25 purpose of the emergency operating facilities and the tech ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336. % s6
2290 18 11 217
- 1 support centers is to give him the support that he needs and
()'DAVbur 2 perhaps to give those groups the guidance as to how to make 3 that decision is what we are really looking to.
4 MR. EBERSOLE: Oh, yes, I agree.
5 DR. SIESS: Does the decision have to be made 6 early enough that it is the operator that makes it?
7 MR. PICKENS: I think we would have enough time 8 that we would not have to have the operator at the board 9 making that decision on a short term basis.
10 DR. SIESS: At what level would it be? In 11 Bethesda?
12 MR. PICKENS: Realistically, I think that the
() 13 governor, the NRC, and everybody else that is going to be 14 sitting out there interested in the accident i.s going to 15 have a say-so.
16 DR. SIESS: All of them are going to have to make 17 up their mind. Then it will be too late.
18 MR. EBERSOLE: In that context, that is why the 19 Browns Ferry fire went on indefinitely because there was no l
20 one there with whatever it took to say this is my plant, I l
l 21 have got to do what is the best thing to do for it.
l 22 He started a committee action, and you know what l
l l
23 it cost. It was just infinite, virtually. That was the l
24 worst system possible, to throw it to a committee. I think
() 25 you are going entirely the wrong way when you just did what ACE-FEDERAL REPORTERS, INC.
- 202-347 3700 Nationwide Coverage 800 336-M86
2290 18 12 218 1 DAVbur 1. you say.
2 DR. MARK: Give the operator a guarantee of a f 3 phone line to Bernero.
! 4 MR. EBERSOLE: Really, an operator is like a i
I
! 5 submarine commander. He has to do what they say in the face i
i 6 of something dreadful. ,
- 7 DR. OKRENT
- I am interested in understanding
. 8 today if you think there exists within each utility enough I
r 9 detailed knowledge of phenomena involved in this range of ,
1, 10 severe core accidents that we have been talking about from ;
{ '
i l 11 within the utility that'one would be able to make relatively i
12 robust estimates of the course of the event, given that we 1
() 13 do vent or we don't vent, and so forth.
i 14 Are there any utilities you think that meet ~that?
15 Do all of.them meet that capability?
f 16 MR. PICKENS: Well, if you talked to the members
[
17 of the Emergency Procedures Guidelines Committee and some of 18 the people who have gotten into this venting issue and i
19 developed it, they feel that they have taken -- with the 20 existing plant capability and the amount of information they 21 had, they have taken their best shot at it.
l 22 Whether or not that is the definitive answer to l
I
~
23 venting and whether or not we vent or when you vent and when j 24 you don' t vent and when you go back internally then to the
() 25 utilities, apart from these guys who represent three i
ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 80433M686
2290 18 13 219 1 utilities on the committees, I know that there is probably a
{ezDAVbur V
2 lot of internal discussions that go on as to whethe or not 3 they agree with what has been put down in the emergency 4 procedure guidelines and what is being put forward in the 5 emergency operating procedures.
6 I think that is why Bob has seen today a little 7 bit -- quite a bit of variance in the way people are 8 implementing the emergency procedure guidelines. So he is 9 saying to you put your best effort forward to identify what 10 you think is the underlying strategy in the emergency 11 procedure guidelines and follow it.
12 So we have really been striving to achieve that.
13 I don' t know if I would say that we are at the end.
14 DR. OKRENT: I guess my question wasn't clea~r, 15 because really it is not so much aimed at being able to 16 determine the emergency procedure guidelines, which I think 17 can be done with a lesser detailed phenomenological 18 knowledge of what goes on in the vessel, ex-vessel, and so 19 forth.
20 MR. PICKENS: I think the Emergency Procedure 21 Guideline Committee would be quick to caution anybody, too, 22 that those were intended, as their scope and charter is to 23 look at the preventative side of core damage. So if you 24 talked about venting in the emergency procedure guidelines, O 25 you would be talking preventing core damage, not once you ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
2290 18 14 220 3-mDAVbur 1 have achieved core damage.
'U 2 They would probably say they would not take 3 responsibility for venting, for what you are doing with your 4 venting actions there.
l 5 There I think further work is needed, and we need l 6 to understand it. We need to understand the releases. We l
I' 7 need to make these judgments before we can say here is the j i
l 8 guidance we are doing. !
i i
! 9 ;
1 l 10 i i !
l 11 l
12 l
lO t3
~
. 14 15 L 16 l
17 !
i 18 l 19 i i
l 20 l 21 ,
22 23 24 O
V 25 i
l l
k ACE-FEL'ERAL REPORTERS, INC. j M2-347 3700 Nationwide Coverage 800-336-66 6 t
7 .-
2290 19 01 221 DAVbw 1 DR. OKRENT: Fair enough.
2 MR. EBERSOLE: Would you agree that you would l
l 3 have to adhere to this age-old concept called'" unity of l 4 command"? There is no escaping that. As in the fire case, 5 I wouldn't be surprised to hear it involved auditors and 6 shareholders before they got through with it.
7 MR. PICKENS: Our insurance company is getting 8 involved, even in our emergency plan right now. Okay.
9 (Slide.)
10 Element 4, core debris barriers.
11 Reduce the likelihood of failure by direct attack 12 on your drywell liner.
(j 13 Re quireme nts . Use practical debris retarding 14 barriers and conserve suppression pool water as a quenc'hing 15 pool.
16 The two aspects of this one, one, would be debris 17 barriers up inside your drywell. The second one is the one 18 which Bob talked about a little bit, which is actually 19 conserving your suppression pool water down in the torus 20 room, if you have failure of your suppression pool.
I 21 In the drywell, the concepts which we consider --
22 I'll move that up a little bit.
23 The concepts which were considered were a plug in 24 the pedestal opening. This plug would be capable of hol6ing r^s
(,/' 25 the inventory of the core inside the pedestal. The second a
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4646
2290 19 02 222 1 AVbw 1 one that was looked at to achieve the same type of purpose 2 would be to increase the sump size inside the pedestal, 3 again, holding the inventory inside the pedestal.
4 Concerns identified with that type of concept 5 are, particularly, if you use something like the plug in the 6 pedestal, you are holding a very deep pool of this coring 7 material. You are not really letting it spread out. You 8 are not letting it'give its heat off. You are not letting 9 it cool. In addition, you are preventing the water from 10 your reliable drywell sprays, if you put a hole in your 11 pedestal opening, from reaching down inside to the pedestal 12 and actually providing any debris quenching.
- l 13 .Other concepts which were considered were a curb L-14 at.the drywell liner, floor junction out at the wall,- where l
15 the drywell floor goes over an intersects with the drywell 16 liner, the curb might be seven, ten inches, twelve inches i
17 tall, some height that would be capable of when you spread I
j 18 the core out over the entire floor of the drywell of keeping 19 it from interacting with the drywell liner.
20 The other thing that was considered, and this is 21 in kind of promoting an even distribution of the coring 22 material was to put in additional openings in the pedestal.
23 Bob described that there is a door which goes 24 through the pedestal in underneath where you can control (m 25 rods and things like that. That provides one exist for t, _)
l l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 i
2290 19 03 223 1 1 the coring material, if you assume it is going to become fAVbw 2 whooshing out in one big bunch. It comes out that one exit 3 door and goes over to the liner, yes.
4 DR. SIESS: Is this covering all MKs or just MK1 5 and MK2?
6 MR. PICKENS: This is covering -- these types of 7 fixes here would be the MKls.
8 Looking at the configurations of the MK2s, the 9 MK2s are pretty well covered, in terms of, they are~ going to 10 direct the corium to go down into the water inside the 11 suppression pool. This isn't a question on MK3s. This is
, 12 primarily a MK1 issue.
f f
fv ') 13 MR. EBERSOLE: May I ask you to just consider 14 something that bothers me.
15 I heard you talk about invoking new spray sources 16 of water, the fire pumps, and so forth.
l 17. Where is the suction point? What limits the rise l
l 18 of water, wherever the water is going over cycle? It is a L
19 little bit like the fact that you have to have a LOCA in a l 20 PWR to make that work. And they don't have anywhere to 21 create it in, because they don' t have any, you can call them
)
l 22 SARs. This is the other case.
23 What is going to happen to the water, as it fills 24 and fills and fills? I recall a lengthy argument about l r~
i (_h) 25 containment. The drywell should ba designed so it won't l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M646
2290 19 04 224 7 sDAVbw 1 burst out like a wet paper bag, but pulverize the level U 2 right clear up to the top.
3 MR. PICKENS: I think that is one of the reasons 4 you see cutting back on the drywell spray. If you put in a 5 back-up system that is capable, when you pull out the RHR, 6 you are pulling out the torus, and you are pulling it 7 through the drywell spray. But when you get to a backup 8 water source, and you're going to an open cycle, you don't 9 want those kind of flow rates going through your drywell 10 sprays, because you are going to run into problems of 11 filling up your torus, making it solid and starting to fill 12 up the d.rywell.
0 f,,_/'? 13 MR. EBERSOLE: The detailed design of the old 14 Browns Ferry plant required that this not fail as you filled 15 it. It retained its spherical shape. You could take it 16 clear to the top, and you found a hole through the LOCA to 17 go. You could fill the whole bay.
18 MR. PICKENS: Drywells were designed to be fil' led 19 all the way up to the top. They are designed to go solid.
20 You can pump them full.
21 MR. EBERSOLE: The structural argument is, you 22 know how that bag does.
23 MR. DIDERICH: I think ours is capable of being 24 filled right up to the top.
() 25 MR. PICKENS: I think for most MKls, you can make l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6646
e 2290 19 05 225 P3DAVbw 1 that statement. You still run into a lot of problems. If U
2 you make the wet well go solid, you are looking at an 3 integrated approach, and the scrubbing and things like that, 4 you can venture in the wet well, if you have made it go 5 solid. You don't want to do that.
6 MR. EBERSOLE: This is getting a little hazy, but 7 I recall, if he filled the hole, he would do it himself with 8 the river and tertiary systems. I haven't looked at this in q
9 a long time.
10 DR. OKRENT: What are the LOCA considerations and 11 the seismic interactions?
12 MR. PICKENS: That had to get into what we are
() 13 going to do in terms of costs and what kind of analyses does 14 this impact and everything else. I mean, just to say, the 15 original concept, when Bob brought us in, was these are all 16 going to be simple fixes, and we are not going to have to do ,
17 a lot of things and everything else.
18 If you look at taking up volume inside the .
19 drywell by putting these curbs in, by putting these plugs in 20 and things like that, you are going to have to go in and 21 look at your LOCA analyses. And you could run into some 22 problems, in terms of higher pressures, because you've got a 23 ' smaller volume in there, and you've got more going in i
j 24 there.
() 25 Seismic reactions. You can't let these things go ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800
- . . . _-336-6646- -- - - - -
2290 19 06 226 rolling around inside the drywell. You've got too much
[SDAVbw 1
\~)
2 equipment in there right now.
3 Those are the kind of things where we looked at 4 it and said, it's not as simple as just taking a chuck of 5 concrete and throwing it, the pedestal open. There is going 6 to be negative impacts. There is going to be things that 7 these interface with, that we are going to have to assess 8 and say, is it okay to let the pressure go higher, or should 9 we not put that in there for a LOCA analysis?
-10 ALARA concerns. This is not going to be a short 11 time period to put these things in. This is going to be an 12 extensive amount of work. You are going to pick up a lot of I~)
\s 13 dope, and is it worth it for what we are looking at? I 14 think looking at all these things, I don' t recally have' a 15 conclusion, or I do have. On the next slide, it says 16 qualitative benefit.
17 DR. OKRENT: Can I ask you about the plug in the 18 pedestal before your take that Vugraph off? I'm not quite 19 sure I understand what form it takes in some of the comments 20 you made. Could you give me a mental picture of what that 21 plug --
l 22 MR. PICKENS: The pedestal is sitting in the 23 center of the drywell, and you have an opening for going 24 down and working underneath the reactor vessel. So you have r)
(_ 25 that opening.
1 l ACE-FEDERAL REPORTERS, INC.
I 202 347-3700 Nationwide Coverage 800 336-6646
2290 19 07 227 h DAVbw 1 The first suggestion which came out -- and the
(_)
2 term Bob used was the Jersey Bouncer. You just take a big 3 chunk of concrete like they put out in the middle of the 4 Turnpike to keep you from bouncing in the other lanes, and 5 you put it there in the opening that goes into that pedestal 6 to prevent the corium from flowing out and going over and 7 contacting the drywell liner.
8 DR. OKRENT: So it would be something that would 9 hold the corium in and the water out.
10 MR. PICKENS: That Ja right.
11 DR. OKRENT: And ycu made the point that the 12 corium then wouldn't have water accessible.
13 MR. PICKENS: Not by the drywell sprays, which is
(])
14 the other element Bob is saying we should integrate with 15 this.
16 DR. OKRENT: It is my recollection, it has been 17 suggested that in the BWR you might possibly, if you had, 18 instead of a rather slow rate, a low rate of corium leaving 19 the vessel, a rather large amount leaving together, that 1f ,
20 this fell into water, you could then have a kind of direct 21 heating event that would be threatening to the containment.
22 And if I understand correctly, this thing would 'ceep you 23 from falling into water.
24 MR. PICKENS: It would keep you from having water
() 25 flowing through the drywell sprays. I am not sure if you ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
2290 19 08 228 T xDAVbw I would have any water from when the vessel failed up there or b 2 not. I don't think I could make the statement that just 3 because I had that curb there, there won' t be any water down 4 inside the pedestal. In fact, the sumps are down there, and 5 you could well have water in your sumps, because that is 6 where you collect your identified and unidentified leakage 7 already sitting there.
8 So the fact now that you put a curb or- don' t put 9 a curb in there will not really impact whether or not this 10 core is going to fall in the water.
11 DR. OKRENT: That is the outer curve or this 12 inner plug. We are talking about Item No 1.
13 MR. PICKENS: Both of these are talking about
(]) ,
14 holding the corium inside the pedestal, okay? These were 15 talking about things that we might do outside. The curb 16 outside the pedestal opening might be halfway between your 17 drywell liner and your pedestal wall and would provide a 18 longer path for the corium to flow, before it would contact 19 the drywell liner. This one would be right at the drywell 20 liner wall, which would prevent direct contact, just by no 21 matter what path it takes. And then the additional pedestal 22 opening is just to promote an even distribution, so that you 23 don't have it all coming out of this one opening and taking 24 a direct path right across the drywell.
() 25 DR. OKRENT: Let me give the follcwing. I will ACE-FEDERAL REPORTERS, INC.
202 -347-3700 Nationwide Coverage 800-336-6646
2290 19 09 229 sDAVbw 1 call it a question.
}O 2 It is my impression that we are still looking at 3 this direct heating event for pressurized gater reactors as 4 a possible way of pressurizing the containment rapidly. It 5 seems to me that one has to keep it in mind for the BWRs to 6 make sure that by doing something, whatever it is, assuring 7 that the sprays are on before the vessel fails, for 8 example, it doesn't, in fact, significantly increase the 9 probability of a direct event.
10 Let's just leave it at that.
11 MR. EBERSOLE: Tell me the configuration of this 12 accident hypothesis, the sequence. I gather you are talking 13 about an event where you have l'ost core cooling, and
(]) ,
14 therefore, you have a core melt, but you've got a little bit 15 of water from someplace for spray. You have, by no means, 16 at 10,000 gpm coming out of those big nozzles tha you might 17 have, had they been there to fill a corium quench function.
i 18 You would have to put a big pump in place to do that.
19 I see where that would not be too good.
20 MR. PICKENS: If you still had your RHR pumps 21 that would supply 10,000 gpm, I don't think you would still 22 be using that to put water in the core to prevent core 23 failure.
24 MR. EBERSOLE: Right.
l .
25 (Slide.)
l t
i ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336 6646
. , ,. . . ._ . -. . - - ~ , -
l 2290 19 10 230 1 MR. PICKENS: The qualitative benefit on our r( w)DAVbw 2 assessment of the dry core debris barriers was low. I think 3 that stems froma lot of the work which IDCOR has done in our 4 talks with them. I think there's probably fewer people that 5 believe it is going to flow like water than that it is going 6 to flow like molasses and with how dependent any design that 7 you put together and say whether or not it is adequate to 8 prevent failure via contact with the drywell liner is 9 dependent on the analytical model of how that debris travels 4
10 across the debris mobility. If you implement the other 11 element of the reliable drywell sprays, then you have a 12 shower or water coming down on top of this core dooris.
I think it would even strengthen the arguments
( 13 14 more that it is going to be solidifying or cooling as it 15 goes across the floor, and it would not say that it contacts 16 the drywell liner. I think we would say that the core 17 debris barrier, at best, right now, in our assessment is I
l 18 probably not something that is necessary. And if you put l
l 19 drywell sprays in, that would add to the argument even mor 20 strongly.
l 21 DR. OKRENT: Your comment is that most plants 22 have the capability. It seems to me, that leaves some 23 plar.t c tnat don' t have, presumably, and I am not sure where l 24 it leaves that issue, if one follows your proposed 1
() 25 resolution, which is to do nothing with the sprays.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coveragc 800 336-6646
2290 19 11- 231
/^gDAVbw 1 MR. PICKENS: Again -- and I think this is an
, (_)
2 integrated approach, if you look at this altogether, if you 3 assist the core debris barriers and the reliable drywell 4 sprays.
5 I have done a lot of talking with Bob, and I 6 think I would be comfortable saying, even with him not here 7 now, he would say that probably core debris barriers, you 8 are right, to have reliable drywell sprays is not 9 necessary.
10 DR. OKRENT: Let me just then leave as a 11 question, right now I have no feeling for the seismic 12 issue. I have no feeling for steam explosions in the 13 drywell and their potential seriousness. It may be trivial,
(]) ~
14 but right now to assume that the material just goes bang 15 into large particles and always with no pressure pulse, that 16 may not be a close enough representation.
17 MR. PICKENS: On the wet well, for holding the 18 water down in the torus room, a survey was done. Most of 19 the plants currently have capabilities to holds water in the 20 torus chamber or torus room. If they allow it to go into 21 the corner rooms, they provide for protection to the 22 critical equipment in the corner rooms.
23 Monticello is an example where we do not have a 24 watertight corner room; however, our RHR pumps and critical
() 25 equipment, the electrical components, are elevated. So that ACE-FEDERAL REPORTERS, INC. .
202-347-3700 Nationwide Coverage 800-336 6646
2290 19 12 232 7 ;DAVbw 1 if the torus failed with the current inventory that is in V there, we would not be taking out those pumps now. If.we 2
3 were pumping at 10,000 gpm a minute through drywell sprays, p
4 and it was pumping down through the drywell, that might be 5 different.
6 But Bob tossed out a number, three feet, as a 7 minimum, and we've assessed that the qualitative benefits 8 are low, because looking at what you are really going to get 9 from this, it varies from plant to plant.
10 For our plant, the downcomers would not be 11 submerged if the torus failed. Our suction lines, however, 12 would still be submerged, if we had three feet of water down
() 13 in the torus room. Other plants, if you had.three feet of 14 water in the torus room, by maintaining the inventory in the 15 torus, would not even touch the bottom of the torus, because 16 they are up much higher.
17 So in terms of just coming out and saying, hold 18 that water there and do it for the sole purpose of, if the I 19 corium makes it down to the downcomer and fails, the bottom 20 of the torus, the water goes out.
21 I think you have to believe first that the corium 22 is going to take that path and actually make it down and 23 f ail the torus, which I don' t think that we believe, if we 24 are saying that we don' t need core debris barriers, because
() 25 we don't think it is going to travel like that up there.
ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 800-336 6646 202-347-3700__._ . _
2290 19 13 233 1 1 We are also saying that we don't think it is
{AVbw 2 going to go down to the downcomers and fail your torus.
3 DR. KERR: Is that because you don' t think it 4 will fail the torus, if it goes down the downcomers or 5 because you don' t think it will go down the downcomers?
6 MR. PICKENS: Both. I think you could fail in 7 the downcomers before you get down into the torus, and you 8 wouldn't actually --
9 DR. KERR: Is this just a feeling or has somebody 10 looked at this to try to get some idea of what may happen 11 physically?
12 MR. PICKENS: In terms of what is going to
() 13 happen, once it gets into the downcomer.
14 DR. KERR: Once it leaves the drywell.
15 (Slide.)
16 MR. PICKENS: You are talking abou* -ing over 17 here, filling up to at least the tip. I think Bob said it 18 is 12 feet. We have 7 inches at Monticello going up the 19 lip, flowing down, traveling into here to fail the torus.
20 And whether or not you are going to come up here, come down,
- 21 you could fail right here to come down, or you might get 22 radiant heating and fail your torus.
23 There is different ways, but I think the ral l
24 question is, is it going to make it out of the drywell and fh q,) 25 take that path right there.
l I
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
2290 19 14 234 1 DR. OKRENT: This is exactly the kind of question
{Vm,DAVbw 2 I would pose to the steam explosion question.
3 Does the steam explosion disperse material enough 4 to get in through the downcomers and into the torus enough 5 to give a problem?
6 I don't know. I think no one looks. One assumes 7 it doesn't occur.
8 9
10 11 12 0 13 -
14 15 16 17 18 19 20 21 22 23 24 25 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage, _ _ 800-336-6646
2290 20 01 235 7 NDAVbur 1 MR. PICKENS: There were some discussions on that
(__)
2 point, and I don't think you get a big glob.
3 MR. DIDERICH: If there was debris lying about 4 inside the dry well, it is unlikely to get sufficient into 5 the vent pipes to cause failure because, although not shown 6 on that drawing, in front of each vent pipe is a metal plate 7 which prevents normal design basis missiles from entering 8 the vent.
9 DR. OKRENT: That may or may not be. It sounds 10 good, except I don't right now have a picture of how this 11 material is dispersed by its postulated steam explosion, 12 whether, you know, it slips by.
() -
13 DR. KERR: The metal plate makes it somewhat less 14 likely probably.
15 DR. OKRENT: I think it may make it less likely, 16 but it isn't obvious that it makes it sufficiently less 17 likely.
I i 18 MR. PICKENS: Again, in the time period that we 19 had, these were at best qualitative assessments. This was a l
i 20 point that was discussed. If it had steam pressure and if 21 it got blown out, would it b.e big globs, would it be little l
l 22 globs? If it were little globs, would it really fail the i
23 downcomer?
24 And I think the assessment from my talking with
() 25 some of the consultants for IDCOR was no, we don't feel ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 20 02 236 1 DAVbur 1 that is a reasonable failure mode, but that is just the 2 assessment which was done.
3 DR. KERR: ' See, if you get a steam explosion with 4 these particles, in order to get a steam explosion with a 5 significant amount of energy they are no longer liquid, they 6 are probably solid.
7 DR. OKRENT: That could be.
8 DR. KERR: They have got to be small in order 9 that you get a steam explosion.
10 MR. PICKENS: There is not enough energy left in 11 the globs that; go out to actually cause the failure from the 12 steam.
O 13 oa oxne"r= 'o= re =9e *t"o = =e a teivetv 14 than I would be willing to because it is possible to have 15 steam explosions af ter a considerable amount of material has 16 first reached the floor, and not all of it is necessarily 17 involved.
18 DR. KERR: It has got to be in little pieces.
19 DR. OKRENT: The part that gives up energy very 20 rapidly is going to be in little pieces, but I am just 21 saying we don't have --
22 DR. KERR: You don't have a rigorous proof.
23 DR. OKRENT: I am not talking about a rigorous 24 proof. I haven' t even seen somebody do -- maybe they have O 2s done te -- whee I wou1d ce11 e ree on b1v srstemetic, ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
2290 20 03 237 1 preliminary scoping of the thing instead of just assuming
} fAVbur 2 you can' t have a steam explosion because it doesn' t take the 3 same size to rupture the vessel and push the head out.
4 DR. KERR: That is certainly true.
5 DR. OKRENT: And they occur with a high 6 probability.
7 DR. KERR: They will almost certainly occur.
8 DR. OKRENT: Okay.
9 (Slide.)
10 MR. PICKENS: Element 5 was training and 11 procedures. The objective is to ensure operators are ready 12 to use the plant features to best advantage in severe
() 13 accidents. The requirements, clear symptom-based strategies 14 and integrated approach using all your systems, removal' of 15 unnecessary inhibitions, and then training and procedure for 16 the operators.
17 Again, to reiterate the commitment that was made, 18 Revision 4 implementation by all utilities consistent with 19 previous post-TMI commitments has been made.
20 MR. EBERSOLE: Is it buried in there that you, f
l 21 that the operators have been educated in the physical 22 cascades that might take place -- you know, the actual I
! 23 events in a progressive, fundamental way?
24 MR. PICKENS: Not just the symptoms, but what the l
() 25 typical events are?
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
2290 20 04 238 DAVbur 1 Dick or Kevin, do you know how much that goes 2 into the actual events? What scenarios are leading up to 3 ,
those events?
4 MR. DIDERICH: For the operator training 5 regarding the emergency procedures?
6 MR. EBERSOLE: The TMI folks didn't know what was 7 going on, the operators. The engineers never told them.
8 MR. DIDERICH: The implemenF.ation of the 9 symptom-based procedures is a big step forward because they 10 don' t have to understand the sequence that is occurring.
11 They recognize the symptoms and take corrective actions on 12 that basis. ,
() 13 MR. EBERSOLE: But they are there. Can they use 14 that knowledge as human beings to take up occasions that are 15 not in all the symptoms? They have to know, when the 16 symptoms don' t match, what actions to take.
17 DR. KERR: That is contrary to the philosophy of l 18 symptom-based. I am not sure I like it either, but that is l 19 the philosophy. Symptom-based procedures says you don't l
20 have to understand what is going on. You look at the 21 symptoms involved, I think.
1 j 22 MR. EBERSOLE: It is a matter of degree, I 23 think.
l l 24 MR. DIDERICH: I have talked to operators or l
() 25 engineers who have been training operators, and they have ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4 336-6686
2290 20 05 239 T~wDAVbur 1 come up with sequences that they have run through
'() 2 simulators, and the operators have successfully mitigated 3 the sequence and brought the plant into a safe, stable 4 state, cold shutdown.
5 When they were all done, they came over and said, 6 vhat in the world was broken? What was that? I didn't 7 quite understand what was going on. All I did was follow 8 the procedures. And they got it all back.
9 MR. EBERSOLE: Of course, that is precisely what 10 happened in TMI-2. They didn't know what they were 11 measuring in the pressurizer water level because of the 12 profile of that particular PWR. They didn't know it was 13 depressed in the core but high in the pressurizer, and they
(])
14 were doing it on a symptom-based sequence.
15 MR. DIDERICH: It may be instructive for you all 16 to join the training session that the owners groups are 17 running under the NRC staff in the emergency procedures.
18 You at 1 east ought to be able to get an executive overview.
19 MR. PICKENS: I think in the first week of 20 October there is -- we will walk you through the emergency 21 procedure guidelines.
22 DR. OKRENT: I don't think there were, Jesse, the 23 symptom-based procedures at the time.
24 MR. DIDERICH: They were a result of.
() 25 DR. OKRENT: So they were not using the ACE-FEDERAL REPORTERS, INC.
, 202-347 3700 Nationwide Coserage 800 336-6646
2290 20 06 240 1 procedures based on this new philosophy, which in fact I
} zDAVbur LJ 2 think is helpful for a wide variety of transients. I will 3 guess that there still are some where there are some 4 complexities, and when you get into the severe core damage 5- area you just don't have much to say.
6 DR. MARK: Could you help me by giving me an 7 example for No. 2 up there?
8 MR. PICKENS: Removal of unnecessary inhibitions?
9 DR. MARK: Yes. What kind of inhibitions are you 10 imputing to the operators, or is it the engineers?
11 DR. KERR: This is Mr. Bernero's language, not 12 his.
() 13 DR. MARK: I know.
14 MR. PICKENS: Lowering the water level or tu'rning.
15 off the water supplies during an ATWS scenario to get your 16 power down when almost everything else in an operator's mind 17 tells him that it is good to put water into the core, and I 18 would say that there are some operators who feel somewhat 19 inhibited right now in taking the action of lowering or 20 stopping flow into the core to take that power level down.
21- DR. MARK: But this will be embodied in your 22 emergency procedures that you are supposed to follow, the 23 new ones. So he isn't supposed to be going around full of 24 inhibitions?
() 25 MR. PICKENS: No, and we are working with them ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage __ 800 336 4 646
-- _ l ___ _- . 202-347-3700_ __ _ __ _ _
2290 20 07 241 7'TDAVbur 1 during training and as we walk through the simulators with
(_)
2 them on the emergency procedures, and we did it at 3i Monticello. We walked them through that scenario 4 specifically just to get their reaction to it and to get 5 them feeling comfortable.
6 That is one thing you need to do, is get your 7 operators feeling comfortable.
8 DR. OKRENT: In fact, I would argue a plant that 9 has to have emergency procedures that tells the operator 10 stop putting water into the reactor and to drop the level 11 has a kind of -- what would you call -- an Achilles' heel, 12 which may be found by some operator who thinks this is going
() ,
13 on but it is really something else. It is not really the 14 way one would ordinarily.
15 DR. KERR: I' agree. Let's open up the main steam 16 isolation valve.
17 (Laughter.)
18 MR. EBERSOLE: Let me ask -- I don't think I have 19 seen this -- another way without reducing the level, that is 20 reducing the pressure and to vastly increase the void 21 volume, and you get out the same entropy rate because now 22 you have got the low density voids.
23 Isn't that maybe more conservative than the 24 approach of depressurizing and building up the voids?
() 25 MR. PICKENS: Indeed, if you listen to Bob Henry, ACE-FEDERAL REPORTERS, INC.
_ Nationwide Coverage 800-336-6M6
-. - . . ___ _ _. 202-3{7 3700 _ _ _
F#'0 20 08 242
(_/ .
1 DAVbur 1 who gave a presentation over at the NUREG-ll50 meeting in 2 Bethesda.today, one of the things he is talking about is 3 depressurizing in conjunction with lowering the core 4 levels.
5 MR. EBERSOLE: But you could be more conservative 6 about lowering the level. You don't need to go as far down 7 because the void density and the heat removal rate -- is 8 there a curve that shows how low you can get the power?
9 MR. PICKENS: Is it 2 or 3 percent which he 10 reached?
11 VOICE: That is the point in one of Mr. Pickens'
() 12 earlier slides when he was talking about venting.
13 Specifically, we have some activities ongoing now with some 14 more detailed modeling to try and really assess what that 15 power level is.
16 MR. PICKENS: As part of the question as whether 17 .or not to vent during an ATWS scenario.
18 MR. EBERSOLE: Vent is a containment --
j 19 DR. KERR: Jess, why don't you let him finish?
20 MR. PICKENS: There was concern about an ATWS 21 scenario leading to failure of containment and venting and 22 what does that do. So the New York Power Authority came out 23 and said we don't think you need to vent. So they have got
() 24 a scenario where they say you don't need to vent for an ATWS 25 scenario.
ACE-FEDERAL REPORTERS, INC.
_.. _ , .202-347-3700_ _ _ , . __ _ __ Nationwide Coverage _ 800-336-6646
I ~2290 20 09 243 DAVbur Now, there is a lot of work going on. EPRI has
' 1 2 got a contract with General Electric right now where they 3 are doing a TRAC G run. As part of that they are looking at 4 this overall approach and what you might do to address the 5 ATWS scenario in lowering the water level, initiating ADS, 6 what are the sensitivities, all that kind of thing, and that 7 is what I was referring to back under that other element l 8 where I said it under review for sizing it.
9 What we are trying to do is getting away from
! 10 sizing it based on ATWS, saying we don' t need a vent, and i
11 that is by lowering the water level.
I 12 MR. EBERSOLE: And doing the ADS? That would be
() 13 in front of it?
14 -
MR. PICKENS: Are they actually lowering thi ADS 15 during the New York Power Authority scenario?
16 VOICE: I guess they are. I knew that was one of I 17 the options.
! 18 MR. EBERSOLE: That may eliminate the need for
- 19 venting.
20 MR. PICKENS: That is exactly the argument which 21 Bob Henry is working on right now when he presented it.
22 The other element on this is a proposed review of 23 Rev. 4 with insights from severe accident studies. That is 24 going before the owners group for a funding request right i
() 25 now, ACE-FEDERAL REPORTERS, INC.
. . _ _ _ _ . em_ _.aee-u._ =m _.__
2290 20 10 244 7"NDAVbur 1 A scope item which I have added onto that -- and V
2 again I want to emphasize that that is to look at the 3 existing emergency procedure guidelines and to make sure 4 that the actions which we are telling the operators to take 5 don't conflict with any actions which you might take if you 6 got into a severe accident sequence. The scope of the 7 emergency procedure guidelines right now stops with 8 prevention of core damage. It does not go into the severe 9 accident realm.
10 It is the feeling of a lot of people out there 11 that, you know, we would not change that significantly from 12 what they are now to continue into a severe accident
() 13 sequence. -
14 But that check still needs to be done, and there 15 is a scope item under this proposed review. I have asked 16 that a scope item be put in there which would identify what 17 further we need to do in order to bring this to a complete i
l 18 package of emergency procedure guidelines and accident
- 19 management procedures.
i 20 (slide.)
21 That again brings us to the conclusion slide, 22 which I don' t really need to go over again. We saw this 23 before.
24 With that, that concludes the presentation which l
() 25 we have prepared on what we have done so far.
l 1
l ACE-FEDERAL REPORTERS, INC.
WrkD07 2700 Nationwide Covers 80433M446
'2290 20 11 245 7~NDAVbur 1 DR. MARK: Thank you.
()
2 Are there further questions?
3 DR. KERR: I am waiting for the judicious remark 4 by the Chairman. ,
5- DR. MARK: That was the first one.
6 (Laughter.)
7 DR. MARK: Well, I at least am impressed by the 8 attention that the BWROG's people are giving this subject.
9 I think this will probably reverberate around 10 amongst at least some -- maybe not ultimately definitive, 11 but at least a better understanding of what needs to be 12 thought about, looked at, pictured, and really be of general l ) 13 use. -
14 DR. KERR: I just hope that it is absolutely 15 useless.
16 DR. MARK: You hope it is useless. That is 17 because you aren't really fond of severe accidents.
18 DR. KERR: I just hope there is never one.
19 DR. MARK: I think it is a good exercise, 20 nevertheless, to lay a lot of dust, I hope, and thank you I
i 21 very much for helping us.
22 If there are no additional remarks, then I will 23 declare this subcommittee meeting closed.
24 (Whereupon, at 4:50 p.m., the subcommittee
() 25 meeting was adjourned.)
ACE-FEDERAL REPORTERS, INC.
__ 202 347-3700. _ , _ Nationwide _ Coverage _ __. _ _ _ 800-336-6646 _ _ _ _ _ _ _
CERTIFICATE OF OFFICIAL REPORTER L)
This is to certify that the attached proceedings before.
the UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEES ON CONTAINMENT REQUIREMENTS AND REACTOR SAFEGUARDS DOCKET NO.:
PLACE: WASHINGTON, D. C.
\ DATE: TUESDAY, SEPTEMBER 23, 1986 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
/
(sigt) -
r (TYPED)
DAVID L. HOFFMAN Official Reporter ACE-FEDERAL REPORTERS, INC.
Reporter's Affiliation (v
-. . - - _- . -. . - _ _ - = _ - - . . ._
9 h
ise SEVERE ACCIDENT CONTAINMENT POLICY INDUSTRY EVALUATIONS i
t i-i O
i i
SEPTEMBER 23, 1986 f
- WASHINGTON, DC i
i 1
3 t
i t
i
~.--~~m- -. .n-, - _ . . . - . - - .-,,nnn---- --- m-n,- -,-- -- . --- ...w. , , - - . , , . . _ ,.-
t
- O
- OBJECTIVES
O PRESENT RESULTS OF IDCOR/BWROG & UTILITY EVAL 11ATIGIS 0F PROPOSED SEVERE ACCIDEllT CONTAINMENT POLICY ELEMENTS o DISCUSS CONCLUSIONS REACHED BY UTILITIES 1
. O l
i i
i O
- --- - - - - . . _ _ _ _ _ _ . _ _ _ . _ _ _ . . . _ . ~ . _ - _ _ . - - - - - - - - - _ _ . _ _ . . . _ _ . . _ _ - - - - . - . - - _ _ - - , . - . . - - . , - -
CONCLUSIONS - BWR EXECUTIVE MEETING O
o AUGUST 19 MEETING - 20 0F 23 BWROG UTILITIES REPRESENTED o CONTINUE CONSTRUCTIVE DIALOGUE WITH NRC o AGREEMENTS:
COMMIT TO IMPLEMENTATION OF REVISION 4 TO EPGS NUMARC CONTACTED TO CONSIDER SEVERE ACCIDENT CONTAINMENT ISSUE AS A GENERIC INDUSTRY ISSUE CONTINUE WORKING WITH NRC TO BETTER DEFINE ISSUES FOR RESOLUTION PROPOSE TO BWROG SEVERE ACCIDENT INSIGHT REVIEW 0F EPG REV, 4 i
I I
i O
au.+
1 I
O IDCOR/BWROG/ UTILITY EVALVATIONS:
o WIDE VARIATION IN ESTIMATED COST o EVALVATIONS PERFORMED ON SMALL NUMBER OF PLANTS 1 l
0 OBJECTIVES APPEAR GENERIC - EllHANCEMENTS APTEAR PLANT SPECIFIC 4
i o QUALITATIVE ASSESSMENTS OF BENEFITS AND NEGATIVE I
IMPACTS PERFORMED 4
i 1
4 i
i - _ - .--.- . . . - . _ , _ . - . . - - - - - - , - . . . . - - .-
~
h ELEMENT 1 - HYDR 0 GEN O
OBJECTIVE: PREVENT HYDR 0 GEN COMBUSTION CAUSED FAIUJRE REQUIREMENTS:
A. 0XYGEN CONTROL (MRK I AND l%RK II)
B. HYDROGEN CONTROL (MRK III)
IDCOR/BWROG EVALUATIONS:
~
o 0XYGEN CONTROL BY NITROGEN INERTING ADEQUATE FOR MRK I AND MARK IIs.
o LIMITING THE TIME DEINERTED UNDER REVIEW.
O MRK III HYDROGEN CONTROL BEING ADDRESSED BY HYDROGEN CONTROL 0WNERS' GROUP (HC0G).
O
O HYDHDGEN CONIPOL OWERS GOOP FIFAEGY 10 ADDRESS SENERE NrmFWIS NO SIATIN HEACKDUT O - m. e- - -
BgmPstR SEPEMBER 11, 1986 l
e O
v -
---~<w- --- - - . . - - - - - - ~ . va - -<v------e, - , , - - w nw , , , , e,me--,, ww.v,mm-_. v.v~~-w s,- --ne w -w ww e n,
d O
G summomo
- Established HOOG Programs to address Hydrogen Rule requirments for
" Degraded Core Accidents".
- Quarter Scale Testing Program
- Analytical effort
- Station Blackout as a hydrogen generation event (HGE) within the context of recoverable degraded cores is an issue being addressed.
- Current HOOG evaluation indicates that SBO is not a credible HGE
- HCDG responding to NRC questions
- %e need for an independent power supply for igniters in the event of an SBO identified by the NRC in the context pf Severe Accidents.
i 4
l O
O
O HOOG to supply the design criteria for a backup power supply to the hydrogen igniters.
- Backup power supply need rot be safety related
- Identify impact of addressing Severe Accidents on the design of a backup power supply.
- Nmber of igniters required in the event of an SB0
-Itke use of existing data base and criteria
- Additional testing, only if necessary, to follow cmpletion of current Test Program - end of this year Responsibility of individual Mark III owners with support of HOOG as required.
- Define backup power supply source -
- Define associated costs
- Meet and discuss with the NRC the details of the design, costs, and benefits of a backup power supply to the igniters
- Decision and timing for pr M ing O
ELEMENT 2 - SPRAYS GJECTIVE: SPRAY WATER T0:
.1, QUENCH DEBRIS (PRIMARY)
- 2. SCRUB AEROSOLS (SECONDARY)
- 3. L0iER PRESSURE (SECONDARY) 14 . COOL VULNERABLE EQUIPMENT (SECONDARY)
REQUIREMENTS:
- 1. SPRAY IN DRYWELL
- 2. BACKUP WATER SOURCES AND PUMPS
- HOSE C ONNECTIONS
- USE OF FIREPAINS O
IDCOR/BWROG EVALUATIONS:
, o TYPICAL SPRAY CAPACITY 5 - 10,000 GPM/ HEADER 0 CONCEPTS CONSIDERED o CONNECTIm TO HOSE STATIM IN REACTm BUILDING
- APPRWIMATE FL(W PROVIDED 200 GPM
- DTS NOT PROVIDE SPRAY
_ _ - . . _ _ _ _ . . _ - . , . - _ .- . - - - - _ . . - _ . - _ _ - _ - - - - - - - - - - - . _ . - - . ~ .
ELEMENT 2 SPRAYS (Continued)
O o CROSS TIES FR0n DiESEt riRE euneS 10 RaR
- APPROXIl%TE FL(W PROVIDED 1/2 0F FIRE PUMP RATING
- CLOSE OFF APPR0XIPATELY 70% T N0ZZLES TO ACHIEVE SPRAY o FLOW RATES IDENTIFIED APPEAR ADEQlRTE i
.; o DEBRIS QUENCHING DOES NOT REQUIRE SPRAY i
o POTENTIAL BENEFIT / RISK WARRANTS FURTHER STUDY O
J O
\
. . , . - , . . - . . . - ~ , , - . - - - - - , . - . , , , , , - - . - - - - . . - . - - - - , - - , - - . . - - - . - . . . . _ _ _ . , - - - - - . - - , . , _ , . . - - - - - - - . - - - - - ,
ELEMENT 3 - PRESSURE O
OBJECTIVES: 1. AVERT UNCONTROLLED OVERPRESSURE FAILURE
- 2. CONTROL RELEASE PATH (SCRUBBING) l REQUIREMENTS:
- 1. SUBSTANTIAL CAPABILITY TO VENT WETWELL
- 2. REMOTE / RELIABLE CONTROL OF VENT VALVE
- 3. ABILITY TO RECLOSE VENT IDCOR/BWROG EVALUATIONS:
O o CONCEPTS CONSIDERED o UPGRADE DUCTING AND STANDBY GAS TREATMENT SYSTEM (SBGTS) TO CONTAINMENT DESIGN PRESSURE CAPABILITY
- NOT FEASIBLE TO UPGRADE SBGTS o HARDPIPED BYPASS AROUND SBGTS o 1%RDPIPED DEDICATED VENT o COSTS ARE C WPARABLE FOR HARDPIPED TTIWS O
ELEMENT 3 - PRESSURE (Continued) 0 0 Veli? SIZING UNDER REVIEW o ATWS o 'DHR i
o NEGATIVE IMPACTS NEEDING FURTHER REVIEW O SECONDARY CONTAINMENT CONTAMINATION
~
o DELIBERATE RELEASE O
O
ELEMENT 14 - C WE DEBRIS O
OBJECTIVE: REDUCE LIKELIHOOD E FAILURE BY DIRECT ATTACK REQUIREMENTS:
- 1. USE PRACTICAL DEBRIS RETARDING BARRIERS
- 2. CONSERVE SUPPRESSIm P0OL WATER AS A QUENCHING P0OL IDC N/BWROG EVA LUATI WS ;-
0 DRYWELL O o CONCEPTS CONSIDERED PLUG IN PEDESTAL OPENING INCREASE SUMP SIZE INSIDE PEDESTA L CURB (UTSIDE PEDESTA L OPENING CURB AT DRYWELL LINER / FLOOR JUNCTION ADDITIWAL PEDESTA L OPENINGS TO PROMOTE EVEN l
DISTRIBUTION o NEGATIVE IMPACTS LOCA CONSIDERATIONS SEISMIC INTERACTICNS
- ALARA CWCERNS
ELEMENT I4 - CORE DEBRIS (Continued)
..O c 00ALITATIVE BENEFIT LOW DEPENDENT ON ANALYTICAL MODELS OF DEBRIS MOBILITY WHICH ARE VERY UNCERTAIN o WETWEU_
o MGT PLANTS CURRENTLY HAVE CAPABILITIES TO HOLD WATER IN TORUS CHAMBER OR PROVIDE PROTECTim TO I
CRITICAL EQUIPMENT IN C mNER ROMS o QUALITATIVE BENEFITS LOW O o NO NEGATIVE IMPACTS o NO FURTHER STUDY WARRANTED I
i O
-.------,.-,---,-,---,---,,,,,-+---,----,,,n~- -m----e --n-- -- - - - - - - ------ - - - , - - . - - - - -- - - - ------n--- ,, - - - -.--- -
ElfMENT 5 - TRAINING AND PROCEDURES
@JECTIVE: ENSURE OPERATCRS ARE READY TO USE PLANT FEATURES TO BEST ADVANTAGE IN SEVERE ACCIDENTS REQUIREMENTS:
- 1. C LEAR SYMPTm BASED STRATEGIES (INTEGRATED)
- 2. REMOVAL OF UNNECESSARY INHIBITIWS
- 3. TRAINING /PRCCEDURES IDCOR/_BWROG EVALUATIONS:
o REV. 4 IMPLEMENTATION BY All UTILITIES CWSISTENT WITH O eRevia;S eGT-TMi COMMirrtENT o PROPGE REVIEW 0F REV. 4 WITH INSIGHTS FRm SEVERE ACCIDENT STUDIES l
l l O
CONCLUSIONS - BWR EXECUTIVE MEETING
.O o AUGUST 19 MEETING - 20 0F 23 BWROG UTILITIES REPRESENTED o CONTINUE CONSTRUCTIVE DIALOGUE WITH NRC o AGREEMENTS:
COMMIT TO IMPLEMENTATION OF REVISION 4 TO EPGS NUMARC CONTACTED TO CONSIDER SEVERE ACCIDENT CONTAINMENT ISSUE AS A GENERIC INDUSTRY ISSUE CONTINUE WORKING WITH NRC TO BETTER DEFINE ISSUES FOR RESOLUTION O -
PROPOSE TO BWROG SEVERE ACCIDENT INSIGHT REVIEW 0F EPG REV. 4 I
l'O
- p- ,
tO I
i ,
i t i
1 i i ACRS SUBC0f7.ITTEE !
! i i
i i ON i
i i CONTAI!1liENT PERFORiiANCE i i
i 4 i i
1 SEPTEMBER 23, 1966 .
r i
i O !
t I
! i 1
R. M. LERNER0, USHRC I
s i
i I
i
! I I
l
~
O ,
t i
f I
--, .--- _ . -v
m . . . . .. ._ _ ...
1
\
O .
GENERIC LETTER ON ,
BWR CONTAINMENT PERFORMANCE THE SETTING: PLANT EVALUATIONS UNDER THE SEVERE ACCIDENT POLICY STATEMENT O
THE KEY REGULATIONS: GDC 16 AND GDC 50 THE SUBJECTS: 37 BWRS WITH PRESSURE SUPPRESSION CONTAINMENTS THE METHOD: A GENERIC LETTER OF REQUIREMENTS TO IMPLEMENT
! CHANGES BASED ON GENERIC EVALUATION e
~
O 1
l L
. .- . ~ . = -.
. ; 2
.. -t ,
NRC SEVERE ACCIDENT
.- POLICY STATEMENT ,
1 i
, o THE MOST COST-EFFECTIVE OPTIONS FOR REDUCING THIS I
VULNERABILITY SHALL BE IDENTIFIED AND A DECISION SHALL BE REACHED CONSISTENT WITH THE COST-EFFECTIVENESS CRITERIA 0F THE COMMISSION'S BACKFIT POLICY AS TO WHICH OPTION OR SET OF OPTIONS (IF ANY) ARE JUSTIFIABLE AND REQUIRED TO BE IMPLEMENTED. -
l 1
i
[O t e IN THOSE INSTANCES WHERE THE TECHNICAL ISSUE G0ES BEYOND j CURRENT REGULATORY REQUIREMENTS, GENERIC RULEMAKING WILL BE 1
i THE PREFERRED SOLUTION. IN OTHER CASES, THE ISSUE SHOULD BE DISPOSED OF THROUGH THE CONVENTIONAL PRACTICE OF ISSUING
! BULLETINS AND ORDERS OR GENERIC LETTERS WHERE MODIFICATIONS ARE JUSTIFIED THROUGH BACKFIT POLICY, OR THROUGH PLANT-SPECIFIC DECISION MAKING ALONG THE LINES OF THE INTEGRATED SAFETY ASSESSMENT PROGRAM (ISAP) CONCEPTION.
I i
1 l .
O
.~. --__. _ , .-- ,. . _ _ _ .. _ . ,- - _ .-. - _ . _ ~ __. ,_. ._ _. -_.._-. _ .. _ _ __ _ _ . _ _
/ 3 O
GDC 16: ,
CRITERION 16 - CONTAINMENT DESIGN, "
--AN ESSENTIALLY LEAK-TIGHT BARRIER AGAINST THE UNCONTROLLED RELEASE OF RADI0 ACTIVITY TO THE ENVIRONMENT AND TO ASSURE THAT THE CONTAINMENT DESIGN CONDITIONS IMPORTANT TO SAFETY ARE NOT EXCEEDED FOR AS LONG AS POSTULATED ACCIDENT CONDITIONS REQUIRE."
O o
O
~
O
.. 4 O .
GDC 50: .
CRITER10N 50 - CONTAINMENT DESIGN BASIS. --AS REQUIRED BY SECTION 50.44, ENERGY FROM METAL-WATER AND OTHER CHEMICAL REACTIONS THAT MAY RESULT FROM DEGRADATION BUT NOT TOTAL FAILURE OF EMERGENCY CORE COOLING FUNCTIONING, (2) THE LIMITED EXPERIENCE AND EXPERIMENTAL DATA AVAILABLE FOR DEFINING ACCIDENT PHENOMENA AND CONTAINMENT RESPONSES, AND (3) THE CONSERVATISM OF THE CALCULATIONAL MODEL AND INPUT PARAMETERS."
O a
D
. . .= .:....:- . :. u ... - - - - - . . .. ..+ -
U.S. B ILIN WATER REACTORS O -
1 e 24 BWR 2/3/4 WITH MARK CONTAINMENT (ALL LICENSED) .
i e 9 BWR 4/5 WITH MARK II CONTAINMENT (7 LICENSED)
I i
e 4 BWR 6 WITH MARK III CONTAINMENT (3 LICENSED)
J l ..
o 1
O
~
~
O
- 6 O INSTITUTIONAL PROCESS e CLOSURE OF SEVERE ACCIDENT ANALYSIS FOR CONTAINMENT NO FURTHER ANALYSIS UNLESS EXCEPTION IS TAKEN.
e SPECTRUM 0F OPTIONS RULEMAKING 50.54F LETTER FOLLOWED BY ORDER
~
GENERIC LETTER FROM DIRECTOR NRR OR DIRECTOR DBWRL O
e PROCESS CHARACTERISTICS
- BASED ON TECHNICAL WORK AVAILABLE, IDCOR, SOURCE TERM, PLANT SPECIFIC WORK, NUREG-1050, NUREG-1150 OPEN TO PUBLIC FOR COMMENT AND PARTICIPATION ti e
-. . . - . . . ~ - . , . . . . . - . ..
7 A BWR - MARK I
) FOR REFERENCE BEFORE ,
e CORE MELT FREQUENCY: 1x10-4/YR A FULL SPECTRUM OF SEQUENCES INCLUDING BLACK 0UTS e CONTAINMENT CAPABILITY: UNCERTAIN AND VARIABLE BUT ASSUME 1 OUT OF 2 CORE MELTS GIVES FAIRLY LARGE RELEASE AFTER e CORE MELT FREQUENCY: 1x10-4/HR IPE FOR FRONT END MAY REDUCE BUT NO CREDIT IS TAKEN HERE
(]) -
e CONTAINMENT CAPABILITY: SUBSTANTIAL ASSURANCE THAT CONTAINMENT WILL MITIGATE CONSEQUENCES, DEGREE VARIABLE FROM PLANT TO PLANT BUT 1 OUT OF 50 CORE MELTS GIVING A FAIRLY LARGE RELEASE SHOULD BE REPRESENTATIVE O
g HYDROGEN CONTROL PROPOSED REQUIREMENTS e CHANGE TECH SPEC. AT END OF OPERATION FROM 24 ,H0VR ALLOWANCE TO 12-HOUR ALLOWANCE OF NON-INERTED OPERATION AI REDUCED POWER e PERMIT 12-HOUR PERIOD AT REDUCED POWER WITHIN THE OPERATING CYCLE TO SEARCH FOR UNIDENTIFIED LEAKAGE RATIONALE O
e DEINERTING TYPICALLY TAKES 4-8 HOURS e LEAKAGE INSPECTION AND MINOR REPAIR CAN BE REASONABLY ACHIEVED IN 4-8 HOURS e REDUCED POWER ( 6 33%) SUBSTANTIALLY REDUCES SHORT-LIVED FISSION PRODUCT INVENTORY AND DYNAMICS OF POSSIBLE ACCIDENTS 9
9
9 DRYWELL SPRAY PROPOSED REQUIREMENTS e REDUCE DESIGN SPRAY RATE (CHANGE N0ZZLES) TO AB.0VT 10% OF PRESENT VALUE e PROVIDE AC-POWERED BACKUP WATER SUPPLY FOR SPRAY AND AC-INDEPENDENT WATER SUPPLY, AVAILABILITY BY REMOTE MANUAL OPERATION OR BY SIMPLE RELIABLE PROCEDURE DESIRABLE '
Oe MAKE ALTERNATE WATER SOURCES AVAILABLE TO COOL CORE DIRECTLY e 90/10 MODE OF RHR OPERATION RATIONALE 11 e WATER SUPPLIES AND EQUIPMENT ARE ALREADY AVAILABLE FOR LOWER FLOWS e LOWER FLOWS PROVIDF. ALL BENEFITS EXCEPT LOW A T DECAY HEAT REMOVAL AND DO NOT RAPIDLY FLOOD CONTAINMENT __
e ASSURED DRYWELL SPRAY SUBSTANTIALLY REDUCES PROBABILITY AND .
O SIGNIFICANCE OF DRYWELL FAILURE OR SUPPRESSION POOL BYPASS
,_ . . _ _ _ _ , , . . _ . - . , ,,_-...,,.--_..-._.._..v .
m .. . , . . . . . . . .... . .
10 PRESSURE CONTROL PROPOSED REQUIREMENTS e RELIABLE CAPABILITY TO VENT WETWELL AT EPG PRESSURE LEVEL WITH OR WITHOUT AC POWER. FOR VENTING WITHOUT AC POWER MANUAL PROCEDURE IN ADVANCE MAY BE USED IF NITROGEN PURGE IS AVAILABLE o VENT OF 18-INCH DIAMETER OR GREATER DESIRABLE Oe ABILITY TO VENT SLOWER SEQUENCES THROUGH STANDBY GAS TREATMENT SYSTEM e BURST RESISTANCE DUCTING IN REACTOR BUILDING TO MINIMIZE COMPLICATIONS RATIONALE e RELIABLE VENTING PREVENTS UNCONTROLLED OVERPRESSURE FAILURE WHICH CAN CAUSE CORE MELT e VENTING WITH DRYWELL SPRAY GIVES GREAT ASSURANCE OF RELEASE MITIGATION .
~
O
- n. . .. .. . .. ..
11 CORE DEBRIS PROPOSED REQUIREMENTS ,
e ASSURE RETENTION OF WATER AT LEAST 3 FEET DEEP IN TORUS ROOM IF TORUS LEAKS ENTIRE CONTENTS DESIRABL E e CONCRETE CURBS OR OTHER BARRIERS WHICH WOULD RETARD DEBRIS ATTACK OF DRYWELL SHELL O
e AVOID LOSS OF REC 0VERY SYSTEMS FROM WETTING BY TORUS ROOM WATER
~1 RATIONALE e DRYWELL FAILURE BY DEBRIS ATTACK IS MADE LESS LIKELY AND LESS SIGNIFICANT BY DRYWELL SPRAY AND VENTING e RETENTION OF TORUS WATER ENSURES DEBRIS QUENCHING AND SHOULD FACIL'ITATE ACCIDENT REC 0VERY O
Table 4.7
SUMMARY
OF RELEASE FREQUENCY OF EACH RELEASE TYPE FOR EACH ACCIDENT C1. ASS: BASE CASE
, "BEST ESTIMATE" CASE h
Accident Class Release Frequency (yr" ) of Release Type Designator Frequency Core Melt.,g)
(yr M g g _L_M, g NCF IA 1.33E-05 7.38E-09 2.58E-08 1.25E-07 6.89E-08 1.81E-07 1.29E-05 IB 6.20E-06 7.07E-09 1.11E-07 5.07E-07 3.56E-07 9.03E-07 4.31E-06 IC 2.60E-06 5.27E-07 2.83E-07 1.61E-08 9.31E-09 2.44E-08 1.74E-06 ID 3.90E-06 3.32E-09 4.60E-08 2.86E-08 5.44E-07 1.54E-06 1.74E-06 II 2.10E-06 4.73E-10 5.39E-07 1.56E-06 - - -
III ' 7.30E-07 5.53E-10 9.06E-10 6.56E-09 4.55E-08 1.28E-07 5.49E-07 l
IV 2.20E-06 5.44E-09 6.55E-07 1.54E-06 - - -
l 1
l Sum 3.10E-05 5.51E-07 1.66E-06 3.78E-06 1.02E-06 2.78E-06 2.12E-05
% of Total Core Melt Frequency 100% _. 7.1% 12% 3.3% 9.0% 68%
Source: Vermont Yankee Contaiir.ent Safety Study - August 1986 f
. ' t-f G G G
~ .. . = .. -
. -4
o' eBLE1 o BOILING WATER REACTORS WITH MARK I CONTAINMENTS
. LICENSED OPERATING PLANT,, POWER . LICENSE NAME LEVEL DATE COUNTY STATE UTILITY '
BROMSFERRY1 3293 12/20/73 LIMESTONE COUNTY Al TVA BROWNS FERRY 2 3293 08/02/74 LIMESTONE COUNTY AL TVA BROWNS FERRY 3 3293 08/18/76 LIMESTONE COUNTY Al TVA BRUNSWICK 1 2436 11/12/76 BRUNSWICK COUNTY NC CAROLINA POWER & LIGHT BRUNSWICK 2 2436 12/27/74 BRUNSWICK COUNTY NC CAROLINA POWER & LIGHT COOPER 2381 01/18/74 NEMEHA COUNTY NE NEBRASKA PUBLIC POWER DISTRICT DRESDEN 2 2527 12/22/69 GRUNDY COUNTY IL COMMONWEALTH EDISON DRESDEN 3 2527 03/02/71 GRUNDY COUNTY IL COMMONWEALTH EDISON DUANE ARNOLD 1658 02/22/74 LINN COUNTY IA IDWA ELECTRIC POWER & LIGHT FERMI 2 3292 07/15/85 MONR0E COUNTY MI DETROIT EDISON FITZPATRICK 2436 10/17/74 OSWEGO COUNTY NY POWER AUTHORITY OF STATE OF NY HATCH 1 2436 10/13/74 APPLING COUNTY GA GEORGIA POWER HATCH 2 2436 06/13/78 APPLING COUNTY GA GEORGIA POWER HOPE CREEK 1 3293 04/11/86 SALEM COUNTY NJ PUBLIC SERVICE ELECTRIC & GAS MILLSTONE 1 2011 10/16/70 NEW LONDON CT NORTHEAST NUCLEAR ENERGY MONTICELLO 1670 01/19/71 WRIGHT COUNTY HN NORTHERN STATES POWER
. NINE MILE POINT 1 1850 08/22/69 OSWEGO COUNTY NY NIAGARA M0 HAWK POWER OYSTER CREEK 1 1930 08/01/69 OCEAN COUNTY NJ GPU NUCLEAR CORP PEACH BOTTOM 2 3293 12/14/73 YORK COUNTY PA PHILADELPHIA ELECTRIC PEACH BOTTOM 3 3293 07/02/74 YORK COUNTY PA PHILADELPHIA ELECTRIC PILGRIM 1998 06/08/72 PLYMOUTH COUNTY MA BOSTON EDISON QUAD CITIES 1 2511 12/14/72 ROCK ISLAND COUNTY IL COMMONWEALTH EDISON QUAD CITIES 2 2511 12/14/72 ROCK ISLAND COUNTY IL COMMONWEALTH EDISON VERMONT YANKEE 1593 02/02/73 WINDHAM COUNTY VT VERMONT YANKEE NUCLEAR POWER 9 .
t a
E3 -
o
()
CHRON0 LOGY e JUNE 16, 1986: MEETING WITH BWROG/IDCOR PROPOSED A GENERIC LETTER, PRESCRIPTIVE SOLUTION, BY BACKFIT e JUNE 30, 1986: VERMONT YANKEE COMMITS TO GOV. KUNIN TO DO A SPECIAL 60-DAY CONTAINMENT STUDY
~
e JULY 25, 1986: BOSTON EDIS0N COMPANY BOARD ANN 0UNCEMENT ON PILGRIM CONTAINMENT e SEPTEMBER 11, 1986: MEETING WITH BWROG TO COMPARE BACKFIT NOTES AND STRAWMAN GENERIC REQUIREMENTS
([) e SEPTEMBER 11, 1986: MEETING WITH VERMONT YANKEE TO REVIEW CONTAINMENT STUDY e SEPTEMBER 23, 1986: NRC/IDCOR MEETING ON BWR/ MARK I ANALYSES e SEPTEMBER 23, 1986: ACRS SUBCOMMITTEE ON CONTAINMENT PERFORMANCE TO DISCUSS HARPERS FERRY WORKSHOP RESULTS AND BWR CONTAINMENT GENERIC APPROACH e SEPTEMBER 24, 1986: ACRS SUBCOMMITTEE ON CLASS 9 ACCIDENTS TO DICUSS BWR/ MARK I ANALYSES AND SEVERE ACCIDENT PROGRAM e OCTOBER 23, 1986: NRC LETTER TO VERMONT YANKEE WITH COMMENTS AND QUESTIONS ON LICENSEE'S CONTAINMENT STUDY e NOVEMBER 13, 1986 (MORNING): TECHNICAL INTERCHANGE MEETING .
~
([) WITH BWROG PRIOR TO CRGR BRIEFING
.O ;
e NOVEMBER 13, 1986 (AFTERN0ON): VERMONT YANKEE MEETING WITH NRC TO RESPOND TO NRC QUESTIONS e NOVEMBER 19, 1986: CRGR REVIEW 0F DRAFT GENERIC LETTER ON BWR CONTAINMENT REQUIREMENTS (T0 BE PUBLISHED FOR COMMENT) e DECEMBER 17, 1986: ISSUE DRAFT GENERIC LETTER ON BWR CONTAINMENT REQUIREMENTS FOR PUBLIC COMMENT e DECEMBER 24, 1986: NRC LETTER TO VERMONT YANKEE TO GIVE TENTATIVE FINAL REVIEW' BASED ON POSITION OF DRAFT GENERIC LETTER e APRIL 1987: ISSUE FINAL GENERIC LETTER ON BWR CONTAINMENT REQUIREMENTS
!O .
I e
~
- O
f:p _ _ . _ . . _ _ . _ _ _ . _ _ _ _ _ . _ . . . _ _ _ - . _ . . . _ . . . . _ _ . _ .
- O ! l t
I CONTAINMENT PERFORMANCE DESIGN OBJECTIVE i ,
i 1 STATUS BRIEFING TO i
1 .
i ACRS SUBCOMM11 TEE ON CONTAINMENT REQUIREMENTS j 1 i
! i SEPTEMBER 23,1986 [
Oi WASHINGTON, DC i
i i
I !
t :
BY i i
i t DR. GARY R. BURD!CK !
0FFICE OF NUCLEAR REGULATORY RESEARCH ,
! i i
U. S. NUCLEAR REGULATORY COMMISSION i i
, i i !
. i Oi l 1 i !
l
O I BACKGROUND l
EDO MEMORANDUM TO RES/NRR TO INITIATE A PLAN FOR DEVELOPING A CPD0 - JULY 29, 1985
- DRAFT CPD0 PROGRAM PLAN TO RES, NRR, ACRS - OCTOBER,1985 BRIEFING TO ACRS SUBCOMfilTTEE ON SAFETY PHILOSOPHY, TECHNOLOGY AND CRITERIA - OCTOBER 9, 1985 INTERVIEW 0F SELECTED EXPERTS FOR THEIR VIEWS ON CPD0 -
NOVEMBER 1985 - JANUARY 1986 INITIAL DRAFT CPD0 OPTIONS AND FRAMEWORK 0F IMPLEMENTATION AFPROACH FOR REVIEW - FEBRUARY 1986 O -
CPD0 PLAN TO EDO - MARCH 1986
- ACRS SUBCOMMITTEE BRIEFING - MARCH 12, 1986
- CPD0 WORKSHOP AT HARPERS FERRY - MAY 12-13, 1986 COMMISSION POLICY STATEMENT ON SAFETY G0ALS - JUNE 1986 ED0 MEMO TO RES/NRR IN RESPONSE TO THE CPD0 PROGRAM PLAN:
CONFORM THE CPD0 PROGRAM TO THE SAFETY G0AL POLICY STATEMENT GUIDANCE - JULY 30, 1986 O
O CPD0 WORKSHOP MAJOR CONCLUSIONS CONTAINMENT IS A NECESSARY ELEMENT OF A DEFENSE-IN-DEPTH PHILOSOPHY A MINIMUM CONTAINMENT PERFORMANCE LEVEL NEEDS TO BE SPECIFIED
- A QUANTITATIVE, AND, TO THE EXTENT FEASIBLE, MECHANISTIC CPD0 IS PREFERABLE CPD0 AND ASPECTS PERTAINING TO IT SHOULD BE WELL-DEFINED AND UNAMBIGUOUS O -
0FE OF THE PROPOSED CPD0 OPTIONS WAS CONSIDERED PROMISING AS A BASIS FOR FURTHER WORK 1
O
O COMMISSION POLICY STATEMENT ON SAFETY G0ALS TWO QUALITATIVE SAFETY G0ALS (INDIVIDUAL AND SOCIETAL RISKS)
TWO QUANTITATIVE SAFETY G0ALS (INDIVIDUAL AND SOCIETAL RISKS)
GUIDELINES FOR PEGULATORY IMPLEMENTATION:
GENERAL PERFORMANCE GUIDELINE:
OVERALL MEAN FREQUENCY OF A LARGE RELEASE OF RADI0 ACTIVE MATERI ALS TO THE ENVIRONMENT FROM A REACTOR ACCIDENT SHOULD BE LESS THAN 1 IN 1,000,000 PER YEAR OF OPERATION
(])
STAFF TO PERFORM ADDITIONAL STUDIES AND DEVELOP SAFETY G0AL POLICY IMPLEMENTATION GUIDANCE O
O IMPORTANT ISSUES DEFINITION OF A LARGE RELEASE CONSISTENCY AMONG THE SAFETY G0ALS AND THE CPD0 DEFENSE IN DEPTH CRITERIA FOR VENTING ACHIEVABILITY O
i l
l l .
O l
l l
. 1 O
DEFINITION OF LARGE RELEASE RATIONALE:
O PPOMPT FATALITY SAFETY GOAL: AVERAGE INDIVIDUAL RISK WITHIN A MILE FPOM SITE BOUNDARY < 5X10~7(EF)/RY (EARLY FATALITIES PER REACTOR YEAR)
O IF R IS THE LARGE RELEASE LOWER LIMIT RISK TO INDIVIDUALS IN
- THE DOWNWIND SECTOP TO BE IN CONFORMANCE WE MUST HAVE R_ <_ 5X10-7(EF)/RY O 6 OR , R < , 8X10-6(EF)/RY ASSUMING UNIFORM POPULATION AND 16 SECTORS, O
O DEFINITION OF LARGE RELEASE RATIONALE (CONT.):
FOR DEFENSE-IN-DEPTH AND CPD0 PURPOSES WE COULD TAKE R =
AND HAVE:
DEFINITION: A LARGE RELEASE IS ANY PELEASE OF RADIONUCLIDES TO THE ENVIRONMENT WHICH WOULD LIKELY CAUSE ONE OR MORE EARLY FATALITIES.
^
10-6 O
P(X> EF) 7 1 EF THE LARGE PELEASE THUS DEFINED SATISFIES THE PROMPT FATALITY SG BY ABOUT AN ORDER OF MAGNITUDE AND COMPLIES WITH SG IMPLEMENTATION GUIDELINE ON FREQUENCY OF LARGE RELEASE.
O
O OTHER CONSIDERATIONS VENTING SEVERAL VENTING-RELATED ACTIVITIES ARE UNDER WAY IN THE AGENCY. RESULTS WILL BE INCORPORATED INTO THE CPDO.
LIMITING VENTING LEVELS ARE BEING STUDIED,
- IDEALLY, VENTING SHOULD PREVENT HIGH PRESSURE BUILD UP AND VENTS SHOULD BE CLOSED WITH HIGH RELIABILITY BEFORE CORE DtMAGE OCCURS.
STAFF CONFIDENT THAT EARLY FATALITIES VERY UNLIKELY FROM RELEASE OF 0.1% OF NONNOBLE INVENTORY.
Q -
RELEASE OF NOBLE GAS INVENTORY LIKELY IN SEVERE ACCIDENTS.
NOBLE GAS INVENTORY RELEASE APPFARS WITHIN PPOMPT FATALITY G0AL.
- POLITICAL, LEGAL, AND OTHER (?) QUESTIONS INVOLVED WITH NOBLE GAS VENTING.
DEFENSE IN DEPTH THE DEGREE TO WHICH PRVENTION VS. MITIGATION TRADEOFFS COULD BE MADE.
O
I 9
O THE CURRENTLY CONSIDERED CPD0
- 1. CONTAINMENTS SHOULD SURVIVE ALL BUT HIGHLY UNLIKELY SEVERE ACCIDENTS WITHOUT OCCURRENCE OF A LARGE RELEASE. THE AGGREGATE FREQUENCY OF A "LARGE RELEASE" 0F RADIONUCLIDES TO THE ATMOSPHERE FOR ALL SEVERE CORE DAMAGE ACCIDENT SEQUENCES SHOULD NOT EXCEED ONE IN A MILLION REACTOR YEARS A "LARGE RELEASE" IS DEFINED AS RELEASE OF RADIONUCLIDES TO THE ENVIRONMENT THAT WOULD LIKELY RESULT IN EARLY FATALITIES.
EQUIVALENTLY, "LARGE RELEASE" CAN BE TAKEN TO MEAN " LIFE THREATENING RELEASE."
- 2. CONTROLLED VENTING OF CONTAINMENT SHOULD BE SUCH THAT RELEASE
() 0F RADIONUCLIDE SPECIES TO THE ATMOSPHERE NOT EXCEED 10 CFR 100 LIMITS.
- 3. EMPHASIS ON DEFENSE-IN-DEPTH SHOULD BE RETAINED. THE CONDITIONAL PROBABILITY OF A "LARGE RELEASE" FROM A CORE DAMAGE ACCIDENT SHOULD NOT EXCEED 0.1 (0.01 FOR NEW PLANTS),
REGARDLESS OF OTHER CONSIDERATIONS.
l
C)
LEVEL OF SAFETY AND MEET THE CPD0, A TO ACHIEVE A HIGH CONTAINMENT DESIGN SHOULD CONSIDER THE FOLLOWING:
VERY LOW PROBABILITY OF CONTAINMENT BYPASS (INTERFACING LOCA OR ISOLATION FAILURE)
SEQUENCES FOR WHICH CONTAINMENT FAILS BEFORE CORE MELT, PROVISION FOR VENTING BEFORE CORE MELT, AND HIGH RELIABILITY OF THE SYSTEM FOR RECLOSING THE VENTS HIGH STACKS FOR VENTING OF NOBLE GASES AUTOMATIC DEPRESSURIZATION OF REACTOR VESSEL TO ELIMINATE POSSIBILITY OF DIRECT HEATING OF THE CONTAINMENT O THE ATMOSPHERE HIGHLY RELIABLF CONTAINMENT COOLING SYSTEMS O
i O
FOR A NEW . PLANT, THE FOLLOWING TYPES OF DESIGN OPTIONS MAY BE CONSIDERED:
INCREASED PRESSURE CAPACITY CHOICE OF CONCRETE THAT WOULD BE LEAST REACTIVE WITH A MOLTEN CORE SHAPE OF CONTAINMENT FLOOR TO MINIMIZE EFFECTS OF CORE-CONCRETE REACTIONS AND SPREAD OF MOLTEN CORE CONTINUOUS MONITORING FOR GPOSS LEAKAGE LOW PROBABILITY OF EARLY FAILURE
(])
FILTERED VENTS, NOBLE GAS DELAY LINES /HIGH STACKS FOR VENTING NOBLE GASES OTHER INNOVATIVE FEATURES I
t
- i l 4 l
9 6
l l
{