ML20151R282
| ML20151R282 | |
| Person / Time | |
|---|---|
| Issue date: | 08/04/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1685, NUDOCS 8808120111 | |
| Download: ML20151R282 (190) | |
Text
. - . .- _
-.~..e . . . . . . . . . _ . ., - - - -
1 m m g v 0 n O UNITED STATES NUCLEAR REGULATORY COMMISSION l
,=======cs-as===s================s===============================
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS In the Matter of:
SUBCOMMITTEE ON ACRS PHILOSOPHY, TECHNOLOGY AND CRITERIA O
t
? ~ ~ .., ,
' 0' -%p., ,_
~T K.
I I DATE: August 4, 1988
~*
- ~ .IVk.if,Ik LOCATION: Washington, D.C. .
PAGES: 1 through 185
=
= am sm e 4 an = = = an = = = me m um mm as am am m m m m m m m as sa m m m m m en sm e == m e = am me ss am en e m m e =
HERITAGE REPORTING CORPORATION O osuntRenwr 1220 L Seest, N.W.,5de det 1
l
- *',4 WasMagton, D.C. 20005 gg81 @[ F. g -
1-15ss (ggg)ggg,4ggg
- i i
1 PUBLIC (JOTICE BY THE 2 UllITED STATES tlUCLEAR REGULATORY COMMISSIOt4'S 3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS l 4
i 5
6 7 The contents of this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards (ACRS),
10 as reported herein, is an uncorrected record of the discussions 11 recorded at the meeting held on the above date.
12 No member of the ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors
? 1 14 inaccuracies of statement or data contained in this transcript.
15 16 17
~
18 - - - - - - - - =- ,, - ~ _ _
mw I 19 f d TRO4 !
~
20 DELLTE B, hWI f
21 REIURN ORIGINAL 70 B. hWITE, o l ACRS-1016-H y 4
22
'1 hanks! g 23 Barbara Jo hhite -
4-1407 ij 24 '
.m m.
25 '
I
%4 p Heritage Reporting Corporation (202) 628-4888
\-
1 UNITED STATES NUCLEAR REGULATORY. COMMISSION
() 2 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8
3 4 In the Matter of: )
)
$ Subcommittee on ACRS )
Philosophy, Technology and )
6 Criteria )
-7 8
. Thursday,.
August 4, 1988 9
Room 1046 1717 H Street 10 Washington, D.C.
11 The above-entitled matter came on for hearing, 12 pursuant to notice, at 8:30 a.m.
13 ACRS MEMBERS PRESENT:
()O 14 MR. DAVID A. WARD Chairman 15 e Research Manager on Special Assignment E.I. du Pont de Nemours &' Company 16 Savannah River laboratory Aiken, South Carolina 17 DR. WILLIAM KERR 18 Professor of Nuclear Engineering Director, Office of Energy Research 19 University of Michigan Ann Arbor, Michigan 20 DR. CHESTER P. SIE$S 21 Professor Emeritus of Civil Engineering University of Illinois 22 Urbana, Illinois 23 MR. CHARLES J. WYLIE Betired Chief Engineer 24 Electrical Division Duke Power Company O 25 Charlotte, North Carolina Heritage Reporting Corporation ten =
2 1 ACRS COGNIZANT STAFF MEMBER:
2 Dean Houston 3 ACRS CONSULTANTS:
4 Walter Lipinski John Lee 5 ;
NRC STAFF PRESENTERS:
6 George Sege 7
Thomas Ryan ,
Leonard Soffer g Fuat Odar David Diamond 9
1 1
10 11 ;
12 O
O 14 15 .
1 1
16 17 1
18 19 20 21
'2
~
O 23 24 O 25 Heritage Reporting Corporation msu
1 . _ _ . - . _ _ _ __ ._
3 1
P R Q g 3 3 p 1 N,g S 2 MR. WARD: The meeting will now come to order.
3 This is a meeting of the Advisory Committee on 4 Reactor Safeguard, Subcommittee on Safety Philosophy, Tech-5 nology, and Criteria.
6 I'm David Ward, the subcommittee chairman. Other 7 ACRS members in attendance are Mr. Kerr, Mr. Wylie, and we 8 expect Mr. Siess to be here.
9 We also have ACRS consultants, Mr. Lee and Mr.
10 Lipinski.
11 The purpose of the meeting this morning is to re-12 view first the status of NUREG 1251, implications of the Q 13 accident at Chernobyl. And second, the NRC research program 14 at Brookhaven National Laboratory which addresses studies 15 with regard to severe reactivity transients.
] ,
16 Mr. Dean Houston on my right is the Cognizent ACRS 17 staff member for the meeting.
l 18 Rules for participation in today's meeting have 19 been announced as part of the notice of the meeting previously 20 published in the Federal Register on July 22nd. A transcript 21 is being kept and will be made available as stated in that 4
22 notice.
O 23 I request that each speaker first identify herself 24 or himself and speak with sufficient clarity and volume so O 25 that she r he will be readily heard.
Heritage Reporting Corporation a
(Set) matte
~
1
'4 1 I We've received no written statements nor requests ,
I
(} 2 to make oral atatements from members of the public. l l
3 I'm ready to just go to the agenda. And I believe ,
( 4 Mr. Sege is going to start off.
5 Do any of the members have anything that they'd 6 like to say at this time?
7 (No response.)
8 MR. WARD: George.
9 MR. SEGE: Thank you, Mr. Chairman.
10 l I am going to start things off by reporting 11 generally on the status of the Chernobyl implications assess- ,
12 ment and including a description of the Chernobyl follow-up 13 research that carry out the recommendations made by the
( ))O 14 Implications Assessment Report, and the status of that work 15 generally.
16 Just to recapitulate the general status that I 1
17 think for the most part members of the subcommittee are al-l 18 ready acquainted with, the factfinding report which NNRCB l 1
19 participated in by several other interested federal agencies )
20 and industry groups, that was issued in January of last 21 year and some limited editions were made subsequently and 22 it was issued in December of last year.
s 23 The implications-of the Chernobyl accident were 24 U.S. commercial nuclear power reactors has been issued for
() 25 public comment, was issued in September of last year. Those Heritage Reporting Corporation (see) 6ss-mes
- _ _ _ . _ _ ~ - _ , , _ - , _ - - . - _ . . _ . . . ~,-- . .
. . . . - ~ . . ,, .. _ ._ _ , , ,.
5 i comments have been received and a resolution of them pre-() 2 pared and the final version of the report is in the final 3 stages and was submitted to the Commission for its approval.
( 4 Previously, as you may recall, the Commission had
~
$ approved the issuance of the draft for public-comment but 6 reserved final approval of the report after considering the 7 revised version that would include responses of the public 8 comments.
l 9 In the course of the preparation of the report we go have had the benefit of several reviews with the subcommittee i
gj and the full Committee and just for reference I have listed j 12 the dates of the resulting advice.
13 The Chernobyl follow-up research plan is a.collec-()o 34 tor for all the different recommendations that were made as l
15 a result of study of the various Chernobyl-related issues '
16 and are described later on. The work on these follow-up 17 studies is in progress now and as Chairman Ward mentioned, l s
l l
18 the particularly significant project of reactivity transients l 19 study will be reviewed with the subcommittee in detail after-20 our general overview.
l I
21 The subcommittee has the full public comments as I
22 well as the draft report resolution of those comments and '
' () 23 I thought I should put on the -- the list of the various
{
I a
24 comments by some sort of categories with comments from the
(]) 25 nuclear industry, NUMARC, Westinghouse, state agencies, Heritoge Reporting Corporation
<= m me
6
)
l Illinois, New York', North' Carolina, three national labs, the 2 public interest groups, the Nuclear Information Resources f
(~}
, 3 Service, Ohio, Citizens for Responsible Energy.
(} 4 And we had received comments from then Commissioner !
5 Asselstine in the course of the earlier; review of the draft i 6 and the Commissioner directed that former Commissioner Assel-- '
7 stein's comments be treated as part of the public comment a 8 rerolution.
9 ANBEK is a private firm that is in the potassium 10 iodine business and they have commented.
i 11 Although the comments were quite voluminous and j 12 extensive, for the most part the changes were limited to l 13 some corrections, some clarifications of statements that were 1
]
(])(:) 14 not that clear, some limited modifications of assessments in 3
i is some specific areas. These were fairly numerous but the ,
16 major aspects of the assessments have not been substantially 17 changed as a result of the consideration of resolution of 18 public comments.
19 The two items that have been of some greater sig-l 20 nificance than the rest are in my view have been the addi- .
- 21 tion of a recommendation on the study of violations of proce-
! i 42 dures, systematic study of the violation of procedures at 2
() 23 U.S. reactors to try to determine the nature and suverity i
i <-
24 and possibly review what regulatory a if ery, might ,
() 25 be appropriate to consider. This wc '# the items Heritoge Reporting Corpt )
n=) = == . j l
7 I that Tom Ryan in discussing the Chernobyl Follow-up Research
(} 2 Plan in more detail in the human factors area will comment 3 on in considerably more detail.
A
\~/ 4 And we deleted the recommendation to study the 5 merits of having a high-level full-time onsight safety
\
6 manager at each plant on the basis of comments and considera- i l
7 tions of those comments which suggested that there is enough 1
8 of a risk of such a step being counterproductive-that we ;
9 thought that we could eliminate it on the basis of the assess- ,
10 ments reported in the assessment report itself without.further 11 follow-on study. The fear was that having one person desig- !
12 nated as the safety person it might detract from the'atten- i 13 tion to safety that everyone whose actions at the plant have a
j
()(:) 14 a bearing on safety must have in mind in the course of con-15 ducting their activities.
4
, 16 The Chernobyl Follow-up Research Plan contains all l
] 17 of the study items recommended in NUREG 1251, the Chernobyl i
18 Implications Report, and is limited to that work. In the 19 report, this work is recommended in the conclusory and re-
- 20 commendation sub-sections of each section dealing with par-1 21 ticular issues and here in the plan these are collected for 4
l 1
i 22 convenience of management and oversight --
i(
. i 23 In all of these the work is that which annexes I
24 to Chernobyl that is reasonably direct and clear and sub- l i
l
() 25 stantial, but we have always recognized that Chernobyl as an Heritoge Reporting Corporation (393) Matte
--, - - _,. ,. . .. .. ..,r,- . _ , . ._. _. --
r_. . . .. - .. , _ , , . . . . , , , . . _ . _ . - -
8 4 o I experience will have effects, substantial effects, in a wide
() 2 - range of regulatory considerations and decisions in the future 3 but where that effect is somewhat remote and indirect we have 4 not included those specifically in the Chernobyl Follow-up i
- 5 Research Plan.
6 The Plan is intended to be closed out in the middle -
7 of next year with'the expectation that to the extent possible 8 we will resolve the matters signaled ~for study by that. time. ,
9 There are a number of items, however, which do not11end them- .
4 to selves to closing out in five years. They clearly require 11 a longer period of study. And in those cases the close-out 12 report will identify the further work that is needed'and s 13 contemplated but that further work will then be carried out
(])O 14 as part of the Commission's regular regulatory development 15 activities and the Chernobyl collector label will at that j 16 time no longer be useful and will not be retained. {
17 The effort consists of twenty-six discrete tasks .
a l 18 of which four are new items and I will identify them later. j i
j 19 And the other twenty-two are quite limited modifications of j 1
1 20 ongoing programs where the purpose of the Chernobyl follow- l l 21 up activity is to ensure that the Chernobyl implications 2
l 22 and the Chernobyl lessons are adequately recognized in'those
! ()
j 23 relevant studies.
24 The total work is, of course, much more limited 4
() 25 than the TMI follow up. Only two NRC staff years are Heritoge Reporting Corporation
, ne an== ,
,e-c.,- ,o- - - , . ~ . . , ,v, x- , .-,. -n-..,e - e .---.-..-,-,e ,,.~m,n--,- --,m --
,,-,r--- .,,7, w p.,, ,,,,m w- , - , , ,ye ,.-n-cw rv w-,-p,n, >nvymr
1 9
1 to be required. We have fairly extensive contract support
('s v
2 of over half a million dollars.
3 The breakdown of the tasks according to the dif-(^T
(-) 4 ferent issue areas are indicated on this slide. As you will 5 note, the operational controls and human factors area has 6 the greatest number of tasks followed by design. And the 7 new tasks are all in the operational controls area. The ex-8 ception of the Brookhaven study on reactivity transients is 9 part of the design area.
10 Almost all the recommended work as almost all of 11 the implications assessment itself has addressed light water 12 reactors and one graphite reacting operation and the one that I
~l 13 is in the design development stage where the subject of one (O n l l4 final chapter in the report and two qualified recommendations 15 for follow-up work in that area are included.
16 I will now briefly just list the different tasks 17 broken down according to different areas and maybe say a 18 few words about some of them and then Tom Ryan and Len Soffer 19 will cover in more detail three of the particular isssuc
?0 areas and then in still more detail you will have the Brook-21 haven Report on reactivity transients.
s 22 In connection with--I think you are familiar with
\_)
23 the symptom based emergency operating procedures development ,
24 in NRR. There the Chernobyl --
persists in assuring that
() 25 the Chernobyl lessons have been taken into account by talking Heritage Reporting Corporation
<an ..
l _ _ __ . _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ _ - _ _
10
(
l about violations, tests ann 9 ! ay views. Criterion guide-j) 2 lines are being developed a- o e 1:eerative venture with the 3 industry. The proposed criterion guidelines of industry are 4 in the process of a review by thel agency and at'some point 5 where the agency with some changes as the staff considers 6 necessary'is satisfied, it will be issued where the Chernobyl 7 considerations will be reflected.
8 In cont.ection with the NRC testing requirements, t
9 this is one of the new projects that Tom Ryan will talk about 10 more. The issue is a study of the benefits of various tests t
11 and the risks that the conduct of a test itself may generate . j l
12 A revision of Regulatory Guide 1.47 concerning l
13 safety system bypass has been in progress for some time and
(})O 14 that, of course, has some Chernobyl. lessons to take into 15 account. As does the study of the adequacy of requirements
! 16 witn respect to availability of engineered safety features i
l 17 particularly that availability under conditions other than 18 full power operations to make sure that we have not incor-19 rectly assumed full power operations to be controlling in 20 those instances when something might be needed at other con-21 ditions that might not be as obvious from a full power opera-22 tion focused study.
)
23 A review of the bases statement, technical specifi-24 cations, has for its purpose to assure that the technical 1
25 specifications are understood correctly in relation to the Heritage Reporting Corporation (set) 6as-4ees
. - - . .. . . _ _ _ . _ _ - . , . _ _ , . ~ _ . - , , _ ..m. ..~._.,_..-_-_,__.-.._.m.._.._.-~._,. -, _
11 I safety analyses that underlie degeneration so that any ex-() 2 ceptions to them or interpretations of them not go amiss, 3 lack of full understanding of what they are supposed to do.
p.
-- 4 MR. LEE: Can I ask a question?
5 MR. SEGE: Yes, sir.
6 MR. LEE: Do you expect through this analysis and 7 investigation the volume involved in the technical specifi-8 cations could possibly go down on all this?
9 MR. SEGE: I don't think that the tech is neces-10 sarily directly connected to this issue, but we have Sam 11 Bryan here from NRR and perhaps he would care to comment on 12 this question.
13 MR. BRYAN: I'm Sam Bryan from NRR. We do have a (v ')('S
'~
14 Tech Spec Improvement Program underway right now and its 15 purposes is to make the tech specs more operator oriented.
16 And improve the human factors aspects of tech specs so they 17 will be readable and usable. More readable and usable.
18 We expect that by applying criterion that has been 19 generated for the purpose of saying one technical specifica-20 tion should remain technical specifications will indeed when 21 they are applied split the tech specs into those that will
,s 22 be tech specs and those that will be controlled by other
( )
%)
23 things. By the licensee's controlled --
24 Several specifications will be removed from the
/~Ni
(,/ 25 toch specs and into other licensee controlled documents, so Heritage Reporting Corporation (202) 426 4044
._ _. _ ..- . _ . ~ . _. . . . _ . . ~
12 I .the answer is, it isn't short, but it is that they will pro--
O 2 beb1v de reauced -- 1 l
3 MR. SEGE: Well, I understood that these days tech 4 specs perhaps make.up at,least two binder volumes in typical 5 PSARs.
6 MR. BRYAN: That's true. I don't know about two 7 binders of FSARs. I can tell you that.some are that thick.
8 But it does include.usually the bases which are really not 9 a part of the specifications itself. They are for explana-10 tory material, 11 The tech specs are being removed. They will be 12 relocated. They will not be technical specifications. They i i !
'Q 13 will be under the control of licensees and some other li- '
14 censee document. They will'still have to be done, but they l
15 can be changed through the licensee's procedures rather than f 16 having to come into the NRC to be changed.
17 MR. LEE: Thank you.
q 18 MR. SEGE: Thank you, sir.
I 19 Risk analyses are focusing on low power and shut-3 20 down operations is another one of the four new study items 21 that are not quite started yet but it's expected to soon.
22 MR. WARD: One question about that, George. See, 23 last week at another subcommittee meeting we heard from Wayne 1
24 Hodges of NRR who was talking about the problems and possible I
i O 25 ee1utione to these with reserd to the decer heet remove 1 to l Heritage Reporting Corporation l
] <no m.m I
i .-
. - - - - - ~ . . , , _ _-.__ -.-..--.---_, --, - , , - . _ , . . , - - - - - . ---_--.--.m , . , - - . , .
13 I get low power or shutdown. Are we talking--are those programs
[]) 2 connected? I mean what he was talking about and what you are 3 talking about here. Is there some--
() 4 MR. SEGE: The decay heat removal is not considered 5 part of the Chernobyl program and the A45 generic issue is 6 not part of the program, but I am not specifically acquainted 7 with the work that you are referring to.
8 MR. WARD: Well, this is really under Generic 9 Issue 99 I guess rather than A45.
10 MR. SEGE: I see.
11 MR. MINNERS: Maybe you ought to tell him what it 12 is for, George. What does 14C cover?
13 MR. SEGE: It's a risk analysis for low power and
({}O 14 shutdown operating modes. The project manager for this pro-15 ject was expecting to be here and I~ don't see him now, but 16 if he comes in later perhaps you could return to it and ask 17 him to comment on it.
18 MR. WARD: We don't have any--there isn't anything 19 on the agenda to talk more about this.
20 MR. SEGE: No. That was not intended to be in-21 cluded today.
22 DR. KERR: Would you say it had to do with acci-0 23 dents that may occur when the reactor--or might initiate d
24 when the reactors shut down? You said it has to do with--
() 25 MR. SEGE: It does not include decay heat removal l Heritoge Reporting Corporation m ean
I 14-
> I I issue.
.i 2 DR. KERR: Well, does it address itself to accidents ,
3 that may initiate because in reading the summary that's the l 4 impression I got. That you --
to address accidents that :
s initiate or might initiate when reactors shut down. It's hard 6 for me to see how that could exclude decay heat removal.
7 MR. SEGE: That's a separate issue for that. But 8 I think Warren Minners.would like to comment on this.
9 MR. MINNERS: I think it's a matter of terminology.
10 Warren Minners of the Research-office. When George says 11 decay heat removal, he's talking about A45 which-I'think j 12 the Committee will remember is basically directed towards qQ 13 the dedicated heat removal system that removes decay heat i 14 after a plant has shut down.
1 is DR. KERR: I think we understand that. ,
j 16 MR. MINNERS: That's what George means by decay ;
17 heat removal.
18 The other Issue 99 that we reviewet, I guess I'd 19 have to talk it over with George, but I don't see why that 20 wouldn't be part of the Action Sequence Evaluation Program ;
! i i
21 and I'd have to go double check but I believe that that kind i >
j 22 of accident would be considered --
- O
- 23 DR. KERR: One of the reasons we're asking is be-l l 24 cause I thought he pretty much had a solution to that problem.
O 2s and if thet vroetem hes etreedy been so1ved-- !
Heritage Reporting Corporation i
_ _ _ _ _ . - _ _ - - _ - ._ _ ~ - - - - . - - - - - !
15 1 MR. MINNERS: We never go over the same ground
() 2' -again, Dr. Kerr. You know that. We're very careful.
3 DR. KERR: I'm trying to be helpful because I O
(-) 4 know you are all busy and concentrate on your day-to-day 5 j responsibilities,and we have the luxury of sitting back 6 thoughtfully to examine'these' things.
7 MR. cINNERS: f I.think this study is to look for 8 accidents oth'er'than the one'which we've already identified.
9 We identified 99. Thats not'the problem. It's to look at 10 low-Power and shutdotn operations to see if there is something .
11 else we might list.
12 DR. KERR: Mr. Hodges maybe should wait before he 13 initiates new tech spec changes until'that's been examined
/~}/~T k
\#
14 more thoroughly because that's one of the things he proposed 15 to do, to add something to tech specs.
16 MR. SEGE: I'm going to make sure that if communi-17 cation doesn't alread-j exist between the two projects, that 18 it is established. The project manager is not here so I g9 couldn't answer for the fact of the' question about communica-20 tion between the two.
21 Assessment of NRC requirements on management and l
22 the accident management area will ba covered in some detail 23 by Tom Ryan shortly.
24 MR. LIPINSKI: Before you take that off, I have a
() 25 couple of commencs. Back up to the testing requirements and l Heritage Reporting Corporation (tet) us asse
16 i
! Regulatory Guide 147, there have been human errors that have )
O 2 been c mmitted, the most recent one is the one at La Salle 2-3 where the wrong valve was operated, en b 4 MR. WARD: What did you say? The wrong what?
5 MR. LIPINSKI: Wrong valve was operated. And 6 that single valve had:effect on many instruments, not just 7 a single one. So it led to confusion in the control room.
~
8 But the other area that wasn't ment 3.oned is maintenance, where 9 errors are being committed in maintenance and what effect those 10 have on safety systems. Because testing is'one aspect but 11 then maintenance is another aspect.
12 MR. SEGE: What I would like to do if it is agree-13 able to you is defer discussion of this question'until Tom O 14 Ryan's turn comes because he will cover the general areas 15 in more detail and he'll have a presentation on your ques-16 tions and will have follow up as necessary.
17 MR. MINNERS: Mr. Lipinski--
1 18 MR. LIPINSKI: Yes.
19 MR. MINNERS: This is a very limited effor't.
20 These are studies which are rather directly related to 21 Chernobyl, okay, and I don't think we discover at any mainte-22 nance related thing lessons to be learned from Chernobyl.
23 Maintenance is an important thing, but it's not something 24 learned from Chernobyl, and this is limited to lessons O 25 1eerned from Chereebv1.
Heritage Reporting Corporation l imun.
)
17 l
/
I g MR. MINNERS: I'd like to address Dr. Kerr's cor.tment
, 2 on Ohrther we should maybe hold up on Generic Issue 99 reso-3 lution until we complete the study here and the reason I 4 want to address that is because I - t hink that's something we 5 should not do. Generic Issue 99 is addressing a very speci-6 fic situation which has occurred frequently. It's a mid-level 7 operation where we've had frequent encounters with lack of 8 decay heat removal for short periods which we believe is a 9 very significant risk. 'We need to resolve that issue and 10 9et something out on the street as fast as we can. I think 11 we've already delayed it long enough. I wouldn't want.to 12 delay it any further pending the outcome of further reviews Q 13 of other potential low-power accidents.
O 14 DR. KERR: I was not suggesting that it should be.
13 I don't know whether it should be or not. My impression--
16 my guess would be that that is perhaps about the only acci-17 dent sequence that you are going to encounter and it may be 18 that the problem is already solved. I don't know.
1 19 MR. SOFFER : That's possible, sir.
20 MR. LEE: Could I ask a question on the last item 21 on the last transparency. It relates to accident management.
22 I was wondering how you are going to coordinate O 23 this effort with the individual plant examinations that all 24 utilities are being asked to--or will be asked to go through
- O 25 ema enee ee vert it vse we 11 eetx enoue ecciaeme memeseme t Heritoge Reporting Corporation m wee.
18 !
I there.
2 MR. SOFFER : In general, of course.
3 MR. LEE: Of course.
4 MR SOFFER : First of all, let me say I think we're 5 intending to come and talk to the subcommittee, I'm not sure 6 which one, very shortly on accident management. But the IPEs 7 are related to accident management in the following way. We 8 would expect that after the plants do their IPEs that there 9 would be some obvious things related to accident management to that should be pulled in"to any potential accident management 11 plans that utilities would then be expected to devise. 'How-12 ever, our accident management program is not waiting for the f) 13 IPEs to be completed although there may have been an implica-0)
\'
14 tion like that in the Integration Plan where it talked about 15 completion of accident management, completion of IPE. But we 16 are already beginning an accident management plan. We have 17 a program--we have two documents, both of which are being 18 reviewed at this time and ;- expect to review them shortly, 19 within a couple of months I think. I think somebody has been 20 in contact with your staff to talk about that.
21 But one is an agency plan that has to do with the 22 regulatory basis and the other is a research plan which has 23 to do with research which was pulled into the regulatory 24 program. But even though those plans are not fully approved O 25 et thie voime, we et111 de11 eve it gruaent to deein work in Heritage Reporting Corporation
<m m
- - . - - . - .~
19 1 certain research areas that were important to, for instance,
(} 2 establishing guidelines and that sort of thing. But it's not 3 going to wait for the IPEs to be completed although the
(-)'
's 4 schedule calls for a final development of plans and review 5 of those accident management plans after the IPE because 1
6 obviously some information from the IPEs would be very useful '
7 in developing accident management plans.
8 MR. LEE:. Is NUREG 1150 going to play a role in 9 this research plan?
10 MR. SOFFER : Absolutely. In fact it's very much 11 going to be an early starting point for examining--because -
l ll 12 certainly there was a lot of accident management information
(]) 13 that came up in the 1150 review and continues to come up and l 14 we are going to use that to the extent--you know, to the 15 maximum extent possible to start our evaluation.
16 MR. LEE: Thank you.
17 MR. SEGE: Tom Ryan, did.you want to add anything 18 at this point?
19 MR. 1YAN: Tom Ryan, from RES and the next speaker.
20
. you please raise that question again when I 21 get up there and start talking about the relevance of Cherno-22 byl or the relevance of this issue to Chernobyl. Because the 23 answer to your question is, yes, we do plan on taking a look 24 at IPEs amongst a number of other things as a basis for
() 25 addressing the Chernobyl issue which in fact says due to these Heritage Reporting Corporation o.n .
, , - ~ . . . , . . . - - . - - - - - - , . - . - . .
20 I accident management plans take into account all the lessons
() 2 we've learned from Chernobyl.
3 DR. KERR: At the risk of being repetitious, I
/~T l k/ 4 would like to comment on the phrase "things learned from 5 Chernobyl." And from Mr. Minners' comment I could get the l 6 impression that unless something specifically happened at 7 Chernobyl, you can't learn anything from Cherno'byl. I'm 8 glad I misinterpreted what he said. Because it seems to me 9 what we need to be looking for as.well as things that actually 10 happened at Chernobyl is thing's that Chernobyl might suggest ;
l 11 to us that-didn't happen there. I think it's unlikely there 12 will be<another Chernobyl, but there may be some' things that
(]) 13 Chernobyl suggests to us that would help us avoid other 14 kinds of accidents.
15 It seems to me that may be as important as the.
16 specifics of what happened at,Chernobyl.
17 MR. SEGE: That is exactly the way that we reviewed 18 the issue of Chernobyl implications in the Implications 19 Assessment and that's how the Chernobyl Follow-up Program 20 is structured.
21 DR. KERR: That wasn't altogether clear to me when 22 I read the report, but I would hope that's the case.
23 MR. SEGE: I did use the term "direct indications,"
24 but that was not intended to be a literal specific feature-() 25 by-feature comparison.
Heritage Reporting Corporation om m me
l 21 1 In the design area the reactivity transients work
() 2 you will hear about shortly. And there are four tasks that 3 -relate to --
unit sites where the Chernobyl implications rx kl 4 are being taken into account in ongoing programs. They have 5 to do with control rooms and disability and contamination 6 outside of control rooms as a result of an accident in another 7 unit on the same site. And also smoke control as a result 8 of an accident in another unit on the same site.
9 As far as shutdown is concerned in future plants 10 we intend to consider complete avoidance of shared shutdown 11 systems except to the extent of-shutting down one unit as a 12 result of a problem in another would not be impaired but-13 would in actuality be helped. And this-is a rather complex
(])Q 14 way of stating it but we will endeavor to study it in a way 15 that would avoid mutual' dependence but not the benefits of 16 redundancy where an objectionable mutual dependance can be-17 avoided.
18 Then last but not least risk assessment is fire 19 fighting with aviation present which is another study that 20 is ef fective in this area.
21 MR. WARD: Let's take, for example, 2.3B, where you 22 were talking about shutting down the systems. You are talking 23 about future plants. What's the product of this cast going 24 to be?
) 25 MR. SEGE: Presumably a set of recommendations Heritage Reporting Corporation
<mun
22 I
that would then be taken into account in standards for re-
{} 2 viewing guides for future plants by the standard designs of 3 specific applications in.the future.
0)
'- 4 MR. WARD: So there'll be a change in the general 5 design criteria or the standard review plan or what?
6 MR. SEGE: Change or interpretation of a general 7 design criteria and standard review plans. I think Warren 8 Minners wants to comment on that.
9 MR. MINNERS: The plan says the work description is 10 "prepare guidance of --
system as part of severe acci-11 dent policy."
12 MR. SEGE: Two tasks in the containment area, con-l').
13 tainment performance,during a severe accident that go beyond x
14 the basis on ,'which the containment .for design and our consi-15 deration -- venting particularly international information 16 as changes, the countries were --
venting is being imple-17 mented for consideration and Len Soffer will comment on~these 18 further.
19 In the emergency planning area that will also be 20 elaborated on further by Len Soffer. As you now, the agency's 21 statutory responsibility in this area are adjacent to those fx 22 of the Federal Emergency Management Agency and in conjunction 23 with FEMA and other agencies are involved, we plan to follow 24 development in the Soviet Union and elsewhere in connection
() 25 with the ingestion pathway of decontamination experience in th e Heritage Reporting Corporation m sm
23 I regions around the plant and the relocation experience for
{} 2 possible lessons for our emergency planning in these areas.
3 In connection with severe accident phenomena, the
(~
'- 4 Chernobyl experience has iridicated there is some stripping. of 5 the surface of the fuel, possibly through oxidation of the )
6 UO2 and that stripping either chemically or thermally from 7 the surface of the fuel was a dispersion mechanism that may 8 not have been taken into account in our source --
studies 9
or at least not to the extent that the Chernobyl experience 10 suggests so that is being studied now. And also the effective--
11 ness of mechanical dispersal as distinguished from the melting ,
12 MR. LEE: The oxidation issue--I wonder why it has 13 not been picked up in our source term study or severe acci-(])o 14 dent research plan so far. I get the impression that this 15 particular aspect E has not been really studied so far. Am I 16 correct?
17 MR. SEGE: I am not sure whether it has been stu-18 died at all but it was not really included in the models that 19 we conventionally used. And I'm not familiar with the prior 20 work on it and I don't know to what extent any study has 21 been devoted to it.
~ 22 Len Soffer, do you know--can you comment further on 23 this question?
24 MR. SOFFER: I'm sorry. I cannot. Ralph Myers is
() 25 the individual that would be able to comment. I'm not Heritage Reporting Corporation
<mnwa l
_ - _ _ _ _ - - _ _ _ _ _ _ _ _ - _ _ _ _ =
24 i
i familiar with it.
O 2 sa. sece: e 1e noseis1e thet Me. Chee 3
has been following this work will be here this morning and O 4 if she comes perhaps she could take up this question, but I 5 don't know the answer.
6 Steam explosion work is not going to be done. It's 7
a study of fuel coolant interactions in severe activity 8
initiated accidents. That work is not intended to be done s'
unless the work from Brookhaven that will be reviewed with 10 you in connection with the reactivity transients suggests 11 that it should so it is in the program only on a contingent 12 basis.
13 MR. LEE: May I try to understand that issue.
14 There is a distinction made apparently between what happened 15 in Chernobyl and what is being considered more extensively 16 in severe accident contact which may' involve a large amount
~
17 of molten fuel that could get into, contact with cold water 18 or something like that. This steam explosion phenomena that 19 you are talking about is something else? Am I correct?
20 MR. SEGE: This would be steam explosions that 21 could result from reactivity initiated accidents if it 22 found that there is sufficient likelihood of the proper con-23 ditions for it arising.
24 MR. LEE: But in the NUREG 1251 report I thought I 25 read some kind of distinction being-made and I'd like to Heritage Reporting Corporation
< => un
25 1 ' understand why the distinction is-made or indeed it is mean-(} 2 ingful to make the distinction. The same terminology "steam 3 explosion" is used, but apparently we're talking two different 4 kinds of steam explosions.
5 MR. SEGE: Yes.
6 MR. LEE: Am I correct in reading the report?
7 MR. SEGE: Yes, that is correct. But I'm not ex-8 pert enough on the subject to clarify the distir.ction beyond 9 what the report says. But we do have clarified language for 10 that particular section.
11 DR. KERR: Who was the expert who made the decision 12 that this would be looked at--the Brookhaven study indicated m
13 that it should be. I mean somebody must have been expert
) 14 enough to make that decision.
15 MR. SEGE: Yes. . The people involved in that were 16 Themis Speis and Theo Theophonos and Carter Salan, Bob Wright, 17 and there may have been others. ThereJhave been others in-18 volved. But incidentally, that section of the report that 19 makes a distinction had some obscurity in it and it is being 20 rewritten, the version that you had is being clarified now 21 and in the final version I hope it is going to be--and I can g 22 point it out to the ACRS staff where the change has been made.
%)
23 MR. LEE: Do I understand correctly in the context 24 that you brought the steam explosion issue up, in the forth-() 25 coming Brookhaven study there is no relation to high heat Heritage Reporting Corporation j (391) 6 Mat 04 l
l
.- _ -- . = -
26 1 content in the fuel itself which may be subject to Super Fund I' 2 critical trends of some kind, which may then get into contact V) 3 with water of some kind. Is this the context that we are O 4 looking at?
5 MR. SEGE: That is approximately. correct.
6 MR. WARD: I gather it. deals more with the poten-7 tial for steam explosion in an intact reactor.
8 MR. SEGE: Right, right.
9 MR. WARD: As opposed to the other studies you are 10 looking at. A molten --
11 MR. LEE: Right, right.
12 MR. WARD: Well, do you think that distinction 13 makes sense? Or do you think it doesn't make sense?
()O 14 MR. LEE: I'm not sure I understand the physics 15 well enough at this stage to be able to say, yes, they have 16 to be differentiated. I thought when you go down to the 17 bottom level, we are talking about the same phenomena. I 18 may be wrong.
19 MR. SEGE: In the graphite reactor area two issues 20 were identified in the possibility of a FRA and that is under 21 discussion with the licensoc now under the IPE program, and 22 then the issue of whether there is enough risk of cracking of 23 the structural graphite for the core to be able to fall away 24 from control rods and consideration is being given to pre-() 25 sumption of structural graphite experiments in this context.
Heritoge Reporting Corporation
<=nn.
27 i That concludes my presentation and I will be
(} 2 followed by Tom Ryan for the human factors parts of the 3 Chernobyl Follow-up Research Program and then by Len Soffer.
() 4 These are general reviews or plans and status and we'll wel-5 come any comments that you would care to offer that.might 6 help us in the conduct of these various activities. And 7 finally when we come to the reactivity transients work, the 8 detail review with you was scheduled at this time on the 9 basis that we wanted to--the subcommittee wanted the meeting to and we wanted the meeting at a time when enough work has been 11 done and there is something meaningful to report, but it is 12 still close enough to--still early enough in the program that
() )
13 there will be ample opportunity to take the subcommittee's
14 comments into account in doing a better job. That is the l 15 purpose of our being here this morning. I 16 Thank you very much.
17 MR. WARD: George, just a minute. Let me go back 18 to one of the topics you've covered. There are a couple of i 19 tasks associated with emergency planning decontamination,' and 20 it might be the pathway protective measures.
21 Apparently after the Chernobyl accident the Soviets 22 have effectively used some sort of polymer substance to D/s ;
23 cover contaminated areas, like land areas and so forth.
24 Is that considered to be an important control mea-() 25 sure? It isn't claar to me that the task is looking at that Heritage Reporting Corporation m.
., . - - . _ .~ , ,-
28 1 sort of thing.
~) 2 MR. SEGE: The task is intended to be broad enough
(~/
3 to encompass anything that was actually.done to see what was O
'/
'- 4 done, what the problems were in doing it, how good the results 5 were, but,Len Soffer will cover this area somewhat more fully 6 but, Len, would you care to add anything to--
7 MR. WARD: If Len is going to cover that,-I'll just 8 wait.
9 MR. SEGE: Okay.
10 MR. RYAN: Good morning. My name is Tom Ryan. I'm It with the Human Factors and Reliability Branch within the Of-12 fice of Research. And my purpose here this morning is to 13 discuss with you five Chapter 1 Chernobyl follow-up issues
(]('t
'd 14 and they have to do'primarily with human factors.
15 Sam Bryan who spoke a little bit earlier is here 16 from NRR to answer any questions you may have with regard to 17 Chapter 1 issues that I will not Se speaking about and as-18 signed to NRR.
19 MR. WARD: Tom, just so I understand. Does this 20 mean that these paths are assigned primarily to your branch 21 or that you are talking about the piece of it that your 22 branch is working on?
23 MR. RYAN: As suggested here, of the five issues 24 I'm going to talk about, our branch has four of them and the
() 25 Electrical and Mechanical Engineering Branch within Research Heritage Reporting Corporation o.u -
--. .=
i 29 1 has the fifth one.
2 MR. WARD: Okay.
{
3 MR. RYAN: Our branch has four tasks, one having to 4 do with willful or intentional violations of procedures, 5 which is one of the new issues.that is mentioned this morning.
6 The second one has to do with-the frequencylof.
7 testing requirements. And the presumed involvement of human
'8 errors, which is also a new issue. Two issues having to do 9 with generally what we consider to be management and accident 10 management.
11 And, of course, the last one has to do with safety i
12 systems monitoring.
13 I'd like to make tnree general statements about what I
()[) ( 14 I'm going to talk about because I don't have a heck of a lot l
15 of time. Firnt of all, what I'd like to do is pretty much j l
16 state the Chernobyl issue itself, that which we are trying to l
17 address ourself to. What the research objectives are. What l 18 the work scope is. How we are going to go about doing it.
19 And finally the products and the-schedule that we have in 20 place for doing it.
21 The second thing'that'I would like to bring to your 1
gg 22 attention, in three of the.five cases I'm going to be talking G
23 about research objectives in a work scope that have two l l 24 phases, and I think it goes to a little bit to what Mr. Kerr
() 25 was alluding to.
Heritoge Reporting Corporation om.m.a _i i
_ . _ _ , . _ . ,, _ , . . , _ . . . . . . _ , - .m___, - . _ _ , . ~ __ - , _ , . , . - , _ . . _ ._
30 i We have found in our case that simply responding to h 2 the Chernobyl issue as it's stated really isn't the end of 3 the story. There's more work that needs to be done. And 4 secondly, there is at least one case where the Chernobyl 5 experience itself has stimulated some new thinking about'a 6 more general subject area.
7 So in three of the five of these I'm going to talk 8 about, you'll be seeing two phases. The~back phase that we 9 are going to respond to specifically to Chernobyl issues and to then follow-up work that we're going.to be doing as part of l 11 our research program to try to bring resolution to_the larger l
12 issue. !
13
-{u 14 Lastly, I'm not sure what material you have that you are looking at there in terms of dates. As some of you 15 know, we've gone through a lot of reorganizing in'the NRC 16 especially in the human factors area. And it has been rather j 17 recently that we have sort of come together as a critical 18 mass, so what we've tried to do is sit down with these issues 19 and come up for you for some realistic dates that we think 20 we can meet now that we seemed to have settled down and can i
21 go pretty much about our business.
22 DR. KERR: You don't think another reorganization 23 is scheduled for the next six months?
24 MR. RYAN: Hopefully not.
25 Okay. The first issue which is known as 1.lD Heritage Reporting Corporation
< =n:.-
31 1 Procedures Violations. Basically the Chernobyl issue itself
( 2 is can we come to grips with the frequency nature and impact 3 of potential or willful violations of procedures. And we b 4 are certainly not restricting ourself to any particular set 5 of procedures, and as a response partially to Mr. Lipinski's 6 question a little earlier, we also are interested in mainte-7 nance type procedures.
8 This particular issue has two phases to it. As I 9 was talking about earlier. Research objectives, of course, 10 are to determine the frequency and cause of impacts of the ,
11 intentional violations and to, if we do find that these are ,
j 12 safety significant, to develop a technical basis for doing n 13 something about it,
()b ,q 14 We also recognize here that there's a need to take 15 the same kind of look at what we might refer to as non-in-16 tentional violations of procedures. Human errors associated 17 in having procedures involvement either because there is 18 something wrong with the procedures or because the people 19 can't read them or they just missed steps in the procedures 20 or whatever. So our follow-on objective, making this part'of 21 our larger research program, is to deal into this issue after 22 the Chernobyl close out.
~
23 The work scope--
24 MR. WARD: Excuse me, Tom. The last one--looking 25 at the causes of frequency of unintentional errors. I mean Heritage Reporting Corporation mmm
32 l
I my impression is you've been doing that for twenty years.
() 2 What's new here?
3 MR. RYAN: Well, I'm not sure we have really been 4 doing it for twenty years in a systematic organized way, 5 taking into consideration everything that's available out 6 there. For example, INPO has a system called HPES.
7 MR. WARD: Right.
8 MR. RYAN: That's a reporting system where they 9 claim the majority of the reports they are getting have to 10 do with procedures as a causal factor. So what we would very 11 much like to do is to bring together and get our arms around 12 most of the information that's out there and use it as a
(]) 13 basis for doing this kind of analysis.
14 DR. WYLIE: There are several recently where two 15 identical safety features have been taken out of service and 16 the plants have been operating. At least from what I could 17 read in these things, it's just a matter of ignorance on the 18 part of the maintenance people that took it out. They didn't 19 even talk to the operators. They didn't follow procedures to 20 do it.
21 It suggests, you know, training is needed in those 22 cases. I think it's a good thing to take a look at.
23 DR. KERR: Just find it.
24 DR. WYLIE: I beg your pardon?
() 25 DR. KERR: Just find them.
Heritage Reporting Corporation
. m ma
_- , _ _ . . _ , _ _ . _ . _ _ _ _ . - , _ .- . ~. . - - - . - _ . . - - , . . - -
33 3
DR. WYLIE:'Well, they did find them, when they 2 found them.
3 DR. KERR: Did they fix them?
O '
DR. WYLIE:Well, they got their attention, I'm sure.
's sl 4 5 But it suggests that, you know, personnel change in these 6 plants and whether there is adequate training behind the 7 plants to assure that the maintenance people and the operators 8 stay abreast of what's needed in the plants is the question.
9 MR. WARD: Well, I don't question that it's an 10 important safety issue. I guess what I was questioning was gg how or why it's a new issue coming out of Chernobyl. You are 12 saying there really hasn't been a systematic look at particu-13 lar errors resulting from the procedures.
({}( ) g4 MR. MINNERS: I would ask the simple question is 15 what is the error rate of industry. I think that's a fairly 16 simple question.
17 MR. WARD: Did you get what Mr. Minners' said?
18 Would you repeat that,. sir?
39 MR. MINNERS: I said I've asked the simple question 20 of what was the actual error rate that you have observed in 21 the industry. There is no answer to that question.
22 MR. WARD: I guess I'm surprised the AEOD hasn't O 23 looked--doesn't have something to.say about the OERs.
24 MR. MINNERS: They have some reports in which they (f 25 looked at tech' spec violations and they can tell you tech Heritage Reporting Corporation m ma
34 1 spec violations. In some narrow areas there are some numbers
{) 2 and there is quite a bit'of research that has been done. And 3 I think the question could be answered but when you ask it of t]' 4 people, there isn't anybody who has looked at all this dif-5 ferent stuff and kind of put it together. You are not going 6 to come up with an answer, I understand that. But you should 7 be able to have some kind of indication.
8 DR. SIESS: Well, one problem is, it's a tough 9 question. And if you ask them how many. tech spec violations, 10 that's a clear cut. But when you say what's the error rate, I
11 what's an error? l i
12 MR. MINNERS: Well, I would expect a human factors 1
() 13 expert to come back and say--define an error, and say the 14 error rate for these kinds of errors is such and such.
15 DR. SIESS: If you want to define it, I'll bet you 16 they can answer it.
I 17 MR. MINNERS: Oh, I bet they can. We've got one i
18 here. Ask him.
19 DR. SIESS: If you let me define the question, I'll 20 give you an answer on anything.
21 MR.MINNERS: Well, that's right. But he might be
- 22 able to give you an answer. He could come back and say, "Hey, 23 for tech specs, here's the error rate." And I'd say, "Well, 24 how about for other errors?" And I think the general answer
() 25 would be, "Well, I don't know for other errors."
Heritage Reporting Corporoflon mam
I' 35 1 DR. SIESS: What about the errors you didn't find?
() 2 MR. MINNERS: That's right. That's another ques-3 tion which hasn't been looked at. And maybe there's no
{}-
\ 4 answer for that.
5 DR. SIESS: It's never going to be looked at.
6 MR. MINNERS: Well, there are other sources--
7 DR. SIESS: The only ones you know about are the 8 ones you detect. There are others that you got away with 9 for weeks or months that nobody knew about.
10 MR. MINNERS: But you can look at different places.
11 I mean we've just looked in the LER. You can go look at logs.
12 You can interview people and INPO has got a system, you know,
() 13 et cetera, et cetera. There may be other sources of data 14 that we haven't looked at. But people are going around and 15 saying, "Hey, human error is a big problem in nuclear plants."
16 I would think that if you were going to make that statement 17 you should be able to have some idea what the error rate is, 18 and then be able to say, "Hey, next year I made the error 19 rate go to this," or "I've fixed it,"' or "I've made it 20 worse."
21 MR. WARD: Okay. I believe you.
22 MR. RYAN: Okay. I'd like to turn to the work 23 scope. First of all, the bulk of this activity is going to 24 be done by an outside contractor. We've in the process of
() 25 letting the contract to Pacific Northwest Laboratories.
Heritage Reporting Corporation (302) 6 4 4004
36 1 During the--what we call the "Chernobyl phase,"
/^ 2 we're going to be doing three things. One, looking at-the
%)'
3 literature on intentional violations. Two, doing an analysis O 4 of the frequency using event reports from a variety of systems 5 we have available to us. And hopefully do some. case studies.
6 That is, take some selected events that we suspect to involve 7 violations of procedures and go out and start talking to peo-8 ple.
9 Now, you might'ask the question why are we doing 10 those three things'. Well, there are two or three reasons for 11 that. First of all, in order to try to respond to this 12 question, we're not sure all three of these are going to be 13 successful. We are hoping one or.two of these are. But tak-r~
U 14 ing the literature review itself, we're interested in moti-15 vation. What motivates people for better or for worse to 16 violate procedures? Because quite frankly there are two 17 sides to this. There is the individual who violates proce-18 dures te degrade the operability of the system for whatever 19 reason. And then there's the individual, and we here about 20 this somewhat frequently, who has a lot of experience in the 21 area. They are confronted with a situation. They know what 22 to do is right, but the procedures do not allow them to do 23 that. And now what happens?
24 So we would like to got a little handle on motiva-() 25 tion.
Heritage Reporting Corporation
< => m
37 I obviously, the second bullet involves what we would (J 2 really like to be able to do. Is that gleaned from a suffi-3 ciently large number or a representative sample of events n
kJ 4 that we know about to be able to label them intentional vio-5 1ations and use that as a basis for answering the Chernobyl 6 question about frequency and possible impact?
7 The third bullet is intended as sort of a confirma-8 tory activity. That is, after having done that, take some of 9 those events. Go out and hopefully be able to get added'in-10 formation on them from the plants at which they occurred, 11 which might help us confirm the degree to which we really are 12 talking about procedures violations. And also, to sort of
/~
"T 13 get us some feel about the motivation question that we've O 14 hopefully gotten some information on in the first bullet.
15 MR. WARD: How do you see this relating to the work 16 on cognitive models?
17 MR. RYAN: My guess is that it is conceivable that 18 we might find a couple of events that might be of interest to 19 us to run through the CES in simulation later. But standing 20 up here, off the top of my head, I'm not quite sure how I 21 would tie those two things together.
22 MR. LEE: May I ask, Mr. Chairman, what cognitive 23 model you are talking about? I'm not familiar with that at 24 all.
( 25 MR. RYAN: Under another research program, Human Heritage Reporting Corporation (set) 626 sees l
\ .
m 38 1 Reliability Researcil Program, we have developed what we call
() 2 the Cognitive Environment Simulation. It's an artificial 3 intelligence based. simulation for analyzing intention-forma-O
\/ 4 tion by operators in nuclear power plants. Not-execution.
5 But how they decide to take one action rather than another.
6 Now, that is on a SUN system at Westinghouse Research and 7 Development Center and tied into a training simulator. A 8 CES makes the decision. It's automatically executed through 9 the training simulator which then comes back and paints the 10 new picture for CES to work with. Because the sequences 11 evolve over a period of time. It's not just a one-step 12 process. The minute ~I do something with a system, things
(] 13 all start changing. So this is basically a simulation we 14 have developed to support' human ~ reliability analysis work and 15 also to answer other human factors related questions having 16 to do with everything from expert systems to just basic 17 questions about the decisionmaking behavior.
18 MR. LEE: It's primarily an expert system based 19 aporoach?
20 MR. RYAN: Well, it's artificial intelligence based 21 approach. It's built on the Eagle shell, AI shell, that was 22 developed by a fellow by the name of Harry Porpoll, who you 23 riay or may not be aware of. Who developed it to support 24 nedical diagnosis. There is a large simulation called
() 25 ' Caduceus" that's used in the medical profession to assist Heritoge Reporting Corporation
< =n
- , -._ _ - - . - . .~ - ._- . _ . ,
39
- 1. in making diagnoses.
O 2 We heve teken thet beeic Eee1e ehe11 end heve bu11e 3 something we call Cognitive Environment Simulation.
4 MR. LEE: Have you gotten anything out of this 5 effort to be able to shed some light on--
6 MR. RYAN: Well, it's going through a field vali-7 dation right now, and early in the next fiscal year we are 8 going to start taking events that have been recommended to-9 us by AEOD. The Davis Besse event, the MacGuire event and to a couple of others and we are going to start running them 11 through the simulation for the benefit of AEOD to attempt 12 to get some better insights into what might have been going Q 13 on as part of the decisionmaking process within those events.
14 MR. LEE: Thank you~. ,
15 MR. RYAN: But that's another problem.
16 Okay. The second part of that is hopefully to 17 use the same methodology to take a look at the unintentional 18 type of errors.
19 With regard to the schedule, we plan to be able 20 to report on the frequency causes and impacts, et cetera, 21 of intentional violations in June of next year. This is q 22 somcwhat of a change. I'm not sure what documents you are G
23 looking at. The date has been set back. And to prepare a 24 Chernobyl close-out report. That is, you know, what do we 25 do about it kind of thing. And then go into our larger Heritage Reporting Corporation (M2) 644000
l
)
40 i
f I research program on the' unintentional errors which is going 2 to be a year-long activity and we hope to be able to be at
{^}
3 a point where we have answers for people like Warren Minners
.O 4 .about human error rates and other things by September of 5 1990, 6
And part of this, by the way, the latter.part of 7 this is driven somewhat by the funding schedule and so on.
8 That's procedures violations.
9 Okay. The next issue, 1.2B, NRC Testing Require-10 ments. And, of course, the Chernobyl issue here is because 11 people make errors, is it possible that we're having these !
i 12 tests and surveillance activities occur too frequently, ;
i 13
(])' therefore raising error rates and really in effect reducing
(')
\- 14 the operabilility of the system. I guess the thought here 15 is like you've got your-new car. You drive it 12,000 miles.
16 You've got no problems. You take it in for a check up. They 17 start taking things apart. 'And it never runs the same again.
18 So are we having these things conducted too often?
19 Basically the researcher objective, of course, are 20 to identify the kinds of tests which may have a potential 21 for having those kinds of things occur. And the second part 22 of that is to develop a technical basis for making modifica-23 tions.
24 This is basically a combination of in-house activity
[]} 25 and contractor work, some work which we have ongoing up at Heritage Reporting Corporation i ann .
1
. _ _ , , _ _ _ . , , . - - - . . . . . . . . . . - , , _ , , ~ . - , --, , _. , _ _ - . _ , , , . _ . _ .- _ I
41 I Brookhaven. They are going to be involved in this.
2 And basically as you can see down here, it's going 3 to be a risk assessment type of analysis hopefully to shed C 4 some light on this particular--
5 DR. KERR: You may-be planning to do this, but it 6 seems to me it'also would be interesting to find out what 7
effect the required surveillance during the full-power opera-8 tion has on safety. There's a certain amount of surveillance 9
that is required. And certainly it's not news to anybody 10 that there are plant shutdowns, transient induced by mistakes 11 made during testing in full power. And that's probably part 12 of your study, isn't it? If it isn't, it seems to me it's Q 13 worth considering as a possible add on.
34 MR. RYAN: Does the statement up here that I have 15 include surveillance?
16 DR. KERR: Iflvou are going to look particularly, 17 it seems to me that surveillance during full-power operation--
18 and, yes.
19 MR. RYAN: That's basically the issue that we're-20 attempting to respond to.
21 MR. LIPINSKI:There's one other aspect of testing. It's 22 i the cost benefit issue.
C, In terms of increased automation.
23 And plant operators insist that humans can do the job as 24 opposed to increasing the automation, and taking on the added O 2' exvemee- o#e or the twia9 enee ~outa ne tatere tias trom Heritage Reporting Corporation -
mam i
42 I 1 this study is whether the human error factor is large and 3
() 2 would increased automation reduce--
3 MR. RYAN: I want to make sure I understand what ,
4 you are saying. You are talking about an allocation for d'.ing 5 operations type tasks or maintenance type task?
6 MR. LIPINSKI: No, routine tests. Just routine 7 tests. Particularly the valves in the plant. The state of 8 those valves are not indicated in the control room. So it's 9 only the operator manipulating the valve that knows what he.
10 has done. No one else ktows unless somebody comes along be-11 hind him and verifies this data. And it's an issue like that 12 as to whether that valve-- or someone else knows that the i
( 13 state of the valve has now~ changed.
14 MR. RYAN: One of the issues a little bit later on 15 that has to do,with the Safety System Monitoring--when I talk 16 about that maybe we can revisit this question and see if in 17 fact, at least in some' degree, we're addressing ourself to l l
- 18 the kind of issue that you are raising. l 19 MR. LIPINSKI
- Okay.
l 20 MR. WARD: Walt, are you talking also about the, for 21 example, the SCRAM. Automatic on-line SCRAM circuit testing 22 with--
(g_)
i 23 MR. LIPINSKI: Well, if you take the entire issue i 24 of testing safety systems, everythir.g that has to be done l
A(_/ 25 routinely--operators have to go into that plant and I'm going Heritage Reporting Corporation .
(202) 626 4000
- - - ,. -__, ,, ,, . ~ _ _ . , _ - . - , _ , _ _ . - _ _ . . . , _ , . . - - _ . - , , , _ . , , . . . .
43 I to take that Dresdon case where the operator turned the wrong The status of that valve itself is not indicated, but
.( } 2 valve.
3 the effects of having turned the valve did reflect in the 4 control room in many different places because it oparated more 5 than just one instrument. And it's that issue as to whether 6 all of those items that must be manipulated to conduct tests 7 on safety systems should be monitored. Or here I am a human, 8 change their -- and restart it. Because you-can get in and 9 turn a valve and if you don't restart it--it's not a civilian 10 reactor. It's the end reactor was the case where they go in, il turn a valve to do Fests on the containment pressure switches.
12 If '.he operator doesn't restart the valve, those switches
( 13 never work. And the valve position is not monitored, so you 14 have to rely on that operator doing his job. Reinforcing him !
1 15 with another guy that can verify that he's restarted the 16 system. And it's that kind of an' issue that pertains to the 17 safety system.
18 MR. WARD: One of the points'on the on-line full-19 power surveillance and testing. There's been quite a lot of 20 thought given--well, review of Japanese requiremente and l
21 Japanese practices in this area. And there's an NRC team l l
22 that visited Japan this spring which looked into that in some i
'O d 23 depth and that may or may not be a useful source of informa- l l
24 tion for you. I can give you the name of the. people later on
() 25 if you want me to.
Her!toge Reporting Corporation (let) 6344006
. _-..,---r___-.. _ _ _ _ , . . , . _ _ _ _ _ . _ . _ - . . .. . - . , - . _
44' I MR. RYAN: Okay. Carl Johnson, I don't know whether j* 2 you know Carl, has the basic responsibility for this. Would 3 you have an objection if I just had' earl get in touch with 4 you?
5 MR. WARD: It just might be worthwhile for him to 6 talk to these people.
7 MR. RYAN: Just to have you direct him.
8 MR. WARD: Yes.
9 MR. RYAN: Okay. The products and schedule, basi-10 cally we plan on preparing an impact analysis by March of 11 '89 and a close-out report hopefully will be completed by 12 June of 1989.
13 DR. WYLIE: Let me back up a minute.
14 MR. RYAN: Sure.
15 DR. WYLIE: Maybe I'm just quarreling with the se-16 mantics here. Performing routine tests. It's really per-17 forming tests, right? >
18 MR. RYAN: Right.
19 DR. WYLIE: I mean :outine suggests those things 20 you just don't do day to day, frequent, repetitive, but in i
21 the case of Chernobyl, that was not a routine test.
4 3 22 MR. RYA!1: Well, I'- not sure about--I think if you G
i 23 look real close at a lot of the literature we have,,. '11 24 find that significant precursors to a lot of probleia , safety
() 25 related or not, have to deal with people who are not doing
- l. Heritage Reporting Corporation 1
(302) M
.-.s _ _ . _ - , , _ _ _ _ . . . _ _ . , . _ - _ _ . . . _ , , , - _ . . . , . . . _ , - . . , , _ . , - _ . . . , , , _ _ _ - . , _ . . . _ . - . _ , . . . . _ _ . _ , , _ . , _
45 I the things that they regularly do.
[} 2 DR. WYLIE: My question is you are not excluding 3 other than routine tests.
4 MR. RYAN: No.
5 DR. WYLIE: Okay. I mean this was suggested just 6 through looking--
7 MR. WARD: Well, Charlie's point is that in fact 8 special tests might be less subject to controls than routine 9 tests are.
10 DR. WYLIE: Yes.
11 MR. WARD: Potentially more dangerous.
12 MR. RYAN: Okay. The third issue that I'd like to 13 cover here is 1.6A, which has to do with the kinds of NRC
- 14 requiremer.cs that are put on managers at plants and basically 15 are they or are they not deficient in the respect of per-16 mitting the managers to do their oversight and respond to 17 abnormal type situations. Too few, too many, wrong mix.
18 Whatever.
19 This is strictly an in-house activity and I think 20 I can respond a little bit to Mr. Lee's question earlier 21 about IPEs. This is a case where we are going to be drawing 22 on and from other programs in'the NRC that have anything to 23 say about management. The IPE program, the Field Evaluation 24 Team work, Performance Indicator Program, and also an organi-() 25 zation in management segment of our overall Research Program.
Heritoge Reporting Corporation i
c=nu
46 I So this is basically going to draw from all of 2
those programs, not only to deal with this issue but.also to 3
deal with the next one I'm going to talk about which you O 4 brought up earlier wat the Accident Management Plans, so 5 basically we have two phases here and for a little bit dif-6 ferent reason.
7 First of all, we've only been asked, as far as the 8
Chernobyl issue is concerned, is there a problem? .And we 9 certainly don't think that's the end of that. That may be 10 the end of the Chernobyl question, but it certainly isn't 11 the end of the issue for us. So basically we have a second 12 phase that says if row we have put some bounds on the problem
(]) 13 if it does exist, how can we deal with the problem? And O 14 that's what the follow-on work is all about which will be 15 part of our larger organization management program.
16 Down here, basically it gets into the Chernobyl 17 phase and refers to some of the other programs that we have 18 ongoing to try to determine if, in fact, there is a problem l 19 and to characterize that problem if one exists.
20 And the implementation phase, of course, as it 21 states up there, is to start taking a look at alternatives l
22 for dealing with those problems. And that gets into better 23 selection procedures for managers. What kind of qualifica-24 tions? Training, oversight, and so on. A whole host, I
() 25 guess, of possibilities.
Heritoge Reporting Corporation
< =i m.
J 47 1 DR. KERR: When you talk about requirements imposed 2 on managers, are you going to restrict yourself to formal 3 written requirements?
4 MR. RYAN: Not if we can determine what the other 5 ones are. There was no intent here to--
6 DR. KERR: Because there are certainly are others 7 that occur in the :.nteraction that takes place, for example, 8 between inspectors and plant management.
9 MR. RYAN: I'm glad you raised that because I was 10 going to respond to something you talked about earlier and .
11 that was that we should not be just restricting ourselves to 12 the letter and intent of Chernobyl. But we should also be ,
Q 13 concerning ourselves with what does this imply to us? And 14 the Chernobyl event itself and some of the thinking we have 15 dono up to now on this whole question of management has been 16 so focused pretty much on between the manager and somebody 17 who does something. When in fact there probably is a large 18 part of the issue that has to do with between manager be-19 havior, from the first-level supervisor up to the plant 20 manager. And we're starting to find that this is a big issue 21 in some of the other work that we are doing on organization 22 and management.
23 So in here a little bit later on during the imple-24 mentation phase, it talks about--that's the follow-on stuff--
l 25 is looking at some of these issues that have been applied to Heritage Reporting Corporation (tet) M4006 i
, , _ _ , - - , , , , - . . . . . - - + ~* - * ' ' ' * ' ' *~ ' ~ ~ ' ' ' " ~ ' ' * '
48
't I us by the Chernobyl experience itself. While they aren't l l
~
' /'s 2 part of the issue itself', just going through this process it !
U 3 suggests to us, hey, we have an area here that we had some 4 sense about before. It has to do with the interaction 5 amongst managers at various levels that may have as great an 6 impact on what goes on than just the relationship between 7 the manager and supervisor and somebody who actually exe-8 cutes it. So this_is a case that kind of fits in to what you 9 were asking earlier.
10 Basically we do plan on doing the problems analysis, 11 Wrapping that up by December of88, this year. And we do 12 plan on being to the point where we have some recomendations 13 for remedial action by June of '89. So that's a possibility
(])On A/ 4 for actually if we're not going to wrap up the report itself 15 until the middle of the calendar year '89, certainly given the implementation could conceiveably be part of it. Even 16 l
17 though it goes beyond the very specific intent of the thrust l 18 of the Chernobyl issue.
l l 19 MR. WARD: Again, referring to the Japanese prac-20 tices here, the Japanse, as I understand, have a system where l
at each plant there is a designated chief nuclear engineer, 21 22 and a chief electrical engineer, and a chief --
engineer 23 or something. And these are not separate jobs, but among the 1
24 engineering managers at the plant, one is designated in each
() 25 of these roles, are certified by the government, one or another Heritage Reporting Corporation
<=3 m.
I
.- . . - - . , . . - - _ . _ . - ~ - . - . .. - ,.
1
~49 I agency of the government. And in order to do that they have
() 2 to pass some sort of requirements including testing require-3 mente. Now, we don't have anything like that in our set of O 4 requirements. The agency insists on STAS who tend to be kind 5 of rookie engineers on night shift. The utilities tend-to--
6 like to_have some of their managers go through SRO training 7 which may or may not be particularly effective. Are you 8 looking at that sort of thing?
9 MR. RYAN: Requirements for managers?. l 10 MR. WARD: Yes, 11 MR. RYAN: Not per se, but--well, let me respond in 12 two ways. With regard to the foreign management practices,
(] 13 Mr. Banjord, the Director of the Office of Research, has asked 14 us to consider a study of management abroad and we are putting 1 1
I 15 together a package for him right now to build on the organiza- '
16 tion of management work we are doing currently, which should 17 help us get into some of this kind of thing.
]
18 A second point I would make. I firmly believe that j l
19 if the Commission persists and wants to have reactor opera-20 tors be degreed, so that they have upward mobility, I would 21 suspect that that might open the door to a number of things. l l
22 One of which would be a possible series of levels of training l 23 like the military has. You go to college. You focus on the 24 technical part of the job. You come in and you get to be
() 25 ready so you are a senior reactor operator and there's thought Heritage Reporting Corporation mm
- . ~ . - , , - . - - . , , - - . , . , ,- _. - - .- - --, . , , . , .
50 1 to making shift supervisor. You go somewhere and you go to
() 2 staff school. And you start, learning about managing in addi-3 tion to some more technical. You get to that level and then 4 start talking about a department head. You go to a' command 5 and general staff college where the majority of the training 6 has to do with management functions. And you are sort of 7 certified then to be allowed to be a department head. And 8 you go to the war college to move from the department head 9 to the plant manager or something like that. So one, we are 10 going to start taking a look at some of these foreign prac-11 tices. Hopefully within the context of the organization of 12 management program and certainly that might lead us in the 13 director that you are talking about.
14 But more immediately, I have the sense that this 15 whole issue of degree requirements for the purpose of upward 16 mobility and professionalism I think opens the door to a lot 17 of possibilities, one of which I've just sort of suggested to 18 you.
19 MR. WARD: Thank you.
20 MR. LEE: A question. In the earlier version of 21 NUREG 1251 there was some consideration given to having some-22 body with a title of--I've forgotten what it was called-- but 23 safety engineer or safety manager or whatever, could have an 24 oversight responsibility for a lot of plant operations and
() 25 safety-related kinds of issues._ But that particular Heritage Reporting Corporation
<mm
51 1
consideration was deleted basically from the final draft.
2 Am I correct?
3 DR. KERR: Mr. Sege, you mentioned that.
f~s 4 MR. SEGE: (Inaudible) 5 MR. LEE: And I'd like to learn a little bit of 6 the pros and cons behind that from somebody like that.
7 MR. RYAN: I think we have maybe two questions.
8 I want to go back to simply responding to the Chernobyl 9 issue. I'm not terribly sure that's completely relevant, 10 but as far as the follow-on work, that we're calling imple-Il mentation and research that goes on beyond it, I don't think 12 we're necessarily limiting ourselves at all.
(]) 13
~By the way, one of the things we're finding out in 14 some of the other work we're doing, we developed a model of 15 the human organization in nuclear power plant and we're find-16 ing that some of the line functions, that is, from BP nuclear 17 down, turn out not to be as important as some of these de-18 partmental functions because of the way the plant is set up 19 and because we regulate what goes on very, very closely. So 20 it may very well turn out that somebody like you are cha-21 racterizing could conceivably be one of the options for im-22 proving the organization and management of a configuration
(
23 control on a nuclear power plant.
24 DR. KERR: Did you get Mr. Sege's comments on why
() 25 they deleted it?
Heritage Reporting Corporation
<= m
52 i MR. LEE; I'd like to learn a.little bit beyond
{} 2 what was stated in the report, the final version of the re-3 port, if I may.
D
'/ 4 MR. SEGE: Well, if'I could comment--it was put in 5 and taken out. Two conflicting considerations that were go-6 ing to be studied. On the one hand, someone at the plant 7
who doesn't have to worry about producing electricity, doesn't 8
have to worry about other things that go on but can be a 9
specialist in safety and have sufficient sttture in the 10 organization for his views to be accepted for action in the 11 safety areas.
12 The down side of it that was going to be evaluated 13 against that, and in fact was, but the evaluation was felt As l you know, 14 --
is that if there'is one. person whose job was 15 dedicated to safety, his presence might suggest to others 16 they don't have to pay that much attention to safety because 17 this guy is going to take care of it and he'll watch out for 18 them. He'll let them know when they are doing something that 19 they shouldn't. That they don't have to worry about safety 20 all that much.
21 It was felt that that objection is very strong 22 because safety should really be an integral part of the 23 attributes of many decisions that are made by many people 24 in the plant in their day-to-day work. It's not that one
() 25 person produces electricity. Another person sweeps the floor.
Heritage Reporting Corporation n.n .
53 1 And a third person worries about safety.
2 , The operation of the plant is for the purposes of 3 producing electricity.- It's_a complex matter. Where the (V3 4 immediate maintenance needs, the longer-term maintenance needs ,
5 the needs for/ producing power: currently ~and the need to con-6 sider the safety consequences.of any of these actions are such
~
7 an integral part of so many positions in the plant that seg:.e-8 gating that-or creating a position that is intended as the de-9 fense in depth for this area, but has the effect of segregat-10 ing this, it was felt that that objection was stronger than 11 the merits of that position.
12 There have been public comments concerning this and Q 13 some of these considerations will go down for public comments 14 and felt that the study of this issue as a result of public 15 comments, the resolution has been sufficient to close it out.
16 There was nothing at Chernobyl that would suggest 17 the merits of such a position. But that is essentially how 18 the proposal for the study came about and the considerations 19 of it.
20 MR. LEE: Do you happen to remember which organiza-21 tion had objections primarily to this idea? Did they consult 22 utility groups?
23 MR. SEGE: Yes, I would have to look at up.
24 MR. LEE: Well, if you don't remember,.it's okay.
25 MR. SEGE: It's in the report. But I don't--
Heritage Reporting Corporet'on
< =n.
54 i DR. WYLIE: The conclusions were based purely on 2 the comments received?
)
3 MR. SEGE: No, no. On reconsideration that was 73
+ ;
' 4 stimulated by the comments received.
5 DR. WYLIE: You know, most utilities have in their 6 organization what they call safety managers. And I think 7
history will show that they've been successful in achieving 8 overall--not nuclear safety. I'm talking about safety opera-9 tion of the whole systems, and they manage safety, and their 10 main job is to promote as managers and instill safety in the 11 organization, and they've been very successful at it. So it 12 is sort of odd that they say naming a high level safety mana-ll 13 ger in the nuclear area would be detrimental.
14 MR. LIPINSKI: You know, the parallel argument to 15 me is quality assurance. That everybody is an organization 16 can instill quality and therefore you do not need any type 17 of quality assurance organization that is separate from the 18 people who are doing the job.
19 DR. KERR: Right on. I agree.
20 MR. LIPINSKI: But yet that's the case. Organizing 21 quality assurance is to oversee anyone else to make sure they
- 22 are doing a quality job.
L_J 23 MR. SEGE: The intent of the high-level dedicated 24 managers was not the same as having a safety specialist at
(~~
t 1 25 the plant. (Inaudible)
Heritage Reporting Corporation
< m> m .=
55 1 DR. WYLIE: You know, if you ask yourself the ques-2 tion who was accountable for safety at Chernobyl--
3 MR. SEGE: Well, the plant manager--
4 DR. WYLIE: The plant manager.
5 MR. SEGE: --was considered to be accountable.
6 DR. WYLIE: But something failed there.
7 MR. SEGE: Right.. He in fact was punished for not 8 doing what was expected of him.
9 DR. WYLIE: Yeah. He didn't do his job basically.
10 MR. SEGE: And safety was considered an integral 11 part of his job.
12 DR. WYLIE: In the U. S. reactors, the plant mana-() 13 ger again is the one that is held accountable for safety.
%/ 14 MR. SEGE: I'm sorry. I didn't hear that, f 15 DR. WYLIE: I say in the U.S. plants, the plant 16 manager is still the one held accountable for safety.
17 DR. SIESS: That's not true, Charlie. They fired 18 the Chairman of the Board at one utility. And 4
19 DR. WYLIE: Yeah, but the day-to-day operations 20 I think it's well recognized that the plant manager is the 21 one that's accountable for safety and where the management
- 22 of the safety conscience of the plant resides.
v 23 DR. SIESS: I've heard a lot of people say it has 24 to go all the way through the system, up to the--
() 25 DR. WYLIE: I've heard them say safety is Heritage Reporting Corporation (ses) utases
. . _ . . . . . . . _ _ - . _ . . . _ . . _ _ . . - ~ . . _ _ _ _ _ _ . . , _ . _ _ , _ . - _ . . . , , _ _ . . - . . _ . - _ . -_. _ _ _
56 I everybody's business. Well, you know, that is saying that O 2 everve av' '" 1"e 1 " b av' bu 1"e==- "o o"e t ecc u"*-
3 able. And you have to have somebody accountable to run the b
d 4 program.
5 MR. RYAN: Okay. The next task is 1.7A, which is 6 accident management. And-the issue here is to make sure or 7 to determine if the accident management planning work that 8 is being has=left any of what we might have learned from 9 Chernobyl.
10 Again, like the other management related projects 11 in house, there are two phases. One phase to document any 12 discrepancies. Phase two, to do what's necessary to fix 13 that.
14 It's here again that we run into this whole ques-15 tion of what might be stimulated by the Chernobyl issue '
16 above and beyond that which we 'have- just been asked to do 17 and this gets us to-the-manager-to-the-manager issue. And l
18 while that may pervade the plant during normal operations, 19 it may even become far more crucial during an accident-related 20 type of situation. Certainly anything that applies to the 21 other issue certainly applies here in spades.
I 22 The work scope again, the Chernobyl phase is basi-23 cally to identify the problem, prepare a report of recommenda-24 tions for factoring these in, and--at any rate, to prepare a O 25 regert of diecregencies end then, ef couree, the img1ementetiom I
1 Heritage Reporting Corporation i I
- cman.=
.- , - _ , . . _ _ . - _ - - ~
57 I phase is to start to look at ways in which we might correct 2 the situation.
)
3 The discrepancies report is due in December of this 4 year which is consistent with the dates I think you have.
5 And by June of 1989, the end of June, we hope to have the 6 recommendations typed report. And that basically goes beyond 7 a simple response to the Chernobyl issue.
8 Are there any questions about.those tasks? I'd 9 like to go on to the task'that the other' branch has, but to certainly to answer any questions about those for which we 1
11 are directly responsible, j 12 Yes.
13 MR. LEE: A_ question. The schedule for the imple-14 mentation of accident management plans by the end of June of 15 '89.
16 MR. RYAN: Right.
17 MR. LEE: I'm not even sure by the time we may have 18 even started the formal individual plant examination processes .
19 MR. RYAN: Well, certainly we're going to--again, 20 like the former one, we're going to use what's there. We're
- 21 going to use what's there from other programs. I didn't mean i
22 to imply that the IPE work was going to be the only source 23 of information. What's available to us, either formally or i
J 24 informally, certainly we will try to take advantage of it.
() 25 DR. KERR: In fact, according to a recent SECY, all Heritage Reporting Corporation im m
. . _ . - _ _ _ , _ . . _ , , . . . - .- _. . . ~ . . , _ . _ . . - . , . , _ . .-
58 I these programs are going to be carefully integrated.
() 2 MR. LEE: So I mean I'm concerned about redundancy, 3 a close-out of this issue, for example.
O 4 'MR . RYAN: Well, my feeling is that if we get down 5 to near December, and, of course, hopefully we'll know a lot 6 more than we do right now, and it looks like this thing should 7 remain open until the middle of the calendar year which is 8 really the kind of sch'edule for the close out, then certainly 9 I don't think we would hesitate to come back and say, look, 10 we would like to extend the Chernobyl'part of this through II June and here are the reasons why. Why we think that that's 12 the thing to do. For whatever reasons there are.
13 MR. WARD: Okay.
( O) 14 MR. RYAN: Lastly there is task 1.3A which involves i
! 15 Reg Guide 1.47. This is an issue that the other branch has i
16 and basically the issue here is to provide explicit means 17 whereby the people in the control room will be knowledgeable l
18 at all times about the status of the safety systems and 19 changes in those systems whether they occur because of some-20 thing happening to the system itself or whether somebody l 21 comes along and wants to take a system down to perform a task 22 to it. Some kind of surveillance or maintenance or whatever 23 it may be.
24 Basically it was decsded that this would be an in-() 25 house effort in which the Feg Guide itself would be revised j
Heritage Reporting Corporation j ,
<wn . ;
.- ,_....,___-.,,__._j
59 I
I to basically accomplish two things. One, that there would l l
O 2 ae some xima e em 1#atcetio=, ecae=1ce1 or oener 1 e- to l
3 the control room people, that a status was changing and also O 4 procedures for coordinating between the control room'and the 1
l l
5 test or maintenance people or whoever else might be involved.
l 6 So that systems are only taken down after this coordination 7 has been accomplished and everybody is on the same sheet of 8 music.
~
9 MR. LIPINSKI: Before you take it off, the last to line there on sign-off, procedures, that relates to that 11 earlier task. Are sign-off procedures'w'orking? Are they 12 adequate? Are human errors creeping in? Should there be 13 more mechanical means to convey information to the control -
14 room?
15 MR. RYAN: Well, you are raising--there are a num-16 ber of issues that are associated with this particular task.
17 One concern--one of the reasons for something like this is 18 the concern of some of the human factors people about the 19 proliferation of enuciators and displays in the control room 20 that we might be getting into a situation where people suffer 21 from information overload. And that by having something 22 q mechanical in there, the control room people observe and V
23 respond to it in some way, shape or form rather than the l
24 sign off on procedures. They exacerbate their performance. !
25 As a matter of fact, I guess the headlines in today's paper Heritage Reporting Corporation
< = > -
- . _ - - _ _ - . . . . - _ _ _ . _ _ _ = . . _ _ - - . , . _.___
60 i talked about the navy incidence which may be a classic. case O 2 or eeoete bet #e 9 1vea eautemeae e=a u#eroce ea 1#rormeti =
3 at such a rate that they couldn't deal with it.
O 4 MR. LIPINSKI: If you'are saying put it in and.
5 it's the wrong information, I could see you are having a 6 problem, but you are going into human factors expert systems, 7 this could be all put_through computers and the information 8 can be digested.
9
~
MR. RYAN: Oh, I don't disagree. What'we were
~
10 talking about here basically .is taking the existing control 11 room and start adding. things to it. So'there's a discussion aboutwhenisenoughenough,andmahbeoneoptionisapro-
~
12 ,
Q 13 cedure whereby two people. sign off:something, one being the 14 control room person and the other being the-task person or 15 whoever it is that's involved, and that be a knowledge on I 16 both sides that there has been coordination and knowledge 17 of what's going on. l l
18 MR. LIPINSKI: Okay. My question is is that system i
1 19 working or are there still errors that are resulting from the ;
i 20 fact that you are relying on humans and humans are committing 23 errors. !
22 MR. RYAN: I guess there probably are cases where--
O 23 MR. LIPINSKI: And the question is how significant 24 is that? Does the system need change?
O s MR. RvAN 1'm eorry thee there's some of these NRR Heritage Reporting Corporation (mmmu
i 61 I people who are not here or the gentleman who prepared-the 2 current Reg Guide.
3 DR. KERR: Well, in a sense, isn't that what this 4 program today--one of the questions it should answer--isn't 5 that the kind of questions suggested?
6 MR. RYAN: That's right.
7 MR. LIPINSKI: Well, you see, this question hac 8 come up before between the reactor designers. I'm going to 9 use a number like 1,000 valves.
10 DR. KERR: A thousand what?
~
11 MR. LIPINSKI: Valves. Not all the valves in the
]
12 plant, but those valves that are specifically related to ,
l
(~) 13 safety that have to be manipulated in order to conduct tests.
14 Some have position indicators on them. many of them don't.
15 And the argument is that you connect all those valves and 16 you end up with a proliferation of wiring and a large expense.
17 Now, to say each one of those is going to go to an enunciator 18 in a control room, I say you are not using modern technology.
19 BeCause that can be put through systems that digest the in-20 formation and only show it to the operators needed. And it's 21 that issue--if what is in place is adequate then you don't have 22 to complicate things by going ahead and putting further auto-(
23 mation. But if we are still encountering problems with human :
24 error that could lead to an accident because of the combina- i (v') 25 tion of things--generally we think one thing at a time is Heritage Reporting Corporation
- i=> = =
l
{
,_m
62 I going to happen. But some of the major accidents we've seen O 2 more tat =se et e time tuet neveee- t vo= eet the r1 9ae 3 sequence of things that happen, we could end up with another 4 major accident.
5 MR. RYAN: Well, keep in mind this is a Reg Guide. !
6 There is a discussion going on within the staff now about the i
~
7 degree to which we should be prescriptive. How much of the 8 burden for making the kinds of decisions and justifying them 9 that you are talking about should be left with the utility ,
10 or the plant itself. And what degree should the NRC people il establish some criteria based on error rates or whatever.
12 So you are more focused on what it is the utilities should O " "" "" " '
O 14 Now, the Reg Guide is going through, as I'll show ,
15 in a minute here, through the process of elevating it to CRGR 16 as a proposed Reg Guide and to the ACRS, and I certainly 17 believe that some of the kinds of issues that you are raising 18 will probably be part of that process.
19 Right now the Reg Guide says that you will accom-20 plish these two things. You will coordinate and you will 21 be knowledgeable. It is not terribly prescriptive. It does
! 22 not say we have done some kind of an impact analysis and
'1, 23 therefore say that you have to go to automation or more or 4 !
4 <
24 fewer displays or strictly keep this a paper and pencil l i
O 25 exercise. ;
i Heritage Reporting Corporation
<= -
c.. - .-, , .- - . . - . - . - - . . . - - - , -c~. - , . .
63 I so I think you have a~very good question. It is an O 2 issue. And certainly hope it gets more ironed out as part 3 of not only the Chernobyl follow up but the whole prolmugation .
O
- 4 process for this Reg Guide.
5 MR. LIPINSKI: One of the things that caused me to 6 think about this is the recent incident of Dresdon 2. Where 7 a single valve was manipulated and I think it affected about.
8 six instruments. And it caused confusion in the control room.
9 And when the operators finally did figure out what was happen-10 ing it was too late and the reactors did SCRAM. But they did 11 not know that particular valve had been manipulated. They ;
12 only saw the result of the manipulation of the valve. It Q 13 reflected in about' six other instruments that behaved that O I4 normally,. "
15 MR. LEE: One specific item. I'd like to ask you ,
i 16 if it was included or if it wan even considered in this con-17 text, which is the safety parameter display system. Is it con- l 18 sidered within the context of this regulatory guide at all?
19 MR. RYAN: Well, it very basically says--it's l 1
20 being left up to the utilities to make the decision. Now, 21 that may be a part of a display that's part of the main con-l
< 22 trol boards. It may be part of the SPDS. It may be an audi-23 tory type of thing. It may be a written procedural kind of 24 thing that we were talking about, the double --
. Cer-O 25 tein1y it.e being 1 eft up to the uti11 ties rieht new. 1e l
Heritoge Reporting Corporation inn. -
l
-. _ _ . . _ _ _ . - _ . - ~ . _ . _ . . _ - - . - . - - -. - - -. . , . . -
64 1 may turn out during the promulgation process of the Reg. Guide,
(} 2 because there are some differing' opinions within the staff, 3 that maybe that will change. But right now there's nothing O 4 that says you have to incorporate this into the-SPDS,'and 5 at the same time there's nothing that says you can't.
6 MR. LEE: Well, could somebody provide us some
~
7 guidance to the utilities at large or is there--
8 MR. RYAN: .That is the debate that's going on over 9 the Reg Guide right now, is how prescriptive, how much guid-ao ance, how much direction should be in the Reg Guide?-
11 MR. LEE: But I'm talking about specific guidance.
12 Does the debate include this issue of safety parameter dis-13 play system specifically? To your understanding?
(])2)
(
14 MR. RYAN: ' You mean directing people to actually i
15 use that as part--
16 MR. LEE: To study-- l 17 DR. KERR: Are you talking about including it as !
18 part of the problem or as-part of the solution? ;
19 MR. LEE: Part of the solution.
20 MR. RYAN: Well, the solution is being right now--
21 if you read the Reg Guide carefully, the draft, it's being i 1
22 left up to a large degree to the utilities. The way it's 23 written right now.
24 MR. LEE: Right.
() 25 MR. RYAN: There is a debate over whether we should i Heritage Reporting Corporation (set) 6asases
-65 1 be doing some extra work and make some of the kinds of--pro-2 vide some of the restrictive guidance that you are talking-3 about. It will consider the SPDS. It-will consider alter-4 natives to cluttering up the display boards. That's part of 5 the process that's going on right now. As a matter of fact, 6 this is what is going on right now. We've going through an
'7 office-level review of this thing to hopefully'get it in, get 8 some kind of a consensus ~to go'before the CRGR and on to:th'e.
9 ACRS by hopefully December.
10 MR. LEE: Do you have any feedback from those utili-11 ties that have installed SPDS's and have used them for some 12 time?
13 MR. RYAN: Their feeling about them?
v 14 MR. LEE: Yes. Do you know--
15 MR. RYAN: Well, I'm sure that they would tout them 16 but it's my understanding that the NRR has done a number of 17 SPDS reviews that haven't come out very well. And there are 18 a few people here from NRR that probably can answer this more 19 specifically than I can.
20 MR. INNERS: Kind of a different question. The 21 SPDS I don
- think was supposed to consist of a --
22 The SPDS is more of a plant variable process wherein--
23 MR. LEE: I understand that.
24 MR. MINNERS: So the experience of the SPDS I don't-25 think is going to answer this question.
Heritoge Reporting Corporation
=
\ _ - _ - - _ - - - - - - -
66
' j.
1 MR. LEE: It may not have a direct bearing on --
(} 2 and so on, but as Walter is indicating, a valve error could 3 show up in terms of critical safety problems. And I thought-I 4 these indications are very crucial sometimes. <
5 MR. MINNERS: Well, I think in the TMI experience the SPDS was proposed to give you some kind of indication of-6 7 what was going on, but you_didn't have to rely on valve posi-8 tion, that you had flow indications, okay. So that you could i 9 have a better understanding of what was going on in your sys-10 tem, il But the Reg Guide 1.47 I think is more directed 12 toward specific need. Things like what valve position.
13 Determining the status of the system.
i+
(](2) 14 MR. LEE: Thank you.
I 15 MR. RYAN: Thank you for your time, l
16 MR. WARD: Mr. Soffer is ready to go.
17 MR. SOFFER: Good morning. I'm Leonard. Soffer from 18 the Office of Research and I'd like to talk a little bit l 19 about some of our follow-up activities for Chernobyl with 20 regard to containment and emergency planning.
l 21 Just to remind you, the objective of this is to
] 22 allow for-feedback of Chernobyl information and insight in
! )
23 these areas to be-reflected consistent with U.S. reactor 24 characteristics into the ongoing staff efforts on containment !
() 25 performance under severe accidents as well as information on Heritoge Reporting Corporation l own. 1
67 1 accident response and consoquence assessment. There are
() 2 several tasks and I've combined them briefly but I'll try to 3 talk about them a little bit in detail, p
YJ 4 With regard to containment, we're looking at basi-5 cally two tasks, containment performance and filter venting, 6 ! 3.1A and 3.2A. And the purpose, as I said earlier, is to 7 reflect the Chernobyl experience where relevant in containment 8 reviews that are being carried out under the Severe Accident 9 Policy.
10 Chernobyl accident, of course, demonstrated very 11 strongly the importance of containment performance under l 12 severe accident conditions and while recognizing the very 13 significant differences between the Chernobyl containment 14 design and that of U. S. reactors, we feel nevertheless that 15 it is important to try to stay abreast of any kind of infor-1 16 mation that we learned from Chernobyl.
17 This is strongly related to existing projects that 18 we have underway. There are existing staff efforts both in l 19 containment performance improvement and individual plant ex-20 aminations that are underway to assess containment perfor-l 21 l mance. l
}
,-, 22 As you may know, there was a Commission meeting 3
( ,
mj 23 approximately two weeks ago where the staff presented some 24 I interim results with regard to suggestions for the MARK I l
rx
()
t 25 containments. There is a Commission meeting tomorrow with Heritage Reporting Corporation Fl m u.-
g
68 1 regard to the individual plant examination efforts that are 2 underway. And both of these will focus to some extent on (v) 3 containment performance. The containment performance effort
- 4 will focus on the more generic aspects. The IPE effort will S focus on the plant specific aspects, and so they are comple-6 mentary to each other.
7 The reason for its being tied to Chernobyl is 8 primarily to stay abreast of any new developments, any new 9 activities that we learn of to make sure that these are 10 appropriately factored into our continuing effort.
Il In addition, there is a continuing activity going 12 on on filtered vented containments. This is something that ll 13 is of great interest in a number of European countries and 14 h we have been following these efforts quite closely. We have I
15 been attending meetings, sending representatives to such 16 meetings overseas. And it is also worth noting that one of 17 the improvements that the staff has talked about with regard 18 to the MARK I's that is a hardened vent capability is for 19 all practical purposes a filtered vented containment. It 20 would employ a filtered vented containment.
I 21 >
The expected impact is we do not believe that the i
I cs 1
22 Chernobyl follow up is expected to require an expansion in x>
23 the existing efforts. And so it is perhaps not fair to call 24 this a research effort. It is a follow-up activity. It is
()%
i 25 a continuing awareness of any new developments in Chernobyl Heritage Reporting Corporation (302) 626 4848 i
69 -
i
- a. I as to_how it might affect ongoing activities that are already I
i n
U 2 in place. i i
3 MR. LEE: Could I ask a question? -
j 4 MR. SOFFER: Yes.
5 MR. LEE: Could you elaborate a little bit more
- I 6 on the French effort which requires filtered venting in 7 operating reactors?
8 MR. SOFFER: I am aware that the French are involved 9 in building filtered venting containments. But I am not sure 1
10 what it is that you would want me to elaborate on.
i 11 MR. LEE: What is the status or what is the' ratio-
- 12 nale or what kind of acronisms do they have in mind? Any !
more details that you can share?
( 13
! v 14 MR. SOFFER: I am not thoroughly familiar with it. !
l '
15 The French are basically building sand bed types of filtra-16 tion units on tops of the auxiliary building. The intent is 17 allow for venting under long-term conditions. They perceive l 18 it as a way of venting after a relative long-period of time i
19 to prevent any gross failure of containment. And they see it 20 as a way of reducing any radioactive releases so as to be j 21 compatible with their emergency plans.
1 l 22 And basically this is the purpose that they are
! (}
i 23 doing it. I think that there is some information that is i
J 24 available on some of the French design if you are interested '
i i
(]) 25 and I can probably send of the information to you.
) Heritoge Reporting Corporation j imm =
,, * ,,------,-,-r,,,,-n-+- . ~ - r-- -,,,,---,,--r- --r-,-y-,-r- y , , ,-v, - - - -y~, --w- wwT.-v-- v*-------w - -
---r ,r-e--- r ==w~
70
, 1 MR. LEE: The action is to be'taken annually, I t
2 presur..o?
3 MR. SOFFER: We understand--you mean when that
. D V 4 thing is to be done? q 5 MR. LEE: Right.
6 l MR. SOFFER: We understand from the French autho-j 7 rities that the decision would be made in Paris if it was
- 8 ever needed. And not necessarily at the plants themselves.
9 MR. LEE: Thank you.
10 DR. KERR: Len, has any consensus been. reached l 11 as to whether'the Chernobyl would have been contained.by 3
12 enlarged dry containment? .
13- MR. SOFFER: I cannot give you a single answer to 14 that, Dr. Kerr. The original that I heard two years 'ago was, i ,
15 yes, it would have been contained by a large' dry containment, l 16 Later on I heard a conflicting answer that it might not have
, 17 been. I don't'know if anybody else has.anything else to add 18 or has any.other information in that regard.
19 DR. KERR: That's'enough. .Thank you.
20 MR. WARD: Let's see. Is there any work going on
- f 21 to provide a better answer to that question or is it just j
22 perceived not to be of any useful interest other than satis-
,O, 23 fying curiosity?
l 24 MR. SOFFER: My understanding is that there is a l 25 continuing effort underway within the Department of Energy to Heritage Reporting Corporation
=
+ ww w 4 - -- ,-~<r , --e--ep-y =-wc - - - - -m.w--,w,w- , -w---r-- -, ,,-r,- a r- , w,- y-pg~+,ww-,---+ y-m,4s----- ,-g,- -,.<- - ,
71 I try to define the Chernobyl accident a little bit more pre-
'(} 2 cisely. And that effort may be able to pinpoint the answer 3 a little bit better. I think the question is basically the rR kl 4 energetics and whether a large dry containment.would have been 5 able to hold that.
l 6 But my understanding is that DOE has an' effort 7 underway to try to refine and understand the '$ernobyl se-8 quence.
9 MR. WARD: Do you have any. idea--i3 that at.one of 10 the labs or--
11 MR. SOFFER: I think there were several labs in-12 volved. I believe that Argonne is probably involved in that 1 13 to some extent. But I'm not sure whether they are'the only
(]O 14 ones.
15 On emergency planning, this is again a little bit 16 of a misnomer. First of it, it is not directly related.to 17 emergency planning. It is properly called accidant response.
18 There are two efforts that are-intended to be followed up.
19 One is the ingestion pathway protective measures. And the 20 purpose of this is to participate,with other agencies such 21 as FEMA, the Food and Drug Administration, EPA and the IAEA.
22 in evaluating the ingestion pathway protective measures that Oa 23 were taken by the Soviets as well as other nations with re-24 gard to food, water, animal feed, and to determine how this ,
(~)\
% 25 relates to potential U.S. actions.
Heritage Reporting Corporation imman
, ,, , ,.-~..,,..,,--,n~. , - , , , , - . . - , - - - , -- - , - , - , . - - - ~ . . ., , , .
1 l
E 72 1
And the work is related--well, first of all, let I
1
{} 2 me indicate what some of the problems and some of the issues 3 that came up. As you are no doubt aware, there were signi-p
\ 4 ficant actions taken by a number of countries following the 5 Chernobyl accident to regulate'the importation of food stuffs 6 from neighboring countries. There were conflicting standards 7 that were put out. There were different standards that were 8 put out. These tended to confuse people. Nobody was quite l
9 sure whether the standards were scientifically valid, whether 10 countries were playing political games were one another, and 11 in the process, a great deal of public confusion was generated .
12 There is a feeling on the part of a number of
/~% 13 authorities that there should be a harmonization of such Cf3\J 14 efforts in the event of another accident and one should try 15 to come along with standards that have a reasonable scien'ti-16 fic basis an.1 that people can agree with.
17 Thi" is an area where NRC is not directly related 18 since NRC would not be directly involved in the regulation 19 of food stuffs-following an accident. This would be.the job, 20 ot ?ourse, of agencies like the Food and Drug Administration 21 and the Department of Agriculture. But we are participating J1 with such agencies. FEMA is heading up this group in an 23 attempt to understand and to see what the implications of I 24 these activities would be.
() 25 The importance is that it would continue to work Heritage Reporting Corporation (202) 6244aes
I 73 i I with the sort of things that we're doing on understanding l l
(} 2 developments in consequence asse'ssment. We have consequence 3 models, for example, that we use to examine the environmental
\
(~%
ss/ 4 consequences of accidents and we look at ingestion pathway 5 measures. And one of the things that we really ought to be 6 saying is whether these activities are going to impact on the 7 consequence models that we're using and how they would impact 8 on it. And that is the primary area I think where it.would 9 affect the kinds of things we're doing. It might affect 10 environmental impact assessments. It might affect cost bene-11 fit assessments, the sorts of things that we do.
12 DR. KERR: Is the idea that one might modify the
(~') 13 current ertimate of what a sou:s term is or one might need "O 14 to modify the modeling of exizting source term as it--
15 MR. SOFFER: What we want to do is we want to use 16 the existing source terms that come out of our severe acci-17 dent research, but we want to take into account any activi-18 ties that are happening in the international community with 19 regard to potentiel changes in ingestion pathway.
20 If there is an agreement, for example, that foods 21 should not be distributed beyond a certain contamination 22 level, we would want to look at the models that we are using 23 and our consequence assessments-and harmonize those with 24 those kinds of levels, generally speaking.
() 25 Yes,-sir.
l l
Heritage Reporting Corporation (102) us-deos
~ _ . _ ..
74-1 MR. LEE: I was intrigued by a statement in NUREG
-(3 U
2 125] that the rainfall control, action that was taken by the 3 Russians along the vecation path is considered defective.
O 4 MR. SOFFER: The rainfall controls?
5 MR. LEE: Seeding of the rain away from the .vaca-6 tion path I guess.
7 MR. SOFFER: Yes.
8 MR. LEE: Is that kind of strategy a part of U.S.
9 evacuation plan as well?
10 MR. SOFFER: I'm not aware that anybody has deli-11 berately planned for a seeding or cloud seeding, but it is 12 certainly something that could be accomplished on fairly 13 short notice I'm sure. It does not take a great deal of )
14 advanced planning to do that sort of thing. But I'm not if aware that it is actually ir. the plans of any nuclear utility.
16 And I'm confirmed of that fact by checking with--
17 MR. LEE: I wonder whether there should be more 18 explicitly reflected--the way I remember reading the NUREG 19 report, it was mentioned somewhat as an afterthought. That 20 the Russians had taken this particular action. It was sort of 21 left dangling there in my opinion.
22 MR. SOFFER: It-was added as an additional comment.
23 It was pointed out by a number of commenters that the origi-24 nal draft statement did not include some of these activities
() 25 and it was decided to add some of the accounts of what th'e Herit'oge Reporting Corporation
<mi m.
h
~ , . __.I
75 1 Soviets did.
2 But I should add that the actions that the Soviets 3 took were not preplanned either. Almost all of those acti-4 vities were'ad hoc. And so it is not clear to me that those 5 sorts of things should be necessarily planned for in U.S.
6 activities.
7 Remember in the case of Three Mile Island, for 8 example, that we flew in an entire air filtration unit from 9 Idaho on less than one day's notice into Harrisburg. And to that was never planned for.
11 MR. LEE: But I mean after having learned these '
12 things, should we consider these actively, explicitly in the 13 emergency planning? I guess that was my question.,
14 MR. SOFFER: I think one'should certainly~ keep 15 those sorts of things in the back of one's head, yes.- I-4 16 think they should be retained within the consciousness of 17 planners, but the question of whether they should be actively i
18 and definitely put into plans is not clear to me.
19 DR. KERR: Ken, related but different topic. Has 20 any thought been given to designation if possible who is in 21 charge in a serious emergency? It seems to me that in TMI 2 22 there was some vacillation or maybe misunderstanding or maybe ;
(:) '
23 just unclear of who was responsible.
24 From what I read about the Chernobyl accident, a
({) 25 centralization of authority that they have in the Soviet Heritage Keporting Corporation
<=>.
15
' 76
/
1 Union assisted in rapid decisions that were carried out fairly
- O-2 vrom9t1v. t wou1d eeem to me thet eome eheueht ehou1d se 3 given to the process to be used in the establishing authority 4 and responsibility'unless in the view of those that'look at' s this~that it's already done with~ sufficient detail .
'- 6 (GO TO NEXT PAGE) 7 8
9 10
, 11 12 O "
O 14 15 16 17-18 19 20 21 22 23 24 O 2s Heritage Reporting Corporation amm T
y- --m es m . ume_& v umir m *w we w wer e w -- - n1 * .---__es-em:r e ew-h- se-w ee-w'wm -
te maw wa m e-
. . ._. - .. . - . .. . . - - - . = _ - --
Y 77 g MR. SOFFER: At the present time, there are no' l
('i 2 plans for that sort of a designation or centralization of
\_) b 3 authority, but that option always remains, of course.
() 4 I think you're asking a question that is a little-l 5 bit beyond my scope. I think that my is.that it--my under- 1 1
6 standing is that at the present time, there are no plans for 7 that sort of a designation or the centralization of authority 8 but that option always remains, of course.
9 DR. KERR: I agree with you that somethings probably 10 have to be ad hoc. The one you have mentioned is probably gg a good example.
12 But in order that things start moving, there needs i 13 to be someone in authority who has enough authority to get i
34 things done.
15 Perhaps, in our form of government, it is impossible 16 to establich that,.I don't know, but it does seem to me that 17 that question is worth raising in the study of the kind that 18 you seem to be doing. ,
19 MR. MINNERS: Maybe I'm not talkhs very; plain, 20 but I have a different perception. I think that TMI, the 21 plant manager had authority for on site action and I think
- 22 those things are~specified in current regulations as to who
%/
23 has authority.
24 DR. KERR: The Governnor didn't have authority to
() 25 order in units of the U. S. Army, for example, which certainly Heritage Reporting Corporation om .. I
- ~ . . . . . - . . - .. ._- . . _ - - .. .. .-
+ -
78 3 seems to be. - TMI II. Emergency--you seem to be saying
() 2 no further action is needed. My impression, from reading 3 about what happened in Russia is that we would not have been 4 abl'e to find the bundreds of tons of carbide as rapidly as 5 they did and get it'there and other things of this sort.
6 At TMI II.you didn't need that sort of thing 7 because it was not the kind of emergency that occurred at 8 Chernobyl. May, I hope we never will need it, but it seems 9 to me that some thought, at least,'ought to be given to 30 that.
11 Maybe you're right, Warren, maybe everything is 12 in good shape.
/'\ MR. MINNERS: And you may be right and our research 33
\# i4 pimi doesn' t include anything for that purpose.
~
15 DR. KERR: I recognize this and-I'm not being 16 critical.
17 MR. MINNERS: I want to make it clear that as 18 currently proposed, the research plan does not include any--
39 DR. KERR: That is the impression that I got.
20 MR. WARD: Dick Van Neal of NRR. In addition to the 21 governor's responsibility as far as taking action off site, 22 there is a federal plan for response to emergencies and with 23 other agencies in the lead, so there would be some organizatio1 24 there for taking action on site.
( 25 DR. KERR: If you look at this and you're convinced Heritage Reporting Corporation (ter) 62s-4aes
3 79 g that the authority did do things that it normally wouldn't.
2 do, I don't have a problem.
3 MR. VAN NEAL: It shouldnt be us.
O 4 on xean: So somebody e1ee ehou1d do ehee eed moe 5
NRC?
6 MR. VAN NEAL: That's right. FEMA has that 7 responsibility in addition to the other federal agencies.
8 DR. KERR: And you're convinced that they have 9 that authority.
10 MR. VAN NEAL: The federal plan has been looked gg at in that regard, yes, sir.
12 MR. LIPINSKI: The general question would be,
(
t g3 has FEMA looked at Chernobyl for lessons learned?
14 MR. VAN NEAL: Well they are, I think, Len will 15 cover that. They are working in a couple of'these areas in 16 that regard.
g MR. SOFFER: FEMA is in the process of looking at 18 lessons learned with regard to Chernobyl and how it should g, affect their emergency planning both for nuclear power plants 20 and or other for the national emergency level.
21 And, in addition, although I can't speak in detail ,
22 I would say there there ar e a ' number of other agencies, specifi -
23 cally the Food and Drug-Administration, Department of 24 Agriculture and EPA, that are also examining the implications O 2s of Chernoev 1 with reeerd to how ehev wou1d gureue their fun =eic ne Heritage Reporting Corporation msu
[ . _ .
. -. _ - .- . ~_. -.- - . . _ . -
'/
80 1 and regulatory activities under. accident conditions.
() 2 MR. LIPINSKI: Now in the report, there are several' 3 responses pertaining to potassium iodine. I don't see 4 anything in your plan pertaining to potassium iodines 5 MR. SOFFER: But there is a federal policy on 6 potassium idoine. The federal policy is' that it should is 7 that it should not be pre di~stributed' a to the general'public, that 8 it should be made available institutionalized workers--
9 instiutionalized individuals and emergency workers and that to any decision on potassium iodine should be made by the states, T
11 NUREG 1251 reiterates that that policy should 12 remain in effect, should remain unchanged.
()('
13 14 You want me to go on?
MR. WARD: All right.
15 MR. SOFFER: The last'two-tasks are on-decontamina-16 tion and relocation and again the major purpose is to partici-17 pate with other agencies primarily in obtaining information on 18 the Soviet decontamination and relocation efforts.
19 It has been alluded to earlier that the Soviet 20 decontamination effort has been a very extensive one. In 21 addition to doing hings'like-cloud seeding, it has been men-22 tioned that they spread a polymer substance to try and act as O 23 a fixative to prevent anv resuspension of radioactive materials
~
24 and we are aware that the Soviets have engaged in very signifi-() 25 cant decontamination efforts.
Heritage Reporting Corporation m .n
. . . , - . . , _ _ . , - , - . - - - , , - - - - - , - - - - _ . --- . . - - . . ~ , , - - - . , - - . , , - , - - . - - - - - , . _ . . _ - . . - , . - .
L' 81-1 There was a medical conference on the aspects of chernobyl that was held in May in the city of Kiev. A number
({J 2 3 of U. S. scientists were invitied to this conference and we 4 have had some informal communications with several of these 5 gentlemen and they have told us some of the efforts that the' j 6 Soviets have done with regard decontamination efforts.
7 For example, they decontaminated athey removed a l
8 pine forest in the immediate area of the site. They allowed 9 it to remain for about a year.to accumulate whatever burden
~
10 of cezium might accumulate and then cleared the land entirely 11 and buried the forest in an area that we're presently not 12 sure of.
13 They have applied fixatives, they have decontaminated
({
14 the site area so that site personnel for the 3 remaining 15 operating units are exposed to rather low doses. They are .
16 building a new city for the workers associated with Chernobyl, 17 Th'ey have set up decontamination check points so that vehicles 18 going from one area do not cross over a decontaminated zone 19 into a less contaminated zone and so on. And they are.
20 probably going to have the larest experiment, if you want to
, 21 call it that, in decontaminating agricultural and forest land 1
22 that we have yet seen. We hope to learn a great deal from i
)
23 that information and we are proceeding, through international 24 channels and"through other federal agences, hoping to gain O)
\_ 25 as much experience and knowledge as we possibly can of this 1
Heritage Reporting Corporation omsnau l
. . _ . . . - , . . . - . - . . . _ _ . _ _ . . ~
(o 82 i effort.
2
( It will be factored into, certainly-the kinds of 3
activities that that we would do-in an accident, if we hoped 4 to learn by it. We also hope to understand and change-any 5 developments that this might have,-again, in assessing the 6 consequencns of accidents, as our models, as_our computer 7 models.are presently designed to do.
8 I think that is all that I have unless there are 9 any other questions.
10 MR. WARD: Thank you very much, Len. Next is--
11 DR. KERR: Do you have any comments on how 12 effective polymar useage was or do you knom how effective-( j 13 it was?
14 MR. SOFFER: I haven't heard any numbers associated 15 with it. My understanding is that it has generally been 16 affective, but it has been used with in combination with 17 a number of other techniques. They have hosed down 18 vehicles, for examples. They divided the 30 kilometer 19 zone into 3 concentric zones and vehicles from one zone cannot 20 generally cross into another zone. You reach a certain check 21 point and everybody gets out and transfers to another vehicle 22 and then the one that is assigned to the contaminated zone 23 just simply turns around.
24 They have used water wash downs. They have used
() 25 some sort of chemical wash down which they have hot been very Heritage Reporting Corporation o.o u-
gl 83 i specific about and we understand that the entire range of
} 2 l methods have been quite effective, but I can't give you num-3 bers for anyone specific.
4 DR. KERR: I was just thinking qualitatively what--
5 MR. SOFFER: Qualitatively, DR. KERR:
~
6 Was the polymer effective enough that 7 we know what it is and it will be available?
8 MR. SOFFER: They have been less than forth' coming 9 in this area. We think it is because it.is related to some to of their chemical warfare activities and they are not anxious 11 to discuss some of these things.
12 MR. LIPINSKI: Question.
13 MR. SOFFER: Yes, sir.
34 MR. LIPINSKI: The zones made~in the reports- -
15 references me to Americans having automobiles, performing 16 their own evacuation. Under those conditions, you wouldn't 17 have wash downs and transfers. That isn't addressed in this 18 report at all?
19 MR. SOFFER: No, it isn't addressed in that report.
20 MR. LIPINSKI: The comparison between what our 21 evacuation consisted of compared to the Russian evacuation, 22 do you think that would be appropriate?
23 MR. SOFFER: To have decontamination of vehicles--
24 MR. LIPINSKI: No. To draw distinctions between
() 25 the way the Ru'sians s handle their_ evacuation versus how we .
Heritage Reporting Corporation von snm
_ . ~ - . . - . . . .- . - - - .- --. . . - .
84 I would handle one of ours?
2 MR. MINNERS: It was my understanding that they 3 had the zones during'the transition. You're not saying that?
Ci 4 MR. LIPINSKI: No.
5 MR..MINNERS: During the evacuation period, they 6 just loaded people up and drove them out.
7 MR. SOFFER: I think the creation--
l 8 MR. WARD: Even now is what I think he is talking about.
9 MR. SOFFER: I think the creation of those zones 10 came along later.
11 MR. WARD: Anything else?
12 (No response.)
t 13 MR. WARD: Thank you, Len. Doctor Odar.
^ a 14 DR. ODAR: I am Frank Odar. I am the Project Manager 15 of the Reactivey Transiets Project-at Brookhaven National 16 Laboratory.
17 I would like to make a small presentation on the 7 18 General Approach and the schedule of the project.
19 We have initiated a project at Brookhaven National 20 Laboratory to study the probability of the appearance of 21 selected reactivity transients and estimate their consequences.
22 We know that the design of the U. S. reactors--
23 but the safeguard reactor also involves the safe operation 24 of the reactors.
25 Our intent is to identify those transients which Heritage Reporting Corporation
<m, m.m. 0
, , . . - . - - . - , , - , , , , , , , . - . . , , _ . ~ . , . . - - , . - . _ , _ , . . . _ . . - , , - . . . , - - , , . - - , , , . . . . . . , . .
'Y . .
i 1 85 !
I would result inllarge reactivity insertion which would need 2 further study and our objective is to bring into question 1 3 previous judgments on the adequacy-of the current line basis O 4 reeceivity eccidene which hee been ueed in the 11 censing 5 Process.
6 Now here I would like to emphasize that the 7 Project really involved identification process, it's a 8 selection process. What the project focus' on is to 9 find out if there are some certain reactivity transients 10 whether or not large reactivity insertions can occur, jg rather than going and investigate into everything else.
12 So our study, will conce on those reactivity 13 transients which have been tabt 'dentified in NUREG 14 1251 and some of them are like L4 injection of un-15 borated water and boron dilution and then atlas"without 16 disturbing emission pump trip and many other transients 17 mentioned there.
18 And if you look at those transients,_in general, 19 you will find that they involve multiple failures. They 20 areprobably part low probability then, but the question is 21 whether or not the consequences would be severe or most:of 22 you had done what was expected, which involved a large reacti-O 23 vity transient insertion, like Chernobyl type.
24 DR. LEE: Are you going to try to rank them, rank 25 all these transients in some order?
, Heritage Reporting Corporation (m) 6st4aes v ,
/0 86 1 DR. ORDA: The project will have two approaches.
i[ ) 2 One of them is the probabalistic affect and the other is 3 deterministic assessment and the information you gain from 4 each side will be fitting to each other.
l 5 For example, we'll look at this without research 6 pump trip and suppose we have a deterministic analysic of this 7 particular transient which involves multiple failures and 8 which is very low frequency event, but suppose that it 9 doesn't lead to burst up reactivity excursion, then this 10 would not qualify, for example, for further study, 11 DR. LEE: You have quite a large number of 12 possible transients that you would like to consider.
13 DR. ORDA: This is what the project is all about.
In other words, we would like to identify'those transients 14 15 with multiple failures whi'ch may lead to severe reactivity 16 excursions. That is a very narrow focused project. I would j 17 like to bring that out right up front.
18 MR. WARD: Question Frank. In the ACRS letter 19 of a year and a half ago, I guess it encouraged a certain 20 part, but we were particularly interested in reactivity 21 transients, not so much that might result from multiple
,f- 22 equipment failures, but we hoped that the study would in-u) 23 clude transients that that might be--might develope because 24 l Of People weren't following administrative controls.
/'s (j 25 DR. ODAR: That's my next topic.
Heritage Reporting Corporation m .a..n
//
87 1 MR. WARD: Oh, I'm sorry, 2 DR. ODAR: The study will emphasize the low power 3 operation. Now let's remember that we have looked at
( ,
(_) 4 design basis type of transients. In other words, they are 5 generally a 100 percent power with single failure and we 6 looked at limiting conditions, but now Chernobyl brought up 7 up possibility of perhaps there are some other conditions, 8 other operating situations where large reactivity transients 9 may occur, so we are going to emphasize the low power opera-to tion, refueling mode of our operation, existence of high to li moderate coefficient in the first cycle, for example, for a 12 very low period of time, you may have that situation, ll 13 particularly operator errors, under all these situations. l 14 Granted that a fellow don't have all the information 15 available on operator errors, but we hope that we can get ,
16 an uncertain event, uncertainty on our calculations which 17 would tell us that this kind of a transient would fall into 18 a grey area or the other question that it can't happen or it 19 may happen and certainly we will be coordinating all of our 20 efforts with the human factors program that is going on at 21 the time.
7- 22 MR. WARD: I hope that operator errors, I hope this N_-]
23 i'sn't going to become a traditior.al PRA style analysis where you 24 write off things because,the "yes's" are experience indicated,
() 25 they're very improbable because I don't think, you know, that's Heritage Reporting Corporation m .n m.
J 2-88 1 not good enough because PRA wouldn't have predicted the Chernobyl transient, for example,
() 2 3 So, are you going to connect the sort of operator t- / 4 error you're talking about here, the sort of work that Tom 5 Ryan was talking about, where there are willful violations 6 of procedures?
7 DR. ODAR: We'll try to identify all--as we go a along, certainly, we will coordinate all of our work that 9 with Tom Ryan's work and the project will include not only go PRA, but also the deterministic analysis too.
11 In other words, even though operator really 12 committed all this errors and let's see what the consequences ll 13 will be. If he's just a little puff, we can't cut him down, 14 then we just put that transient away, it's not going to ful-15 fill the concentration, but on the other hand, if it is an 16 excursion, then we would like to know.
17 MR. MINNERS: The plan says the attention will be 18 focused on sequences and positive liks up that advise in 19 connection with deliberate contests and safety features 20 and because it includes human error, admission, omission--
21 DR. LEE: If I may ask again. How the NRC is
, 22 going to use the results that you are obtaining from this l ) !
l 23 study? Some of the events are highly probable. We know what 24 consequences we are going to be talking about without doing
(/ 25 any study, in my opinion. Some may be in the gray area that Heritage Reporting Corporation
<m>m4.
13 89 need to be looked at carefully, but some of the more severe
() 2 kind of reactivity transients, we know what the consequences 3 will be and at this stage,.what is your perspective?
r)
(_/ -4 DR. ODAR: That,'s the' criteria that we will be 5 talking about and. Ken is going to present a criteria and 6 you will see a combination of a probability or estimated 7 probability and the consequences involve '. from a deterministic 8 analysis for those particular transients.
9 DR. LEE: It is a low probability event but 30 large severe consequences we're talking about?
11 DR. ODAR: If it is very very large and everybody 12 , agrees that it's 10 - or something.
13 DR. LEE: How can you get such a small number like
~
14 10 ?
15 DR. ODAR: There will be some uncertainty, certainly.
16 Some transients may fall into a gray area because-of'the human 17 error, but we would like to go ahead with the project and do 18 it and see whether or not certain transients need further 19 attention. I think the aim of the project at this time is to 20 identify those transients that we should be concentrating on, 21 rather than resolve it at this point.
22 Brookhaven will present, the criterions 23 that were just mentioned, I think the slides'also show that -
24 the decisional, that the criteria has not gone through manage-() 25 ment'and the completion, therefore it is decisional.
Heritage Reporting Corporation (2e2) 626-4ees
. . . _ - _ . . , _ . . . . . _ . . , , , _ . _ _ ~ , , -. . _ _ , _ , .. .. _ _ _, . . .
d 90
, I would also like to show some preliminary re-2 sults tat you can look at and question about human errors v
3 if you would like.
) 4 The schedule for completion of the project is 5
early 1989 and that, report will be due in December of this 6
year and we hope we will be able to identify whether or not 7
those transients belong in one bin or the other.
8 MR. WARD: One more point that I pointed to in the 9
letter earlier, we suggest that you might want to look at 10 these transients in terms of what we call levels of events
,, and I see that as kind of a different perspective of the 12 pr babalistic perspective and the idea is that probabalistic 13 estimates areprobably not particularly good for predicting 4lF 34 human performance of this sort and we're suggesting that 15 some sort of another type of systematic analysis might be 16 developed. For example, where you talk about the levels of 37 error or something. How many levels of procedure to control 18 r how many levels of maloperation does an operator have to
,9 perform to get into a given reactivity transient.
20 Does he have to make a series of 5 errors or 2 l 7, ) erros or are you going to have to systematize the study l 22 in that way?
,S 23 DR. ODAR: We have a very large fault tree analysis g for one of the transients and we've addressed many of the
~
(v ') 25 failures. As you increase the number of failures, certainly Heritogo Reporting Corporation m .m.
19 '
91 1
the probabalistic number that you can calculate, I mean, l
2 it's ridiculously low at some times.
3 MR. WARD: Yes, but they're not meaningful either.
4 DR. ODAR: Right.
5 MR. WARD: I hate to hear people get too hung up 6 in this sort of work on saying, well, we calculate it's 7 10 ' , it's ridiculous and we're going to forget it.
g MR. MINNERS: I think the purpose of PRA is to 9 do a systematic analysis of what faults have to occur to 10 get you into a bad situation. It seems to be the best method gg that we have found.
12 The problem comes with what number you hang on 13 that. You pick out a number, but you need to apply some 14 judgment and--
15 MR. WARD: No, I don't think that's appropriate at g3 all for this sort of study.
1 l
17 DR. KERR: What do you think, that the PRA is 18 something like a probabability of willful violation in the j 39 regulations?
20 MR. MINNERS: You could put that in the PRA. )
l 21 DR. KERR: I know you can put it in, but will it have ,
l em 22 any significance? What is the probability of a willful viola-
%_l 23 l tion?
24 MR. MINNERS: You'11 have to take a guess at that,
(),
25 but the PRA can calculate the significance of a willful vio -
Heritago Reporting Corporation m u.4
lh 92
- lation which is just as important as the probabability.
2 DR. KERR: The PRA, ^ o me , is something--
}
3 DR. SIESS: You'rt talking about the failure l' >
1! 4 modes and effects modes.
5 DR. KERR: An effects analysis is what you're 6 talking about.
7 DR. ODAR: I think you may find out that many of 8 the transients, despite the errors, might not meet the burst 9 of reactivity and some of them will fall in the gray area, 30 by gray, I mean the ones that we have been looking at.
11 MR. MINNERS: Well, I think John's point is valid.
12 Most of the consequences of these events are pretty well known 13 at this time. You don't intend doing a lot of computer 14 code analysis on the consequences. They're pretty well know.
15 It's the problem of, hey, how can you get increases.
16 MR. WARD: Right.
17 MR. MINNERS: liow do get into these situations 18 and that's what I call PRA and if a PRA doesn't look into 39 that, it's not worth calling a PRA.
20 MR. WARD: But the problem with conditional PRA 21 methods is that the trees tend to get pruned based on the 7_.s 22 estimates of probability that are put in the various steps k.)
23 and as Bill has pointed out the willful violation probabilities 24 really aren't very well understood and so it might be over-(,) 25 looking sequences that you shouldn't be overlooking.
Heritogo Reporting Corporation can u. . .
14 93 1 MR. MINNERS: I think I agree with you because I 2 think I'm going to be happy to keep looking at those' violations.
[
3 I.think there has been a tendency to do this. There are pro-O 4 cedures to follow'and that's just not the case. Whether the 5 violation is wil'lful'or. unintentional or whatever, it doesn't-6 make much difference.
. 1 7 .MR. WARD: Maybe'a little close to home. the opera-8 tors at Chernobyl had noble intentions.
9 MR. MINNERS: They had noble purposes. The road 10 to hell is paved with good intentions.
Il DR. ODAR: That's all I have'.
12 MR. WARD: Any questions.
/"T 13 Thanks, Frank. .
s r 14 Next will be Dave Diamond from Brookhaven.
15 MR. DIAMOND: Good morning, gentlemen, I'm David 16 Diamond from the Department of Nuclear Energy at Brookhaven 17 and with me this morning- are Bob Fitzpatrickland Charlie 18 Hsu who will be assisting me in this presentqtion.
19 What we're going to do this morning is:to talk 20 about the project which. entitled "Reactivity Accidents,"
21 and tell you exactly what this project entails and what we 1
, 22 have done to date and what we hope to be doing in the future. I 23 I will spend some time repeating in a little more 24 depth what Frank Odar has said about the general objectives i
() 25 and the scope and methodology of this project.
Heritage Reporting Corporation mem
.l{
94 g I think it is necessary to spend a little time on 2 that type of an overview and then I will discuss specific 3 reacitvity event and I have borken them down into three 4 different classes on this particular slide.
5 One are PWR sequences with boron dilution,Lwhere 6 we have done a cbrtain amount of~ analysis already and we have 7 PWR reactivity events which,I call beyond design basis events.
8 I will explain that when we come to that,and then PWR reactivity 9 events.
10 And I wil try td go into detail explaining what we
- have done for a number of events, but also I will try to touch 12 base with all possible events, even events where we haven't 13 done very much in the past because I think one of the ideas 34 behind this meeting is not only for me to present to.you what 15 we have done, but also for us, at Brookhtven, to get -some 16 feedback from this subcommittee on your judgments as to which 17 events might be important or which directions we ought to be 18 headed in this project.
19 And I will then, at the end, of course, give a 20 brief summary and we'll have time for a general discussion.
21 The purpose already stated of this project is to 22 reconfirm or bring inte, question previous judgments an the O 23 adequacy of current reactivity design basis accidents, how 24 easily reactivity accidents have been of interest since day 1 0 25 te ene 1edeeerv. caermodv1 suee soue ue iato reen1=x189 ene Heritage Reporting Corporation
< mma.e
l
[$ l )
95 1
entire subject and seeing whether everybody has been properly 2 covered.
t 3 The scope of the study is to firstly develope
( : 4 possible sequences'of events and I will go into'some' detail 5
on how e are doing eac of these items, to determine conse-6 quemces, in some cases develope probabilities and also as a 7 secondary objective, to ' define the= criteria for events which 8 will need further study.
9 Now the study has aEfocus as'a result of Chernobyl go and I think as I-read through these, items, it's quite' obvious 3j how Cherncbyl has shaped this focus. One, we're interested in 12 sequences that arise in connection with deliberate by passing or disabling of any safety feature. We're interested, of'
(]) 13
(,) y course, in sequences whose caused include human error and 15 we're interested in low or zero power or testing as the 16 initial conditions prior to the reactivity event.
17 We're also interested in sequences that' involve 18 possitive moderator temperature coefficient, again, br l
19 obvious reasons connected with Chernobyl, sequences with in-20 adequate response of the shutdown system and finally, the 21 emphasis tends towards sequences where there is a relative 22 large reactivity insersion.
23 Now, the idea is to eventually be able to make a 24 judgment about these sequences and this schematic gives you an
() 25 idea of what our thinking is along these lines. And, of course Heritage Reporting Corporation
<mm.
n'3 l 96 i this type of thinking is really quite generic. There are 2 many NRC problems that are approached in this way.
[O s
3 If we just consider the fact that we have a number 4 of sequences or, if you can assume that any arrow is some 5 initiating event which representes a total sequence, possible 6 sequences.
7 We want to have some kind of a screening criteria 8 which will enable us to judge which sequences might, in the 9 future, be--it might be necessary for NRC to spend more 10 effort studying and this is, again, one of the general ob-11 jectives of this project.
12 Now, this screening criteria is not fixed at the 13 moment. We have a general idea of what type of criteria 34 we would use. We would use a representative frequency of 15 an event on one axis and consequdnce along another axis as 16 defining a matrix which would be useful for a screening 17 criteria.
18 Now the consequences on this particular chart 19 are related to the core. We are not going to look at anything 20 beyond what might happen to the core. And, generally, what 21 we do is we consider well, from this point to y-1 represents
,s
, 22 consecuqences that are usually acceptable--excuse me, that LI 23 are acceptable to the NRC. For example, no DNB.
24 Between Y-1 and Y-2, we have consequences which are 1
7s ,
L
.) 25 not acceptable. These are consequences which are something l Heritogo Reporting Corporation im> u.4m
pt.
97 I like a core melt. This is part of what MRC generally studies
'( ) 2 in the Severe Accident Program, so_this range here would be 3 core melt or, in the case of a reactivity accident, an increate s] 4 in peak fuel, ample peak beyond the current limits.
5 And, finally this other category of consequences 6 up here might be something a little more catastrophic than 7 what is currently considered in the Severe Accident Program, 8 namely, a fragmentation ~of fuel, a Chernobyl like event 9 where the core was disrupted-in a very short amount of time.
10 he lost coolable geometry.
11 So generally these types of events are most 12 interesting. If we were to find.those events that fall into l 13 this category are what is in a general consequence range 14 or of most interest.
15 When we talk about frequencyies, in general, we 16 talk about very low frequencies if we can think of increasing 17 crequency along the axis here, x-1 could be a number like, 18 for example, 10 ~ per reactor year. Something below which 19 most of the time, NRC is not concerned with the consequences 20 of the accident.
21 The range between x-1 and x-2 could be where curren:
3 22 NRC severe accident work is of most interest or is concentrated
-7 -5 23 now. Say it. he range of 10 , 10 ' per reactor year.
24 So it is in this range where there are a number of
() 25 Herituge Reporting Corporation
< =nm.
, - _ . - - - . . _ _ _ . . . , . . . .. ..-m, _. . - - . , , - , - - . - - , _ - - , . . _ . - - _ . . . - _
)1-98 i severe accidents that are currently considered by NRC. and 7, 2 those, of course, would fall in here.
3 And finally, about x-2, we have frequencies that g
j correspond sometime to design basis accidents and I should say 4
5 at the outset that the types of events that we're looking at 6 in this setting will probably not fall anywhere in this 7
high frequency range.
g But the idea behind this schematic is just to give 9 an idea of how one would approach the making of the judgment.
10 DR. KERR: Excuse me. Let me see if I got the right 3, idea. The idea I got is that you're going to use a product 12 f frequency and consequences to make a choice, is that the 13 right idea?
g MR. DIAMOND: That's correct. In someway, I hope 15 this answers Mr. Lee's questions.
16 DR. KERR: I get the impression that Mr. Ward was 37 q not thoroughly conviriced that exposure was the only way in 33 which one cught to make choices. Did you hear the same thing
,9 I did?
20 MR. WARD: Yes. Well, you know, I think this 21 generally is appropriate strategy but coming back to my prob-l 22 1 lem, I would rather see you put something other than frequency
(./
23 on top of that system.
24 MR. DIAMOND: Well--
[]
L .)
25 MR. WARD: Let me describe what I am trying to say l
Heritage Reporting Corporation m usa,ss l
}
99 1
here. I would see it in terms of something more like, maybe l
(v~'; 2 it's too simple to see, the number of mistakes and errors l 3 that would be required where over on the right, what you could L.) 4 say is the high frequency event there might be only one misteke 3
or error over 2, 3, 4, 5 or 6.
6 Because the problem is, you know there have been a 7
couple 2 or 3 dozen PRAs that have been done in this country g so far and they all say tnis isn't any problem, there's no 9 point in making a study, reactivity transients don't contribute 10 to risk.
gg So if all you're going to do is plug in the 12 traditional PRA kinds of numbers there on the bottom, I'm not 13 sure this study is going to be particularly informative.
34 The PRA sort of studies would have said the Chernoby l 15 accident coulnd't happen, but it did happen and I know you're 16 going to say, well nobody did a PRA of Chernobyl but I'll bet 17 you if one had been dona, they never would have predicted the
- l 18 high probability of the sort of accident that occurred there, 39 lbecause you're not looking through the right lens.
20 I mean, if you insist on looking at it as if you 21 really know the probabilities, I think you're going to miss
, 22 something here.
e 23 MR. MINNERS: Okay. It was predicted.
MR. WARD: Well, something like that.
24 I Well, you know, in a very rough way.
I 25 MR. MINNERS:
Heritage R3 porting Corporation m m4.
i E$N d 100 i
! MR. WARD: Right l 2 MR. DIAMOND: Let me--
3 DR. KERR: You're not using PRA by the way.
MR. DIAMOND: Let me make some comments on what you 5- have said. I think that, in the past, when reactivity
~
l 6 accidents have been addressed, primarily by making engineer- ;
I 7 ing judgments rather than by doing detailed probabalistic l 8 studies and so I.think-that'in that respect, this study 9 is some progress.
10 With regard to using absolute numbers, I agree with 11 you that one should not rely too heavily on the quantitative i
12 results that come out of this type of study, but I think it f' 13 forms a good framework for making. judgments and I think that's l V 14 one of the reaons,'of'soursel that you have not--
15 DR. KERR: Excuse me. Can you tell me what the ;
16 statement means that one chould not rely on the numbers but 17 it forns a good framew:rh for making judgments. What does 18 that mean? l 19 MR. DIAMCND: Okay. Well what that m.ans is that 20 I have not put specific numbers on this chart. I will not 21 put a number which says that yeah, if the probabilityLis 22 less than x-1,-then we can discard it. I think we have to 23 look at the probability and then also make a judgment based j l
24 on the suggestion--we're suggesting that well, perhaps it !
() 25 involves so many steps or perhaps we had to add something to Heritage Reporting Corporction
< =n.
I
/
75 101 I to the quantitative numbers that come out of this study.
I'i 2 MR. WARD: I think of it in terms of levels of
,)
3 defense and depth that have to be violated. At some point
/
4 j the defense and depth that y6u have described in different 5 way--
6 DR. LEE: That's a very interesting idea. I 7 thought about that.
8 MR. DIAMOND: I think that the primary focus of this 9 study is te get the information out and the secondary focus 10 is then to make these judgments and I think with these judg-11 ments, that we want the input of everyone who can think of how 12 one should make a proper judgment given this amont of informa-13 tion. This is certainly not the--
14 DR. KERR: You say the primary purpose is to get 15 "this information". What is "this information"?
16 MR. DIAMOND: Information on what sequences might 17 be, what potential consequences might be and what the 18 probabalistic results would be or deterministic results.
19 DR. KERR: So you are going to consequence calcula-20 tions?
21 MR. DIAMOND: Both--to some degree some consequence
- 22 and probabalistic results and I will show you that in a 23 minute.
24 MR. LIPINSKI: Let's stick to t12 BWR case without qj 25 the resort pump trip. In order to try to get an estimate of l l
Heritage Reporting Corporation (202) 42 4 4448
nh V
102 the frequency, you'll have to look at the hardware in detail,
(~'T 2 Particularly to see what kind of human errors could be com- l
(.;
3 mitted to get a better estimate of what that probability
- s
- 1) 4 is.
5 Do you plan to go into the hardware details?
6 MR. DIAMOND: Not for that particular event because 7 it--
8 MR. LIPINSKI: Well, how will you find the frequency?
9 MR. DIAMOND: We won't calculate the frequency for 10 that particular event and--
11 MR. LIPINSKI: Well, we know the consequences 12 before you start.
13 MR. DIAMOND: Well, yes and no. You know the con-T g4 sequences, certainly long term consequences. What is of 15 interest here is, can you get a large reactivity insertion 16 if you have events where there is a failure of the recircula-17 tion pump to trip.
l 18 MR. LIPINSKI: Well, it's not going to be a large l
19 reactivity insertion, it's going to be failure of the reactor 20 vessel. l 21 MR. DIAMOND: Okay. And that is not what we are l
,7 s, 2. addressing. We have to limit ourselves, to some degree, 23 and--
24 DR. KERR: Wait a minute. Are you implying that n
( ) 25 you get a failure in the reactor vessel without--as far as v
Heritage Repcrting Corporation (202) 628 4440
?1 103 3
your activity input?
(~ 2 MR. LIFINSKI: If you don't trip the resert pumps, C/
3 you have to have an atlas with loss of--
t)x L .. 4 DR. KERR: I understand that. I had always thought 5
that the reason you got the big pressure build up was because 6 you also got a big power build up because you got a positive 7 reactivity insertion.
8 MR. LIPINSKI: Right. i l
9 DR. KERR: Is that not the case?
l 10 MR. LIPINSKI: Right. You get the reactivity gg input as the pressure builds up.
g2 DR. KERR: Yes.
13 MR. LIPINSKI: So there is a reactivit. input W g4 giving you a power input, reactivity input--
15 DR. KERR: Okay.
16 MR. LIPINSKI: Power input and eventually you get to 17 a pressure that you don't have enough relief capacity for it 18 and the vessel fails or some other component opens up in gg that primary system to relieve system pressure, but it is 20 ef fectively a reactivity input that causes the auto catalytic l
l 21 failure.
,~ 22 Now, to get an estimate for the frequency, you have )
()_
23 to look at how many systems are involved that get you into 24 tnat particular condition.
(' '; 25 MR. DIAPOND: And as I said, the emphasis here is Heritcge Reporting Corporailon
<m> uom
104 1 on events where one is going to get a rapid power excursion. l l
(~] 2 Now we do--this is not to the exclusion of all events that LJ 3 involve power increases and as we go through, there -ill be
, i
\/ -
4 a number of events that I talk about where--we're not talking 5 about a very rapid increase in power excursion, but we are 6 talking about an uncontrolled increase in power.
7 With regard to the lack of research pump trip, we 8 have looked at that from the point of view of atlas and the 9 "oint of view of the consequences in the short term and 10 whether they could lead to a very rapid power excursion and 11 found that they cannot and therefore we placed it lower on our 12 list of priorities and will not be doing a probabalistic 13 study of it.
14 I can tell you--
15 DR. LEE: With regard to the transi)nt, in my 16 opinion, it should come clos'st to what happened at Chernobyl 17 because it will be associated with the void coefficients or 18 collapsing a void that would drive positive reactivity.
19 There has been some public interest or questions 20 regarding this particular accident and I don't think we have 21 a clear cut answer to give to the public either way, that es 22 this kind of transient is so improbable and consequences are D
23 so minor enough that we don't have to worry about it.
24 DR. MINNERS: Why do you cey that, John? We just
[) 25 got finished with the Atlas rulemaking which was inspected Heritage Reporting Corporation
<anu.au
20 105 1 decided upon. I don't understand that statement. You may 7,
( ) 2 disagree with what we've done, okay, but I certainly think v
3 that we have reams of paper which we analyzed probability of L' 4 consequences of Atlas event.
5 DR. LEE: But that was before Chernobyl and I think 6 were are looking at the lessons--
7 MR. MINNERS: I don't think Chernobyl has anything 8 to do with analysis.
9 DR. LEE: It has, in my opinion, because when you're 10 through Atlas, we were looking at more of a probable kind 11 of transient, but we are looking at more of a probable kind 12 of transient, but we are looking at more of the improbable 13 kind of trc.nsient now, we are forced to because--
MR. MINNERS: We did event trees on Atlas and wo 14 15 looked at all the events and supposedly the fixes had gotten 16 all of the branches of the tree down to reasonable levels.
17 MR. LIPINSKI: Let me comment about Atlas and 1
18 DWRs. Resert pump trip wats offered by General Electric with 19 assurance that with the pump trip, the reactor was going to 20 onerate at stable rower. Laballe II brings that into question l l
21 now, cs 22 MR. MINT 4ERS: That's being investigated, okay.
f 23 DR. LEE: Yes.
24 MR. MINNERS: That particular point is being
'tj 25 investigated, but that has nothing to do with Chernobyl or Heritage Reporting Corporation
<m> m.
0 3
106 I very little.
2 MR. LIPINSKI: It does with respect to the activity
<J 3 transients now. Offering a fix on the so called auto g)
( 4 catalytic reaction on Atlas on a BWR. With the resert pumps n w, w don't have to concern ourselves with the DWR, 5
6 nullifying the LBW activity transients as a result of having 7
triped those resert pumps.
8 MR. MINNERS: Well, I think as Dave has been talkinc 9 about, Lasalle and that and the involvement, I don't think 10 it's something you leans from Chernobyl is that you have the
,, possibility of instability. You have to read very deeply 12 into the Chernobyl analysis to come up with this.
13 DR. LEE: But the failure of resert pump coupled
,4 with the main stream isolation of closure could lead into 15 some failures and in my opinion that is, again, closest to 16 the reactivity transient of the Chernobyl accident.
17 MR. DIAMOND: But that particular sequence is well 18 known and so I don't think that we're adding anything, l9 resert pump trip versus Atlas and as you study this part of 20 i Atlas--
1 l 21 DR. LEE: Sure. But how can you account for the l 22 kind of transient in vour screening criteria? I guess that
(-)
i 23 is really my question. '
24 MR. DIAMOND: Where would it--for example, where
) 25 would something like that fall?
Heritage Reporting Corporation m u. .
107 i DR. LEE: Yes.
[1 L.)
2 MR. DIAMOND: Well, it would, as far as consequencer 3 go, it falls into this general middle area here where there
- r.
(' 4 are--currently people are concerned abouy severe accidents 5 and as far as frequency, I guess it would lie somewhere in 6 this central region here where NRC does have--already has a 7 policy or has been addressing Atlas and these types of events a for a long time.
9 DR. LEE: Then in your opinion, we should go back 10 to Atlas or boiling water reactor and say we cannot assume 11 that resert pump trip isn't reliable anymore or can we 12 still assume that the resort pump is so reliable that--
13 MR. DIAMOND: We are not doing anything in any 34 probabalistic work to look at resert pump trips and see 15 whether recent information changes NRC's perspective on resert 16 pump trips.
17 What we have done is that we looked at situations 18 where there has been no resert pump trip to see whether you g, could get into a rapid reactivity excursion and whether the 20 lack of resort pump trip could lead to a reactivity accident 21 rather than to de mvere consequences rormally associated with e 22 an Atlas event.
( ;
23 DR. KERR: How do you define a reactivity accident?
24 MR. DIAMOND: Well, it depends in what context.
() 25 DR. KERR: You define the context and then define l
l Heritage Reporting Corporation
??
108 1 accident.
MR. DIAMOND: All right. Well one type of re-( 2 3
activity accident or rapid reactivity accident like the m
4 current accidents that are a part of the safety analysis 5
are rod ejection accident, in a PWR in which you increase power on.
the order of--in the time frame on the order of a second, 6
7 that is something that I would call a rapid reactivity g accident.
9 Anything in which there is the potential for fuel f ragmen-a on for--
10 yy DR. KERR: So that sort of a criteria, unless 12 you get fuel fragmentation, it's not a reactivity accident, 13 is that what you're saying?
34 MR. DIAMOND: No. I'm saying that that would be 15 the sort of thing that would be of most concern. I would 16 say that any rapid rise in--unexpected rise in power could 37 be defined as a reactivity accident, yg DR. KERR: I guess I'm not sure what failure pump 19 trip would fall in that category.
20 MR. DIAMOND: All right, if I said that, what I 21 meant was that it would not fall in a category where it 22 would be--where it would yield severe consequences along the 1
23 lines of what I was talking about namely core disassembly 24 or fuel fragmentation, that sort of thing, n
) 25 DR. KERR: So there are severe reactivity accidents Heritage Reporting Corporation (2e21626 4444
v'~
109 g and then there are not quite so severe reactivity accidents
(' 2 and the pump trip is in the not quite so severe category.
3 MR. DIAMOND: Yes, I mean, in the context of
,. ~
4 todays talk, all the events that I am talking abcut are 5 lumped together under reactivity accidents, but only some 6 of them are really to be noted that they require special 7 attention in the future.
8 DR. KERR: Okay.
9 MR. WARD: You know, I think you put it well in 10 the beginning of your talk, you said that we have been gg worried about reactivity accidents since day one and, of 12 course we have, that's a Number 1 concern.
13 But that Chernobyl has tightened our heightened 14 our awareness or something and I think what that means is 15 we want to go back and kind of relook at some of our 16 traditional--some of the things we set aside, let's look 17 at those again and make sure we really examine them jg thoroughly.
39 What I hope you're doing and what we're most worriec 20 about, I think what you're saying along the consequence line 21 there, above y-2 is something very severe where there might bc
, 22 fuel, disassembly, fragmentation or something like that, and Nb
23 you have said that a BWR Atlas without a resert pump trip y would fall between y-1 and y-2.
I 25 MR. DIAMOLD: That's right.
Heritage Reporting Corporation imnaw.
- /
/
110 1 MR. WARD: That's the conventional wisdom and
(~') 2 that's what Atlas studies showed. I would hope that what you b'
3 would be looking is are there--is that really a good all in-
\_' 4 clusive assumption or are there other combinations of 5 timing or fuel site opener or other status cf plant or other 6 things that might go wrong in the plant that would make that 7 sort of--push that sort of incident up into--above y-2 8 and therefore should we go back and re-look, for example, 9
at the Atlas Rule or should we re-look at this that or the 10 other thing.
11 Okay, if that's what you're doing, I think that's 12 good, but it hasn't come clear that that is the sort of 13 thing you're doing.
14 MR. DIAMOND: That's what we're going, in general, 15 but with this particular event that you talk about, namely 16 Atlas without recirculation pump trip, we're not going back 17 and doing any additional probabalistic work because there is 18 considerable literature on Atlas probabilities and two, l 19 what we have decided to focus on was whether this type of 20 event could lead to a rapid reactivity excursion. We focused 21 on that aspect of the problem.
i f- 22 DR. LEE: Dave, if I may add one more comment here.
',\
~
23 The reason that bo+ h Walter and I are jnterested in this 24 BWR research into the pump trip failure is that I believe it 25 (l is one of the very first, if not the only auto catalytic Heritage Reporting Corporation m .a..
- d 111
! reactivity transient that I can think of in light water
() 2 reactors. Rod ejection is the rapid rate of the insertion 3 that you can count on negative model temperature close to t
1J 4 containment.
5 In boiling water reactor transient, this particular 6
transient is not a mechanism we can count on other than 7 that or failure or disassembly or something. That is what I 8 am concerned about.
9 MR. MINNERS: It turns around? It's not auto 10 catalytic, 11 MR. LIPINSKI: It's quick positive because the 12 fuel compensates for tne slow rise and eventually circum-13 stances--
14 MR. MINNERS: Just like a boiling water' catalytic.
15 DR. LEE: Uncontrolled reactivity, but there is no 16 inherent feed back mechanism that we can count on and that is 17 distinction that I am trying to make.
18 MR. MINNERS: Ours went to infinity without a trip g i
39 and it leveled off.
20 MR. LIPINSKI: They didn't trip, but this is Atlas 21 now, okay, this is Atlas, s
,- s 22 MR. DIAMOND: There are other instances of this I
~ -
)
23 type of--
24 MR. LIPINSKI: If you don't have enough steam
) 25 capacity, you are generating power at a greater rate than Heritage Reporting Corporation I
(kJ) 6264448 ll
Ah 112 I which you can dispose of it and the pressure goes up and
) 2 the voids go down and the power goes up.
3 MR. MINNERS: We understand that, okay.
ss 4 MR. LIPINSKI: It doesn't level off.
F 5 MR. MINNERS: It's acetodic. If you have a strong 6 enough vessel, okay--
7 MR. LIPINSKI: We're looking at todays vessels.
l 8 I understand there's an analysis that showed you needed a 9 reset pump trip to survive that transient.
10 MR. MINNERS: I understand but--
11 MR. LIPINSKI: But there are lots of ways you 12 get over pressures in plants. ?It sounds like you think 13 this is the world's strangest accident. I think that's 14 purple language which is not justified. It's a bad accident, 15 okay. We evaluated the accident and in certain circumstances 16 the pressure does go up, It doesn't go up auto catalytically) 17 it goes up acetodicly and probably goes above--certainly above 18 the design pressure.
19 I don't think it's proper to be saying some unique 20 accident that has auto catalytic reaction, I don't think--
21 l DR. KERR: It's auto catalytic in the sense that gm, 22 when the pressure goes up, the void goes down and when the J
23 voids go down, the power goes up and when the power goes up 24 and when the power goes up, the pressure goes up and that
, 25 is what an auto catalytic--it may not continue to be auto Heritage Reporting Corporation
<mmn.
31 113 l
1 catalytic forever, but for awhile it's auba catalytic, at l
f~)
G 2 least as I understand the meaning of the term.
3 MR. LIPINSKI: Why don't you call it positive feed I';
L/ 4 back, 5 DR. LEE: If you want to call it that, you can.
6 MR. DIAMOND: As I say, let me just repeat what I 7 said. We have looked at these events from the perspective 8 of getting a large rapid power excursion, but for the long 9 terms behavior in such events, we felt that that was covered to by research that had been done for Atlas and so we have not 11 done any probabalistic studies with regard to the reset 12 trip pump failure plus Atlas.
13 MR. LIPINSKI: I'm confused because on this list 14 it has Atlast without reset pump trip that your analysis 15 has been predicated on.
16 And based on your discussion, there's nothing here 17 to do.
I 18 MR. DIAMOND: Well,we did. We did some calcula-l t
19 tions because we were concerned, not just with recirculation i 20 pump trip, but also with the possibility of relief valve 21 failures and any mechanism which would tend to keep the power g- 22 up and, as you say, take disadvantage of the positive feed
' )l 23 back that one gets by collapsing voids.
24 So we were concerned with that particular event,
(,,) 25 but we addressed that problem by doing some deterministic Heritage Reporting Corporation om u.
9 ;.
114 1 calculations to see whether we could get a large energy 2 deposition in the fuel in a short period of time, not to look
)
3 at the long term behavior which we agree would certainly leac o
4 to a severe accident, but which we felt was covered by other 5 Atlas research that had been done in the past.
6 DR. KERR: Was this done in the context of your 7 view that the university staff found the Atlas without pump 8 trip.to be an acceptable accident?
9 MR. DIAMOND: No.
10 DR. ODAR: David was just going to prepare slide 11 25, 26 and 27 where all this analysis is presented.
12 MR. MINNERS: Why don't we just say, we're going to 13 get to those and just go through with his presentation, 14 wouldn't that be better?
15 MR. WARD: Dave, are these questions going to be 16 answered later on or--
17 DR. KERR: Promises, promises.
18 MR. DIAMOND: I think that I will give essentially )
19 the same answers later on. I think we, you know, I would 20 like to reserve until that time of the presentation, to dis-21 cuss this a little bit further to see whether our direction 1
,s 22 should change or at least, you know, it's the opinion of the (J)
R 23 ACRS subcommittee that it should change.
24 MR. WARD: Okay. The hands of the clock are pointing
) 25 straight up. Let me ask you. You're kind of in control of Heritage Reporting Corporation m .u m.
l
ej 115 g the agenda. I think we want to quit by 4:00 o' clock. We
/ ' also want to take an hc r for lunch.
) 2 3 Would this be as good a time as any to--
-~
(_) 4 MR. DIAMOND: Well, I would suggest that I finish 5 the general remarks that I am going to make on the methodology 6 and then before we get into specific sequences we could break 7 for lunch.
8 MR. WARD: Fine, let's do that.
9 MR. DIAMOND: The general methodology is shown on go this slide. One of the most important parts is the event 11 sequence definition and how we determine what sequences 12 might potentially lead to a reactivity event of concern and j 13 what we have done is, we've got an obviois extention of l
y presently analyzed events. In other words, there are a number 1 1 15 l of reactivity events that are part of safety analysis. It's '
16 obvious that if you have hypothsized additional failures 17 during those events that you might get into trouble and so 18 that type of approach to the problem is being considered.
39 We have also made judgments based on plant procedurc s 20 by looking at the way that the plant is operated at low zero 21 and power under testing conditions. We're trying to understar d
- 22 where the sufficient errors and f aults can be made which could t i 23 lead to a reactivity event of concern and, of course, we are 24 using the suggestions of others.
, 25 of course this work is based initially on the work Heritage Reporting Corporation imm .
n5
?
116 Howard Ritzens in UUREG 1251 and we talked to other peopic T~'i 2 who had suggestions for event sequences that might be of con-Q'/
3 cern.
4 Then the next step is under probabalistic 5 analysis or deterministic analysis. Notj regard to determinis-6 tic analysis, we don't intend to do any large scale compute 7 code calculations. We don't have the resources for this.
8 What we try and do is either extrapolate existing analysis 9 or look at the work of others who are also looking at tnese 10 events and if necessary do some selective computer code 11 calculations and the ida is to try and--there are many event 12 sequences and the idea is to use our time efficiently and 13 our resources efficiently so what we try to do is the 14 following:
15 We try to, after identifying a particular sequence 16 try to decide is it easiest to look at the quantifying 17 the frequency of that partricular sequence or is it easier 18 to quantify the consequences.
19 If, for example, it's easier to quantify the 20 frequency, then we will come down here and ac whether the 21 result is sufficient and by "sufficient," I mear., though
,s 22 we have looked at the event trees and the fault trees and we
( I
(/
23 see that the resulting core damage frequency is very very 24 small and we have given some judgments as to how many steps l' ) 25 are involved or used some other means of making a judgment Heritage Reporting Corporation cmi umn l
i!)
117 1 and if we can then make a judgment, then we say the job is
(')
LJ 2 finished and--
3 MR. WARD: That's where you eliminate the 4 Chernobyl event for further consideration.
5 MR. DIAMOND: I hope not.
6 MR. WARD: That's what I'm worried about, see?
7 MR. DIAMOND: I understand your concern. I don't 8 know how I can say it, but I wouldn't eliminate the Chernobyl i 9 event, but that says that, for example, the example I just l
l 30 gave, the frequency was so low as to not merit--
l gg DR. KERR: Wait a minute. What's the hatched area l
32 on that?
13 MR. DIAMOND: The hatched area would be an area 34 where there would be no further interest.
15 DR. KERR: Why could you eliminate it just on I
I frequency?
16 17 MR. DIAMOND: Well, you're right.
18 DR. KERR: There is zero. I don't know what your 19 scale is, but the frequency could be zero and I've still got 20 some unhatched areas up there.
21 MR. DIAMOND: That's correct.
, 22 DR. KERR: I assume that that goes up 45 degrees
- u. -
23 for some reason.
24 MR. DIAMOND: That's correct. One would have to
/^) 25 make some additional judgments in order to, at that point, say Heritage Reporting Corporation
<mnw
Id-118 3 that--
() 2 DR. KERR: You got two parameters, you can't 3 eliminate it'on one.
4 MR. DIAMOND: Right.
5 DR. SIESS: So I'm not sure what's on that fancy 6 thing with all the boxes and diamonds on it.
7 MR. DIAMOND: It doesn't show up there. This is a--
8 DR. SICSS: That looks pretty binary to me.
9 MR. DIAMOND: Yes, In this diagram it's certainly p3 binary, but one has to add in, that this is a general scheme 11 of how one would approach the problem, but one has to input 12 other factors as well.
l 13 For example the fact that you had some additional 14 information from other sources that the consequences were 15 really quito severe, then in that case you wouldn't even 16 I
though you had a low probability the study at that point.
17 So this gives you a general way of approaching the 18 problem, but there is some flexibility built in here.
19 DR. KERR: Provoded you don't follow that general i l
20 methodology.
21 DR. SIESS: I always wondered, did you really do it 22 this way, did you really go about it this way or did you just
! ~.i ,
x '
23 draw this diagram to explain to us what you did?
24 MR. DIAMOND: Well this is the--we haven't done--
(_) 25 DR. SIESS: Good.
Heritage Reporting Corporation im> u.-
n ')
l 119 1 MR. DIAMOND: We haven't--I should say we haven't
(, 2 finished this work. I think that this diagram is a good way 3 of explaining how we are trying to not limit the scope, bat
/ N
'- 4 limit the amount of work that we will do as part of this pro-5 ject. We have to make judgments as to what should be done 6 and what sequences should be looked at. We can't look at 7 overy single sequence and we can't do consequence and l
8 ' frequency analysis for every sequence.
9 MR. LIPINSKI: Let me review your diagram. In 10 order to put something in that matrix, I need two parameters.
11 MR. DIAMOND: Yes.
12 MR. LIPlNSKI: If I start out on the left, I 13 identify 3 procedures. .
I 14 MR. DIAMOND: Yes. - l 15 MR. LIPINSKI- I can't add anything to the matrix 16 unless I know what its consequences. So how can I-- i i
17 MR. DIAMOND: Yeah, unless you have information l 18 from some other source.
19 MR. LIPINSKI: That could mean you're going to quan-20 tify the consequences through some other source. Therefore 21 you don't have this path to go directly to the matrix on a
- 7. s 22 l single piece of information.
(_) I 23 i MR. DIAMOND: Okay. We will not quantify--
24 MR. LIPINSKI: As a consequence, he can't get to r
(3 ! 25 that matrix--
Heritage Reporting Corporation m u. a..
r
- f.
120 MR. UARD: I i ll give you an example. I think this 2 box where it says "add to matrix" can be really labeled 3 "think about it" and then the arrow going down say it's the 4 matrix.
5 MR.. DIAMOND: I think this gives us a framework 6 which is a good frainework for thinking about ' these things.
7 If we wanted to define.an.end point here for frequencies, 8 so that below this frequency or below that criterion, it 9 wouldn't matter what the consequence was--
10 DR. SIESS: That's.not on your matrix. You got 11 two variable things there. There has to be this and that 12 to be of.no consequence. You got nothing on that matrix 13 that tells me single plan converted by that shaded area.
14 MR. WARD: Shade the--okay, put a shade mark 15 above the orbit.
la MR. DIAMOND: Let's say we have some flexibility, 17 this diagram is a little too simple and that it worked like 18 that. Then it would, I think,; satisfy your objection that 19 we can-- l 20 MR. SIESS. Not my objcction. You're.the one that l
21 drew it. I'm just trying to let.you stick.to it. It's a 22 perfectly valid representation of the way I think. But it's' l O 23 the way that you're using it that I can't understand.
24 MR. DIAMOND: You know, I'm trying to be systematic O 25 or to now voo r temetic eneroeca to the vreb1em witae=t Heritoge Reporting Corporotion
(=nu
)
, _ - . , _ _ _ . . _ , - . _ . , _ . _._...,m. . - . . ___,.__.-_...,._;.~.--,.....-_.. - , , . _ _ _ , . . . _ _ , . , . . - , ,
ad
{ 121 I giving you, or at least I have tried, without giving you the impression that we're too rigid in using this formalism.
(') 2 3 DR. SIESS: Well, you know, you put a cut off on
,3 s! 4 frequency below which you're not going to look at the con-5 sequences. Just like Mr. Ward said, you're going to rule out 6 Chernobyl because it was a very low frequency event by any-7 body's cauculations.
8 Not after it happened, it's not low frequency. If 9 you would ask somebody to predict the odds of somebody doing to all those things wrong, it would be--well--and yet it hap-11 pened and the consequences were terrible. i 12 MR. MINNERS: But we did consider, in the design of 13 our reactor, Chernobyl type accidents and we didn't put 14 that in because we don't allow positive count coefficients.
15 I disagree with you.
16 ,
DR. SIESS: I didn't have anything to say about our 17 designs. I am talking about his methodology. Sure, we 18 included Chernobyl type accidents.
19 MR. MINNERS: We could have--If you put positive 20 moderator coefficient up there, okay. We wouldn't care what 21 the frequency was, we don't allow it.
7m, 22 I MR. WARD: Yes, but you're presuming-- l t !
23 DR. SIESS: I don't see anything on there that 24 says allowable. I see consequences and frequency.
() 25 MR. MINNERS; But the black is allowable.
HeritoGe Reporting Corporction (302) 428 4468
122 MR. WARD: Your argument is presuming that we f~/d x_
2 don't have any problems, because of our design practices and 3 everything and that is probably right, but what he is doing
(_) 4 is checking that one more time.
5 MR. MINNERS: Right.
6 MR. WARD: And so you can't go back to the argument 7 that everything is okay because he just scratched a study, i
8 l MR. MINNERS: The question I want to address is 9 the statement that has been made several times is that we to would never consider a chernobyl accident because it's too 11 low a probability.
12 DR. SIESS: Then why are we doing this research?
13 If you know--
14 MR. MINNERS: Because we want to see--for instance.
15 Analogous to having a positive moderator coefficient that 16 we don't know is there.
17 i Whethe.r we have some procedure that will let you 18 l get into some activity situation which we didn't realize we gg have here and .give you an analogous situation such as 20 Chernebyl, we're checking our previous analyses. That's all 21 W2'ra doing.
s 22 ; We're very much saying that a situation analogous
(' ) !
23 1
to Chernobyl is possible if we don't design the reactors 24 correctly so we want to go back and double check with our
(~~ ;
()
25 i reactors for design correctness.
Heritage Reporting Corporation m u. ..
-. . ~_ _ _ _ . . . _ _
a
' t(1 123 i
- MR. LEE
- But I thought also you need to go back 2 and look at some of the events of high consequences, but with '
3 low Probabilities or frequency, are we not?
4 MR. MINNERS: Yes, we're going to look at some-of 5 those. We're going to-look at rod injection accidents.
6 Those are very high--can be very'high concentration accident ,
7 Probabilities, but we think they're low probability.
9 ( *
. r
, 8 MR. LEE: But it's more than just a double check-9 ing to make sure that our previousIanalysis was correct, in 10 my opinion.
11 MR. MINNERS: Yes.
12 MR. DIAMONDS: Bece.use we're searching for new 13 events that perhaps were not considered in the past.
34 MR. LEE: Right. '
i 15 MR. WARD: That says we're looking for new ways 16 of getting to the consequences which may have always been 17 recognized as hypotactically possible.
4 18 MR. LEE: Right.
19 MR. MINNERS: We designed out a lot of these ,
20 accidents. Okay, you want to go back anx say, hey, did we 21 really do what we thought we did.
22 MR. WARD: The Soviet designers thought they had O 23 designed a control valve--
24 MR. MINNERS: And they thought that procedures were O 25 eaeauete e r co=tro1 e=a theete e1 o Pere e oer eeerca to i Her!toge Reporting Corporation mu f
. . , . . . ..-..-~.,,-,,.-.--,.-,--,-,,.-~..,,.,,_,-n, - - - , , . , . . . , . - ~ . - - - , . , - - . . . , - - - - . - _ ,,.,,-,-<---,n - - - . , ,
'q $
124 l l
say, hey, do we have some accidents which ware--
, 2 MR. WARD: We need to get off this. What it keeps !
3 coming back to, we're afraid that you're too immersed in the ;
(~\
v 4 PRA culture and we thing you're thinking too much about all '
s the current fashionable religious aassumptions that PRA is 6 making and we're worried that if you get too hung up on that 7 perspective, that maybe you're going to miss something that i
^
8 we're afraid might be important.
9 SIESS: If the consequences are great enough, some 10 of us don't get a heck of a lot of comfort from PRA.
1 U MR. MINNERS: Oh, I agree with you.- If they came up 12 and said, hey you could have a nuclear explosion at a power i 13 plant, I think we would get very nervous -about whether we 34 had--
15 MR. SIESS: And you really would not be concerned 16 with what the probability was, you would be concerned with, -
17 gee, what would have to be done to make it happen and how 18 can I be damn sure that doesn't happen. ,
g9 MR. MINNERS: I think we have to be aware of 20 the probabilities because we know they're accidents and we i 21 know there are rod ejection accidents. he can be very 22 serious and we're basically relying on the probability ap-23 proach that we have, either operational controls to limit 24 those or--
25 MR. WARD: I don't think we're--we love PRAs, we l
Heritoge Reporting Corporation l tsee) usases
-__~__
qd 125 I love rational attempts to estimate risk and really understand
~
[xj') 2 things.
l 3 All we're saying is that we don't think that the N/ 4 tools of the PRA and traditional data bases and assumptions 5 are up to the task that you people got here. We don't want 6 to see you relying too heavily on this.
7 MR. MINNERS: I think you're giving Dave a double 8 message because he puts up a little PRA slide there, you 9 give him a hard time and then he goes on and says, hey 10 we're not going to apply this rigidly, we're going to have 11 a little flexibility, which is what you're saying, and you 12 give him a hard time about that, so I think you're giving him l 13 a double message.
14 MR. DIAMOND: I think sorebody made this point 15 earlier today, perhaps it was Marren Minners, that at least 16 PRA---PRA is more than putting numbers into fault trees.
17 PRA is developing those fault trees and to develope those 18 fault trees, one has to understand--the TNIG ic going to 19 have to understand the engineering procedures, the operating 20 procedures, the technical specifications, the maintenance 21 procedures, et cetera, et cetera, et cetera and I think--
-y 22 so it's all part and parcel of the same--
(a 3 23 l MR. WARD: That's true, I agree that that is very 24 right and it is.
() 25 The problem is you're talking about screening Heritage Reporting Corporation l c mi ... .
ll0 ;
126 i
l here and your screening on frequency and that is the problem, i
t'^1 2y We don't believe the frequercies. We don't think screening
(/ [
3 on frequency is a sound idea for this issue. Sometimes it r
(_y) 4 is, sometimes it has to be, sometimes it's wrong.
5 Even if you don't estimate the frequencies 6 explicitly, you're screening a vast area on the frequency.
7 MR. MINNERS: Maybe it's only semantics, you seem to 8 be saying levels of defense and I interpret that--that's what 9 I mean by frequency. You don't like the word, "frequency,"
go you should go along with--
gg MR. WARD: Well that's very different. You know, I g2 think levels of defense is kind of more primative, but I don't 33 think you know enough about these sorts of events and that's g4 why--
15 MR. MINNERS: You got to look at--
16 MR. WARL: I was hoping you would come up with 17 something better.
gg MR. MINNERS: We can look at the frequencies and g9 recognize the uncertainties.
20 t, MR. DIAMOND. This is a good stopping point. ,
l 3 21 MR. WARD: We didn't mean to give you a hard time,
,_ 22 it's just that this is very important and we're doing the best
() we can.
23 MR. I' 'NSKI: Before we leave, back to the BWR 24
[) 33 relief valves. Relief valves have to be tested. In order to Heritage Reporting Corporation (242) 426 4484
ll 127 I test them you have got to gag the lower valve in order to get (v'; 2 the higher pressure valve there and I would be very curious 3 to see whether there has been any failures to remo're gags 4 on relief belts because again, whether the belt works in a
?
5 BWR sequence is a very important piece of equipment.
6 If the gag has been left on through human error, 7 it's not the same event that was analyzed previously.
8 P'. MINNERS: I agree, There are also human 9 errors . -setting the set points.
10 Mn. MARD: Okay, Dave, if you think we should 11 stop, let's take a break for lunch and come back at 1:20.
12 (Whereupon at 12:15 p.m., the subcommittee 13 recessed for lunch to reconvene at 1:20 p.m., s 14 the same day.)
15 ,
16 18 19 20 21 l
, 22
'\ _ /
23 24 g
V 25 l
Heritage Reporting Corporation
<m>m.m
1 l
128 1 AFTERNOON S'E S S ION
) 2 (1:25 p.m.)
- 3 MP. WARD: Mr. Diamond.
4 MR. DIAMOND: I said earlier that the event 5 sequences that we were going to look at-that I had grouped 6 into three separate categories. l l
7 (Slides shown.)
8 .M R . DIAMOND: And this is the first set' of evrtts 9 that I would like to discuss. And these all have in common 10 the fact that they involve boron dilution and-PWR. And so 11 there are seven general categories here. And as we shall 12 see as I go through them, each of these categories, there 13 are several sequences that are associated with each of these 14 categories.
15 I vill not read this list, and I will just start )
i 16 off by discussing the first item,.which is-adding diluted i
17 accumulator water during refueling. And for each of these l l
18 . events, what I will try and do is to briefly describe the 19 sequences that we are interested in, and what if.any 20 analysis we are doing, either probabilistic or 21 deterministic.
22 For this category, what we are talking about is 23 after the accumulator water has become diluted in its boron l
24 concentration, and during refueling the vessel head is off,
)
25 there is only a single motor operated valve between the Heritage Reporting Corporation
() (202) 628-4888
129 1 accumulator and the vessel.
2 And so we consider two separate sequences, A and 3 B. In the first sequence, water leaks into the vessel 4 through that one valve, and this is from one accumulator, 5 and slowly fills the lower plenum, so that the lower part of 6 the vessel has highly diluted water. Eventually, the water 7 rises to the level of the bottom of the core, and the core 8 becomes critical.
9 $
And the boiling of water in that lower part of the 10 core all of a sudden greatly increases the flow into the 11 core, and this essentially sucks up the unborated or the 12 diluted water that is in the lower part of the plenum, and 13 you have the potential for a rapid reactivity excursion.
14 The second type of sequence in this category is O 15 since the accumnlator is still at pressure and you have this 16 one valve between the accumulator and the vessel, we now 17 assume failure of that valve with the opening of that calve 18 which then leads to the diluted water rapidly entering the 19 vessel, and again the potential for a reactivity event, t
20 depending on how that water eventually finds its way into 21 the core.
22 With regard to the analysis that we have done 23 .ere, we have dcne some probabilistic analysis. We have 24 'oth faults trees and event trees for these events, and z5 I will go over them in a few minutes. And I have decided lieritage Reporting Corporation O (202) 628-4888 l _. _______ _ _ _ - - _ _ - - _ _ - - - . _-
I 130 1 that we would not do any dete ministic analysis. That would, of course, be a difficult problem'to calculate. And 3 in this case, we will have to rely on the resu]*.3 of the 4 probabilistic analysis in making judgments, or deterministic
~
5 analysis will have to be done in the future. Yes.
6 MR. LEE: In this particul ar probabilistic. 1 l
7 analysis, what in your opinion are the most uncertain ,
8 parameters or difficult parameters to get a handle on? l 1
9 MR. DIAMOND: Well, certainly the operator !
l 10 failures are certainly uncertain numbers. And I guess that 11 therearemanynumbersthat.hointothefault' tree. But 1 12 think that as with all of these events that it is usually ~
13 the operator actions that are the most critical, and-to 14 define the cut set which leads to a problem.
15 MR. LEE: But in this case, Scenario A, what
[
16 operator error would be the crucial one? _ s i l
17 MR. DIAMOND: Which gets.you the diluted )
l 18 accumulator water. Do not forget now that the accumulator l 19 under normal circumstances has to have 2000 PPM of boron.
20 So in order to get the boron concentration in the 21 accumulator down to what we would call a critically low 22 number involves a number of actions which we put into our. j 23 fault trees. And I will get to that. I am not finished 24 with that subject.
25 But before I get into that probabilistic analysis, Heritage Reporting Corporation (202) 620-4880
131 1- let me just make a quick statement about what we are talking 2 about when we talk about dilution of.the accumulator. And I
)
3 wanted to show you what we would censider to be the highest 4 boron concentration for a potent ial problem in these 5 scenarios. And we are talking now about refueling
)
6 conditions.
7 And in order to get a worse case estimate here, 8 the assumptiora that we make are that one has 2000 PPM of 9 boron in the vessel initially. And since this is during a 10 refueling outage that one has'to have the minimum 11 shut-down margin in the vessel of 5 percent. So if we 12 assume that they are just abiding by the minimum 13 shut-down margin, then that would be roughly equivalent to 14 500 PPM of boron which would have to be removed from the
[} 1: vessel in order to achieve criticality.
16 So in other words, if the vessel water were 17 replaced, if the boron concentration of the vessel water 18 would change from 2000 to 1500 PPM, then we are saying that 19 we could'have a potential problem with the reactivity 20 accident.
21 -
Now naturally, yen have to add a little bit more 22 than this in order to have an addition of reactivity. And 23 again, this is a simplified analysis that does not take into 24 account the fact that there is a certain probability that 25 you would have more than 5 percent shut-down. But just to IIeritage Reporting Corporation (202) 628-4888
- _ .. - - . - . - . - .-- -. - - - .~. . -.
132 1 give you an' idea of what a limiting dilution might be.
() 2 And therefore, tl-is means that-either the 3 refueling water storage tank or the accumulator must be 4 diluted to at least 1500 PPM for a potential problem.
5 'MR. WARD: There is some potential'that'you would 6 have less than 5 percent shut-down margin too.
7 MR. DIAMOND: Yes. Now the first of these, t 8 mentioned that there were two sequences. And in order to 9 get an estimate of the frequency for the potential 10 reactivity event, we drew this event tree which first states 11 that the reactor has to be in Mode 6 in a refueling mode.
12 That one out of the four accumulator MOVs has to leak. And 13 that there has to be a failure of circulation in the vessel.
14 Because if there is sufficient circulation in the vessel.
() 15 then this scenario will not take place.
16 You will recall that what I am postulatint here is 17 this scenario where you slowly build up in the-lower plenum.
18 And then after the core starts to go critical, you get a 19 rapid increase in diluted water into the core. So if you 20 have a large circulation due to the residual heat removal 21 pumps operating in a high decay heat rate, then it is 22 unlikely that you would get into that situation. And any 23 leakage of accumulator water into the vessel would mix with 24 the boron in the vessel, and this would be a very slow type 25 of event. So therefore, one has to have some sort of Heritage Reporting Corporation (202) 620-4000
133 l
1 failure of-core circulation in this mode. l
{} .2-3 And finally, of course, one has to have failure of the accumulator boron to be.at the' tech spec value. As I 4 said, it has to be at a critically low value, which we 5 estimated for ont case to be at about 1550 PPM. And this 6 event tree, of course, requires an analysis of this 7 particular probability here, which is the probability that 8 the accumulator boron has been diluted.
9 The fault trees for that type of event are given 10 in an appendix to the hand-out which members of the 11 subcommittee have. And I will not go through the details of 12 those fault trees. They are there for you to look at. But 1
13 obviously, those fault trees volve both failures of 14 various systems instrumento, valves, pumps, etcetera, and 15 also failures to take proper action, failures in following
(])
16 procedures, etcetera, in order to come up with a particular 17 conditional probability here that the accumulator boron is 18 critically low.
19 All right. So here we have a cituation where 20 without actually saying that this will lead to a massive 21 reactivity insertion, we look at the frequency, and we ueo 22 that the frequency is a relatively low number. And I guess 1
23 that another way of looking at it, the way that we were 24 discussing earlier, would be that there are many levels of 25 failures that would have to take place to contribute to the Heritage Reporting Corporation (202) 628-4888 O
i i
134 1 end point here. Now this is the sort of thing.that we are 2 trying to do for a number of events.
f(])
3 MR. LEE: On the probability calculation a.little 4 bit, suppose'now we are to say that the boron concentration 5 error of 500 PPM was reduced to let's 200 PPM, would'that 6 probability go up substantially or change at all in your 7 analysis?
8 MR. DIAMOND: Would that go up. It would probably 9 not go up. Because for the accumulator, let's see, the tech 10 spec if what, 1725 PPM. So I would say that any boron 11 concentration below the' tech spec would have the same 12 probability of occurrence. But above that' number, the 13 probability would be greater obylously, because you are not 14 violating a tech spec.
(') 15 MR. LES: But then I could also have a boron 16 concentration all tie way down to 500 PPM instead-of 1500 17 with the same probability and the same event' tree.
18 MR. DIAMOND: Well, not all the way'down, no.
19 Because then, of course, there is a time factor in here as 20 well And that would take a longer amount of time-to get 21 leakage into the accumulator, which would get you all the 22 way down to 500 PPM. So I think that there is probably a j q
23 range of concentrations where the probability is roughly 24 flat. And at high concentrations, the probability j 25 increases. And at low concentrations, the probability again i lieritage Reporting Corporation (202) 628-4888 i
a i
a
135 1 then decteases.
() 2 MR. LEE: So some t'ime factor has been 3 incorporated into your analysis?
4 MR. DIAMOND: No, no time factor has been 5 incorporated in this. This is a relatively simple approach 6 to this particular scenario.
7 MR. LEE: But this particular probability 8 characterization has more to do with the violation of'the 9 tech spec?
10 MR. DIAMOND: Yes.
11 MR. LEE: All right.
12 MR. DIAMOND: That is a very important component 13 in here. There are actually two ways of diluting the 14 accumulator. One is when one fills it. The water for the
() 15 accumulator is obtained from the refueling water storage 16 tanks. So one way is if the refueling water storage tank is 17 diluted and you fill the accumulator correctly but with the 18 tank water that has the wrong boron concentration. And the 19 other way is by leakage through check valves downstream from 20 the accumulator, the back leakage into the accumulator. And 21 that is probably the more likely route.
22 MR. LEE: How then did you come up with an 23 estimate of the probability of the tech spec violation in 24 the particular kind of scenarios?
25 MR. DIAMOND: Okay. The details of that are in Heritage Reporting Corporation
() (202) 628-4888
.1 H
1 136-
'l the fault trees which again are in'the appendix. And it is
(])' 2 a matter of going through and seeing.what.are'the paths into 3 the accumulator, and then seeing how any of those paths 4 might contain diluted water. And as I say, one path is the 5 normal fill path which is'from the refueling water storage 6 tank. And another path is from leakage, back leakage 7 through check valves that c placed between the accumulator 8 and the vessel.
9 -MR. LEE: I guess that I'did not make myself 10 clear. I understand the scenario of the sequence of paths 11 involved. But for each one of them,.I am not sure how one 12 can come up with a tech spec violation so easily, or how do 13 you do it.
14 MR. DIAMOND: I am not saying that-it is so easy,
() 15 but one has to get failure rates and leakage rates either 16 from the literature. Or if they are not1available in the 17 literature, by making estimates. For example,.the ,
18 literature has a lot of -- by the literature, I mean that 19 the LERs are filled with examples of situations in which the- 1 20 accumulator was found to be diluted. And so that'is one 21 source of information on what can happen, and a' source of 22 information that can be used to quantify part of the fault 23 trees.
24 Because in addition, of course, to the fact.that 25 you.might have a path for dilution,.there are procedures, i
IIeritage Reporting Corporation !
) (202) 628-4888 1
137 1 surveillance procedures, which check various' things in the
() 2 accumulator like level pressure and boron concentration.
3 And there has to be failure of that surveillance as well.
4 MR. LEE: Thank you, s
5 MR. DIAMUND: The other scenario is a scenario 6 where you essentially have the accumulator blow down into-7 the vessel. You will recall that I said there is only this 8 one motor operated valve between the accumulator and the 9 vessel. So now if you have this blow down when the 10 accumulator water is diluted, then the question is if you 11 have the potential for a reactivity event.
12 And again here, we have not studied'the 13 phenomenonology. Because obviously, one has.to have an 14 appropriate slug of water come through the core in order to
() 15 get a rapid excursion. And that is the sort of thing that 16 we are interestod in, a rapid power excursion.
17 So for this event, we have to have this one'out of 18 the four accumulators, and the MOV has to open. And then 19 the accumulator boron has to be diluted. So this number, of 20 course, is the same number that you saw on the previous 21 slide. And the result of this calculation for potential 22 reactivity event is shown on the slide here.
23 So now these are two events that involve a diluted 24 accumulator. And there is also the question of can there be 25 an event in which the refueling water storage tank is i
- lieritage Reporting Corporation l
() (202) 620-4888
138 1 diluted, and that water gets-into the core during refueling.
2 For this particular. event, there is still work to
}
3 be done here. We have to consider two different 4 sub-sequences i.ere. One, it takes into account the fact 5 that water from the RWST is normally added to the vessel 6 cavity when one removes the head for refueling. So one has 7 to devise a scenario in which that is done improperly. And 8 the other is if there is some sort of inadvertent injection 9 into the vessel.
10 Now these event trees have not been defined yet or 11 quantified. But the fault tree has already been quantified, 12 because the fault tree for the refueling water storage tank 13 dilution is a part of understanding the accumulator. water 14 dilution. So as you will see, there are a number of ;
{} 15 scenarios here that are tied to either the accumulator or l 16 the refueling water storage tank being diluted. -;
17 Again our intent here is to concentrate on the' 18 probabilistic analysis for these events. And we do not 19 currently plan to do any deterministic analysis.
20 Now there are a series of events'that have to do l l
21 with getting diluted water into the core during a LOCA j l
22 event. And the first set of these are as the result of the 1 23 ECCS water being diluted. And again we come back to the 24 same thing. This assumes that either the accumulator water 25 has been diluted or the refueling water storage tank has lloritage Reporting Corporation (202) 628-4888 O
139 1 been diluted. And then that water is rapidly pumped into
(} 2 the vessel during a LOCA.
The fault trees for dilution have been quantified.
3 4 They were germane to the previous two sets of sequences.
t 5 The event-tree for the sequence has also been quantified, 6 and I will show you that result momentarily. And again 7 here, we are not planning on doing any deterministic 8 analysis.
9 MR. LEE: The reason for this decision is yot have 10 arrived at the frequency of 10 to the minus 9 per reactor 11 year and some low values like that?
12 MR.' DIAMOND: Actually, the primary reason'is that 13 we have to make some decisions at this stage, because the 14 project will only go on for so long. If we determine after
/~T 15 looking at the probabilistic work that there ought to be
'% J 16 some deterministic calculations.done in the future, then we 17 would recommend that for a follow-up study. But at this 18 stage, I think that the preliminary results would show that.
i 19 the frequencies are rather low. I do not think we are ready i 1
20 to make final conclusions that this means that no 21 deterministic work should definitely be done.
22 MR. WARD: If I could ask another question.
23 Do you make that determination that a 24 deterministic analysis is required because you know that 25 there are some calculations of this sort of transient that Heritage Reoorting Corporation (202) G20-4008 O
140 1 exists that you are familiar with?
2 MR. DIAMONDS. No. No, I'do not know of any that ~
3 exist. And therefore, it' would be a complicated calculation-4 with a system thermal hydraulic code in combination with 5 neutronics in order to understand what kind of power 6 excursion one might have.
7 MR. WARD: So for none of these do you really have 8 much information on what sort of consequence of power 9 excursion there would be?
10 MR. DIAMOND: That is right.
11 MR. WARD: llave -you made some judgments?
12 MR. DIAMOND: That is right. In this first group, I
13 there are no decent calculations. The only thing that one l i
14 can do is a back of the envelope type of calculation which 15 shows that given a sufficient boron dilution equivalent to a
(])
16 dollar's worth of reactivity or thereabouts, that you could 17 hypothesize that it comes in in the right way, and you would 18 certainly get a power excursion. Maybe.more than a dollar, 19 but equivalent to some dilution.
20 MR. LEE: May I come back to this question of the ) i 21 accumulated boron dilution frequency. In a certain area of 1
22 the report, you have seen some instance like this, am I. '
23 correct?
24 MR. DIAMOND: Yes.
25 MR. LEE: liow many occurrences would you say that floritage Reporting Corporation (202) 620-4888 O
l0
+ '
141 1 you have seen over the last several years?
2 MR. HSU: Porty or fifty.
3 MR. LEE: ,
Over the past ten years or so?
4 MR. DIAMOND: Those are not tech spec violations, 5 the forty or fifty referred.
6 MR. LEE: Then I come back to your number of 7 1.7 times 10 to the minus 4, whether that is realistic.
8 Because if such an incident occurs once in the lifetime of a 9 plant, it will give you a larger number. ,
10 MR. DIAMOND: If it occurred once?
11 MR. LEE: In the-plant life, forty year life.
12 MR. DIAMOND: But it is detected, and then it is 13 corrected generally. So in order at any given time for it 14 to be undetected or ignored, the probability is lower.
15 MR. LEE: So if there has been let's say forty 16 instances like over the past ten years, we could be talking 17 about -- ,
i 18 MR. HSU: Detected.
19 MR. LEE: Detected, but still should we rule them 20 out from this calculation of frequency because they have all 21 been detected?
22 MR. HSU: No.
23 MR. DIAMOND: It is included in there. But do not 24 forget that not of those examples would lead to a reactivity 25 accident. Because the tech spec limit is above-the limit Heritage Reporting Corporation O (202) 620-4888
142
'1 1 where you would get into trouble. In other words, if the
() 2 tech spec limit is 1700, this simple calculation here would i
3 'show that you do not get into trouble until you get down to 4 below 1500 PPM.
5 DR. KERR: So the number that you are using for 6 probability then is rather arbitrary, it is a good 7 engineering judgment of what you think would be a reasonable 8 number?
9 MR. DIAMOND: The number for what, the' bottom line j 10 number for the conditional probability that the accumulator 11 is dilute'd?
12 DR. KERR: -What the likelihood is that this l 13 accumulator discrepancy will cause problems.
14 MR. DIAMOND: I am sorry, the number? I
() 15 DR. KERR: Well, the number of times that the 16 accumulator is diluted is clearly much larger than the 17 number that you are using in your fault-tree.
18 MR. DIAMOND: Yes.
19 DR. KERR: The number that you are using in the 20 tault tree must not come therefore from experience, but from 21 an extrapolation of experience to what you think to be those 22 instances that might cause difficulty, or am I missing 23 something? j i
24 MR. DIAMOND: I think that that is safe to say. j 25 The numbers are not used directly. ;
i l
Ileritage Reporting Corporation (202) 628-4088 i
143 1 DR. KERR: So the numbers represent not (n_) 2 arbitrariness maybe, but good engineering judgment or 3 something, they are based on experience.
4 MR. FITZPATRICK: If I may comment here. We try 5 to use the cata as we found it in the LERs. The data that 6 says forty to fifty times in the last ten years that have 7 been low boron concentration, the cut set that includes 8 things like that also says that people fall to act on it, 9 etcetera. So that is what is driving down the probability.
10 That is an initiating event type of thing.
11 DR. KERR: But you do not know how many times 12 people fail to act on it, do you; I mean that becomes a 13 judgment call, does it not?
14 MR. FITZPATRICK: We have a handle on some of (3
\_/ 15 that. Part of what we found here and the leading 16 contributor to that quantification is actually where we have 17 found that licensees have known that they have had a fuel 18 leakage and have ignored it for months at a time.
19 And it appears to use from the literature search 1
20 that we have done that they have made a judgment that it was l 21 cheaper apparently to pay the fine than to shut down the l
22 plant and fix it. And that particular item right now is the l I
23 leading contributor to the quantification of an event like 24 that.
25 DR. KERR: I am not questioning your use of 73 lieritage Reporting Corporation
() (202) 628-4888
144 1 engineering judgment. I am simply saying that that'is what -
j 2 it appears you have done. You do not really have data for
}
3 this event.
l 4 MR. FITZPATRICK: We have LER data which has been 5 put into the model. l 6 MR. MINNERS: Has any. accumulator ever been 1
7 diluted down to 1500 PPM?
8 MR. DIAMOND: Yes, a couple of times.
9 DR. KERR: Is that the data that we used for those 10 numbers?
.11 MR. DIAMOND: The data that.we used is-generated l 12 by looking at the detailed fault tree.
13 DR. KERR: I do not see how data can be generated 14_ by looking at a fault tree.
(} 15 MR. DIAMOND: I am sorry. The bottom line number 16 that we use. l 17 DR. KERR: Yes, okay.
18 MR. LEE: So for example, the bottom line to me is 19 if there have been two boron dilutions below 1500 PPM over 20 the past ten years which would translate into approximately-21 a 1000 reactor years of operation. So it would be two times 22 10 to the minus 3 at least.
23 MR. MINNERS: It does not work that way, John.
24 MR. LEE: Pardon me.
25 MR. MINNERS: You are not doire, the calculations lieritage Reporting Corporation (202) 628-4000 O
- m. , , _ _ _ _ , _ _ _ . . _
._.,..,,....m , , * - . , _,._ , sm-,. , ,,,_4,, e
145-1 correctly. You can'only be in that condition for so long. l
{) 2 3
This is not a frequency that they are calculating.
MR. LEE: That is what I am trying to understand, 4 .when do'we account for this time factor or time' window..
5 Then my-question.is suppose for various reasons of 6 uncertainties that you come out with a number like 4 times 7 10 to the minus 7 per. reactor year for the first one that 9 you are talking about, the leaking accumulator valve, would j 9 you change your decision regarding whether you should look 10 like at this more or not?
11 MR. DIAMOND: I think that the reason that I said 12 that we were not doing any deterministic calculations in 13 that area was that those are difficult deterministic 14 . calculations to do. And as far as making a judgment as to l 1
{} 15 whether they should be done in the future, if the numbers 16 came out higher, that would certainly make that judgment
.17 more difficult, but it would have to be done in 18 corroboration with NRC staff. ;
19 MR. MINNERS: I do not know what our contractor is 20 saying. I hope that he is not giving-you the impression 21 that just because calculations are difficult that we are not l
22 going to do them. My assessment of the situation is that we !
1 23 calculate a fairly low probability on it and it is not worth 24 looking at on that basis.
25 Are you. telling me that if the calculation only Heritage Reporting Corporation (202) 628-4888
, . _ _ _ _ _ ~ , . _ . - _ , _ _ _ . _ . _ .
146 1 took S100 that you would do it with,that probability?
() 2 MR. DIAMOND: Yes, probably. '
3 MR. MINNERS: You do not have my philosophy.
4 MR. DIAMOND: Because that would give me added 5 confidence.
6 MR. MINNERS: So what is the confidence ofEthe 7 probability of the difficulty in doing the analysis. I 8 would agree. That if it only cost $100, you would go out 9 and you would probably do a simple calculation. If the 10 probability was higher and it cost less, maybe you might do l
11 it. But based on the probability and the difficulty of 12 doing the calculation that it is not worth it, i
13 MR. DIAMOND: But it is not just difficulty of )
l 14 doing the calculation. i 1
() 15 MR. LEE: But alco one needs to consider some of ,
16 the alternatives involved in these estimates. That is what 17 I was trying to understand.
i 18 MR. MINNERS: I always consider uncertainties when 19 I discuss probabilities. That is a given when you talk 20 about probabilities. In fact, it is a given when you talk 21 about deterministic calculations. They are also uncertain.
22 Su uncertainties are always considered. ,
23 MR. LEE: But only implicitly apparently.
24 MR. MIllNERS : Yes. I personally think that most 25 of these cases doing an explicit quantitative analysis of fleritage Reporting Corporation
() (202) 620-4888
147 1 uncertainties is a waste of time. Because the uncertainty r^3 2 ic so large that it does not tell you anything.
(/
3 MR. DIAMOND: For the LOCA evento in which one 4 injects diluted water.
- 5 DR. KERR: Excuse rue. Are you workin' o some 6 extent with this man who just spoke? I gather tuat he 7 thinks what you are doing is a waste of time. ;
8 MR. MIllllERS: I think that what he is doing is 9 correct. He has got a low probability and an expensive 10 calculation, so why'do it.
11 DR. KERR: I thought you said calculating the 12 probability was a waste of time.
13 MR. MIllNERS : 13 0 , I said calculating 14 uncertainties. John asked me if we were explicitedly
{} 15 calculating uncertainties and I said no, I firstly think 16 that in most cases that they are a waste of time.
17 DR. KERR: Okay. I misunderstood you. I thought 18 that what .Tohn asked was if the uncertainties were being 19 considered other than implicitly.
20 MR. MINilERS: And the answer is no.
21 DR. KERR: Yes. But I do not think that he has 22 requested that you try to calculate in detail what the 23 uncertainties were. What he was suggesting I think was that 24 they be considered in the process of deciding whether you 25 needed to make a calculation.
Heritage Reporting Corporation (202) 628-4000 -
- 1. . . .. .
hk 148 l>
1 UR. MINNERS: You know that we consider
() 2 uncertainties. That is practically calling us incompetents 3 if we do not consider uncertainties. l 1
4 MR. DIAMOND: For the LOCA scenarios, we are !
S talking about large break LOCA. And on this slide, we are i 1
6 interested again in what the highest boron concentration 7 would be to have a potential problem, a potential for a 8 reactivity event. Or equivalent to what the minimum !
9 dilution would be to cause criticality.
10 Before we gave some numbers'for refueling 11 conditions. This is now for large break LOCA. The
.12 assumption that all rods remain out during a LOCA and at the 13 beginning of the cycle. And based on an assumption that the 14 plant is operating with a boron concentration of about
() 15 950 PPM of boron under normal operating' conditions and 16 taking into account the fact that when the power decreases :
17 during blow down and cold water-comes in as you get the-18 reactivity of 2.5 percent which is a power defect which is 19 equivalent to approximately 250 PPM, it means that you-would 20 have to have less than 1200 PPM in the core in order to 21 consider some sort of potential reactivity event.
22 So this number is a little bit lower than the 23 previous number which I said was about 1500 PPM. .And 24 actually, it means that the refueling water storage tank 25 must be diluted down to at least-1200 PPM or that a minimum i
i lleritage Reporting Corporation
{} (202) 628-4888 i
. _ _ . . . . . - _ _ . , _ . - , . . . . . . , . . - , _ . . _ . _ _ , _ . _ . ~ , . . , . _ . . . _ _ . - - , - , , , _ _ . , . . _ . . . , . . _ ,
149 1 of t' o accumulators must be diluted to deliver a mixture of 2 less than 1200 PPM. Because now during LOCA, we are talking 3 not about one accumulator leaking-into the core. But of 4 course, we are talking about all accumulators blowing down 5 into the core.
6 .This.is getting a little bit ahead of the game.
7 But it alco means that you would need a very large amount of 8 unborated water via the sump or via steam generator'tubef ;
9 ruptures. And this is getting a little bit ahead of the 10 game, so I will defer discussing it for now. If.there is 11 reactor trip or considerable burn-up, then a larger boron 12 dilution is required.
13 MR. LEE: May I ask one more question?
14 MR. DIAMOND: Sure.
r' 15 MR. LEE: Have these event trees-that you have 16 generated for these few transients, have they been -
17 represented in some of the existing PRAs for pressurized 18 water reactors?
19 MR. DIAMOND: No. The fault treP7 and the event 20 trees that I have shown so far, the answer is no.
21 MR. LEE: Why not?
22 MR. DIAMOND: Because I do not think that people 23 have considered these types of reactivity events as being 24 credible or sufficiently credible to he looked at in a PRA.
25 I do not know, that is my guess.
Heritage Reporting Corporation (202) 62P-4880
_,-,,c,,, nn , . - - , y , - - ,
n.
i 150 1 MR. LEE: Are far as you know, these event treen
/
(_m) 2 and fault trees are completely new?
3 MR. DIAMOND: As f a r' a s I know. Now so saying 4 that, what I am going to do is put on an event tree which 5 actually comes from a PRA. But superimposed on this event 6 tree are the events that we were just talking about, which 7 were not in the original analysis. And I should mention 8 that everything except what relates to a reactivity accident 9 on here comes from the Byron PRA. We are trying to use 10 Byron and in general the Westinghouse four loop plant as the 11 model for the analysis that we are doing for PWR.
12 Now this event tree, except for what we added 13 which I will explain in a moment, is a typical large break 14 LOCA event tree. Thi.s is the initiating event. This 3s the e,
k_. 15 availability of the refueling water storage tank as a source 16 of ECCS injection. I will skip this one. This is the 17 safeguards actuation signal. This is operation of low !
10 pressure ECCS injection. And this is the containment spray, 19 and containment coolers, and operation of low pressure ECCS, j 20 and long-term cooling mode.
I 21 So those are the events that are normally on the 22 LOCA event tree. And what we have added on here is several 23 things. But what we are going to talk about now is the 24 possibility that the accumulator water is unborated. And in 25 that case, what we are assuming is if the water is unborated Heritage Reporting Corporation
() (202) 628-4888
151 l 1 that you go directly to a potential reactivity event.
O 2 And the other branch that I want to discuss now in
(/
3 if the refueling water storage tank vator is at a critically 4 low baron concentration, again you would branch to a 5 potential reactivity event. And the end point probabilities 6 are given on here. One is on the order of 10 to the minus 9 7 per reactor year, and the other 10 to the minus 7.
8 And again these conditional probabilities here 9 were obtained from the fault trees which are given in the 10 appendix. So those fault trees have been serving multiple 11 duty i.n that t. hey are related to a number of different 12 scenarios.
13 MR. LEE: Defore you take off this transparency, 14 regarding this particular one involving refueling of the
() 15 storage tank problem wi.th the failure probability of 16 1,6 and 10 to the minus 7.
17 MR. DIAMOND: Well, we want tc look at that a 10 little bit more closely. All of these numbers at this stage 19 are certainly preliminary. And since thav is the largest 20 number that we have come up with so far, we will certainly 21 scrutinize it, as we will with all of the numbers tliat we 22 are generating.
23 MR. LEE: Given that number, have you made a 24 decision regarding the deterministic analysis or transient 25 analysis?
Ileritage Reporting Corporation
(~) (202) 628-4008 RJ
l l
152 l
1 MR. DIAMOND: Again I do not think that we have 1
() 2 made any final judgments. At this stage, our judgment is 3 based on the fact that these calculations, the deterministic -i l
4 calculations, would be very difficult to do.
5 MR. WARD: Is the-project budgeted for any 6- deterministic calculation?
7 MR. DIAMOND: Yes.
8 MR."WARD: All right.
9- MR. DIAMOND: As you can see, there are a lot of 10 sequences that we are going through. .And so we are making _
l 11 judgments as to what would be the most efficient way to 1
12 proceed. Let me point out two othar events on this i 13 particular event tree, which I now want to talk about, which 14 deal with LOCA and with getting diluted water into the core. -
() 15 One of these is if one assumes that there is 16 multiple steam generator tube ruptures as a result of,the 17 LOCA event, then one goes to a core melt situation. This 18 was already on the Byron PRA, but not for the purposes of 19 talking about a reactivity accident.
20 I am sorry, I am getting a little bit ahead of 21 myself. Let me come.back to a couple of other sequences on 22 this chart in a few minutes. !
23 (Continued on next page.)
i endrct 24 )
25 lieritage Reporting Corporation
(} (202) 628-4888
,.+,s , , - , . , ----gn-, ,,--,us--pe , y- w eg 4 g e e,
- - - - - - - . . + .,, ,,. ,
l 153 t/3 1 The next LOCA that I wanted to -- the next LOCA
( 2 where there is a possibility of getting diluted water into 1 3 the core'that I wanted to discuss was a LOCA when the sump 4 water was diluted.
S Since the long term cooling -- post-LOCA-long term' 6 cooling comes from the sump water then there is a question 7 of, can you have a reactivity accident if that water becomes 8 diluted.
i 9 There'are several sources of unborated water which 10 might be postulated to get into the sump. Essentially, ,
11 cooling water, high pressure fire protection water, primary 12 makeup, and component cooling water. And this was an issue 13 that was identified in an LER about a year ago, a little 14 bit -- it was early in '87 by Sequoyah and they went back O(/ 15 and did some probabilistic analysis to see whether this 16 might be a problem.
17 They determined all the various sequences to get 18 into a problem, and there are several sequences involved 19 because of the fact that there are several sources of 20 unborated water. And they concluded that no remedial action 21 was necessary. l 22 In other words, essentially what they said was 23 that getting unborated water into the sump was an incredible 24 event, I think I'm using their word.
25 Now that analysis could be reviewed as part of N
IIeritago tieporting Corporation O (202) 620-4888
1 l
154 1 this project to see whether it is applicable to other 2 plants.
3 However,' it should be noted that if you think 4 about the deterministic side of this question that if the 5 water was diluted you would need, first of all, .a large 6 amount of unborated water in the sump in order to dilute
, 7 what was in the vessel.
8 And the criticality in the vessel would be a slow 9 event. And therefore, it seems like the judgment will be 10 that this type of event is not worth pursuing any further.
11 This kind of event could also be represented on 12 thi; event tree. The last event here le the boron 13 concentration-and the sump water is diluted. And if it-was 14 then there is the possibility of this slow recriticality 15
({) leading to core melt, but again, I think that this is going 16 to be one of those sequences where we can use engineering 17 judgment to eliminate any further analysis.
10 The last event dealing with a LOCA and dilution of 19 the water during a LOCA comes about by assuming 20 consequential steam generator tube ruptures. And then the l
- 21 secondary at higher pressure than'the primary causes l 22 unborated water from the secondary can flow into the vessel.
23 Again, this is a vintage 1987 problem in that 24 there was a series of letters to the NRC concerned about l
25 this problem. So this problem has been addressed by the PRC l l
i l
Heritage Reporting Corporation l (202) 628-4000
[} .
- . .~ , _. . , . _
. m- _ . , , . ._.. _ ._.. - _ . _ _ - - - . _ - . _-, -
1 155 1 durjag this past year, and I believe, dropped as a -- has a
() 2 generic issue or -- I chink that's what it is potentially.
3 Now there are two things that are of concern here.
4 One is the fact that when one has a steam generator tube 5 rupture there is this potential that the water that gets 6 i.nto the primary can slou down the reflood t age and that 7 can lead tv either peak clad -- unacceptable peak clad 8 temperatures; and that's a thermal-hydraulic problem.
9 The problem that we are concerned with is, of 10 course, that this unborated s.ater gets into the core and 11 caures a power excursion.
12 Lie work that was done oier the pa ;t year at NRC 13 shos ad that the core melt frequency would be on the order of 14 10 to the minus 8 per reactor ye r, taking into account the
() 15 thermal-hydraulic effects of the secondary water getting 16 .nto the primary.
17 '
A discussion I had with R. Riggs of HRC said that 18 that work is to be extendert to account for the dilution 19 question. And so I think URC itse]f will be generating a 20 core damage frequency for this number.
21 It should be noted that very large flows would be 22 necessary, In other words, larc numbers of steam generator 23 tubes wou' a- -
ut be ruptured to get sufficient cold water 24 into the c- u And .
did occur, return to power weuld 25 be a slow w A
- M Q6 6g h tie . -
Corporation 1 porting
((} ,.02) 628-4888
'I
_ ~ - - _- - - - - - - -
L I l I l
.156- i
! 1 DR. KERR: What is the meaning of.the statement, j 2 "For' Byron PRA any steam generator tube rupture leads to-3 core damsge?"
4 MR. DIAMOND- The Byron PRA already had on.it a
'S branch where there were multiple steam generator tube 6 ruptures. And this branch--- assuming that --
l 7 .DR. KERR: I was curious about the meaning of the 8- term "leads to." Do you mean there is a path through which 9 one could get to, you don't mean it is automatic.
10 MR. DIAMOND: Yes. Well,_according-to this event 11 tree.
12 DR. KERE: Well, sometimes the dictionary meaning-l 13 of "leeds to" means that if this happens, this happens.
14 That is not what you mean, I think.
() 15 MR. DIAMOND: No. I meant that the Byron 16 analysis --
17 Ca. KERR: If one could follow through a sequence 18 of events which the certain probability produces core ;
19 damage.
20 MR. DIAMOND: Yes.
21 DR. KERR: Okay, that's all I need to know.
22 MR. DIAMOND: They just gave a simple branch here 23 where there were any 3 team generator tube ruptures it 24 would lead to core melt.
25 The sixth category-is the set of sequences that Heritage Reporting Corporation (202) 628-4888 i
)
157 1 involve a startup of a reactor coolant pump _after an j }_ 2 improper dilution. So now we are no longer talking about 3 accumulator and refueling water storage tank diluted and 4 causing a problem. Now we are' talking about a different 5 type of sequence. And t here are three that we are currently 6 looking into.
7 One is that you postulate that you lose your pumps 8 during the dilution process. And this occurs when there is 9 no decay heat available to cause mixing, so the dilution 10 continues and the diluted water accumulates and then you 11 restart your pump and ge' a slug of diluted water coming 12 into the core.
13 This might also be postulated to occur if you had 14 some kind of misoperation of loop stop values.
() 15 Finally there is a sequence-that we need_to 16 investigate which we have not_done anything with yet-but are 17 aware of it, and it is a dilution due to a boiling i 18 condensation sequence where the pump't are -- where you lose
> 19 the pumps and as a result the power coolant mismatch does 20 cause some boiling and that steam condenses in some other 21 t. ct of the primary system and that condensate has a lower 22 horon concentration which then gets back to the coro.
l 23 The reason that I'm aware of this particular 24 situation, although I really don't -- have not studied it 25 yet, is thet it is a situation that is being addressed by ,
Heritage Reporting Corporation (202) 628-4888 O 4
~
158 1 the French NRC or NRC equivalent.
,/
k-) 2 And as a matter of fact, both sequence A and 3 sequence C are being addressed by the French with a system, 4 primary system -- thermal-hydraulic calculations.
5 The preliminary results that I have heard of 6 indicate that this will not be considered a serious problem; 7 and therefore at this time we are not considering doing any 8 probabilistic analysis until we see the results of this 9 deterministic analysis for itepis A and C.
10 And item B was supposedly addressed by 11 Westinghouse; we have not examined that analysis yet, but 12 that is obviously something that will have to be done.
13 MR. LEE: This particular type of accident could 14 have larger consequences then some of the others we have
\
t~)m k- 15 been looking at? Accumulator dilution and so on. 1 16 MR. DIAMOND: No, I think these are all similar in i
17 that you -- theoretically you think of a slug of diluted I l
18 water coming in and displacing the borated -- the highly i 1
19 borated water that is in there. '
20 What will actually occur, of course, depends on 21 the particular event and the thermal-hydraulics of the l l
22 system. They are similar in some respects, but the 23 thermal-hydraulics will really determine whether one is 24 inconsequential or another one might indeed be a very 25 serious excursion.
gj
(_
3 Heritage Reporting Corporation (202) 628-4888
a 159
- 1 MR. LEE: I guess the reason for the question was 2 to understand again your decisionmaking process a little bit
{~s')
u 3 better. So you are not sure whether this particular one, 4 let's say, scenario A potentially had a larger consequence 5 or not, at this stage, compared with the leaky valve 6 situation.
7 MR. DIAMOND: Yes. At this stage I can only say, O based on what I have heard f rom the l'rench, that it will not 9 lead to severe consequences. But that's, you know, at this 10 stage -- I mean, there is nothing written that I can point 11 to.
12 Now we have gone through in a little bit of detail 13 six of these categories. The last category here is -- there 14 is no slide for this. This is an extension of normal buron
/~'- 15 dilution event. Dilution from the chemical volume control O
16 system. And the extensions might be with pocitive moderator 17 temperature coefficient with inadequate scram and/or with no 10 operator actions.
19 I'm not sure yet what we will do here, but any of 20 these faults vouJd lead to a very slow power increase. And 21 from that point of view it may not be of much interest.
22 For example, if you have an uncontrolled boron 23 dilution during refueling, and I think it takes somethina 24 like two hours before you achieve criticality and some sort '
25 or core thermal problem. So the fact that you might have a lleritage Reporting Corporation r^s (202) 620-4000
\
v) .
160 1 positive MTC might hasten that, but in general I think
/~S V 2 events in this category will not be of much interest to us 3 in this project.
4 Now I would like to move on to another class of 5 events which I am calling "beyond design basis events," and 6 there are four here and I will start to talk about the first 7 one of these which is the rod -- the PRW rod ejection i 1
1 8 accident. l l
)
9 There are two sequences that may be of interest '
10 here. One is the accident caused by a rod which has an 11 excessive worth to either tech spec violations or to the 12 occurrence of a special physics test.
13 Thn second type of sequence involves multiple 14 ejections.
O k/ 15 As far as probabilistic analysis goes, none ic 16 currently planned for either one of these. Of course, with 17 regard to multiple ejections this involves a determining 18 mechanical failure rates rather than determining -- I'm 19 sorry, it involves determining mechanical failure rates and 20 components that don't usually fail. It makes it a 21 particularly difflcult problem.
22 Some thought should be given to common causes.
l l 23 For example, a heavy load dropping on through the missile l
l 24 shield and on to the vessel head.
25 As far as deterministic analysis goes t .v should
() lieritage Iteporting Corporation (202) 628-4888
_----------__---J
161 1 be sufficient existing parametric studies which will give us 2 an answer to what kind of rod worth is necessary in order to
('])
'w.
3 have an excessive enthalpy; and then we would look at 4 existing physic calculations to see if one could get that 5 type of rod worth. In other words, a rod worth which gave 6 you an excessive enthalpy rise.
7 Part of this will be to, also to look particularly 8 at the room where people are doing physics tests and the 9 core configuration is in a non-normal state and rod worth 10 may be different than what would be calculated under normal 11 conditions.
12 DR. KERR: What is meant by saying, ncn-currently 13 planned but some thought should be given to common causes?
14 MR. DIAMOND: Yes.
15 DR. KERR: llow does one give some thought to
(']'
16 common causes?
17 MR. DIAMOND: What I mean by that is, for example, 18 the heavy load problem. I wonder if that has already been 19 addressed by MRC and whather people have already looked at 20 what kind of heavy loads would penetrate into the stessel j 1
21 head, and what kind of restrictions there are on moving )
22 heavy loads during operation.
1 23 DR. KERR: That's just a reminder to you to give 24 some further thought to it then?
25 MR. DIAMOND: Yes. And the reason for this is lleritage Reporting 1:orporation (202) 628-4888 G,r3
)
l l
162 1 that our final report, we would like to be able to not only
./~'N 2 say what we have done but to give information relevant tc L) 3 this general question that has been gleaned from all 4 different sources.
5 The second type of beyond design basis event are 6 steamline break events. One could postulate that instead of 7 one steamline breaking, one have two steamline breakings, or 8 that there was inadequate shutdown margin because rods were 9 -- because tech specs on rod insertion limits were violated 10 or seme other reason.
11 Thirdly, this is not really a sequence here but it 12 is merely a note, that you may have to consider other than 13 steamline breaks as the worse overcooling event to --
14 relevant to a reactivity excursion. This is not well 15 thought out at the moment, so let me juet concentrate on A (V^'s 16 and B at the moment.
17 With regard to the first sequence here, two i
18 steamlines involved, ap.in I want to refer to some French l l
19 results. Their calculations showed that if you did a best 20 estimate calculation with two steamlines broken at the same l 21 time that the overcooling would be a less severe situatien 22 than when you assumed one steamline broke and you did a 23 standard licensing analysis where you used very cons'ervative 24 assumptions.
25 So in other words, there are best estimate with fleritage Reporting Corporotion 3 (202) 628-4888 d
163 l
.1 two steamlines involved gave less severe consequences than 2 .their licensing calculation with one steamline involved. So 3 it doesn't look -- aside from addressing the question of 4 what the probability is, it doesn't look like this would he 5 an important question. And again, this is-.a return to 6 criticality, not a rapid power excursion.
-? And item B, of course, is also not a rapid 8 reactivity excursion inadequate shutdown margin, but we may 9 give some additional thought to it at this stage. I don't 10 know what we will do.
t/4 11 Now there are two other PWR events on this chart, 12 bank withdrawal which is normally analysis. But now if we 13 assume that we have a positive MTC or inadequate scram the ,
14 event could become more severe. But.again, it would be.more .i 15 severe in a different way than is of interest to us; it 16 would not lead to a rapid excursion, but only to a slow 17 power rise.
18 Similarly, thermal-hydraulic transients in which 19 one has a positive MTC to deal with are seen as events which -
20 may lead to a power cooling mismatch and, indeed, might even i 21 lead to a severe accident but these would generally be slow !
22 events.
23 We will look at these in more do?all than I am
? l
- 24 givino you just now, but at the moment they do not look like i 25 they will be the events that we emphesize.
i l
lieritage Reporting Corporation O- (202) 628-4888 1 I
_ . . . . _ . _ . -._._ . ..-. ..I
- .. . . . . ~. --
164 1 DR. KERR: Would that be anything based on the-() 2 likelihood, the small likelihood of being then postured-3 moderator temperature coefficient or'do you think that it 4 won't-need to be looked at?
5 MR.-DIAMOND: No, we will look at these,-but what 6 I'm saying is that the -- everything that we have seen so 7 far suggests that, number one, that these are not rapid a
8 events that we should concern ourselves about. And some of
' ~
9 them even -- may not even be of much interest in terms of 10 consequences.
11 We will look at --
for example, there have been 12 some calculations here that relate to testing conditions.
13 If one does some natural circulation tests in a PWR where 14 the scram. system is partially disabled and one gets from
() 15 that condition, then gets into a loss of feedwater, 16 initiating event.
17 There was a question as to whether that could lead I t
1 18 to a serious -- whether that could lead to serious 19 consequences. The reason I mention that particular event is i 20 that I know some work was done for EPRI on that sequence.
21 As I say, in general we tend to look in a little 22 bit more depth here, but I don't think it will be an area I
23 where we will find too much. 1 24 Now we can get back to a question which arose this 25 morning on some BWR events. And there are seven categories, lleritage Reporting Corporation 4
() (202) 628-4888
165 1 some of which I will discuss in some detall.
2 Firstly, there-is the BWR rod ejection accident 3 which has never been considered seriously in the past. It 4 had been, of course, considered at one time. The approach 5 here will be to do some deterministic analysis'by looking at
~
6 existing rod drop calculations as a function of rod speed 7 because rod ejection accident in the BWR is essentially a 8 very fast rod drop accident in a BWR.
9 .This should be suitable for attaining estimates of 10 consequences; and we will see where that leads.
11 MR. LEE: llow much of a difference in rod speed 12 are we talking about, between rod ejection and rod drop?
13 MR. DIAMOND: Oh, it would be quite a large 14 difference. Somewhere the rod drop accidents occurs'on the 15 order of, I don't know, a second or a couple of seconds.
16 How fest a rod would be ejected from a BWR, I'm not sure. I 17 guess it would be slower than a PWR. It would certainly 18 be smaller than the rod drops.
19 MR. LEE: So it should be shorter than two seconds 20 time frame?
21 MR. DIAMOND: Yes.
22 MR. LEE: Do you think there was a substantial 23 difference in the consequences due to this difference in the 24 rod speed?
25 MR. DIAMOND: There may not be, and therefore, lleritage Reporting Corporation O (202) 628-4888
'l l
166 1 this may be of no more consequence than the rod drop
() 2 accident. !
3 With regard to rod drop accidents there had been 4 some -- the concern, of course, is that-the rod.that is 5 dropping out would have excessive worth. This is the most 6 likely reason for having a rod drop with consequences --
7 with unacceptable consequences.
1 8 There had been some probabilistic analysis done_by 9 Howard Richings a lot time ago which showed very~ extremely 10 low probabilities for exceeding the design basis accident-11 limits; and this should be checked to see whether it is i
12 still relevant. !
13 MR. LEE: Also Brookhaven performs on BWR rod drop ,
14 analyses a few years ago.
() 15 MR. DIAMOND: Right. And those existing 16 parametric studies can be used to estimate the potential 1
1 17 consequences and see whether there can be -- to see what 18 type of rod worth one would need to exceed current limits, 19 and whether current rod w th were in that range.
20 So we will do some of those here; some i
21 deterministic and_some probabilistic work.
22 Now we get into the overpressurization events.
l 23 The concern here, of course, is that with overpressurization i l
24 you collapse voids, and that is a reactivity insertion. And i i
25 the concern is that the design basis overpressurization i !
i Heritage Reporting Corporation
()
1 l (202) 628-4888 i
167 1 events don't get you into trouble. Your consequences are
/]
(/ 2 always within acceptable limits.
3 But if we postulate that perhaps we will have no 4 recirc pump trip and/or perhaps that there will be safety 5 and relief valve failures and/or perhaps there will be no 6 scram or scram will be delayed, then perhaps one could 7 exceed certain limits.
8 And the types of limits that we were most 9 interested in were the limits that apply to rapid reactivity 10 events. So we were interested in seeing whether the power 11 excursion would lead to energy deposition in the fuel that 12 was excessive. And therefore we were interested in the 13 short term behavior of these events. And that's why earlier 14 I said we were not involved in what would obviously be the
/~T
(_/ 15 severe accident which would occur during the long term part 16 of some of these scenarios.
17 So we used the BNL plan analyzer. Here is an 10 example where a thermal-hydraulic calculation is very simple 19 to carry out. And we calculated a couple of 20 overpressurization events, one where turbine trip without 21 bypass was the initiator and one with MSIV-closure, and we 22 assumed failure of these systems and found that the energy 23 addition in the first 10 seconds was less than 14 full power 24 seconds. It we assume that all went into the fuel, then 25 that would correspona to an average enthalpy rise of 60 Heritage Reporting Corporation
()
7s (202) 628-4808
l I
160 1 calories per gram. l
(]) 2 And if you take on a peaking factor to that, then 3' the peak fuel enthalpy would be higher than that. But as. l 4 I'm sure you are aware it's 280 calories per gram where the 5 energy deposition is of concern.
6 Now,-the -- ,
7 MR. LEE: Excuse me, but this also means that 8 during that peak of the power, average power could having.
9 two times, at least, rated power?
10 MR. DIAMOND: Yes. Let me show these traces.
11 This one was for the turbine trip without bypass, without !
12 scram, without recirc pump trip, and with one oank of reliet j j
13 valves inoperative. '
14 This is the relative power versus time for the '
() 15 first 10 seconds. And this is the pressure during the first
~
i 16 10 seconds. Normally, if you do have the reactor protection !
1 17 system we get a little bit smaller rise, and then the l I
18 turnaround would be reactor power going down to decay heat 19 levels as reactor trip became felt.
20 What happens here, of course, is that the power ,
i 21 hangs up at about 1.5 times nominal power. And most of the 22 energy contributed -- most of the energy during this first 23 10 seconds then comes from this region in here rather than
- I 24 from this initial power pulse.
25 Now, o f. course, this is still -- this says nothing lieritage Reporting Corporation
(} (202) 628-4888
169 1 about what happens after 10 seconds where you still.have ,
() 2 this power which must be dissipated in some heat sink, and 3 also says nothing about this pressure which is going up, 4 almost up to 100 bar at the end of 10 seconds, and 5 obviously, depending on how many relief valves you assume i
1 6 are inoperative you're starting to get'into the pressure 7 limits for the vessel. But that was not of interest in F
8 these calculations.
9 MR.-LEE: This calculation has been done with 10 dimensional --
11 MR. DIAMOllD: Yes. Just a simple -- it's a 12 BWR-4 model and naturally one could do parametric studies 13 and vary all hinds of the reactor characteristics. For 14 example, void coefficient, th!.ngs like that.
() 15 The most important thing is that we assume here 16 that cne does not have recirc pump trip, one does not have l
17 relief, but one has limited relief, and yet, the energy 18 addition here in those first 10 seconds is not what we would
, 19 consider to be significant from a reactivity incertion point 20 of view. From a reactivity accident point of view.
21 This is the same transient but we plotted minimum 22 CPR, which is this curve, and core average fuel. temperature.
1 23 And again, although as we go towards 10 seconds the minimum 1
l 24 CPR is getting close to one which might indicate fuel clad j 25 problems. And the core average fuel temperature is still j l
l
- lleritage Reporting Corporation
, () (202) 628-4888 i
I 170:
1 rising. The important thing is that, in this first-time
) 2 interval here we would not. expect sufficient energy J. deposition to cause any kind of fuel fragmentation or a 4 severe accident.
5 I have included in the handout the same two types 6 of charts where the lnitiating event is an MSIV-closure.
There is probably no-need for me to put those up on the 7
8 screen; they show similar results to the situation for the 9 turbine trip without bypass.
10 The last or almost last, the. fourth BWR event is 11 the -- comes about when the core is operated in a region of 12 instability. When one has operation in the low power / low 13 flow r' .. then there is this potential that the core will 14 be unstable. This is the situation tnat was observed at 15 LaSalle.and also at other plants.
16 And one has diverging oscillation; and the concern 17 is one of two things, either -- if you don't have trip, this 18 is an ATWS event, then the power will continue to increase 19 without limit until something catastrophic happens or if one 20 does have trip but prior to that trip the frequency
- j. 21 amplitude increases so rapidly that the power excursion 22 which occurs before trip adds energy to the system which 23 causes fuel fragmentation or some excessive fuel enthalpy 24 rise: then that is of interest.
25 We are not planning on doing any probabilistic IIeritage Reporting Corporation
() (202) 628-4888 1
I l
II 171- )
i 1 analysis at the moment, but there is,some work being done j
) 2 with computer code calculations. This was alluded to 3 earlier, there is an RES project which is being put in place 4 at Brookhaven, also, which we'll.use the 5 Ramona-3B code. The Ramona-3B code has a three dimensional 6 representation in the core, and the expectation is that this 7 will-be a good representation and a necessary representation 8 to understand these instabilities. -,
9 DR. KERR: Do you have any reason to believe7that 10 Ramona can model this?
11 MR. DIAMOND: We are hoping that it will; yes.
12 DR. KERR: But other than hope you don't.
13 MR. MINNERS: We have asked them to try to do some 14 validation studies. There are some experiments. .
15 DR. KERR: I wasn't being critical, I just 16 wondered if on the basis of something or other there is ,
17 reason to believe that the code will do it or you aren't 18 sure or what.
19 MR. DIAMOND: The only reason I would hesitate is 20 because these calculations havo not been done in the past.
21 It is expected that the code should be able to calculate 22 these types of oscillations. I mean, it calculates many
- 23 other --
24 DR. KERR: Because it's a good Brookhaven code 1
4 25 that gives you a warm feeling or because it's peculiarly IIeritage Reporting Corporation
() (202) 628-4888
4 112 1 suited to this kind of thing. I mean, I asked because you i
. () 2 know better than I, a lot of codes have been thought to be 3 capable of calculating things and they haven't.
4 MR. DIAMOND: That's true. That is why I did not 5 jump in and say, absolutely yes. But I see no reason why it 6 can't, based on the fact that it has calculated oscillatory 7 behavior for ATWS calculations; I would imagine.it would be 0 able to calculate these types of oscillations generated by 3 being in a low power region.
10 MR. RICHINGS: Howard Richings, NRC. Other ccdes 11 of a similar nature have been used by the Japanese to 12 calculate a number of actual experiment and/or occurrences 13 of these type of oscillations in various foreign reactors.
14 So at least from that experience we know that it can be
() 15 done, and under the conditions on which they occurred.
16 I want to say one thing, Dave, I don't think these 17 should be noted as necessarily leading to diverging 18 oscillation; these oscillations, as best we know them at the 19 moment, or . limits like oscillation, they don't change in 20 amplitude unless some characteristic of the reactor changes 21 in emplitude. And that's what was occurring at LaSalle, 22 some things like flow was slowing changing, limits like 21 oscillation big enough to scram.
24 MR. DIAMOND: Right. But there is always a high 25 probability th- there le going to be some reactivity Heritage Reporting Corporation
() (202) 628-4888 i . .
173 1 adiition mechanism while one is in these limit cycles which A
tj 2 will -- which has the potential to make them divergent.
3 MR. LIPINSKI: Well, unfortunately LaSalle scram 4 took place, but that recire pump trip runback was put in 5 there to plant the scram You could feel the scram going 6 into recirc pump trip and go into diverging oscillctions, 7 then you have a problem.
O MR. MINNERS: You don't get diverging 9 oscillationc.
10 MR. RICHINGS: You may or may not have diverging 11 oscillations depending what the system conditions go to.
12 MR. LIPINSKI: But his point was that other 13 conditions in the system did diverge and finally offered a 14 trip. But if the trip does not take place when you run
(~/
q_ 15 back, then you can get into these divecging oscillations, 16 then you could have a problem.
17 MR. RICHINGS: My point was that, generally you i 10 don't expect diverging escillations as said; you expect the 1
l 19 plant to go to some equjlibrium conditions and there will be '
20 oscillations of sudden magnitude for those conditions.
21 LaSalle was slightly pulling off as time went by 22 and flow was changin.g as time went by; and therefore the l 23 oscilletions change in amplitude a little bit.
24 MR. LIPINSKI: But we never sat here long enough 25 because the trip took place to find out whother it was going 11eritage Reporting carporation
() (202) 618-4000
r . - - . - - ~ ..
'174 i
1 .to go to an equilibrium oscillation, limit cycle.
() 2 MR. RICHINGS: True.
3 MR. LIPINSKI: So you need a good model to ,
4 determine where you would have been. The other question is, 5 how you validate your 3D core model because the data that- ,
6 came from LaSalle didn't give you special information; it 7 was only point information.
8 MR. RICHINGS: There was information that the core 9 was oscillating basically uniformly from the APRM readings.
10 MR. LIPINSKI: But I understand those APRM 11 readings were delayed in time; they were filtered.
12 MR. RICHINGS: Not the APRM readings, not just 13 start track readings.
14 MR. LIPINSKI: The start track readings gave you-() 15 3D special information?
16 MR. RICHINGS: Those were not 3D. What I'm saying 17 is, the information on reading -- people reading the APRM 18 were basically that the APRMs were in phase with the APRMs ;
19 and of equal magnitude,. that they were all reading about the 20 sarae magnitude as you would expect from everything being in 21 phase.
22 If things were not in phase the APRMs would have 23 read significantly different from the LPFMs.
24 MR. LIPINSKI: So all channe19 were behaving 25 coherently. ,
-l i
I 1
lieritage Reporting Corporation (202) 628-4888 4
I l
175 1 MR. RILHINGS: Yes, they were behaving coherently.
- c. 1
(-) 2 As you expect these oscillations to do so unless there is 3 some asymmetry in the core.
4 MR. MINNERS: I think Dave's point is that, the 5 first problem that you probably get into is DMB, okay. High 6 reactivity, enthalpy conditions are probably not a problem.
7 DMB to me is an economic problem.
8 MR. DIAMOND: This laat item here, we have 9 somebody from the finished NRC who is coming to work with us 10 on this proble*a as well . It is a concern in a number of 11 different countries.
12 DR. KERR: Is there something significant about 13 the fact that the person is finished or just that you are 14 getting some outside help?
(m
(- 15 MR. DIIMOND: Significant? No, only that this is 16 a question which is being considered in many other 17 locations, as opposed to some of the other events that I 18 have gone through which have not been a concern elsewhere.
19 Item number five, a multiple rod withdrawal event, 20 again, the potential for an accident if one postulates 21 enough failures. But again, an accident that would be a 22 slow power rise, and therefore it is of less interest in 23 this project.
24 The general question of transients with either 25 inadequate or no scram or this question of flushing of boron gs Heritage Reporting Corporation
(_,) (202) 628-4888
176 1 during an ATWS, again, are -- seem to me to be subsumed by 2 the general ATWS question and the work that has already been 3 done within ATWS. And it is not clear if this project 4 should do more work in this area.
5 DR. KERR Your view-that the seventh item there 6 has been analyzed carefully in the ATWS study?
7 MR. DIAMOND: Item seven?
8 DR. KERR: Yes. t 9 MR. DI AMO!!D: No, I don't think it has .
I 10 DR. KERR: It has only been subsumed then, it 11 hasn't been studied?
12 MR. DIAMOND: Yes, I don't think it has been 13 addressed because we, at one time, Brookhaven was attempting 14 to do some deterministic calculations and unsuccessfully. So 15 I know it is a problem that is still'around.
16 DR. KERR: And that is the reason you don't think 17 it is worth looking at or did I misunderstand you?
18 MR. DIAMOND: No, no. Only that it is -- we say 19 that it falls in ATWS and if the ATWS question is separate 20 from the reactivity accident question, then maybe we 21 shouldn't look at it. I don't know, this is something that 22 hasn't been definitely decided.
23 MR. MINNERS: You are running up the problem we.-
24 always had with the ACRS, if-they don't like it we narrow 25 the scope. They're always adding -- they want it bigger Heritage Reporting Corporation O (202) 628-4888
-I 177 1 scope than we give. We're just keeping this-scope on
() 2 reactivity accidents analogous to Chernobyl. We are not 3- looking back at every possible accident that can happen. 1 4 It's a limited scope connected to Chernobyl-and 5 reasonableness.
I 6 MR. LIPINSKI: Was there a separate score sheet 7 that's being generated on analyzed events in BWR?
8 MR. MINNERS: Yes. That's right. If anybody has 9 got a problem or they think is a problem, you send'it to Tom 10 King and he will evaluate it whether we should restudy. If l 3 ;
11 we find new information that's how we evaluate whether we 12 ought to put significant resources in evaluating. I 13 MR. L1PINSKI: Now number seven doesn't get ,
14 analyzed as part of this effort, does it get sent to tameone !
() 15 else to look at?
16 MR. MINNERS: I don't think anybody is looking at 17 that now. !
18 MR. LIPINSKI: It will fall into a crack. [
3 19 MR. MINNERS: Well, no. It was considered during 20 the ATWS considerations. The only question is, should it bo ;
1 21 considered again. I don't call that falling into a crack.
22 MR. LIPINSKI: But your comment was that you 23 didn't treat it because you weren't able to bandle the
]
24 problem; is that correct? !
25 MR. DIAMOND: We didn't treat it with Heritage Reporting Corporation (202) 628-4888 i
1 1
170 1 deterministle calculations. But I think, so has been 2 recognized at NRC, that the flushing of the boron during an 3 ATWS could lead to some problem.
4 MR. LIPINSKI: Well, at this point I don't 5 understand what treatment it received then.
6 MR. MINNERS: The ATWS has certain requirements 7 and we feel that those are adequate to keep the risk at a O reasonable level. And we don't require any particular 9 additional methods to prevent flushing the boron from the 10 system. That was the decision that was made.
11 MR. LIPINSKI: But the ATWS fix assume that you 12 borated the water.
13 MR. MINNERS: Yes.
14 MR. LIPINSKI: But now there's a subsequent (3
s_/ 15 sequence that says you can unborate the water before you get 16 those rode in.
17 MR. MINNERS: llow are you going to do that?
10 MR. LIPINSKI: Well, that's what we're asking him.
19 He has identified that as a possible reactivity addition.
20 DR. KERR: You unborate the water without having 21 some boron carryover, don't you.
22 MR. MINNERS: That sequence was considered during 23 ATW5 24 DR. KERR: It was analyzed.
25 MR. MINNERS: No. You reserve analyze for doing a lioritago Reporting Corporation fs) v (202) 628-4888 l
)
- = .- . .
179 1- computer code or numerical tabulation code, but analyzed,'it
() '
2 was cons'idered. Whether that was sufficient or not, I don't 3 know, but that's the basis that the ATWS decision was made.
, 4 on, on that kind of a consideration.
5 'Ncne if somebody thinks that separation was 6 inadequate --
7 DR. KERR: No, I don't know it was not --
8 MR. MINN3RS: I don't.either.
-- but I look at those rod ejection 9 DR. KERR:
10 accident to-BWRs has,certainay been considered before, and 11 the rod drop accident has been considered before, power 12 excursion due to operation of regions of instability has ;
13 been considered before.
14 MR. MINNERS: I guess we haven't gotten across to
() 15 you the purpose of this exercise, you take the Chernobyl 16 experience which is basically a reactivity transient, a fast 17 transient in which you get core damage, and see if we have 18 any of those we have overlooked, okay.
19 It is not to go back and look at every possible ;
20 accident sequence. l l
21 DR. KERR But I had assumed that --
22 MR. WARD: Well, why don't you classify seven as 23 one of those instead of --
24 MR. MINNERS: It's not a fast transient.
25 DR. FERRs I had assumed that you were going to IIerit age Reporting Corporation (202) 628-4888
- . . 3 ._,,.--,_<4 --.,-,,ye.y-- ,y, . - , , _
100 1 1cok at accidents that could involve unexpected or perhaps (s.
\_) 2 rapid insertion of reactivity. And I also would have 3 thought that if Chernobyl reminded you of some things that 4 may be in retrospect, needed additional attention, if you 5 would, and ignore that.
6 MR. MINilERS: Well, I guess we're drawing a line 7 at a different point.
8 DR. KERR: I think that's clear.
9 MR. MIllllERS : I mean the line has to be drawn 10 somewhere and you're suggesting to drawing it a little 11 differently.
12 DR. KERR: I'm trying to find out where you are 13 drawing the line and why, at this point, I have not made a 14 decision as to whether you were drawing it at the wrong 15 place or not. I'm just trying to find out --
16 MR. MINNERS: I guess we have drawn it between six 17 and seven.
18 DR. KERR: And you haven't drawn it, it seems to 19 me, on the basis of any safety considerations; just on the 20 basis of whether it has to do with Chernobyl or not.
21 MR. WARD: 11 0 , on whether it's fast or not.
22 MR. MINNERS: Well, I think that's correct. If 23 you had come up to me and, because of Chernobyl and people 24 had started to think positive, and suddenly you discovered 25 an accident which hadn't been even considered at all and had ileritage Reporting Corporation dgs (202) 628-4888
( __ _ _
. . -= - -
181- l 1 significant consequences but wasn't directed to,Chernobyl,-
2 yes, I think people would start to move.
3 But the safety consideration was.-- sequence seven ,
4 up there had been considered. So we don't think that they 5 -- once again, it's a matter of judgment.
6 DR. KERR: But why the safety consideration of 7 number one bas also been considered'and in a lot more detail 8 than number seven.
9 MR. MINNERS: -But number one is much more closely 10 related to Chernobyl than number seven.
11 DR. KERR: But it has also been studied a lot 12 more.
13 MR. MINNERS: Fine. That's how we drew our line. I 14 DR. KERR: So you prefer to -- j 15 MR. MINNERS: We're trying to explain to you where 16 we drew our line; do you disagree with that.
17 DR. KERR: No, I'm trying to understand why that 18 you look at an accident, possible accident and it has been 19 studied a lot more and is probably much better understood I 20 than one that has not been studied very much, and therefore i 21 is not very well understood.
22 MR. MINNERS: I'm not sure I would agree that the 23 rod ejection accident has been studied a lot.
24 DR. KERR: Well, it has been analyzed in a good
^
25 bit of detail, I certainly believe.
l i
IIeritage Reporting Corporation O (202) 628-4888 1
- - ~
,.,,.,--,--.,,.,-,-,n .---c. e..nm,----__,--.w.n.---,,,, , , - . . , , , . . -,;
102 1 MR. MINNERS: I'm not the expert, but he seems to
(',N) 2 be making the implication that the rod ejection accident 3 doesn't have big consequences, and I'm surprised to hear 4 that.
5 MR. DIAMOND: In the past I think it was 6 considered an incredible event in a BWR and not considered.
7 MR. MINNERS: I don't kncw of any analyses of a 8 rod ejection accident in a BWR, So I don't think it has 9 been studied in the past.
t/5 10 MR. LIPINSKI: What is your definition of a last 11 transient, is that where you are drawing the line, one 12 second?
13 MR. IIINNERS : We are talking about something that 14 can put enough to the rod that can damage the core.
(~.
(_) 15 MR. LIPINSKI: Over what period of time?
16 MR. MINNERS: I'm not talking about a core melt, 17 when you slowly come up and just melt the thing down. We're 10 talking about something -- I don't know, I guess I don't 19 have a quantitative definition.
20 I think what he has explained to you, to drive a 21 quantitative definition from that. Something in a few 22 seconds.
l 23 MR. WARD: Dave.
24 MR. DIAMOND: To summarize, we are looking at l
25 several different reactivity accidents in this project. The lleritage Reporting Corporation ,
(~'; (202) 620-4000 l u-l
- p r 183 1 efforts to date have concentrated.on doing probabilistic
() 2 analysis for.the various PWR boron dilution events'and doing
. 3. deterministic analysis for BWR overpressurization events, t
. 4 Primary results show previous judgments on
~
1
.5 reactivity accidents sufficient, but we haven't finished the 6 job yet. Work is continuing. I would say that the three 7 areas where we~ expect to do considerably more' work are, one, 8 looking at sequences that add diluted refueling water.
9 storage tank water to the vessel. The other is looking at 10 beyond design basis, PWR rod ejection accidents. And the 11 third area is BWR events caused by instabilities.
12 A report will be written, it is due December the 13 let and we will document BNL work and essentially the state.
14 of the art with regard to reactivity accidents.
() 15 DR. KERR We have been told, I think, that the 16 LaSalle incident is being studied in some detail and that 17 there will be a report due out in October or September . Is 18 this part of that or is this a separate study. -
19 MR. DI AMOND,: No, separate.
l 20 DR. KERR: This is a separate study, okay.
21 HR. WARD: Who is doing that?
i 22 DR. KERR: It is being done by HRR, I think. !
23 MR. MINNERS: Brookhaven is doing it.
24 DR. KERR: Brookhaven is doing it.
'25 MR. MINNERS: Research is paying for it because l
lieritage Reporting Corporation l
[} (202) 628-4888 j 1
--+ * %.- w m - et-y--- 7g' y, y -
yem$w-M-My-v-,+,-q q,r--w-ggygv- -- ye -
-gr*w-W *e+4m--+wg-ve+--w-W #9aM* we g- t 4wie-g- nvyi
'164 1 NRR asked us to.
- i) 2 DR. KERR: Dut_it is not this. This is a separate 3 study.
d' 4 MR. MINNERS: What is this?
5 MR. DIAMOND: Yes.
6 DR. KERR: This study of BWR instabilities like 7 LaSalle.
8 MR. MINNERS: No, that's the same study.
9 MR. DIAMOND: Well, it's this same study, let me 10 explain. This question of BWR instabilities'_was on our list 11 from the outset, way back when. It was always something 13 that we felt was germane to reactivity accidents.
13 LaSalle didn't come along, I guess, until March 14 and then it became a greater interest because of the general
() 15 question of BWR operations.
i 16 So now there is a separate RES project that is-17 looking at the question by doing sorae thermal-hydraulic 18 neutronic coupled calculations.
19 DR. KERR: What you referred to here is the study I
20 we heard earlier that would be -- okay. So this is going to 21 be, to some extent, pulled out of Chernobyl and be written 22 up as a LaSalle --
i 23 MR. MINNERS: Yes. If we had it happen at one of 24 our reactors, how we would treat it.
25 DR. KERR: I was just curious. l t
i Heritage Reporting Corporation
() (202) 628-4888 i
i 1
105 1 MR. LIPINSRI: How many channels do you simulate
/~'
(;s 2 in your 3D simulations for channel coherence?
3 MR. DIAMOND: Well, any number, but generally --
4 MR. LIPINSKI: There's a limit, you're certainly 5 not going to have 100.
6 MR. DIAMOND: Generally they have 191 neutronic 7 channels which represents a quarter -- a quadrant of the 8 core; and maybe on the order of 25 thermal-hydraulic 9 channels.
10 That is all I wanted to present.
11 MR. WARD: All right, thank you very much.
12 (Whereupon, at 3:16 p.m. the meeting was 13 adjourned.)
14
/~S l
's / 15 j 16 17 !
l l
18 19 l
20 1 21 22 23 24 25 lieritage 11eporting Corporation
(-) (202) 628-4888
. . _ . . ~ - - - - .
1 CERTIFICATE 2
(
s* 3 This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the matter of:
5 Name: ACRS Subcommittee on philosophy, Technology & Criteria 6
7 Docket Number:
f 8 Place Washington, D. C.
9 Date: August 4, 1988 10 were held as herein appears, and that this is the original 11 transcript thereof for the file of the United States Nuclear 12 Regulatory Commission taken stenographically by me and, 13 thereafter reduced to typewriting by me or under the direction 14 of the court reporting company, and that the transcript is a 15 true and accurate g! d of th foregoin proceedings.
16 /5/ IV/ (\ fw Aff IRWId'L. FENB 17 (Signature typed):
18 Official Reporter 19 Heritage Reporting Corporation 1
, 20 21 I 22 I 23 24 25 l
O(_, Heritage Reporting Corporation l (202) 628-4888 l l
i
1 CERTIFICATE 3 This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the matter 5 of:
i 6 Name: Implications of the accident at Chernobyl 7
and NRC research plan to address implications 8 Docket NJmber:
9 Place: Washington, D.C. '
10 Date:
August 4, 1988 11 were held as herein appears, and that this is the original 12 transcript thereof for the file of the United States Nuclear 13 Regulatory Commission taken stenographically by me and, 14 thereafter reduced to typewriting by me or under the
() 15 direction of the court reporting company, and that the 16 transcript is a true and accurate record of the foregoing 17 proceedings. r'~)
18 /s/ Oh 19 "" *
(Signature typed):
20 official Reporter 21 Heritage Reporting Corporation 22 23 24 25 Heritage Reporting Corporation
() (202) 628-4888
M\
SD INTRODUCTORY STATEMENT BY THE SAFETY PHILOSOPHY, TECHNOLOGY AND CRITERIA SUBCOMMITTEE - CHAIRMAN REPORT AUGUST 4, 1988 The meeting will now come to order. This is a meeting of the Advisory Committee on Reactor Safeguards Subcommittee on Safety Philosophy, !
Technology and Criteria.
I am D. Ward, Subcommittee Chainnan.
The ACRS Members in attendance are: W. Kerr, C. Siess and C. Wylie.
Also in attendance are ACRS Consultants: J. Lee and W. Lipinski, The purpose of this meeting is to review: (1) the status of NUREG-1251 (Implications of the Accident at Chernobyl) and (2) the NRC research program at BNL to addres:; the implications in regard to severe ;
reactivity transients, f Mr. Dean Houston is the cognizant ACRS Staff Member for this meeting. .
p, I V The rules for participation in today's meeting have been announced as part of the notice of this meeting previously published in the Federal Register on July 22,1988(FR27781).
A transcript of the meeting is being kept and will be made available as ;
stated in the Federal Register Notice. It is requested that each !
speaker first identify himself or herself and speak with sufficient clarity and volume so that he or she can be readily heard, i l
We have received no written comments or requests to make oral statements l from members of the public. l (Chairman's Comments - if any)
We will proceed with the meeting and I call upon Gecrp Sege (RES). l l
l
, O !
i y-- -
v ,- , , . - - . w-m . - .- ,c .
-.s-w. - - - ~ r-, -= . - - -- - - - - -e-r_,