ML20195E985
| ML20195E985 | |
| Person / Time | |
|---|---|
| Issue date: | 06/03/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1518, NUDOCS 8606090318 | |
| Download: ML20195E985 (194) | |
Text
OR GKAL e
O UN11ED STATES NUCLEAR REGULATORY COMMISSION IN THE MATTER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS COMBINED SUBCOMMITTEES ON SEVERE ( C/OSI-9 ) ACCIDENTS AND NUCLEAR PLANT CHEMISTRY LOCATION: WASHINGTON, D. C. PAGES: 1 - 156 DATE: TUESDAY, JUNE 3, 1986 i s ac0t.0:; !
g60 ree - ; -
...,, l
[CRSOFREECDPY <q, Jo Xot Remove" rom ACRSOffice :
n' ACE-FEDERAL REPORTERS, INC.
CfficialReporters 444 North Capitol Street d g Washington, D.C. 20001 oQ (202)347-3700 NATIONWIDE COVERAGE
l l l
l
() PUBLIC NOTICE BY THE UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS TUESDAY, JUNE 3, 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 dat'b.
i 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 i
O ,
i
__-..__,_-._.-_,_..,..,__,.__..,~_._.,.___.,___,,,_._,,.._,_-,__,,,y,..___,.,_m__,__, _. . . - , t --, ___ - . . . , , _ _
CR27025.0 1 BRT/sjg h- 1 2
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4 COMBINED SUBCOMMITTEES ON 5 SEVERE (CLASS 9) ACCIDENTS 6 AND 7 NUCLEAR PLANT CHEMISTRY Nuclear Regulatory Commission Room 1046 9 1717 H Street, N.W.
Washington, D. C.
11 Tuesday, June 3, 1986 12 The combined subcommittees met at 1:00 p.m., Dr. William O 13 Kerr presiding. .
14 A RS MEMBERS PRESENT:
15 DR. WILLIAM KERR MR. JESSE C. EBERSOLE 17 DR. J. CARSON MARK 18 MR. GLENN A. REED 19 MR. DAVID A. WARD 20 21 ACRS CONSULTANTS PRESENT:
- 1. CATTON 22 M. ORRADINI 23 P. DAVIS 24 25 DEAN HOUSTON, HERMAN ALDERAMN, ACRS Staff Members ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Covernor 800 33H646
27025.0 2 BRT A
V 1 PROCEEDINGS 2 MR. KERR: The mee ting will come to order. This 3 is a combined meeting of the Advisory Committee on Reactor 4 Safeguards, Subcommittee on Severe Accidents and the 5 Subcommittee on Plant Chemistry. My name is W. Kerr, 6 Subcommittee chairman on the Subcommittee on Severe 7 Accidents. Dr. Moeller, Subcommittee Chairman of Nuclear 8 Plant Chemistry, is unable to be present.
9 Other ACRS members here are Ebersole, Mark, Reed 10 and Ward.
11 We also have as consultants, Catton, Corradini 12 and Davis.
13 The purpose of the mee ting is to review a final 7-(> 14 draft ot NUREG-0956, " Reassessment of the Technical Bases 15 for Estimating Source Terms."
16 Houston and Alderman are the cognizant ACRS 17 Staff members for this meeting.
18 The rules for participation in today's mee ting 19 have been announced as part of the notice of this meeting 20 previously published in the Federal Register on May 16, 21 1986. A transcript is being kept and will be made 22 available as ind ica ted in the Federal Register notice.
23 I ask that each speaker identify himself or 24 herself and get as close to the microphone as may be 25 feasible.
O ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336 6646
27025.0 3 BRT 1 We have received no written comments f rom 2 members of the public nor have we had requests for time to 3 make oral statements by members of the public.
4 You will recall that af ter a series of meetings 5 extending over a period from May through December 1985 to 6 consider a draft of this report, the ACRS transmitted a 7 report dated December 1, 1985 to the Commission. The 8 report pointed to what the Committee interpreted as a 9 number of deficiencies in that draft that we had considerod, 10 and concluded that the calculational methods described in 11 the report were not in the form considered by the Committee 12 a ppropr ia te for use, at least as significant f actors in the 13 regulatory decision process.
14 Since our report and af ter a consideration at 15 public comments, the original dratt has been extensively 16 revised by the NRC Staff.
17 Today we discuss that revised version. in a 18 meeting of the full Committee to be held later this week on 19 Friday af ternoon, this will be f urther discussed, and the 20 Staff is asking for a report by the Committee commenting on 21 the version of 0956 that we will discuss here today.
22 I would ask those of you participating as 23 Subcommittee members and consultants, be prepared to give 24 me your comments on, first, whether the new version has 25 been appropriately responsive to the comments made by the ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwde Coverage 800-336-6M6
i G
i 27025.0 4 <
BRT a
l v. 1 Committee in its letter of December 12th; or, if you feel 2 the Sta f f shouldn' t have responded at all, you can say that, l
3 too; your view as to the confidence to be placed in and the i
j 4 use or uses to which the current version should be put; any 1 5 changes, if you can identify them, which in your view would !
6 improve the report -- changes, additions, whatever.
l 7 Since we are being asked to write a report af ter i .
a 8 this mee ting, I will need these comments from you at the '
] 9 end of this meeting, written in legible form, or at least 10 very shortly thereafter.
11 Are there any comments by members of the i 12 Subcommittee at this point? Any comments by consultants? j 1'
13 I hear none. M. Silberberg? <
- (:) 14 I should add, as a matter of logistics, that I 1
j 15 do not intend that the mee ting last beyond 6 :00 p.m.
- 16 MR. SILBERBERG
- Thank you, Dr. Kerr.
! 17 Good afternoon. We are indeed pleased to be 18 here today to help you with your review of the final report,
]
{ 19 NUREG-0956.
20 Today's mee ting with the Subcommittee is a very 21 important milestone for us. First, we have taken the 22 comments in your December 1, 1985 letter on draft j 23 NUREG-0956 very seriously, and have attempted to fccus on 24 the concerns in that letter and have tried to be responsive 25 to them. As we proceed through the rest of the detailed ,
, t L C) !
i j ACE-FEDERAL REPORTERS, INC.
- 202 347 3700 Nationwide Coverage 80 4 336 6646
27025.0 5 BRT 1 presentations, why, I believe we will have a number of 2 places, as we go through this, where , in fact, we address 3 those concerns. We think we have addressed those concerns 4 quite explicitly.
5 In developing NUREG-0956, we were tor tuna te in 6 working with our team of principal investigators f rom our 7 source term contractors and our inde penden t sta f f 8 consultant. Just to get under way, I'll show that now.
9 (Slide.)
10 From Bechtel it was Denning and Gieseke , and 11 Kress from Oak Ridge, Powers f rom Sandia and Pratt from 12 Brookhaven. And our inde pe nden t statt consultant to the 13 source term work over the past several years has been CE) 14 Dr. Isbin.
15 You can put the lights back on.
16 These people are here with us today, over on the 17 left. They are actually sitting in the place where the 18 Staff normally sits. But we feel that's appropriate 19 because, in fact, over the past number of months they have 20 worked very closely with us on providing valuable advice 21 and input on both the content and structure of this report.
22 We feel that the laboratories have an investment in this 23 work, in the work that they've done, and we feel that a 24 part of this report, they each have a little piece of -- of 25 it.
O ace-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Cmcrage MXb336 6646
27025.0 6 BRT 1 I would also like to note that at the direction 2 of the EDO and the office directors, we have added another 3 dimension to the final review process for NUREG-0956. In 4 parallel with our internal NRC review and the ACRS review, 5 which are occurring, in fact, right now in this week and 6 the previous week, we have asked our contractor 7 laboratories, namely, Battelle, Sandia, Oak Ridge, 8 Brookhaven and Idaho, INEL, to review the document to 9 confirm the tollowing points, in a letter that was 10 specifically addressed to the program managers at the 11 laboratory:
12 1. Does NUREG-0956 correctly portray the 13 contribution of each of the organizations participating in O 14 the source term work? Does the laboratory stand behind its 15 own work as incor pora ted in 0956? And, finally, has the 16 laboratory identified any significant technical flaws in 17 the docum'nt? Last week we visited several of these labs 18 to tacilitate this eview, which in fact is under way now.
19 Now, in developing the final draft of NUREG-0956, 20 the Sta f f had what we thought was a considerable challenge, 21 in order to be responsive not only to the ACRS, but to the 22 broad variety of public comments that we had received on 23 the document. And I will say that these comments were 24 substantive and right on target, and we fel t we had a 25 strong obligation to address them as f aithf ully as we could.
i ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-(M6
.. . - . . _ . - . . _ - - . . . _ - . . - . . - . . = . . . . . . .-
4 l
j 27025.0 7 BRT i 1 Part of this challenge was also represented in a
}
2 need to present a balanced, accurate report on the l
3 technical basis for estimating source terms and their '
i 4 limitations while walking a fine line to avoid
, 5 overstatement.
1 j 6 There were a number of important messages and i 7 points that we had to grapple with, that we wanted to I !
j 8 communicate or cover or deal with in some way in the (
l j 9 document, that we feel were at the heart of the objective J
j 10 and the scope and the purpose of this report and its future l
l 11 role in the Agency's business.
4 j 12 I would like to note a few of these now, but I I 13 would also point out that these will all come up again j
( 14 throughout the presentations that will follow mine. The y i
1 15 will come up in the context ot conclusions, recommendations '
l j 16 and points that are being emphasized. And I am going i
- 17 through some of these now just to kind of set the stage for l 18 that, to indicate where we wanted to be f air to the state j 19 of source term technology as we understood it, but also to ,
i 20 try to deal with a product that we had , that, in fact, has ,
l 21 a number of applications. So some of these points are as 22 i tollows, and I'll kind of read them as statements because ;
1 23 later on they appear in a similar way.
i 24 The methodology for source term estimation, l i
j 25 namely the Source Term Code package , is more than a set of l j !
2 Y
! ACE FEDERAL REPORTERS, INC. !
l, 202 347-3700 Nationwide Coverage 800-346M6 ;
' , - , ~ . . - - . , . . . . - - . - .--------.----.----,,--,--,..L
27025.0 8 BRT 1 codes. There is, in fact, a supporting basis of data 2 emerging trom the research program which provides improved 3 understanding of phenomena and insights on severe accidents.
4 These data and models represent major progress 5 and advancement in source term technology over that which 6 has been available in the past, namely, TID 14844, and the 7 methodology and approach available trom the WASH-1400, the 8 reactor safety study. But at the same time we know they l 9 are impertect and we are aware of areas where further 10 improvements can be expected f rom the ongoing research.
11 And these areas have been discussed in the document.
12 Nevertheless, these advances are so significant, 13 we teel, that the new source term methods are recommended 14 for use by the NRC Sta f f to reevaluate source terms in 15 regulatory applications, provided uncertainties are taken 16 into account tor each application.
17 We believe that the Source Term Code package 18 provides a much better basis for use in such applications 19 than does TID 14844, 1962 and WASH-1400, 1974. But we also 20 had to be care f ul not to make judgments on the adequacy of 21 the now source term intormation with its current 22 unce r ta in tie s in the specitic applications. Such judgments 23 have to be made by the user tor each specific application.
24 We believe that is something that should follow this work 25 as the source te rm information is used.
( .
ACE. FEDERAL REPORTERS, INC.
202 347.1700 Natlon*ide Coserage Nxk336-6M6
-. - = - - . . _ _ . - - - - -_ -- _ .
! 27025.0 9
! BRT i
1 It was also necessary for us to clearly state 2 the boundaries of scope and objective for NUREG-0956
[ 3 relative to NUREG-ll50; namely, where does NUREG-0956 end?
4 So we tried to do that and we will be discussing that again 1
i 5 in some detail in the speakers that follow me, and you will
} 6 have to j udge whe ther we have done that properly. We feel 7 tha t we have .
] 8 In today's presentation, Jocelyn Mitchell and i
j 9 Ralph Meyer f rom the Staf f will summarize purpose ,
i i
10 potential uses and limitations of NUREG-0956; the content i 11 of the report in terms of its major changes; conclusions i
j 12 and recommendations; and finally, provide some remarks on 13 the S ta f f re sponse to the ACRS letter and public comments I
)
14 in some specific ways.
15 Now, I would like to close by noting that in i
i 16 tact -- and this, I think, helps place today's meeting in i
' t 17 perspective -- that the Source Term Code package in fact '
i 18 has two immediato uses that are now starting to go forward. e 1 19 One, as you are aware of, the Source Term Code
- 20 package has been used for the NUREG-1150 work, which is now l 21 being -- that report which is now being pulled together and I 22 drafted. And, in tact, I believe the Subcommittee had a 23 presentation on NUREG-ll50 laut month in Albuquerque.
'. 24 So, its initial use, the initial use of this i
25 qg ,intormation is right there. The next use, which is in fact
! ($) !
ACE FEDERAL REPORTERS, INC, 202 347-37m Nadonwide Coverage En))mM
i l
27025.0 10 BRT 1
1 close at hand, can be found in the implementation plan for 2 the severe accident policy sta tement, SECY 8676, which 3 clearly spells out the intent of the Statt to use the 4 Source Term Code package, and other source term-related 5 information, to develop new source terms to replace TID 6 14844. And, in fact, we are working with the Staff now, as 7 the regulatory statt moves out to take the baton trom us 8 and do that.
9 Le t me sum up , Mr. Chairman, by saying that wo 10 would very much appreciate having a letter from the 11 Committee that deals with the subject of the final report, 12 NUREG-0956; we would appreciate having the le tter , in the 13 month ot June.
O 14 I should note that Dr. Ross has noted this 15 morning that, hope f ully, with the letter available this 16 month, he would very much like to provide a copy of that 17 letter in this document as an appendix. As you recall, in 18 NUREG-0772, back about almost tivo years ago to the month, 19 l we similarly took the letter that was provided en un by thn 20 Committee and that was' published in NUREG 0772. We feel 21 that it is appropriate that your comments accompany this 22 document. Thank you.
23 MR. EBERSOLE: Mr. Chairman? I would like to 24 get away f rom the academic aspects of this and into the 25 practical problems that I think Glenn would say the tolks
()
t ace FEDERAL REPORTERS, INC.
] 202 347 3700 Nation *ide Coverage 800 33M6M
27025.0 11 BR7
,O
(_) 1 in the trenches have to face.
2 I have had occasion to ask several operators at 3 different plants a question like this: You have had 4 some thing happen and you have , so far as you know, the 5 beginnings of core molt. You don't know how severe. It 6 may be that half of it is gone , maybe j ust the top 1 foot.
7 What are you going to do? Are you going to pour water on 8 this incandescent portion? Are you going to deprive it of 9 water? What is your action? And almost to a man they say, 10 I haven't the foggiest notion.
11 That's a today problem and I should think that 12 might be one of the first products of a group like your own, 13 to investigate it, to tell the operator in the trenches 14 what to do.
15 MR. SILBERBERG: Dr. Obersolo, that's an 16 excellent point. In fact, we havo grappled with that issue 17 at meetings like this in the past when we have discussed 18 the broader severe accident research program that's 19 described in NU'tEG-0900, Revision 1; and, in fact, wo 20 believe that one of the objectives of our program, although 21 we haven't -- certainly a goal, and we are trying to 22 address it, but we could not provide the answer today -- is ,
23 in fact, to answer that question.
24 MR. El3 ERSOLE : To use llal Lewis' basis for 25 answering questions, it you were emperor, what would you O
ACE FEDERAL REPORTERS, INC.
202 347 37a) Nation *ide Cmcrage Am33MM6
4 i
27025.0 ,
12 BRT 1
f I command them to do? Somebody has to tell them what to'do.
1
! 2 You wouldn' t think they should decide for themselves?
I j 3 MR. KERR: I don' t think he expects an answer to j
} 4 that today -- 6 I
j 5 MR. SILBERBERG: Your charge is well taken.
{
j 6 MR. EBERSOLE: There should be something as the j
- 7 basis for action.
1 i 8 MR. KERR: This, of course, is not the first !
9 time this question has been raised. That's the reason I 10 doubt you'll get an -answer today.
11 MR. EBERSOLE: No, I keep hammering at it every
- 12 chance I get. I would have made an answer up for myself i
i 13 and I know what it would be. If I were an operator, I'd
- 14 pour water on it.
i 15 MR. SILBERBERG: That's certainly one answer.
i i
16 Except we would say, given that, is there any caution with j 17 that? That would be the response, but given that situation, j 18 is there anything we need to find f rom the program that
{ 19 might determine some bounds as to where that might be a problem.
20 A lot of that information is just starting to 1
- 21 come in, so the first response would be your response; sure.
l 22 MR. EBERSOLE: So this may be forthcoming, some i 23 change to that?
I 24 MR. SILBERBERG: Ye s . And we will continue to I
25 charge the program with that request.
{
()
1 ACE. FEDERAL REPORTERS, INC, 202 347 3M Nationwide Co<erage 800 336-6646
1 l 27025.0 13
{ BRT l
sl 1 MR. EBERSOLE: Okay.
4 2 MR KERR: Other questions?
3 MR. MARK: I believe you said the report in the 4 form we have it has gone out to the labs who had major !
! 5 input.
i l 6 MR. SILBERBERG: Yes.
1 7 MR. MARK: They will come back with some i 8 suggested modifications, possibly; maybe serious ones,
, 9 maybe trivial ones. Those will go into what will really be l 10 the final 09567 j
j 11 MR. SILBERBERG: Certainly that's the process.
12 Hopetully it will be more minor than major, but that's yet 13 to be seen.
. ( 14 i MR. MARK: In any event, this document might be t
! v j 15 changed a little bit before it is really final? !
I i
! 16 MR. SILBERBERG: Some. I think that's a fair i 17 s ta teme n t . ,
18 MR. MARK: In the report, I believe there were a
! 19 numbat of places where you reter to poccibic cxperimental 1
20 results or measuromonts or observations which are not quite 21 in hand yet, but will come r maybe in months, maybe in a
_22 year, and which will call attention to the need for further i
23 changes in the report, possibly, or improvements.
24 MR. SILBERBERG: In the methods, yes.
25 MR. MARK: Or might merely provide confirmat(on (
O 1 ,
l ACE FEDERAL REPORTERS, INC. ;
1 202 347 3700 Nationwide Coserase mn34664 1
. _ - - - - . _. _ . _ . = _ - . . . -
i I
l 27025.0 14 l BRT l l
1 that things fit very nicely and we are very happy with it. I 1
l 2 MR. SILBERBERG: Certainly. Yes.
I 3 MR. MARK: When and where would you picture 4 those things appearing? As supplements to 0956 or in 1150?
! 5 They should not be forever detached from this because this, l
6 thing is a thing which is at some stage --
1 7 MR. SILBERBERG: No, that's correct.
j 8 Since the schedule for NUREG-ll50 is sort of ?
i 9 close on the heels of this document, it will largely report i 10 the state of science as of that report, plus the issues 11 that we laid out in chapter 6 will be dealt within some
] 12 additional depth in what we call issue papers in 1150. But j
~
13 then 1150 goes out on the street for comment in the early i
O 14 fall, and so, two things -- two places to pick this up:
15 One would be to pick up additional changes or new 16 information that has como about when the final NUREG-1150 [
f 17 is produced, to say: Hero is where we are now, now 18 information. If it's important enough, we could certainly l
} 19 have it as a supplomont or an, ydu know, a stand-alone 20 report. But right now wo are showing, you know -- here it 21 is '86. I was going to say in 1988 -- and that's only two 22 years away and wo started this over throo years ago, so two 4 s
, 23 years isn't that f ar -- we show on the books an update in i l
j 24 1988 of NUREG-0956. Given all those things, et ce tera , e t f
- 25 cotora, here's how things camo out as of 1988. So right O .
I I
i ACE FEDERAL REPORTERS, INC. !
j 202 347 3700 Nationwide Coserage M k3 M M 46
l 27025.0 15 BRT l
1 now we have such a milestone. Ralph, do you want to add 2 something to that?
i 3 MR. MEYER: I just wanted to say that we really l
4 view the report in near final form and we are not holding l 5 it up at this point for something imminent. >
l 6 MR. MARK: No. I realize that. And the changes i 7 which I mentioned that might come from the labs are 8 presumably almost textual rather than -- '
9 MR. MEYER: I wanted to point out that we have i
10 already gone through a major review and revision with the 11 labora torie s . This occurred about a month ago. So this is 12 a final review at the lab with the lab management, all of 13 which were not involved as these people behind you were in i 14 that process.
15 MR. MARK: I understand that. My point is not l l 16 really raising an objection. This is a status report, i 17 necessarily.
18 MR. MEYER: Yes. E i
19 MR. MARK: And the sta tus will -- [
t i
20 MR. SILBERBERG: Change. l 21 MR. MARK: -- move , slowly perhaps, but somewhat, I 22 and it deserves to be kept in sight somehow. !
23 MR. SILBERBERG I agree and I think we L
24 understand the spirit of your comment. t 25 MR. KERR Mr. Ward? I t
( ,
1 ace. FEDERAL REPORTERS, INC.
202 347 37m Nationwide Coverage *n)4Mi46
(__________________________________ -
_ _ _ ._- . - . . - . _ _ __ _ _ . _ -~ __ _-_ . . _ . _ - - __ ._
! 27025.0 16
' BRT i
MR. WARD:
~
l l You want a letter from the Committee
! 2 this week. I'm not sure whe ther Dr. Kerr or the Committeo I
l 1
3 has any particular problem with that. I guess we'll see. ;
'4 But that's so you can issue the report in table form f i
! 5 shortly? Or is this come action that is awaiting issuing 1
! 6 the report or awaiting the ACRS letter? !
i 7 MR. SILBERBERG: No. We feel that the action is i . ,
! 8 to publish the report in July. Okay? By mid-July. l 4 6 1 l j 9 Fine, that's what we are going to meet. We feel l 1
4 i
j 10 it would be appropriate to have your letter in hand betore 11 we publish. That would be, we teol, not only appropriate ;
l t
i 12 but, I think, a good situation and, as I mentioned, even to ;
i !
j 13 the point of publishing the letter in the report.
[
] 14 So, we go for the milestone. The milestone we ;
15 are going to meet. And the letter, hopef ully the letter !
. 16 will be with it. It for some reason it isn't, we don't i .
! 17 think that would hold un up.
1 18 MR. MARKt They'd just like it.
19 MR. KERR OLher questions? l 20 Suppose I wanted to get a copy of the Source i
21 Term Code package , the one that is presumably the latest
{ 22 version of yours, 0956. !!ow would I ask for it? What i
j 23 would I call it?
I i i 24 MR. SILVERHERG You'd call it the Source Term !
l ,
j 25 Code package. !
j C:)
i I
t
.i ,
I ACE FEDERAL REPORTERS, INC. [
202 )47 3700 Nationwide Cmerage N0lb)M M46 ;
, l l
' l 27025.0 17 '
! BRT l l i i v 1 MR. KERR: Could I get one today?
i 2 MR. SILBERBERG: Yes, you could get one.
j 3 MR. KERR: I mean, is this a report, for example, 4 that describes BMI 210?
I 5 MR. SILBERBERG: Do you want to add to that, ,
i j 6 Ralph? Any of the mechanics of that?
i l 7 MR. MEYER: The code is in its reference form.
[
8 It is available today. The document will be published >
1 i 9 sometime a little later this month, the document being a i
10 code manual to go along with it. ;
j 11 MR. KERR: In the report there's a reference to 12 "a number ot organizations 'have used the Source Term Code l
13 package." Does that roter to the latest version or to O 14 various versions that have existed? Or none of the above?
! 15 MR. MEYER: There are more than two i
16 organizations --
i
! 17 MS. MITCilELL: Only mentioned two, Brookhaven I 18 and Battelle.
1 j 19 MR. MEYER: Btockhaven and Battelle have used t J
l 20 the Source Term Code package and are using the current 21 reference version of the Code.
l
} 22 MR. KERR Those are the two organizations that 23 have used it, what you would now refer to -- what you would 1
i 24 say is the Source Term Code package? ,
l 25 . MR. MEYER: That is correct. And that's the i i
i 4
(:) l I
i I ACE FEDERAL REPORTERS, INC.
j 202 347.)?ts) Nationwide Cmerage 80lk))64 M 6
i -
1 i ,
4 27025.0 18 l
{ BRT ,
1, 1 Source Term Code package that would ' be used in NUREG-ll50.
2 MR. KERR: The second question. Mel, you refer j 3 in the report and you also read a statement that said the i
j 4 regulatory Staff -- you recommend that the regulatory Staf f f 5 make use of calculations in the Source Term Code package, ,
i i
6 provided they took uncertainties into account. I did not i
i 7 see in the report any advice on how to take uncertainties 1
i 8 into account.
I j 9 Does this mean they should be aware of l
10 uncertainties and should use their best judgment? Or do t
) 11 you have advice on how to take them into account?
12 MR. SILBERBERG: It's going to come up again, 7
] 13 but Jocelyn may want to --
1 I
a 14 MR. KERR: Then let's take it in the normal i
! 15 course of the presentation.
16 MS. MITCHELL: I will.
,i 17 MR. CATTON: In the report it references 3.01 as i i 18 being the reference to the code manual. Do you have a -
2 19 little work to do on the references?
l 20 MR. HEYER: We have all the work in the world to -
i 21 do on the reterences. We should have pointed it out.
22 MR. CATTON: Where is the code manual? ;
1 !
23 MR. MEYER: The code manual is in Columbus in J 24 draft form about ready to go to the printers. It hasn't i
) 25 been issued , but it will be issued and we will have a i
i j ACE FEDERAL REPORTERS, INC, j 202 347 3700 Nationwide Coverage 800 336-6646
27025.0 19 BRT
(-
~J l proper reference citation in NUREG-0956 at that time.
2 On most of the re ferences in the review copy of 3 NUREG-0956, they have not been properly listed at the end 4 of the chapter.
5 MR. CATTON: There's one in there, reference 104.
6 MR. MEYER: There's probably one in there that 7 says " reference Lisa." These are just cryptic little notes 8 so we can put in the proper reference citations in the next 9 week and a half.
10 MR. KERR: This is in an effort to prevent 11 information from leaking to foreign powers that might be 12 useful?
13 MR. MEYER: No. It's in an effort to meet the O 14 schedule.
15 MR. KERR: Mr. Corradini, you had a question?
16 MR. CORRADINI: No.
17 MR. MARK: Have they run the Source Term Code 18 package on Chernobyl yet?
19 MS. MITCHELL: It doesn't treat graphite.
20 MR. KERR: There was an interesting comment on 21 TMI 2. It says, "Only recently have the thermohydraulic 22 models been available ," something to that ef fect, which 1 23 assume means this has also been used to look at TMI 2.
24 Ma ybe that will come up.
25 MR. SILBERBERG: It will. We plan to do O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Cover. ige 800-336-6646
=. ~. - - .-- . - - _ -. ._. . --
4 l _27025.0 20 i
BRT 1 something about that.
l 2 (Slide.)
3 MS. MITCHELL: This is a figure that was in the
) 4 draft report, stopping right here, and it has been changed 5 to add to it the information about NUREG-ll50. Throughout,
-)
! 6 the report tries to draw the line between what does I'
7 NUREG-0956 do and what will 1150 do. Everything always i
3 8 builds to the left so that everything builds on everything 9 that came in the past. But the source term revisement 10 study, NUREG-0956, will lead to some source term-related 11 changes, but there are a lot of plant regulatory 12 implementations that will await the risk implementation and
) 13 uncertainty information that is coming out of the 1150 i
- 14 study.
l 15 I brought the list from Zoltan. This is one of 16 Zoltan Ristocci's Vugraphs that gives the potential source 1
l 17 term changes, and I think he's used it with this 18 Subcommittee in the pa st .
l i 19 These three, the near-term: Revised treatment 20 of accidents in the environmental impact statements has 21 already been done with the issuance of south Texas' draft 22 environmental statement. They are considering the use of 23 spray additives in PWRs and suppression pool credit for -
, 24 BWRs.
25 Those are the only three that are the source
- O - -
- ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 21 i BRT 1 term -- so-called source term-related changes. The rest of 2 these, in the intermediate and long term, will await the 3 information from NUREG-1150 on the risk of the five plants 1
] 4 and the uncertainties.
5 MR. EBERSOLE: May I ask a question on the last 6 line? Is it contemplated that control room habitability 7 will have a new base line for the dose potential of a j 8 severe accident beyond the present one , which implies that i 9 the containment is almost pe r fe c t?
10 MS. MITCHELL: I think you are going to have to
! 11 talk to the people from NRR. l 12 MR. EBERSOLE: Okay.
} 13 MS. MITCHELL: As I say, this is Zoltan i
14 Ristocci's slide. He has been down talking about what the y 1
15 intend to do, and I'm sure when he comes back he taay have i
16 some information.
I 17 MR. EBERSOLE: We have taken a very dismal view 18 of the source term inside the containment, 1400; but then 19 we have canceled it by the proposition that the containment 20 is virtually perfect and it may be an inverse of this 21 process that we have to face.
22 MS. MITCHELL: You may be right. No discussion --
23 no quarrel with that.
I 24 (Slide.)
25 This is also a figure that was in the draft 4
h
/\CEJFEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80fb3346646
27025.0 22 BRT i
U 1 report, or its predecessor was in the dratt report. It has, 2 again, been changed. Is it just my eyes or is this really i 3 badly out of focus?
4 MR. KERR: That's good.
5 MS. MITCHELL: Be tte r?
6 The pieces of information that we now say are 4
7 the support for NUREG-0956 are somewhat dif terent from the
. 8 items that were on that chart before. Specifically mi ssing ;
i 9 from these little bubbles here that were there before are 10 the reports of the containment loads working group and the 11 containment performance working group, with the idea that
- 12 the behavior of the containment is something that, for
! 13 these fission product release and transport codes, like the 14 others of its class, are determined outside the scope of l 15 the calculation and input to it, that is, a failure of I
i 16 pressure, a failure of time -- are de termined f rom
[ 17 architect engineering evaluations or wherever, sources of i i
18 information all over. And a containment of entry would be 19 constructed , and the fission product release and transport i
j 20 calculations are designed to provide information about bins 21 of sequences that represent sequences and events on the l 22 containment event tree so that we do not assess the r
1 -
23 behavior of the containment, we assess when a criterion is j 24 met. And we assess what its impact, then, is on the 1
J 25 release to the environment. But we do not determine the i
($) -
i
{
ACE-FEDERAL REPORTERS, INC.
. 202 347-3700 Nationwide Coverage 800-336-6646 ~
[
27025.0 23 BRT 1 criteria that are input to the code.
2 However, there are a lot of other things that 3 provide backup for the source term reassessment study, 4 beginning with the development of the original codes in 5 BMI-2104; the beginnings of the uncertainty studies. This 6 was a study in depth, but for only three sequences for two 7 plants, so it was not very broad.
> 8 The status of validation report from Oak Ridge, 9 as you will see in just a few minutes -- and I don' t want 10 to steal the thunder from the next speaker, but we did try 11 to pick up information that was either in this report or 12 was developed subsequent to that report, to make the 13 document now a technical basis document and not just 14 a ttempt to include the status of validation by reference.
15 The report that will contain the new sequence 16 analyses done by Battelle Columbus Laboratories will be 17 published at the time of the publishing of NUREG-0956 tinal.
18 We 'also foun'd it necessary to institute a 19 quality control or quality assurance procedure for all of 20 the source term calculations. ,
21 MR. KNRR: Excuse me. Does that terminology 22 imply that at lea,st insofar as this study is concerned, 23 quality con' trol and quality assurance are synonymous? Or 24 are they two separate activities?
25 MS. MITCHELL: There were several activities O .
s
' ACE-FEDERAL REPORTERS, INC. l 202-347 VT00 Nationwide Coverage 800-336-6646
' ' )
, .-. - - . _ . .__ _- ,. __.J.._ _. , _ _ . . , , _ ..
27025.0 24 BRT
/ 1 that belong in there.
2 MR. KERR: The word I meant to use was " separa te . "
3 Are quality control and quality assurance being treated 4 se para tely?
5 MS. MITCHELL: I'm not sure we've thought out a 6 definition.
7 MR. MEYER: I'm not sure it's clear on the 8 distinction. Quality assurance is a procedure and quality i 9 control is what you do to implement it?
. 10 MR. SILBERBERG: Dr. Kerr, I think " assurance" 11 could just as well be substituted the re .
r 12 MR. KERR: You are not distinguishing between 13 the two in the language you used? That's all I wanted to C) 14 know.
l 15 MR. SILBERBERG: We'll explain what we did.
16 MS., MITCHELL: What we did was several things.
17 The first was to verify the code. That is, having received 18 '
the FORTRAN statements that represent the Source Term Code 19 package, they went through and attempted to 'say: Given 20 that this code wanted to solve these equations, then these 21 FORTRAN statements will indeed do that.
22 The second thing was, given that we put up' this 23 code on our machine as it was given to us and run-the exact j 24 same input, we do, -indeed, come out with exactly the same l 1
25 answer at the end. l l
($) ,
ACE-FEDERAL REPORTERS, INC.
202-347-3700- Nationwide Coverage 800-336-6646
_ _ _ _ _ ~_ _ __ . - - _ . _ . . _ - . . - . _ __.._,,-_ __ . _, _ _ _ . - , _ _ _ _ .
27025.0 25 BRT
(~s v 1 And third was an effort to look at the input for 2 the Source Term Code package runs. Does this really 3 represent the size of the Peach Bottom containment and not 4 represent the size of some other containment? Are the heat 5 sinks what we think? Is the initial inventory of fission 6 products what we would conclude?
7 So, in that sense it was really an audit 8 calculation. They looked at the input for all of the 9 sequences and they went through and redid from the 10 beginning, looking at assumptions for four calculations.
11 You will see the results of two of them reported in chapter 12 4 of NUREG-0956.
l j 13 MR. WARD: That sounds like QC to me. I don't l 14 know what you think.
t
- 15 MR. KERR
- I just wanted to know if they were i
i 16 distinguishing between the two. The answer was no, they 17 seren't.
18 MR. DAVIS: I have a related question, 19 Mr. Chairman. Your slide implies that that sort of quality 20 control was employed in the 0956 calculations, but I 21 noticed on page ES-2 and also 1-6, the statement that you 22 do not endorse the source terma calculated in this document.
23 MS. MITCHELL: Yes, this is true.
24 MR. DAVIS: Because they have not been reviewed.
25 MS. MITCHELL: There are indeed some aspects of n
\-)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 26 BRT
~
l the input that are controversial.
2 MR. DAVIS: That will always be the case.
3 MS. MITCHELL: Yes. So that when you have to 4 put in the size of the pool of corium on the floor of the 5 containment af ter the core has melted through the bottom 6 head of the vessel, that particular input is subject to a 7 lot of interpretation, a lot of disagreement, and that will 8 influence the result that you get f rom the source term.
\
9 To the extent that, indeed, the size of the 10 containment that was input to this calculation does, indeed, 11 represent Peach Bottom, we stand behind it; that we wanted 12 to model a V sequence that had a hole in this particular 13 piping of the size that we will stand behind. But in the 14 context of whether or not the pool of corium on the floor 15 should have been wider ane? shallower or narrower and deeper, 16 is something that should be reviewed in the context of 1150.
17 MR. DAVIS: Okay. I had a little trouble 18 understanding what you meant by "not endorsing the results" 19 because you used the results to draw some conclusions. For 20 example, saying that these results don't confirm the ANS 21 conclusion that across-the-board reductions are appropriate.
22 It troubled me a little bit that you don't endorse the 23 result and yet you use them to draw conclusions.
24 MS. MITCHELL: Most of the definitions of the 25 sequences came from the people from 1150. They wished to O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 27 BRT
("h
\/ 1 have this particular sequence and that particular 2 containment behavior investigated. To the extent that some 3 of the controversial input was, indeed, spe cified for us, 4 we make no statement about whether that is the most likely.
5 If you had 1000 TMHB prime accidents, would they in the 6 majority of the cases go the way we have evaluated? Or are i 7 we looking at what other people may consider to be a very 8 small, unlikely branch of the whole tree?
9 If it is possible that the corium should have
., 10 been narrower and deeper than the input that we have I
l 11 specified, to that extent it will change the source term 12 results that we get and that is something that 1150 will 13 have to answer.
14 MR. DAVIS: Thank you.
15 MR. KERR: Is that perfectly clear?
16 MS. MITCHELL: We put down the three reviews, r
- 17 the kinds of reviews that we had. This particular review 2
18 was held over a matter of years during the development
'1 19 stage.
j 20 Mike, did you have a question?
21 MR. CORRADINI: It goes back to this a little 22 bit. Is part of your QA -- do you require the people that 23 are doing calculations with this group of tools to develop 4
24 what I would call a calculational workbook that verifies --
25 that has input all the calculations they need to get the i O
l l
l ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 l
_ , _ - . . _ _ _ _ . _ , _ _ __ _ - - - _ _ , ,_. 1
27025.0 28 BRT 1 input to do the calculations and then where they got the ir 2 information from?
3 That is, I'll turn it around. If Westinghouse 4- came in to NRC to show that something with ECCS was working 5 properly for some sort of limit, they would have to show 6 some sort of traceability back to drawings for the 7 structure of the plant or set po in ts , e t ce te ra . Here you 8 have a very -- now relatively rigid or relatively fixed 9 network of calculations, so if you put in the same numbers 10 you'll get the same result every time.
11 What I'm more concerned with, now that you have
, 12 done that -- or partly concerned with, is how do you get ,
13 the official input numbers, and have you gone through any QA 14 of the input?
15 In other words, if I want to know the volume of 16 Zion, for example, how do I calculate it? What drawings do 17 I refer to?
18 MS. MITCHELL: Yes, I know what you are saying.
19 For that kind of thing we do have two organizations that 20 look at that, number one, to find it, say, from the FSAR, '
21 what i s the free volume in the containment; and the other e g 22 to check it to make sure that they agree that it does, 23 indeed, re pre sent the plant that you want to model.
24 MR. CORRADINI: So then, you do have a check and 25 a double-check?
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
- -- . = . - . . . - - _ - , , _ . - _ . - . , . , . . , - . , . . , _ . - . . . - , . . . . . . . -
27025.0 29 BRT O 1 MS. MITCHELL: Right. Right.
2 The American Physical Society Review is now 3 published in the Review of Modern Physics, and we have made 4 use of their words of wisdom in changing some of the focus 5 of the research and how we view what is the use of this 6 methodology.
7 We had a five-month public comment pe r iod . The i 8 letter that you wrote to us in December is listed among 9 those comment le t te rs . So this is where -- although you i
10 are not members of the public, exactly, that is where it is 11 binned, as we bin everything in source term world.
12 We did receive comments from 68 individuals and 13 organizations, comprising about 700 pages of input and ,
14 roughly 1600 individual comments. So a lot of people had a 15 lot of advice that they thought we should take.
1 16 MR. KERR: Is the purpose of this slide to l
- 17 demonstra te that the work is on a very firm foundation?
4 18 These represent the foundations? I'm not sure, what is it 19 we are getting at?
20 MS. MITCHELL: It is the type of foundations 21 that we do have. The details of the foundations you'll get 22 a lot in chapter 3, not in the discussion today, but
]
23 physically in chapter 3. But it's just to show -- one of 4 24 the criticisms that was leveled was that it was -- the l 25 whole document before was just a bunch of codes and we want
()
4 t !
4 ACE-FEDERAL REPORTERS, INC. )
202-347-3700 Nationside Coverage 800-336-6646
__ _ _ . . _ . . _ _ , _ _ _ _ . . _ _ _ _ _ _ , , - _ . _ . _ . ~ . _ . _ _ . , _ _ . _ - - . . _ . _ _ . _ .
27025.0 30 BRT
(-
g-}s l to show that this is not the case, it is not just a bunch 2 of codes. There are other things backing it up.
3 Ralph, do you have an additional comment?
4 MR. MEYER: Yes. The source term reassessment 5 is not described in its entirety in one single report, l 1
6 NUREG-0956. All of the projects ind ica ted in the balloons 7 there, I guess there are eight of them, are directly l l
, 8 involved in the source term reassessment project and exist 9 in separately published documents. These are, if you will, 10 the building blocks of the source term reassessment, all of 11 which is brought to te rm , hopef ully in some coherent 12 summary fashion, in NUREG-0956.
13 MS. MITCHELL: I just want to close with a O 14 statement of what we believe NUREG-0956 does and what we 15 believe NUREG-0956 does not do.
16 What we do, is we draw conclusions about the i 17 source term analytical procedure and about the example 18 calculations that can come from them. We also recommend 19 that the Source Term Code package can be used to evaluate 20 regulatory practices which in the past have been built on 21 the reactor safety study or TID 14.
22 844 provided -- and I think Mel also emphasized i
23 that -- that uncertainties also are taken into account.
24 Now, you want to know where are the uncertainties and how 25 do you take them into account?
4 O .
ACE-FEDERAL REPORTERS, INC.
202 347 Nationwide Coverage 800-336-6646
_____.-...___..____,___-..-3700, _ - _ . _ _ _ , _ . . _ . . - - _ - . _ - . . . _ . - - . _ ,
27025.0 31 BRT s 1 The uncertainty evaluations that started with j 2 this one in the KUEAST study are be ing followed in two 3 programs, one at Brookhaven National Laboratory, which is a 4 follow-on but with some important dif ferences from the 5 QUEST study, and one that is being followed at Sandia for 6 the risk calculations for NUREG-1150. Those we do not 7 report on in NUREG-0956, the methodologies.
8 MR. KERR: I probably did not make my question i
9 clear. The advice was, I thought, or the statement, that
, 4 1 10 regulators use these methods provided they took the
~
11 uncertainties into account. I could interpret that to mean i
! 12 that they are ready to use now if you take the 13 uncertainties into account. What I looked for was i O 14 some whe re that says: Here is the way that we recommend you 15 take the uncerta in tie s into account. I did not find that.
16 MS. MITCHELL: You didn't and you won't.
17 MR. KERR: Then that sort of means to me that 18 the advice is a little bit gratuitous , because if you tell 19 somebody to use it if they take the uncertainties into 20 account and then you don't give them any advice on how to 21 ta ke the uncertainties into account, it strikes me it's not 22 extremely useful.
23 MS. MITCHELL: The problem is that I don' t 24 believe there is any one simple answer, because there are 25 10 areas. The uncertainties will, indeed, hit the
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27025.0 32 BRT
- 1 different areas in different ways.
2 MR. KERR: I'm simply asking, if you tell NRR, 3 let's say, you can use these, but take the uncertainties 4 into account, it seems to me a natural question is: Where 5 do I go to get your advice on how to take the uncertainties 6 into account? What you are telling us, I think, is we 7 don't know how to take the uncertainties into account.
8 There isn't anything wrong with that answer if it's the 9 right answer. But it seems to me it is not very useful l
10 adv ice to someone to say, you can use this if you take the 11 uncertainties into account, and then tell them, but we a
12 don't know how to take the uncertainties into account.
13 MS. MITCHELL: I can visualize for the 10 areas 14 this could be nine different uses of the uncertainties.
15 In some cases you will find that there is no 16 decision that is able to be made , because the uncertainties 17 are too large for you to make any useful decision. We are 18 working with the regulatory Staff to try and iterate on 19 this, to try and use --
20 MR. KERR: But the implication seems to be that
- 21 the NRR people will know more about how to take the 22 uncertainties into account than you will. That strikes me 23 as being a little --
24 MR. WARD: I guess I can see -- you are saying 25 that the treatment of uncertainties is an applications O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
. _ . - - _ _ . . _ ._ - . _~. - - _ , - _ . _ _ , . . . . - _ . _ , _ _ . . _ _ . _ _ _ , , . - _ . , - .-
27025.0 33 BRT 1 problem?
2 MS. MITCHELL: Ye s .
3 MR. WARD: Even at the NRR. Yet you need to 4 supply NRR with what you know about uncertainties.
5 MS. MITCHELL: There are two f acets to the 6 problem. One is the uncertainties as they exist in the 7 analytical tools that have been given to the user; the 8 second is the application to which somebody wants to put I
9 them and both people had better have input into the process
} 10 of using it to make any changes.
j 11 MR. SILBERBERG: I think what Dr. Kerr is
, 12 asking -- I think there's an answer to it -- forget about j
13 the specifics of the application , that obviously they'll 14 have to do for each application. But is this a how-to?
I' 15 How would they go about getting some of this information?
i 16 And I think the answer to that is the first place that they 17 would look that has explored the bounds of the se 18 uncertainties, and I think you have seen it, is in the 1150-19 work.
20 1150 will have as a state of the art an 21 assessment of what the uncertainties look like for these 22 various plants and these various sequences, and one can use 4
23 that to get insights to ' be able to say -- I'have some idea --
24 I guess what Dr. Kerr is saying: How do they get the ir 25 arms around uncertainties, magnitudes -- ,
O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage - 800 336-6646
. ~ _ -. _ - _ - - - .
27025.0 34 BRT 1 MR. WARD: I can see where you might say 1150 2 can provide some assessment of the importance of 3 uncer ta intie s , but I don't see how you can get any 4 information on the uncertainties from 1150.
5 MR. SILBERBERG: There has been some attempt, 6 preliminary and limited, to actually take a look at that
- 7 and I think maybe Joe Murphy wants to add something on that i 8 that I overlooked.
i
) 9 MR. MURPHY: I think what you said is what we
.)
1 10 hope to do. What we hope to do in 1150 is to give a i
j 11 characterization of the uncertainties, more of an i
! 12 engineering approximation of what reasonable bounds are 13 rather than precise crystal estimates, because of the O 14 amount of subjectivity that is involved. We described thic
'I 15 in some detail in the Subcommittee meeting last month.
16 In 1150 we hope to have a section in there that 17 will describe how to use this in the regulatory process.
18 This will build of f of what we have already written in 1050 I 19 on the same subject.
20 The problem you run into is the way you handle l
21 uncer tainties dif fers with each issue as it comes up. In 22 some issues you can reach regulatory judgment with a high 23 degree of uncertainty and you can put it in several 24 different ways. But as you shift from issue to issue, the
, 25 way you handle the uncertainties in making the regulatory O
ACE-FEDERAL REPORTERS. INC.
202-347-3700 Nationwide Coverage 800 336-6M6
. .. . =, . . . . . -- . --_ .-
4 27025.0 35 BRT i
i 1 decision changes. We tried to address this in 1050, and 4
1 2 when we got all done it was about the best job we could do, 3 but it sounds like motherhood.
4 We are going to try in 1150 to give five 5 specific examples of how we think uncertainties should be 6 considered in the process, to try to explain the words that 7 are more philosophical that are in NUREG 1050.
8 MR. KERR: Thank you, sir. I should comment 9 that it has always been my understanding that motherhood is 10 fairly certain. Fatherhood may be uncertain, but 11 motherhood is much less so. Mr. Catton?
12 MR. CATTON: Just a comment. It seems to me the 13 document we have describes the codes. The people who wrote i'
14 them and who put them together were contributors to this
- 15 document. It seems to me that the people who did that are 16 the ones that ought to say whether their work is good, bad, 17 or indifferent and somehow out of that comes an uncertainty.
18 I don't understand this separating the uncertainty away i
i 19 trom the tool itself and letting somebody else do it.
i 20 MS. MITCHELL: The people who are involved in 21 having developed the fission product release and transport 22 methodology are the very people who are involved in the l
23 1150 attempt to make a quantitative assessment of the 24 importance of uncertainties. So, indeed, they are involved 1
25 and indeed they are bringing to bear their best judgments
, D)
\,-
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800 336-6M6
.- .- . . . _ - - . . . .. -.- -- - - . - - _ . - ~ __
27025.0 36 BRT 1 of what is the sta te of the art, so what you say should be 2 done is, indeed, be ing done.
3 MR. CATTON: I may have a misconception on how 4 to write reports, but it seems to me if you write a report 5 it ought to be comple te . If you describe a tool, part of ;
6 its comple teness is what you think of it. If you don't 7 have it, in my view, it's incomplete. Tha t's the only 8 conclusion I can come to.
9 MS. MITCHELL: We have asked for the sta temen t 10 f rom the individual laboratories that were involved as to 11 whether or not they do stand behind what we said about 12 their work. So they will be on the record with their 13 opinion of it. The problem is that PRA methodology comes O 14 in pieces, beginning with the assessment of the core melt 15 frequency and going on through to the consequence i
- 16 calculations at the end. This particular document is one 17 piece and one piece only, which leads me into my last point 18 that I wan ted to make
- What d s NUREG-0956 not do?
f 19 MR. CATTON: But the title says, " Rea sse ssmen t i
20 of the Technical Basis." The technical basis has to 21 include the uncertainty or you haven' t done this. It !
22 should have another title.
23 MS. MITCHELL: This is no one uncertainty.
24 Something that may be unimportant for one plant and one 25 sequence and one purpose, may be the driving factor for
! ()
1
! ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
t 27025.0 37 BRT O 1 another purpose for that exact same plant. This is a 2 problem, that there are more views of uncertait;ty.
3 Mike , did you have a question?
4 MR. WARD: I think you are using the term i 5 " uncertainty" maybe in a couple ot different ways. You 6 seem to be talking about uncertainty in the regulatory 7 application where at other times you talk about -- people 8 talk about uncertainty in the scientific data base , source 9 te rm -- those are two different things.
10 MS. MITCHELL: One is a subset of the other, for 11 the regulatory basis, using it in the regulatory arena, 12 then the uncertainty in the source term science is one i
! 13 aspect along with what is the volume of this containment
,' 14 and what is the geometry of the pool, ex-vessel.
1 15 MR. WARD: It seems you have a responsibility in 16 0956 to talk about uncertainty in the source term science, j 17 just se para te .
l 18 MR. MEYER: We do indeed.
i 19 MR. SILVERBERG: We talk about it quantitatively l 20 not qualitatively.
21 MR. WARD: And some of the basis for that has to 22 come from the Sandia report; right?
23 MS. MITCHELL: Yes.
24 MR. SILBERBERG: But the fact of the matter is 25 the quantification of uncertainty, in other words, the I
l ACE FEDERAI. REPORTERS, INC.
{ 20 347-3700 Nationwide Coverage 800 336-6646
27025.0 38 BRT (D
(/ 1 procedure, discipline, work necessary to do that is in fact 2 really j ust getting under way.
3 In other words, Joe Murphy referred to 4 subjective uncertainty in the discussion ot getting people 5 who did the work together to make judgments of ranges and 6 so forth. But, in fact, as stated in the report, we 7 discussed the area -- areas of uncertainty, and the 8 concerns. But their quantification is, in fact, at a very 9 early sta te .
10 MS. MITCHELL: Yes.
11 MR. SILBERBERG: And that's a statement of fact.
12 MR. KERR: For those who have not been involved 7-13 in all the preceding meetings, we have immediately said
'x_/
14 that we thought 0956 ought to have quantitative estimates 15 as best you can get of uncertainty. We have repeatedly 16 been told that that will be done in 1150. I don't know 17 whe ther this is schedule-driven or order-driven or 18 something else, but I think that's the way things are. We 19 continue to believe and ought to make the point, but I 20 believe the people that prepared these two reports made a 21 decision that only qualitative decisions about unce r ta in ty 22 will be in 0956, and qualitative will be in 1150 or i 23 subsequent reports, if it becomes too difficult to complete 24 before 1150 is released.
25 MR. WARD: Does it make sense to you that the j
()
I ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8 % 336
__-__--___.__-___________-_-____-_-____--_--_-____-__-_-__-_______-______-_-6686_ - - _ _ _ _ _
27025.0 39 BRT
,~
k-) 1 quantification or discussion or estimate of uncertainty in 2 the source term science will come out of the 1150 work?
3 I'm asking you.
4 MR. KERR: I have said repeatedly in news 5 reports, le t te rs , informally, every way I know how, that it 6 does not make sense to me, but there we are.
7 MR. MCYER: May I comment on that, please?
8 MR. KERR: Certainly.
9 MR. MEYER: First of all, let me point out, I am 10 not aware of any other methodology that we use in the 11 licensing practice that has quantitative uncertainty built 12 into it. I think that this is probably one of the first s 13 attempts to seriously quantify the uncertainties.
14 As Jocelyn has tried to point out, the 15 quantification of uncertainty is, in fact, related to the 16 application.
17 The first application of the Source Term Code 18 package is the risk study. And the risk study is to be 19 re ported in 1150.
20 Here is an application. It's specific. We now 21 know what the goal is and how to approach the uncertainty 22 evaluation, and it is being done. Uncertainties are being 23 estimated quantitatively for the analyses in this 24 application of the Source Term Code package.
25 , For all subsequent applications, we recommend
(")
m ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27025.0 40 BRT 1 that they likewise be assessed - quantitatively considering 2 the application. We also offer our services to NRR and anyone 3 else in the Agency who tries to do that, in order to help 4 them accomplish that.
5 I don't think it is out of place that the tirst 6 appearance of a quantitative estimate of uncertainties 7 appears in the first application of the Code package.
8 MR. SILBERBERG: 1150 is not going to have 9 quantitate assessments in it. That's a summary 10 sta tement --
l 11 MR. WARD: I still think you are mixing up 12 uncertainty in the data base , in the science , and j 3 13 uncertainty in the regulatory application, which results 14 partially --
15 MR. MEYER: I'm sure there is some confusion on 16 that, but I thought we have -- we have, in fact, addressed 17 uncertainties in the phenomena, in a qualitative manner, 18 throughout the report. We discuss the modeling of physical ;
19 phenomena. We have discussed -- we believe candidly --
20 areas where our knowledge is limited and therefore where 21 our models are uncertain, because we don' t know how to 22 build a model that accurately describes the phenomena.
23 Those are discussed in the report. We go so far as to l
24 catalog them, to tabulate them, to plan a research program 25 to try and address those areas. That's the phenomena part O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
_. - - .~ -_ _ __
27025.0 41 BRT t .
! 1 of it, the qualitative part of it which we do. We do that
- 2 with judgment, and end up in our judgment, believing that 3 the Source Term Code package is in fact a fine tool for l
i 4 today's analysis of source terms.
i 5 Then we go further to try and quantify the 6 uncertainty in this first application.
l 7 I think this is an admirable effort in trying to 8 nail down the uncertainties.
j 9 MR. KERR: Le t me see it I understand. People 10 who are doing the 1150 study will have to take into account
{
11 uncertainties and a lot of things, including the scientific 2
12 bases for 0956. Many of them have worked directly in the 13 organizing of 0956 activities, so they know how to do this t
l 14 from firsthand experience. But suppose you turn this over l 15 to somebody other than National Laboratories? How are they 16 going to know what the uncertainties in the scientific 17 bases are, because they won' t find it in 0956?
18 MR. MEYER: That will be a big problem. I think 19 we have chosen our words carefully, such that we are not 20 suggesting that this package of codes be forced on the I
21 industry or some independent applicant. It is being 22 recommended for use by the Agency. And we have the 23 support --
24 MR. KERR: It's not a question of it being 25 forced upon them. We have spent a great deal of ef fort and
()
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6 l
27025.0 42 i BRT i 1 tunds as taxpayers, to develop a method which should be 2 useful to anybody who wants to use it. Af ter all, we are 3 not trying to keep secrets from people. We are trying to 4 make reactors safe or at least to see how safe they are.
5 We aren't suggesting that it be forced on anybody. But it 4
6 seems to me it isn't unreasonable to suggest that this be :
7 at least scrutable to people who want to use it.
j 8 MR. MEYER: I think it is scrutable to such 9 people. I think quantifying the uncartainties is a very l
10 difficult job.
4 i i 11 MR. KERR: Of course it is. It's expensive .
> l l 12 MR. MEYER: -- and would be extremely dif ficult
. 13 for a user without support from the National Laboratories.
1 O 14 MR. KERR: And therefore it ought to be as i
15 understandable and usable to as wide an audience as .
1 I
j 16 possible and not just continue to reside in the National 17 Labs.
r j 18 MR. MEYER: I agree with that and I think we
! 1 19 have gone absolutely as f ar as the state of the art will l ,
20 permit. '
21 MR. KERR: Let me come in with something i
i 22 peripherally related. .
23 You said this is the first time in the-1 24 regulatory arena that any effort had been made to quantify, '
25 and if you look at that statement again, you'll realize it I
(2) i ACE-FEDERAL REPORTERS, INC. ,
202-347-3700 Nationwide Coverage 800-336-6646
1 27025.0 43 BRT f ,
4
! U 1 might not be correct.
, 2 For example, the design of containments, for j
3 example, for given particular temperatures and pressures,
- 4 use codes which have been in existence for years and which I 5 have built in to them a great deal of allowance for ;
i 6 uncertainties, as evidenced by the fact that these things ,
7 were designed for pressures of 50 psi will probably 4
8 withstand pressures up to maybe 125. So it isn't correct 9 to say the uncertainties have not been taken into account.
10 MS. MITCHELL: I don't think we said that. He j 11 said that they don't know --
$ 12 MR. SILBERBERG: In a formal way.
j 13 MR. KERR: But they have been taken into account 14 quantitatively by those codes and the ' development of those 15 codes has been done carefully over the years and there's 1
16 quantitative considerations.
i 17 MS. MITCHELL: I think if that were true and you i
18 went to the architect-engineer and said: What is the i
19 pressure at which this containment would fail, they would i
) 20 be able to give you the answer.
l 21 MR. KERR: That's not what they are supposed do.
22 They are supposed to give a design that will operate with
< 23 high assurance at a given pressure. That's what they have
- 24 done over the years and that's what they can do.
25 If you give them a different task, they might be
- ACE-FEDERAL REPORTERS, INC. ,
i 202-347-3700 Nationwide Coverage 800-336 6646
27025.0 44 BRT O 1 able to do that as well and, indeed, that's what you are 2 trying to do and that's what makes it new. It doesn' t make 3 the uncertainties any less important. It makes them 4 different.
5 But the AEs have never been asked to design for 6 failure. They have been asked to design for nonfailure.
7 MS. MITCHELL: This is true. But I think if 8 they had a handle on exactly what margins and uncertainties 9 were inherent in the method,.they would be able to answer 10 the other question.
11 MR. KERR: Well, maybe you think that. But the 12 point I want to make is that they have made use of 13 uncertainties in dealing with the task that they have 14 undertaken and they have done it. And the uncertainties 15 have been very quantitative -- they have developed them 16 over the years. But they have not been asked to design for 17 a failure pressure. And it's a tougher task, but it's also 18 one of the things, it seems to me, that makes it desirable 19 that every bit of information that's developed, we ought to 20 somehow have available to a wide audience of use.
21 MS. MITCHELL: In a way your example is an 22 excellent one for the f act that it is_ difficult, without
)
23 knowing a particular application, to say anything about 24 what the uncertainties are; for one application, that is 25 designing for not failing, you have uncertainties in one O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8043E6646
27025.0 45 BRT
/G
(-) I way. If you are designing or want to know when it fails, 2 you have uncertainties that hit it in another way.
3 MR. KERR: I agree with you wholeheartedly, and 4 I don't think you could have done anything very meaningful 5 about uncertainties without taking some examples. That 6 would have taken longer, it would have taken perhaps more 7 sources, but it seems to me the result would have been much 8 more meaningful in terms of being able to say to NRR: Here
, 9 is the way we would recommend that you do it. Hete's on 10 example.
11 It's easy to second-guess people , of cc irse .
12 MR. SILBERBERG: Mr. Chairman, I would like to 13 ask Dr. Powers from Sandia, who in fact, managed that and 3
\
/
14 therefore has in s ig h ts , perhaps he can put some of this in 15 perspe c tive .
16 MR. KERR: Dr. Powers?
17 MR. POWERS: I'm Dan Powers from Sandia 18 Laboratories, and I do want to agree with what Mr. Ward 19 says, that we should distinguish between the uncertainty in 20 the sc ie nce , calculating source terms, and uncertainties in 21 those computed results. Beca use there is an element of 22 science and this is an element of using large computer 23 codes to describe the plant. That point was well made. I 24 think Jocelyn's point, too, is when we get down to 25 computing the results for a particular plant one has to Oa ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4646
27025.0 46 BRT
] 1 recognize the applications you are going to make.
2 I would like to turn to this question of 3 summarizing the uncertainty of the science and just point
]
4 out to the Committee what a formidable job you are talking 5 about.
f 6 In the QUEST study we attempted to take into I 7 account and quantitatively evaluate the level of
- 8 uncertainty we had of about 12 major physical phenomena;
, 9 things like cupproccion pool performance; things like a
10 release of fission products from the fuel. That i
f 11 documentation of the technical bases for uncertainty i 12 studies on just 12 topics occupied about 610 pages of i
13 written material.
O 14 In order to no that careful quantitative j 15 summarization of the uncertainty in many, many aspects of 1
i 16 the physical phenomena that are required to take into 1
17 account means you have to have a formidable document. I i
18 think just to keep the document tractable, the authors are 19 forced to summarize what we know about uncertainties in j 20 very concise fashion and refer you to where we have source i
j 21 of more quantitate evaluations, and those are the completed 22 QUEST study, the ongoing 1150 study, and the recently
-,I i, 23 started QUASAR study.
24 In talking about uncertainties, I think it's 1
l 25 important for the Committee to recognize one does not just j
) ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nation *ide Coverage 800-336-6646
27025.0 47 BRT
(/ 1 go out into the lab and measure an uncertainty.
2 Uncertainty inherently means something you do not know.
3 In the QUEST study we did attempt to identify at 4 least three areas of uncertainty, though one was collapsed --
5 two of them are collapsed into a single area. One is 6 uncertainty of inputs, which is a job that is user -- only 7 the user can really handle uncertainty in his inputs. He 8 can only make an assessment of the uncertainty of his 9 inputs. It's not something the codes can have built into 10 it beca use the code does not pre suppose the application.
11 Uncertainty in the models, which, as Dr. Catton 12 pointed out, really the authors of the models ought to be 13 able to make some estimate of what assumptions they made
' }
14 and consequently what uncertainties ought to be a ttached to 15 the results of their models. The results in NUREG-0956, I 16 think they do exactly that.
17 I would like to point out, for one of those
'l 18 model, the VANESA model, the description of the assumptions 19 and consequent uncertainties that one ought to think about 20 in judging the model, occupy something over 300 pages of 21 written text. Consequently, a complete summation in 0956 22 again appears impractical.
23 Finally, we identified there were uncertainties 24 in things we just simply haven't taken into account. That i
25 is, not only models of physical phenomena but aspects of
.i
- O a 8 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33MM6
27025.0 48 BRT
- 1 the coupling of physical phenomena, and those aspects of 2 the things -- simply, we don't know what those things are, 3 are -- neither.summarizable nor quantifiable, in any kind 4 of mechanistic sense. That's the only point I'd make.
5 MR. KERR: Mr. Catton? I believe you had your 6 hand up first.
7 MR. CATTON: Just reading through the report, 8 312, first sen tence in the last paragraph, "Many thermal
- 9 hydraulic hypothocos arc wcll undcratood and models are 10 statistically supported by data base."
11 In another place it says that " direct code to 12 data comparisons."
13 Ye t , nowhere where these statements are made do O 14 I find any references to what was done , where it was done, 15 or how relevant the comparison was. To me that's a lack of 16 rigor in the document. Considering its importance, I think 17 you ought to fix that. That's a statement.
18 MR. KERR: Mr. Corradini?
19 MR. CORRADINI: It really comes down to the user. !
20 Are you envisioning that j ust Brookhaven and Battelle will i 21 be users or are you-looking for wide use?
22 MS. MITCHELL: They are today the only users of I
23 the Source Term Code package; however, there are other i 24 people that use all of or part of the Battelle codes.
25 The British have used the TRAPMELT a great deal.
1
(:)
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 49 BRT 1 So, no, we don' t envision that only Brookhaven and Battelle 2 will be the people who use the Source Term Code package.
3 MR. CORRADINI: I understand what is being asked.
4 If, tomorrow, I said I wanted to use this thing, I would be 5 afraid to. One, because once I got it, I wouldn' t know 6 what to do with it. There's no documentation of what's in 7 it. I know individual pieces, but there's no documentation 8 of what's in it, how I go about starting, and the normal 9 method I would accumo is, if a laboratory had Lv do it, 10 they'd say, send me all of this stuf f and send me some 11 input that works and I'll try to understand what works and 12 modify what works a little bit to get it going on a
- 13 different problem.
G 14 MS. MITCHELL: That's what we did when it was in 15 Columbus to Brookhaven is we sent along the standard 16 problem so that you put up the code and you run that input l 17 and you get the same answer.
18 MR. CORRADINI: When I think of uncertainty, I 19 think of a lot of those uncertainties. But the first one 20 is, do you even have the ability to do the calculation with 21 this very monstrous tool or groups of tools.
22 So one things I saw which was missing here , and 23 Professor Catton showed it to me in terms of a hand 24 calculation on table 3-4, is you show the code package and 25 the hand calculation.
em a
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationside Coverage- E,00-336-606
27025.0 50 BRT
{T, Al 1 For example, if you feel it's impossible or 2 should not be in this document to designate the 3 uncertainties of certain pieces of physical processes which 4 may be important in one application or another, it would be 5 very useful as an appendix to just explain exactly how this 6 hand calculation was done for this specific example , as 7 different to the code package calculation. Literally an 8 appendix that went through detail by detail how the hand 9 calculation was done. That might give the potential user 10 some indication of what he might have to go through if he 11 did hand calculations.
12 On a side point, they may choose to do the hand 13 calculations.
14 MR. MEYER: Let me reply to both your comments.
15 The first about a documcnt that tells you what's in the 16 code, then how to set it up and run it, is being prepared.
17 It's just about ready to go to the printers. I want 18 further detailed information -- the people from Battelle 19 can tell you about it. It has sample problems in it, 20 inputs, outputs, run the code, see it works.
21 The hand calculations are also documented in a 22 report that's about an inch and a half thick that will be 23 issued as a NUREG CR report prior to the publications of -
24 NUREG-0956. It may already be at the printers, I'm not 25 sure, but it's imminent.
O ACE-FEDERAL REPORTERS, INC. l 202-347-3700 Nationwide Coverage ' 800-336-6646
i 27025.0 t 51 BRT 1 MR. SILBERBERG: But in fact it exists.
2 MR. CORRADINI: Then I'd agree with what Ivan 3 said. In a lot of places you make statements which 4 probably have snme basis in a past calculation document, 5 but you have to be real careful to addend references 6 because if Yt's. not going to be a reference here, there has
/ ,
7 to be trace,1b111ty back. .
There were a lot of places where 8 there was_not, traceability -- where we couldn't see 4 w .
9^ traceability.
i 10 MR. KERR: Did you want to make a comment?
11 ,)ilR; BAGCHI: No. I just wanted to make note <
12 this report was available, soon to be-available.
,, .. ~ r 13 . MR. KERR: Re ferring to?
O 14 MR. BAGCHI: The users guide for the Source Term 9
Codo p'ackage.
- 15 x.
t 16 MR. CORRADINI:- nnd th5tm is a repor's on the j .
17 hand calculations. .,
s -
18 MR. CATTON: 'I would really like tD+see a. copy 1 -
lA of the reports on~ the hand calculations. There is no 20 reference in 0956.
21 MR. MEYER: The references aren't fixed up yet.
k, x 22 but it will be all fi.xed up.
l 23 MR. KERR: What I mean is could somebody -- if
~ '
24 we order a report, could somebody tell us what to order on
, 25 the hand calculations?
+
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage -
'800-336-6646
- ,~ . . , . - . , - . .-. -, _ _ , , - . . - . - - -
27025.0 52 BRT O. 1 MR. MEYER: .'m sure we can provide you with a 2 draft copy of that report if you'd like it but it's going
, 3 to be published within a few wee ks.
4 MR. KERR: It does not now exist in published 5 form?
6 MR. MEYER: That's correct.
7 MR. KERR: Mr. Davis?
8 MR. DAVIS: I had a specific comment on 9 uncertainty that you might want to consider in the rewrite.
10 In this section you have some comparisons between the 11 Brookhaven and Battelle results. These are tables 414 and 12 415. And you say the results are well within the 13 unce r ta in ty.
O 14 MS. MITCHELL: Yes.
15 MR. DAVIS: I don't know what that means. Which 16 uncer tain ty? Because you show ruthenium release to be four i
17 orders of magnitude ditferent in the two calculations.
18 MS. MITCHELL: -There was a change in the 19 technology in between one, the result of the BMI 2104 code i
! 20 sweep, which looked at the available experimental data and 21 chose to use its model to hit some values that were decided 22 later to have been cases where the oxidation could have 23 been taking place, and so what you are seeing is the 24 vaporization of ruthenium oxide. And other data were l
25 available where it * * '1 be more nearly the kind of i i /~T
(/
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646 I
, __ _. . _ - _ _ _ - . _ __. -- - _ . _ . .- _ . , . _ , ~ _ . , _ , . _ _ . _ . . - _ . _ . .~.'
i- !
i
- 27025.0 53
- BRT 1 environment that you might find inside the vessel, so that 2 this was a change in the CORSOR M coef ficients such that
- 3 they do give you a four orders of magnitude. There are 4 data that would support both of those.
5 MR. KERR: So the uncertainty is four, as a 4
l 6 magnitude?
7 MR. DAVIS: You don't know which answer is right.
8 Is that correct?
9 MR. MEYER: We believe the lower answer is right 10 because of the oxidation potential in the --
11 MR. KERR: What we are saying, the results being i
12 uncertain, this would seem to imply the uncertainty is as 13 far as magnitude.
O 14 MR. MEYER: It is indeed. If you look at the 15 f igure that shows comparison of CORSOR code with some data, 16 you'll see ruthenium data spreading up 2, 3, 4 orders of j 17 magnitude.
18 MR. DAVIS: I'm not arguing that the magnitude 19 could be that large. I'm just trying to get a feel for a
20 what you mean by these results being within the range of 21 uncertainty. You are saying either answer could be correct; l 22 is that right?
23 MR. MEYER: Dr. Powers is eager to answer this.
24 MR. POWERS: This is a very important issue.
25 Yes, there's orders of magnitude different between i C:)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 54 BRT p
k) m 1 ruthenium releases, but both numbers are damned little.
2 Those are the same numbers, as far as source term 3 consequences of any type, as far as ruthenium release. ". he 4 differ between 8 times 10 to the minus 7 and 9 times 10 to 5 the minus 3, that's a trivial ruthenium release.
6 What that says is within the uncertainty they 7 get exactly the same numbers.
8 21R. DAVIS: Maybe that needs to be explained in 9 the te x t . But le t me poin t out , if you have other parts of 10 this source term reduced, then the ruthenium release c'y 11 not be trivial in terms of health consequences. It's 12 trivial in the context of what else has been released, but 13 it may not be trivial if some of these others are lower.
O 14 The other aspects of this is on page 377. You 15 say that, "The QUEST study for the TMLB B prime series 16 sequence, "which is the same series showing this large 17 ruthenium results dif ference , you say it's " brought on the 18 order of a factor of 100."
19 Does that statement mean that for all species 20 the uncertainty is a> factor of 100? If so, then it doesn't 21 --
it's not consistent with the 10 to the 4th.
22 MR. MURPHY: What has been done in 0956 here is 23 they are repeating the conclusions drawn from the 24 conclusions in volume, 84-10, the second volume of the 25 OUEST study. ;f n
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
27025.0 55 BRT f
O 1 In that QUEST study they have a -- they 2 specifically defined the source term for the purposes of 3 the summary, uncertainty as being the total number of 4 curies escaping the plant as a f unction of time. So that 5 factor of 100, which is -- it is correctly stated -- is on 6 the order of a factor of 100 because it varies from about a 7 tactor-of 40 early, to something over 10,000 late, but by 8 late everything is so low, it doesn' t matter. So it's a 9 round number, but it only refers to the total curies coming 10 out of the plant.
11 MR. DAVIS: Independent of the radionuclides?
12 MR. MURPHY: It's the sum of all radionuclide 13 leaks coming out of the plant.
O 14 MS. MITCHELL: I just wanted to mention four 15 things, five things that NUREG-0956 won' t do.
16 MR. KERR: Only five things it won' t do?
17 MS. MITCHELL:~ Only five. It does not propose 18 regulatory changes. There were a lot of people who felt 19 that it did propose regulatory changes and that publication 20 of the document was indeed equivalent to acceptance of 21 changes in the regulations.
22 It does not discuss risk. We took out the 23 comparative risk appraisal, which was made up of a' ,
24 hodge-podge of the information that we had 'at tre time, to 25 try and put into perspective . what the source term changes O ;
l l
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
27025.0 56 BRT 4
(~h s/ 1 might be, and it doesn't work very well and it will be I
i superseded by NUREG-ll50 very shortly, so we took it out.
2 3 It does not evaluate event f requencie s . Those l
4 are also 1150. We were asked to evaluate the Source Term 5 Code package specific sequences, and whether they are large 3
6 contributors to core melt frequency or minor contributors i
7 is not an issue for 0956.
2 8 It does not quantify the containment event trees, 9 I think I mentioned this already.
1 10 And another we have discussed indeed already is 11 it does not endorse specific values of the source terms
, 12 with the restriction that we did do a quality assurance j
- 13 review at Brookhaven, but the f act is that the de finition l
14 of the sequence is indeed open for discussion in 1150.
! 15 MR. KERR: Doesn't have a specific source te rm ,
, 16 but it does say it should be used to calculate a source 17 term to replace 14 A44.
18 That I recognize covers a wide range --
4 19 something different than 14 A44.
20 Questions and comments? Thank you, Ms. Mitchell.
j 21 MR. KERR: Whenever you are ready, Mr. Meyer?
1 22 MR. MEYER: We are here , on the agenda , content
! 23 of the final report.
t j 24 I know you are inte re sted in these three "
l j 25 subjects which we have already discussed,'in facts.
4
()
1 ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800 336-6646
! 27025.0 57 BRT J
1 (Slide.)
2 This is right f rom the agenda.
3 I did not structure the presentation along the 4 lines of the agenda , but rather at this point, would like 5 to walk through the document to identify what is in it, 6 particularly in contrast to what was in the draf t report 7 that was published last summer. We will, in fact, come 8 back to the subject of technical bases for the Source Term 9 Code, status of the code package , and uncertainties, if we i
10 wish to discuss those more than we already have , we can do i 11 that as we come across them again.
12 So, what I would like to do now is to start 13 walking through the report, chapter by chapter, and 14 identify what the changes have been.
15 (Slide.)
1 16 I think this is useful, partly because I believe ,
i 17 were you all familiar with the draf t report and partly 1
18 because the comments that we've received are understandable 19 in terms of -- I mean the comments were made on the draft 20 report and I would like to show you that we've responded to j 21 those comments and produced a final report that, hope f ully, i i 22 resolves many of the problems.
i l 23 The first two chapters of the report basically 1
l 24 discuss background and have only minor changes compared i
l 25 with the draft report issued last summer.
(:)
i
- ACE-FEDERAL REPORTERS, INC.
I 202 347 3700 Nationwide Coverage 800-336 6646
27025.0 58 BRT
(,.
t
's >) 1 Section 1.3, the purpose of the documen t, has 2 simply been revised in a minor way to try and clarity what 3 we are up to.
4 In Chapter 2 we have added three new sections, 5 basically to address new material that is available now or 6 was only available in an earlier stage a year ago. It was 7 a section on IDCOR work and other source term calcula tions.
8 There are other reviews of existing calculations by some 9 other organizations like Stone & Webster and a couple of 10 tore ign programs, including the OECD review; and finally a 11 section that describes the forthcoming risk assessment 12 study, NUREG-1150, which we have already talked about 7s 13 considerably this morning.
()
14 So I would describe the changes in chapters 1 15 and 2 as minor, and would move on to chapter 3, where the 16 changes in the document are quite major.
17 *
(Slide.)
18 This is just straight f rom the table of contents 19 of the review copy that you have. The organization of 20 chapter 3 is quite similar to the draft report of a year 21 ago, but the chapter really has been completely rewritten 22 to accomplish several things.
23 Firstgof all, there was a comment fairly widely l 24 made, and I believe made explicitly or implied by the ACRS 25 comments, that the dratt report of a ' year ago did not O .
ACE-FEDERAL REPORTERS, INC.
t 202-347-3700 Nationwide Coscrage 800-336-6M6
i 27025.0 59 l BRT 4
1 present much technical material and therefore did not 2 satisfy the reader as a document presuming to provide 3 technical bases. Jocelyn indicated earlier that in our 4 draft report we had attempted to do that by reference, by 5 citing these reports that formed the building blocks of the 6 source term reassessment study, and we had, in f act , an Oak 7 Ridge report that gave the status of validation of many of 8 the codes in the source term sheet of codes used at that 9 time.
10 MR. KERR: Do you inte rpre t the Oak Ridge report 11 as having said that the validation was adequate? Or i
12 incomplete or partial?
13 MR. MEYER: I would say, at this time, that it '
14 is out of date. Some of the insights --
15 MR. KERR: That's enough. I don't want you to 16 stop something important --
17 MR. MEYER: We still cite it as a reference. I l 4
18 think it contains a lot of good insights into the status of l 19 some of the codes, but there has been --
- 20 MR. KERR
- At this point you would not use it as 21 something which will tell what the current state of 22 validation is? I r
23 MR. MEYSR: We would not use it as the sole t
24 basis of the status of validation, and have attempted to I
25 incorporate directly into NUREG-0956 some code to data l
( ..- -
tw I i P
, ACE FEDERAL REPORTERS, INC. .
f 202-347 3700 Nationwide Coverage 800 336-6646
27025.0 60 BRT n
( )
- ' 1 comparisons, and other information from experimental 2 programs that will provide at least a basic data base, or 3 information base for assessing the adequacy ot the codes.
4 Now, I realize in a single document with a large 5 general readership, that we cannot present in-depth, a 6 large number of code-to-data comparisons and other 7 validation results. And, in fact, for many of these codes ,
8 since they are severe accident codes and we have only 9 recently started reaping results from our severe accident 10 research programs, at this point in time the validation of 11 the severe accident codes is still quite lim i ted .
12 And, so, what we have attempted to do is to
, 13 incorporate some validation comparisons directly into the O, g 14 report in order to try and give the reader some basis for 15 agreeing or disagreeing with our conclusion 'that the --
16 that the codes represent an adequate s ta te of the art 17 methodology for doing work in this area.
18 MR. CATTON: I would like to just comment on 19 your section on MARCH 3. It says the core melts at 2277 20 degrees, then it says tem'p erature not allowed to" exceed 21 2277 degrees, then it says if the molten mass exceeds 75 d
22 percent, the molten core falls into the lower head.
23 Can you reference something somehow somewhere so 24 a person could decide whether they agree or disagree with 25 that? Why not 80 percent or 50 percent? Why not 3000? ,
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 L
27025.0 61 BRT O 1 Reading the document, just all of a sudden they 2 are there and then you are on to the next paragraph and you 3 are telling me about the many thermohydraulic processes --
4 MR. MEYER: Mr. Catton, the discussion of the 5 maximum temperature in the report is related to a phase 6 diagram in the report, rela ted to observa tions made in the 7 PBP program and from TMI. Its adequacies and inadequacie s ,
8 I think, are fairly candidly discussed.
9 MR. CATTON: What about the 75 percent?
10 MR. MEYER: The 75 percent is another matter. I 11 don't know, perhaps Rich Denning would want to comment on 12 it?
13 MR. CATTON: I don't want to hear it. I just 14 think in the report it ought to either give me 15 justification or refer me specifically to the document 16 where the 75 percent is arrived at.
17 MR. MEYER: The referral is there. The 75 '
18 percent is part of the meltdown model A in the MARCH code.
19 It is described in the 2 code manual which is referenced in 20 the report.
21 MR. KERR: But the MARCH 2 code is not the Source 22 Term Code package.
23 MR. MEYER: The MARCH 2 code, with modifications 24 that will be explicitly described in the Source Term Code' 25 package manual, comprises that part of the Source Term Code O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
4 l 27025.0 62 BRT s .
1 package.
i 2 MR. CATTON: See, that 75 percent is a key 1
3 assumption.
4 MR. MEYER: Of course that is.
5 MR. CATTON: Because that leads to the 6 containment heating and all sorts of things that are a
- J
)
7 result of having to deal with 75 percent.
8 MR. MEYER: That's right.
i 9 MR. CATTON: I think you ought to be dealing I
10 with 30 percent.
l 11 MR. DAVIS: IDCOR says 20 percent.
i l 12 MR. CATTON: And my guess is as good as yours.
13 Justify your guess. Your guess is not justified in MARCH.
- 14 It's just assumed in MARCH.
15 MS. MITCHELL: It's assumed in these 16 calculations.
17 MR. CATTON: To be the same as MARCH 2, where it 18 was assumed to be 75 percent.
19 MS. MITCHELL: It's an input number, just as 1 20 surely as the width of the core on the floor underneath the 3
21 vessel. If that kind of thing is the reason why the the l t
! 22 numerical values, the specific numerical values are ,
I> !
23 reviewed in NUREG-1150, this is something that is uncertain !
, 24 and is being evaluated, and the core meltdown. model is part l
i 25 of the uncertainty analysis that's being done in NUREG-ll50.
i . .
}
I ACE FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 800-336-6646 -
.. . . _ - -_ . -- - = - _.
i 4
- 27025.0 64 j BRT p
v 1 discussion or at least a specific reference where the l
2 discussion exists ought to be in the report.
3 Mr. Meyer says it's in --
1
] 4 MS. MITCHELL: Chapter 3.
1 i 5 MR. KERR: -- the MARCH 2 documentation, but we 6 are told the source of the code package is not 2 anymore, j
j 7 it's something else.
8 You see the difficulty one has following the l
9 road map -- of course, there's a lot of roads, but it seems 10 to me because of that it's extremely important to give ones j 11 who intend to dig into this where the information is found.
12 MR. CATTON: I thought perhaps I owed you an
, 13 apology, so I looked for it again, but I don't. It's not j 14 in here. You refer to picking the central range between i
15 1925 and 2475. knd that's how you picked this value. '
16 Somehow there's got to be a rationale behind that other 17 than just averaging. There's got to be. If there isn't, 18 it certainly decreases any confidence one might have. And I
l 19 you start off this paragraph about the principal unresolved 20 uncertainties, I think you should say a little bit more
- 21 about what the impact of those principal unresolved 1
j 22 uncertainties are. These particular ones happen to drive l' 23 the whole business.
-24 Enough.
25 MR. KERR: Just?
f l
O 1
I.
! ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336 6646
1 l 27025.0 65
- BRT 4
1 O- 1 MR. BAGCHI: Was it made clear the 5 percent is i i
2 a user input number so you are not forced to use it if you i
2 3 don't like it?
4 MR. CATTON: I understand MARCH is one great big 1
j 5 string of inputs and I could whatever I want. But if i
j 6 that's the case, I don't need the code.
i 7 MR. KERR: Any other relevant comments? Please l
1 8 proceed, Dr. Meyer.
9 MR. MEYER: You asked for a road map to
! 10 understand how MARCH fits into the course term code package. !
11 MR. KERR: Wha t I asked for was a road map of 12 references that deal with rather key issues in the code l
I 13 predictions. It was that to which I referred. This 5
- 14 diagram is helpful, but that isn't what I was referring to. -
i j 15 MR. MEYER: Well, to continue with chapter 3.
f 16 (Slide.)
17 I sense you are not too satisfied with what we l 18 have done but we have, between chapter 3 ard chapter 6,
- 19 where we have introduced what I think is a substantial 20 amount of technical information not prr.viously there, 1
4 21 included more than 20 figures and half a dozen new tables 22 that do try to relate the results coming out of our severe 1
- 23 accident research experimental programs with the codes and i
24 the code package , and what was on our part a serious
}
25- attempt to put some substance behind the discussion of
{
1 i
1 l ACE FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Co$erage 800-336-6M6
.,._ -...__ _ _ _ ,.,,,..~ .,__ . - . _ _ _ _ _ . _ _ _ _ , ._, _ , . ,_ -._-_ _ _ __ _ ,___.. _ __., _
?
l 27025.0 66
! BRT 1
1 codes.
j 2 MR. KERR: We should acknowledge, I certainly i 3 want to acknowledge, that I think"you have done that.
}
l 4 MR. MEYER: The material is leafed in to each 5 section dealing with a major code in the Source Term Code 6 package.
j 7 (Slide.)
- 8 At this point I had planned just to indicate i
{ 9 that this figure identifies the computer codes that were in 10 the original BMI-2104 suite of codes which, in fact are, 4
j 11 with some modifications, the building blocks of the current i
i 12 Source Term Code package.
i 13 The Sour .e Term Code package is a direct and 14 close descendant of the Battelle suite of codes in BMI-2104.
i 15 What I would like to do is just indica te to you --
16 MR. EBERSOLE: Before you take that down, may I i
17 ask a question?
i j 18 MR. MEYER: Yes, sir. '
I 19 MR. EBERSOLE: Considering the potential for the i
20 evolving set of events as you go into core molt, I l 21 mentioned earlier that I don't know what the operator is i 4
22 going to do, and apparently you don' t either. So I just j 23 only ask you, are these codes prepared to take the thesis i
- 24 that a part of the core melts and the operator reinvokes 25 cooling or a larger part melt and he reinvokes it in a i C
- )
I t i
i !
ACE FEDERAL REPORTERS, INC. {
202-347 3700 Nationwide Coverage 800-336-6646
. . . . _ . - - - - - _ - _ . _ . . - _ - - - . _. . - . ~- .- -
i j
i
'27025.0 63 BRT I
! 1 I don't know whether they happen to be taking 0 percent and --
i l 2 MR. CATTON: We heard about 1150 and how the l 3 uncertainty analysis has to be done. But that isn't going i
j 4 to get at it. If that 75 percent is important somebody has i
j 5 to be digging into what goes on in the core melt process.
j 6 MR. SILBERBERG: I believe we described in the 7 analysis that the core melt analysis and the analysis of !
8 that phenomena is an area of some study under the ongoing j 9 research program listed, in fact, in chapter 6 as in fact i
j 10 an open issue. It's one of the major sources of 4
j 11 uncertainty in the whole calculation is that number, and 12 there's a research program trying to get information on l 13 those phenomena so we can do a better job than just picking ,
j 14 5 percent.
i 15 MR. MURPHY: It's. looked at specifically in Othe
! 16 0UEST study, understand what the source terms consequences I
- 17 were, at least in a sensitivity fashion. They adjudicated i
18 the range of parameters by the kinds of core detail i j 19 mechanistic analysis I just spoke of. So it's an example I
t 20 where the type of uncertainty study, in reference to the
}
i 21 document, without going into great detail, has been of 22 assistance in adjudicating the accuracy of the.whole i
- 23 package. t
]
j 24 MR. KERR Since that's such a crucial i
25 assumption, crucial number, I think that perhaps a bit more l
i i
i l ACE FEDERAL REPORTERS, INC.
202 347 3700 ~ Nationwide Coverage 800-336-6646 j
l 27025.0 67 BRT O 1 ste pwise con tex t? Is it prepared to take the range for
! 2 both these considerations that are possible? From full 3 core meltdown to just the upper one foot?
4 MR. MEYER: Jocelyn, and then let's let 4
i 5 Rich --
6 MS. MITCHELL: This is one sequence where in 2
l 7 formation given in chapter 4 talks about a particular q
] 8 scenario, particularly the Sequoyah scenario, for an i
9 induced hot lake scenario where temperatures are going to
- 10 heat up, having introduced the hot lake failures, pressure l
11 go up and accumulators go in ang; it's the sort of scenario l 12 you have in mind.
l 13 MR. EBERSOLE: That's an automatic one. I'm 14 talking about a progress, though, which is complete. Like, f 15 say, 10 stages of melt and recovery of coolant.
16 MR. CATTON: Je s se , they can't do it.
j 17 MR. EBERSOLE: If they can' t do it, then where
. 18 are we ge tting?
19 MR. CATTON: They don't understand the core melt 20 process, whether it's crust or not crust, water is going to 21 get in or not get in; these things are just not understood.
22 MR. EBERSOLE: What I see 'is a repetition of the 23 old large LOCA study work, in which a hypothetical large 24 LOCA was created and then umpteen billion dollars and 15 25 years was spent on a problem recognized now as purely
()
i 1
ACE FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coserage 800 336-6M6 i_,__ _-.__2_.__ _ _ . - . - -
.---_,_.1,--___-,--- . _ _ _ _ _ _ . . _ , _ _ _ . - . . . _ _ -,_... -
4 27025.0 68 2
BRT
. V 1 hypothetical.
t 2 MR. CATTON: 1.5 billion.
t 3 MR. SILBERBERG: Let ne point out, as we tried 4 to point out in chapter 6, the detailed look, more 5 mechanistic look at core melt progression, as partial l 6 termination, partial flooding, are the subject of what we i !
, 7 call our detailed mechanistic codes and experimental work
< 8 going on to support that.
9 For example -- and we note in the documents j 10 specifically that the use of a code packet by itself for 11 all cases is -- j ust doe sn' t wor!t. You need to support it 2
12 with additional research, which is there , which isn' t used
! 13 to calculate it. In other words, everything -- every i
! 14 single sequence , but . it's there to handle tha t situation.
j 15 For example, SCDAP handled that at different i 16 stages and you get insights from that that one could use.
l 17 MR. EBERSOLE: I just ,want us to revisit the old j 18 large LOCA. I see a color -- it seems to look like that.
19 MR. CATTON: The large break LOCA physics in 4 20 many ways is simpler and used appendix K to put an umbrella l ,
21 over it. .
j 22 MR. KERR: Mr. M[yer, please continue.
23 MR. MEYER: I would like to ask Rich Denning I
l s
, 24 from Battelle --
l 25 MR. DENNING: Rich Denning from Battelle. I
! () .
- ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage B).336-646
.- . . . - . . , . - - - - _ . - . . - - - -,- - -- - - - -, ,...,-.- - _ .- - - - ~ . , ,- . ,.
27025.0 69 BRT -
l s 1 would like to respond a little to that. I think you should 2 recognize that the purpose and intent of MARCH is to handle 3 the situations. You talk about partial recovery, this sort 4 of thing, of the source term and code package.
5 The operator actions, although they are not 6 explicitly modeled in the Source Term Code package , can all 7 be taken into account by -- you have to, before you run the 8 code, speculate as to what the operator is going to do.
! 9 The in te n t is to be able to do that.
j 10 But Ivan is also right, and some other people ,
j 11 that when you get into these partial recovery accidents j 12 there are some very complex phenomena that a code like l 13 March can't handle very well and that's why you want to 14 develop codes like SCDAP, to look at the se in more de tail.
i 15 But it's important to analyze the types of 16 accident sequences that you are talking about in the 17 consistent coupled fashion that the Source Term Code 18 package can do. It's just that we have to look very 19 carefully at the models and we do have to recognize there 20 are significant limitations in their ability to predict 3
21 hydrogen generation or this type of thing under"those l
22 circumstances. We can do that type of sequence. It's just 23 a matter of how well we can do it.
i 24 MR. EBERSOLE: It's the sort of analysis that 25 hypothesizes that once you start core melt, everybody runs 1
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27025.0 70 BRT n/
A- 1 for cover, that nothing takes place by deliberate voluntary
, 2 action. A most unreal state of af fairs, as evidenced by 3 the Chernobyl case. I hate to see a large endeavor based 4 on an unreal hypothesis. That's the gist of my problem.
5 MS. MITCHELL: I'm not sure that's a criticism ;
6 of the Source Term Code package.
7 MR. EBERSOLE: I don't know what it is. [
j 8 MR. SILBERBERG: The assumptions and sequences
] 9 that have been done.
10 MS. MITCHELL: I think there are operator 1
11 actions that are assumed to take place. You are right. -
12 You have to postulate what is the operator going to do and i
13 then go to try to find out what might happen.
l 14 For instance there is, in chapter 4, one 15 comparison of sequences with and without venting. The 16 specific plant is Peach Bottom and the specific operator 17 actions are as close as we can get to that particular 18 plant's operating procedures. But this is'an investigation 19 of what would it mean, what could it mean in the source 20
- term world , if the operator at Peach Bottom successf ully' 21 vented at a particular time? ;
t j 22 MR. EBERSOLE: But he vents af ter the disaster j 23 has occurred, which is a very nasty state of af fairs, and i 24 he 's presently committed not to vent, I believe , to i 25 preclude the accident trom having occurred, a most
! ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
27025.0 71 BRT
> 1 unsatisfactory state of af fairs in its own right.
- 2 MR. KERR
- Although this is an extremely 3 important se que nce , I think our discussion probably ought
) 4 to stick with 0956. I think your general comment, Je s se ,
1 i 5 is that you are not certain that this is going to be very 6 useful.
7 MR. EBERSOLE: Before I do a good job on 8 something I've got to find out whether it's worth doing.
9 That's all I'm trying to do.
10 MR. KERR: Please continue , Mr. Me ye r .
11 MR. MEYER: Most of you will remember the 12 Battelle suite of codes here. Le t me indicate some things 13 that were not done real well in the original suite of codes O 14 which we had an opportunity to remedy and have done so, in 15 the Source Term Code package.
16 The March code , for example, MARCH 2, has in it a i +
17 fission p'roduct release subroutine for in-vessel fission i
! 18 product release , it's called FP loss, and the core concrete 19 interaction subroutine called INTR.
- 20 Neither of these were up-to-date models, they 21 ' were recognized to be in error, and externally to March, we g ,
22 were running better codes in the code package but had an 23 inconsistency because some of the analysis was done with 24 the" older codes and some with the new codes.
25 You will see that we have eliminated the INTR o
ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 202 347 3700 800-336-6646
27025.0 72 BRT '
I subroutine and FP loss subroutine and used the be t te r 2 models in the MARCII code.
3 Also the TRAP-MELT code and MERGE code were not 4 coupled in the original analysis.
5 TRAP-MELT would evalua te the transport and 6 deposition or tission products within the reactor coolant 7 system; the deposit and decay heat was not fed back to the 8 thermal hydraulics in the BMI-2104, and"' t ha t was something 9 addressable in the near term and that has been improved in 10 the Source Term Code package.
11 Also, the content used in the calculation was 12 MODl, that had a slurry viscosity problem that became well 13 recognized. The core conversion, MOD 2, during the course
'A" /
14 of the study, was made ava ilable , and we have in fact i
15 ! switched to the CORCON MOD 2 version. And the Source Term h
16 l Code package, then, is comprised of those codes f rom the 17 Battelle suite of codes with modifications as indicated on 18 this slide.
i 19 It is not one single integrated code but a set 20 of a few coupled codes, the largest be ing the MARCii3 codes, 21 comprised or MARCil2 that has been modified, principally to 22 install CORCON MOD 2 as L replacement for INTR and CORSOR.
, u 23 These are a'11 described in the code manual that will be 24 published within the next few wee ks.
25 MR. CORRADINI On 4-14 where we previously V()
ACE. FEDERAL llEPORTERS, INC.
202 347 1700 Nationwide Coverage Mn336-6646
27025.0 73 BRT m ) 1 asked about the Brookhaven and Battelle calculation, were 2 these all done with --
3 MR. MEYER: I can't hear you.
4 MR. CORRADINI: Table 4-14, attachment 459, the 5 LOCA at Surry and the T&W prime at Surry, calculations for 6 Brookhaven and Battelle, were they both done with the 7 Source Term Code package like that?
8 MS. MITCHELL: No.
9 MR. CORRADINI: So part ot the reason you got 10 differences is that the models are dif ferent?
11 MS. MITCHELL: Yes. I think we pointed tha t out.
12 That was the whole reason tor doing this.
13 MR. CATTON: What about the rest ot the 14 calculations? Do you clearly note at the bottom of each 15 , table which version was used?
16 I
MS. MITCHELL: I think the text normally sa/(.
17 But that might be an editorial comment to put it on the 18 table itselt.
19 MR. CATTON: The tablos and tigures ought to 20 stand by themselvos.
- 21 MS. MITCHELL: The wholo reason tor doing this 22 particular audit and calculation that's roterred to on 23 table 4-14 was to assoss, given the old suite of codes and 24 taking one sequence that they intended to uso for 1150, 25 what would you got if you ran it with the Sourco Term Code O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Ntionwide Coverage Nn34/M6
4+-*A-_ a. W m. -.=i+=- a W. A _
.,4 -
4 ,4 b # -----.E<---e .e > 4, 4. e -
4=w 4- m. J _ _- - - -
1 j 27025.0 74 t
BRT l
j l package? '
f 2 You find out that, yes, there are different 3 3 numbers. It turns out there's a mistake in this table f 4 -- there are other mistakes in the tables which I have 5 subsequently found, but there is still a difference between .
6 the two. And it is due to feedback that you get out of the t
7 merged TRAP-MELT interface.
l 1
j 8 MR. CATTON: Isn't it also because the earlier i
i 9 calculations were done with one model and the later ones j
j 10 are done with the later one? i l I 11 MS. MITCHELL: The y a re . But the important <
12 comparison in this case, there are two things about this l
i 13 particular calculation, besides the ruthenium mentioned in i
O 14 the te x t s one is the feedback you get, and you get a i I 15 '
ditterent answer for the volatile materials which are
] l
! 16 deposited in the pipe , and in one case the feedback !
i l 17 indicates that they would move on. In the other case , ;
I !
18 there was no feedback and so they remained in place.
! 19 That's MERGE and TRAP-MELT, the interface between MERGE and l
)
20 TRAP-MELT. '
}j j 21 The COftCON in this particular sequence is not
- I j 22 quite so important.
- 23 There was one other change that was made in the i
j 24 understanding of the sequence itself , and that was the idea '
i j 25 that Surry is indeed a subatmospheric containment and we
!O 1
I l
I j ACE FEDERAL REPORTERS, INC.
202 347 3Rio Nationwide coverage 800 336 6646
]
27025.0 75
) BRT j '
1 now have a pipe about this big (indicating) that leads back i
j 2 into the containment. It's going to take a long time for 3 the pressure equalization back into the containment to get s
l 4 to the containment for the containment to be at a high t
t j 5 enough pressure to push it back out again.
I t
6 MR. CATTON: Shouldn't this report be based on a I
7 consistent set of calculations?
i 8 MS. MITCHELL: This particular table was 9 specifically put in there --
j 10 MR. CATTON: I'm not referring to that, but just 11 in general. The report should be based on consistent data.
- 12 If what Brookhaven has is the final word, it seems to me i
13 that's what this report ought to be based on. It shouldn't I i
O 14 have bits and pieces of earlier models.
15 MS. MITCHELL: If you'look in the table, the I 16 comparison table with the reactor safety study, you will l 17 not find the Battelle calculation. You will find the
', 18 Brookhaven calculation as the one we feel is the better 19 calculation.
i 20 This specific table was put in here to e 1 . .
3 21 illustra te the point of the audit calculation. Without 22 looking at two ditterent calculations you can' t make that .
j 23 point.
l l
l 24 MR. CATTON: I understand. But if I look at i
25 table 426 -- no, that's 416.
i C:)
I J
l i ACE. FEDERAL REPORTERS, INC.
j 202 347 3700 Nationwide Coverage 800-336-M46
- 27025.0 76 BRT I
1 MR. KERR: Your point is there ought to be a !
. i 2 consistency of calculation. I think you have articulated.
. 3 MR. CATTON: Several ways, but I'm not sure it 4 was understood. I'm still not.
I' 5 MR. KERR: I'm sure Ms. Mitchell is aware of 1
6 your view.
i 7 MR. KERR: Mr. Meyer, I'm looking for a 8 convenient break point somewhere near 3:00.
9 MR. MEYER: Any time. '
I 10 MR. KERR: Is this okay?
i j 11 MR. MEYER: Sure.
i l 12 MR. KERR: Le t 's ta ke 15 minutes.
13 ( Re ce s s . )
l ,
14 MR. KERR: As soon as you can by heard above the l
! 15 uproar, Mr. Meyer, we are all ears.
! 16 MR. MEYER: Let me pick up with an >
17 identitication of the changes in the final document, late f 18 in chapter 3.
4 4 I
19 (Slide.)
i-
]
20 Af ter completing our discussion of the codes-I
] 21 that comprise the Source Term Code package , there were in 22 the, earlier report some subsections that talked about l
23 validation and uncertainties which have been expanded in j 24 the final report. We now have a section that deals with i 25 both verification and validation, where we define
}
i i
! ACE FEDERAL REPORTERS, INC.
- 202 347 3700 Natlonwide Coverage M33&M
- - . - _ _ _ - - - - - - - - - ~ - _ _ _
! 27025.0 77 i BRT l- l~ verification to be a, loosely speaking, quality control 1
1 2 effort, where one attempts to verify that the code is 3 actually performing the job that it was intended to perform; I
j 4 that is that the coding statements have been written 1
5 prope rly , that the code actually will run on the machine 6 and produce answers to sample problems, et cetera.
- 7 In addition to that, we have -- we defined 8 validation to be a test of the modeling, to see that the 9 model is valid; that the model adequately represents the 1
1 10 phenomenon or adequately describes the phenomena that are i
! 11 of in te re st .
t 12 Now, the verification ef forts described in 0956 l
'l 13 are two: One that has already been mentioned, done for us
(:)-
14 at Brookhaven National Laboratory, to look at the input 7
l 15 values for the sequences analyzed with the Source Term Code
]
i 16 package for NUREG-ll50, and also to look at the code itself; i
i 17 the other verification exercise which hasn' t b'een mentioned '
i l 18 - yet is something that we refer to colloquially as the hand '
] 19 calculations done at Battelle, and these are described in a 1
I 20 little detail in NUREG-0956. '
l <
] 21 What was done there is to take one aqp,ident ,
1 i
j 22 sequence that ha'd been analyzed with the Source Term Code i
1 23 package and to walk through a number of important time i
, 24 steps in the analysis, where it was possible to do an
} 25 approximate closed form analytic solution or to use some
(
I
! ACE-FEDERAL REPORTERS, INC.
!- -202 347 3700 - Nationwide Coverage 804 336-6646
27025.0 78 BRT
(')
k- 1 conservation principle to check the code as it ste pped from 2 one point to another through this sequence.
3 A fairly large number ot such so-called hand i
4 calculations were done by Battelle for one Peach Bottom I
5 sequence, I think it was, and are reported in a NUREG CR 6 report that will be referenced and will be published by the 7 time 0956 goes to press.
8 The validation section in Section 3.2 of 9 NUREG-0956 is very brief and simply points to two places.
10 It points to the validation work that has been heaved into 11 the previous discussion of the individual computer codes, 12 and then it points to a chapter 6 of the report, where a 13 systematic validation program is being executed in NRC's 14 severe accident research program. That is described in
~
15 chapter 6 in a little more detail.
16 Now, in the draf t version of NUREG-0956, with 17, regard to uncertainties, there was a description in 18 reference to the QUEST study performed' by Sandia National 3
19 Laboratory. We now have expanded this section to, not only 20 talk about the OUEST study, but to talk about the so-called
. 21 OUASAR study at.'Brookhaven, which is just getting under way; j 22 and also to briefly comment on the uncertainty estimates. ,
23 that are being made for NUREG-1150. We discussed this 4
l 24 subject earlier today. I don' t have anything else to offer i
25 at this time, unless you wish to discuss it some more. I 1
i '
C:)
4 J
ACE-FEDERAL REPORTERS, INC, 202-347 3700 Nationside Coverage 800-336-6646
- ,. i ,,s -,
27025;0 : y 79 BRT< **
w ,,
, 1 (Slide.)
2 So let's move on, finally, to another chapter ,
~3- tg,i:hapter 4 of the document. Chapter 4 presents some 4 sariple} calculations for selected accident sequences. By 5 ancf large , most of the sequences commented upon in chapter 6 4 tiave been calculated with the Source Term Code package in 7 its final referGnce version as it is being used for 8 NUREG-ll50.
t 9 In that regard, chapter 4 has been largely 10 rewr'itten-from -- in comparison with the version that was 11 in the published draft of a year ago. In chapter.4 of the '
12 dracE version.of NUREG-0956, the sample calculations were
~
13 the calculat. Iions reported in BMI-2104, where they now are ,
14 for the most part, calculations done with the code paTkage.
15 Three sections have been added to chapter 4, two 16 of which are' rather sub'stantial. One is section 4.8, 17 integratod.rplease from fuel in-vessel and ex-vessel, is a 18 relatively minor ef fort. It j ust gives' a little additional ,
19 information you might be'able to relate to as one follows a
20 the sequence through , the accident event. ,
21 Section 4.11 and 4.13, however, are significant 22 additions to the reporr. that represent new work performed 23 in the l'ast year.
24 Section .11 talks about the+ quality assurance 25 program, we called it a qual'ity contiol program. We did
~
.i -
l
?
w ~;
.,;{#-
's +
.- s ACE-FEDERAL REPORTERS, INC.
, y 202-347 3700 Nationwide Coverage 80 4 33 & 6646
l 1
27025.0 80 BRT 1 not make any distinction between those two terms; and the 2 audit calculations at Brookhaven.
3 I guess I would like to point out that there 4 actually were two types of quality control measures that we 1
5 are referring to here. One was a checking effort, where we 6
did go through and check all of the input tor all of the 7 sequences calculated with the code package.
8 In addition to that we took four sequences and i 9 completely reanalyzed them, setting up the problem from 10 scratch at Brookhaven in a manner that they understood to l
11 be the same way that Battelle had run the problems. And i
12 running four sequences in their entirety, as an audit of f 13 the Battelle wor k. And that's describad in section 4.11.
- k3) 14 'Now, as you know, since we published the draf t 15 report there has been a lot of discussion about the 16 chemical form of iodine at various locations in the plant.
l 17 We did not have a very extensive discussion on this at the 4
18 time we issued the draf t report. There have been, in fact, 19 some new results that have emerged since we published the
. 20 report that have probably made the ' issue a little less 21 , comprehensible than we thought it was.
22 We are now in the process of trying to resolve 23 what the chemical forms of iodine are. We have continued a
24 to use the assumption that the predominant. form of iodine 25 is cesium iodide for the calculations that we have done ,-
O f
ACE-FEDERAL REPORTERS, INC.,
, 202-347 3700 Nationwide Coverage 600-336-6646
M 27025.0 81 BRT 1 but we have performed a parametric analysis using dif ferent 2 assumed fractions of cesium iodide and other volatile forms 3 of iodine. And the results from the parametric studies are i
4 described in section 4.13.
5 (Slide.)
6 Chapter 5 described --
7 MR. KERR: Mr. Meyer, I want to find 4.13, 8 because I looked for this specifically.
9 MR. CORRADINI: Page 469, parametric analysis.
i 10 MR. KERR: I found the title , but it seems to me 11 that at one time or another the comment was made , back l 12 during the Subcommittee mee ting , that it would be useful, 13 since iodide, at least previous to TMI-2, had been O 14 identified as a rather key item, to go'through and find out 15 the dif ference between what was predicted in WASH-1400 and 16 what was predicted by the code package.
17 It seems to me there's some sort of general ,
i 18 statements here about parameter variations but there isn't 19 anything like that at all, so far as I can tell.
20 MR. MEYER: I'm sorry. I didn't understand the 21 question.
22 MS. MITCHELL: We did not go back and try to
, 23 figure out what WASH-1400 me thodology would have said. Oak l 24 Ridge is developing a chemical package that could be added 25 to othee codes like March, if it was found to be desirable ,
LO v
! ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4646
27025.0 82 BRT
- \
~ l and that is the parametric study that was done for which 2 the numerical results are 4.72. The tables are on 472.
3 In this case, this is one sequence, the station 4 blackout sequence, for Surry, for two assumptions relative
. 5 to containment behavior: One is the early containment 6 f ailure and the other is a late containment failure.
7 MR. KERR: I think one of the reasons one of us j 8 thought this would be interesting is because it seems to me -
1 9 much is made about the fact that improvements nave occurred.
10 One improvement is the treatment of iodide. And I would 11 have been sort of interested to see what the dif terence was 12 in the results. Maybe that's not feasible.
i
- 13 MS. MITCHELL
V
-) For cesium iodide, the comparison 14 with what WASH-1400 said for certain sequences are 'in here.
15 The problem of going back to isolate each and every one of I
]
16 the dif ferent changes without having the feedback of one to 17 the other is not, I don' t think, necessarily so useful.
18 This is the result of making a bunch of changes, not just 19 one at a time. ,
s 20 MR. KERR: But you are convinced that what you 21 have here is a lot better?
22 MS. MITCHELL: Ye s .
23 MR. KERR: This must mean that it's a lot 24 different.
I 25 MS. MITCHELL: Why is that? I don't know that O
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
_ _ _ _ . _ - _ _ . . - _ , _ _ . _ , . ,,_ _ . _ .. _ . _ _ , . ~ - - , , -- - , . - . - , _ _ . . - _ , - _
l
)
27025.0 83 BRT l
1
(- 1 that follows.
t 2 MR. KERR: If the results are not very much 3 different, it may be a lot more complicated and a lot more 4 expensive , it doesn' t strike me as a lot better if it gives 5 about the same results. You developed a new method. If it 6 gives the same results as the old method, where is the 7 improvement?
8 MS. MITCHELL: Because it gives the exact same 9 answer for one sequence doesn't mean that it's better or 10 worse or not better.
11 MR. KERR: All I'm saying is I have been told 12 that the new methodology is a lot better.
13 MS. MITCHELL: Right.
14 MR. KERR: To me it probably means it gives 15 answers that are a lot easier to understand or a lot more 16 accurate or something.
17 MS. MITCHELL: I think both of those.
18 MR. KERR: If the answers are not very much I 19 different --
l 20 MS. MITCHELL: They might be for some sequences 21 and might not be for other sequences.
22 MR. KERR: Precisely my point. It would be 23 interesting to me to pick something out that would 24 illustrate the dif ferences produced by this new method.
25 MS. MITCHELL: We made a handful of what we O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 84 BRT
(-) 1 consider to be major changes and we ran the entire package 2 with all of those changes in place and made a comparison in 3 here with what we could make a comparison for the two 4 plants, Surry and Peach Bottom, for which we had results in 5 the reactor safety study.
6 Now, to say which of those changes made the 7 specific dif ference , I think, is not quite so usef ul.
8 Should we make the changes one at a time and 9 avoid the feedback Kinds of ef fects that you might get by 10 making several changes at once? Should we go through and 11 leave boil, as the only, or hot trap, as the only one that 12 was used? We just felt the most logical thing is to take 13 the whole new package and make the comparison.
Cs1
'~
14 MR. KERR: Maybe what I'm asking for is so 15 simple that it doesn' t make sense . But to take a key 16 contributor to risk, in WASH-1400, iodine was identified as 17 a very key contributor. It is, I gather, now, not looked 18 upon as necessarily the key contributor it was earlier.
19 It would have been nice, it seems to me, tor one 20 to pick out sequences, and say, aha, here is a different 21 way this new package treats -- using iodine as a surrogate, 22 maybe something else is be tter.
23 MS. MITCHELL: The only way to make tha t 24 comparison is to hold everything else fixed and change one 25 single assumption at a time.
(v-~)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6M6
27025.0 85 BRT O 1 MR. SILBERBERG: Like , cesium iodide , for example.
2 MS. MITCHELL: That means developing a 3 methodology where you would go back and do it one tiny 4 piece at a time.
5 MR. KERR: What I want is impossible. Okay.
6 MR. MEYER: I think you can see f rom the table 7 it's not producing the same answer for each of the 8 radionuclides and for all of the sequences.
9 MR. EBERSOLE: On the subject of iodide, does :
10 this contribute to the question of ingestion? .
l 11 MS. MITCHELL: I beg your pardon? I didn' t hear !
j 12 your question. '
l i 13 MR. EBERSOLE: Ingestion. -
O 14 MS. MITCHELL: It's a cloud problem and an 15 inhalation problem.
16 MR. EBERSOLE: In what sort of degree is it one ,
17 versus the other? Mostly a. problem due to cloud shine or 18 ingestion? -
,I 19 MS. MITCHELL: In the early exposure it's cloud f
20 shine. In the latent radiation it is ingestion.
I 21 MR. EBERSOLE: What portion of the cloud shine {
t 22 does it present? -
23 MR. EBERSOLE: There's a Sandia problem that ,
i 24 goes through all of those , one at a time, and tries to talk .
-25 about its contribution. I don' t happen to know those . !
O t I
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 -
1 l
27025.0 86 l BRT w l 1 MR. EBERSOLE: Is it 5 percent? 10 percent? 50 2 percent?
3 MR. KERR: It depends on how much iodide you are l
4 calculating will be released. That's one of the 5 differences, I presume, to -- in the two packages.
6 MR. EBERSOLE: Okay.
7 MR. CORRADINI: But there is in that Sandia 8 re por t , I know exactly what you are speaking of, a set of 9 assumed contributors, by Strip and -- what's his~name --
10 Dan Albert, but I seem to remember that iodine was 11 approximately 20 to 30 percent for early deaths, and cesium 12 was dominant for latent cancers.
13 MR. EBERSOLE: Is this from ingestion?
14 MR. CORRADINI: Under the assumption in the 15 report that they had equal amounts released?
- 16 MR. EBERSOLE: Was the e f fect due to ingestion 17 or to shine?
18 MR. CORRADINI: I can't remember.
19 MR. EBERSOLE: It seems that would be important.
20 MS. MITCHELL: For latent ef fects it's normally 21 the ingestion, inhalation and ingestion deposition. But 22 for the early fatalities it's the cloud shine and/or ground 23 shine.
24 MR. EBERSOLE: Okay. Thank you.
- 25 MS. MITCHELL
- If you would like me to I can
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
27025.0 87 BRT s- ) 1 look up that Sandia report number.
2 MR. EBERSOLE: I'll look it up.
3 What I have been hearing at Chernobyl and then 4 that crazy state of af fairs in Chattanooga, the iodide -- I'm 5 trying to get a handle what that means as an inhibitor to 6 the dose problem.
7 MR. KERR: Good point. Please continue.
8 MR. MEYER: Chapter 5. Chapter 5 describes the 9 review of the source term work , and the major change in 10 chapter 5 really is the addition of the section on the 11 public comments.
12 We received a large number of public comments, 13 over 700 pages of comments from about 68 individuals and 14 groups. We have looked at each and every one of the 15 comments carefully and found that we can group the comments 16 into roughly 68 categories, and have done so for the 17 purpose of providing a reply to be placed into the final 18' document. Actually it's placed in appendix C.
19 (Slide.)
20 In chapter 5, the brief section, 5.7, describes 21 the review process and makes reference to appendix C, where 22 we have collected paraphrased statements of the comments 23 and Staf f replies to the comments for the vast majority of 24 comments that were received. This is how we dealt with the 25 public comments received on the draft report that was O
ACE-FEDERAL REPORTERS, INC.
202-347-3700.. Nationwide Coverage 80 4 336-6646
27025.0 88 BRT
/~
l published last summer.
2 In the draft report this was a chapter 6, 3 dealing with a risk perspective. As mentioned earlier in 4 the discussions, all of the material in the draft on risk 5 assessment and on containment performance have been removed 6 from the report because the report really does not, and did 7 not at that time, deal in a complete or comprehensive 8 manner with those subjects. And so we now have removed 9 that section. And so chapter 6 now becomes the research 1 10 chapter, which also has had some significant modifications i
11 compared with the earlier version of the report.
1 12 (Slide.)
g 13 In the earlier version of the report as we have J 14 done here, we talk about the areas of uncertainty in source 15 term analysis and we describe areas, not only identifie,d by 16 ourselves as a result of this project, but also by several 17 other groups that have been working in the same area; 18 namely, the American Nuclear Society identified a list of 9 18 research topics related to source terms, where the y. .
20 thought research should be continued; the industry degraded 21 core rulemaking group, IDCOR, and the NRC have met on a 22 number of occasions and produced a list of so-called NRC 23 IDCOR issues, which are topics that either have major 24 impact on source term analysis or represent topics on which 25 the NRC and its contractors have difference of opinion with O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 89 BRT
()
s 1 IDCOR and its contractors about modeling assumptions.
2 Also, the American Nuclear Society in its study 3 came up with a short list of items that needed additional 4 attention.
5 If you look at all four of these lists of topics !
6 and items and suggested research areas, they are very 7 consistent and do, indeed, I think, indicate with consensus 8 where the shortcomings in our technology are , and in the 9 first part, actually section 6.2 of chapter 6, we have 10 discussed eight of the se issues which we view as the areas i
4 11 of major uncertainties and have discussed them not only in 12 terms of their importance to the source term analysis but 4 13 also in terms of research results that have been done in O 14 those areas that have a bearing on the status of these 15 issues.
l 16 So chapter 6 you'll notice a change in the title; 17 it said "Further Research," now it just says "Research."
18 Chapter 6 has been revised in the attempt to describe the 19 status of some current research areas as well as pointing 20 to future research areas.
21 In section 6.3 and 6.4, we describe planss for 22 further research in the areas of source term and I
23 containment loads. These, in fact, are taken directly from l 24 NUREG-0900, Rev. 1, the report issued this spring, .that j 25 describes the revised severe accident research plan.
! C:) -
a
- l ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
27025.0 90 BRT
- 1 I can tell you that section 6.3, plans for 2 source term and containment loads research, corresponds to 3 Section 3.2 of NUREG-0900, Re v . 1, and, in fact, consists 4 of the paragraphs titled " Planned Activities" from 5 NUREG-0900, with some editing. The editing is just 6 basically to try and make it more readable for'a broader 7 audience than 0900 might have been.
8 MR. KERR: Mr. Meyer, 0900 has a good bit of-9 discussion about research through FY '87. It talks some 10 about things that should be done in FY '88, but only in 11 very general te rms , so general that one certainly could not 12 evalua te what was to be done. At least I couldn' t evaluate 13 what is to be done.
14 It seems to me the same thing is true of 0956.
15 That puzzles me a little bit because we are right now in .
16 the process, at ACRS, we are now in the process of being 17 asked to comment on the research for FY '88. If this is 18 the most detailed description that we have of what is to be 19 planned it is going to be difficult to comment on it.
) l l
20 That's not the principal burden of our discussion today )
21 but --
22 MR. SILBERBERG: In fact, I think you had raised 23 that issue, and we can speak somewhat to that tomorrow. It )
24 does tie in with what we learned from things like 1150.
, 25 This is a close relationship. Your observation is correct.
- () l I
i ACE-FEDERAL REPORTERS, INC. '
202-347-3700 Nationwide Coverage 800-336-6646 l
27025.0 91 BRT
'(D x/ 1 MR. MEYER: Our intention here is simply to 2 recognize that there is a research plan that has, in fact, 3 been focused on our weaknesses in source term containment 4 loads area.
5 MR. KERR: 0900 says virtually nothing about l 6 anything past FY '87.
7 MR. MEYER: I guess that will have to be 8 discussed.
9 MR. SILBERBERG: That is in fact the subject for 10 tomorrow.
11 MR. MEYER: But our intent here is to be 12 consistent with 0900 and not to invent a new wheel.
13 (Slide.)
14 Finally, before we have conclusions and 15 recommendations, we are going to walk through the 16 conclusions and recommendations presently -- let me just 17 comment. You have heard us say that we removed f rom the 18 draft report, material on containment behavior and risk.
19 There were several appendices that supported chapter 6.
20 Those have also been eliminated. In their place we have 21 two new appendices: One that describec the IDCOR-NRC 22 technical issues that I mentioned a few moments ago, and 23 one that provides these generalized comment statements and 24 replies in summarizing the public comment period tha t we ' ve i
25 gone through.
O ACE-FEDERAL REPORTERS, INC.
g 202-347-3700 Nationwide Coverage 800-336-6646 L
27025.0 92 BRT O' 1 Now, what I would like to do is start working 2 our way through the comments and recommendations which 3 appear both in the executive summary and in chapter 7 of 4 the report. They are identical in both places.
5 MR. KERR: Le t me suggest , Mr. Meyer -- are 6 these going to have a lot of -- have they had a lot of 1
i 7 influence on the technical content of the report?
8 MR. MEYER: The conclusions result from the
! 9 report. We tried to look at the four major chapters of the i 10 report, chapters 3, 4, 5, and 6, and, chapter by chapter, 11 to see what conclusions we drew from that material. These i
12 are those conclusions.
13 MR. KERR: What do you plan -- if you don't mind C) 14 my asking for a road map -- what do you plan to do with the i
j 15 public comments? Are you going to say here's the comments 16 and what we did and go back to the report? Or are you just 4
17 going to show us how you have entered them?
18 6 MR. SILBERBERC: In today's meeting?
19 MS. MITCHELL: Are you asking what we are doing 20 today?
21 MR. MEYER: I hadn't planned on doing any of 22 them today.
23 MR. KERR: What is it you said were you going to 24 do?
25 MR. MEYER: Conclusions and recommendations.
1 O
s !
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I 27025.0 93 BRT 1 .
1 MR. SILBERBERG: Comments will be treated 2 lightly.
. 3 MS. MITCHELL: Treated in a brie f f ashion. Not a
4 treated lightly.
5 MR. WARD: Disregard it.
6 MR. SILBERBERG: Pardon me.
7 MR. MEYER: I'm going to stand up here for a few -
8 more minutes as we go through a half dozen conclusions that 9 came from chapter 3, which is the chapter tha t I wrote .
10 Then Jocelyn will take over and continue the discussion of i 11 the conclusions.
12 Conclusion 1 may either strike you as profound 13 or trivial. It is rather an abstract concept but, in fact, 14 it's one that we think is important. We are now talking 15 about an approach to source term assessment where you don't 16 simply make sweeping assumptions or calculate in isolation 17 two or three ef fects and then try and combine them to get a
, 18 source term. We are talking about an approach where we 4
19 attempt to at least semi-mechanistically analyze events 20 that occur in a severe accident, and to the extent that's l 21 practical, to couple the se together becau se the phenomena
- 22 do feed back on each other. This approach, we believe , is 23 one that is now widely accepted. It is the same approach
! 24 as being used by IDCOR and EPRI in their attempts to do 25 current source term analysis.
O 1
! ACE-FEDERAL REPORTERS, INC.
, 202-347-3700 Nationwide Coverage 804336-6M6
-_ , _ . . _ , _ . _ ,_ - ~ . . . - . . _ . . , , . ~ , . . _ . , _ ,_,_ . _ , _ . .
27025.0 94 BRT b's> 1 MR. S I LV. .RB ERG : I should just point out: We 2 don't think it's trivial.
3 MR. KERR: It was he who used the language, not i
- 4 us.
5 MR. SILBERBERG: Right.
{ 6 MR. MEYER: The last slide was yellow and this 7 one is blue.
8 (Slide.) -
9 This one is blue. There's a significance to 3
10 that because we have made a wording change compared to the 11 wording in the review copy in front of you. We made these 4
12 changes very recently on the basis of some review that 13 we've had between May 23rd and today. I don't think they j
O 14 change the sense of any of the comments, but I think maybe 15 they are a little clearer and a little smoother, pe rha ps .
16 Conclusion 2 says that -- oh, excuse me, I have 17 to interrupt myself one more time. We renumbered a couple 18 of the early conclusions. This was -- this used to be 19 conclusion 3. l 20 "The complexities ot source term phenomena have l i
21 re sulted in sophisticated computer codes and large amounts q 22 of numerical information that require the use of quality I l
23 assurance measures." j i
24 This is a conclusion based on our experience.
l 25 We had a lot of difficulty early on, particularly with the c-(-)
ACE FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800-336-6646
27025.0 95 BRT t
\s>i 1 Battelle suite of codes as we were using it, in manually 2 handling large quantities of data and in using codes that 3 had not been fairly widely circulated and put through the ir 4 paces by other users in a critical mode, such as in the 5 inde pe ndent program we have up at Brookhaven.
6 We discovered quite a number ot errors, both 7 numerical errors and coding errors, and have come to the 8 opinion that in dealing with large codes that we must 9 necessarily deal with, to cover this whole range of source 10 term phenomena, that it is going to take a conscious ef fort 11 and a quality assurance program in order to do it with 12 reasonable amount of accuracy.
g- 13 MR. KERR: I'm very pleased tha t you reached Q) 14 this conclusion. It puzzles me a little bit that you had 15 to go through a lot *of this to reach it. It seems to me 16 you almost would have reached the conclusion at the 17 beginning of the process. I expect you did , really. I 18 would suggest that you change one word, simply because of 19 the conno ta tion it has in the regulatory arena. I really 20 think you are talking about quality control. And I don't 21 want to quibble, but quality assurance has come to mean 22 some thing very specialized and I don' t think it's what you 23 are talking about.
24 MR. MEYER- Jocelyn, will you make a note of I
25 that?
O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6M6
27025.0 96 BRT 1 MR. KERR: I think what you have developed are j 2 quality control measures, which I think are what you need.
3 MS. MITCHELL: We did, essentially, to reach 4 that conclusi6n in the draft document -- I think it has 5 been restated. It used to be a lot longer but did talk 4
6 about quality assurance required. It was a conclusion.
7 MR. KERR: It's not a big thing. If you object i 8 I won't --
I 9 MS. MITCHELL: No.
i
, 10 MR. MEYER: Conclusion 3, which is number 4 in 11 the report in front of you, with minor wording changes, i 12 says that " Validation- of NRC's Source Term Code package" J
13 -- now we are specifically talking about the code package l
O 14 we described earlier - "is substantial in some areas but i'
15 limited in others; more comparisons of computer codes with 16 experimental results are needed."
17 This is a fine line to try and walk. It's a 18 difficult topic to deal with. We believe that we are 19 somewhere like halfway up the learning curve. We are not 20 at square one. We do have some validation. We do have 21 some confidence in our codes.
22 We also recognize that the validation is.
l 23 incomplete and that the uncertainties remain large and that l 24 more work is needed.
25 MR. KERR: If it is permissible for me to go a 0
- ACE-FEDERAL REPORTERS, INC. l 202-347-3700- Nation *ide Coverage 800-336-6646
.~ . ____ _ . . _ , _ _ _ . _ _ _ , _ , _ _ . . . , _ _ _ . , ,_ . ,,, _ __ ,. ,,l
27025.0 97 BRT
'- 1 little further than this and read from the report , under 2 that particular conclusion someone has written: " Validation 3 against larger scale , more prototypic experiments is 4 necessary to achieve a desired level of confidence."
5 I want to ask you what the desired level of 6 confidence is, but do you have some mechanism in place 7 which will eventually tell you what the desired level of 8 confidence is when you reach it? I'm not trying to be 9 facetious here because we have spent a lot of ef fort on 10 this research activity for the last five years or longer 11 and I don't see anything that would give an indication that 12 would say, here's where we are and here's what we need to 13 go to get where we want to be, except for the general O 14 s ta temen t. It implies there's a desired level of 15 confidence.
- 16 MR. MEYER
- I don't know how to quantify that.
17 MR. SILBERBERG: There would have to be some 18 cri te ria .
19 MR. KERR: But if you are going to plan a
- , 20 research program at this point with five years of research 21 in back of you, it seems to me somebody ought to start 22 thinking about where it is you think you are going and how 23 you will know when you get there.
24 MS. MITCHELL: We do know where we think we are 25 going today, based on where we believe the major sources of
()
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
I 27025.0 98
~BRT n
kJ 1 uncertainty are in the te chnology . As we try to apply it, 2 we will find out that things are good enough over here and 3 maybe we don't need any more research in that area, but 4 things are not good enough and we are not going to be --
5 MR. KERR: Do you know, in a particular area, 6 where things are not good enough, what would be good enough 7 if you got there?
8 MR. SILBERBERG: Today I'm not quite sure we 9 know that.
10 MR. MEYER: Bill, I don't think this is a 11 scientific question. I think it's more a societal question.
12 MR. KERR: It's not a scientific question at all, 13 but it's a question you better ask if you are going to do 14 scientific research with limited resources.
15 MS. MITCHELL: As we say, relative to one of the 16 other conclusions, that -- maybe it's not in the 17 conclusions, but answers to public comments -- that surely 18 things as we start to look will require, again, refocusing 19 the research program. I don't think the research program 20 is the same as it was nor are the emphases. The emphasis 21 on dif ferent areas has been changing over the years.
22 MS. WEGNER: Find things out we can make changes 23 on where we want to go. As we use it, we can find things.
24 MR. KERR: Almost anybody around the table would 25 say one should probably not stop, I would say -- let me put O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 99 BRT f%
\~ l it that way, we should not stop the NRC research program.
2 But I don't think it should continue to do the same thing 3 year after year. It it does, then I would be in favor of 4 stopping it. If it is not to do the same thing year after 5 year, then one has to define when you get finished with a 6 given program and can release those r$ sources to work on 7 something else.
8 I don't see any indication in this document or 9 in 0900, that anybody has made any ef fort to do that. I
- 10 may have missed something.
11 MS. MITCHELL: I think we are starting. Until 12 you start trying to apply it, you don't know.
7x 13 MR. KERR: But af ter five years or six years and U 14 a lot of millions of dollars, one ought to have reached 15 some point at which one could say, these problems, we' ve 16 done enough.
17 MR. SILBERBERG: Yes. We have. I thought 0900 18 -- and maybe it didn't do it well enough -- said there's an 19 awful lot on the front end of the accident that we have a
, 20 lot more confidence in, on the beginning of melt 21 progression.
J 22 MR.' tKERR: I don' t know what the front end of 23 the accident is. Up to now I thought it meant until you 24 start core melt. To me severe accident research is core 25 melt.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 100 BRT O- 1 MR. SILBERBERG: I meant up to the time when you 2 are initially making -- in other words, making hydrogen and 3 getting to the initial temperature peak.
4 MR. EBERSOLE: I've got to ask Bill, what is 5 core melt?
6 MR. KERR: Core melt to me means the core is i
7 completely molten.
! 8 MR. EBERSOLE: It doesn' t mean the top six 9 inches?
10 MR. KERR: No. Everything. That's the Kerr 11 definition. It is specific, unambiguous. It may not have 12 any significance, but it's unambiguous.
13 MR. EBERSOLE: It's the double ended pipe 14 rupture. .
15 MR. CATTON: It has the same amount of 16 significance 1
17 MR. KERR: You asked me what I mean. That's my 18 answer.
1 19 MR. EBERSOLE: It's the large LOCA in the sense 1
20 of the context. And, again, it's probably minute.
21 MR. KERR: Jesse, I'm not saying it has an 22 importance. I'm saying I can define it.
23 MR. EBERSOLE: I know, but don't you have to say
, 24 something that's important?
25 MR. KERR: I'm trying to talk with these people
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27025.0 101 BRT
( )
x/ 1 about another topic and I'll discuss this with you at 2 another time.
3 MR. EBERSOLE: I hate that word --
4 MR. KERR: There are reasons for that, Je s se ,
5 but I won't bring it up now.
6 Well, I made enough of that point, but I do 7 think, almost in order to justify future research, you 8 probably thought it through, I just don't see it in the 9 documents. I should say where we might go and we think we 10 have an end. When you get there you may have discovered 11 enough in the process that you would have been all wrong --
12 MS. MITCHELL: Where we do get the list of g-) 13 significant issues and major uncertainties is out of V
14 getting insights from codevelopers and users and 15 experiments --
16 MR. KERR: We have now been at this process 17 since 1980. That's six years. We ought to at least be 13 able to reach some fairly specitic conclusions about what 19 should be done next and a' bout how long it will take.
20 MR. SILBERBERG: We have. In fact, in 0900, as 21 a result of --
22 MR. KERR: There's nothing in 0900 that tells 23 you about how long it will take beyond fiscal '87.
24 MR. SILBERBERG: What it says, now identified 25 where to re focus our e f forts. Others have reviewed the o
O ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Ceverage 800-336 6M6
J 1
27025.0 102 BRT 4
1 program and said the same thing. Here's where we find 2 ourselves. Now the question is, when do you know that you i
l 3 have enough of that, reach a certain point? That's 4 somewhat an iterative process, along with 1150 and the user.
5 MR. KERR: At this stage you can't say what the d
6 desired level of confidence is at all, as it is used under 7 conclusion 4.
i 8 MR. MEYER: I think that's a correct observation.
9 I really wanted to make what I think is a serious point, 10 and that is that the level of confidence that we need is 11 not a scientific process that -- I don't think we can say 12 in advance. It has to do with public confidence and with s 13 concurrence of our peers. When we get there and we can get i
14 the various groups who have inte re st in this kind of j 15 analysis in the regulatory framework to agree that what we j 16 have is adequate , then I think we can stop work in that j 17 area. I don't know how to forecast that in advance. We i 18 obviously aren't there.
j 19 MR. CATTON: If you can't have confidence j 20 yourself, how can you expect your peers to? If I don' t
! 21 believe what I've got --
1 j 22 MR. KERR: It has no quantitative significance 23 at this point or even qualitative significance because a 24 there's no way you know of at this point that you will know l
l 25 when you get there.
ACE-FEDERAL REPORTERS, INC.
) 202-347-3700 Nationwide Coverage 800-336 6646
4 27025.0 103 BRT i
1 MR. MEYER: That's my feeling, yes.
2 MR. DENNING: I'm Rich Denning from Ba ttelle , we 3 really, with the new 1150, have a way to start to make some 4 gauge, what is an adequate level of confidence.
5 The uncertainty analysis in 1150 is directly 6 applicable to the kinds of questions that we are asking.
7 We can' t answer the question without looking at 8 applications, and 1150 is an application. It looks at 9 risks and cost / benefit tradeof f s and the size of 10 uncertainties and we get to see what are the contributors 11 to the uncertainties. For example, nine of the key severe 1,
! 12 accident issues actually appear in there. We can look and 1 13 see how important are those uncertainties to a decision of, I
14 would you back something to an existing plant?
15 I have heard you ask this question for the last 16 two, three years, and there hasn' t been an answer as to how 17 to finally get a handle on how good is good enough? I 18 think we really have a tool now, through 1150, the ,
19 beginning. We are finally at a level at.,w3ich we can see 20 how important are these uncertainties to regulatory ;
21 questions.
22 Now, 1150 is just the first application and we 23 can only look at those questions that 1150 addresses. But 24 we can explicitly see in 1150 how important some of these 25 issues are, and we can drop some of them. It's an
()
I ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
i j 27025.0 104 i BRT 1
0 ~
v 1 excellent tool to prioritize research.
2 Just step 1. But it is a tool to do that. So I
{
} 3 wouldn' t completely go along with the completely i 4 qualitative view that Ralph has here. I think we finally
- i j 5 do have a quantitative way to prioritize our research
$ 6 program, to drop things, to really focus on those things 7 that are important and affect decisions because of the-
$ 8 uncertainties that they represent.
I
{ 9 MR. KERR: Mr. Corradini?
10 MR. CORRADINI: Just one thing. Rich, you have 11 a po in t . We have 1150, but we have to recognize it is j 12 subjective. It relies on a bunch of people sitting in a 4
i 13 room arguing about it or discussing it, not the code 1
- 14 package, and then coming to some sort of agreement as to a 1
i j 15 range. So I'm still thinking about the need for how much J
16 research is enough? If you feel you couldn't do it on a 1
I 17 broad basis, is this -- is there some physical basis that 18 you could give as an example? If you can't do it for i -
19 everything, could you do it for a few things where you ~ can l ,
20 say, here's an example, this is enough. This is put to bed.
ll l 21 And this is not enough and give reasons how far you should 22 go. i 3
j 23 MR. SILBERBERG: ~ Let me give you an example, a l
! 24 practical example of a case where there was some debate i
{ 25 over the control rod failure in melting model and how l
I
()
a j ACE-FEDERAL REPORTERS, INC.
3 202-347-3700 Nationwide Coverage 800-336-6646
- - _ _ _ . _ ~ _ _ , _ _ . - - - - _ - . , _ - . _ _ _ _ _ _ _ _ . _ _ . , . . , - _ , _ , , - _ - - , . . _ - _ ,
27025.0 105 BRT
' 1 control materials might par ticipa te in heating up and 2 becoming aerosols and becoming part of the transport 3 process.
4 As we pointed out in 0900, I believe, Dr. Kerr 5 -- and it's an area that we believe is -- because of 6 reaching some research point is closed -- we determined 7 from several experiments, one was in PBF, the other was 8 ACRR, that in fact the insights were that the control 9 material, basically, melted and went down and was not in 10 the same sequence -- same time ' sequence heating up the rest 11 of the core. Therefore, you don't use all the silver 12 aerosols, et cetera , et cetera, et cetera. ,
13 That was a physical, you know, experimental and 14 a specific answer to a question -- to a problem. We sa id ,
15 yes, okay. Now we know how to treat that. We have some 16 backup, some confidence, so I won't model it. I won't 17 consider modeling the aerosols that way. And the community, 18 reviewing the experiments and discussing them said: Yes, 19 that's basically it.
20 So that's one example. That is one very. _
21 specific example but nevertheless, it's an example.
22 MR. KERR: Is that enough?
23 MR. CORRADINI: That's fine.
24 MR. SILBERBERG: Let me apologize for not giving
. 25 the answer Rich gave , becauce it turns out that's one of O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
(
27025.0 106 BRT f'n
's / 1 the things we were going to discuss tomorrow, Dr. Kerr.
2 1150 is a start in a direction of trying to get a measure, 3 an approach for measuring how far you have to go with some 4 of these uncertainties. But it's only a start and it's not 5 in a vacuum. You have to deal with other things.
6 So, let me say that I think your point is well 7 taken and I think we should revise the statement in there 8 that reflects -- we should revise that.
9 MS. MITCHELL: The 1150 is definitely the 10 subjective process, asking peoples' best judgment on issues.
11 However, it isn't the only effort there is.
12 The QUEST used the then-existing code sui te to 13 draw evaluations, in-depth, for a very limited set of 14 sequences, and the QUASAR in place, in progress now at 15 Brookhaven, is using the Source Term Code package for the 16 same kind of thing. So it's a multi-pronged effort on the 17 uncertainty analysis, not just the single program.
18 MR. CORRADINI: I understand that. But be 19 careful. Because the thing said -- if you are familiar 20 with it -- about 0956, is you just don't want to focus in 21 on the code alone because you yourselves in the last 22 presentations -- well, the last Vugraph we have in our 23 pac ke t here, somewhere, is the things that are not in the '
24 Source Term Code package. The biggest one there is 25 direct --
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
- y. + -.
,g - - .
. N- ,~ n r y 't
]. ~
27025.0 , s . .' 107 BRT -* . #
. a.,
1 MS. MITbHELL: This thing is trying to be taken 2 into account in the other. There is a multi-pronged e f fort 3 on it and"not just one single effort.
1 4 MR. CORRADINI: Then my last point is that, I 5 was' writing this down, I might as well say it. In chapter 4
6 6, I thought, what was the purpose of having chapter 6 on 7 0956, is there something there that isn't in revision 1 of 8 09007- If it isn't - .if the answer is no there isn't, if a .
9 the whole thing is encapsulated in 0900, take out chapter 6.
10 ,,
MR. KERR: It's somewhat more extensive than
^
! 11 0900. ~
i ,
12 MS. MITCHELL: The front part is more than 0900.
back, section 603, is out of 0900.
- o. 13 14 s Th had tried in the past to incorporate things solely by The problem is we 15 Yt eforence and it doesn't work very well. So you want to j 16 have at least something for people to look at.
17 MR. CORRADINI:' The only other thing I was going 18 to say, what I would try and,- include ,' if you can ' t do it in 19 a broad approach for all\ physical processes, a few examples 4
20 of things where we have gone far enough and things we i 21 i haven't gone f ar enough and how far should we go would be 22 useful. --If you can't do it for everything, if you could do s l 23 it by example for a few, I think it weald by useful. '
24 MR.~ ^ KERR: -Please continue',1Mr. Meyer.
]
! 25 MR. MEYER: Conclusion 4 is that "The current O
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage M336-6646
. . - ~ - - - . , . .. -
I 27025.0 108 BRT 1 severe accident data base and the phenomenological models 2 in the NRC's Source Term Code package represent a major 3 advance in technology, compared with the reactor safety 4 study."
5 We describe in chapter 3 a number of areas where 6 advances have been made. We recognize the limitations of 7 wha t we have , but notwithstanding , I think we have come a .
8 long way.
9 (Slide.)
10 This one is really important. It says "The 11 Source Term Code package does not incorpora te intentional 12 biases, a feature that was believed to have been 13 incorporated in previous me thods." Furthermore , we don' t 14 think it is possible "to provide an inherently conservative 15 margin in a Source Term Code that is to be used for many 16 sequences and many purposes."
17 Previous methods have been thought to have built 18 in conservative margins. And perhaps for a single program, 19 risk assessment,-for example, you might be able to figure 20 out which way to bias the models in the code in order to 1 21 overpredict the source term and by this be conservative.
22 What we found is that even for a single 23 purpose-like risk assessment, we can find two sequences
- 24 where -- for e. mple, if you bias the in-vessel fission i
25 product release high, that is, if you designed your CORSOR O
4 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-33 6 6646
27025.0 109 j BRT 1 model to overestimate in-vessel releases, we can point to 2 some sequences where this would lead to larger source terms ;
3 and others where it would lead to smaller source te rms .
- 4 In other words, if the in-vessel releases were 5 all scrubbed through a suppression pool, and you had an 6 ex-vessel release component to go along with it and you 7 were able to get it all out early, this might lead to what 8 you would call a nonconservative source term. So we simply 9 believe that it is not possible for a code that is going to 10 be used for many different sequences, not only for risk 11 assessment but for equipment qualification, which may 12 completely different senses of uncertainty -- of 13 conservatism than releases to the environment, tha t it is O 14 not possible to fix this thing up so it's always going to 15 be conservative. So we tried not to do that.
16 This, then, leads directly into conclusion 6 17 which says -- recognizing that the uncertainties are large, 18 and that we haven' t been able to bias the codes in a way 19 that would guarantee that we are going to have conservative 20 or bounding results, that the importance of the 21 uncertainties must somehow be quantitatively estimated for 22 each type of application that's going to be made of the 23 code.
- 24 (Slide.)
25 At this point I would like to turn this over to (y
, (_/
4
- ACE-FEDERAL REPORTERS, INC.
j 202-347-3700 Nationwide Coverage 800-336-6646 I _ .-- . . _ _ . . , .- -_. _ - - _
27025.0 110 BRT O 1 Jocelyn, who will continue right through the conclusions.
2 MR. KERR: She's going to continue discussing 3 conclusions?
4 MR. MEYER: Yes.
5 MS. MITCHELL: With that taken together, the 6 kinds of sequences that were analyzed with the BMI-2104 7 suite and the source term code package have represented a 8 wide range of the important parameters that govern the 9 source term calculations.
10 (Slide.)
11 And, theref ore , we have , indeed, exercised more 12 than just a wide range, that it is not likely that you can 13 find a lot of sequences that would be outside of the range.
O. 14 We have looked at high pressures and low pressures and 15 sapidly developing accidents and slowly developing 16 accidents. This is not to say the source term results 17 would not be different if you ran it for another sequence.
18 But only that the kinds of conditions have been looked at.
19 The main reason for saying something like this 20 is that we were beat about the head and shoulders for not 21 having considered such types of sequences as earthquakes 22 and we believe that, although we did not consider a source 23 term for a sequence that was initiated by an earthquake ,
24 that the kind of phenomena that could be encountered would 25 be similar. And we put another example of this in the new i ()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
27025.0 111 i BRT
- 1 document.
2 MR. KERR: Excuse me, before you start with that, 3 is the significance of the last phrase , " these calculations 4 have exercised the code over a wide range of conditions,"
5 that they have demonstrated the code will run. You don't 6 know whether the results are any good or not, but at least 7 it will run over those wide range of conditions? Is that 8 what is meant by " exercised"?
9 MS. MITCHELL: That plus a little bit more. We 10 certainly haven' t been able to validate the code for all 11 the kinds of sequences.
12 MR. KERR: I'm not trying to be critical. I 4 13 just want to know what the term " exercised" means.
14 MS. MITCHELL: It means we have run it. It i 15 strictly means exercised. It does not go off in to an 16 unphysical situation.
17 MR. KERR: It doesn't blow up or quit or 18 anything hike that?
19 MS. MITCHELL: Right.
20 MR. KERR: So this isn' t a statement of accuracy 21 or confidence, it's simply a statement that it will run 22 given a wide variety of inputs?
23 MR. SILBERBERG: It's more than that.
24 MS. MITCHELL: I think it's more than that.
25 That is true, it will not blow up.
()
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
._m . _. ___ _ _ _ . _ - , _ . _ . _ . .. . ..... ~ __ _ . . . __ . _ . _ _ _ __
27025.0 112 BRT O
\# 1 MR. KERR: What is the little more, then?
2 MR. SILBERBERG: Yes. The little more is people 3 looked at it and the results do not appear to be 4 unreasonable , or, you know, nonsense.
5 In other words, as you go back and look at the i 6 result, and say, except for this situation and these inputs 7 this is what I got , and you can stand back at it and get 8 some insights that say: Yes, this is not a physically 9 unreasonable situation.
10 MR. KERR: If that's what you mean, I don't 11 think that statement says it.
l 12 MR. SILBERBERG: I don't think it says that, but f- 13 to be f air to the people at Battelle --
b 14 MR. KERR: I'm not trying to be fair or unfair 15 to people at all. I'm trying to find out what the writer 16 intended.
i 17 MR. SILBERBERG: If the exercise says what- you 18 think it means, then it's wrong.
19 MR. KERR: I was asking what it meant to whoever 20 wrote it. That's all I was trying to de termine .
21 Indeed, it either means -- I'm not quite sure 22 what it does mean, other than that the code runs.
f 23 MR. MEYER: I understand what you mean. I'll 24 make a note . Well do something.
25 MS. MITCHELL: We'll work on the phrasing.
() '
r l ACE-FEDERAL REPORTERS, INC.
j 202 347 3700 Nationwide Coverage 800 336 6646 L __ . , _ _ , . _ . _ . . . - - _ _ _ . _ . _ . . _ , _ - - _ _ _ . _ , - - - - - . . _ . , - ~ _ . - - - . ~ - . - _ - -
)
i 27025.0 113 BRT i l
(D v 1 (Slide.)
i I l 2 This is also a little bit of a change, and 3 basically compared with the document that you have from the j i
- 4 23rd, it is basically the other shoe from the quality l j 5 control conclusion, out of the section on the codes. l l
6 This says that you need to look also at the j
7 representation for plant and the representation of the 8 sequence, and it requires a great deal of care and
.1 j 9 a tten tion to a great deal of detail.
i 10 (slide.)
11 I think this is basically not changed over a l
12 similar conclusion in the dra f t of the document, and we gg 13 found a lot of cases -- I think others have also found a
\_)
j 14 cases -- where the calculations depend a great deal on the 15 details of construction or the design, and that we believe I
16 that it will make development of generic source terms very i
17 difficult.
! 18 (slide.)
l 19 We have found that it the containment lasts for 20 some hours following the secession of input, aerosol input i 21 to the containment, that you can reduce the source te rms 22 and we find that whether there could be -- whether the
{ 23 containment survives or fails or is bypassed because or the
! 24 sequence definition, that it is a major factor affecting j 25 the source te rm s . This, again, is pointing out that we do
(
1 1
ACE-FEDERAL' REPORTERS, INC.
202-347-3700 Nationside coverage 80 4 33 & 6646
4 27025.0 114 BRT 1 not analyze the performance of the containment. We do 2 analyze in this when a criterion for failure such as 3 pressure or time or temperature has been exceeded. We 4 calculate when that has taken place in the sequence and 5 what its effect is on the source te rm .
6 MR. EBERSOLE: Before you put that down, may I
, 7 ask you for an elaboration on the definition of source te rm?
8 I've always envisioned this as a host of terms,-one being 9 what gets out of the pens, what gets out of the vessel, 10 what gets out of the containment. Because I see and read 11 about all of these things being called source terms in both 12 an intermediate and terminal state. So what do you call a f- 13 source te rm?
(_/ 14 MS. MITCHELL: As used in this document, it is j
15 released f rom the plant.
16 MF. EBERSOLE: Relea' sed from the plant, the 4
17 tence?
I 18 MS. MITCHELL: Outside of the containment, edge 19 of safeguards building --
20 MR. EBERSOLE: Plant or fence? You have to tell 21 me. To me it's a variant, depending on what you consider 22 as a boundary, and I never have seen a definition of what
! 23 it is.
24 MR. KERR: Did you ever get a satisf actory 4
25 answer from her?
(:)
f ACE-FEDERAL REPORTERS, INC.
j 202-347-3700 Nationwide Coverage 800-336-6646
27025.0 115
.BRT O' 1 MR. EBERSOLE: No.
2 MS. MITCHELL: Released f rom the plant itself, 3 the edge of the containment or, depending on the sequence ,
4 released from say the safeguards building or reactor 5 building --
j 6 MR. EBERSOLE: Again we see the ambiguity.
7 MS. MITCHELL: -- it's released to the 8 environment. I'm trying to define the environment. The-9 environment begins when it leaves the plant, either the 10 containment or the place where it leaves the plant.
11 MR. EBERSOLE: Inside t he fence? Because I'm l
l 12 going to be looking at the opera tor dose -- '
13 MS. MITCHELL: You bet. One of the -- i C)
~
i 3
14 MR. EBERSOLE: -- tha t's inside the plan t.
i 15 MS. MITCHELL: You bet. ,
16 MR. EBERSOLE: I think more of ten than not it's l :
17 identified as that which escaped containment and that !
i :
18 doesn't say whether it's at ground line or the top or f 19 whether it went through the auxiliary building and killed i 20 all,the employees.
i :
21 MS. MITCHELL: This is an input to the code. !
22 If you wish to look at the containment bypass '
23 sequence like the V sequence , it is the definition of the ;
j 24 sequence. Is this going to take place in a pipe that goes f 25 into the safeguards building? Isn't that incorpora ted in !
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I 27025.0 116 BRT
- 1 the definition?
2 Furthermore, when you say that the containment 3 will fail at 120 psi, you have to say it will fail at the 3
4 knuckle of the dry well. Or when I vent it, I vent it at 5 the 18-inch line which is located right over here. Those 6 are in the definition.
7 MR. EBERSOLE: Could it be that which emanates 8 from a broken pipe that looks into the building and looks 9 into a molten core?
10 MS. MITCHELL: We use certain results in certain 11 applications. You don't necessarily, for equipment 12 application, want to look at what is out in the parking lot.
13 You want to look at what is airborne in the containment and 14 those are results which you can get out of exercising these 15 codes.
s 16 MR. EBERSOLE: Let me give you one scenario. It 1
17 can be a molten core , looking through an open pipe into the
] 18 auxiliary building, within which is the control room.
19 MR. KERR:. Do you now want to talk about 20 sequences or do you want an answer?
21 MR. EBERSOLE: I want to know what a source term 22 is. It seems to me it's a great variable. It seems it can 23 be anything.
24 MR. KERR: The report says -- it is re pe a ted . I I
25 think everybody would agree --
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationside Coverage 800-336-6646
27025.0 117 BRT 1
1 MR. EBERSOLE: When somebody says " source te rm" 2 I never know what they mean.
3 MR. KERR: Since this has been brought up, let 4 me go back to the conclusion. I think it's 9. That's 10 5- you have.
6 " Source terms were found to depend strongly on f
7 plant design and construction details," and in the last i 8 phrase, "thus making development of usetul generic source 9 terms dif ficult."
10 As I read the total report, I might _ interpret 11 that to mean or to be replaced by " impossible."
12 What I want to find out is whether you have 13 concluded that there may not be any such thing as a source 14 term? If you have, this is very important.
15 MS. MITCHELL: A generic source term?
16 MR. KERR: Well, I think before this rese,trch 17 began, there was at least some feeling that there might be 18 a generic source term. Af ter all, WASH-1400 was done with 19 the idea that you could take a couple of plants and they'd 20 become surrogates for a large number.
21 Well, that was felt, perhaps, to be a little 22 oversimplified. But, then, further research has been done 23 with the idea that you can make some general statements 24 about a large number of plants by looking at a small number.
25 Now, as I read this report, I could easily draw ;
l l
1 ACE-FEDERAL REPORTERS, INC. I 202 347-3700 Nationwide Coverage 800-336-6646
I 1
- 27025.0 118 l BRT O 1 the conclusion, I think, that you have prethy much i
2 concluded that you've got to go through and calculate a 3 source term, whatever that means, for every single plant, 4 in order to get any significant information about the risk 5 of operating that plant.
6 If you have concluded that, that is significant 7 enough that I think the report ought to say that. It hints 8 at it, but it doesn't say it.
9 MS. MITCHELL: I think you are picking out one 10 specific purpose. I think if you want to get through --
t 11 MR. KERR: I'm picking out a specific purpose ,
12 which is calculating the risk due to operating power plants.
13 That's a pretty important purpose.
I O 14 MS. MITCHELL: You said a particular plant. It 15 you want to state what is the risk of operating a 16 particular plant, I think you ought to look at that 17 particular plant.
18 It you now want to use this for a manufacturer 19 to be able to say that the equipment that he -- t ha t is i'
i 20 being designed by the company is adequate to be used for 21 safety purposes in nuclear power plants, it may be 1
22 difficult to develop a generic source- term which will i
23 adequately describe what is necessary for MARCH 3 24 containments or large dry containments. This is what we 25 are trying to say.
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage . 804336-6646
. _. .~ . _ . _ _ . . . ~ _ -_ _. _ ._ .- _ -.
27025.0 119 BRT i
! 1 MR. KERR: Suppose we are doing risk J-l 2 calculations for 100 plants. My question is: Can one talk 1
3 -about -- is it your view at this point that one can talk 1 4 about something which one gets from a calculation done with 5 at most five or 10 plants, and make sensible predictions ;
l- 6 about the risk of 100 plants.
- 7 MS. MITCHELL
- If you are going to do that, you 8 ought to do replicates of at least some of them to find out. t i
- 9 I don' t know the answer to that question. '
10 MR. SILBERBERG: It's possible? You don't know ;
]
11 the answer?
l 12 MS. MITCHELL: We have only one ice condenser j 13 plant. How different are they? You don't know until you 1
- 14 start looking at the second one. L
) 15 MR. KERR: I read repeatedly that calculations j
16 up to now would indicate that source terms are very 17 plant-specific, not once but many times. '
i 18 MS. MITCHELL: Yes. ,
1 19 MR. KERR: On the other hand, we are embarked on 4
20 a program as enunciated in the severe accident policy 21 statement, for example, which could be interpreted to say &
1 22 that at least when that policy statement was ' written there i
l 23 was a feeling on the part of some people that one could use i !
j 24 calculations done on 10 or so plants and make sensible l 25 predictions about the risk of 100 or so.
1 i
! ACE-FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Coverage 800-336 4646 t_____ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .
27025.0 120 BRT 1 It seems to me at this point if you have reached 2 a tentative conclusion that do or do not make sense, tha t ' s 3 a fairly important prediction.
4 MS. MITCHELL: I don't think reached that l 5 conclusion because we have not looked at a replicate of a I
6 plant.
i 7 MR. SILBERBERG: There's a dif ference between l
, 8 " difficult" and "does not make sense."
l
) 9 MR. KERR: Difficult sometimes is a euphemism l 10 for impossible, especially when engineers write , because 1
l 11 they don' t think anything is impossible .
i 12 MS. MITCHELL: Having written this by myselt, I
, 13 did not mean impossible.
14 MR. KERR: What do you mean by " difficult"?
15 All of the calculations you are making are difficult. This 16 whole subject is extremely difficult. If you really just j 17 mean difficult, I don' t know the significance of the 18 statement.
19 MS. MITCHELL: No. I meant if you want to look 4
20 at what you wish to make an estimate of the source term
) 21 trom two plants of nominally the same design, that for l 22 instance, if they were ones that happened to have sequences 23 that resulted in large core-concrete interactions and they 24 had dif ferent concretes, then the source terms may not be --
}
25 they probably won' t be the same. They won' t be the same.
l 1 l ACE. FEDERAL REPORTERS, INC.
1 i 27025.0 121 )
BRT i I
i 1 And, so, if you wish to have a source term that i
] 2 is generic for all plants, even all plants of that kind, i
j 3 you may have to add a great margin of uncertainty and it 4 will be dif ticult to justify generic source terms.
5 5 MR. KERR: Le t me ask this , then. I think you j 6 have told me that you, at least as an individual -- perhaps i
j 7 you three -- do not have enough information yet to make a l
1 8 judgment that one could or could not use the calculations
- 9 of, say, 10 plants to predict with some confidence the risk 10 ot 10 plants. Is that a reasonable sta tement?
l 11 MS. MITCHELL: Yes.
12 MR. KERR: Would you be able to predict what you j 13 would need to do before you could reach a conclusion? I'm
- ( t i 14 not asking you to tell me --
15 Ms. MITCHELL: I haven' t been asked to make a
- 16 judgment, but at least I would go out and try to replicate i
17 the same plan --
2
- 18 MR. KERR: I'm sorry, if you have not been asked j 19 to make a judgment then I want to ask you, because this
)
20 whole program has been aimed at dealing with the risk of 21 opera ting plants.
22 MS. MITCHELL: The severe accident policy 23 statement talks about the search for outliers in individual
- 24 plants. That certainly is a rec'ognition that there may ,
! l j 25 very well be things out there where --
1 l
J t
I ACE FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Coverage fKn336-6646
27025.0 122 BRT O 1 MR. KERR: But tha t is in the context of an 2 assumption that one can get information about the risk of 3 100 or so operating plants by calculating in detail, risk 4 contributions of about 10, I think.
5 MS. MITCHELL: I don't know that that follows.
6 MR. KERR: Am I misinterpreting things?
7 MR. SILBERBERG: Dr. Kerr, perhaps Joe Murphy 8 can help a little on this discussion, since he's --
9 MR. KERR Joe, am I misstating?
10 MR. MURPHY: It's a complex answer, so I'll talk 11 tor a little while.
12 If you --
13 MR. KERR: I didn' t make my question complex 14 enough.
15 MR. MURPHY: If you want to know, with as much 16 precision as we are capable of doing it, what the risk of 17 an individual plant is, the only way you can do that is 18 with plant-specitic analysis.
19 MR. KERR I don't think you can do it with 20 plant-specific analysis, but that's another subject. What 21 I'm interested in is an analysis that.will tell us 22 something about 100 or so operating plants. That is what 23 we are interested in, isn't it?
24 MR. MURPHY: Let me talk first about the design 25 trequency characterization.
O ACE. FEDERAL REPORTERS, INC.
202 347 3700 Natlonwide Coverage 800-336-M46
-- . . _. = ._. .- . - ..
27025.0 123 BRT O 1 I think from the work that we've done and the PRAs 2 that have been done and the industry base line study, we 3 have a very good handle on what are the kind of features 4 that exist in a plant that cause-outliers that lead you to 5 a higher core melt frequency.
6 I think we are on the way of being able to 7 develop a way of searching for that type of thing. I think 8 IDCOR has made a very good first start in their individual ;
9 plant evaluations. The kind of analyses we .have done on
! 10 the base line studies, which are essentially mini PRAs, are l
11 a very good way of looking at this.
12 We know now that we have to worry about things 13 that actually Mr. Ebersole has been telling us for years, 14 things like the service water systems and component cooling 15 water systems and the AC and DC power dependencies.
16 We know the type of problems in those areas that 17 can lead us to a high core melt frequency.
18 In terms of getting into the actual risk 1
19 calculation, we know the elements for a given source term.
- 20 We know the impact of population evacuation and that sort 1
21 of thing. That's reasonably well in hand. We have the 22 area in the middle of what about the source term itself and i
23 how that varies from plant to plant.
24 Much of that is very plant-specific and is going
- 25 to take plant-specific analysis. But even there we know a t
(
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
2 27025.0 124
, BRT l
i j
O
\' 1 lot of things that we can extract.
2 For instance, if you are worrying about 3 something like equipment qualification inside containment, i 4 we are in much better shape when we are talking about 5 release from a primary system into containment than what 6 happens inside the containment itself. So it's possible _ to 7 get some generic source term defined now.
1
- 8 MR. KERR
- But look, if you are really trying to 9 qualify it, you look and ask yourself what is this stuff i 10 going to be used for and what's the worst condition to i
11 which it is going to be exposed or condition l'm willing to
! 12 face, some probability, and_you put it in thero. You don't
]
13 analyze a particular plant for sure.
14 MR. MURPHY: That's exactly what I mean. We
, 15 have the ability now, from what we have at hand and with a
- 16 reasonable understanding of where we are, to be able to
) 17 envelop the situation and define a criteria for that.
1 18 MR. KERR: I'm not talking about that. I'm 19 trying -- tell me if I'm wrong, but I thought we were l
20 embarked on a program that was an effort to predict the j 21 risk of operating plants, among other things. At least i
22 thare was some feeling that by calculating carefully the
- 23 behavior of a few, 10 or so, one might then be able to 24 extrapolate and get some idea of the risk of 100, whatever l 1
25 you want to call it; to me this is that you have to be able .
l l
1 J
t ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-336-6646
27025.0 125 BRT 1 to talk about some sort of generic source term.
2 MR. MURPHY: I think what we are really going to 3 get is more a negative than the positive statement. You 4 just made a positive statement about what we are getting.
5 MR. KERR: A negative statement would say you 6 can't do this.
7 MR. MURPHY: I meant it in a different way. I 8 think what we can do is we can identify those plant i
9 teatures that would allow -- that would cause a plant to be i
10 different from what we have calculated for that family. I
! 11 think we are at that stage where we can do the analysis, j 12 f rom all the work that has been done, we can identify what 13 features will get us into problems on a plant that hasn't 14 been analyzed.
15 MR. KERR: This report says that source te rms 16 are very plant-specific. It doesn't say class-specific.
! 17 You are telling me , I think, that these are classes of f 18 plants, that you calcula te for those classes. Is that it?
]
19 Or are you misinterpreting what you are saying?
20 MR. MURPHY: You are misinterpreting me. There l 21 are classes but unfortunately they vary from issue to issue.
- 22 MR. KERR
- I'm getting myself back in a
! 23 situation where you've got to do the calculation for every l 24 plant; right?
25 MR. MURPHY: If you want to know what the risk f
l ACE. FEDERAL REPORTERS, INC.
1 202 347 3700 Nation *ide Coverage R00 3346646
27025.0 126 BRT 1 for that plant is you have to do the calculation. If you i
2 want to know the features of that plant that could get me 3 into trouble and identify features that could bite me , I r i 4 think I can give you a reasonable list with a little more i
- 5 work, as to what kind of features of a plant could cause it
! 6 to be outside the range aof the re ference points.
7 MR. KERR: Le t me go back., If we don't want
- 8 risk, it seems to me that you are sure doing a lot of work 3
] 9 to try to become quantitative about what happens when a i
j 10 core melts. If all we want to know is whether a core will l
- 11 melt or not, that I think is fairly straightforward. We l
l 12 have been investigating for about six years, hundreds of i
i 13 millions ot dollars, something which I thought was an i
14 effort to become quantitative about fission products that j
l 15 will be released and therefore contribute to risk. Am I I
16 missing something?
I f 17 MR. MURPHY: You have to_ understand what you j 18 have by looking at a series of plants, doing your analysis,
! 19 coming to a reasonable appreciation of what the risks are
- 20 with those plants and the associated uncertainties in l 21 connection with them. Then, from that information, you can l 22 look at that body of information and say
- What would cause i
23 it to be different? What would cause it to be higher?
24 MR. KERR Do you have to do that for every j
25 plant?
]
i i
s ACE. FEDERAL REPORTERS, INC.
j 20L347 3700 Nationwide Coverage 8003)(>6646
27025.0 127 BRT i 1 MR. MURPHY: I think you can do that more or 2 less generically, ask the question: What would cause it to [
3 be higher.
4 MS. MITCHELL: The policy statement is going to 5 require it for every plant.
6 MR. MURPHY: In a way I'm agreeing with you and
- 7 in a way I am disagreeing with you.
4 8 MR. KERR: I do not have a set opinion on it.
9 When I read the report, it seems the report could be 10 interpreted as stating, forget about calculating a source l 11 term or two or three source terms; you really have to do it 12 for every plant because the results are going to be very 13 plant-specific. I don' t know whether the plant means to i
i 14 say that or not.
15 MR. MEYER: Dr. Kerr, we have only calculated i
4 16 five plants. From the five plants we have calculated it 1
17 doesn't look like it's a sure thing that you can calculate 18 these generic source terms, but we certainly, on the basis 19 of five plants, haven' t ruled it out. It just looks like I 20 it's going to be a dif ficult job.
21 MR. KERR: My question, then, is what are you 22 going to do next to determine whether you -can or cannot?
1 23 That's a Murphy question.
24 MR. SILBERBERG That's a Murphy question.
25 MR. MURPHY: Forget plan X. I think I have 4
, ACE FEDERAL REPORTERS, INC.
. 202 347 3700 Nationside Coverage 8%3366M6
l 27025.0 128 BRT I O- 1 enough information to say, do I expect 'it to be dif ferent 2 from what was analyzed today, based on a preliminary survey 3 of that plant.
4 If the answer is: Yay,_ verily it's different, 5 and now I want to know what the source term associated with 6 that plant is, there's no way you are going to get it 7 without doing a plant-specific analysis. But the 8 identification that it is different is something I think we ,
9 can get out of it.
10 MR. KERR Mr. Corradini?
11 MR. CORRADINI Can I ask you a question since 12 you are up? The IDCOR purpose as I understand it is a 13 little more than that. Not only are they trying to 14 identify it, they are looking for some features and then ,
15 say okay, you are inside or outside the class because of 16 these key features. But the presentation we got f rom them 17 a while back is they went furthkr than that into the 18 containment loads and performance and said, now we'll try 19 to e s tima te the ef fects of those dif ferences. Rather than i 20 redoing a complete plant analysis for that plant, they 21 tried to get an ef fect of those dif ferences.
22 Am I misinterpreting?
23 MR. MURPilY: My interpretation, IDCOR gave the 24 NRC a briefing last month, three or four weeks ago. I l l
25 unfortunately had a time conflict and I could only sit in l O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage WA336 6646
__ _ __ . __ ___ _ _ ~, . _ _. _. _ _ _-
27025.0 129 BRT I
9 1 on the session on the analysis of the accident trequency.
- 2 And I've read most of what they submitted on the t i 3 Westinghouse-type plants.
i j 4 There, the suggestion is a pretty good search
) 5 for outliers by looking at what has been done in t he pa st .
6 I think they made a very good first attempt on that.
7 I can' t comment on the containment or source ,
4 l 8 te rm . I missed those parts of the discussions.
! 9 MR. WARD: You mean their approach is limited to 1
1 10 outlier, as f ar as -- you mean unusual sequences or as far j
11 as probability of core melt outliers or what?
i 12 MR. MURPHY: Well, tor a detailed discussion of 13 what their approach is I'd.rather IDCOR gave it but my ,
i 14 interpretation of it is --
15 MR. WARD: That's what I want to hear.
j 16 MR. MURPHY: -- that they are giving, for the 4
17 analysis of the accident frequency, it consists of i
- 18 constructing si.nplified fault trees and doing a very i
I 19 simplified study that would tell you if you are about to 20 get into trouble. Actually, there's not too much 4
21 ditterence between what we've done in our base line 22 analysis and what the IDCOR approach is. f j 23 There are some dif ferences in that we take it
] 24 through to the end and calculate frequencies which j j 25 basically they don't do. But in terms of all the accident i
1 ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 3 % 6646
27025.0 130 BRT h)
'/ I sequences -- but the basic process that they go through is 2 very similar, the thought process is very similar.
3 So, properly tuned -- I haven' t done a detailed 4 evaluation of it -- I think it could give you a readable 5 handle as to whether there was a system dependency, either 6 hard-wired or procedure-oriented, that could get you into 7 trouble.
! 8 MR. KERR Further questions?
, 9 Please proceed.
10 MS. MITCHELL: The little paragraph in the 11 document that explains this conclusion calls out the tact 12 that the American Nuclear Society found that they were 13 persuaded that there were large factors of reduction that 14 apply to presumably both plants, but all accident sequences 15 and all chemical element groups compared with the reactor 16 safety study.
17 We do not tind that that is based on the 3
18 analyses which we have pe r f ormed to da te . So we conclude 19 that generalizations at this time are ina ppropr ia te .
20 MR. KERR I think I know what you are driving 21 at but it seems to me that that statement doesn't get 22 across. Again, it's written the way you and I write as 23 engineers.
24 Put it up and let me show you what I have in 25 mind.
O ACE-FEDERAL REPORTERS, INC, 202 347 3700 Nationwide Coserage 80433MM6 L
. - . . . .- - .- _~
! 27025.0 131 BRT 1
1 (Slido.)
2 " Generalizations are inappropriate since large 3 factors of reduction in source terms were not found for all 4 sequences."
5 What ANS was trying to say had to do with risk.
6 They were sort of saying that there was enough reduction 7 that the risk was really low on it. I think you are in a 8 position to say that you don't think very broad, general 9 . conclusions about risk can be drawn on the basis of 10 calculations you've made. Maybe you are not saying that.
11 MS. MITCHELL: I think their statement was about 12 source terms. They have concluded that the source terms 13 were reduced, which, of course, leads further on to risk 14 reduction. But they did conclude that the source terms 15 were reduced, a factor -- at least a factor of 10 compared 16 with reactor safety study.
17 MR. KERRt If I were an ANS reporter here , I'd 18 look at that and say, yes, but for all important source 19 terms they certainly are reduced.
20 MS. MITCHELL: That wasn't the statement they 21 made.
22 MR. KERR But I'm saying, an adherent looking 23 at that statement could say, all you are saying is it 24 wasn't found for all sequences. That means if you found it 25 tor 99, but didn't find it for one , your statement is still O
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336 6646
t 27025.0 132 BRT 1 true. I don' t believe you found that big reduction in 99 1
2 out of 100 sequences.
3 MS. MITCHELL: This is true.
4 MR. KERR: Maybe not even in 50 out of 100. All 5 I'm saying is I believe that that statement doesn' t really 6 carry the meaning of what you think you found.
7 MR. DAVIS: For all important se quence s?
I 8 MR. KERR: Something or other, which, I J
l 9 think --
4 10 MR. WARD: For many important sequences.
4 11 MS. MITCHELL: --
that the sequences are 12 importan t is the fact that we specifically do not address
~!
13 in NUREG-0956, the probabilities of the sequences. So 14 therefore, I make no statement whatsoever on the importance 15 of the sequences and I do not have a justification for i
- 16 stating risk or event frequencies.
17 MR. KERR: I'd be willing to guess if sequence 18 led to early containment failure would be more impor tan t 19 than one that didn't lead to containment failure at all.
20 MS. MITCHELL: I think they meant importan t in 21 the probability sense.
22 MR. KERR: Importan t in the risk sense, which is' 23 what we are looking at. I 4
24 If you think it carries the meaning, okay. I 1
25 was just suggesting --
i
(:)
, ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationmide Coverage 800-336-6646
27025.0 133 BRT
. 1 MR. MEYER: I made a note . I know what you are 2 driving at.
3 MR. EBERSOLE: May I ask, in the matter. of i 4 con ta inmen t failure, I think most of the time you are 5 talking about containment failure from cause matters inside 6 the containment, high temperature , high pressure , whatever.
4
- 7 What contribution did you find in a relative 8 sense between initial containment, that is, the preaccident 9 conditional failure -- it already had a leak in it, didn't 2
10 close the valve , j ust simply didn't work -- what was the 11 variance in source terms from that cause versus the spread 12 of releases of radioactivity to containment proper? Wha,t
~
I
- 13 was the variance in the source te rms?
! 14 MR. KERR: Jesse, they didn't look at that 15 question at all.
t 16 MR. EBERSOLE: I have a suspicion tha t the 17 greater variance would probably be on initial containment.
18 MR. KERR: 1150, I think, does take tha t into 19 account. The likelihood you'll have an open event in tha 20 containment, inadver te n tly , 1150 will consider that.
4 21 MS. MITCHELL: We did look at what the source 1 22 term would be in the BMI suite of codes for the failure to 4 23 isolate , assuming a given f ailure size , what is the source 24 term.
25 The probability for any plant that that might O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336-6646
27025.0 134 BRT O
1 actually be the case , we do not evaluate.
l 2 MR. EBERSOLE: The reason I asked that it was 3 found historically, going back to 1969, that containment 4 vent valves couldn' t even close in a LOCA.
5 MR. KERR: They certainly didn' t look at that 6 question. 1150 ought to but this doesn't get to it.
7 MS. MITCHELL: Even though we did have an 8 extensive review, most of the commenters felt that we 9 needed further review and additional publication for peer 10 review of the models and the validation for those models in 11 the Source Term Code package.
12 (Slide.)
13 We do have -- we have asked our contractors to
~
14 publish, prepare things in a form which would be acceptable 15 for peer review journals and to submit it for publication, 16 though that is a fairly long- term ef fort.
17 As we use the technology in dif ferent 18 applications, it will be reviewed. I did want to mention 19 in connection with the point of uncertainties, the method 20 that is developed by the research and regulatory people 21 together for any use, any particular use of the source term 22 me thodolog ie s in the regulatory area, the way of getting 4
23 the uncertainties for that will be reviewed; should any 7
24 potential changes arise from it, it will be in the public 25 arena.
i ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
i 27025.0 135 BRT O 1 '
- ~
(Slide.) ,
2 ~ MR. KERR: If it were made available to you, if 3 you could get that, would you review it?
1 4 MS.'MITCHELL: Sure would like to.
5 . nn. KERR: If you could get MAAP.
6 MR. SILBERBERG: It's always. a question of 7 timing. We would like to have done that several years ago.
8 MR. CATTON: There have been presentations made 9 on it at the IDCOR mee tings -- r 10 MR. SILBERBERG: Necessary but insuf f icient .
11 MR. KERR: We OUght' to be able to persuade IDCOR 12 to' release it, it seems to me. It would seem to me they 13 would want to make it available to the NRC.
O 14 MR. CORRADINI: You are saying the code manuals 15 is not enough.
16 MR. SILVERBERG: Some of the points we have been i
~
17 trying to make here about, you know, quality control, et
. 18 l ' ce tera , e t cetera , the same thing applies. Let me say that 19 a code like MARCH was thoroughly used by many pedple and
~
20 there was what I call a complete -- a MARCH asses ment 21 program which was sponsored by us back about, I don't know, 22 four or five years ago, that really went into it. _
23 People were talking about line by line in the i .
24 code. That's wha t we mean. I don't think it's 25 ur reasonable.
O -
? ,
' ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 136
- BRT O 1 MR. CATTON
- But things like the kind of 2 physical modeling --
3 MR. SILBERBERG: A lot of that, certainly had a 4 lot of insights from the discussion.
5 MR. KERR: As long as they have a reason not to !
, 6 review it and they use it, we are getting to a point where 7 they wouldn' t have an excuse . We ' d be be tte r ' of f --
8 MR. CORRADINI: Can I ask a question? The 18 9 issues with IDCOR, those are of ficially settled or in the 10 settling process? Because they essentially arose from the 11 five meetings we had with IDCOR on this.
1 12 MS. MITCHELL: On MAAP during that time?
13 MR. CORRADINI: But, so --
1 14 MR. SILBERBERG: Many are unsettled. Of the 18, 15 some are settled but many are not. Appendix B addresses 16 that.
17 MS. MITCHELL: 14 are settled or not se t tle d ,
. 18 Ralph?
- 19 MR. MEYER: There are 14 issues that now remain 20 unresolved.
21 MR. CATTON: There are only four settled in the 22 last years?
23 MR. MEYER: I don' t know if four have been 24 settled -- one or two have been se ttled.
25 MR. CATTON: One of the -issues was the aerosol O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 ' Nationwide Coverage 800-336-6646
a 27025.0 137 BRT I
1 behavior of silver in the control rods. That one was 2 settled. Agreement was reached based on experimental 3 results and actually the NRC's model is the IDCOR model in 4 that particular case.
5 MR. WARD: Le t 's see . Are decisions going to be 6 made by the Agency, under the severe accident policy, 7 dependent on some conclusions that you are receiving from 8 the IDCOR program, using MAAP?
9 MS. MITCHELL: I can' t answer that question.
10 That's a regulatory question which you'll have to save for 11 Zoltan, when he nets here.
12 MR. KERR: The people in this agency are 13 reluctant to speak to each other across boundaries.
14 MS. MITCHELL: This is not true.
15 (Slide.)
16 MR. WARD: But NRR is fully cognizant that MAAP 17 hasn' t been reviewed by research, the research office?
18 MR. SILBERBERG: I think that's -- yes.
19 MS. MITCHELL: Ralph alluded ' to all the lists of 20 all the people who work in the source term area, have a 21 list of areas that they find uncertain and important in 22 uncertainty. Even though there may be 10 from one or eight 23 f rom the other, they are basically the same, either 24 coalesced or taken apart to make more or fewer. They still, 25 basically, agree on the areas that need further research.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 138 BRT O 1 (Slide.)
, 2 And the corollary to this is that we have 3 revised recently the focus of the severe accident research 4 plan and we belief the work has been focused on the areas 5 of the technology that need further study.
6 The basis for this statement is basically the 7 agreement of all of the researchers on what are the 8 important areas.
9 (Slide.)
10 This is -- this borders, perhaps on a statement i 11 ot f act rather than a conclusion based on the work in there. ,
12 But, with all of our other conclusions about the necessity i
13 of validation and verification, we just wanted to note that 14 we do have programs in place.
15 In this context, the fact that there's a small "s" 16 and "t" is important. In this case it's not just the 17 Source Term Code package, which we write with a capital "S" ,
18 and capital "T", but the source term codes that the agency 19 has. Be cause if you intend to tune your faster running 20 codes on the basis of insights that you get out of medical 21 progress or SCDAP to name some codes that were mentioned 22 earlier, it is necessary also for a validation and 23 verification program for those codes.
24 - So, we do have a small "s" and "t" on the 25 verification and validation.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 -
27025.0 139 BRT O 1 (Slide.)
2 We had three recommendations in the draft 3 document and people felt free to pick and choose among our 4 recommendations and to quote some of them back to us and to 5 ignore the existence of others where, in reality, we meant 6 them to be taken together. They were, in our mind, a set, 7 and that one does not come without the other.
8 So, we have gone to runon sentences and one very 9 long recommendation so that people would be le ss inclined.
10 We will not stop them from doing it but they will be less 11 inclined to take a piece , one without the other. I think 12 we've talked about all of these points. We do recommend it.
fg 13 (Slide.)
V 14 As the tool -- this is the tool for the NRC 15 evaluations of source terms in the regulatory applications, 4
16 provided the uncertainties are taken into account for each 17 application. And I think we've talked about them and we 18 certainly will have more dialogue on the uncertainties. We 19 do have several types of programs ~in place. Those that are 20 looking broadly like the 1150 and those that are trying to 21 look deeply like the QUEST and the QUASAR studies.
22 We feel strongly the uncertainties, like we said 23 before, are dependent on the purpose. Therefore, what we 24 are doing for NUREG-ll50 will only give the uncertainties 25 for its purpose. 'Its purpose being the risk assessment of ;
4 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 140 BRT
~
l five plants.
2 We need to work very closely together in the 3 research area and the regulatory area to develop the 4 uncertainties as the regulatory people move forward, and I 5 do want to no te , again, that the levels of uncertainties 6 and the method for obtaining them, will definitely be a 7 subject of public review should we get to any case where 8 regulatory changes are, indeed, proposed.
9 We also note that you can get additional 10 information from the detailed mechanistic codes, like SCDAP 11 and MLPROG, the data bases for all of the codes from our 12 severe accident research program and the forthcoming risk 13 assessment in NUREG-ll50.
14 MR. KERR: So a user other than the NRC Staff 15 and these two national laboratories, at this point, is 16 really not in a position to take uncertainties into account, 17 other than qualitatively? Is that a fair statement?
18 I don' t mean f air. Is that a true statement?
19 MR. MEYER: I think that's pretty close.
20 MR. KERR: Okay.
21 MR. SILBERBERG: That's it.
22 MS. MITCHELL: I don't want to go into the ,.
23 comments except to just put up table 5.3 out of the 24 document that you have.
25 (Slide.)
1 ()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 141 BRT
%- 1 This is just a list of the bins. We bin 2 everything in the source term world and we have binned all 3 the comments.
4 MR. KERR: For the record we will stipulate 5 table 5.3.
6 MS.'MITCHELL: I wanted to note one here called 7 advisory groups. In that bin are the comments made by many 8 members of the public that quoted the ACRS letter of 9 December 1985. They quoted mainly two specific concepts 10 out of that le tte r . One is that the work represents a task 11 well begun but f ar f rom comple ted. And the second one was 12 that it, in its present form, should not be given much 13 weight in the regulatory process.
14 I guess we agree with both of these. I think 15 that we know full well that the task is not comple ted with 16 a statement of the major sources of uncertainty. One has 17 to realize the task is not over. However, we have well 18 begun. We have made progress. And I believe that it does 19 not -- where we are today does not preclude use in the 20 regulatory arena. I think that it is an iterative process, 21 that we have something now which has reached a relatively 22 stable situation which represents new insights, and that we 23 ought to go and look and see how good a tool it is and 24 iterate , again, on fine-tuning those things that we find we 25 cannot make decisions using this tool.
O ACE-FEDERAL REPORTERS, INC.
202-147-3700 Nationwide Coverage 800-336-6646
27025.0 142 BRT 1 I truly believe that the methods we have today 2 are better than the reactor safety study and better than 3 the assumptions of TID 14-844.
4 We do totally agree that the fission product 5 release and transport technology, which is what we have 6 discussed in NUREG-0956, is only a piece of the puzzle for 7 looking in the regulatory arena.
8 The other things that we talked about at this i 9 meeting that are important are event f re que nc ie s , behavior, 10 performance of the containment, and a method for assessing 11 the uncertainties for any given application.
12 I think that the comparison that we have made g~ 13 with the reactor safety study is about the strongest 1
\_
14 s ta tement that the results today will support, and that is:
15 There is no single multiplicative factor that will tell you 16 about the dif ference between the calculations of today and 4
17 the calculations of the reactor safety study for all 18 sequences, both plants, and all chemical element groups.
19 Do you have any other statements that you want 20 to make?
21 MR. KERR: Other questions?
22 I don't know to whom these questions or comments 23 should be addressed, but you can decide.
24 First, on page 313, there is a statement that is 25 important enough that I think it probably should appear-on O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 ,
27025.0 143 BRT
- 1 the front cover somewhere. It says the following: "The 2 principal unresolved uncertainties in MARCH involved core 3 melt." I think that's an extremely important statement. I 4 would give it more prominence.
5 Then on page 312 there is a discussion of the 6 difficulty of integral experiments being compared to MARCH 7 e xpe r imen ts , and reference is made specifically to TMI-2.
8 It would seem to me, indeed one would certainly want to 9 compare -- want to see if the MARCH package could predict 10 to TMI-2, but I find the description -- a statement which I i
11 don't understand. Maybe can you help me, " Thermal 12 hydraulic descriptions of the TMI-2 accident are only now 13 being developed."
N',g) l 14 I would have thought MARCH would have had all 15 the stuff in it to describe core damage.
16 MS. MITCHELL: You are talking about what was 17 the TMI-2 accident. How many gallons of water.
18 MR. SILBERBERG: I agree, that statement by 19 itself needs to be corrected. Let me qualify it. It's not 20 clear.
21 MR. KERR: If we can't now describe the one i
i 22 severe accident we have so far had --
23 MS. MITCHELL: We don't have a description of i
24 the accident.
25 MR. SILBERBERG: Please let me explain. There O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
,-. - -.~...-,,,,
27025.0 144 BRT (m
k-- 1 is an exercise being sponsored by the OECD, just gotten 2 under way, called the TMI-2_ benchmark calculation of 3 comparison. Worldwide. People have been invited worldwide 4 to participa te.
5 We, the NRC, are participating , both in the 6 context of, let's say in the context of CORCON, SCDAP or 7 the whole package.
I 8 Now, the first step in that program is to 9 develop a se t -- being developed in Idaho for the OECD 10 group, a definition of -- in other words, for people to all 11 start at the same place. That is what this means. It is 12 the thermal hydraulic description that people will be using 13 for benchmarking to start their analysis, will start from 14 that information.
15 Do you want to add something?
16 MR. DENNING: I think you are missing some key 17 words that will help them understand. It's the boundary 18 conditions. How much makeup flow is there, what times; how 19 much letdown flow is there, what times? It's really the 20 boundary conditions which have been very uncertain in the 21 past and they 'are now trying to develop a standard --
4 22 MR. SILBERBERG: Problem.
23 MR. DENNING: A standard problem, but also a 24 good characterization of what really happened as the 25 thermal -- as the boundary conditions.
() l
, 1 l
ACE-FEDERAL REPORTERS, INC.
, 202-347 3700 Nationwide Coverage 800-336-6646
I 1
l 27025.0 145 {
BRT 1
/ '
l 1 MR. SILBERBERG: We need to restate that, so 2 that we state basically that, in fact -- that work is going 3 to be done.
4 MR. KERR: If we are serious about this code's 5 ability to calculate severe accident and we don't know how 6 to calculate the one severe accident we have had -- I'm not 7 trying to be critical of anybody, I know it's difficult --
8 MS. MITCHELL: We will participate in that. I 9 just talked to the people last week out in Idaho who are 10 developing what Rich calls the boundary conditions that 11 allow the description of the accident; they intend to get a 12 best estimate scenario. That is, for this many minutes, 13 this many gallons were put into the vessel and that many 14 gallons were let out between such a period and something 15 else.
16 They will also try to get the uncertainties on i 17 that of what is the minimum flow during this time period, 18 and what would be the maximum flow. And you can 19 investigate how important those uncertainties might be to 20 the answer. But they are only just now describing the 21 accident in detail.
22 MR. KERR: Then I found it interesting to i 23 compare the table on page ES-6 with the table on page ES-4 24 or vice versa , where one had areas of improvement and areas 25 of uncertainty.
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Covesage 800-336-M
27025.0 146 BRT 4
4
\d 1 There's a remarkable correlation between those.
2 MS. MITCHELL: Yes.
3 MR. KERR: For example, one of the significant 4 areas of improvement is treatment of chemical forms of 5 iodine and other fission products.
6 And, number 7 under the areas of uncertainty, is 7 iodine chemical form.
8 I conclude that improvement means I increase the 9 uncertainty, which is a good way to do research, 10 incidentally, because then you dig up more research 11 programs.
12 And, number 3, " improved data base for in-vessel 13 melt progression, hydrogen generation"; core melt 14 progression and hydrogen generation is one of the major 15 areas of uncertainty.
16 I could go on but there are about seven of these 17 that have almost a direct correlation.
i 18 MR. MEYER: I think the main point to get from 19 these is the areas where we have made improvements, in most 20 cases where we have still large uncertainties are the 21 important areas. These are the areas that have a strong 22 impact on source term results and we have , indeed, made 23 major strides in those areas.
24 Uncertainties haven' t increased. We no longer 25 treat iodine as noble gas.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 147 BRT n
1 MR. KERR: Since the whole purpose of the code is 2 to predict core melt, it seems to me that has to be the 3 principal area of un ce r ta in ty.
4 MR. MEYER: I didn' t realize your comment was 5 facetious. I 6 MR. KERR: It's not f a ce tious . ;
7 MR. SILBERBERG: He's saying it's obvious. !
8 MR. KERR: What I'm saying is it would seem to 9 me one would want to say a little bit more. I agree with 10 you, these are the important areas. But what is there that 11 is dif ferent about -- you could have written down the areas 12 of uncertainty almost before you started. Because they are, 13 obviously, and they are still areas of major uncertainty.
14 One could conclude from that that very little 15 progress had been made and I don't think that's necessarily 16 the case. But it seems to me the tables don't make that 17 very clear and the text doesn't make it altoge ther clear.
18 MR. MEYER: I would welcome a suggestion. We 19 had problems with this and I frankly don't know how to 20 treat it.
21 We could walk through this table of major 22 advances and tell you specifically what we had in mind,
- 23 each step along here because these are not just wishy-washy 24 s ta teme n ts. And we could similarly go down the other table 25 and tell you specifically where the hard spots are. They O
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
27025.0- 148 BRT C- 1 are in the same areas.
2 MR. KERR: You've got to do the latter to plan a 3 research program. Now, when I go over to look at the 4 research program I see generalizations that are not much 5 more specific than that table. Somebody has got to do 6 something specific. Maybe it has already been done, but if 7 it has been done, then it seems to me this chapter on 8 research is where it ought to be. ,
i 9 What I'm saying is, I believe that you made a l
] 10 lot of progress in five years , but it is hard to see the 11 effect of that progress, for example , on the new research i
- 12 that is being done.
13 Now, let me go beyond this. I personally think 14 this version of the report is a vast improvement over the i 15 last version, and, it seems to me, you have been responsive 16 to the comments we made earlier. There are some things in 17 here that I would like to see dif ferent, but I think this 18 represents a lot of work and I think it's much better than 19 it was before. I don't know what you think. And I don't 20 think it's much better because of our comments, I think 21 it's much better because you've done a good deal of work on 1
22 it and I think it is clearer and has more useful 23 information. l l
24 The things that I have picked on today are , I 25 think what you want -- well, I wouldn't want you to --
O ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
2'7025.0 149 BRT l rs
\ l anyway --
2 MR. SILBERBERG: Say it.
3 MR. KERR: What you want are the things that 4 need changing. Not the things that are okay, I hope.
5 Other comments?
6 MR. CATTON: Do you want us to make comments on 7 those three items you made earlier?
8 MR. KERR: If you've them written down that's -
9 enough. If you want to emphasize something specifically, I 10 welcome that.
11 If you think the written word tells me what I 12 need to know, I don't necessarily want you to tell it to me 13 unless your writing is illegible.
i O. 14 MR. CATTON: It's not good but it's legible.
15 MR. DAVIS: I agree with your comments about the 16 quality of the report and about what we can say at this 17 time.
18 One thing I think that would improve the report 19 is to add a small segment on what the relative significance 20 is of the radionuclides that make up the source term, 21 because right now a reader has no way of determining 22 whether the comparison between the calculations have any 23 significance or not unless there is some indication of 4 ,
24 which of these materials are biologically significant.
25 In other words, we talked about it be fore , is a
, ()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage M 33 N
27025.0 150 BRT
(~h V 1 10 to 9th difference in ruthenium important or not? And 2 you can't tell whether the arguments over tellurium 3 behavior and iodine behavior are really signiticant unless 4 you know whether these parts of the source term are 5 important to risk. And I think the document that Rich 6 Denning reterred to has in it information that could be 7 directly lifted into this report to give the reader some 8 feel for which radionuclides are really important sources.
9 The second and last comment is regarding a more 10 minor problem, on page 4-6. At the top of that page
.i 11 there's a discussion that troubled me somewhat. You say 12 that thermal hydraulic calculations for small and 13 intermediate break LOCAs showed that the steam generators 4
O 14 were effective in removing decay heat f rom the core and
~
15 therefore they were not core melt sequences.
i 16 MS. MITCHELL: They were sequences that do not 17 lead to core melt.
18 MR. DAVIS: For small and intermediate break 19 LOCAs the steam generators will never be sufficient by 20 themselves to prevent core melt because eventually you are 21 going to lose the primary system inventory and you cannot 22 prevent core melt and, in fact, in most PRAs, small break 23 LOCAs are more significant than large break LOCAs.
24 MS. MITCHELL This is the G sequence, this is 25 the S-1-G sequence , that is an intermediate break LOCA with O
ACE-FEDERAL' REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
27025.0 151 BRT
,o
( )
\/ 1 the loss of containment heat removal. So that the ECC is 2 corking along just fine. So you are making up the 3 inventory. The problem is, if you didn' t remove the he a t 4 f rom the steam, by virtue of -- by means of steam 5 generators, then it would be in the containment and would 6 eventually overpressurize the containment because you 7 cannot remove any heat from it.
8 MR. DAVIS: Maybe I'm taking that statement out 9 of context then.
10 MS. MITCHELL: No. For example, what this whole 11 paragraph is, is about one se quence , the A-G sequence . Why 12 did we look at the A-G sequence? Becausc they felt that fS 13 the containment -- loss of containment heat removal might V
14 be a problem, and that it would be a problem for all of the 15 sizes of breaks. So they ran the thermal hydraulics for S-1, 16 S-2, and A.
17 MR. DAVIS: Okay.
18 MS. MITCHELL: And you are looking for the S-1 19 and S-2, can you remove enough heat through the steam 20 generators that it doesn't deposit in the containment.
21 Now, when the containment fails, everything is 22 going along just fine , even on the A-G sequence , until the 23 containment fails. At the time you depressurize , then I 24 presume we lose injection. It's at that point you can' t 25 make up the inventory anymore?
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
l l
27025.0 152 BRT l
1 MR. EBERSOLE: Suppose you deliberately vent?
2 Is the fuel damaged prior to this sequence?
3 MS. MITCHELL: For the A-G sequence?
4 MR. EBERSOLE: Is there a venting prerogative i 5 without the problem of venting contamination?
6 MS. MITCHELL: In the A-G sequence , it is a fail 7 first, melt la te r . That is, it's doing just fine until the
! 8 containment fails and you lose ECCS.
i
- 9 MR. EBERSOLE
- We have a problem here both in 10 the legalistic context and practical context. There is not i
11 now in place any permissive on the operator to vent to save 12 his soul and his plant.
13 MR. SILBERBERG: Mr. Ebersole, I believe the O 14 last I heard was that NRR was in fact trying to develop 15 -- not develop, but at least codify some procedures along 16 those lines.
17 MR. EBERSOLE: It's been a process going on for 18 at least 10 years.
i 19 MR. SILBERBERG: Well, they seem to be getting 20 very close. At least the impression is that it is --
21 MR. EBERSOLE: For the record, I would like to i
22 say I sure would like to see it af ter all these many years.
23 There's nothing now in place that gives an operator the j i
l 24 permission to release radiation in order to save the day.
{
25 Any radiation, I mean.
i i
1 t !
1 ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
l l
27025.0 153 l BRT <
l r~s '
1 MR. KERR: Pete , does that conclude your comment?
2 MR. DAVIS: Yes. Thank you.
3 MR. KERR: Incidentally, Mr. Davis has pointed 4 out the B sequence is mentioned several times. Mr. Davis i 5 points out, I believe, that some analyses indicate that i
6 even if you expose that low-pressure system to full system 7 pressure, it won't burst.
8 MS. MITCHELL: So, fine. That may or may not be 9 the case . I don't know.
10 MR. KERR: Well, it has some fairly significant i
11 effect on source terms, doesn't it? Whether it does or not?
12 MS. MITCHELL: If we are discussing the j v 13 frequency --
14 MR. KERR: We are not discussing frequency at 15 all. If it doesn't burst, it doesn' t burst.
! 16 MS. MITCHELL: The frequency of that bin is se t l 17 to zero.
18 MR. DAVIS: The IDCOR analysis shows --
4 19 MS. MITCHELL: IDCOR didn't look at Surry. It 20 looks at Zion.
21 MR. DAVIS: If IDCOR looked at Surry they would 22 probably come to the same conclusion at Zion because the 23 pressure in the two plants is the same, 600 psi. IDCOR now 24 concluded that the pipe would not rupture , but the 25 low-pressure coolant pump seals would fail, giving. you a
(~)
4 ACE-FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coverage 800-336-6646
27025.0~ 154 BRT I small break LOCA, which would extend the accident 2 substantially and gave you an entirely different source j 3 te rm .
4 MS. MITCHELL: It would indeed give you a 5 .different source term. There's no doubt about it, if you 6 had a S-3 sequence instead of the V sequence out in the 7 safeguards building, you would get a different source term
! 8 from the one that is here. Please speak to 1150, when they l
9 come, about what is the frequency.
4 10 MR. KERR: We 'll speak to 1150. We j ust wan t
! 11 the word to get around. We are not sure ycu people talk to i
12 each other.
4 13 MS. MITCHELL: They heard.
f~s d 14 MR. EBERSOLE:
1 Was it the seal that failed or .
l 15 the suction part of the casing?
16 MR. DAVIS: Seal.
i 17 MR. EBERSOLE: Generally pumps have a 18 low-pressure capability on the suction side of the casing.
]
19 Was that okay?
20 MR. DAVIS: I don' t know how extensively they 21 looked at it. The IDCOR analysis has _ not been presented 22 for review.
j 23 MR. KERR: It may be worth looking at.
24 Further comments or questions? Of these - people?
I 25 Well, again, we appreciate your coming. What I A)
\.
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27025.0 155 BRT O'. I would suggest that you do on Friday --
, 2 MR. SILBERBERG: That's what I would like some 1
3 help on.
4 MR. KERR: The presentation you made today I 5 think is appropriate for this meeting. It might be more 6 ef fective Friday, if you go through our report and say what 7 you have done about what we said in the report. You'll get 8 a lot of questions, you know from previous experience. But 9 if you could hang it on that, it seems to me the full 10 Committee would be in a better position to consider what 11 they wanted to say in the letter, or what they will decide 12 about the letter, rather than try to present the whole 13 thing.
1 O 14 MR. SILBERBERG: Then the report should be the 15 ve h icle .
j 16 MR. KERR: You think about it. If you decide
. 17 there 's a be tter approach, it's okay with me. You are 18 trying to get something across.
19 MR. SILBERBERG: That's fine. The only question, 20 is there anything in addition to that that would be usef ul?
21 MR. KERR: Not from my point of view. You know, 22 be prepared to respond to whatever.
23 MR. SILBERBERG: If this will get the job done 24 be t te r , then that's what we'll do.
i 25 t1R. KERR: This is just my guess.
()
.I ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 33 6 6646
27025.0 156 l BRT MR. SILBERBERG:
1 You know better than I. It 1
i 2 sounds like the way to go.
4
- 3 MR. KERR
- Okay.
i
- 4 MR. SILBERBERG
- Fine. We'll go that way.
i, 5 MR. KERR: Anything else? ,
j 6 I thank all of you..
7 (Whereupon, at 5:20 p.m., the meeting was 8 adjourned.) j 9
- i. 10 11 f
4 4 12 lO i
14 ,
15 i
, 16 i 17
! 18
! l l 19 i !
20 !
\
21 l 22 i 23 24 25 O
. ACE-FEDERAL REPORTERS, INC. -
202 347 3700 Nationwide Coverage 800-336-6646 t
---....--,-,..--e- ,-.m,.., ,,--.m--, .,,,,m.,,,-_,-ec-,y.,-r-,-w-w-.mmmm. .,y .wmy,vm,--......--,~-4,-y,wr-w-w--m..n3-
CERTIFICATE OF OFFICIAL REPORTER O
v 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 SEVERE (CLASS 9) ACCIDENTS AND NUCLEAR PLANT CHEMISTRY ,
DOCKET NO.:
PLACE: WASHINGTON, D. C.
DATE: TUESDAY, JUNE 3, 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) _ - A IM (TYPE JO BREITNER Official Reporter ACE-FEDERAL REPORTERS, INC.
Reporter's Affiliation O
i 1
3 l l
. \
l '
I i
i .
1 ,
ACRS SUBCOPNITTEE NETING .
j ,
1 ON 1
l l
NUREG-0956 FINAL REPORT JUNE 3, 1986 t
M. SILBERBERG !
1 l
i j J. MITCHELL 2
{. R. MEYER '
i i
i l
l OFFICE OF NUCLEAR REGULATOPY RESEAPCH I t
i i
i i
s
._,.,,.---___,,...___,.._._.___,_______,_.__.__m.__._ _ . _ _.__ _ _ .
1 O
I l 2. Introduction
. O i
i n
1
- 1 O
I l
l
.c-,. -- , , , , . , , - - - - - , - - .--,n,--..-- - , . , . . , - . . _ _ , _ . - _ _,_n,
O e
PRINCIPAL IfNESTIGATORS R. DENNING, BCL J. GIESEKE, BCL T. KPESS, ORNL D. POWERS, SNL T. PRArr, BNL O
INDEfBDENT CONSULTANT H. ISBIN 1
1
l l
O I
- 3. Purpose, Potential Uses and Limitations of NUREG-0956 O
O l
Source Terms in the Regulatory process 1960 1970 1980 1990 Sandia Siting Study NUREG/CR 2239
/
Source Term Reassessment NUREG 0772 Severe Accident Severe Accident Research Program Risk Rebaselining Program DNL Accident Consequence Study WASH 740 k % m Reactor Safety Source Term Risk Study Reassessment Study Assessment Study
\ WASH-1400 N U R EG-0956 NUREG-ilio TID-14844 \
In-Plant Release K Assumptions Containment WASH 1400 Source Term Release Tables Analytscal Procedure Rin Assessrnent Performance Credit Sitin R I tions Emergency Planning
" " Planned Regulatory Planned Regulatory "P Implementation Regulatory Guides S)e no implementation Design Basis Accident PRA Analyses Assumptions figure 2.1 History of source term assessment and relationship to regulatory process.
O O O .
O POIENTIAL SOURCE TERM CHANGES 4
NEAR-TERM. INTERMEDIATE LONG-TERM
~
REVISED TREATMENT EMERGENCY PLANNING SITING OF ACCIDENTS IN EIS REMOVAL 0F SPRAY CONTAINMENT LEAK RATES ACCIDENT ADDITIVES (PWR) AND INTEGRITY (Ag,T) MONITORING O
SUPPRESSION P0OL ENV. QUALIFICATION CPEDIT (BWR) 0F EQUIPMENT
\
SAFETY' ISSUE EVALUATION (0",OL ROOM HABITABILITY AND AIR FILTRATION SYSTEMS !
O .
l
?
- - _ - -- , , , - , _ _ . . , , . . , . - . . . . , -- . , - . . , , , - , . , _ . , . , . , -,-,--.---n ,-.n-
Source Term Reassessment Program Relationships Source Term Analysis ReSearch ' Regulation BMI 2104 l l
Uncertainty Estimates SAND 84 0410 ,
Status of Validation ORNL/TM 8842 New Sequence Analyses m Source Term Changes in NUREG/CR-4624
- Reassessment Study 3 Source Term Based (NUREG 09561 Regulations l
Quality Control l NUREG/CR-4629 I Technical Expert Peer Review )I l BMI-2104,Vol VII .
l APS Review Severe Ac;ident i Reviews of Modern Physics Risk Rebaselining l Severe ccident Program "
Implementation l
Public Comments 3 NUREG-ll50 l NUREG-0956 y Figure 2.2 Relationships of the NRC source term reassessment program with regulatory im mentation.
l O l
\
i l
1 1
i l
1 i
i l
1
- 4. Content of Final Report O
I
]
I O ,
I I
e b
n..,_,_,n., - .
- 1. STATUS AND APPLICATIONS .........................................
1.1 Applications of Source Terms ................... ...........
- 1. 2 Research and Technical Issues ..............................
Qgg +1. 3 Purpose of This Document ...................................
V 1.4 Contents of This Document .................................. -
- 2. HISTORICAL PERSPECTIVE OF SOURCE TERM ASSESSMENT ...............
2.1 Accident Consequence Study (WASH-740) ......................
2.2 Maximum Credible Accident (TIL-14844) ......... ............
2.3 Reactor Safety Study (WASH-1400) ........ ..................
2.4 Review of Source Term Technical Bases (NUREG-0772) .........
2.5 Sandia Siting Study (NUREG/CR-2239)................ .... ...
2.6 German Risk Study ..........................................
2.7 Sizewell-B Source Term Estimates .... .. ..................
2.8 Current Source Term Reassessment (NUREG-0956) ..............
' 2. 9 IDCOR and Other Source Term Calculations ..................
)[qtu3 2.10 Reviews of Existing Calculations ..........................
2.11 Risk Assessment Study and Its Associated Uncertainty Study (NUREG-1150) ........................................
. G O 9
O O O
- 3. AN INTEGRATED APPROACH TO SOURCE TERM ASSESSMENT.................
3.1 NRC's Source Term Code Package..............................
3.1.1 Overall Analysis for the Reactor and Containment (MARCH) .............................................
3.1.2 Detailed Flow Rates in the Reactor Coolant System (MERGE) .............................................
3.1. 3 Detailed Core-Concrete Interactions in the Containment (CORCON) ................................
3.1.4 Fission Product Generation in the Fuel (ORIGEN) .....
3.1. 5 Release from the Fuel (CORSOR) ......................
3.1.6 Retention in the Reactor Coolant System (TRAP-MELT) .
- 3.1. 7 Release from the Core-Concrete Melt (VANESA) ........
3.1.8 Aerosol Deposition in the Containment (NAUA) ........
3.1.9 Retention in Water Pools (SPARC) ....................
3.1.10 Retention in Ice Condensers (ICEDF) .................
3.2 Verification and Validation of source Term Code Package.....
3.2.1 Verification.........................................
3.2.2 Validation...........................................
3.3 Uncertainties in Source Term Analysis.......................
3.3.1 Sandia QUEST Study...................................
3.3.2 Brookhaven QUASAR Study.............................
3.3.3 NUREG-1150 Estimates................................
3.4 Summary Evaluation ........................................
i
O Fission Product Thermal Hydraulic Transport j ;
Behavior
______________q l ORIGEN MARCH I
I Fission Product ,,,,
Inventory in Fuel Bhst M l
E h" Reactor Coolant System Molten Core, and
=
gl l el CORSOR l Containment l
g al i l l
8 Release l Retained Tl from Fuel ; in Fuel l ij i l I e ' i I i
I .
l l
Reactor Coolant System - - ___
Detailed Temperature. l Transport and e Pressure. and Flovv in Reactor Coolant System l R et e n t.on j i j l
l_ _ _ _ _ ______
____a l l
r i i I l CORCON I VANESA a
- l r i E 3 Release from =- Detailed Core . J Core Concrete Mett Concrete Temperature and interact.ons 3l 8
4 l l
"l l
!l , ,
j j N AU A. SPARC, ICEDF l
Containment Transport l l and Retention l
l L____ __ _________.;
1 r Refease of fission product $ to the environment: Source Term O
Figure 3.1 .BMI-2104 suite of codes as used in the source ,
term reassessment.
( .
I O lNPUT l
' Plant and Sequence Description i
i i f MARCH 3 (M ARCH 2, CORCON MOO 2. CORSOR.Mi l
i f y
TR AP-MELT 3 VANESA (TR AP-MELT 2. MERGE)
O 1r
> N AU A/SPARC/ICEDF e
' f Release of Fission Products to the Environment: Source Terrn Figure 3.2 Source term code package.
O
,,,.--&,y. . - ---,--.----.-mm,. ,-~yy.,--. - - - - . - - .-y- ,-myy.er-,+<--r.-~. --wy,,+-,, . -,w-w,--g,,m-..-c-
F
- 4. EXAMPLE CALCULATIONS FOR SELECTED ACCIDENT SEQUENCES.............
4.1 Summary Description of Accident Sequences Analyzed .........
4.2 Fission Product Inventory ............... .. ...............
4.3 Fuel Temperatures ..........................................
4.4 Fission Product Release In-Vessel ..........................
4.5 Upper Plenum Temperatures ..................................
4.6 Retention in the Reactor Coolant System ....................
4.7 Fission Product Release Ex-Vessel ..........................
Ide*3 y 4.8 Integrated Release from Fuel In-Vessel and Ex-Vessel .......
4.9 Containment Leakage and Failure ............................
4.10 Aerosol Retention in Containment ...........................
} ga) ; 4.11 Quality Assurance Program and Audit Calculations ...........
4.12 Summary of Sample Source Term Calculations and Comparisons with the Reactor Safety Study Results ....... ... . ......
A Euh -
- 4.13 Parametric Analysis of Ef fect of Iodine Chemical F orm . . . . . .
4.14 Uncertainty ................................................
4.15 Conclusions ................................................
9 O e .
N i
4 1
1 7
1 i
i i
i 1
- i
- 5. REVIEW OF SOURCE TERM WORK ......................................
4
,i
. 5.1 Technical Expert Peer Review ............................... '
! 5.2 American Physical Society Study Group ......................
- 5.3 Industry Degraded Core Rulemaking Program ..................
j 5.4 American Nuclear Society ...................................
5.5 Other Organizations ........................................
y ' 5. 6 Other Reviews ..............................................
j , 5. 7 5.8 Public Comments ...........................................
j Summary .................................................... .
1 1
1 4
i i
J 4
3 3
4
't t.
4 I
- 6. RESEARCH ........................................................
6.1 Identi fication of Source Term Issues. . . . . . . . . . . ............
P' 6. 2 Status of Research in Important Source Term Areas...........
6.2.1 Natural Circulation in the Reactor Coolant System....
6.2.2 Core Melt Progression ...............................
)) e.sg> 6.2.3 Steam Explosions.................. ..................
6.2.4 High-Pressure Melt Ejection............ . ...........
6.2.5 Core-Concrete Interactions...........................
6.2.6 Hydrogen Combustion..................................
6.2.7 Iodine Chemical Form.................................
6.2.8 Fission Product Revaporization...... ................
w 6.3 Plans for Source Term and Containment Loads Research........
6.3.1 In-Vessel Melt Progression Research..................
. 6.3.2 Fission Product Behavior Research.. .................
ggg 6.3.3 Containment Loads Research...... ....................
6.3.4 Ex-Vessel Fission Product Behavior Research..........
0900 6.4 Severe Accident Computer Codes... .. .......................
b.1 6.4.1 Two-Tier Code Strategy................ ..............
6.4.2 Verification and Validation..........................
6.5 Summary.....................................................
9 O e . .
O O O '
~
i i
i 4
- 7. CONCLUSIONS AND RECOMMENDATION ..................................
j 7.1 Integrated Approach to Source Term Assessment...............
! 7. 2 Insights from Selected Accident Sequences...................
7.3 Review of Source Term Work..................................
7.4 Research ...................................................
7.5 Recommendation .............................................
APPENDICES l A. BASIC REACTOR SYSTEMS AND CONTAINMENTS ..........................
4 ' P B. IDCOR-NRC TECHNICAL ISSUES ......................................
l N '
- C. REPLIES T0 PUBLIC COMMENTS ......................................
1 s
t i
I 4
i i
l J
l c
l l
1 i
O l
i I
l l
S. Conclusions and Recommendations O
1 l
O .
L I
L. . - . ,--_._._---.---,-_,.,.,____,._,,,._.,,,..,,,_..,-...-.-.,.-..,,_,,-.n..__,,.__,,._,-.,-.,,-_.-..n. - _ - . , _ _
C0f1CLUSim 1. Aff APPROACil TO SOURCE TEPN ASSESSMEt!T US!fJG A FPAMEWORK 0F COUPLED At!D IffTEGRATED CmPUTER CODES HAS BEEll DEVELOPED; Tills FRAMEWOPK CAf4 ACrrfftDATE IMPPOVEMEf1TS EXPECTED FR0f1 ONGolflG RESEARCH.
O O O .
D 9 O O O CONCLUSION ?. THE COMPLEXITIFS OF SOURCE TEPM PHEtKNENA HAVE REStJLTED IN SOPHISTICATED COPFUTER COCES AtlD LARGE AMOUNTS OF NltERICAL It! FORMATION THAT REQUIRE THE USE OF QUALITY ASSURANCE MEASUPES.
L___ _ _ _ - - - - - -
CmCLilSI0ts 3. VALIDATIori OF THE NRC'S S0tfPCF TERM CODE PACKAGE IS
. SUBSTAffTIAL Ifl SWE AREAS BUT LIMITED If1 OTHEPS; MORE COMPARIS0fl9 0F COMPUTER CODES WITH EXPFRIMEtlTAL RESULTS ARE flFEDED.
o O O O .
O O O l -
l CoficLustoftI I. Tm cuRREllT SEVERE ACCIDFitT IMTA BASE Af:D THE PKi10MEfl0 LOGICAL MDDELS Ifl THE NRC's SOUPCE TERM CODE PACKAGE PEPPESEffT A MAJOR ADVNICE Irl TECHf.0 LOGY COMPARED WITH THE PEACTOR SAFETY STUDY,
CONCLUSI0fi5. THE SOURCE TERM CODE PACKAGE DOES t'OT IrlCOPPOPATE IffrEffTIorlAL BIASES, A FEATURE THAT WAS BELIEVFD TO HAVF BEEft If.CORPORATED Ifl PPEVIOUS METHODS; IT IS t0T POSSIBLE TO PROVIDE NI If1HERENTLY CoffSERVATIVE MARGit! Ill A SOUPCE TERM CODE THAT IS TO BE USED FOR MAtlY SEQUENCES NID MAfW PURPOSES, O O O l
i ,
j O O O 1
i t
i :
1 I
i ;
i 1 I
I '
i l- CONCLilSION 6. (DJCERTAINTIES IPI SOURCE TERM TEGfl0 LOGY ARE LARGE, AND TIEIR IPf0RTANCE PilST BE OllAffTITATIVELY ESTIPMTED FOR EACH TYPE OF ;
APPLICATI0Fl.
. t I- !
1 l
- I 4 a i i l
e 4
i 1 !
1 ,
4 :
! i I
- k. i i I I
i i
i ;
- i i !
1 1
i
~
i
.. 1
CofJCluSr0f4 7. A LAPGE flUMBER OF SEVEPE ACCIDEffT SFOUEf1CES HAVE BEEN ,
Af1ALYZED WITH THE NRC'S SOURCE TERM CODE PACKAGF NID ITS PREDErESSOR, THE EMI-210l4 SUITE OF CODES; THESE CALCULAT10flS HAVE EXEPCISED THE CODES OVER A WIDE PAf1GE OF CoffDITIOf1S.
O O O
t
~ '
I 6 o o i
i r
i t
j -
i 1 t I
I i -
t i
t i !
J i e !
l CONCLUSION 8. SEVERE ACCIDFfJT SEQUENCES APE COMPLEX, AffD EXPERIEfJCE
- i. -
. HAS SHOM1 THAT THIS COPFLEXITY PEQUIRES CARE AND ATTEtiTION TO i i .
r i
DETAILS ON TIE PART OF A COPPUTER CODE USER TO REPRESEllT TtF PLMT i I
i Af0 THE SEQUENCES ADE00ATELY Ifl TIE CODE INPUT.
4 s
1 i
- I a
! I I i
! f i
i i
1 .
a :
i l [
u . , _ _ _. _ _ -- - _ . _ _ . _ - _ _ . - .
C0f!CLUSIOft 9. SOURCE TEPMS WEPE FollflD TO DEPEf!D STP0flGLY Ofl PLN1T ,
DESIGft AfiD C0f!STRUCTIOfl DETAILS, THUS MAKifJG DEVELOPMEtlT OF tlSEFUL GEf4ERIC SOURCE TFRiS DIFFICULT.
O O O
. . .. t
.O O O t i e \
fs 1
} i I
i t
t CONCLUSION 10. CONTAlf#ENT PEPFORMANCE (SURVIVAL, FAILUPE, OP BYPASS),
t i
i-milch IS NOT E CHANISTICALLY ANALYZED IN SOURCE TERM CODES, IS A ,
MAJOR FACTOR AFFECTING SOURCE TERMS.
I t
I i
i i
4 i
i i
i
- ----t_ __.- -- _,-- .....--.--,- .-..~-,.,. ..-- -..- .. - -.---,- ...
C0f1CLIIStof: 11. THE NIALYSES PERFORMED TO DATE WITH TllE SOURCE TEPM CODE PACKAGE SUGGEST THAT GEf1EPALIZATinfis ARE INAPPROPRIATE SitlCE LARGE FAC-TORS OF REDtlCTICt1 IN SUURCE TEPt1S WERE ?!OT FOUfJD FOR ALL SEQUEtlCES AS REPORTED Ifl OTHEP STUDIES.
O O O .
O O O i
i t I
i
)
t 1
- i. ,
i :
i i j ,
t i
l i -
i J*
CONCLUSION }2. THE PRC'S SOURCE TERM TECHNOLOGY HAS BEEN EXTENSIVELY ;
i P
REVIEWED; NEVERTELESS, MOST CO M NTERS FAV0 PED FURTHEP REVIEW AND
- ADDITIONAL PUBLICATION OF THE MODELS IN THE SOUPCE TERM CODE PACKAGE.
I I
l a ;
I f
)-
I i
i l
4 4
i !
t
- 3 ,
i-1 i
j .
C0f!CLUSinti33. ADDITIONAL METijoDS OF SOURCE TERM ANALYSIS EXIST, PARTICULARLY It' THE MAJOR INDUSTRY SOURCE TERf4 CODE PMP, BUT TilESE METHODS ARE NOT AVAILABLE TO TlfE NRC AND, DIEREF0PE, ARE fl0T DISCUSSED Irl DETAll OR EVALUATED Ifl THIS PEPORT.
~
S O O O
O O O l
l COfJCLUSI0r! 14. THE REVIEWERS At4D THE NRC STAFF ARE Irl GEt'EPAL i AGREEMEfJT Off AREAS It1 THE SOURCE TFF11 TECHNOLOGY THAT f!EED FURTHER RESEARCH.
i I .
t l
CorlCLUSinti15. THE NRC'S SEVERE ACCIDEllT RESEARCH PLAM HAS BEEll RECEf4TLY REVISED Af1D WORK liAS BEEN FOCtISED IN AREAS OF Tile SolJPCE TERM TECit0 LOGY TilAT NEED FURTHEP STUDY, W
O O O -
.o oo o, 7
CONClusroN 16. FORMALIZED VERIFICAT10f! AND VALIDATION PROGRAMS ARE t
Ifl PLACF FOR THE NRC'S IfJTEGRATED SOURCE TERM CODFS, e
ia i
r THE SOURCE TERM CODE PACKAGE IS RECOMMENDED AS NI INTEGRATED ANALYTICAL TOOL FOR NRC EVALUATION OF SOURCE TERMS IN REGULATORY APPLICATI0t!S PPOVIDED THAT UffCFRTAltlTIES ARE TAKEt! INTO ACCOUNT FOR EACH APPLICATI0tl. ADDITIONAL If1 FORMA-TIOf! Of1 l!!1 CERTAINTIES CAN BE OBTAINED FROM NRC'S DETAILED MECHANISTIC CODES, THEIR DATA BASES, FUDTHER RESULTS FROM NRC'S SEVERE ACCIDENT RESEARCH PROGRAM, AND THE FORTHCOMING REASSESSMEtIT OF RISK FOR SIX PfFFREtJCF PLAtlTS, NUREG-1150.
.p i
Y T
w d
W~
O O O _
s
l
~
O
- 6. ACRS Letter and and Public Comments O .
O
I Table 5.3 Categories of comments on Draft NUREG-0955 lll Topic of comment Accident Sequence Advisory Groups Aerosols Applications Most Be Known Bias Changes in Source Terms Chemistry--Iodine Chemistry--Radionuclides Consider Other Studies Containment Performance Control Rod Materials Core-Concrete Interaction Core Melt Progression and Modeling Debris Coolability Decay of Isotopes Decisionmaking Definition of Terms Documentation Equipment Survivability External Events Fuel Burnup Generalizations High-Pressure Melt Ejection Hydrogen Combustion Hydrogen Production Inadequate Research Industry Effort Natural Circulation NUREG-1150 Operating Experience Operator Actions Outdated Technology Peer Review Probabilistic Risk Assessment Regulations--Relax Regulations--Recommendations Report--Recommendations, Revaporization/Resuspens ion Risk -
Secondary Structures Skill Needed To Use Codes s Steam Explosions Technical Content Of NUREG-0956 Thermal Hydraulics TMI-2 Data g Uncertainties--General W Uncertainties--Quantified DRAFT 05/22/86 5-16 NUO956 CH 5 NEW VERSION
< -