ML20245J132
ML20245J132 | |
Person / Time | |
---|---|
Site: | Limerick ![]() |
Issue date: | 07/27/1989 |
From: | Office of Nuclear Reactor Regulation |
To: | |
Shared Package | |
ML20245J116 | List: |
References | |
NUDOCS 8908170516 | |
Download: ML20245J132 (194) | |
Text
{{#Wiki_filter:_ _ __ - _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _
, . . , ,, 4 0
UNITED STATES NUCLEAR REGULATORY COMMISSION OFFICE OF NUCLEAR REACTOR REGULATION In the Matter of: )
)
LIMERICK SEVERE ACCIDENT ) MITIGATION ISSUES ) O Pages: 1 through 130 Place: Rockville, Maryland Date: July 27, 1989 m.,====== ....================================="""""""""""C HERITAGE REPORTING CORPORATION O opuuReetm 1220 L Street, N.W., Suke 600 Washington, D.C. 20005 (202) 628-4888 a90st7001c. s90 sis PDR ADOCK 05000352 P ppc
F .m si l'. b~ (~)
-- (j UNITED' STATES NUCLEAR REGULATORY COMMISSION . OFFICE OF. NUCLEAR REACTOR' REGULATION In the Matter of: ) ) -LIMERICK SEVERE ACCIDENT )'
MITIGATION ISSUES ) i Thursday. July 27, 1989 j Nuclear Regulatory-Commission i
~ .11555 Rockville Pike Rockville, Maryland' The above-entitled matter came on for hearing, pursuant to notice, at 1:14'p.m.
BEFORE: GENE Y. SUH Project Manager PARTICIPANTS: BRAD HARDIN, RES/DRA RICHARD BAPRETT, NRR R.L. PALLA, NRR E SIDNEX FELD, RES/DRA JIM FULFORDj NUS Corp. A. J. MARIE, PECo JERRY PHILLABAUM, PECo AJIT K. BHATTACHARYYA, PA/ DER /BRP E. S. CHELLIAY, NRC/RES/DSR Heritage Reporting Corporation (202) 628-4888
2-7- N w
,d PARTICIPANTS: .(cont id)
ANN HODGDON,.NRC/OGC TREVOR PRATT, BNL.
- TIM MARGULIES, NRC CHARLES.HINSON,_ RPB/DREP/NRR'/NRC JOHN N. RIDGELY, NRC/RES.
i CHARLES FERRELL, RES/NRC i ROBERT-M. RADER, Conner & Wetterhahn E.-ROBERT SCHMIDT, NUS Corp.
~i R. M. KRICH, PECo '!
1 IL
!O \ ' Heritage Reporting Corporation (202) 628-4888
- _. -_-:_-______ .- . _ - _ - _ - -- -- a
j 1 3
,A~y l Q' ' 1 EBQQREQ1EQR 2 MR. SUH: I'll begin the meeting.'
3- This is a meeting between the NRC staff and the 4 licensee,' Philadelphia Electric Company, held at the' request 5' of the NRC staff, to discuss the recent licensee submittal' 6 with regard to severe accident mitigation design 7 alternatives. 8 As you can see, a transcript is being kept for the 9 meeting,.and~ copies will be made available'and placed in the 10 public' document room. 11 For the purpose of the transcripts, speakers 12 seated at the table will not have to identify themselves to 13 the reporter. I.ask, however, that each speaker-not seated 14 at the. table identify himself or herself and speak loudly { 15 enough so that the reporter will bet able to make an accurate 16 recording.and so that others may hear what's going on. 17 I've passed around two attendanco sheets. Please 18 sign in on one of the two sheets. 19 Before getting t. tarted with the' technical l 20- discussion and to kacilitate the conduct of the meeting, I-21- would like te Lok the people in the room to identify 22 themselves. 23 I'll start with myself and maybe go around the 24 room, . go around the table this way, and then maybe start on 25 the back rost and go around.
- Heritage Reporting Corporation (202) 628-4888
[
- i ~l My name is Gene,Suh. . I'm a Project Manager in NRR 2 with the NRC.
3 MR. HARDIN: I'm Brad Hardin from the Office.of 4 -Research. 5 MR. BARRETT: I'm Richard Barrett. I'm Chief of 6 the Risk Applications Branch, NRR. 7 MR.-PALLA: I'm Bob Palla. I'm'in the Risks 8 Application Branch of NRR. 9 . MR. FELD: I'm Sidney Feld. I'm an economist in 10 -the Office'of'Research. 11 MR..FULFORD: Jim Fulford, an engineer with NUS. 12 MR. MARIE: I'm Al Marie, Supervisor of the Risk 13 Assessment Branch of PECo. 14- MR. PHILLABAUM: Jerry Phillabaum, Philadelphia 15 Electric, Risk Assessment Branch. 16 MR. KRICH: Rod Krich, Philadelphia Electric, 17 Limerick Licensing Branch. 18 MR. SCHMIDT: Bob Schmidt, NUS Corporation. I'm 19 . Manager of PRA Activity. 20 MR, RADER: I'm Robert Rader, an atterney with the H
. 21 . law firm of Connor and Wederhaun, counsel for Philadelphia l 22 Electric.. I'm here as an observer.
23 MR. KLEIM: Dennis Klein frcm Bechtel Corporation. 1
.24 MR. HEARN: Jim Hearn, likewise from Bechtel 25 Corporation.
Heritage Reporting Corporation
'. (202) 628-4888
i I l' 2 5 t' - l L k-), 1 MR. FERRELL: Charles Ferrell, Office of Research, , 2 NRC. 3 MR. RIDGELY: John Ridgely, Office of Research, )
.i 4 Severe Accidents Branch.
5 MR. HINSON: Charles Hinson, Radiation Protection I l 6 Branch, NRR. 7 MR. MARGULIES: Tim Margulies, Risk Application 8 Branch, NRC. 9 MR. PRATT: Trevor Pratt, Brookhaven. 10 MS. HODGDON: ~i'm Ann Hodgdon, Senior Trial 11 Attorney in the Office of General Counsel for Limerick, 12 MR. CHELLIAY: E.S. Chelling, NRC. 13 MR. BEATTACHARYYA: I'm Ajit Bhattacharyya, (') 14 Pennsylvania Electric, Radiation Protection. 15 MR. SUH: Okay. Thank you. 16 It seems like there is a representative from 17 Pennsylvania. Let me just make the statement. For the 10 membera of the public in attendance, this is a meeting 19 between the NRC staff and the licensee. For those members 20 of the public in attendance, does anycrie fec3 they need to , 21 make a statement at this tracting? 22 (No response. ) 23 MR. SUH: Okay. I didn't hear any comments. 24 Let's move into the meeting. 25 The purpose of the meeting is to discuss the
/'} \-
Heritage Reporting Corporation (202) 628-4888
._ ______-___-____a
B' t
-) 'l 6
D
-(s/ 'l information provided by Philadelphia Electric Company in its 2 June 23rd, 1989,. submittal.
- 3 That submittal included information on dominant E .
4' accident sequences, cost benefit analyses of various
'5; mitigation alternatives for severe accidents, and 6 improvements made to the station since the preparation of '7 'the 1982 probablistic risk assessment.
8 In our review of the submittal, the staff had some 9 questions, which were compiled as a list for discussion 10 items, which we've shared previously. 11 ~Before we start the meeting, I'd like to ask' Rich 12 to make some comments. 13 MR. BARRETT: I'm Rich Barrett. (}. 14. I would simply like to make one plea, and that is 15 for everyone to make the most efficient possible use of the 16 time that we have here. 17 The purpose of the meeting, as Gene said, was to 18 gain additional information and clarification from the 19 utility concerning the basi.s for the submittel,.the recent g i; .l 20 enbn.itt al . , y 21 We formulated scme specific questiens which are
< 22 some of them narrow, some of them rather broad, which we 23 would like to have used as the basis for the discusaion, but a
24 I would like to point.out that we have a limited' time, and I-25 would like to.ask the participants from the NRC to keep this~ L Heritage Reporting Corporation D .
, (202) 628-4888 l
7
,r m
(/ 1 in mind, and when you're asking questions, limit those 2 questions to areas where we truly need further clarification 3 from the utility. 4 MR. SUH: Okay. Thanks, Rich. 5 MR. KRICH: We are here to provide as much 6 information as we can to help you understand our situation. 7 MR. SUH: Okay. I think the way we would like to 8 conduct the meeting, the way I' d like to conduct the meeting 9 is work off of the list of discussion items, unless there's 10 a preference on the part of Philadelphia Electric. There 11 isn't. 12 I would suggest for my part, I'd like to handle -- 13 let's start from the top. [) . 14 Question Number 1. Let me go ahead and read the
%j 15 question and then I'll let the staff member to expand on 16 that, if necessary, and then ask for response.
17 Question Number 1 or Item Number 1, rather, is a 18 comparison of the core damage frequencies for each accident 19 class and the original and updated PRA, and the reasons for 20 the changes by accident class between the original and 21 updated PRA, including changes in models and assumptions, 22 hardware improvements, and credit for recovery actions. 23 Okay. Bob Palla, did you -- did we want to say 24 anything further by way of clarification on that?
~
l 25 MR. PALLA: No. I think that's fine. Heritage Reporting Corporation ( )x i ( (202) 628-4888 l l
c',,
> 8 .d~ !
d-)/1- l' MR. SUH: Okay. ! r . l 2- MR. PHILLABAUM: Bob Schmidt'from NUS would like 3f to' answer that question. 4- MR. SCHMIDT: I'll. pass out material here.
- 5 We have vugraphs about the material. I'm not sure
.6 chow clear they'll be on the screen.
Looks like that can be ; I
- 1 7 read.
ff This is:the breakdown of the class frequency by ! 9 initiator type, if you will, that went into the total that's ; 10 in~ Table 2-2 of the submittal, broken-down by internal fire j
.)
11- and flood. Flooding includes some other than internal i 12' flooding. .I'll cover that in a minute. Total being 1.03E. i 13 to the minus 5.- Talk about each one of these now in a )
) 14' little more detail.
15 The next is a list of'the internals by, if you (l l 1 l 16- will, detailed typical transient or other initiator LOCA 17 initiating events. The total again is the same ao the total q i 18 on the prior vugraph or 5~.9E to the minus 6. This is the l i 1
- 19. current best estimate for Limerick. -{
L 1
~20 You can see the contributions about a little more 21 than a third offsite power, a third other transients, ATWS
- 22. about twenty percent, and then the other LOCAs and we've 23 added random vessel rupture.
f 24 Go on from that. I address the question more 25 . explicitly in the next vugraph. j%. Heritage Reporting Corporation U (202) 628-4888 o . _ - _ _ - _ _ _ _ _
4 ), :1 For internal initiators, I've compared the 2- PRA/ SARA number, and this is really the PRA number. SARA,' 3_ ! severe accident risk assessment, was based on SARA. The
-4 current assessmentiand some of the reasons for the change.
5 We have not done,.if you will, a point-by point comparison L 6 of why therchanges. 7 The plant PRA has gone through a number of 8' revisions, including a complete loading on to the 9 interactive work station and PRA work station that can make 10 changes and so forth, and the models have all been updated 11 and integrated and from that standpoint, it's very difficult 12 to look at any one. thing and see what the difference is. 13 But in the Class 1, which is a -- these are 14 transients where the injection has failed, either high:or.
}-
15 low pressure injection, number has gone down by about a 16 factor of three, and.it's due to a number of things. The
-17 emergency operating procedures were explicitly included in 18 the' current update on the original PRA. They basically 19 didn't exist at-that time, except there may have been a very 20 first cut at them, but training has been factored in, that 21 in interpreting the procedures, in understanding what the 22 operators do in performing even a liability assessment, we 23 talked to the people, operators, training personnel and so 24 forth.
25 That's had a fairly significant impact on the J . Heritage Reporting Corporation (202) 628-4888 s-- _ - _ _ _ _ _ _ _ _ _ . _ - - _
10 oN k) 11' ' result. Loss of offsite power modeling has been. changed
~
E, 2- completely, has updated in things such as battery life, this 3 kind ofLthing.
/ 4 Initiator frequencies have changed. We'll talk-5- more about that a little later, and EDS modifications in 6 terms of-hardware changes certainly are factored into it.
7 I think there are a lot of detailed hardware 8 changes in there. That's one that sticks out as being 1 9 discussed before. 10 The Class 2, which is a loss of heat removal from 11 the containment, has gone down. It's gone down'because the
-12 current assessment includes plant performance and data 13 experience learned in Unit 1 during its start-up test
(} 1 41 program, recognizing the experience there in terms.of on 15 transients, how many relief valves open and so forth. 16 Initiator frequencies have changed and venting has 17 been included in the current analysis. Venting per the 18 procedures, per the EOPs, but only in the context of a heat - 19 removal mechanism, of relieving the pressure to prevent 20 over pressure rupture, and it's been modeled in accordance 21 as the procedures indicate the operators to do after 22 discussions with the operators on the procedures they would 23 follow. 24 I think one item is that based on those 25 discussions, there have been changes in the EOPs to indicate y Heritage Reporting Corporation (202) 628-4888 V-___ __ - -
s; j D i 11
'1 'a better way of doing'it. So,-that's E:en kind of E. fall-2 out.
l 3' The Class 3 and 4, these'are basically the 4 original PRA/ SARA Classes. Class 3 and 4 are both ATWS 5 classes in SARA and PRA. The total'has gone down slightly. 6 Class 3 has gone down considerably. It's gone down. 7 Initiator frequency has changed because of. initiator 8 frequency. The EOPs have very detailed instructions and 9 procedures for handling.an ATWS lowering level, this kind of 10 thing. Those have been factored into a very detailed ATWS 11 analysis, and that's the revised modeling, and then the. 12' . lowering of the MSIV closure set point from Level 2 to-Level 13 1 has made -- has an impact, particularly here, in terms of 14 ability to cope with an ATWS because it maintains the
-(}:
15= condenser as a heat sink for a longer condition. 16 Class S is a -- basically, it's a rapid vessel 17 failure. It's class. The number is used in PRA and SARA 18 was 2.7 minus 8. The current number that's in here is 1 19 minus 8. The difference between the two is strictly a ratio 20 of mean to median for an error factor of 10. It was 21 included just to say that we've included it, to point out 22 that it's not included in NUREG 1150. It's discussed there, 23 my understanding, and written off as a contributor. We 24 thought we wanted to carry it along. The number is 25 certainly subject to a lot of discussion on what the true Heritage Reporting Corporation O' (202) 628-4888 l 1
.i
_777 c. ,
.q ^
- s1 12
- 1. number:is.
* .2 MR. BARRETT: I would like to ask,a. question'about -
lL Class.l'. 4 'I recognize ~you don't have a complete breakdown of 5 how much each )ne of those changes contributed, but could t 6 you -- are the; in that' order or any order presented? Are 7 'the EOPs.the most -- 8 MR. PHILLABAUM: I guess the point we're trying 9: to mak'e was we did not d.o a sensitivity investigation to-
- 10- look at the impact of those things individually. All we did 11 was try'to update the'PRA to reflect all the changes that 12 have been going on. So, what we know'is the integrated 13 impact of.those things. We just don' t know on an item-by-14 item basis what the biggest. impact is.
}(}.
15 Probably the emergency operating procedures we're
'16 following probably did make the biggest impact, but the' ADS 17 modification logic change was another big thing. It's just 18 'hard -- you don't know unless you run through tne numbers. .19- MR .' SCEMIDT: Loss of offsite power is a more 20' easily identifiable piece. I think it was about a factor of 21 three, which is about total. So, it's hard to say, you 22 know. I think my guess is clearly initiator' frequency, we 23' could do the number. We just haven't done it, but they're 24 all about equal.
25 The EOPs and training are important along with the Heritage Reporting Corporation
' O l' (202) 628-4888
_ y
'l .
13
, il' ADS modification.itself in reducing the X' failure to-2 'depressurize probability, for example.
3 I think the percent contributions are very similar L4 to the' earlier assessment. .l 1 5 MR. PALLA: Has the human reliability analysis 6- changed in any1 significant way? Can you point out any key 7 recovery actions that are now credited that weren't credited-8- before? i I 9 MR. SCHMIDT: I think the human reliability 10- . analysis has changed in detail. There have been key areas. 11 If you're familiar with the ATWS EOPs,'those are all human 12 actions, and to incorporate those require basically a whole 13 -new human reliability analysis that wasn't in the earlier 14 version. ( 15 In other areas, they've been factored'in. I don't 16 .think, though, there were any new recovery actions added 17 that were not in, except, for example, in the ATWS EOPs, 18 lowering a level, if you call that a recovery action or not. 19 .I don't really call it a recovery action. 20 MR. PALLA: Well, how about use of diesel-driven 21 fire pumps for core injection? This is credited. 22 MR. SCHMIDT: This is credited only in the venting 23_ analysis in terms of things that the operator could do if -- 24 to provide sources of water from outside reactor enclosure, 25 but not in a detailed. There are many sources of water that
. Heritage Reporting Corporation (202) 628-4888 1
_. -__ ___-_________________________ - - Q
, . . 14 , ; [^'g '.
1 As/ , 1 'the procedures specify. 2 So, from that standpoint, that supply of: water 3 from outside the reactor enclosure is associated with 4 venting for Class 2 wasn't included before. . 5 MR. PALLA: That's the answer. 6 MR. SCHMIDT: Yeah. The venting -- things 7' associated with ventingLare new. Anything associated with
- 8. that. I can't think of anything else.
9 MR. PALLA: I have.a question. You made'a 10 presentation, I guess, the utility people, you made a 11 presentation to ACRS, and presented a profile of breakdown 12 of core damage frequency, which is somewhat different than 13 what we saw in the June 23rd submittal. () 14 MR. SCHMIDT: Different being? 15 MR. PALLA: Some of the frequencies were different 16 for ATHS and transients, not in a major way, but I just want 17 'to understand what we have. 18 MR. SCHMIDT: I can explain. You've got the 19 numbers. 20 MR. MARGULIES: I've got the pie charts. I 21 MR. SCHMIDT: These are the detailed numbers, and l-22- the only difference between this and which the pie charts 23 were based on are associated with turbine trips. At the 24 transient with turbine trips and in the ATWS, where the 25 transient, the turbine trip frequency has gone down from, I
^
j7' Heritage Reporting Corporation (202) 628-4888 o
r
'15
. -A' l' don't remember the number, to 2.55. We cover that later. 2- ..That's.the only change. L 3 MR. PALLA: So, really,7what we are looking at in-
'4 the most recent submittalLis the 1988 update with;only that 5 ' change?
6' MR. PHILLABAUM: With only that change. 7 MR. PALLA: The. turbine trip frequency change? 8 MR. SCHMIDT: Right. 9 MR. PHILLABAUM: And I think if you go back.to the 10 Ltranscript of the ACRS subcommittee meetings, we indicated 11 that we were going.to make that. 12 MR. PALLA: You made that observation. I just
.13 wanted to make sure that that was the only change.
() 14 MR. PEILLABAUM: I' m just going to show you 15 quickly. It really hasn't changed that much. This is the 16 updated numbers that showcd up and presented to ACRS, and 17 this is what happens when you use the revised numbers 18' showing up down here. It's a little -- it's pretty tough to 19 see the difference. 20 MR. SCHMIDT: But it is solely due to the turbine 21 trip frequency. That'e the first pie chart that they l 22 presented, was based on this same table, except the turbine
'23' trip core damage frequency here and here was higher.
24 So, these numbers, the ATWS frequency percent. 25' contribution has gone down slightly, and the transient has 1 Heritage Reporting Corporation 1 (202) 628-4888
16
' hv 1 :gone1down even less to give these numbers.
2' MR. CHELLIAY: I guess we do see some differences 3 to what.you are showing.here, in particular, on class 4 sequences estimate,fand what you have provided on Table 1-2 L 5; of your July 17th submittal they verify on the class 4 l 6- sequence. frequencies. 7 MR. SCHMIDT?, 'This is not -- Class 4 total is -- .) 1 8 for internals is 1.05E to the.minus 06, 9~ VOICE: Is that 06 or 07? 10~ ~MR. CHELLIAY: I'm going to bring to you a handout 11 right now class 4 -- 12 MR. SCHMIDT: Some of these are Class 3 down here. 13 That's the difference between that total. 14 MR. CHELLIAY: No. I'm looking at ' your first page
.( ).
15 of your Table 1-3. Page 1-3 of your current presentation,
.16 you said on class 4, the current estimate is 1E minus 7. I-17 think it's more than that.
18 MR. SCHMIDT: I think there's a typo on that. 19 MR. CHELLIAY: I think so. I think it's 1 E minus
-20 6. Because I'm comparing your current presentation with 21- what you provided to us in Table 1-2 of your submittal.
22 MR. SCHMIDT: The second number current /4 should 23 be 1.05E to the minus 6. 24 MR. CHELLIAY: Yes. That's what I thought. 25 MR. SCHMIDT: That's correct. Thank you. Heritage Reporting Corporation O (202) 628-4888
. _ -__-__ _D
q 1 17 m l' ' (_) 1 MR. SUH: Is that the one that's MSIV closure set ) 2 point, Bob? Is that the number you're referring to? 3 MR. SCHMIDT: Yeah. It was all done rather 4 hurriedly.
- l 5 This number 1.05E-06 --
6 MR. CHELLIAY: Bob, then the question is, why the i 7 frequency changed from the previous PRA estimate to now 8 minus 67 9 MR. SCHMIDT: From the previous PRA? 10 MR. CHELLIAY: Yes. 11 MR. SCHMIDT: What do you mean previous? The 12 November ' 88 update or the PRA that was done by -- 13 MR. CHELLIAY: The PRA is part of SARA? 14 MR. SCHMIDT: Why the change? Okay. All of those 15 reasons changed it. The number has gone up principally -- 16 let me make first a statement. l 17 I think the current analysis of ATWS is 18 conservative. The dominant contributor, if you will look at 19 the dominant sequences that are shown in the submittal, 20 Table 1.2, which lists all the sequences, the Class 4 ATWS 21 sequences there have a U prime failure, which is failure to 22 prevent over-filling of the vessel. 23 MR. CHELLIAY: Okay. 24 MR. SCHMIDT: The current analysis takes into 25 account the EOPs which instruct the operators to do a whole r'g Heritage Reporting Corporation x ,/ (202) 628-4888
t 18' lli variety of things.such as: lowering level:-- 12 MR. CHELLIAY: Yes. 3 MR. SCHMIDT: -- and also requiring 4 depressurization -- 5 MR. CHELLIAY: Yes. 6- MR.'SCHMIDT: -- to keep the suppression-pool' 7 temperature below the capacity temperature limit. U 81 'MR. CHELLIAY: .Yes.
. 9 MR. SCHMIDT: One-of the effects of that that is 10 probably more. predominant in Limerick than other plants, 11 because Limerick has an automatic slick system, --
12 MR. CHELLIAY: Yes. 13 MR. SCHMIDT: -- is'the potential for when you 14 .depressurize an open relief valve for low pressure system 15 injection to overfill the vessel and dilute the boron. 16 MR. CHELLIAY: Yes. 17 MR. SCHMIDT: Trevor may be able to say there's 18 good analysis that I'm not aware of, but the analysis of 19 these partial boration, flooding, level, all.of these 20 interactions are very'hard to assess. There's' transients
- 21. that occur, a lot of things.
22 We have, I think, a very conservative analysis of 23- that potential, and you'll see the dominant failure, 24 dominant failure mode for ATWS is the E prime failure to 2.5 prevent overfilling the vessel. Heritage Reporting Corporation (202) 628-4888
f, f M. ' h 19 O L/' 'i, . R f f ..-l o That: perception.of what happensfdidn't: exist when
~ , .21 Lthe PRA.was'done.. .The EOPsJdidn't exist. -Depressurization, H 3 directions.to depressurize, directions to lower-level,
- 4 didn't exist. So, it's. basically a-new analysis that's been 5 done,.and the trend is, indeed, upward ~for that particular.
6- . transient. 7 :What has. happened is the EOPs now strive 1to keep 8 the core cool and give the. operators a lot of instructions 9 to inject water in the vessel, where, in the:PRA/SARAL 10, analysis',;those'didn't exist. So, what happened was the 11' high pressure injection system failed, HPCI and RCIC. failed 12 due to high temperature'in the suppression pool. Therefore, 13 most of'the time, you got a Class 3.
-- 14 MR. CHELLIAY: Yes.
15 MR. SCHMIDT: And in a smaller fraction of time, 16- you got Class 4.
.17 Well, now with EOPs, you tend to do that. So, , -i -l 11 8 that's.the reason. It's a very complex story. .19 MR. CHELLIAY: Yes. I do agree. That's what I 20 'was predicting in the olden days. I was not really quite <21 sure what you would be looking for in this p~rticular a Class 22 4:which really dominates the early photography of this 23 particular facility. i 24 Could you answer then my question? In this 25 situation is the EOP hurting you or helping you, so far as Heritage Reporting Corporation .O. (202) 628-4888
.20-(sJ l' 'the'likblihood of the. full d'amage --
2- MR.-SCHMIDT: .I haven't made that comparison. I 3 don't know. It helps you in some respects. There are some 4 adverse impacts in the other. We have made recommendations , 5 to Philadelphia Electric to true knot the EOPs, if-you will. 6 They have been looking at these, and this is something that 7 they could do, but it's hard to make a balance. 8 MR. PALLA: Could you just explain a-little bit' 9 more about the cause of this operator failure to properly 10 control? Is it that he does not throttle back the injection 11 pumps when -- after he's depressurized and is attempting to 12 inject? It's an uncontrolled -- it's cold water that's not 13 properly maintained? 14 MR. SCHMIDT: I don't know what the dominant
) -15 individual contributor is, but it is the reactor is 16 depressurizing,'either intentionally, i.e. he's doing it 17 intentionally, or -- and the procedures are very clear on 18- this. To depressurize and be careful about.all this. Or 19 he's doing -- it's occurring unintentionally because there's 20 either a stuck-open relief valve or there are relief valves 21 that are left open for other reasons. So, they open the 22 relief valves.
23 When the boron comes in now, power goes down and 24 the pressure can go down fairly quickly. The low pressure 25 pumps then can come on and, unless they're careful, very Heritage Reporting Corporation (202) 628-4888
. s_
1
t
- 21
- p. . . -
V 'if quickly, because you have a very large capacity, which is 2- normally good.to refill'the; vessel, and that's the kind of 13' thing. o4 I don't know. I mean, we've looked at a lot of 5- different scenarios. They're all; included in the analysis.
~
6s 'What's the. dominant particular one, I don't know. 7 MR. PALLA: And those sequences are assumed to goL 8 ..to. core melt'even though -- 9 MR. SCHMIDT: Those are seen to go to core melt. 10 MR. PALLA: Do you get like a pressure or power 11- ' oscillation or something? 12- MR. SCHMIDT: You get pressure and power 13 oscillations. The net effect is assumed that'the power is
'() '14: higher than-it would be. The assumptior is that.you dilute 15 the boron, that you put water down the steam lines and get 16 ' rid of the boron that's otherwise needed to maintain it-sub-17 critical. Therefore, you have a long-term shutdown problem.
l 18 Now, there are procedures.in place for' adding 19 boron from other means than the EOPs. .When this analysis E 20 was done, those were fairly ineffective in terms of time. l-21 Time they took. So, the assumption is that.by loss of 22 boron, power would be higher than it would be otherwise for 23 a number of hours. Consequently, the containment would be - 24 - temperature and pressure would be higher and that would 25 lead to failure. Heritage Reporting Corporation (202) 628-4888 l _ _ _ _ _ _ _ - _ _ - - _ - _ _ ~
u 22 l if3< 1 ik/J 1' I think that's a conservative assessment. We 1: o ;2 stopped it.at'that, point, though. It's a question of how 3- far you'want to keep going. There is no venting credit-l 4 given for that, and that's - particularly for these partial 5- borated cases, you-might -- venting might be affected or 6- wouldn't be for the full case where you have no boron in at
~
7 all or it might not be because-less gives you higher power. 8 MR. MARGULIES: I have a question. We noticed that station blackouts contributed'-- 9 10 ~you indicated that there had been some improvements. Could 11 you elaborate on that, in particular battery lifetime and 12 the capability to cross tie or recharge. batteries? Could 13 you elaborate on that on a station blackout scenario? ()- 14- MR. PHILLABAUM: Let me just back up. Previously, 15 the PRA/ SARA, there were no station specific procedure to 16 follow.- What they just,have now is the station blackout 17 procedure, which they can follow and takes credit for doing 18 things, such as the thing.that's identified as opening of 19 the HPCI and RCIC room doors,-provide alternate ventilation, 20 making sure you lower the pressure of the vessel to try to 21 extend the amount of time you have for heat-up, those kinds 22 of things, trying to conserve your CST water, transferring 23 back and forth from the suppression pool to the CST to 24 maintain cold water as long as possible. 25 Battery life has been assessed to be -- the Heritage Reporting Corporation O' (202) 628-4888
[ l 23 f
,P~ .
, -1
. 1 designed battery. life in Limerick is eight hours per each of.
2 the'four divisions,-without load shedding. We didn't take 3 credit in the' procedure for the load shedding that they have 4 identified in our station blackout response -- I- 5 MR .' PALLA: Is that true even in the updated PRA? 6 You have no' credit? You have load shedding procedures that 7 don't'have -- 8 MR. PHILLABAUM: We have not taken credit for the 9 . extended battery life. -We said, well, their design -- this 10 battery -- the design would be an eight-hour capacity for 11 their loads that they' re supplying and that's where we just 12 said fine, we'll take that eight hours and assume that they 13 have no load shedding to extend the battery life.
- .14 .MR. SCHMIDT: Original PRA what was the battery-15 life?
16 MR. PHILLABAUM: Original PRA was assessed to be 17 four hours, assumed to be four hours. 18 MR. SCHMIDT: . What are the changes that you're 19 seeing?
- 20. MR. PHILLABAUM: Additionally, the question was, 21 do we take credit for cross tieing or jumping? No, we 22 don't, although in our station blackout submittals in 23 response to the rulemaking, we plan to have procedures in 24 place to cross tie diesel generator sets from one division 25 to another as necessary.
1 Heritage Reporting Corporation O- (202) 628-4888 , I
- . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ l
24' p [. 1 '-
. ~ 1- 'So,' diesel generators.can repower up the. battery-2 Echargers, so you can, resupply'DC power that-way as well.
3 MR . KRICH: In response to the station blackout l.
.4 rule, we classify ourselves as a one-hour alternate AC e
5' plant. We would'have alternate AC within one hour. [ 6. We have not taken credit fer venting during a 7 station blackout scenario. 8 .MR. MARIE: It depends on the accessibility of the 9 timing that you'want to t ry to align the valve to vent 10 containmer.c. 11 MR. SCHMIDT: I think it's possible, but there's 12 no credit for venting in blackouts. m 13 MR. PALLA: These would be manually-operated f 14 valves? Do you have air bottles or something? 15 MR. MARIE: Because there are AC valves in station
.16 blackout, right, you have to go out remotely and change the-U '17 valves.
18 MR. PALLA: Same situation as Peach Bottom? 19 MR. MARIE: Exactly. 20 MR. PHILLABAUM: Okay. Maybe if we could switch
' 21 ' now to ATWS, --
22 MR. SCHMIDT: Excuse me. Did that answer all the 23 parts of your question, because I think you were all asking 24 about loss of off-site power? ) 25 MR. PHILLABAUM: And you have indicated HPCI and Heritage Reporting Corporation (202) 628-4888
= _ _ _ _ _ - _ _ . - _ _ _
pr t d- 25
%( /
i 1 RCIC coolin'gs, improvements:you've made -- J 2 MR..PALLA: 'Are those-things within the-first' L3 column, though or after that?
'4 MR. PHILLABAUM: Those things have been 5 -incorporated into the current. numbers. Again,Lpreviously, L 6 there1was no station blackout procedures. So, a lot'of 7 ' assumptions,.more assumptions rather than what actually is. .8 in place at Limerick, was_done originally. In '82, Limerick-9 didn't have-a license, .you know. They were still in the 10 process of developing'their procedures in the next couple of 11 years.
l12_ MR. SCHMIDT: Alternate' room cooling was included 13 in the original PRA, given the likelihood of achieving _it has been reassessed considering now that there are f() 14 15'- procedures.
.16 MR. PHILLABAUM: Procedures in place in training.
17 MR. SCHMIDT: Training and so forth.
-18 MR. MARGULIES: Just a little bit on 19 pressurization. What is your current indication of the '20 values, probability values for successfully venting 21 overfilling? .22 MR. SCHMIDT: In ATWS?
23 MR. MARGULIES: Yes. You talked about them , 24 earlier.
'25 - MR. SCHMIDT: It's from a very small number to 1
j f p' Heritage Reporting Corporation (_/ (202) 628-4888
s - ~l 1 26 1
;K. ' /w I .1% like .5, I'believe. Very small. number. -High pressure, p~ l2 injection systems.have a Level 8 trip which prevents them L
33 . continuing. Low pressure systems do not. For low pressure
~ ~ .41 systems,.I. don't recall the low end of that, but I think'in 5- -some situations, we've assumed that the -- assessed that the 6- likelihood of~ preventing overfilling can be as bad as 7 something11ike .5, and I-would have to confirm that:
8 MR. SUH: In the interest of time, I'd like to' 9 move on, but let me ask a bookkeeping question first, and I 10 think it's the same question:that was asked earlier. 11 The basis that I'm operating out of is I read your
- 12 ACRS presentation. It seemed to me that just talking about 13 internal events, the core damage frequency from the PRA, 14 original PRA, was 1.45 to the minus 5. Then, your ACRS --
j} 15 before the ACRS, you talked about the 1986 Level-1 update,
- 16 the 1988 Level 1. update. And that got you down to a core 17 damage frequency 6.69 to the minus 6. Okay.
18 There's a number of reasons why it went from 1.45-19 minus 5 to 6.69 minus 6, and I won't go through them,~but 20 except'to ask, can we go back and parse these out and say 21 pay'for the eightieth logic improvement that you made, can 22 we go and figure out what risk reduction was achieved on 23 that one thing and what it costs to have done that and go 24 back and do that? 25 MR. PHILLABAUM: Now, I think the answer is no Heritage Reporting Corporation O' (202) 628-4888 i
-_--____-____a
( 27L _.pn
- k. 1 because1what we've done is we've'gone and put this on a PC
, .2 work. station. So,.you don't have the '85 --'we have the '82 ,
3 model,Lwhich is not -- it's not integrated together, fully-4: linked.together like the present model is'. So, it would be 5 going back in-timetand then putting things in one at a time 6 and see what their impact is, and that would be pretty 7 difficult to do, 'especially-looking at.the procedures that 8 :they,go through everything, just infiltrate, you know, all
- 9. the --. pretty much the whole PRA.
lL O I don't think it's really very easy to do. 11 MR. SCHMIDT: It could be a number game of what 12 was X-before and what is it now and'what's the impact of 13 that,:but there are a lot of pieces.that contribute that and 14 there are an awful lot of pieces that contribute.to all of 15 the sequences. 16 Take one sequence, TQUX, a lot of things 17 contribute to X and Q, U and B are changed. So it would be ps 18 very misleading to do it.
-19 MR. MARIE: And to try to show a Delta for any 20 single item will require a certain basis and you have to 21 assume what else was done or was not done, and it's left.
22 It would be kind of a numbers game. It's kind of a 23 synergistic effect when you combine them all and see the 24 bottom line and try to pull out a portion of that and assign 25 it to any part is going to be difficult.
-r Heritage Reporting Corporation ; (202) 628-4888
28 1 MR. SUH: Okay. Thanks. 2 One administrative item. I guess, Jerry, I guess 3 you showed a slide earlier. Could you please just keep those I 4 on the side that we can maybe make a copy later? t 5 MR. PHILLABAUM: Sure. 6 MR. BHATTACHARYYA: Is there any estimate for the 7 uncertainties of those numbers? 8 MR. SCHMIDT: I believe that's later on. Let's 9 hold off until'we get to Question 2 and then we'll talk 10 about it, sir. 11 MR. SUH: Can we move on? 12 MR. MARGULIES: One more quick general question. 13 Did you use the Maris data in the quantification? 14 MR. SCHMIDT: For what?
}
15 MR. MARGULIES: For the PRA -- 16 MR. SCHMIDT: Failure rates? Most of the failure 17 rates are generic. The diesel generators are plant 18 specific. The turbine trip frequency is now plant specific. 19' There are a few plant specific, but, generally, it's 20 generic. The plant has been running about three years. So, 21 most of the plants are generic. o 22 Question on fire initiators. Changes. First 23 column is PRA/ SARA. Second column is current. You see that 24 the Class 1 has gone up. Class 2 has gone down. Class 1 has 25 gone up or changed for a number of reasons. Some changes Heritage Reporting Corporation ((~g _/ (202) 628-4888
_ - _ _ _ _ - - _ _ = - _- ._
- p. #.
.~ .
29' a}?l;
? 11 ?made it go.down; others'have made it go up. Plant design.
2: At the-timet the'PRA was done, the Appendix R modifications 3J were in a~ state of flux and being negotiated. So,.this
- 4 number is based on thefcurrent. It's~ Revision 11'to the Fire-Protection Evaluation Report', which has improved things.
S 6- . generally. 7 New initiator:and new fire' suppression data from 8 the'Sandias' Scoping Study have been included in here. Those
- 9. tendLto make it go up.and then the integrated model that 10 'went'into:the' internal events analysis have been included in
- 11. the fire analysis,ti.e. if the fire fails a certain set-of 12 equipment',-the probability that the core melts due to that, 13 the-fire has to include random failure, and the random 14 failure probability'that leads to core melt or the condition.
([
.15 of probabilit'y of core melt given all these random failures 16 are based on the internal events model.for this' equipment 17 failed that had.been failed due to the fire.
18 So,_all of that combines and yields a slight 19 increase. We still believe that's conservative for reasons 20' we'll talk about in another q astion. 21 The Class 2 again, the loss of heat removal 22 sequences have gone down. Changes occured because of the 23 Class 1 items plus the fact that venting is accounted for in , i 24 Class 2.
.25 MR. BHATTACHARYYA: In the, revised current Heritage Reporting Corporation Ot (202) 628-4888 i
.r . 30 7y [ -b I l' estimate,'you have fire'supression or something?'
, J2 MR. SCHMIDT: -Fire' suppression?
- 3 MR. BHATTACHARYYA
- .Yes.
4 MR. SCHMIDT:- Yes. There is manual fire 5- suppression analysis and the extent that it's covered, when-13 'we talk about uncertainties,-we cover that, if I.can defer 7- .that.. Okay?. What was done there. In one of the later 8 questions, we cover fire. Okay? 9 MR. BEATTACEARYYA: Could I ask a question one, j 11 0 more time?. 11 .You have two kinds of systems. One kind has a 12 'three hour fire rating, one kind has a one hour fire rating. 13- 'Is that'the same plant configuration you have at this time?- () 14- MR. PHILLABAUM: Could you repeat that? 151 MR. BEATTACHARYYA: Okay. What do you estimate 16 based on your plant configuration there you have, one-plant 17 system has a one hour fire rating. The'other plant safety 18 system has a three hour fire rating.
.19 My question to you, the actual plant configuration 20 right now, are they the same thing or something else?
21 MR. SCHMIDT: I don't know the answer to that 22 implicitly. I'm not the fire protection expert. 23 MR. PHILLABAUM: Are you basing this on what was 24~ in the FSAR?
-25 MR. BHATTACHARYYA: No. On the PRA.
Heritage Reporting Corporation
..O4 - .
(202) 628-4888
._____m._ _ __
p _ . . . . , 31 =-
.(
E A - ' 1 MR. SCHMIDT: There have been a let of changes
; 2 since'the-original PRA was done.
p 3 MR. BHATTACHARYYA: Exactly. The first. time you 4 gave over the high fire' frequency I suggested to you some
-5 . plant' modifications. We did that. We brought'down the. fire
+ 6 frequency. Now, you people at-PRA again far under 2R. t 7 That's okay, you know, but I'm trying to find out still the 8 same modification'is there or both plants are clear of fire
'9 rating? .10 MR. PHILLABAUM: We have to get back with you on 'll . that'one. We have to check that one.
12 MR. CHELLIAY: I'd like to get that answer. 13' MR. BHATTACHARYYA: That would have a greater-
~ .. 14 impact on fire. See, these things are important because 15 some of the SAMDAs are based on the class 1 sequences. So, 16 we want to be.somewhat -- we want to get a feel on what'is a 17 realistic estimate on class 1 sequences. Particularly, '18 some of our CPI programs, some of the mitigation devices are 19 concentrated ~on this class 1, you know. So, I'd like to get 20- what is the realistic frequency estimate.
21 MR. SCHMIDT: This fire frequency for Class 1 is, 22- -we believe, conservative, and we have a list of the reasons 23 why in a later question. 24 MR. BHATTACHARYYA: Okay. Thank you. 25 MR. SCHMIDT: The third category of initiators is f"'g Heritage Reporting Corporation (,/ (202) 628-4888
'Y. Y; ^
32 ry.. 5I 1 flooding and:other initiators, other meaning, otherLspecial
' 2: initiators, which I think we'll talk about more later,. but 3 the originalLin SARA; the Class 1 and 2 frequencies, due to 4 ' internal flooding, were shown,-estimated to be less than 5 5 minus 7 less than 7 minus 8.
6 'We've updated that. analysis and the current 7 ' estimate of flood, internal flood initiated sequences, 8 and 8 10 to-the minus B. There are other special initiators which 9 total about 9 minus 8. They're all lower than that. So, in 10 the assessment.of the SAMDAs we basically added those two o 11 and rounded it up to 2 minus 7 to add in a, if.you will,
'12 -account for that. They're not major contributors.
13 In fact, in SARA, I think flooding was included as 14 not a contributor. (f
'15 MR.'BHATTACHARYYA: During the current' flooding 16 analysis is.this detailed in nature, or a scoping analysis 17 like what we did during SARA unit.
l l 118 MR. SCHMIDT: It's a little more detailed in that 19- it does look at the height of floods and what equipment 1 20 might be damaged by that height and in one area, there is an 21'- instrumentation that's there to warn of flooding and that's L 22 factored into the analysis.
'23 So, it is fairly -- it's not just a very broad 24 screening analysis, but I think it's still conservative. ; 25 Well, that answers Question 1. .e Heritage Reporting Corporation (202) 628-4888 1
l' l 1 _________-_______---___.__a
s 33
'. 1- MR. SUH: 'Before we go on to Question 2, Bob,.do i.
l 2' you have some' estimates of the uncertainty associated with i 1 3 the core damage frequency estimates.on which the PECo/SAMDAL 4 assessment is based? 5' MR. SCHMIDT: .The simple answer to that.is no.- 'Ho 6 uncertainty analysis has been done for the current estimates 7 that are used in the SANDA analysis. The last uncertainty ~ 8 analysis performed was done in SARA. There was an estimate. 9 It was estimated by Brookhaven in their review of the PRA, 10' .and an estimate does exist for Peach Bottom in NUREG 1150. 11 Certainly, the most complete analysis, I guess, of 12 uncertainty. 13 I'm presenting those because I think they're 14 , relevant when you look at numbers that have been generated 15 for. uncertainty. The Peach Bottom analysis in NUREG 1150 16 has an internal events core damage frequency mean of 17 something like 4, 4 and a half and 10 to the minus 6. We're 18 ' talking about 6 times 10 to the minus 6. Roughly that kind 19 of sequence is dominant. 20 .But to recapitulate the uncertainty factors that 21 were estimated in various places, the key here is that the - 22 - in each of these equations, the first number is the error
- 23. factor in the upper direction, the ratio of the 95 24 percentile to the median. The other number is the ratio of 25 the median to the 5 percentile, and then the bottom number Heritage Reporting Corporation (202) 628-4888
34
/^\ 'L/ ~ 1 is the product of those two or the ratio of 95 to 5 2 percentile.
3 From that, you'll see that they're reasonably 4 consistent. Total range varies from 25 to 50 or 37, and I 5 think you can say an error factor median of 95 somewhere in 6 the range of 6 to 8 or 6 to 10 is certainly a reasonable 7 number. 8 Now, if we're to do this for Limerick today, what 9 would we get? I'm pretty confident we'd get a number in 10 that range. I don't think there's anything so unusual, so 11 overwhelmingly dominant, that it would change that. 12 If you get, I guess, one sequence of extremely 13 large fraction of the total core damage frequency, then you () 14 might go be shifted towards the uncertainty for that, and 15 when you get a lot of sequences, they're all talking pretty 16 much the same set of generic data, same uncertainty factors, 17 and, so, I think these are reasonable estimates. 18 Fire. You see what was estimated there. 19 Consistent numbers, but we do not have a, if you will, a 20 Limerick PRA current PRA specific uncertainty analysis. 21 MR. MARGULIES: Tim Margulies. You recognize that 22 PRA is a snapshot of the design and operations of the plant. 23 What if we try to use those for operations? In a sense, as 24 I understand it, that's kind of a liability maintenance 25 program, a kind of ongoing program. Kind of like a leash r' Heritage Reporting Corporation It )T (202) 628-4888
L i i L 35 l f
. f' l .
I L .
/
1- .for uncertainty when you're operating the plant. 2 MR. MARIE: I will.try to address.that. As.you 3 mentioned, we are developing at our RCM Program, at this 4 stage, that's^in-the formulation stage of developing a pilot 5 . program to be putLin place next year, we have established a 6 PRA Program plan, which has' developed specific objectives, 7 and this was described to the ACRS committee in that 8 presentation. 9 'And that has included input in the involvement of. 10' parties from the stations, all parts of the nuclear. 11 organization. So that people are in tune with the objective 12 of.PRA. And we have had on going training programs for 13 personnel in the departments of the nuclear organization and l (} 14 -the training has been provided to over 600 personnel up to 15 date. 16 We try to keep information current through 17 issuance of routine newsletters to supervisors and managers 18 throughout the nuclear organization,.providing information 19 on updates in terms of what's going on in the industry as 20 sell as updates on our analysis. 21 If we perform an update, as we do for the PRA, , 22 periodically, the results and insights are summarized in 23 memoranda and circulated to the management throughout the 24 nuclear organizo. tion. It's a program to keep people aware E 25 and informed. Heritage Reporting Corporation O (202) 628-4888 1
36 1 MR. SUH: I would like to move on to Question 3, 2 and I'd like to have a break right after this question. 3 Number 3 is how NRC review comments, 4 recommendations per NUREG 1068 on the original PRA were 5 incorporated within the revised PRA, and the rationale if 6 certain recommendations were not included. 7 MR. KRICH: Jerry Phillabaum will answer this one. 8 MR. PHILLABAUM: NURLG 1068 referenced Brookhaven 9 NUREG CR3028, which had really identified three main areas 10 of disagreement when they reviewed the Limerick PRA. One 11 general area was deficiencies they felt between the 12 commonalities that may exist between systems across the 13 various events in the event tree plus perhaps some missing 14 links between support system and front line systems. . (-[') 15 The second main area was some disagreements 16 between some quantification of some failure modes that were 17 included in the logic models for the systems, and the final 18 area was -- the main area was e disagreement about the event 19 frequencies that were used. 20 Concerning the first general area of the 21 dependencies, what has been accomplished is to go back and 22 reverify the accuracy of the support system front line 23 system linkage to make sure we have accurately reflected 24 that. 25 Secondly, we address some of the dependencies that (' Heritage Reporting Corporation (202) 628-4888 i
l V, L
'37 lf)-
O' k . 1 werel identified across events on our event trees. We 2 ' accommodated all these deficiencies, except for the 3 dependence of the high pressure injection systems on the 4' decay heat dependence between U and W functions. 5 We have'made -- at this point,-we have made a 6 sensitivity investigation of that dependence and believe
'7' .that it's not very significant, and that's one of the things 8 we.can do in the future, but at this time, we haven't 9 completed that incorporation of that one change. Everything 10 else, though,:has been accommodated and is reflected in our 11 updated PRA numbers.
12 Finally, as far as the incorporation of the 13' dependencies go, it's important to emphasize that we now l()' 14' have full fault tree linking, which, using the code that we 15 have now, .does account for commonalities across events. So 16 that deficiency has been corrected. 17 Moving down to the frequency of the initiating 18 events, this shows the original -- this shows the comparison 19 between the transient initiating. event frequencies used-20 originally versus the Brookhaven assessment versus what was
- 21. done in the update with a highlight to show as far as the
'22 June submittal went, the one change to reflect the actual 23 Limerick experience through its commercial operation 24 history.
25 We'll note the situation there, and we believe, Heritage Reporting Corporation O- (202) 628-4888
38'
~
il based bn1 Limerick, again based on actual Limerick;
.2. . experience, that the numbers we have used in our updated PRA 3- - are' supportable. Again, the low frequency. initiating event, 4 transient events,. are based on generic EG&G data.
k 5 MR. PALLA:- Is that MSIG thing a typo'or is that 6' correct? 7 MR.-PHILLABAUM: 0.23. 8 MR. PALLA: That is:a zero?
'9 MR. PHILLABAUM: Right. What we've done on -- we 10 separated out -- we categorized some of the things that.were 11 considered as MSIV closure into loss of feed water and loss-12 of condenser vacuum to better account for these unique- -13 challenges in plant responses to those initiating events.
14 MR. SCHMIDT: Totals about one. The difference is
-{ } .15 not that big.
16 MR..CHELLIAY: Still the way we define the MSIV 17 Closure, where we can not get the feed water within 30 18 . minutes from'the transient. Isn't it? So that we can not 19 include it in the recovery? 20' MR. PHILLABAUM: No. Recovery of'the reopening
- 21. from the inside of the enclosure is assessed differently, 22 it's assessed as a probability. But the probability varies, 23 that's one of the reasons that we separate it out
- 24. MR. BHATTACHARYYA: That's what I'm trying to get 25 at.
Heritage Reporting Corporation O. (202) 628-4888
i k p 7-39 1 MR. PHILLABAUM: See, before these were lumped i 2 together and that's one of the main reasons they were i ! 3 segregated here, to better account for the possibility of. y 4 recovering feed water and reopening or determined bypass i J: 5 valves when'these different initiating events, and for that 6 reason, that's why they're. assessed a little differently. I'- 7 MR. BEATTACHARYYA: Do you have some improved 8~ procedures'to reopen the MSIV by balancing the condenser 9 vacuum? 10 MR. PHILLABAUM: We still have to balance the
- 11. ~ differential, pressure between one side of the MSIVs to the 12 other. That's still a requirement.
.13 MR. SCHMIDT: The EOP specifically talk about and
() 14 direct the operator to restore the condenser as a heat. 15 sanction. So, there is a lot more guidance to do that. 16 MR. PHILLABAUM: And you always list feed water'as 17 one of the primary injection sources as well no matter what. 18 It says to make sure to try to re-establish feed water. So, 19 that's one of the things that we'll be trying to consider 20 when they go down there. The alternatives. 21 Lastly, the -- 22 MR. MARGULIES: Do you recall the kind of 23 breakdown on the loss of ofrsite power in terms of the 24 percentage of the plant center that is stability related or 25 weather.related?
/ Heritage Reporting Corporation (202) 628-4888 I
____.___..._____._a -
l r 40
'1 MR. PHILLABAUM: What I wanted to do was there's a 2 question, I think it was 10,-I'll'say Question 10, talks j
,1.
? about -- we interpreted it to mean that kind of a. thing.
! 4 So, 'I' d . say, that that - kind of information will- be provided . .i j S' then as'far as what' contributes to the' frequency of loss of q 6 'offsite powerfand how we develop the. recovery. : 7 -MR. MARGULIES:- Okay. ! L8 MR bHILLABAUM: The last major' category, I don't , 9 have any overheads for, and that's'the corrections and l 10 modifications for Brookhaven necessary to correct some of 11 'the input. Amd that's what the rest of that handout j 12 describes. ; 13 It' s ,yoint-by-point comparison of the comments
), 14 in NUREG CR3028 versus what we've done, and-with a few 15 exceptions, we've tried to accommodate all the Brookhaven !
16 identified shortcomings. There's just a few on the first 17 _page here, failure to recognize that the low pressure 18 injection systems may be required if the second relief valve 19 causes pressurization. That -- we've looked at that in 20' conjunction with the previous comment about the dependency I 21 of U on W. 22 So, that's being addressed at this present time, but now we haven't resolved it yet. That's why it's carried 23 24 the way it is. It's not included in this latest update which 25 formed the basis of.our submittal to your questions. I Heritage Reporting Corporation j O' (202) 628-4888 1 ( l
41.
.:\-(?}1 1 I don't want to go over these in detail. .I think 2 they're pretty self-explanatory. If anybody does have any 3 . questions sometime, I'llfbe glad to answer them.
4 MR. SUH: Just one question, Jerry. When -- this 5 is the subsequent -- it says resolution of the updated PRA. 6 The updated PRA being the 1988 PRA or - . 7' MR. PHILLABAUM: Well, 1988 PRA and the one that 8- -- it also is the same as what was submitted for the 9 response to the questions received. So,.the '89 submittal. 10 The one you just got. The same applies for that. There's 11- no difference. 12 MR. SUH: So, this isn't back in the 1983? This 13 was very recent? () 14 MR. PHILLABAUM: This is what -- well, this -- the 15- comments are based on the '81-82 PRA. Those are the 16 comments. The resolution is as it stands today essentially. 17 MR. SCHMIDT: It's incorporated in the time period 18 from ' 83 through ' 88. 19 MR. SUH: What's the difference between the 20 handwritten and typewritten? 21 MR. SCHMIDT: Well, I tried to make -- there were 22 some things I thought would be -- that needed to get 23 clarified a little better. So, that's why I wanted to make 24 -- this is what I use as kind of my checklist and there were 25 some things I wanted to go ahead and add. That's the only (' Heritage Reporting Corporation (202) 628-4888
A, , 42 s ' .- L(,A 1 : difference. 2 MR. PHILLABAUM: That, and the short amount of
.3 time that'we had to get it to you.
L 4 MR. BHATTACHARYYA: But the original was for a 5 scheme. This is good. I think what you gave this is the 6 review where I performed at NRR as far as the internal and 7 external review. But the first thing right now refers to 8 1068. I'm right? Please correct me. Your Question Number 9 3 is the 1068, which is the staff review of the PRA as.well m 10 as the NUREG CR 3028. In the staff report we asked you to 11 consider certain improvements, plant improvements, like ADS 12- logic change, which you have done I think.
.13 MR. PHILLABAUM: Thst's right. , -. 14 MR. BHATTACHARYYA: I think that is the question.
15 This answer is good. This is your rebuttal of our BNL and 16 the staff comments on your original PRA. You agree with 17 this, no doubt about it. But the question right now, it' 18 reads 23. It refers to NUREG 1068, which is a staff NUREG. 19 Which review comments in the sense that if you made a 20 comment, these are the comments you answer. Plus we 21 suggested to you three improvements. We would like to get 22 an answer to what.you did. One of the improvements the ADS 23 logic, which you have incorporated, which you said 24 previously, what happened to the other two? 25 MR. PHILLABAUM: I have the NUREG, but I can Heritage Reporting Corporation Oj (202) 628-4888 l
. _ _ . - ._____.-_.m.___-___
r 43 4
' <~)
s- -1
-probably just tell you. Part of the other two 2 recommendations.
3 MR. BHATTACHARYYA: That's the question we are 4 discussing. 5 MR. SCHMIDT: Why don't we look at it at the 6 break? ; 7 MR. BHATTACHARYYA: All right. 8 MR. KRICH: Let me make sure I understand. 9 As I read this question, I think my understanding 10 of it was that you were asking us to discuss how the NRC 11 review comments on the PRA were addressed. 12 MR. BHATTACHARYYA: 1068' comments. 13 MR. PHILLABAUM: Right. But regarding the PRA, () 14 not necessarily regarding how the plant is designed. 15 MR. HARDIN: Well, I can certainly see how this 16 question would be read that way. 17 MR. PHILLABAUM: I think that's how we prepared 18 the answer.
- 19. MR. HARDIN: Having read it that way, I think you 20 prepared an excellent answer.
21 MR. BHATTACHARYYA: Answer, yes. 22 MR. HARDIN: And I think with regard to rules, the 23 question that is being posed, I think some of the 24 information you certainly gave us as a response to Question 25- 1, and along with your assertion that it's difficult to take t Heritage Reporting Corporation (\_/] (202) 628-4888 i
- - _ _ _ _ _ _ _ _ _ _ _- I
r.g.: ,o
- - - ~ -~- - -- - - - --
x Af; '.m y s L ' 44
& ,w l1 . individualcitemsfand quantify them.
t .2 ; ;MR.'PHILLABAUM: Yes. 3- MR '. HARDIN: However, I think it would be very
'4'- Luseful during the~ break if you could -- if you are aware of -5 -which specific items are being. talked aboutj if.you could
- 6 try; to the best?of your.' ability, to give'a qualitative-i
'7 answer at least or' semi quantitative answer to.what;the
- l. ,
8 impact of the 1068 recommendations are. a 9 MR. PHILLABAUM: Sure.
, ~ 10l MR '. HARDIN: And also the.second part of that. ~
11 question,.if any o'f those. recommendations for improvements 12 to the~ plant were not incorporated in the plant, what was
~ '
l- . . . 13 the rationale for not incorporating those? 14 MR. BHATTACHARYYA: Yes, that is the question.
~ 15 Thank you.
16 MR. SUH: Okay. Was that the answer that you had?- 17- Jerry, did you finish your answer? 18 MR. PHILLABAUM: Yes, I did.
<, 19. MR. SUH: All right.
20 Why don't we take a break and try to get back at
- 23. 2:307 We're running out of time real quickly.
.22 (Whereupon, a recess was taken.) - 23 24-25 Heritage Reporting Corporation j (202) 628-4888
o 45 kJ 1 1G1. SUH: Let's get started again. I guess where 2 we left off the rule had a question on the NUREG 1068, three 3 recommendations as I understand it. And, Jerry I guess you 4 might have something on that. 5 MR. PHILLABAUM: Looking at Table 8 on page 8-3 6 you have identified the improvement that you would obtain by 7 ADS initiation logic change. Yes, we did make that change, 8 to avoid the dependency of the high drywell pressure signal i 9 in compliance with TMI action on 2K-318. I 10 The second idea was to improve design to achieve 11 alternate method of HPCI/RCIC room cooling. We did that by 12 developing a Limerick specific procedure that identifies the 13 pathway they need to set up to accomplish room cooling, ( ). 14 which involves -- one thing it does involve opening the HPCI 15 and RCIC doors, plus some additional doors to establish that 16 natural circulation. When that was done, and the 17 calculation was done in the room heat-up rate, that was 18 found to be satisfactory to prevent the over-temperature 19 failure or over-temperature conditions within the room. I l 20 Again, one other thought is we have with the i 21 Limerick specific procedure again to react to a station 22 blackout situation, we have made a more realistic assessment 23 using that procedure as an outline of what actions would be 24 accomplished. A more realistic assessment of the Limerick 25 equipment capabilities. I just want to reiterate another (')
\./
Heritage Reporting Corporation (202) 628-4888 i
l 1
\
L.. 46 !; 'b 1l thing-is the increased capability from.that assumed in the 4 , / 2 PRA for the battery life. i [
- 3
- I think that hopefully addresses the different i
4 insights. i
- 5 MR. PALLA
- One question about what I thought was L
the third improvement, apparently made: the increased 6 7 availability..of containment sprays. j f 8 MR.'PHILLABAUM: What we have not yet done -- That [ 9 was a-recommendation. A credit for containment sprays'was 10 -never taken in the initial Limerick PRA. The idea was to I. 11 make sure that procedures were in place to use them 12 properly, that the operators could accomplish that. 13: Operators do receive training that is included in () '14 the ' Limeri ck specific EOP's on when they should be used to 15 control containment parameters. We still have not taken 16 credit for it until we try to -- We have not made a rigorous
- 17. analysis of the credit that would be obtained by doing that.
18 MR. PALLA: But, you have to get credit though in 19 -- 20 MR. SCHMIDT: The plant has always had the 21 ' capability to spray under the drywell. In the original PRA 22 there was no credit given for that. In the acsessment of 23 SAMDA's, the EOP's do tell them to spray, and we have given 24 them credit for certain classes of accidents. We have 25 assessed the dominant sequences in those classes. What's Heritage Reporting Corporation (202) 628-4888
'A' q ' ~l) 11 .the likelihood that the sprays would be available, i.e., for*
2 a's'tation blackout they are not available, or--their 3 availabilityLis dependent on the probability of. recovering 4 : power-from the time the core melts until you need the spray.
- 5. You need the spray at some later time.
6 And, we have made that assessment and adjusted the 7- risk profile, if you will, downward to account for the 8 benefit that you get from the sprays. For TQUX type 9 sequences, sprays are available from the beginning. They 10- are there, they haven't failed, and they are therefore Ell available. So we did factor it'in in those kinds of 12 situations. 13 MR. PALLA: So, you have given full credit -- y '14 MR. SCHMIDT: In the assessment of SAMDA, we have 15- given credit for sprays. Yes. Appropriate credit. 16 MR. PALLA: If they are not there, obviously it's "17 not viabte to include. 18 MR. SCHMIDT:- Yes. Yes. 19' MR. BHATTACHARYYA: One other comment I think you 20 did address was the support system failures contributing to 1. ! 21 class-one type sequence frequency estimates. Therefore, we I 22 were interested in these kinds of programs which would 23 increase the availability of this spray system, because it's 24 so late in the sequence you would have a lot of time you 25 could recover-something. Heritage Reporting Corporation O- (202) 628-4888
(; ; r ,
+
N s AB 43
.1 - MR. SCHMIDT: I'think this was the gist of 2 accident. management.
3 MR. BHATTACHARYYA: Exactly. 4 MR. SCHMIDT: And, I think to some extent that has 5 been accomplished. We try to make a realistic estimate of' 6 :the benefit-the existing spray system would provide. And, 7 that's included in the base risk, if you will, for which the 8 SAMDA's are assessed. b 9 MR. PALLA: It wasn't originally in the -- E 10 MR. SCHMIDT: It wasn't originally in the PRA or-11 SARA -- 12 MR. PALLA: In the PRA now. 13 MR. SCHMIDT: It is in the current assessment. Is it in the November '887
. 14 MR. PALLA:
15 MR. SCHMIDT: November '88. stopped at core melt, 16 and we are talking _ post-core melt. In November ' 88 it was 17 just the core damage frequency, level one, internal events 18 only. It's something to remember. It was an internal 19 events only, level one PRA, and the level two part 20 consequence, fire, flood and all that -- no coherent update' 21 had been done since the SARA analysis. 22 MR. PALLA: Okay. Now you've got credit for its 23 effect on the containment. Have you taken credit for the 24 fission product removal? A DF factor of some sort? 25 MR. SCHMIDT: The major credit is it would prevent Heritage Reporting Corporation
- 0; (202) 628-4888 i -- - ___ A
49 1 l'C i j, ~yL
'%/ 1. failure,-and that's the big' benefit. I think that I ~
2 subsequent eredit~-- I think the.only credit that we really
. 1 o 3 assessed:is the fact that the' spray would prevent- I .4 containment failure, and therefore reduce the source. term. j J
Si In that'way, it.is true that the spray would remove. fission j 6 products. If you have a ruptured containment already and 7- you spray,.then there would be some benefit. That was not 8 given credit.for. 9 MR. SUH: Why don't we move to question number 10 four? The models used to develop the release fractions.for 11 each of-the accident classes; plant damage states; and to 12 develop on-site' consequence estimates.
.13 MR. KRICH: Okay. Jim Fulford from NUS? .14 MR. FULFORD: I'11 run through these pretty 15 quickly here. The three slides -- source, consequence and 16 the risk. The' underlying source terms were calculated using 17 1982 available technology and are recorded in the 1982 PRA.
18 Essentially the pressure temperature driving conditions were 19 calculated with a code called INCOR, which you might look
. 2 0. upon as some. kind of a predecessor of March 1, or a 21~ variation'of it. And, release fractions were done with p 22 CORRAL using primarily Wash-1400 bases and assumptions.
23 The consequences were calculated in the 1983 SARA 24 -- there's a typo there -- and were calculated in CRAC2
- 25. using the CORRAL calculated source terms, and the Heritage Reporting Corporation 3O (202) 628-4888 L
7----__-- I 50 I 7 x/ 1 containment failure modes -- got the two numbers 2 interchanged'-- based upon the '82 PRA, and the results from
! 3 that were characterized for the SAMDA assessments as 4 accident class conditionals. For example -- and, I will use 5 this example in the next question -- given the occurrence of 6 a Class 4 accident sequence, the conditional 50 mile total 7 population exposure is 2.7 x 10 to the minus 7 person-rems 8 per occurrence. So, a conditional values we got from the !
9 SARA. 10 MR. MARGULIES: Could I ask you a question, Mr. 11 Fulford? 12 MR. FULFORD: Sure. 13 MR. MARGULIES: This is going to be real specific, 14 Do you recall the interdiction criteria that were assumed in ( o) 15 the person-rem calculations? 16 MR. FULFORD: No. I don't. I would have to refer 17 you back to the SARA for those details. 18 MR. MARGULIES: That's important. 19 MR. PALLA: Has something changed since the SARA? MR. KRICH: Not in terms of this. No. Not in 20 21 terms of conditional risk. 22 MR. FULFORD: In terms of the risk, we use the 23 accident frequencies and what we call in other slides the 24 " current results", including the fire, flood and internal l 1 p 25 initiators. For example, the Class IV frequency is 1 times l Heritage Reporting Corporation (^>\ (- (202) 628-4888 l l
II g y - 51' ,1-
- 1. 10 to.the.minus.7 per year. We use the SARA conditional 2 accident class' consequences'as on.the previous slide, so the L
public risk is the; accident class frequency times the
, 3 k 4 ' conditional class consequence. For example, for a Class IV, H
,- 5~ :the'off-site exposure is the product of those two numbers,
- . 6 - 11.e., three' person-rems per year. That's the unmitigated 7 number.- So, it's '82 source terms, '83 consequences,'and 8 ' current' class frequencies.
9 MR. SCHMIDT: Unmitigated by any additional' 10 SAMDA's. Mitigated by whatever the. plant presently has. I 11- MR. FULFORD: That's our response to discussion of 12 item four. 13' MR. SUH: So, does that sort of answer your [f(f -14 question, Bob, about whether they took credit for scrubbing 15 from the containment spray system? I didn't know whether to 16 carry that as an open question. 17 MR. PALLA: No. I wanted to just understand what 18 we had here and whether it was different than the earlier 19 study. And, that's the answer. 20 MR. SUH: So, you got the answer? 21 MR. PALLA: Yes. 22 MR. SUH: Okay. Thanks. This question number
'23 five, how the' risk reduction benefits averted man-rem -- in 24 other words, averted man-rem estimates were derived for each 25 SAMDA and.in particular for the core debris control devices
< i' Heritage Reporting Corporation (202) 628-4888 c____- _ =__ _- .
$ < l$ ?
w
,s r 52
- 1 A)-- li -considered. A more complete. description of the techniques g
2' considered for core _ debris control. f5 3 MR. FULFORD: I'll continue with the answer to One is a detailed. c4 that. I am going'to hand out two things. p w 5 listing 1for each one of the SAMDA's; how the risk benefits 6 were achieved. And,.another onelis the two page sort of 7 explanation.
- f. 8 In general, for each SAMDA -- this equation.is on p , ; ; 9: the first.page.of the larger 15-page handout. But, in words, L -i 10 for each SAMDA'took'the risk reduction and its averted b
11- person-rems -- which is dependent both on the accident class
- 12 (i)1and the.SAMDA (j) -- is calculated as the accident .13 sequence frequency per year, based upon the class (i) times
- j ' 14 the probability of. mitigation -- and, I will explain this on j 15 the next slide -- by SAMDT., which depends both upon the 16 p ClassLand the particular SAMDA timos the difference in the 17 population dose -- unmitigated minus mitigated -- in the
' 18 class, person-rems. And, that depends upon the class, and 19- sum it over all the classes to give the reduction for each 20 -SAMDA. . 21' MR. PALLA: When it was mitigated did it generally 22 fdisappear, oriwas there just aspigned to a less severe --
23 MR. PALLA: 'Look at the last bullet here on this 24 one.- In the majority.of the cases, we said that it was most 25 -- it was completely effective so that there was no Heritage Reporting Corporation (202) 628-4888
%w_ u__ i____..___________m_ _ . . _ _ _ _ _ _ _ l
53 I ')
1 mitigated does. But, in some cases we did say that there L 2 were certain physical reasons for saying that we couldn't 3 mitigate all the dose. But, in the majority of the case, )
4 yes. 5 In the probability of mitigation we also put in an 6 assessment as to whether or not it actually would be 7 effective. We tried to judge whether it would be effective 8 and then, having got that out of the way, looked more 9 physically at what the degree of effectiveness would be. 10 And, on the -- 11 Let me ask you a question. Now,
! MR. PALLA:
12 accident classes are composed of many sequences, and a 13 mitigative measure may apply to really just a subset of i 14 those sequences in a particular class, or in perhaps several s. 15 classes. Did you sort through the leading sequences within 16 the bin and assess the effectiveness of the SAMDA for the I 17 sequences and then work from there? 18 MR. FULFORD: To a degree, but the accident 19 classes are -- these are mitigation devices for when core 20 melt has occurred. And, a lot of the differences do wash 21 out by the time you get to this point. We did look at -- 22 MR. PALLA: They are not all purely mitigated. 23 MR. FULFORD: They are not all, and that's why 24 some of these are not zero. The details are in that 15-25 pages where we did that. (] Heritage Reporting Corporation
\ss (202) 628-4888 l
L -
I J 54 1 ') Y 1- MR. SCHMIDT: Remember these are, in almost all (
)
2 cases, new devices added to the plant rather than the use of I 3 an existing system. Therefore, dependency on -- for j ' l 4 example, the heat removal system has got its own AC power, i 5 so you don't have to worry about sequences in which there is 6 no AC power available. So, that reduces that kind of 7 consideration. It still was in mind. They knew what the 8 dominant sequences were in each class when these assessments 9 were made. 10 MR. PALLA: So, you basically treated them as a 11 class and not really as a sequence. 12 MR. SCHMIDT: We didn't go through sequence by 13 sequence. When we looked at the effectiveness of the spray n 14 system to get to the baseline risk -- the existing spray -- () 15 when we did that, we looked at the dominant sequences 16 because here it was the effect of the spray and existing 17 equipment capability, and we had to look at what was the 18 availability of the spray given each sequence. But, for a 19 purely mitigating device that has almost no dependency on 20 the existing plant, then the purpose of the grouping is 21 indeed to have a common phenomenological base under those 22 sequences. 23 MR. SUH: I guess I'm hearing two different 24 things, but the way I am understanding it from Bob and 25 yourself is when you needed to go to the sequence-specific (N Heritage Reporting Corporation
\ (202) 628-4888
55 A' 1 level you did. You went down to the sequence level. But 2 when in your judgment you decided you didn't have to go down 3 that fine you didn't. You treated it by class. 4 MR, SCHMIDT: That's right. 5 MR. SUH: So, in essence the answer is you did do I 6 it by sequence, because you determined.you didn't need to in 7 certain cases. 8 MR. SCHMIDT: Right. 9 MR. SUH: Which has a follow-on question. I don't 10 know that you have the answer to this, but in NUREG 4025 11 they assumed perfect mitigation, I believe, for a number of 12 their SAMDA's and the question is: Do you know whether they 13 just treated accident classes, or whether they treated r (m) 14 accident sequences? You may not have an answer at all. 15 MR. FULFORD: I don't know the answer. If you 16 want it, I guess we will have to try the 4025 -- 17 MR. SCHMIDT: I think they treated functional 18 sequences, but certainly not -- TQUX/TQUV -- but, they may 19 have done it in class. I really don't know. I can look in 20 the report and find out. 21 MR. SUH: No. That was my question. I was just l 22 asking if you knew. , 23 MR. HARDIN: When you assigned these mitigation ] 24 probabilities from .99 down to zero -- I was just curious -- 25 did you have a group of people take part in that and take an t) Heritage Reporting Corporation () (202) 628-4888 , l l l i I
l bgc- 56 ys . '4 SI .1- average'?..How did you do that? Was'that done by one person? 2 MR. SCHMIDT: It was done by one person with 3 review by a couple'other people. They are reasonable. This 4 was not a scientific group solicitation process. 5 MR. HARDIN: Did you find anything-that you 4: 6~ assigned a value of less than .5? 7 MR. SCHMIDT: Yes. 8 .MR. HARDIN: Do you recall? 9 MR. SCHMIDT: I think Jim's got a table of those. 10 MR. BHATTACHARYYA: Excuse me, Bob. When you 11- presented some PRA results to the ACRS members last time -- 12- Yesterday I was going through some slides. I noticed that 13' you do have an expert panel in Philadelphia Electric Company. () 14 who is looking into some other PRA changes; PRA type 15 analysis. Given that, I am wondering -- this kind of SAMDA 16 analysis -- was it reviewed by that expert panel? I think 17 there are three or four people in that panel. 18 MR. MARIE: The analysis was reviewed by at least i 19 one other member of that panel. Bob is one member of the 20 panel and Ed Burns, who is another, member, also reviewed it 21 and provided his input. 22 MR. BHATTACHARYYA: So, anyway, it's not really 23 one man assigned the probabilities. These are sort of a 24 consensus of at least two or four distinguished individuals. 25 MR.. MARIE: Yes. Yes. Heritage Reporting Corporation ! g (202) 628-4888 l l 1 l
o' O .57 g, A[ 1) MR. PALLA: Was the analysis -- 2 MR. BEATTACHARYYA: If you had different opinions F JL Lyou had to work it out. I: X 4 MR.'PALLA: Excuse me. Was the analysis evaluated-
.5 .by'them,'or just.the submittal?
6 MR. FULFORD: I might comment that Ed, whenever he 7'- had a question, called uscup:and asked us for the reasons 8 why we did this. ~So, I would say it was the deeper. reasons' 9 for'it. 10 MR. KRICH:- I think there was.a lot of t-11 interaction. I'm not sure -- 12 MR. FULFORD. I would also add that they were, in 13.. my opinion,.done in such ways to give the'most benefit to
'14 the SAMDA.
15 --MR. PALLA: Yes. But, I was just asking was the
~
16 document they reviewed the final document that they gave an c 17 overview kind.of a review to, or whether they participated 18 in the whole process. 19 MR. FULFORD: Participated in. 20 MR. SCHMIDT: There was a discussion of what we 21 were doing, and.the' credit given. There was not a detailed i 22 -- we didn't,have the detailed calculations, but did have an 23 intermediate product meeting. 24 MR. BEATTACHARYYA: Did you have a chance to talk 25 to people like the sort of people in the country, I guess
^' Heritage Reporting Corporation ' (s (202) 628-4888
58
,~. ;\ 'j ' 1 Jamous people in the D.C. area?' Like FNS and Castonburg?
2 People who have done previous work? Have you had any chance 3 to talk to them when you were doing such analysis? 4 MR. SCHMIDT: Not exclusively about this analysis. 5 No. We have talked to them about these-kinds of problems, 6 but not about this work. 7 MR. BEATTACHARYYA: I'm talking about the 8 effectiveness of this mitigation. 9 MR -. SCHMIDT: No. We didn't. We had no 10 discussions. 11 MR. FULFORD: This is an example to explain the 12 15 page handout. The first one is for the rubble bed core, 13 and the other one is for the dry crucible core retention n- one, two, 14 system. It's given by the accident classes:
!v) 15 three, and four. And, Pm was the effectiveness -- not the 16 effectiveness, the probability, .25 and .25 and zero and 17 zero for the rubble core. For example, for class two, it 18 has no effect because the over pressure containment failure 19 still occurs. We said it wouldn't affect that class.
20 We assessed that it was only a 25 percent chance 21 of success for the class one and class three, because we 22 thought that some debris can still remain in the drywell 25 with that type of design and still cause containment 24 failure. So, in those cases where the probability was zero, 25 we didn't assess what the mitigation was, because it was (~N Heritage Reporting Corporation
's-) (202) 628-4888 .
l
--^-----.m. _ _ _ _ _ _ _ _
7, ; . . , , 37 nn .
- )
; d ~ ?O ,7 .
59 G ?
'i s .1 irrelevant.
2' In the:other two cases, we said it wasl fully. .
}
- 3 effective. I introduced this mitigation effectiveness l
. i 4- factorLjust'for clarity. IIf it's fully effective, it's one. .5: If it's not1 effective, it's zero. And, it's defined there e,,
6 -at.the. bottom of the pagegin' terms of the equation on the
'7L mitigating'page.
8 And, then the-risk reduction factor has in it.the 9 accident' frequencies and the mitigated and unmitigated
;10 population-doses'from'the formula on'the first.page, and the
, 1 i E 11 data:that'sLgiven? ~in the other handout. 1
.12- MR. PALLA: Did you assess any credit if the 13' device' delayed containment failure? It sounds like you did h) 14, ..not.
15: MR. SCHMIDT: No.
'16 MR. FULFORD: I can't think of a situation like
- 17. that.
18 MR. PALLA: Well, for example, an ATWS vent or a 19 vent that does' release some -- 20 MR. FULFORD: We took'it beyond the time horizon 21 of the PRA. .I don't think we followed on to that. In the 22 . case of class four for the dry crucible, we have an
'23 ' 'over pressure and over-temperature containment failure where 24 we believe that the sprays that are associated with that 25 concept would'still be. effective. So, we reduced the source p
Heritage Reporting Corporation q (202) 628-4888 4 1
ap.; ;.. ,
? L g' <
60-Y f>}
"4/ .. ' 11-- -term inithat case, .because it's' highly likely they will be
- 2. : effective. But', the sprays were only, partially' effective in k- 3. that case. '
4 MR. SCHMIDT: That other handout.goes through each'
~
[ 5 one of the'SAMDA's; discusses each one.of the classes; and' 6~ the parameters for each class. 7 MR. SUH: Does'that include a more complete n 8 Description of the core debris control techniques? g 9 MR..FULFORD: No. I interpreted that'to mean in i 10 terms of the risk reduction benefits. If you are'asking for l .. 11 more description of.the core debris, there will be another
- n 121 material from Bechtel in a'later question.
L '13 ~ . MR . SUH: .Okay. 14 MR. PALLA: If you'have more information -- 15- VOICE: Like designs and things like that. 'Is 1' 6 - that what this last sentence' refers to? 17 .MR. SUH: For question five.
.18 VOICE: Well, we will be providing more --
19 MR. SUH: . Well, I just thought I would -- What I J20. was handing out was -- Oh, okay. 21' .MR. KRICH: I was wondering where you got this 22 from. The gentleman back there gave me this at 23 MR. SUH: L 24 the break, and we just made some copies. 25' VOICE: The information I am referring to is in
,& Heritage Reporting Corporation 1
V. (202) 628-4888 c
1 i 61 Y' 1- that document. 2 MR. KRICH: The primary document should arrive 3 today, f 4 MR. SUH: Oh, sure. I had assumed it was the same 5 thing. It's different?
'6 MR. KRICH: It's the backup.
7 MR. SUH: Did you want to hand this out? 8 MR. KRICH: Yes. U
- 9. MR. SUH: Oh, 20 MR. SCHMIDT: We have answered five now?
11- MR. SUH: If you guys don't mind, I would like to 1:2 move to question 11. I was waiting for these handouts to be 13 delivered. Maybe if we could please move to that one. Let
. c) f l'4 me read it while that's -- Item 11. Underlying assumptions 15 and bases for cost estimates as appropriate for each 16 mitigating system cost for the following elements.
17 MR. FELD: Gene, I wonder if I could answer it. 18 MR. SUH: Sure. 19 MR. FELD: Since preparing this question, I 20 received the Bechtel cost estimating study. Which I was 21 able to go through, and my feeling is at this point that it 22 might be more beneficial for my time to ask some very 23 specific questions about some of the underlying subsystems, 24 just to get some very general clarification and better 25 understanding about the assumptions that you have embedded S Heritage Reporting Corporation i tJ k (202) 628-4888 I l l
[ 1 m , k K' p( . ~jj 62 '
\/ '
11 Ein.your report; And, maybe I can simply review this and we ] l y 2' ,can.-- 3l MR. SUH: Oh, okay. Unless they want to go 4 .through. l 15 . MR . .FELD: If they would like.to go through it. 6 .. MR. SUH: Part of it anyway.
- 7. MR. KLEIN: What my plan was is just.to give you a 1 8- brief summary of what's in there so that -- j l
9 MR. FELD: All right. Why don't you do that, and l
-10 .then I_can hopefully follow up.
11 MR. KLEIN: All right. Basically, the base-12 . assumptions and general descriptions of the design concepts- . l 13 are-in-the main report that had been previously submitted to
! /~ '
j 14 you, and that's what you have had a chance to look through 15 already. 16 What this supplemental report is is to give an 17 additional detailed breakdown correlated to the specific 18 points that you had asked about in question 11. We weren't 19 able to provide all of the facets that you were asking for 20 because we had not developed our cost estimate in the same j
-l 21' _ format that you had asked the cost to be formatted in. So, .j 22 we'have gone'through and done an explanation of how we have l 23 incorporated those costs into our estimate, even though we l 1
24 haven't been able to break it down in the same itemized l 25 fashion that.you had asked for. Heritage Reporting Corporation (202) 628-4888 j i l l l I l
r 63 {~}
\- / 1 We have included at the front there a general 2 description of the process that we had went through in 3 developing this estimate.
4 MR. FELD: I think the questions that came to my 5 mind when I went through your more detailed report would 6 most likely still be valid, even given the submission that 7 you just made. So, I wonder if I could just turn your 8 attention to page 36 where you basically run through the 9 underlying assumptions that you have employed in your cost i 10 analysis. And, I should emphasis that I am really not 11 interested in exhaustive detail in response to these 12 questions -- just to give me a general feeling of some 13 magnitude or the treatment that you have employed.
.I ) 14 At the fourth bulleted item down you indicate that 15 you have applied productivity adjustments for certain labor 16 activities. And, I am trying to get a sense of the 17 magnitude of those adjustments -- whether we.are talking 18 about a 50 percent premium, or a factor of two, or how large
[ 19 a -- these productivity adjustment factors that you have 20 employed, do they tend to significantly alter the labor 21 hours associated with tasks? l 22 MR. KLEIN: Jim Hearn is our lead estimator for 23 this effort, so let me ask him to answer that question. 24 MR. HEARN: Okay. In evaluating the productivity l 25 adjustments, we developed essentially two factors for ['; Heritage Reporting Corporation K- (202) 628-4888 1 l l' l. l l l E .. _ _ _ _ _ _ _ _
p [" ' - 64 g Ep//
^- '
l'
~ -working first of all inside the PAB, the security fence. .?
2' ;AnyLwork'inside=that fence'was pro-rated up at 10 percent - - 3 an-additional.10 percent. Work that required' dress-out: b' 4- protective: clothing; potential use of masks; and other 4 I 5~ precautions for contaminated situations were marked up an l . . E < 6 additional'50 percent. e i 7 M1. KLEIN: All right. g 8 MR. HEARN: Now, to further clarify,.we also had i
- 9. to go through the estimates and determine what work, first 9
10 of all, was'insid'e the fence; which work was outside; and, 11 of the work that was inside, hcw much of that was in the 12 building in an area that would be considered contaminated
- 13 and require.those adjustments. So, it was all built-on
- 14 various: percentages of the total work.
15 MR. FELD: Okay. Fine. I guess the next bulleted' 16' item refers to employee training efforts, and I am trying to 17 .get a general sense as to how large a dollar amount you' 18- attributed to training activities and, more specifically, if 19 I go to your cost estimating tables that follow these 20 underlying assumptions, I am curious as to where this 21' training cost was included -- in which cost category? 22 MR. HEARN: Okay. If I can direct your attention 23 to the handout that we brought with us today. If you look = 24 on, for example, page four. It's hand-written at the bottom-25 of the page. It says section H training costs -- so, this (~ : Heritage Reporting Corporation
\
(202) 628-4888 i
v . 65 1 1 .is.for.' option.A-1,'which is the heat removal. You can see p; 21 it says that the Bechtel general employee training is g-
-3. included in:Section C of this-particular document, although I '4 . we'did.tell you how much money we put in the estimate.
H 5 In this' case here, it's $552,000, for this
- 6. particular option. Now, you will find that each one of them
'7- .has a line item like that, and we give you that detail in g 8' 'the body of this report.
p 9 MR.-FELD: All right. And, when you say it's
.10. included in Section C -- or,. in case of PECo training in 11 -Section J --
12 MR. HEARN: .I call that direct labor. Direct 13 manual labor. It's included in those dollars. So, if you 14 .look back on page three -- in Section C under labor costs, (}
-15 under direct labor you have the total dollars of 16 $10,519,000. It's in that figure there.
17 MR. FELD: -All right. And, when I want to take 18 this total labor cost for this particular option and relate 19' it back to the tables that appeared in the primary report, [ 20. am I'to assume that these labor costs are embedded in the 1-21 construction --
'22 MR. HEARN: Yes.
23 MR. FELD: -- description categories? 24 MR. HEARN: Yes. That's accurate. L 25 MR. FELD: Okay. The next bulleted item that I am I' l: ( ; Heritage Reporting Corporation (202) 628-4888 1 l' I
-----_a_ . . _ _ _ _ _ _ _ _ _ . _
e 66 ,s \_/ 1 interested in is the direct' input from PECo staff. Again, I 2 am curious as to where these items enter in the cost 3 categories that appear in the main report. Am I to assume, 4 for example, that since they are input from PECo staff and 1 5 presumably reflect PECo's effort that they go into station 6 and owner costs? 7 MR. HEARN: Yes. That's correct. 8 MR. FELD: All right. So, even though there is a 9 construction element here -- or a field engineering effort 10 here -- those elements ard all subsumed within the station 11 owner category. 12 MR. HEARN: That's correct. And, if you look 13 again on page four, under item J, "Other Costs" -- We have a () 14 section on here PECo costs. I believe you should see the 15 very same categories there and the associated dollars that 16 were estimated for each of those. 17 MR. FELD: All right. Fine. And, is it possible 18 for you to give me a brief description of the types of costs 19 embedded in two of the elements -- I and C and rad waste. 20 What types of activities are captured there? 21 MR. HEARN: Okay. I and C stands for instruments 22 and controls. It's mostly calibration of instruments and 23 controls. This also would include a portion for training; 24 procedure writing; and procedure revisions; and that's about 25 all I can recall from I and C. ("% Heritage Reporting Corporation ( (202) 628-4888
bs: ' [',' f-
... , 67 / . ~
k_/. ( :1-MR. FELD:- All right.
- 2. MR._HEARN: Now, rad waste on the other hand F
l 3- includes costs for laundering of protective clothing;
'4. -handling of rad waste materials -- I would assume some 5 decontamination' materials are in~there also.
6' MR. FELD: All right. Again referring back to 7 -your. cost tables, Jean you tell me which cost category does i 8L PECo's health physics support fall into,' and whether h 9: occupational exposure has been estimated and included in
, - 10 .these estimates?
11 MR. HEARN: Okay. The health physics costs you 12 will find.under the station owner category. Likewise on. 13' .page<four you will see that there is a line item for health
) 14- physics there. So, it's consistent with your other 15 question. .16; As far as the cost for exposure -- no. That was 17 'not'a portion.of the estimate per se. We did not establish - 18F a cost fi'gure and likewise an exposure generated from'doing 19' the work. So, we have not-included that in the estimate.
I 20 MR. FELD: Do you have e.ny sense of what kind of 2 11 occupational exposure is associated with these different 22 options? 23 MR. HEARN: Not at the moment. Although I believe I we have been asked to pursue that'for you. 24' 25' MR. FELD: All right.
. Heritage Reporting Corporation !
(202) 628-4888 i
68 A 1 MR. .KRICH: Well, we are pursuing that, I think, ! 2 in terms of the recent order and we have also -- at least i 3 within PECo -- have' looked at some numbers based on the unit 4 one, measurements taken during unit one operation, or during 5 unit one refueling. 6 MR. SUH: When do you think you will have that 7 information available?
- 8. MR. KRICH: Certainly before August 2nd.
9 MR. SUH: I was going to ask on these -- and, I 10 guess you were already ahead of me, but we were very 11 ' interested in how many hours are required for each of these. 12 Is that shown here, by the way? 13 MR. HEARN: Let's look at, I believe, it's page 14 three. Under labor costs, I believe we were requested to
)
15 identify the hours; the labor rates; and the costs. And, 16 under Section C you will see direct labor and indirect 17 labor, the total of which is, in this case here, 406,953. 18 That's the total manual labor -- the craft labor involved 19 with performing this particular SAMDA. In addition to that, 20 we will generate additional field non-manual labor for 21 130,000 hours, for a total labor effort of 537,000 hours in 22 this case. 23 MR. FELD: But Gene, that is not the hours that 24 they are in a radiated environment. J 25 MR. HEARN: Pardon me? I'm sorry. I didn't -- l
/^\ Heritage Reporting Corporation l
( .) (202) 628-4888 l
- . - _ .__ _- ____ a
a: .
. ~ . 69 o.
H -') 1 , MR.-FELD: No. .I~think you were trying'to.get at
-2 :these numbers to do something.with occupational exposure, d '3- and these hours do not necessarily correspond to the work : , I 4 that's'being.done in'a radiated environment. That would be 5 some' smaller' subset of that. I 6- MR. HEARN: That's correct. That is a percentage 7 of these.
- 8. MR. SUH: Is that broken out-anywhere in here?
9 MR. HEARN: No. It is not. 10 MR. SUH: Okay. And, similarly, I guess the same
- 11. answer for -- I had a specific question on that. So, Ron, I 12 take it you are looking for the dose rates that might be
.13 seen for the. work -- the number of hours in a particular I '14 area of containment and so forth.
15 MR. KRICH: Yes. 16 MR. SUH: Okay. Let me just leave it'at that..
'17 MR. FELD: The last bullet on page 36 speaks of l18 the regulatory costs. And, I would like to get a 1 91 qualitative sense of the types of cost elements that you 20' have taken into account in that cost category. '21 MR. KLEIN: That was just a percentage base cost.
22 We assumed that to cover all the regulatory costs, which 23 .would include NRC costs, as well as Philadelphia Electric 24 Company's costs in developing things like tech-spec changes 25 and such. That.would run 25 percent of the overall Heritage Reporting Corporation
.O+ . (202) 622-4888
I 70
<m ' .]
1 engineering cost. 2 MR. FELD: In addition to tech-spec changes, did 3 you envision any other types of regulatory costs -- I guess 4 meetings and so forth? 5 MR. KLEIN: Right. Meetings, discussions -- we 6 assumed we would have to go through a detailed approval 7 process for each of the SAMDA's, and have fairly extensive 8 discussions with the NRC on the design concepts. Probably 9 end up going through at least a two phase approval process 10 on that with the NRC. 11 MR. FELD: On the top of page 37, you provide your 12 assumptions on replacement power cost. I would like to just 13 make sure that my interpretation is correct. If the down () 14 time associated with the modification is 13 weeks or less, 15 you took no replacement energy cost penalty. 16 MR. HEARN: That's correct. 17 MR. FELD: And, if it exceeded it, you -- For 18 example, if a six-month down time was identified, you 19 deducted 13 weeks for the scheduled outage and only took the 20 difference. 21 MR. HEARN: That's correct. Multiplied by an 22 average figure per day. , 23 MR. FELD: Can you tell me what that average 24 figure per day is? 25 MR. HEARN: We used an average of $850,000 a day
/~h Heritage Reporting Corporation kl (202) 628-4888
L. 71'
~
- O MM 11: for the purposes of assessing it. We. realize that. figure L2. '. changes from year ~ i?89 through 1990 through year ' 94, over-L
- 3 that. time frame.
4 MR. FELD: Is that for both reactors? 5 MR. HEARN: Yes. 6: MR. FELD: It's a 800,000 -- 7' MR. HEARN: Per unit.
.8 'MR. FELD: Per unit.
9 MR. HEARN: Just for one unit. 10 MR. FELD: Per unit per day. I had some 11 difficulty in reconciling the values that appeared for the 12 .' . few. options in which replacement energy cost penalties were 13' being taken. I must admit I had to eyeball the amount of. () -14 ' incremental'down time based on your schedule there. But,' it' 15' did appear to me that the dollar amount per day was
- j. 11 6. consistent between the different' options. Assuming.that one
, 17 of them required one month down time; another three; another
'18 six -- that'was how I basically eyeballed it. I couldn't
- 19. identify anything close to a constant dollar amount per day 20 that would produce the number that you are providing.
21 MR. HEARN: Okay. In the answer to your Section I 22 of the request that we received the other day -- replacement 23 energy costs -- you asked us to respond with the number of 24 days and the associated costs. So, you will find for each 25 of the options we have given you the number of days -- ( Heritage Reporting Corporation N. (202) 628-4888
t 72 7-
.1 MR. FELD: All right. Can you refer'me to the.
L 2 page? l 3 MR. HEARN: For example on page four'-- 4 MR. KLEIN: Turn to page 11, because that's the 5 first SAMDA which has a replacement energy cost. 6 MR. HEARN: Very good. On page 11,.under item I, 7 replacement energy costs, number of days 28 days -- under 8 the associated costs 23,800,000. So, that 28 days relates 9: directly to that particular schedule for that option. 10 MR. FELD: And, if I took 28 days times $850,000 11 times two, I should derive the $23,800,000 figure? 12 MR. HEARN: This should be just for one particular 13 unit. ri 14 MR. FELD: This is one unit? im) 15 MR. HEARN: One unit. Yes. 16 MR. FELD: All right. I will go back. Moving 17 down page 37, you indicate on the next bulleted item that 18 you are going to multiply the operation and maintenance 19 costs by 40 years. I am wonde. ring why you are doing that if I 20 you are presenting this on a present worth basis. 21 MR. HEARN: The method that was used to calculate 22 the operating and maintenance costs -- we went through a 23 fairly detailed review by department with Philadelphia 24 Electric to determine how much cost would be generated per l 25 year on the average. And, the method used is we used that l (~} Heritage Reporting Corporation (> (202) 628-4888
- 73 >
,3
(-) 'I as the average for all 40 years and used a present worth 2 calculation for what it would take to finance that cash flow 3 now to support 40 years of supporting the system. 4 MR. FELD: Can you give me an indication of what 5 types of activities you were costing and envisioned under 6 operation and maintenance, aside from perhaps annual 7 maintenance and inspection? Were there any other elements 8 that were being captured there? 9 MR. HEARN: Of the items considered I can tell you 10 the departments, for example, I and C department; rad waste; 11 test engineering; and maintenance. I would say maintenance 12 is probably the majority of the cost. 13 MR. KLEIN: That combined with rad waste. (} 14 MR. HEARN: Yes. Which is what we expected to see 15 generated. Maintenance can cover a pretty wide spectrum of 16 activities inside the plant between maintaining pumps; 17 repairing pumps; fixing leaks; cleanup; painting -- there's 18 a whole list of items we could probably go through. Again, 19 this is based on an evaluation by the plant personnel of 20 what they felt it would take to maintain these systems given 21 their knowledge of other similar systems in the plant, and ' 22 their knowledge of what their operating budgets are now. 23 MR. FELD: The labor rate that you provided in the 24 main body of the report -- and I guess the ones that appear 25 in the submission of yesterday and today -- are these to be Heritage Reporting Corporation ' N (202) 628-4888 I l i l
E ' V' to v m 74 g s v-o interpreted'as fully;l'oaddd labor rates? 1-
. 2L MR.-HEARN: That's our cost per hour . It ' includes, 3- --
4' MR. FELD: That includes fringe benefits and 15 foverhead:-- 6 MR. HEARN: -- fringe' benefits and union costs --
- 7. those types of things.
8 MR. FELD: A little bit further down you speak , 9 about contingency. factors, and I am curious if they have 10 only.been applied to the core catcher and crucible designs, 11 ;or contingency factors are applied across all options. 12 MR. HEARN: All options have:a contingency 13- applied.
.() ' 14 MR. FELD: And, is it possible for you to give me 15 a rough idea.of the percentage of the total cost that's.
16 being. captured by contingency factors?
- 17. MR. HEARN: We have given you the dollar figures 18 .in the evaluation.
19 MR. FELD: All right. Fine. 20 MR. KLEIN: In general that was 25 percent.
' ;21 MR. FELD: Twenty-five percent?
22 MR. KLEIN: There were some exceptions to that. 23 MR. FELD: All right. When you say that you have 24 taken 10.25 percent escalation per year to arrive at the
- 25. present worth, is it another way of saying that you have Heritage Reporting Corporation 01 (202) 628-4888
i . , u-
~75" U ,/ l ~ ' discounted these costssat-10.25 percent?
2 MR. HEARN:~ Yes. j 3' MR. FELD: And now,'if I were to go'to the cost E4 tables in the main cost report, is it. reasonable to 5 interpret the procurement category as' incorporating all of' 6: the equipment and material costs associated with the 7 construction?' L 8 MR. HEARN: That's correct. 9 MR. FELD: And, the construction is essentially , 10' the labor installation costs? I 11 MR .- HEARN: That's correct. 12 MR. FELD: All right. Rod, I think I would like 13 to request a break. Oh, .I'm sorry. 14 MR. HARDIN: I was just curious to ask'a general {) 15 question. 16 MR. SUH: Please. Go ahead. 17 MR. MARDIN?: When you have done this cost 18 estimating, did you assume that any parts of these systems 19 are safety grade, or is it all assumed to be'not required to-20 'be safety grade? 21 MR. KLEIN: Well, we stated our assumption 22 concerning that in the report, and what we state is that 23 they are not safety related unless they would adversely
'24 impact an existing safety related system. Then it states if 25 we tie in our pool cooling system in our option we came up Heritage Reporting Corporation (202) 628-4888
__ __ ____a
76
,3 w 1 with, we tie in to the existing RHR section line. So, that 2 portion that ties into it down to the isolation valve will !
3 be safety related. Beyond the isolation valve, it would be 4 non-safety related. 5 MR. HARDIN: You accounted for those tie-ins in 6 your costing? 7 MR. KLEIN: Same for a seismic. Right? i 8 MR. KLEIN: Yes. Safety related; seismic category 9 one, up to the isolation point. Non-seismic, non-safety 10 related. 11 MR. HARDIN: You raised that question, John. Is 12 that -- 13 MR. RIDGELY: This is John Ridgely. Yes. That's 14 what they had said in their original submittal in response ( ). 15 to the questions, and that's the way it should be. 16 MR. SUH: And, John, the question about one unit 17 or two units -- is that pretty clear in your mind? 18 MR. RIDGELY: Yes. It is. 19 MR. SUH: Okay. 20 MR. RIDGELY: It's one unit. 21 MR. FELD: I do have one other general question,
- 22 and I know that you reference the R and D Associates report
\ 23 in your cost analysis. And, obviously when you costed what 24 may be identified as comparable systems, the cost t 25' differences between these two studies are quite dramatic. l I Heritage Reporting Corporation (]) ( (202) 628-4888
I
. .77 d 3N gs - ' l- Wouldlyou care to comment on the scope of the work.and the. .i '
2, Kspecificitycof the1 systems that you are costing relative to. 3 what R and D Associates.did? 4 MR. KLEIN: By scope of'the. work, the process we
'5 went through and the detail we got into in looking.at it?
I . 6 MR. FELD: I guess the' detail and for example,
'7: under'the first option where you identified a:need for a 8~ building, which I imagine is a fairly costly item. And, I 9 think if you'were treating this issue more generically you '10 may.not focus on those kinds of secondary types of cost 11 ' items.
12 MR. KLEIN: Let me start out by throwing up a 13 slide that is basically the first page on our supplemental (()
~
14 process,. just to tell you the process that we went through
~I '15 on this, and then I will get a little more specifically~to 16 it.
17 (Slide) 18 We started out with NUS providing the basic design 19 bases for each of these options. After a review of RDA and 20 looking at that in light of the PRA efforts done for 21 Limerick'. We also gathered much of the available SANDA 22 information via our licensing staffs. 23 We then assembled conceptual design teams by 24 bringing in people from our various Bechtel offices into 25 Pottstown who are familiar with severe accident mitigation a Heritage Reporting Corporation
\ (202) 628-4888 l .
l
{ o 78 y l' features; combine them with people familiar with Limerick . 2 ~ unique features, plant layout and such; and had a' conceptual J 3 design team that looked at a fair amount of detail within 4 the time frame allowable on this. [ 5- This was done in a relatively_short time frame --- p 6 two to three week time frame -- however, we wanted tc make 7 sure it was done in a consistent well-thought-out manner so 9 we could get the base well-meaning estimates available 9 within that time frame. 10 MR. KRICH: On the basis that we were trying to
- 11. respond.to your RAI within.the 30-day period.
12 MR.'KLEIN: So, the conceptual design teams 13 established the design concepts using the NUS input. They
) 14 did use the RDA report that was available, as well as the 15 other available information that had been gathered by our -16 licensing groups. These design concepts were then reviewed 17 by PECo and NUS to make sure we were consistent with what 18 was being done on the risk reduction side, and adjusted as - 19 ~ required.
20 At that time, the conceptual design teams worked 21 with our discipline engineers. We got our construction 22 engineers involved, as well as estimators, to work through 23 and come up with material quantities for the various 24 commodities; man-hour estimates; and schedules. 25 We also tied in with PECo station, as we had ' Heritage Reporting Corporation (202) 628-4888 l J l
.P, '.
s
. 79 I~ ~ ' \ .. 1 indicated,: to get their estimates on the' owner costs. This 2 data was then'taken-'by.our estimators and put'into cost o
3 estimates and a. draft report-issued whichiwas reviewed by. 4 Bechtel management, NUS and PECo; comments' resolved; and a E' 5 final report issued, which was submitted to you. 6 This. supplemental issue gives some of the 7 additional-details.that we had come up with as part of that 8 process. As an. example of the detail we got into, if you 9 turn to page five on the supplemental report, that shows how E 10- we did break it down by commodity -- coming up with. quantity 11 estimates for each of the' commodities, and then material 12 estimates associated with that. This is broken apart by-
-13 discipline, so for each of the SAMDA'snthere is a separate '14 . list of commodities by discipline.
15 MR..KRICH: I don't know if that answers the 16 question, though. I think what you are asking is why was 17 there such a big difference between our numbers and the R 18 and D. 19 MR. FELD: 'Yes. And, one of the things that I was
-20 trying to get a feel for was that the scope of the design 21 that was being costed here was far broader than what was 22- being looked at in the earlier report. I am trying to --
23 MR. KRICH: It's hard to say how R and D did their
-24 work, but I guess from what we have shown you, I think it's 25 a pretty thorough job, so we feel pretty confident about the L
Heritage Reporting Corporation T (202) 628-4888 l l
K U ~ 80-E 3. r 1 s U . 1 -work'that Bechtel has done. I can't tell you I feel 2- confident about what was done by R and D because I don't 3 know.- 4 bet. SUH: Any other questions? l 5 MR. FELD: No.
< 6- MR. CHELLIAY: Excuse me. I have a general 7 question. Do you have rationale to make this 10 percent 8 discounting? Can you describe some of the reason you 9 estimate.on this discounting business? .10 ' MR. HEARN: We have taken a couple of different-11- approaches in developing those percentages. The first one,-
12: I will tell you, is mostly from experience in having
.13 : performed work at a number of operating plants, and plants
() 14- under construction. This plant is currently under full
'15 security and will soon be radiation controlled, if not 16 already in most areas a radiation controlled situation. -17 We have compiled experience in that the 18 productivity losses involved with badging in and going 19 .through security -- you can quantify it in a number of 20 minutes per day lost going through portal monitors; through I
21 security frisks; having to have tools and equipment frisked 22 to get out of the building. There are very real 23 productivity losses involved with performing the work. i 24 In addition, we have' performed studies that have 25 gotten to a very great level of detail in evaluating how Heritage Reporting Corporation
.O (202) 628-4888 i
i
p i 81 I
/~s\'
s i i K/ 1 long it takes a person to dress-out; to frisk; to undress; 2 to handle a piece of radioactive waste. So, we have gone 3 into further detail over the years and have altered those 4 numbers, and they are used at a number of different plants. 5 MR. CHELLIAY: Thank you. 6 MR. SUH: I would like to take a break, and maybe 7 we can stop recording and come back in about 15 minutes. 8 (Whereupon, at 3:28 p.m., a brief recess was held, 9 and the meeting resumed at 3:52 p.m.) 10 MR. SUH: We apologize for the long delay. I 11 wanted to. pursue the question that I was asking about the 12 man-rem estimates. It is my understanding just looking at 13 the information provided on cost estimates that you did not () 14 have a specific line item for man-rems incurred in 15 installing or in maintaining the SAMDA's, and given that 16 that is a cost item that may be relevant to our review of 17 your submittal, I would like to just pose that as a 18 question. 19 When you do have the information, we would ask 20 that you please send it to us. I think the way I will do 21 that is just maybe in our meeting minutes that I will 22 prepare, or in some other form, I will just list specific 23 questions that have come out in this meeting. So, the 24 specific question is, for the installation of these SAMDA's, 25 what was the time assumed for installing it; the time
/~ Heritage Reporting Corporation
! k-} (202) 628-4888 t
P' b ,
~
82 D-l \s/ ^1 assumed'for maintaining it on a yearly basie; the dose rates 2 that one can assume for where that~ activity is taking place l l'- 3 so that we can arrive at some sort of man-rem estimates. 4 MR. HEARN: Can we clarify the word time? Are you 5 referring to hours expended? 6 MR. SUH: I suppose. Yes. 7 MR. HEARN: Okay. Rather - than duration in days. - 8 The hours is what we would most likely use'to generate -- 9 'MR. SUH: The hours are fine. 10 Sid, did you have a question'on this specific 11 item, or do you have enough information on that? 12 MR. FELD: Which item is.that? No. I'm fine on 13' that.
.() 14 MR. SUH: You are fine on that.
15 MR. FELD: I'm perfectly fine. 16 MR. SUH: Okay. We may have another staff member 17 come in'later and ask another question. He's not right 18 .here,uso if you will excuse us we will interrupt at that 19 point. Otherwise, I think -- 20 MR. KRICH: Excuse me, Gene, before we go on. I 21 think we may need to make a clarification, or have you done 22 it already? 23 MR. KLEIN: No. I would like to make a 24' clarification. 25 MR. SUH: Okay. Please. f' Heritage Reporting Corporation (202) 628-4888
o ; k i
= l' 'MR. KRICH: On the presentation that was:given.
f 2 MR. KLEIN: We were discussing replacement power
]
3 costs, and whether that was for a single- unit or: double 4 unit. All of our estimates -- both in the report we
) ' 5 submitted a-day ago, and the supplemental'. report are put 6- together on double unit costs, except:for the replacement 7 . power costs. We were_just looking at'it. We realized that 8 we have.not multiplied that by a factor of two to be.
9 ; consistent'with the other costs. So, we have underestimated 10 under replacement power costs, l'l' 'Three out.of the'10 SAMDA's that had replacement' 12 power costs. 13 MR. FELD: And, in those instances the replacement () 14 power costs that you.are showing are essentially half of 15 what they should be. I 16 MR. KLEIN: That's correct. 17 MR. SUH: Okay. Is that it? 18 MR. KLEIN: Yes. 19 MR. SUH: You had a question, I think. 20 MR. CHELLIAY: Yes. I was wondering to what 21 extent you have. considered the material and environmental 22 qualification of these systems, because my feeling is they 23 will be be quite esoteric, they will have to function under 24 a relatively harsh environment. 25 MR. KLEIN: Well, one of the assumptions that we Heritage Reporting Corporation 7 A (202) 628-4888
p
;_:-v j; g4 , 'R L IN,l - 1 putninto.the report ~that we submitted a day ago was that'the 2 equipment would not be environmentally qualified unless it 3 'could adversely impact . safety related equipment. By=that we 4 mean.we would not put it under the 10 CFR 50.49 program 5! that's. required for safety related equipment. As previously 6 stated, we would not consider this as safety related 7 equipment.
8 MR. CHELLIAY: I wasn't thinking of relevance. I 9 am more thinking of the selection of the material. 10 MR. KLEIN: In selection of material we did 11 ' consider that it would have to stand up to the environments
'12 that it-would see and would have to function in. And, in 13 some cases, we made some assumptions about what type of We weren't able to get
() 14 material could function in there. 15 into.a great deal of detail. And, we have called out that 16 in some of the uncertainty statements in there, that we have 17 assumed carbon steel piping is going to function in this. 18 We are not going to assume we are going to some exotic 19 material and base.our cost on that, but'rather more a 20 standard material for purposes of the cost estimate. 21 MR. SUH: Okay. Sid, did you get the discussion 22 you needed on item 11? 23 NR. FELD: Yes. There was a second part to that 24 which dealt with the other mitigation systems that the staff L 25: identified. L. L
. Heritage Reporting Corporation n (202) 628-4888 L
1
p
'85 H
L r-. D]- 1- MR. SUH: Oh. Okay. 'Could.'we.do item.12 now? 2- .Let me read it'. It's table 1 identifies a number of Mark II: 3 improvements under consideration as part of the NRC 4- containment performance improvement program, a discussion of
- 5. views on the risk reduction potential and expected costs for 6 each of these commodities.
7 MR. KRICH: .I think'between Al Marie ano I, I 8 think we can respond to this. You' realize that these are 9 items that,.as'we understand, are now currently under 10 . consideration by the NRC as potential improvement items. 11 And, we are not within the scope of the remand for what.the 12 ASLB has established to be the scope of the hearing. 13 MR. SUH: That's correct. As far as the ASLB
-(f 14- proceeding is concerned, they put out.a memorandum and order-15 which limited the litigation on the Intervenor's contention, 16 just the six that your June 23, 1989 submittal addressed.
17 MR. KRICH: Okay. And, the other part of that is 18 -- you understand that these are under consideration now 1 91 within the NRC. 20 MR. FALLA: As generic improvements. . 21 MR. KRICH: As generic. Yes. We understand that. T" ' 22 MR. PALLA: And, we were looking to see if you had 23 any from a plant specific point of view what you thought of 24 these. 25 MR. KRICH: What we will give you now is just
' Heritage Reporting Corporation -- (202) 628-4888
( ) t I $ . , L86 ffN ~ N 4 l' basically;where we are with respect to these items.then.
.- 2: .MR. PALLA: .-Okay.
6 AP ~ 3' MR.' MARIE: We haven't performed a cost' analysis g :4'- offthese specific items, so I.can't-give you any specific-
- c. ?
5 ' result'on.that,.and'I' don't want'to. speculate. However, I-I ' 6 can, provide-some comments on how some of these things might 7 pertain:to Limerick, if you care to have these comments. 8, First,. I note that the intent here was to take the t s 9: first few items in combination. We might benefit from a h: 10 little'more' explanation of the overall intent of this 11 combination before I proceed with my comments. 12 MR. SUH: Sure. John, would you?
' 13 MR. RIDGELY:. Yes. This is John Ridgely. I am 14- the one who created this list you are talking about. Yes,
- 15. .indeed,~ these are some of the items that we are considering,
'16 but1the question was.what are the generic items that we are -17: considering for Mark II's -- hardware'related -- that might
- 18. be cost beneficial for Limerick, and these are the' items I' 19- 'came uptwith.
20 And, the reason:for taking the first three 21 together, which is the diesel generator; a low pressure 22- backup water supply system; and a short RBCU for i 2MF -comparability is, for the severe actions that we are 24 basically.considering, is'a station blackout. A station 25 blackout with a loss of all AC power -- you need a motored Heritage Reporting Corporation (202) 628-4888
t-87-1 force, hence.the diesel. 2 .These, by itself, . with no other -- water supply or
-3 way of getting water in or out is not.much good. So, then '4 you.need a backup water supply system to put-water either 5- ~into containment sprays -- if the core is on the floor -- or 6 put'it.in a' reactor vessel if you can still put it in there.
7 If you are going to contain the spray route because in 8 ' severe accidents,.you assume-core on the floor, the Mark II 9 containments, as we understand them, are susceptible to 10 . failure by excessive water in the wet room with atmospheric-11 .over pressure.
'12 Therefore, we. feel that you will need,a way to get I ~13 the water out of-the suppression pool, and that's where we
() 14 think.short' operability in the RBCU to get the water out. 15 Also, there has been a study done which says that 16 operability of IUK:U could eliminate the need for all venting 17 for sequences such as TW, because it provides adequate 18 continuous cooling for the suppression pool. So, basically 19 that's why those three are lumped together, because they. 20- seem to need each other in order to be assured. In the Mark 21 I work that we did, we saw this dependence very strongly in 22 the analysis that we have done, and that's why I lumped the 23 three together, h O
'l 1
(_)- 87 1 force,.hence the diesel. 2 These, by itself, with no other --' water supply or 3 'way of getting water in or out is not much good. So, then 4 you need a backup water supply system to put water either 5 into containment sprays -- if the core is on the floor -- or 6 put it in a reactor vessel if you can still put it in there. 7 If you are going to contain the spray route because in J
'8 severe accidents, you assume core on the floor, the Mark II i
9 containments, as we understand them, are susceptible to 10 failure by excessive water in the wet room with atmospheric 11 over-pressure. I 12 Therefore, we feel that you will need a way to get' 13 the water out.of the suppression pool, and that's where we
)
14 think short operability in the RBCU to get the water out. 15 Also, there has been a study done which says that 16 . operability of RBCU could eliminate the need for all venting 17 for sequences such as TW, because it provides adequate 18 continuous cooling for the suppression pool. So, basically 19 that's why those three are lumped together, because they 20 seem to need each other in order to be assured. In the Mark 21 I work that we did, we saw this dependence very strongly in 22 the analysis that we have done, and that's why I lumped the 23 three together.
- 24 ( Continued on next page) 25 O Heritage Reporting Corporation (202) 628-4888 I
_ _~__...__._-_-_________-_A
88
..+
kN1 'MR. MARIE: Let me run.down each item and I'll v 21 offer some information. At Limerick, as you know, we have
.four diesels perfunit in an eight diesel plant,-in 4 , comparison 1with the. plant referenced here, which is a four .5 plus a swing' diesel.
L6. Limerick.has'five transmission lines with three 7 offsite power sources, four reliable sources, offsite power 8- sources. We do have on the' Philadelphia Electric system 9 numerous flatbed trucks with diesel generators. These are 10' located for-system usage-and would be available for Limerick 11 in the event of a long-term statien blackout. So there 12 would be a travel time of one or two hours to get them to 13 the site. And the --
;( )14 MR. PALLA: These are there. You indicated they 15 were there? Where are there?
16 MR. MARIE: If they are not onsite they're 17 located on our system. Okay? Normal location is in 18 Philadelphia. Now the ratings on the diesels'are from 200 19- to 250 kilowatts. We have not taken credit for these in the
.20 PRA.
21 One other item that should be included, I think, 22 and I would like to comment that these diesel generators, 23 relative to the cost, is, you must also consider as far as 24 the cost to maintain and test a long term diesel, which 25 would be a present worth of that long-term maintenance cost. Beritage Reporting Corporation (202) 628-4888
89
- V1 And.to. perform that cost --
2: . Moving on to the low pressure backup water L ' l3 supply, I'd like to note that we presently do have the 4 ability to tie the diesel fire pump via hookup to a fire 5' . hose to the RWCU for injection to the vessel when, at low 6 pressure. WeLpresently have a procedure to accomplish this 7 and the' operators are trained in this procedure. And we 8 have not taken credit for this material also. 9 MR. RIDGELY: Do you know what the delivered flow
, 10 rate of the vessel is?
11 MR. MARIE: We don't have a specific analysis to 12 to-find that value. We have estimated it to be on the order 13 of 200 to 300 CPM, but we don't have a calcuation for that. The capacity for that pump is 2500, but, of course, there's [( ) 14 t -15 line loss. So it may be substantial for removal of long
- 16. term KD. After the first two or three hours, when the K 17 power has been droped down to a percent of rating.
18 MR. PALLA: Can I back you up? Do you any kind 19- of, you talked about the diesels on the flatbed truck? 20 MR. MARIE: Yes.
'21 MR. PALLA: Would they, would anyone know what to 2 .2 do with them once you got them to the site?
23 MR. MARIE: Okay. There presently are not 24 specific procedures to tie them in. This would be worked 25 out kind of ad hoc on the spot, during the long-term. D d Heritage Reporting Corporation (202) 628-4888
I-90
'5 1 MR. PALLA: Okey.
2 MR. MARIE: Again, that is the primary reason why 3 we had not taken credit for it in the PRA. 4 MR. BARRETT: Than am I correct in saying that 5 perhaps the only cost might be the cost of having some kind 6 of guidance provided to your technical support people in 7 order to making them, reminding them that these things are 8 available, giving them some guidance as to how they might 9 best be used in certain situations, and then writing 10 procedures as to how you might go about tying them into the 11 equipment that they might -- 12 MR. MARIE: That's a possibility. 13 Okay, on the third item, RWCU, it's important to note that RWCU can take sunction from the suppression pool, ( )14 15 and that's why we' re somewhat confused by this. 16 MR. CHELLIAY: I got confused, too, and then I 17 read this, okay. 18 MR RIDGELY: Do you have any water system that 19 can provide -- that can take the water out and pump it 20 someplace else? 21 MR. MARIE: Yes. Water can be transferred to the 22 CST by way of HPCI or RCIC, to the condenser by way of 23 suppression pool cleanup pump, and to rad waste by way of 24 RHR. 25 There are procedures. Also relative to moving Heritage Reporting Corporation (202) 628-4888
W.
.. - 91 g 9'
- . T 1~ suppression poo1~ water, and performing a cost benefit,-one
!.1 2 .must also consider the downside risk'of moving potentially
- l. j 7
'3' highly contaminated water from the suppression pool. And if 4J this were.used for post vessel. failure, of' course you'd have E5 !. 'signficant' levels of activity being transported outside.
I 6' ; containment, and potential leakage to that enclosure. 7 MR. PALLA: Can I just back you up on your water 8- ' systems that you transfer with? The sequence that this-c ,
'9 option'is after, which is station blackout, are.these 10 systems going to be available?-
11 MR. BARRETT: HPSI and RCIC are steam- driven. f 12 But you won't, it won't handle.the station blackout that's 13 going to core melt, no. MR. MARIE: Well, vessel failure, yes. Not post-()14
' 15 core melt, right.
16 VOICE: Although I thought that these were being 17- mentioned as a possibility during PW sequences.. 18 MR. PALLA: That's true. PW is a very small 19 period.
- 20. MR. SCHMIDT: That's the purpose of having the 21- three connected. Item 1 is the diesel that brought power to 22 the others, that's why.the three are there. From an 23' operability standpoint of requiring power on it.
24-MR. FULFORD: Have you looked at the cLse of
' 25 where you might vent air rather than water?
Corporation Heritage Reporting (202) 628-4888 l
}
d 92
.)
hw/ .1' MR. RIDGELY: The problem.with venting air is the 2 very large uncertainty when suppression pool bypass will 3: occur.- It could be bypassed almost immediately after the 4 fail to never. bypass at all. And the problem would then be-5 any place. :You. won't be able.to that unless it's clean air.
-6 But the problem as I understand it, which is try and get the 7 water.out, is as I understand the draft four of EPGs, 8 there's guidance in there that tells the operator that'he needs to terminate injection into the drywell, .get the water 9
10 level just above some level. And this is, I presume, 11 because of the potential failure mode. And so the object 12- here is to never get the water at that level but'to 13 continuously spray in the drywell, fission products and remove heat, and if it's core on the floor, potentially
. )[14 '15 arrest core. flow and provide core cooling.
16 MR. MARIE: Yes, and by the way, when you talk 17 about pumping out water with nasties in it, the object here 18 is to prevent tank failure. You're concerned about where 19 you it, just can't dump it out on the street. But with the 20 object of saving containment, that seems to be a secondary 21 concern. 22 ,MR. RIDGELEY: Do you have any thoughts on Item 23- 47 24: MR. MARIE: Yes, Item 3 -- I was just mentioning 25 that that would need to be considered though as a down side. Heritage Reporting Corporation (202) 628-4888 l-t-
.4 -
J. s 93-O
. k/ il MR . RIDGELEY:'-Right. I 2- LMR. HARIE: In Item 4, I'm assuming you mean the 1 ~F : spent fuel pool?-
4 HR.~RIDGELEY:' No. 5- MR. MAR 1E: Wait a minute. I'm sorry. What 6 pool?
- 7- MR.'RIDGELEY: From the spent fuel pool and'put 8 'it over the.drywell bin?.
'9 MR. MARIE: Yes, right, okay. The description 10 here matches the suppression pool. Thank you for the 'l l , clarification. Presently, in the system, you can't flood 12- that' area because of HVAC openings to that region. So it 13 would require modification with the ability to remotely
( f 14 ' block and isolate those in the event of a sequence. So that
~15: would be, you know, additional modifications. It's hard to 16 do that.
- 17. MR. RIDGELEY: I'm not sure I understand what you 18' just told us. The.drywell head, during normal operation,
.19 and it's on place'and the bolts are tight.
20 MR. MARIE: Yes. 21': MR. RIDGELEY: Above that is the missile sheet?
- 22, MR. MARIE: Yes.
R .23 MR. RIDGELEY: Between that space it's serviced 241 by the ventilation system?
~ .25- MR. MARIE: Yes, that's part of secondary Heritage Reporting Corporation L (202) 628-4888 h
r-
'. - _ - _ _ _ _ _ _ _ ..t____-------.-----
1
~ 94
- /%"
1 containment'and it's maintained regularly. r 2- < lGl.,RIDGELEf: Okay. I learn something new ]t 13 everyday. Do you know if that is the same of~all' Mark-II? l 4 MR. MARIE: It's t' rue for Limerick and Peach 5 Bottom. 'I really,can't^ speak for the rest.
.6 MR..RIDGELEY: Thank you.
7E MR. SUH: ,What is the function.of these 8 ventilation HVAC systems?' H 9: MR. MARIE: To maintain that region of negative
-10 pressure as part-of secondary containment. Also provides h
L 11 cooling'to that area, some cooling to that area. l- 12: And lastly, I' d like to comment on the fact, the L 1'3 - potentiality of scrubbing of, through a head leak, head f14 gasketLleakage, would have to be assessed and some type of 15L -efficiency assigned to it, which we believe it would be 11 6 difficult to assess without being able to define the flow, 171 the flow area, and so forth, with a burping or belching type 1;8 action as the head were to lift, we could envision large 19 flow for short periods of time, essentially displacing the 20 water, pumping large bubbles and getting very little 21 scrubbing. This would not be a controlled dispersed, 22' . dispersion of the gases through the water where we believe
~23 quantifying a significant amount of scrubbing. That would 24 have to'be evaluated.
25 MR. RIDGELEY.: We were going to analyze it. A Heritage Reporting Corporation (202) 628-4888
U. i b, i 95 11 MR. MARGULIES: Follow up on that. What is the 2 likelihood-from -- 3 MR. MARIE: Okay, would you like to address that? L4 MR. SCHMIDT: We can address what the PRA 5' assesses in terms of containment, the' assumptions in the 6 existing analysis for that leakage.
- 7 MR. SUH: That.was Question 6.
8 MR. SCHMIDT: Yeah, that's the next question. 9 MR. BARRETT: Let me just make sure about the 10 question about the inability to~ flood. Does that mean that
~ .11 you have total inability to have.any heat transfer from that 12 head, say if you were'to pump water in there? I'm trying to-13 get some sense of how high these ducts are? In other words, could you have some heat transfer from that?
( ' )14 15 If the concern were mainly the over temperature 16 failure as opposed to scrubbing.- Or are they in position in 17 such a way that if you wanted to get heat transfer from that 18 head you would actually have to actually spray the head
'19 itself?
20 MR. KLEIN: It's been a good many years since 21 I've even looked at that, so I can't just definitively say, 22 although my recollection is that they are located lower down 23 and we do have to block them before they could plug up the 24 reactor wall. I believe they are lower down. But I can't 25 conclusively state. Heritage Reporting Corporation (202) 628-4888 i I i _ _ _ --________-______D
rr 96 1 MR. RIDGELEY: Could you provide the drawing that 12 shows us that detail? 3 MR, SUH: Maybe just reference that. L 4 MR. MARIE: We'll' check it out. p.. 5 MR. SUH: So, Joh,.that was'an SR figure for the
.6 drywell head area and ventilation ducts? Is that the 7 question?
8 MR. RIDGELEY: Well, actually, I'm looking for 9 more detail, but I'm not sure that the article has that 10 level of detail. I'm-not sure. I 11 MR. SUH: 'Why don't we start with'that. 12 MR. RIDGELY: Sure. One other item I'd like to 13 ask,'which is not really on the list is, I understand that.
'drywell to wetwell vacuum for Mark 1 have been found in'all
( ) 14 15 kinds of distorted and odd positions in normal operations. 16
~
And I believe that it's the same type are at Limerick. Do 17 you have any plans or any procedures or anything in 18 addition, to be sure that those vacuum breakers will indeed 19 seal when called upon?- 20 MR. KRICH: Those vacuum breakers are tested 21 every 30 days, first of all. There's, of course, there's-22 four pairs. They are redundant. And then, during every H23 - refueling outage, they are tested to, surveillance tested to
-2 4 - ensure that, in fact, they do lift off their seats at half a 25- pound Delta P, and go open at one pound Delta P. So if Heritage Reporting Corporation (202) 628-4888 1 i
i l' l I. . . . ..- - . . . . .
(
- v p--
r, t- .- V 99
.O 1~ they're --
2 MR. RIDGELY: Be sure that they close? 3- MR. KRICH: Yeah, I mean, there's'a limit switch. 4 And'so.you have'a red and green light in the control room, L 5- 'So every 30 days you do a test and you actually stroke the. L 6' valve and it.goes from red to.two lights to one light, and
- 7 it tells you it's.' full open. And then you, if somebody
~8' comes back, you'11 get the' sequence of what in fact-it tells 9 you'it's-closed. 'You only get the close light when it's on,.
10, when the valve is on its seat.
>11 These are monster valves.
12- MR. RIDGELEY: Motor, is it? 13 MR. KRICH: No, no. 14 MR. RIDGELEY: How do you --
.15 MR. KRICH: Oh, there's a pneumatic operator, I'm 16 sorry, yeah. .17. MR. BARRETT: There are four pair of two valves 18 e a ch ,' so the failure of one valve is not --
- 19. - MR. KRICH: That's correct.
20 MR. BARRETT: The failure of one valve to close, 21- that's not a failure of the -- 22 MR. KRICH: Of the system to prevent bypass. 23 That's correct. Does that' answer your question? 24 MR. RIDGELEY: Yes, thank you. 25 MR. SUH: I.want to' move on to 6. I want to read Heritage Reporting Corporation (202) 628-4888
--- -,7 l 1 98 { C1 it for the record. Assumptions used to develop information. l 2 and the values used in PRA; a) probability of magnitude of 3 drywell' head leakage; b) probability of'suppressionable 4 bypass; c) fraction of sequences in which the reactor' fails
- 5. to -- and' d) likelihood of accidents while being de-inerted 6 and continued challenges due to de-inerting while venting.
.7 MR. SCHMIDT: As we indicated earlier, the 8 evaluation that is done of the SAMDAs is based on the source 9 term analysis, i.e., the containment analysis in the PRA and 10 cited sections.here where it's explicitly covered. And the 11 figure reference for the' containment of entry.
- 12 I have attached to the handout the containment of 13 ~ entries. They are extremely difficult to read. I pencilled in a few of the numbers on there. But these come up in the
( )14 15 questions that were answered. This is the CETs for Class 1, 16- 2 and 3. And then there is a CET for Class 4. 17 The specific questions that were asked, the first 18 one was the probability and magnitude of drywell head 19 leakage. The CET that I just showed, those included 0.5 20: probability of a leak that prevents rupture. And then that 21 leak, that .5 is subdivided into half, large, greater than l 22 100 percent per hour and half small. And if you study the 23 tree, you'll see that. I think it's a Delta failure mode B. 24 MR. PALLA: Let me make sure I understand what 25 you're saying though. The question said probability and Beritage Reporting Corporation (202) 628-4888 f
99
/~N i \_) 1 magnitude of drywell head leakage, not drywell. What are j I
2 you answering? 3 MR. SCHMIDT: I understand that. There was no 1 4 explicit treatment of haad, of where this leakage would be, / 5 whether it be drywell head or be elsewhere. When the CET j 6 was developed, and obviously this is a very simplified 7 containment event tree compared to the ones that are 8 generated in the last couple of years, yet one of the 9 potential locations was drywell head leakage. And that went 10 into the fact that there was a good chance of leakage. Now 11 be it not treated explicitly, be it a drywell head or be it 12 elsewhere. So we can't say that there's a .5 probability of 13 drywell head leakage. The assessment done was that there was a .5 probability of leakage. ( ) 14 15 MR. PALLA: Well, there's a structural analysis 16 somewhere that looks at various locations and thcir 17 challenge. 18 MR. SCHMIDT: And the FRA that was done, there 19 was considerable analysis done. And in one of the 20 appendices, I think, of the PRA, there was work done. In 21 fact, by CB&I, if I can remember, a Senior Vice President of 22 Philadelphia Electric getting on the phone and telling CB&I 23 they' d better shape up and do it. There was a lot of 24 argument about it. 25 The, again, the analysis wasn't done to the kind ( () Heritage Reporting Corporation (202) 628-4888
j 100 g k-) 1 of analysis that is done today. It was kind of, it was j l 2 assessed that there would be a .5 probability of leakage, J 3 drywell head or other places. And the impact of where that 4 was is treated in this way. Because it would go, some of 5 them would be large, some would be small, and there was a I 6 likelihood of being cleaned up by the treatment system. 7 MR. PALLA: Do you know where the most likely 8 location is? I 9 MR. SCHMIDT: My view is the most likely location 10 is there, but I don't recall any detailed analysis done of 11 assessing the probability of leakage in various locations. 12 That was beyond the state of the art at the time -- 13 MR. PALLA: But the loading at various locations and the structural response at various locations was part of ()14 15 the structural analysis. 16 MR. SCHMIDT: There was a structural analysis 17 done by Bechtel of the integral performance, and the strains 18 were higher in certain locations. 19 MR. FULFORD: It was an over-pressure analysis. 20 MR. SCHMIDT: An over pressure analysis. Now 21 what would be, failure modes be, with today's understanding 22 of core melting phenomenon is perhaps another story. But 23 again, we didn't do anything new and nothing new has been 24 done for Limerick since the PRA in this area. But in the 25 evaluation of SAMDAs, this was what was used basically. The Heritage Reporting Corporation (202) 628-4888
, in F
f 101 m
' .f 3 1\_/ 1 probability of various' containment failure modes was'given 2 basically on these-containment of entry.
t
. 3- The' question of probability of suppressionable 4 bypass, suppressionable bypass was considered in the PRA 5 only as.a result of containment rupture, in the PRA.. In 6 other words there was no separate treatment in the t
7 containment of bypass, except if the drywell ruptured. If 8 the rupture of the containment occurred in the drywell, then L 9 the pool was bypassed from that' point on. That was 10 ' included. 11 We do expect bypass, as a result of drain 12 failure, at, and there should be an approximate sign in 13 front of the'6 there, in approximately 6 minutes. And this was treated in the PRA and analysis of the, and there are
- [J-'Y14 15 pictures of the drains that are in the drywell, inside the 16 pedestal. space.
17 A conservative -- 18 MR. PALLA: It's in the PRA but it's not in the 19 PRA? You kind of said it's considered only -- 20 MR. SCHMIDT: The phenomenon of drywell bypass at 2,1 six minutes was looked at --
- j. 22 MR. PALLA: But not really included?
23 MR. SCHMIDT: But it was not included in the 24 event tree, for example. You see, there's no particular 25- statement of bypass in the event tree.
-( ) Heritage Reporting Corporation (202) 628-4888 L
L
F r
-102
(~\.
' \f 1 MR. BHATTACHARYYA: Excuse me, Bob. They did-2 look.into some-line fai3ure, even to start out the' sequence l:
1: + L3 itself. We didn't specifically say that it would be 4 scrubbed, for example. We had some sequences. 5 MR. SCHMIDT: 'You're probably referring to 6 seismically-induced sequences, which.-- 7 HR. BEATTACHARYYA: Precisely. They don't have 8 to be seismic. I can put here something to vibrate, I can 9 vibrate the windows, too. It was not done. Nobody did it 10 in the' analysis. 11' MR. SCHMIDT: Seismic sequences can lead to some
- 12 other failure modes which bypass the pool condition. Those 13 were considered but the effects of other reasons were not.
Now the conservative source terms that were used ()14 15 and developed in the PRA and SARA means that the impact on 16 risk of this bypass is very small. If you look at the 17- containment of entry, and I realize this is hard to see, 18 there are a gamma, a gamma prime,.a gamma double prime, 19 these three sequences right here. 20 A gamma containment failure sequence is a drywell 21 failure. A gamma prime failure sequence containment failure
~22 is a suppression pool failure above the water, pools in the 23 waters there, so everything that goes out through the break-24 is through the pool.
25 The drywell failure, the only thing that goes Beritage Reporting Corporation (202) 628-4888 l _ _ _ _ _ _ _ _ . _ _ _ _ . _ i
7-==- L ! L i 103 q T_ / 1 through the pool is basically the melt release, what happens 2 before the things that goes through the SRVs into the pool. 3 And the gamma double prime containment failure 4 mode is a rupture of the suppression pool below the water 5 level, and you lose the water. Those are the three 6 containment failure modes. 7 The impact of this bypass can be estimated by 8 taking the gamma prime, i.e., the vapor space failure mode, 9 which is the ones where the pool is completely affected, and 10 assign them a gamma or a drywell failure, source term. When 11 you do that, you get a 5 percent increase in population dose 12 for the SAMDA, as evaluated for the SAMDA. You take the 13 population dose, the base line population dose that was used to make that analysis and it works out to be about a five ( ) 14 15 percent increase. It makes about a 15 percent increase in 16 Class 4 source term. And then Class 4 source terms, or 17 risk, is about one fourth of the total. 18 So it's not a large contributor. Part of that is 19 because the PRA source terms are quite large. I don't 20 remember the numbers but if you look them up, they're large 21 numbers. I mean .4's and .5 iodine kind of numbers. 22 The pool bypass was accounted for in determining 23 the effectiveness of SAMDAs, i.e., a vent, non-filtered vent ; i 24 did not provide fission product attenuation because the pool 25 was bypassed. And if you read the words in the June 23 () Heritage Reporting Corporation (202) 628-4888
l 104 { " ~} submittal, this.is mentioned.
' 11
! : 2' MR. PALLA: How did you, if you made your
- 3. judgments of SAMDAs based on, at the class level. If you 4 didn't look at sequences in the Limerick plant, how could 5~ you --
6 MR. SCHMIDT: But every class behaves, sequences L 7 are'only different in how you get them in the, core melt, not-8 what happens-after core melt. i 9 MR. PALLA: But aren't some Class 1 sequences 10 some will bypass an'd some will not? 11 MR. SCHMIDT: I think it's very, I think they're 12' probably both the same. You can have a high pressure 13 injection versus a low. MR. BARRETT: I'd like to interrupt at this point
. ( ) 14 15- because we are, it is now 4:30, after 4:30, and we've had a 16 great deal of interesting discussion here back and forth 1
17 about these questions. But we now have questions 6, 7, 8, 9 18 and 10 to-cover. And the schedule, at least, says we're 19 going to be only here for another half an hour. 20 The primary purpose of this meeting is really for 21 you to communicate to us the answers to the questions as you 22 see them. And so I'd like to make two requests. First, I 23 'would like to ask the NRC staff to hold questions until we 24 have gotten to the end of all the discussion of these five 25 questions. Unless there is something about the discussion. Heritage Reporting Corporation (202) 628-4888
7_,_-_--- L 105 s f~ ; ( )/ 1- 'from the licensee that y'ou absolutely don't understand, and : 2' then we can briefly ask for a clarification. But I want, I f 3 don't think we can have any more of the kind of give and 4 take that we've been able to have so far. And then I would 5 ask that, if at all possible, that you could speed up the r 6- presentation so that we can get through a little more? 7 MR. SCHMIDT: Okay, fine. 8 MR. BARRETT: Thank you.
'9 MR. SCHMIDT: -I think that covers B, C, what 10 fraction, I think it's clear here. Fifteen percent of.the 11 CDF are TQUX sequences, type sequences. Another 14 percent 12 is from sequences, long term station blackouts when-13 batteries fail, battery failure itself from time zero, these 14- kinds.
A 15 The last question, or last sub question, 16 likelihood of accidents while de-inerting and containment 17- challenges.due to'de-inerting following venting. In the 18 containment of entry, here I utilized .01 as the 19 unavailability of inerting. One percent of the time the 20 plant is running de-inerted. This is based on data from 21 Peach Bottom and it's actually a text back in Limerick. And 22 de-inerting following venting has not been specifically 23 evaluated by us. The owner's group, I think, has done work
. :2 4 on that and it is available.
25 MR. PALLA: Just a quick question of () Heritage Reporting (202) 628-4888 Corporation
n.. 106
'l clarification, on the fraction of the sequences that fail at p 2 high pressure.
3 MR. SCHMIDT: It's a fraction of core damage g 4 frequency.
'5 MR. PALLA: The total times that the vessel 6 breaches, is it'15 percent at high pressure? I'm confused 7 by the 14 percent.
8 MR. SCHMIDT: This is the fraction of the core 9 damage frequency at high pressure, not the vessel failure. 10- Vessel failure-goes beyond. It has to do with okay the TOUX-11 sequence occurs where you get failure at high pressure. But 12 now, between the time of core melting and vessel-failure, he
.13 does depressurize. That's not included in here. That would
() 14 reduce the probability. The total is 29 15 . This is the upward estimate. 16 percent of the core damage frequency, would be at high 17 pressure because they haven't.depressurized or the', you 18 don't have DC power, basically. DC power is required to 19 hold the valves open. 20 Now from the time that you start get, when the 21 water level drops-and the core starts being damaged, to the 22 time.when the vessel fails, there is time to recovery and 23 depressurization in there. And you haven't assessed that. 24 But that is something that would need to be assessed. 25 MR. BARRETT: Ready for question 77 Heritage Reporting Corporation (202) 628-4888
107 D) Ih t 1' MR. SCHMIDT: Yes. Question-7:was'the 2 - conditional. early fatality estimates for the four dominant 3 sequencesLidentified in Table 1-4. First of all, you should 4 have received a change to that table, where'there are I 5 think'six sequences now. I've listed those sequences and u 6 .the frequency of them, the class they are associated with 7 and the conditional early fatality risk, in units of early. 8 fatality per accident, per occurrence, of that sequence. 9 The first'ene is the seismic, since the original submittal 10~ question 11, we included seismic. It's sequences, the 11 seismicly induced vessel failure, it's split between class 3 12 and class S; class 3 is'the trivial contributor, class S is 13- a'significant contributor. But that's where the number came ()14- from. If you mul'tiply the frequency times the conditional
-15 risk or the conditional population dose, not population 16 dose, just early fatality, you get 49' percent of the total.
17 The rest of the sequences are all ATES sequences. 18 They are significantly, the second part of the question, is 19 how they've changed from the original, ATWS which has been, 20 as we indicated, completely reassessed. And here is a good 21 example. Those should not be B's, those should be U's. Let 22 the record show that those are U primes rather than B-23 primes. And all of thoso U-prime is there to prevent 24 . overfilling of'the vessel.
- 25. MR. BHATTACHARYYA: And that's one of the
) Heritage Reporting Corporation (202) 628-4888
108 1- reasons, Bob,-I asked this question. I suspect that there 2 was something going on and that was one of the reasons I 3 asked this question. 4 MR. SCHMIDT: The U-prime probability, I looked 5 it up -- 6 MR. BHATTACHARYYA: Okay. 7 MR. SCHMIDT: -- there in, I think, one sequence 8 N that is like .5, but most of them are about .1. And you'll 9 see these sequences occur, generally, when you read the 10 description, there's a stuck-open relief valve, or the 11 reactor is depressurized, and you get some low pressure 12 pump, be it condensate pump, be it other pumps inject and 13 it's not caught by the operator in time. MR. BEATTACHARYYA: Excuse me. This is based on ( ) 14 15 yearly operation or late relocations. 16 MR. SCHMIDT: This is based on the SARA 17 assumptions, which had a three speed. Now the seismic one 18 has in it a fraction of the sequence treated just like in 19 SARA. There's a fraction of the sequences had a delayed 20 evacuation or slow evacuation, just like SARA, okay? 21 MR. BARRETT: Question 8. 22 MR. SCHMIDT: Question 8 is the fire sequences 23 and conservatism. And the best thing is probably to do 24 nothing more than generally discuss it and, these are the 25- dominant sequences, the area in the plant that is, fire area 1
,~
(_) Heritage Reporting Corporation (202) 628-4888 l I
L I q 109 f 1 J.1 .4 4Jis,the safeguard system access area. And the dominant [ '2- contributors, that frequency-that's given to F-44, QLAT, 1 L k 3 includes cable fires,. panel fires, transient. combustible 4 fires. And it includes what we called in SARA' fire grow i 5 stages 1, 2 and 3. L 6 Tried to pull out what are the dominant 7 contributors.. It's cable fires in fire. grow stage 2 in that 8 room. In fire area 2, it's cable fires. Fire grow stage.3 9 and fairly similar value would be fire grow stage 3 and 10~ panel fires. Going through it that way. Because there is 11 different levels of conservatism depending on cable fire, 121 what's the source of it. Cable fire is the overall dominant 11 3 contributor. [ l'4 Now in your handout, the two next pages are lL5. mislabeled in terms of page number. There's obviously one 16 that starts 'the discussion and that's labeled A-3. This 17 goes through a number of the reasons. I'm not going to read 18' them to you. I think they're pretty self explanatory. 19 We did use the Sandia study as the basis which 20 . corrected, for example, Brookhaven's comments on the SARA 21 analysis, on the initiator, and so I think those are all 22 covered. I think they're conservative. For one reason, as 23 we point out, that the plant has got IEEE-383 cables. We 24 didn't credit that. We used very short propagation times 25 and a lot of conservative functions relative to how the
}; . -v Heritage Reporting Corporation (202) 628-4888
_ _ _ . ___--_-______-_____-_____a
/9 110 ' /~T
_1. ' damage progressed. I think if,. talking about the
-2 uncertainty, I think the uncertainty in this would be down 3 strongly.more than up. And it's interesting, if you'look at 4' the NUREG 1150 uncertainty numbers that I put up there for 5 fires, they're strongly down. They're more uncertain down 6 than up. And I think that perhaps is the state of the art.
7- I think the Sandia people, we've talked about this, would 8' agree. 9; Unless there are specific questions on those, . I'm 10 not going to talk any~more about those. Because then I 11- think that takes care of question 8. And then we have L12 . question .9.
.13 Question 9 is the reduction term and trip frequency. Since commercial operation began February of
()14' 15 '86, -- and pardon the typos -- through May of ' 89, and when 16 it went into shutdown at that time, or when we did the 17' analysis, but in that time frame, Limerick 1 had 8 turbine 18 trips, 4 manual, 4. auto. Frequency, therefore, of-turbine 19 trip in this time period was 8.5 over 3.33 or 2.55 per year,-
'20 or it's in that time interval. February '86 through May l-21' '89, they're 3.33 years and the 8 trips.
22- The .5 has the effect of a Bayesian update. My 23 understanding from our expert is that, effectively, that
, H24. adds'a half a transient, so you get the right number by ~25- dividing 8.5 by 3.33, and that's the value.
Heritage Reporting Corporation (202) 628-4888
t
- l' .The new transient frequency, total transient
'2 frequency, is shown here, 6.69'is consistent with what was 3' shown earlier. This is the breakdown. You can'see the '4' \
change' form'5.6 to 2.55. This November ' 88 update, this you 5 see is the only-change in initiator frequency in that it is 6 that value of 2.55.
- 7. MR. BHATTACHARYYA: This doesn't look - you I 8 assume at more than 90 percent power this is how it happened. Because I am interested the ATW sequences.
.10 MR. PHILLABAUM: No. Except for extremely low' 11 power levels we included turbine trips from lower power 12 levels, much lower. There has only been 1 turbine trip 13 automatic SCRAM that occurred at 100 percent power.
[ 14 Everything else has been much lower. 7 15 MR. SCHMIDT: This is conservative. Most of them 16 have been at lower power. 1*7 MR. BHATTACHARYYA: -Well, based on two, two and a 18 half years experience you say that there was only one 19 turbine = trip at more than 90 percent power, but that a lot 20 of things at low power levels. 21 MR. PHILLABAUM: Lower power levels, right. 22 MR. BHATTACHARYYA: I see. The Unit 2, would it 23 it have the same turbine control system as Unit 1, where you
- 24. got the experience?
'25' MR. PHILLABAUM: Unit 2 is very similar to Unit A-V Heritage Reporting (202) 628-4888 Corporation
112 (/ ' '1' 1. With very few exceptions, there's no changes with the
- 2. turbine control system.
3 MR. BHATTACHARYYA: Okay. Because what we are 4 finding out today, like for example, in the south of Texas,- 5 Unit 2 data is. showing higher than Unit 1. In fact, our 6 . chairman Lando Zech, brought this up during the approval of 7 the license. That's why I asked this question, what is the 8 rationale in making use of Unit 1,. with 2 years of 9 commercia1' experience, to Unit 2. TMI, for example, it
-10 happened'unfortunately.
11 MR. PHILLABAUM: The one thing to point out, the 12 ' 88 update is based, that 5.6 excludes firstLyear 13 information from the report. Now the Limerick' actual () 14 experience doesn't include the first year of operation, which would have expected that first year to be higher than 15 16 normal. So I hope we've kind of compensated. 17 MR. BEATTACHARYYA: Well, you say that the 18 control system -- 19 MR. PHILLABAUM: Exactly the same. 20 MR. BEATTACHARYYA: Okay, all right, thank you. 21 MR. PALLA: Could you explain why you just chose 22 to update the turbine trip? Why not revisit the other ones 23 as well? 24 MR. SCHMIDT: They're very small numbers and in l 25 three.and a half years, you do a Bayesian update of zero I L Heritage Reporting Corporation (202) 628-4888 L l- _ _ _ . - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
1 \} 113
' /~ .
(~ \T/' 1; - occurrences,.and in-three and a half years you probably 2- d woul'n't get'a'significant change. 3 MR. PALLA: Did you do that and decide not to 4 change -- o [. 5 MR. SCHMIDT: : No, we didn't do it, and -- 6 - MR . PHILLABAUM: Well, we looked at it and that's 7 exactly.the rationale we came'to, that there wasn't enough 8- experience. We had no, the only experience that was 9 occurring is turbine trips. And based upon no failures or 10- no incidences in the time period, there was just ne 11 information. So we continued to use -- 12 MR. SCHMIDT: You wouldn't get a significant 13 change in11t. That was the decision made. MR. PALLA: You didn't have enough data to- { } 14 15 support it? 16 MR. PHILLABAUM: Yeah. 17 MR. SCHMIDT: Right. Could have done it and 18 would have gotten a change in the number. But said it 19 wasn't worth it. I mean, that was the decision. 20 MR. PHILLABAUM: The second part of Question 9 21 asks about assumptions'related to loss of outside power. We 22 determined that to be the frequency and that's the second -- 23' MR. SCHMIDT: Part of the passout was a copy -- 24 yeah, there is overhead for this.
~25 FR. PHILLABAUM: I really didn't -- just a Heritage Reporting Corporation (202) 628-4888
W-H 114
; ,, . n .
I' A . portion that describes the analysis of'the data that w/ l 1 2 supports the determination of the. frequency of loss of the 3' offsite power to Limerick and its recovery,'which is really 14 based on PJM grid experience. 5- -MR. BHATTACHARYYA: Did you make use of any e 6- fossil data? 7 MR. PHILLABAUM: That's been revised. Now we're
- 8 only looking at the nuclear plants within our system.
9 MR. BEATTACHARYYA: Okay. So no more fossil
'10 data?. 'll MR. PHILLABAUM: No more fossil data was mixed 12 in,. that's right. The section is pretty self explanatory.
13' I really didn't plan on going into -- MR. SCHMIDT: That's the' frequency, and then this {[14 15- the recovery and distribution. So that's all in what we've 16 given you. 17 MR. BHATTACHARYYA: Do you have any data on loss 18 of offsite power -- 19 MR. PHILLABAUM: That's also part of the, that's 20- the last couple pages, talk about the possibility of you're 21 in a turbine trip and you get the subsequent loss of an 22 offsite power situation. , 23 MR. BHATTACHARYYA: Okay, all right. Thank you.
'24 MR. SCHMIDT: Any more questions on Item ??
25 MR. PHILLABAUM: Question 10 asks for discussion Beritage Reporting Corporation
~
(202) 628-4888
> < .. J 115 j 1- regarding any new sequences modeled on the revised PRA such 2' as sequences.resulting from steamline breaks,.outside 3 containment.
4 Again, what I've elected to do is just provide 5 some information from the updated PRA describing for 6 steamline breaks outside containment, our assessment of 71 that. For the interfacing LOCAs, you go through the 8 rationales, describing what the potential of that situation-9 would be and determined that it's not.very likely. So it's 10 not been a, found'it unnecessary to do a quantification of 11 that. 12 -Secondly, the other new sequences that have been-13 . considered result from some special initiators that-we're
-not, that have been included in the updated PRA. Internal f( )14 15 flood we've talked about already. The other new special -16 initiators that were not looked at previously are the 17 reference leg, weaker break, loss of surface water, loss of 18 all surface water, the loss of a single DC bus, high drywell
- 19. temperature, loss of,insulant air, loss of single AC bus.
- 20 I'd like to note the last three of these, again, were 1
21 treated in a fashion similar to the interfacing LOCA where L 22 we looked, made the assessment of their likelihood and said, 23 "Well, high drywell temperature is going to be not more than 24- ten to the minus 8. Insulant air in this range, same thing 25 with the loss of AC bus. The other ones we have done Heritage Reporting Corporation [ (202) 628-4888
' 116 -11 quantification'.of the impact of what these initiating '2 events, special initiating events, have on core damage 3l frequency at Limerick.
4 MR. SCHMIDT: Let me point out that the 8 minus 8 5 was generated for the SAMDA submittal in an update post-6 November88. The rest of them are in the Itovember ' 88. 1 l 7^ MR. PHILLABAUM: Right. They' re ' unchanged and 8 that's where,. earlier on, you saw an asterisk saying L 9 flooding was included with other special initiators. 'These 10 are those other special initiators. And as-you can see, 11 these are very small contributors. 12 MR. SCHMIDT: The November '88 update had an 13 internal flooding number in it, but it was higher, it was very conservative, and we just took some of that t] 14 15 conservatism out. 16- MR. BHATTACHARYYA: Okay, as far as this, I am 17 currently starting a HPCI RCIC steamline break outside the 18 containment. You do have a normally open valve steam goes 19 to the top. At Limerick those valves were closed during 20 downstream break, isn't it? 21 MR. PHILLABAUM: That's correct. 22 MR. BHATTACHARYYA: Okay. 23 MR. BARRETT: I'm sorry. I think we were passing
.24 out the papers when we covered interfacing.
25 MR. PHILLABAUM: What I wanted to do was try not Beritage Reporting Corporation (202) 628-4888
117 j U [ 11 ltoltake'up any time discussing this. What we've done is 2 we've assessed, I just wanted to provide the description i' L3- that's in the updated PRA of our assessment for the 4 likelihood of having a situation with'the, you know, getting 5' outside the containment. And. based upon the Limerick 6 design, the likelihood of a situation occurring, it's 7' evaluated to be a very low likelihood,,so that we have not
- 8~ done a, we have not quantified, other than to say it looks-9 .like it's going to less than'10 to minus 7 per year if the 10 situation happens. No special event tree was created.
11' MR. SCHMIDT: The only assessment that's done for'
'12 Limerick is described in here and it's not a quantitive 13 assessment sequence. )14 MR. SUH: Was that-Question 107 15 MR. SCHMIDT: That was 10, 11 and 12. That's it.
16 MR. KRICH: I think we've gone through all the 17 items -- We're ready to stay as long as it takes to answer
- 18. questions.
19 MR. SUH: .I think what I'll do, at this point, 20: I'd like to, just for our mutual understanding, go through 21' the questions -- 22 MR. KRICH: Questions, I was going to do the same 23; thing. Okay. 24 MR. SUH: -- that I've written down, and maybe I L25 can add and subtract. I've got four of them, and number 1 Heritage Reporting Corporation (202) 628-4888
t 118 g,
'l is with regard'to the fire initiator,.w' ell core. damage 2 frequency estimates require fire initiator. The question 3 was asked did the cable'for both-safety system' trains, do 4 both trains have three hour fire ratings?
J -5 MR. KRICH: Would you repeat that? 6- MR. SUH: Do both safety system' trains have three 7- hour fire rated cases? That is the question and I think 8 you, when that question was raised,.someone indicated that 9 you will look into that.
~
10 MR. KRICH: We'll have to get back to that, yes. 11 MR. SUH: Okay. 12 MR. SCEMIDT: The analysis was done on whatever 13 is-there. Be it three hours, be it one hour for the second. I) 14 -So it was done-based on whatever is there. I just don't 15 know what'is physically there. 16 MR.-SUH: Okay. That was number 1. And number 17 2, that I have, is for the SAMDAs that you considered, what l 18 occupational exposure would be incurred in installation and 19 in recurring maintenance? 20 MR. KRICH: Right. y 21 MR. SUE: Including the basis for your estimates, 22 such as installation hours, location of the work pe formed 23 and dose rates in those locations? 24 MR. KRICH: Installation hours, dose rates and? 25 MR. SUH: Right. Installation hours, location of n' U Beritage Reporting Corporation (202) 628-4888 h
-_ a
L F 119 p;; ; 'l the work performed.--- It
!. , 2. MR .~ KRICH: Location'of the work.
- 3 MR. SUH: ---and the dose rates'in those
' 4' . locations.
5 Question'3'that I have is a reference to drawings 6 'or other' drawings for drywalls, for the drywell' head area 7 and.the associated HVAC systems in the immediate area?
, 8. MR. KRICH: Yes.
9 MR. SUH: Question. number 4 is one that we 10- haven't brought'up, but here we.need to bring out, which.is 11 in the same vein that we're considering these containment
- 12 performance improvement programs for, that we're looking 13 into, the containment performance improvement ideas. It l' occurs to me that one other option,. alternative by a broad
?( f 14 15 definition of the word. alternative, would be for Unit 2 not l.
16 to operate and to replace the energy production from Unit 2 p 17 .from non-nuclear energy sources. 18 And a question in that regard I turn over to Mike 19 Mazmick who's project manager in NRR. And Mike, could you 20 elaborate on what exactly sort of information? 21 MR. MAZMICK: I just wondered if you had any idea 22 of where you might be obtaining that additional power? 23- Would it be just in units that you have or outside pool? 12 4 And if you could give a mix of whether it be coal or what? 25- MR. KRICH: Right. I can't give you one now but, Beritage Reporting Corporation (202) 628-4888
y 1.. 120-
-y.
Ad 1 ;in fact,Lwe're looking into that now. And I think we re . 2 going to .take the assumption that it's PJM and whatever the. o 3 mix is-for PJM. Okay. 4 MR. .MAZMICK: m :5. MR. SUH: ' Mike, for my benefit, could you sort of. 6 restate what you-just'said. 7 MR. MAZMICK: Basically, what I'm looking for is 8 the source of the. replacement power and the mix by different 9 types of generation, whether it.be coal or. nuclear or' hydro.
'10 And whether or:not'it's within their system or outside. !! 11 MR. SUH: Okay. And would you need to have 12 references to the environmental effect from those sources, 13 or not?
MR. MAZMICK: Ideally, that would be great. U()14
-15 MR. SUH: If your environmental report refers to 16 those. environmental impacts from non-nuclear generating-17 sources that~would be useful.
18 MR.. KRICH: Okay.
- 19: MR, SUH: Those were the four questions. Well,.
20 these are in.our C' questions. Let's go through the 21 questions and if you have a comment, let's get.back to the
'22 comment, if you don't mind.
- 23. MR. KRICH: I don't want to beat a dead horse, 24- Jim, but I'm not sure I understand the first question. Does o25- everybody else understand that first question.
) Heritage Reporting Corporation j (202) 628-4888 !
E , 121 M :' V ' . 1'- MR. SCHMIDT: The' concept of'how you're going to
'2i meet Appendix R to protect both' independent trains by both -3 three hour. barriers,,or a three hour barrier in,one and- '4. something'less in the other. And that is something that I ~
L5 think the fire protection guys will know the answer to. .
~
6- -MR. SUH: I had a very simplistic question from 7- my. viewpoint. Maybe I need it clarified myself. Just a 8' simple question.
'9 MR.~KRICH: Yes. Maybe I was making more out of 10 it. Three hour _. fire rated barriers. Yes. .11 MR. BATTACHARYYA: Because in the submittal there i 12 is'here conservative model, I duly agree. But we tried to 13 characterize in our write-out.what degree of annual fire ~ -(]
v 14-frequency went up. 15- MR. KRICH: Okay, okay, thank you. 16 MR. SUH: Did you have any additions or 17 subtractions from'that, Ron, or not? 18 MR. KRICH: From this list? 19 MR. SUH: From this list of four that we just 20 went through? 21 MR. KRICH: Let me check one thing. No. 22 MR. SUH: I'm sorry, Ron?
'23 MR..KRICH: We're set. No, we don't have any j 1
I' 24 other.
'25 MR. SUH:- You don't have additional questions?
Heritage Reporting Corporation (202) 628-4888 t
'122 .3 1 MR. KRICH: No, thank you. +
12- RMR. BARRETT: I think what I'd like'to.do now is-3 we have a group'of people here who are experts -- each of
.4- - whom is charged with the responsibility of a particular area 11 5 of'our risk assessment and our assessment of risks n '6: associated with various-SAMDAs. - And 'what' I would like tio do 7 is. invite each'one of them, in turn, to ask'any'important-8' last minute questions that they feel'are crucial to their c.
p 9 review.
.10 And so.I would like to start with.our core damage 11 frequency person, that's Mr. Chelliay. Do you have any last 12 minute. questions?
- 13. MR. CHELLIAY: Comment? No, you have answered my questions satisfactorily. I did have five questions, you
-{)14 15 answered all five. -I'm very happy with that.
16 MR. BARRETT: Thank you. In the area of so-17- . called backend of the containment performance fission 18 project analysis, and handling estimates we've been working 19L with Brad Hardin of the Office-of Research, supported by. 20- Trevor Pratt at BNL. Would you gentlemen like to pose any
'21 last minute questions that are of high importance to you?
22 MR. HARDIN: 1 was just curious about something 23 somebody else brought up. You said that you weren't taking g 24 any credit for containment spray and fission product 25- removal. I was just wondering, I presume you' don't feel u
. Beritage Reporting Corporation l (202) 628-4888 l
l
y . [, :
'l 123 ]
(1 that'you gained much Ny doing that, but I'm wondering is
- 2:. that true?. Or have you looked at that.having significant
, 3 risk reduction?-
4J MR. SCHMIDT: -I'm-not the expert.. The expert.who' 5 did this chose to assess the impact of spray as a prevention
'i 6 of a' failure mode, primarily over temperature failure, which 7 seems to be the one that is of most concern now rather than
- 8 over pressure.
9 Clearly,- a spray will have, even in an open 10 containment,;will have'some mitigating effect. He tended to-11' think that was secondary to the prevention of failure and, 12 therefore, didn't include it. The magnitude of it, I don't 13 know, and I. don't think he explicitly evaluated it. I know in the past we have done assessments that, ( f 14 - 15 yeah you have a spray, an existing spray, and even if the 16 containment's open, you get a benefit of it. And so there 17 is a benefit there. But compared to the prevention of 18- failure modes containment, we thought it was smaller, so'we 19 didn't include it. 12 0 Our major impact of spray, what we called a 21 spary, was just really getting water in. And you get 22 benefit from just getting water into the containment, pour 23 out, to come out on top of the corium, for example, even if 24 it isn't spray. So we're trying to avoid the questions of 25 spray pattern and a lot of these kind of issues. Really,
)' Heritage Reporting Corporation (202) 628-4888 l
i . . . - . _ _ _
35 1 < 124 ft"'y 1 kdjll? .the: impact?of.what we're looking at for. spray was the
- 2 getting.of water on top of the corium, as the major effect,.
p 3 the major beneficial effect. .I think that's consistent with 4 the BW owner's. group positions on spray and injectionLof-5~ water, post-core: melt, post-vessel rupture for:the core-on-6 'the-floor situation. 7 MR. FULFORD: I might add to that, remember all 8 the. pipes in the place and the presence of biological shift 9 around the vessel. 10 MR. SCEMIDT: That there's a lot of unsprayed 11 areas.
.12 MR. HARDING: Thank you.
13 MR. BARRETT: John Ridgeley is helping us to identify SAMDAs and to identify the weaknesses and strengths
'( ) 14 15 of SAMDAs. John, do you have any additional questions you'd 16 like to ask in addition to what you've already. learned?
17 MR. RIDGELEY: No, I don't.think so. Thank you. 18 MR. BARRETT: Okay. Finally, Sid Feld has been
-191 helping us with-the issues related to cost and cost. benefit ~20 analysis. Sid, do you have any questions?
21 MR. FELD: Yeah, I have one additional question. 22 In your calculation of the replacement energy costs could 23 you tell me what capacity factors you assumed for' Limerick 24 Unit:2 and, more specifically, whether you assumed that that i' 25 capacity factor was variable as the new reactor matures, I, Heritage Reporting Corporation (202) 628-4888 1 i
'y i 125 .w hether there is=some-lower level of energy generation.
ll
]
2: assumed'during'the initial period?
- 3, MR. HEARN: I can't answer that for you and I ,
'4' would have to refer back to the individuals.who provided us
- 5- with the figures and give you their' breakdown on that.
6' MR .' KRICH: What we had gotten from Philadelphia
- 7. Electric was replacement power costs as a function of' time
'for the next three or four years.
8- And we took an average of 9 that~ cost over the period that our schedule showed would be
. 10 needed 'for that specific standard to be installed. - 11 . MR . HEARN: I'm sure within that calculation is . 12 'some evaluationsLof power levels along the lines that you're -13 .asking.
l()l4 MR. FELD: Okay.
' 15 MR. SUH: Sid, for my benefit, let me, restate . 16 the question again and please correct me. What capacity 17 factor was assumed for Unit 2 in terms of replacement energy 18 costs? -19 MR. SUH: Okay. ' 20 MR. KRICH: Did you have an add-on to that? ' 21. MR. FELD: And I was specifically concerned i . 22 whether there was any variability assumed in that capacity 23 factor as the reactor matures? . 24 MR. SUH: Okay.
I.
~ 25 MR. KRICH: And you would like that quantitated?
7 Beritage Reporting Corporation (202) 628-4888
\
1 '
126
\/ 1 FELD: Yeah, if you a precise capacity factor for 2 the first year and then --
3: MR. KRICH: All right. 4 MR. BARRETT: Okay. Bob Palla, who is leading 1 5 this task group, has a couple of final questions that he'd
- 6. like to ask about SAMDAs.
7 MR. PALLA: Okay. I had a kind of a question of 8 scope as far as which SAMDAs were considered in detail. Now 9 apparently the RDA-identified SAMDAs were given a rigorous 10 costing analysis, and were carried further than anything 11 else was or other mitigative factors that were considered 12 and whether similar or some other type of analysis was done 13 that deals with them In particular, the LEA contention r( )14 identified a number of fixes and you say in your June 23 15 response that these were initially considered. And then you 16 go on to say that you looked at the SAMDAs identified in the 17 RDA Study. I' d just like to know what you did with the LEA 18 contention items? 19 MR. SCHMIDT: We did not look at the LEA, you're 20 talking about a submittal they made to the Board. 21 MR. KRICH: Let me take a stab at it. 22 MR. SCHMIDT: All right. 23 MR. KRICH: The approach we took was, and we took 24 this approach from the beginning with the ASLB was that we 25 considered the RDA SAMDAs to be what the NRC Commissioners , i IT () Beritage Reporting Corporation (202) 628-4888 I
, 'i o ~ , 127
(. L %/ 1. .had ordered the Board'to reconsider. And those were the 2' only ones that we looked at, really, in any detail. 31 LEA, we got together with LEA and compared notes ) l 4' 'and found that we, that some of them fell together,.that-5 th'ey were saying the same thing. .And so,-in our evaluation 6 .of our seven SAMDAs, we were essentially covering some'of p 7 the LEA SAMDAs. LEA then submitted another 84 so-called 8 SAMDAs. We didn't really look at that. And they also- 3 9 submitted some things that, like spent fuel pool axis, that 10 we considered completely outside the scope, that was not 11 looked at. 12 MR. PALLA: Okay. Second part of the question. 13 MR. KRICH: That answers the question? Okay. MR. PALLA: That answers the first part. The ( )14' 15 second part is did you look at alternative ways to meet the 16 objective of the six that you did look at? I know that 17' you've assumed a lot of support systems are totally 18 independent here: the suppression pool cooling option that 19 you have, you have three power' systems involved in that. I 20 don't know how much cost is associated with it, but it looks 21 like a big ticket. 22- I know that, I'm not sure whether or not that 23 power system is really needed. The risk reduction that you 24 get for it, do you really need to provide power? Are there 25 station blackout type sequences or any other sequences that l Heritage Reporting Corporation (202) 628-4888
i: [1 s t 1 you'have power. I'm wondering if you considered any p. 2 subsets? I
.3 MR. KRICH: No, no.
4 MR. SCHMIDT: No.
'5 MR. KRICH: Let me give a little reason for that..
6 I.think if you look at the chronology of what we've gone 7 through, it all happened very rapidly. And we needed 8 someplace to start from. And what we thought was the 9 appropriate place to start from was the RDA identified 10 SAMDAs.
-11 -I.think that the progression of the hearings will 12 flesh out maybe, we need to look at modifications. What we 13 started with, what we said was, here is the gross SAMDA.
Now let's look at that first. If there can be enhancement l( )14 15~ or modifications of that'SAMDA, we'll look at that as we go 16 along, but let's get some place to start from. And we 17 needed to do that because there such a large contention as 18 to what was in the scope of the hearing. p 19- MR. SCHMIDT: We did make one change. And 20 that's, if I remember right, their dedicated heat pool. 21 moving system was two of them, and we said that's
-22 ridiculous. Now, did we carry that thought to the power and 23 all that, we felt that that was so ingrained, that thero 24 SAMDA was an independently powered, completely self-25 sufficient heat system. We said there wasn't any point in Heritage Reporting Corporation (202) 628-4888
( . 4:c. b , P 129-Nf L ., .. v 5\ f 11 ; two of'them because'the benefits.would just be past thec l2. point'."of. diminishing returns.
> 3 MR. FULFORD: I believe there was one'other one -
4' also, with regard'to the water supply to the pond. We're 1 5. using. existing pipe. 6 MR..SCHMIDT: Yes, there were'some things like 7 ..that. , We"did not basicallyLbreak it'down an'd say, "How 8I could we' achieve'that same function.other ways?". We-did.not-L- J9 'do that.. 10 MR. KRICH: .There just hasn't~the time nor the -- L11 .we foresee.that'as potentially happening now as we go 12 through the test period. 13- < MR. PALLA: Okay.
.14 . MR ~. SUH: My understanding is there's no other
, 15 comments from the NRC staff. M:b. Bhattacharya, do you have
- 16 a comment?
117 MR. BHATTACHARYYA: Yes, the reason I wanted to 118 -jump'in is because it relates to-my. expressions. When you 19' are doing this evaluation on alternative sources of power, 4 20 would you look at the need for power, also? Is that part of 21 the question? 4 22 MR. SUH: No, that's not a part of that. We'll 23 take the comment under advisement, but the question as I-
,, 24 understand it to be answered is no. That was not part of J25: the question.
Heritage Reporting Corporation (202) 628-4888 u___ ________=.____________..__ _ _ _ _ _ _ l
v -- _- - j 130-1 The. question.was if you needed to-replace the -
?2 power that unit 2 would produce,' how would-you do it? .What.
3D mix , of : energy : source? - 4 MR. BHATTACHARYYA: My question was really 5 directed at Mike, trying to clarify this question.' I . . 6 MR. SUH: Okay, I think we're done unless.there's 7 any comments from your group. p
'8 MR. KRICH: Well, we hope we gave you as much 9' information passed in the time'as possible.
10 MR. BARRETT: I was simply going to say that we h 11 know that this was a difficult set:of questions, posed on a 12 ,very short time frame. And on top of that, I think we hit (; . 13 you with some. difficult questions here face to face. And I think that it's certainly an impressive response and it will
-()14 15 be.very.usefulsto us-in completing our risk study. And we '16- 'thank'you for coming down.
17 -(Whereupon, at 5:15 p.m., the meeting was 18 adjourned. ) 19-
'20 ;21 22 23 '24
[ o 25 Heritage Reporting Corporation (202) 628-4888
w . L ,
~! ~ 1; CERTIFICATE l l2 b
(r: 3 This is to, certify;that the attache'd proceedings before the r ;4 United States; Nuclear Regulatory. Commission in the matter b1 !I of: 6 'Name: Office of Nuclear Reactor Regulation.
!" Limerick Severe Accident Mitigation Issues 7
a p 8 Docket Number: F -9. Place: Rockville, Maryland 10 Date: July 27, 1989 l_ 11 were held as herein appears, and that this is the original 12- transcript thereof for the file of the United States Nuclear 4
'13 Regulatory Commission taken stenographically by me and, f( ) .14' thereafter reduced to typewriting by me or under the 15 direction of the court reporting company, and that the 16 transcript is a true and accurate record of the foregoing 17 proceedings. , ,
18 /s/ b<-,'Y>> 4.bn 't'4
// (.
19 l(Signature typed) : Irwin L. Coffbhberry 20 Official Reporter
'21 Heritage Reporting Corporation 22-23 -2 4 -
l-L 25 li ' '(]) Heritage Reporting Corporation (202) 628-4888
k I y-
-yyr, -
4 3, p
.4..,.
f i l
*i ,
pj., J. ' ). .'. ) P p k
.j \_'
tf
? . l I
l '
' Sl l
Ie 1
- a;- =.' i .
l l . . - - - - - --- - - - - - - - _ _ _ _ _ _ _ , , __ _
L Table L1
SUMMARY
OF THE BNL COMMENTS ON THE 1981 LGS PRA
!pd AND THEIR RESOLUTION IN THE UPDATED LGS PRA COMMENTS RESOLUTION The success criteria used in the LGS No Action
- rhese, have. kee e % ed PRA represent realistic requirements and ' 4, j, ,,g, ;,, ,,,,j y,j f,5 , ,
they do not correspond to safety analysis report (SAR) criteria. The criteria were o a c c455 M M *J *# developed from analysis contained in con 4a6 a c,5+ ked a ava l . NEDO-24708. o The success criteria for the This remains an open item. NEDO 24708 transient initiators are considered indicates RCIC is adequate for . injection reasonable based on the content of until the RPV is depressunzed; however, this document, except for the the Open Item that termmt is that the assumption that RCIC is capable event trees do not require L.P. injection of supplying adequate vessel water following successful H.P. injection for the makeup to an isolate reactor with SORV. an SORV. [pg. 2-5] o Additionally, the success criteria No Action: This has been performed as for ATWS scenarios require part of the GE design record file and O verification adequacy of to each aetermi== system the or have si=ce dee= usea i= ot8er ra^s-function for mitigating these events. (pg. 2-5] The partitioning of transient initiators into four groups was reviewed and considered acceptable. Specifically, o The treatment of initiating events No Actiont Lost, of feedwa-fe ad in the LGS PRA was more g , ,, , p e,, den se,- y a e u u ,n g s-w e/ow,_ realistic than in the RSS and 3 g g,g 4,, Grand Gulf RSSMAP. because .S atCfesca+l cha lle-lc s fo and re s pon s e. o f +A s- p a a+- o Some initiators included in the Big No Action: These have been added in c I Rock PRA were not explicitly the updated PRA. I treated, i.e., Loss of instrument air I - Interfacing LOCAs
- Steam line break outside Contalument i + a. s i., a- m , 4. m a m m ,. w .n i, _ ,, a no % +ka.y a e.-H o n is n e eded . {
S-728510-012 13 l 032389D89F l
Table L1 (con't)
SUMMARY
OF THE BNL COMMENTS ON THE 1981 LOS PRA
- AND THEIR RESOLUTION IN THE UPDATED LGS PRA 'COhDdENTS RESOLUTION o Additionally, the_ following initiators ' developed by the reviewers - were not explicitly addressed in the LGS PRA: - Loss of DC power - No Action: Now included in e updated PRA - RCP seal failure following - No Action: PWR issue an SBO Pipe breaks in aunliary - No Action: Floodmg examined in buildings and instrument updated PRA tube LOCA Scram discharge volume - No Action: Exammed by NRC LOCA and GE on Generic Basis: ;] frequency much less than IE-6/yr. - Loss of component cooling - No Action: Incorporated in loss water of SW - . Loss of instrument and - No Action: Incorporated in loss control power of DC However, the reviewers stated that the No Action: Verified by updated PRA initiating events ' not treated in the LGS would not significantly affect the total core damage frequency. [Pg. 2-10]
The LOS PRA neglected potentially No Action: The hnhng of fault trees and the subsequent perfonnance of important dependence in the accident sequence quantification process. This is Boolean manipulations of the resulting due to the fact that the functional fault expanded ute account for system and trees were used in isolation to quantify functional dependencies in the Level I the probability .of failure of the PRA. i corresponding functions. [Pg. 3-10] O S-728510-012 I4 032389D89F
~E l
Tcble L1 (con't) ;
,. y V
SUMMARY
OF THE BNL COMMENTS ON THE 1981 LOS PRA l AND THEIR RESOLUTION IN THE UPDATED LGS PRA l COMMENTS RESOLUTION 1 The impact of the omission of fault trees for the following systems was not : i evaluated, - but determmed to ' be potentially important due to' the system - interdependence with frontline systems; [Pg. 3-12] j 1 RPS - No Action: RPS fault tree h added.- i Plant air - No Action: Dismissed on ) judgement that plant air is not a 1 major support systems. Turbine enclosure cooling - No Action: Included in Loss of ; water SW for updated PRA. ; q-Q - Reactor enclosure cooling -
. No Action: Included in Loss of j water SW for updated PRA. '
i d Generally, the fault trees appeared to the , reviewers as being complete and accurate. ! However, BNL revised some rnodels. These changes are described in Table 1.2. [Pg. 3-12] BNL ' disagreed with the value used for No Action: BNL has since cht.nged this i the probability associated with the estimate; simulator data by NRC/RhGiP 1 cognitive human error involving failure to (NUREG/CR 4834) also supports the use depressunze the RPV (event 'X'). [p. 3-7] of the original PRA estimate and even lower values. The LGS PRA has used an HEP derived from a sophisticated HEP model and has performed sensitivity studies to confirm the contribution to uncertainty. b v h l- S-728510-012 I-5 032389D89F [ l
~ _ _ ~ . - _ ,__ = _= _ = = = .
Table L1 (con't)
}]
SUMMARY
'OF THE BNL COMMENTS ON THE 1981 LGS PRA AND THEIR RESOLUTION IN THE UPDATED LGS PRA COMMENTS RESOLUTION Particular cognitive human error probabilities which were modeled in the LGS PRA fault trees adjusted by BNL include: [Pg. 3-17] FEEDWATER
- Failure of the operator to No Action: Duplicate event identified by reset - and . restan the FW BNL has been deleted.
c . system
- Failure of the operator to _ Ope _n: Values not changed close RFPT steam exhaust butterfly valves Failure of the . operator for Open: Values not changed f.
bypassing a failed scaling steam pressure regulation
" ADS - Failure of- the operator to No Action: Value . conservatively set to line up instrument air to the 0.1.
ADS valves RHR
- Failure of the opemtor to Open: Not Included open common valves MOV-67A and MOV-67B.
BNL identified cogmtive human ermrs l
. leading- to common mode failures which were not included in the system fault tree f models: - [Pg. 3-17]
JO l S-725510-012 1-6 032389D89F f I , j
ifT}l Table 1.1 (con't) . ,
- v. ;
SUMMARY
OF THE BNL COMMENTS ON THE 1981 LGS PRA P .AND THEIR RESOLUTION IN THE UPDATED LGS PRA - COMMENTS RESOLUTION p FEEDWATER Operator fails - to start. No Action: ' Failure to start mechanical mechanical ' vacuum pump vacuum pump is included with same given SJAEs - fail to probability suggested by BNL (0.02). mamtnin condenser vacuum. ADS
~
Miscalibration of- core No Action: Instruments are different for sprays and RHR pump each set of valves and are not judged to discharge pressure sensors. have a substantial common cause failure. BNL. change not incorporated because common cause is judged to be most - applicable within the RHR system - and within the CS system. This latter is accounted for. O - Miscatidration of tank levei sensor No Action: Commen mode miscatibration of tank ' level sensors is included, probability of failure is.1E-3. System ' dependence _ between functions No Action: The updated LOS PRA has were not always addressed (e.g., functions used hnked fault trees to explicitly model Q and W both include the PCS system the commonalities between systems and and-functions V and W include the LPCI functions. and the RHR systems, respectively, which share some hardware). [Pg. 3-21] Some functional dependencies were o
- Even+ + r ee.s n . t omitted from' the LGS PRA model (e.g., % pen c. d, dependence of the HPCI and RCIC l systems on the suppression pool temperature). [Pg. 3-21]
Dependencies of frontline systems on No Action: Common dependent failure support systems were not " carried over" modes affecting multiple systems are L I across functions included in the linked fault tree scheme of the updated LGS PRA. O S-728510-012 I7 032389D89F
a Tcble I.1 (con't)
.p ,
d
SUMMARY
OF THE BNL COMMENTS ON THE 198:. LGS PRA AND THEIR RESOLUTION IN THE UPDATED LOS PRA COMMENTS RESOLUTION ;
)j System physical dependencies were No Action: - Considered by BNL to be '
covered only marginally in the LOS PRA. outside the scope of the PRA. [Pg. 3-22] Component physical dependencies were not included in the PRA. [Pg. 3-23] Component functional dependencies were No Action: Lmked fault trees are being not included in the PRA. [Pg. 3-22] used in the updated LGS PRA to explicitly account for the component functional dependencies. The vapor suppression function as used in N o Ac.+;oa
- L O C. A e.v e.n+ f re.a s the RSS was not included in the LOCA chan3e.d +o Inclade, vagor sop-event trees. [Pg. 3-23] pre.ssion func.+ ion.
The emergency coolant function ability No Action: WASH-1400 and subsequent was not included in the large LOCA BWR PRAs nave concluded that this O event tree. [Pg. 3-23) event is not appropriate. The frequencies of the initiating events No Action: The initiating frequencies determined by. the BNL approach differ, have been updated using the latest as shown in Table 4.1 from those derived available data, but editing out the first in the PRA. [Pg. 4-5] year of commercial operation. This has been extensively discussed and used in current BWR and PWR PRAs. The probability of the common cause No Action: The failure probabilities of failure of all four diesels used in the diesels have been reviewed and revised PRA was lE-3, whereas, BNL calculated 1.9E-3. [Pg. 4-7] The value of 2E-3 used for the "X" event No Action: See discussion of "X" values in the PRA was regarded by BNL to be supported by simulator data. optimistic. [Pg. 4-8] O S-728510-012 I-8 032389D89F
-._. . . ~ _ . _ _ . . . _ . . . . ~ . . .
f
/
u C o $ n V d 5 e 1 n [ s a H t e P 2 2 1 1 1 I A 2 1 6 4 3 I ! G E l ( fff ( l I u H 3 { A A 0 lil il il u A D A D W p u 1 D l l l l ll D 0 i r 1 0 0 2 1 l 4 3 666 3 ) ) CBA ) ) N 4 DDD A p It 1 H 3 DB0 E 3 { tfC ( ( tft tfC { 0 urh 4 2 orh orh 2
- oa oa V ea 0 0mn 1 1 1 mn g A B 0sn g 1 g 0 0 l H
. 0 - )
1 10e - - 00e u L 10 e d 4 6 . d f d 4 ) ) 1 0 ) 0 C 1 1 h a n g wfal bflT (sl srtffi D e asl waA a d vtI a hn a ath raoh amh eere tsierp on E s bai cep sioe eisi lal 5 aan cfd u u lr alts tls tp d u C l httl tgv o le aa sl a uae uid
.mrauf ei l
uat sp pra uad a lu orti seome soc h f lp t ku eral refo s (fot ocu p hl eai rsn p osu ntt to. ara 0aA tIall 1tWruu uft aa tif ro e f a l p l 0 a L S G jt) mu l pmtd ee o oe l l n ee aw a$ae0 f f p ue( rt sr Vtdw !f 4 R ssb ar . rd thei hea oaa s s to0 A tal de e o 1 ens f edef1 i e e tt vd 4 s s or 0 f fp o oo t 2cea 2ua iu t a yrj of sa oAa Ila ord )p d tbaha u ee f uPv v tn d p d sl es l 4sr t tc ac sIea hdve ble es t
. s th e read eyd v p ap 0ua ha ro r si l onc l acupa I 1rp e tu usut alt lorag r e n au1a ea p1 b g euo g io uulre e f r cf tt h vt 6l hrm e enaa e ara of ng ef au e a ssa t ut3 s iet lr f iwbc t c oes o1 lg o da o n Vaec urc nd r a uur es mr e
ueot e l o - cco ceo l rl n dl 1 uau klu pfetc eat ss ey cf ha oiot 4uun oet in aov oc g re o to ai f a uh 5sst ro rs p l e ) e p f e lru arb t e nl i u 0h a . f esf t en t y rg o oo o Ih8st
.e aawp s c r ae 0a p fr .n r i te a l 1g ae ws od r
ss o . a r a s s f s - s e t o o if el I L1 pf V sa GR _ v sA ua a x h m SE _ a uc mi l i pl S - A l u bc to pl u il nu ls T PI r u u e cr i ce RM e s h AE _ ent ir se le iv a N w w u te _ a e. e e do l n s FT s d it ef yt A no r d b UI n f c e . v ii e LN - o l s a ns e T o r ut n c c n t da o u l n T c i er t u uo d R s e
.d t n dt E h c e l E a v h e S n S a b g J a n r .d t e
_ e A c g e u a - .d E u e k .d s .d e r s
E ~
- a. = .
ag
& = W e = W l .d bi .d M
o c [ e,
> 't.3 .J "O O
C
=="
Q C
.= >
W g Jl. . O *= *= ILI W g u =- . *d b.
. . .U i= = .
s.- u ~ .d o .2 &c W 3 > 3 tu 3 g = N-J . - u. u
.E == . . - . , -- u C < u 6.
O "O Q . Q . . 6W as
- g . . J g-
= gg eft *T y . in. .g g - .
w5 xc .,. uo
===. 2. . =- . =o . - e. .e ,. .,.,., ,. . .e c - ..a.
a mr a. -
.u e .o =a =e g .e .
p. w w= 6 .Q *= W ll> *d == u
. .= w e
8 .O. e. I =
= - - . m
- 4. 6.
m ..
.o.v ., ...o n w .= = .= -- .C C. .b .2 6. . e.d - . .. . . . "..J . .= . 1 .
v . r= -. u u. ... . r= . .oe
.a..
e- os
-o . .., u.o p .- . . u . .e .o .,
J ---. . .om
- c. c.
Q .u . .s b . .s. a .
.. 3 o. .a o.
o=
. - .< -- a. m. . .
r, -
.n . ... .a. o. .
o.
= = < ..
w . .==- e -
.,. -- .= = - a. = w - a. ,. - =, .* = e. .. '.- .. = . . .-e.
i, E . ,. 3. . . . .o p>
. ^
o
- . . ii! .. . -. . ,.s... . . ~ . .
n .. , - ... .. ==, . 3-m; .. . .. . ==.
.w...
J - w
. o 3-Z_ 3- . . .. a.
o .9. , a . 3 w.- mi. .
.a :li! .... . ,~J ,,
P o s.t
. ~ . .
W
-r . .C. C 6 . C. *J h 09 .O C.
( Wl
.C 95 ~as., .c M. f".9". .C .
n. i $.$. .g .o N b -
=
2 . E M. N. O u" E.. g.. .wg c . N O Q
.=
4 w .1..O=.C= = e =. .T. w p . me - y - 8 M O W9
.C .o C S $ .g "" 3 .uG .e we c w g 4 C c
g-
.c =y 2 gyn . 4 Q 2.C at c .=. 0 .d= . as C g cm C gcg W o..=.
3 i es g art. % 2 .g .g se en=g e.re 26 gg . g "J a
. we esig w wg 8EP sp g .2 g. #g*
3 m g lll3
,=g 55 .- #ww- # # ." ~. a ww z w ,, , .
ns .h 889
.n.
43 .S.e
.O ut f.9 Sb i
5 El
=.e ef.
c
.=. o .A t *5 z= 4 e w'i l o w *' bi W.
3 c. O.C S.728510-012 I-13 g 032389D89F 1
i' 11 _. n _ e . o id r . d d t e a d e e ad e d ru g u d m s u bl e e n l S c l i c r- a c l n E u h n n ai E c i R l , i c i c s s T T f a t o i g s e ir mos m NL n o r m I U h c e l u en A r l de TF t e e i os N i . w t a m E A. wd a f k MR se s g nn d e oa TP t u e s mt A - ES il u "R e m RG mc l a 0 h of in T Co TL Li V " r f' o f - ot - oe o o t orf s r ul
- t moo ee rn ve ni g = e ec .
d f 6 w rusn eh 7 id en 3o u e usee l p gte2 eo bi l t l nl l eie S i pl o awsf ltll oa f og io.k a E aei h ( o a n n E f eae pf r ct ne6 geof rp ei f wa R r a sat Ru 8 rl rst oes at ae Gl n uo o T fh es the wsrg i a nal o i er l o ic ft W eP a t T t eftl cv 4 sep
- nr afs s1 L fr e
see L vol a ooun ore 1 t or0 o ush aut U o uv l cle f p0 t) ca ces u i c pr t ccf
) A cs auth e ltaI .
it u ef . aeo d e P alt I.ifP) 0e gnss ou e u deeNg d ob n u oce 1ee8e tse n A telt bla fal0t kv6r l oev chl s(g o pt i t R 8 8 t h omeVvlt ha ecta v t u!l t . t uwa n P- 8 uset ppv eo s Il t o p5 c ppr o l nel e vtd u otf o9 o b . c S i r ieat l eea dt o d. l1s n lCl s e e Ll r ( G i h d pva c a3tlf v- ap grgee dee eve dSae e L eerss ts 2 1 l e N I C S E D eIsi dCed dFia A(tl b: pmh 0 ras e1 eai I3 o et adaso h rh al Coctc dl r dac Avs jlse dll e Aaas G b a S. ) T P 4
)
43 3
)
G - - - ) 0 3 N E o d00 e11 4
- 1 -
D l u a 0 A l I A g u g: : n53 t n 2 1 s V . a2 1 2 l l h o I ( C ( t1t ( ( L N I [ H lIIl i I A It H ePPr t l l il ii Pr ) ) 2
)
I tADt Do 1 B D T llIl AB 0 D U U ooGo 11 D A A P WHiHl 6$ A ( ( N 55$5 Vt ( I 080B II 1 DB 4 I E 4 9 G d d 1 2 A n n 1 P a a 3 5 H t f 5 f
- 5 I I C C R L P il S I B
?' d $ g 6 y _g #
8't$O$*n \li l\' (,lll f! i
4:
~
RELEASE / CONSEQUENCE MODELS I.- RELEASE FRACTION
~o Release fraction (source term). calculations in 1982 PRA, o- One representative sequence per accident class.
o In-Plant P-T conditions from INCOR (INCOR = BOIL + PVMELT + INTER + COMTEMPT-LT)
- a. o Release fractions . from CORRAL' (Wash-1400) using INCOR data, and various containment failure modes.
l l :. 4-1 1 t
=- - _ _ _ _ . .
i ( . '~i
-k . RELEASE / CONSEQUENCE.MODELS II. CONSEQUENCES-o- Consequence calculations in 1982 SARA.
o Consequence results from CRAC2. PRA/ CORRAL release fractions Containment failure modes based on 1983 PRA o Consequence characterized as accident class occurence conditional. For example, given the occurrence of a Class IV accident sequence, the conditional 50 mile total is
.2.7 x:107 person-rems (per occurrence).
O. 4-2 l' L L O l..
III ' l f6 L. ! It ' ' a g
^'
RELEASE / CONSEQUENCE MODELS L. I III.- RISK i- . o Accident frequences in updated (1989) PRA results.
-- Includes internal, fire, and flood initiators.
(;. For example, the sum of the estimated fregency of occurrence of all Class IV accident sequences is 1.05 x 10"I/ year. o SARA conditional accident class consequences. o public_ risk estimated as: Accident class Conditional Class Frequency Consequence f-For example, for Class IV ace
\_S / -offsite' exposure = 1.05 x 1.Cf,idents, /yr x 2.7 x 107 person-rems = 3 person-rems / year (unmitigated)
() 4-3 l ___________a
*='% +
i i g/ RISK REDUCTION BENEFIT I. METHOD o Risk Reduction Evaluation RR = F i x [ P, x (C,i - C,5 ) ) i Where: c Fi = accident sequence class frequency for class i P, = conditional probability of mitigating this sequence with a specific SAMDA C , = the conditional consequences (population dose in man-rem) for the unmitigated sequence in class i C,i = the conditional consequences for the mitigated sequence in class i
,e3 o For each SAMDA (j), estimate for each accident class (i):
i/ Risk Reduction, averted person-rems i,j as Accident Sequence Class Frequency / Year i times Probability of Mitigation by SAMDA i,j times (unmitigated-mitigated) population dose for sequences in class, person-rems i o Sum over all classes to obtain total risk reduction benefit for given SAMDA (
'^^
5-1
l' i
/
ixs RISK REDUCTION BENEFIT II.
! MITIGATION EVALUATION o
Probability of Mitigation, Pm Based onprogression. accident engineering evaluation of SAMDA and Numerical table: Probability assigned according to the following Qualitative Assessment Assioned Mitication Probability Very likely to be effective
.99 Highly likely to be effective .95 Likely to be effective .9 Indeterminate .5 Somewhat unlikely to be effective .25 Unlikely to be effective .1 Very unlikely to be effective .01 Impossible (or extremely unlikely) O.
to be effective o Consequence Mitigation Effectiveness O Majority of cases: mitigation that is, SAMDA considered capable of complete mitigated population dose = 0 Some cases: fission product transport paths resultAssessment of actual mitigatio incomplete mitigation effectiveness, that is,in assigning an mitigated population dose > 0 i 5-2 i L__ ---
b
+ '0 ,
RISK REDUCTION BENEFITS III. EXAMPLES RUBBLE BED CORE RETENTION SAMDA CLASS Em MEF R.R. NOTES-1 .25 1.0l 12 Some Debris remains in DW.
~2 0 -
0 OP CF occurs 3- -.25 1.0 1 Same as Class 1 L 4 0 .- 0 OP CF occurs
- Total 13 Person-rems /yr.' averted DRY CRUCIBLE CORE RETENTION SYSTEM SAMDA CIASS Bn MEE 'R.R. NOTES
- 1. .95 1.0 45 Prevents OP/OT CF 2' .9 1.0 1 Prevents OP OF 3 .95 1.0 1 Same as Class 1 4
.95 .37" 10 OP/OT CF occurs, but DW sprays effective Total .57 Person-rems /yr. averted 'MEF = (Cumi -
Cmi)/Cumi, all in person-rems is mitigation
. effectiveness factor- " for example, MEF=(2.7 x 10 7- 1. 7 X 10 )7 / 2. 7 x 10 7 r
( 5-3
\_3/
(- 1^ . L __.. _--_-_-------- _ _
jm C.) _ CONTAINMENT ANALYSIS GENERAL: o_ DISCUSSED IN SECTIONS 3.4.5 AND 3.5.4 OF LGS PRA. o CET FIGURE 3.5.6 l ( ) 1 L 1 L l i 6-1 f (] l
[I-CONTAINMENT ANALYSIS l SPECIFIC' QUESTIONS:
'A. 0.5. PROBABILITY OF LEAK TO PREVENT RUPTURE 0.5'LARGE > 100%/HR.
0.5 SMALL < 100%/HR. B. . SUPPRESSION POOL BYPASS CONSIDERED ONLY AS RESULT OF CONTAINMENT RPUTURE IN~PRA/ SARA EXPECT BYPASS AS A RESULT OF' DRAIN FAILURE AT 6 MIN. CONSERVATIVE PRA/ SARA SOURCE TERM MEANS' THAT IMPACT ON RISK IS SMALL. IF_ GAMMA PRIME (VAPOR SPACE) FAILURE MODE IS ASSIGNED A GAMMA (DRYWELL) SOURCE TERM THERE IS ONLY A 5% INCREASE IN POPULATION DOSE.
.b -
POOL BYPASS WAS ACCOUNTED FOR IN DETERMINING
}, . EFFECTIVENESS OF SAMDAs.
C. APPROXIMATELY 15% OF CDF IS FROM TQUX (HIGH PRESSURE) TYPE SEQUENCES. AN ADDITIONAL 14% OF CDF IS FROM SEQUENCES WHERE ALL DC IS LOST. D. PRA UTILIZES 0.01 AS UNAVAILABILITY OF INERTING. DE-INERTING FOLLOWING VENTING NOT SPECIFICALLY EVALUATED. 6-2 l ______________J
j,: - U l
- f"%
U
~ !,l i =
ja]! IIg i si l l il .i .!
- l. fi ri 15 5! = =
C Is ti l! I li ! !l a it !I !:1111111 g5 .l I. I. I
. E. .I !. I . .... IIII L i .
5 -
. '. . a : i . 't ..
il . 8! E
& .Q 3 2- 5 C
Ig*f 5 'I 'C be j- %. %.. -8
-c t ?r \alse 4 s i
NlI . - sa
.I : *N :
ea 65 N i I I I j !I ! 9o b> d. c - *= e-3 I . 5 l I-
- 1 I W I * *
~
j:rg b: c g*. Ie
*U ill! i
- 18 I,
,I e
- 8- {. I
{gf 3: Ir I - r m i 2]!11
- 1. s
.i. i. .
l -
)!' l5 ! 5 j 8 < $ ! l[fi-EII 1*! I f"
l
): s
- I !sl: yl Il a]I t
3
- s. l:I IIiEEg3 s
i r;Il d I
- l: ! =j
+
i. li ' -I I'l s !! isi 1;p!!!!!!
*j w 1 6 BI
( , 3-114 Parv. 4, 6/82
, !, ?. . i : f-
.j : ;g .... l t, - '!!I,I- I : I: :
f; '5 i ,, 8 s' .. (. l. UIF. :
.i.lI' a . .,..,I.lylv.
- ~
{ i. l. II ,l'il
- t. 1. .wi.
gg .gg i
,t..'I 1.:
As ., g
.a B l i..83.8 P !E g, E o g 3 s. e i i .i l'.I I I. I v . .
v . . I t' I 5 . -J -*'i2 = .
.a e - l .
s . =
.'- c' li I _.* _
3c O
- v- ?.!e:l. .a 5 g !'* : : -
_w
*I: '
i!I f E frfI :. e e
'5 ae i i c$
3!j.:: - I I o I -!:II 6, g i i i 2Em
; oe oo ..f 1 II 5
L La a c-a
- g. [. co 4s o>-
>w w .;1 . ~~
3
.!:eg II e
om ee ce mv i*i
- 8h; r:l!
3! *s ; . r u+
, I. -
n
.*-l .c Mi, hi eyt .ir i .
II* s
- 1:! ]},
M. m I .= 25I :.
- .. o w
fd v - (!! It5 . 2! 3 s m-
- [tj way Sg $ a
- f. .
sI .e s: s.a .e g
<: r5!
u. F*
<tt <
Il,lis61c::I-! II r - 5 3* .
!n.I :r:1. g !1 a !
l s.
- ..l.
I !'8 3 li : 2 l' 8
![ ~
w i
!!..lli i..I!'5.!!.!
5 5I s 3-115 Rev. 5, 9/82
__m.
/ I I f' ..,a" DOMINANT SEQUENCES EARLY FATALITY CONDITIONAL EARLY ' SEOUENCE.. FREOUENCY CLASS FATALITY RISK.
TSRPV' 4'.8E-07 3 (1.6E-07) 0.58 S'(3.2E-07)' '599
, TCP2 LHV ' 2.0E-07 4 173 1 . TMP2 LHV .' 1.2E-07 4 173 p: TTPPV' 1.2E-07 4 173 - TCP2V' 7.1E-08 4 173 ~ TTPPLHV' 5.SE-08' 4 '173 7-1 O
L -_ ---___--_.____' _ - _ . _ :i
h: ' 4 x FIRE RISK ANALYSIS
)
DOMINANT .j DOMINANT SEOUENCE AREA CONTRIBUTOR l l'" F44QUV SAFEGUARD SYS CABLE FIRES (FGS 2) l ACCESS AREA 1 F2QUV' 13KV SWITCHGEAR CABLE FIRES (PGS 2) ROOM. CABLE FIRES (FGS 3) PANEL FIRES (FGS 2) d F45QUV CRD. HYDRAULIC EQUIP CABLE FIRES (FGS 3) AREA CABLE FIRES-(FGS 2) F47QUV RWCU COMPARTMENTS. CABLE FIRES (FGS 3) AND. GENERAL EQUIP CABLE FIRES (FGS 2) F2QUWFWECC 13 KV SWITCHGEAR CABLE FIRES (FGS 2) PANEL FIRES (FGS 2).
!O O e-1 l
- l-__.-_.-_______
f
.(g' /..
c .
. o MOST OF THE FIRE INITIATORS WERE ASSUMED TO RESULT IN AN MSIV CLOSURE AND THE MSIV-CLOSURE EVENT TREE WAS p
QUANTIFIED FOR EVENTS D & F. THIS IS THE WORST CASE SCENARIO. o THE FIRE SUPPRESSION PE3BABILITY CURVE IN NUREG/CR-5088 IMPLIES THAT THE TIME USED WAS THE TIME REQUIRED TO COMPLETELY EXTINGUISH THE FIRE. IN REALITY, AS SOON AS SUPPRESSION ACTIVITIES COMMENCE, THE TIME BEFORE CABLE DAMAGE IS EXTENDED DUE TO THE REDUCED HEAT FLUX. o THE DOMINANT' FIRE SEQUENCES ARE, IN GENERAL, SEQUENCES IN WHICH THE FIRE PROGRESSES TO FIRE GROWTH' STAGE 2 e l
\ (I.E., DAMAGE TO ALL UNPROTECTED CABLES) . NO CREDIT FOR THE AUTOMATIC SUPPRESSION SYSTEMS WAS GIVEN IN THIS EARLY STAGE AND ONLY MINIMAL CREDIT,WAS TAKEN FOR ANY SUPPRESSION DURING THIS TIME PERIOD. PORTIONS OF THE AREAS ARE PROTECTED BY AUTOMATIC SUPPRESSION SYSTEMS WHICH COULD SUPPRESS THE FIRE EARLY.
i O 8-2 l
m_.__.m_.___-m.--._.2m._ _._m. m_ ._
p , i L 1 CONSERVATISM INCLUDED IN LGS DOMINANT FIRE SEQUENCES-o FIRES.WERE ASSUMED TO DAMAGE ALL CABLES INITIALLY WHICH
.ARE ASSOCIATED WITH THE.SDM IN WHICH THE FIRE STARTS.
s o FIRES WERE ASSUMED TO DAMAGE ALL CABLES IN ALL UNPROTECTED SDMs IMMEDIATELY IN FIRE AREA 2 AND IN 10 MINUTES FOR OTHER AREAS. o NO ATTEMPT WAS MADE TO IDENTIFY THE CRITICAL AREAS WHERE MULTIPLE ' UNPROTECTED SDM CABLING RUNS WERE IN REASONABLY CLOSE PROXIMITY.
- l
. A. o THE BASIS FOR THE 10 MINUTE PROPAGATION TIME AS COMPBRN I CALCULATIONS ASSUMING THE MINIMUM CABLE TRAY SEPARATION.
o THE MAJOR CONTRIBUTORS TO THE DOMINANT FIRE SEQUENCES WERE, IN GENERAL, CABLE INITIATED FIRES. THERE IS A LACK OF BACKGROUND INFORMATION ON THE HISTORICAL FIRE DATA
- l. REPORTED IN NUREG/CR-5088. THE THREE FIRES USED TO DETERMINE THE CABLE FIRE INITIATING FREQUENCY WERE QUESTIONABLE AS TO THEIR APPLICABILITY SINCE THE TYPE OF CABLE INVOLVED WAS'NOT KNOWN. ALSO LGS USES IEEE-383 RATED CABLING EXCLUSIVELY AND AS SUCH MAY NOT BE AS SUSCEPTIBLE TO CABLE INITIATED FIRES.
O 8-3 -. - . __ _ _ b
4-l?$
, . (f -
TRANSIENT FREQUENCY l o- SINCE COMMERCIAL OPERATION. BEGAN-(FEBRURARY.86 THROUGH MAY . 8 9 ) LGS ~ HAD 8 TURBINE _ TRIPS (4 - MANUAL /4. AUTO) . FREQUENCY OF TURBINE TRIP IS'8.5*/3.33.OR 2;55/YR. (* 0 ' 5 EFFECT ' OF BAYESIAN UPDATE WITH NONINFORMATIVE PRIOR) o NEW TRANSIENT 2 FREQUENCY 6.69 VS 9.74. ( 9-1
V;:: l( .
;- TABLE 2-1' l
SUMMARY
OF THE FREQUENCY OF TRANSIENT INITIATORS Frequency. Initiq1;pI iPer Reactor Year) l Nov 88 Uedate. pig 3ent Uedate Turbine Trip 5.6 2.55 Manual' Shutdown 3.2 3.2 1 MSIV. Closure 0.23 0.23 Loss of Feedwater O'.19 .0.19 Loss of-Offsite Power 0.074 0.074 Inadvertent Open Safety Relief Valve 0.07 0.07 Loss of Condenser Vacuum O.38 0.38 Total 9.74 6.69 0: o l 4
j V 4.0 SAMDA EVAIDATIONS 4.1 Methodology The Severe Accident Mitigation Design Alternatives (SAMDAs) proposed are listed in Table 4-1. An input into any decision on the need to install any of these SAMDAs is an evaluation of the value and impact or benefit and cost of the SAMDA. The major benefit of a SAMDA is the reduction in severe accident risk that J the SAMDA provides. The usual measure of risk utilized is the I mean population dose (i.e., person-rem) integrated out through 50 ) miles of the plant. This is consistent with past NRC value- ! impact analyses practices (Ref. 14). The population dose risk ; reduction (person-rem was converted to a dollar benefit using ,
$1000/ person-rem as the monetary equivalent of a unit dose. j (Refs. 14 and 15)
Hence, the annual risk reduction benefit was calculated as: Annual Benefit ($)= Annual Risk Reduction (man-rem / year) X
$1000/ man-rem 1 The present worth of the annual benefit was calculated using the following formula (Ref. 15):
(1 + r) t _1
/~' PW = Ca = 9.56 Ca \
r (1 + r) E Where: Ca = the annual benefit ($) r = the annual discount rate (.1025 from PECo) t = the remaining plant life (40 years) The risk reduction potential for each of the SAMDAs considered in this analysis was evaluated for each accident class and for each t release category associated with that class. (Definitions of l SARA accident classes and release categories are contained in Reference 1). The basic approach to evaluating the risk reduction potential for a SAMDA was to estimate the probability i that an accident sequence in a given class and release category would be mitigated by a specific SAMDA and to assess what the population dose would be for the mitigated sequence. The risk reduction for a SAMDA was evaluated as follows: RR = Fi x [Pi x (Cumi - Cmi)3 i i where: Fi = accident sequence class frequency for class i 1 q < V Pi = conditional probability of mitigating this I sequence with a specific SAMDA l 46 1 l
( - I
.c L
i- TABLE 4-1 SEVERE ACCIDENT MITIGATION DESIGN ALTERNATIVES EVALUATED E o POOL HEAT REMOVAL SYSTEM A separate independent dedicated system for transferring heat from the suppression pool to the b spray pond utilizing a diesel driven 3,200 gpm pump and heat exchanger without dependence .on the Station's present AC electrical power .or other systems. The diesel is cooled with water tapped off the spray pond suction line. o DRYWELL SPRAY A new dedicated system for. heat and fission product removal using the Pool Heat Removal System' described above to inject water into the drywell. o CORE DEBRIS CONTROL (" CORE CATCHERS") Two techniques, either a basemat rubble bed, or using a dry crucible approach, to contain'the debris'in a known stable condition in the containment. o ANTICIPATED TRANSIENT WITHOUT SCRAM (ATWS) VENT A large wetwell vent line to an elevated release point to remove heat added to the pool in an ATWS event. o FILTERED VENT Drywell and Wetwell vents to a large filter (two types
- gravel or enhanced water pool) to - remove heat and fission products.
o LARGE H2 RECOMBINER Independently powered recombiners to remove H2 from the containment in the long-term after a severe accident. o LARGE CONTAINMENT VACUUM BREAKER To restore containment pressure to atmospheric level through 20" valves in certain severe accident cases where a vacuum has been produced. e 47
)
h k/
= the conditional consequences (population dose in Cumi man-rem) for the unmitigated sequence in class i j i
Cmi = the conditional consequences for the mitigated j sequence in class i The rationale for the selection of the mitigation probabilities and the mitigated consequences for the individual SAMDAs are I presented in the following sections. Several broad generic j assumptions were employed which impact all SAMDAs. These are listed below:
)
General Assumotiong I q
- 1. The probability for mitigating steam explosion and hydrogen burn containment failure sequences (release category OXRE) was assumed to be zero (for all accident classes where the SAMDA does not prevent core melt).
- 2. Seismic and large reactor vessel rupture sequences were assumed to be unmitigated.
- 3. The mitigation probabilities were assigned based on the following assessment strategy.
e^ Qualitative Assessment Assioned Mitication Probability b' Very likely to be effective .99 Highly likely to be effective .95 Likely to be effective .9 Indeterminate .5 Somewhat unlikely to be effective .25 Unlikely to be effective .1 Very unlikely to be effective .01 Impossible (or. extremely unlikely) O. to be effective
- 4. Class 3 sequences characterized by failure to shutdown the reactor with loss of core coolant injection are very similar to the Class 1 (loss of core coolant injection following a transient or LOCA initiator) sequences. In the SAMDA benefit analysis it was always assumed that the Class 3 sequences were mitigated to the same extent as the Class 1 sequences by a specific SAMDA.
- 5. All risk values (man-rem / year) are rounded to integer values.
- 6. Seismic population dose risk was not included as specified by NRC question 2.
Cq 48 i
y ~
' (~/
Evaluation of Benefit of Each SAMDA 4.2 j
~4.2.1 Dedicated Suppression Pool Cooling System (DSPCS)
I class 1 Secuences The DSPCS is unlikely to be effective in mitigating Class 1 loss j of ~ core coolant _ injection sequences since no mechanism 'is g provided for preventing drywell overtemperature failure following j vessel rupture. Furthermore, this SAMDA does not provide for any > mitigation of radionuclides release to the environment. j Mitigation- Probability (Pm ) = 0.1 If the DSPCS is successful in preventing containment failure then the. accident source term is very small. Mitigated Sequence Consequences (Cm = 0 man-rem) RR1' = 8.84 x 10-6 x 0.1 (5.4 x 106 - 0)
= 5 man-rem / year Class 2 Secuences The DSPCS likely to be effective in preventing steam is overpressure failure and core melt for the Class 2 sequences.
Mitigation Probability (P m) = 0.9 If containment failure and core melt are prevented no consequences are expected. Mitigated Sequence Consequences (C m) = 0 man-rem The Class 2 risk reduction is then approximately: RR2
=
1.75 x 10-7 x 0.9 (9.3 x 10 6_o)
= 1 man-rem / year Class 3 Secuences Class 3 sequences are similar to Class 1 sequences (mitigation probability and mitigated sequence consequences are the same as '
for Class 1). Hence, the Class 3 risk reduction is approximately: RR3
=
2.73 x10~7 x 0.1 (5.4 x 106-0)
= 0 man-rem / year
( 49
l 1 73 (j Class 4 Secuences It is extremely unlikely or impossible that the DSPCS will be effective in mitigating ATWS sequences. The design heat removal capacity of this system (~ 45 MWt) is far below the heat f production rate during an ATWS ( ~ 10% of full core power or 330Mwt). Hence, this system will not prevent containment overpressure failure or core melt. Furthermore, this system provides no mitigation of the radionuclides released during the accident. Mitigation Probability (P m) = 0. Risk Reduction (RR 4) = 0 man-rem / year Summary-Dedicated Suppressio11 Pool Cooling System Class Risk Reduction (man-rem /vear) 1 5 2 1 3 0 4 0 Total 6 p 4.2.2 Enhanced Drywell Spray System (EDSS) b Class 1 Secuences The EDSS is likely to prevent both containment overpressure and overtemperature failure for Class 1 sequences since the drywell air space and the core debris are provided with a cooling spray of water. Mitigation Probability (P m) = 0.9 If containment failure is prevented a very small or zero source term would be expected. Mitigated Sequences Consequences (Cm) = 0 man-rem The Class 1 risk reduction is then approximately: RR1
=
8.84 x 10-6 x 0.9 (5.4 x 106.o)
= 43 man-rem / year Class 2 Secuences i i
The EDSS is likely to prevent containment failure and core melt since it provides the containment heat removal function which has (qj been lost for these sequences. 50
n
'( ) _
Mitigation Probability (P m) = 0.9 If containment failure and core melt are averted then the consequences will be zero. Mitigated Sequence Consequences (Cg ) = 0 man-rem The Class 2 risk reduction is approximately: RR2
= 1.75 x 10-7 x 0.9 (9.3 x 106-0) = 1 man-rem / year Class 3 Secuences Mitigation probability and the mitigated sequence consequences are similar to Class 1. The risk reduction is then approximately:
RR3
=
2.73 x 10-7 x 0.9 (5.4 x 106-0)
= 1 man-rem / year Class 4 Secuences This system has an insufficient design heat removal capacity to r prevent suppression pool heatup, steam generation and containment
(_S / overpressure failure for ATWS sequences with power levels near 10% of full core power. However, assuming that the EDSS system can survive containment failure it will provide some mitigation of the radionuclides release due to spraying of the drywell gas-space. The probability of mitigating the fission product release by spraying the drywell is: Mitigation Probability (P m) = 0.9 We assume that spraying of the drywell gas space will reduce the source term (and offsite consequence) for these sequences to that of the OPREL release category. This reduces the overall consequences by a factor of approximately 1/3 from their unmitigated values for Class 4 sequences. Mitigated Sequence Consequences = 1.7 x 10 7 man-rem The Class 4 risk reduction is approximately: RR4
= 1.05 x 10-6 x 0.9 (2.7 x 107 _ 1,7 x 10 7) = 9 man-rem / year n
51
~j y v Summary-Enhanced Drywell Spray System Class Risk Reduction (man-rem /vear) 1 43 .2 1 3 1 4 9 Total 54 4.2.3 Rubble Bed Core Retention System Class 1 Sequences The floodable rubble bed system is judged to.be somewhat unlikely J
in preventing overtemperature drywell failure since no cooling is.
-provided.for the debris that does not relocate to the rubble bed from the drywell pedestal. area. However, this ' system should reduce the probability of gross overpressure' failure of containment by.providing cooling to the majority of the debris. . Mitigation Probability (Pm ) = 0.25
- If containment failure is prevented the source term will be very
.small.
Mitigated Sequence Consequences (Cm ) = 0 man-rem The values result in an approximate Class 1 risk reduction of: RR1
=
8.84 x 10-6 x 0.25 (5.4 x 10 6 0)
=- 12 man-rem / year Class 2'Seuuences The rubble bed system does not prevent overpressure containment failure or core melt for loss of containment heat removal ,
sequences and its mitigation potential for these sequences is very small. Mitigation Probability (P m) =0 1 l' RR2 = 0 ma'n-rem / year Class 3 Secuences l L Class 3 Sequences are similar to Class 1 sequences (mitigation g' probability and mitigated sequence consequences are similar).
.O Therefore, the Class 3 risk reduction potential is approximately:
i I 52 i L c_____.______________._. _ __
b- RR3
=
2.73'x 10-7 x .25 (5.4 x 106-0)
= ~1' man-rem / year Class 4 Secuences-The rubble bed system does not provide any mechanism for removing the heat load ' generated by an ATWS event and will not prevent pool heatup, steam generation and overpressure failure of the < containment. Hence, containment failure and core melt are not prevented in this class of sequences.
Mitigation Probability (P m) =0 RR4 = 0 man-rem / year Summary - Floodable Rubble Bed Core Retention System Class Risk Reduction (man-rem /vear) l' 12 2 0 3 1 4 0 Total 13 4.2.4 Dry Crucible Core Retention System Class 1 Secuences The drywell spray and independent heat removal portions of the dry crucible system can remove the heat generated by the debris (both debris relocated to the crucible itself and remaining in the drywell) and it is very likely that both overtemperature and overpressure failure of containment from steam generation or noncondensible gas generation from debris concrete attack can be prevented. Mitigation Probability (P m) = 0.95 If the syster prevents containment failure then the source term will be very small. Mitigated Sequence Consequences (C m) = 0 man-rem The risk reduction potential is approximately: RR1
=
8.84 x 10-6 x .95 (5.4 x 106.o)
= 45 man-rem / year Class 2 Secuences If'the system is activated early in the accident sequence then it is capable of removing the decay heat being injected into the 53
i P ';. i, b1 '
'; e ,
1, 7x . 16 ' OC 1 suppression' pool'and can prevent containment overpressure failure.
.and core melt.
Mitigation Probability'(Pm )'= 0.9-If - containment failure and core melt are , prevented then the j, ' source' term is essentially.zero. Mit'igated. Sequence Consequences = 0 man-rem
~.The risk reduction is approximately:
RR2
=
1.75 x 10~7 x 0.9.(9.3 x 10 6 0) ,
= 1 man-rem / year Class 3 Secuenegg Similar -mitigation. probability and mitigated' ~ sequence consequences apply to Class 3 as to Class 1 sequences.
Consequently, the risk reduction potential for Class-3 sequences is approximately: RR3. = 2.73 x l'O~7 x .95 (5.4'x 106-0)
= 'l man-rem / year ]b.!
Class 4 Secuences The'. dry crucible retention system does not have the heat removal capacity'. to ' prevent containment overpressure ' f ailure from steam production during an ATWS sequence. Hence, this system will not prevent containment failure or core melt for Class 4 sequences. The system can, .however, mitigate the radionuclides release by
. spraying'the drywell atmosphere and attenuating radionuclides in the drywell atmosphere. As for the enhanced drywell spray-system -it'isiassumed that the source term for Class 4 sequences can be.
reduced to the equivalent of the OPREL release category.
= Mitigation Probability (Pm ) = 0.95 (for scrubbing radio-nuclides in drywell atmosphere)
Mitigated Sequence Consequences (C m) = 1.7 x 10 7 man-rem The risk reduction for Class 4 sequences is approximately: RR4
= 1.05 x 10-6 x .95 (2.7 x 10 7-1.7 x 107) = 10 man-rem / year Summary - Dry Crucible Retention System 54
n U L f} v Class- Risk Reduction (man-rem /vear) 1 45 2 1 3 1 4 3 Total. 57 4.2.5 ATWS Vent Class 1 Secruences Following vessel failure the core debris will drain from the E
' vessel onto the lower drywell' pedestal floor. The core debris is then expected to attack the drywell pedestal drain line plate and open .a pathway between the drywell and wetwell air space; effectively bypassing the suppression pool. Consequently, even if . venting is employed to protect the containment against overpressure containment ' failure the post-vessel failure radionuclides releases would be unmitigated by the pool.
Furthermore, the ATWS vent does not protect _the drywell against overtemperature failure due to residual debris in the drywell. Consequently, the probability of successfully mitigating class 1 q sequences with the ATWS vent is considered very unlikely. p-V Mitigation Probability (P m) = 0.01 If pool bypass and drywell overtemperature failure are avoided and the vent is employed to prevent containment overpressure failure then radionuclides will pass through and be mitigated by the suppression pool resulting in a fairly small source term. It is esti.tnated that the consequences would be intermediate between the SAF.A . LEAK 1 and LEAK 2 release categories. Mitigated Sequence Consequences (C m ) = 7.6 x 105 man-rem
.The Class 1 sequence risk reduction is then approximately:
RR1
=
8.84 x 10-6 x .01 (5.4 x 106 - 7.6 x 105)
= 1 man-rem / year Class 2 Secruences The impact of the existing wetwell venting capability in mitigating Class 2 sequences has been considered in the PRA analysis. It is indeterminate whether an independent, hardened, high-capacity vent system will provide additional benefits. l Mitigation Probability (Pm) = 0.5 0 If used during Class 2 sequences the ATWS vent will prevent 4' 55
I overpressure ' ' containment ' if a' lure and core melt and will'- reduce o uthe' consequences to effectively zero.
-Mitigated Sequence consequences (cm) = 0.
6
- The estimated risk: reduction is approximately:
1RR2 . =
.l.75 x 10~7'x .5 (9.3 X'106-0)
F =
'l man-ren/ year L ' class 3'Secuences-The mitigation ~ probability and. mitigated sequence consequences y for Class 3 are~similar to Class 1 sequence results.
L The Class 3 risk : reduction potential is approximately:
.RR3 =
2.73 x.10-7 .x .01 (5.4 x 106-7.6 x 105)
= 0 man-rem / year
[
' Class 4-Secuences .The! optimistic assumption is made that it:is likely that the ATWS vent will prevent. steam overpressure failure and' core melt. THis =
presumes .that core coolant injection can be ' continued until
\
reactor. shutdown efforts are successful.
' Mitigation Probability (Pm ) = 0.9 If containment failure and core melt 'are prevented by he ATWS vent then the consequences from the_ mitigated ATWS sequences will be very small. ' Mitigated' Sequences Consequences (Cm ) = 0. man-rem / year The Class 4 risk reduction potential _is approximately:
RR4
=
1.05 x 10-6 x 0.9 (2.7 x 107-0)
= 25 man-rem / year Summary - ATWS vent Class Risk Reduction (man-rem /vear) 1 1 2- 1 .
3 0 l 4 21 Total 27 i
-(
56
)
l
i l [ e s, U 4.2.6 Filtered Vent System l This section summarizes the benefits for both the gravel bed and nulti-venturi scrubber filtered vent systems, i 1 Class 1 Secuences The filtered vent system will prevent overpressure containment failure. However, it is indeterminate as to whether the filtered vent will protect against overtemperature drywell containment failure. Mitigation Probability (Pm) = 0.5 l If containment failure is prevented then the filtered release of non-noble gas radionuclides will be very low. The consequences of a successfully mitigated sequence can be assumed to be equivalent to release category LEAK 2. 5 man-rem Mitigated Sequence Consequences (Cm ) = 1.5 x 10 The potential risk reduction is approximately: RR1
= 8.84 x 10-6 x 0.5 (5.4 x 106 - 1.5 x 105) = 23 man-rem / year
(' - ') Class 2 Secuences The filtered vent system is likely to prevent containment failure and core melt for Class 2 sequences (it is assumed that it is much more likely to be effective than the existing wetwell vent capability). Mitigation Probability (P m) = 0.9 Since use of the filtered vent will prevent containment failure and core melt the consequences of the mitigated sequences are effectively zero. Mitigated Sequence Consequences (Cm ) = 0 man-rem The risk reduction potential is approximately! RR2
= 1.75 x 10-7 x 0.9 (9.3 x 10 6_o) = 1 man-rem / year Class 3 Secuences r~ Mitigation probability and consequences are similar to Class 1.
V} 57 L.. -. . . .. . .
. I
y ' RR3
= -2.73 x 10-7 x .5 (5.4 x 106 - 1.5 x 105)
L = 1 man-rem / year
' Class 4 Secuences . The. filtered vent system have insufficient capacity to relieve L
~ the_ steam generation rates from an ATWS event at 10% full core power and will not prevent containment overpressure failure .or core melt.- [ Mitigation Probability (P m) = 0.
'The risk reduction potential for Class 4 sequences is then: .RR4 = 0 man-rem / year-Summary - Filtered Vent Systems Class ' Risk Reduction (man-rem / veer) 1 23 2 1 3 1 4 0 Total 25 LO 4.2.7 Large Hydrogen Recombiner This system does not prevent (early) containment failure 'or mitigate radionuclides release for any identified accident sequence. It is viewed as more of a long-term accident recovery system than a short-term accident mitigation system. It is judged that the risk reduction potential for this system- is small.
4.2.0' Large Containment Vacuum Breakers A qualitative assessment by the Boiling Water Reactor Owner's Group (Ref. 16)' of the conditions that would lead to large 1
-negative pressures concluded that such conditions are not expected following recovery of normal containment ' heat removal 1 and termination of venting. Additionally, the reinforced concrete-Mark II containments such as Limerick are not expected to - f ail even for pressure differentials exceeding twice the design differential pressure of 5 psid (Ref. 16). Therefore, the vacuum breaker would not mitigate any accident sequences currently identified.
58 i _ _ . _ __D
. f' \
4.3 Summary of Cost Benefit Results The costs and benefits of the mitigation systems :are summarized in Table.4-2. The table provides the following: Benefit:. .The. estimated risk . reduction in dollars per year calculated from the estimated man-rem per year. averted by the mitigation device (see section 4.2)_ times $1000
.per man-rem.
Total ~ Benefit: 'The present worth in dollars of the yearly benefit assuming a 40 - year plant life and a 10.25% discount rate. L- Total Cost: The total cost of the mitigation device including construction costs and the present worth of annual operating costs over a 40 year plant life. These results are from reference 17. In reference 17, the costs - were estimated for installation at 2 units and were divided by 2 to obtain a per-unit cost. Benefit /. Cost f- Ratio: The ratio of the total benefits to total costs. A A '
-value creater than 1.0 would' indicate a cost beneficial mitigation device.
Cost / Man-rem ; Averted: The cost per man-rem averted. These values were ' calculated as the total cost times $1000/ man-rem divided by the total benefit. A cost less than
$1000/ man-rem would indicate a cost beneficial mitigation system.
The results presented in Table 4-2 show that none of . the mitigation systems examined are cost beneficial. In fact, the results indicate that no mitigation system is within an order of
-magnitude (factor of'10) of being cost beneficial.
f O 59
i h TABLE 4-2 COST / BENEFIT COMPARISON COST / l TOTAL TOTAL BENEFIT / MAN-REM l MITIGATING SYSTEM BENEFIT BENEFIT COST COST RATIO AVERTED y Dedicated Suppression
$25,000/Yr- $239K(1)$25,600K Pool Cooling .009 $107,000 Enhanced Drywell- $54,000/Yr $516K- $46,500K(2) .011 $ 90,100 . Sprays $27,000K(3) . 019 $ 52,300
- Rubble Bed Core $13,000/Yr $124K $38,400K .003 $310,000' Retention Dry Crucible' Core $57,000/Yr $545K $119,000K .005 $218,000' Retention ATWS Vent $27,000/Yr $258K $ 3,900K .066 $'15,100 Filtered Vent $24,000/Yr $229K~ $11,300K .020 'S'49,300 (Gravel Bed) 3r y Filtered Vent ~ $24,000/Yr $229K $ 5,700K .040 $ 24,900 fj (MVSS)
Large Hydrogen $ 0/Yr $ 0 $ 5,200K .0 - Recombiner Large Vacuum Breakers $ 0/Yr $ 0 $ 0 .0 - 1 Denotes that the item is in thousands of dollars 2 'New drywell spray nozzle distribution header 3 Use of existing drywell spray header O. 60 ___________________m
1 L cf -- L J( .1 3.4.3.2 ' Reactor Pressure Vessel' Failure - n %. L Disruptive. failure of the reactor pressure vessel is included in-L the Limerick analysis at 10-7
',3.4.3.3 ~ Interf acing LOCA Thus far, .the LOCA initiators identified in the Limerick probabilistic evaluation are within the spectrum of' LOCAs which are_ typically considered in- the FSAR. These LOCA initiators involve.unisolatable primary system failures inside containment, as such, these breaks result in a transfer of primary system fluid into the dryvell and eventually to the suppression pool, and a requirement for coolant makeup and containment heat removal.
In addition to these sets of accidents, there is a class of postulated ~ events which could result in a loss of primary coolant intoLthe reactor building. 1The differences present in this class of l events from the' LOCAs inside containment include the 1following:
- 1. Isolation of the_ break is possible in' order to limit the-release of fluid to the reactor building.
- 2. In the event of an unisolated break, there may be a high
' ('y environmental stress produced on equipment in the v- reactor building; therefore, equipment used for reaching a safe stable state may be compromised.
The_ frequency of core damage associated with the following large LOCA events outside containment could contribute to the overall core damage frequency:
- 1. Steam line or main feedwater breaks outside containment (within the reactor building)-.
- 2. Breaks in the HPCI/RCIC steam supply or pump discharge lines.
p 3. Interfacing LOCAs in lov pressure systems. 3.4.3.3.1 Approach The evaluation of the large LOCA outside containment in terms of potential core damage frequency can be evaluated by considering two separate categories of effects:
- 1. Prevention
- 2. Mitigation O 3-108 c__ _
i <si i
)
I I ACCIDENT CLASS FREQUENCY BY INITIATOR l (PER YEAR) TOTAL CLASS INTERNAL FIRE FLOODe (TABLE 2-2) 1 4.44E-06 4.2E-06 2E-07 8.8E-06 2 1.42E-07 3.3E-08 - 1.7E-07 3 2.73E-07 - - 2.7E-07 4 1.05E-06 - - 1.1E-06 1S - - - S 1.0E-08 - - SUBTOTAL 5.91E-06 4.2E-06 2E-07 - TOTAL 1.03E-05 9
- Includes Other Special Initiators h 1-1
a-
' Prevention of a LOCA outside containment has two aspects:
1.1 Prevention of .a pipe. or component rupture outside containment.
- 2. Isolation of the failure from the primary system.
- Mitigation is'necessary.for successful execution of'the remaining
= key core and : containment functions if 'the event cannot' be prevented- including scram, coolant makeup, and containment heat removal.
3.4.3.3.2 Limerick Unique Features. There are a number of Limerick unique features that minimize the.
'importance of this initiator at Limerick. These features include the following:
- 1. Cycling of- the interface valves (LPCI and core spray.
injection valves'are cycled on a longer test interval than many other plants; i.e., each refueling outage rather than monthly during power operation).
- 2. 316 stainless steel minimizes the chance of stress corresion cracking induced pipe failures in steam lines and feedwater lines.
() 3. Highly compartmentalized reactor building with steam reilef panels located at precisely the location of possible interfacing LOCA minimizes the potential impact of such.a LOCA on the reactor building equipment which can be used for safe shutdown.
~4. Check valses in the low pressure injection systems are either not air-operated testable check valves or methods.
x of positively assuring they are seated when the reactor is pressurized are available. 2.4.3.3.3 Dunntification There are two types of initiators that can be discussed as subgroups within the LOCA outside containment category. These l l two LOCA initiator types include:
- 1. Pipe rupt. es in high pressure lines attached to the L
primary system which are run outside containment. l
- 2. Interfacing LOCAs induced in low pressure pipe connected to h2gh pressure primary system pipe.
-O.
3-10o
'Large -pipe ruptures. in;:high pressure pipe include nain steam j]/.%, /
lines, feedwater' lines, and HPCI-lines. Other smaller, diameter lines are notL considered as significant challenges to safe shutdown.
- The. frequency of a pipe rupture in the high pressure primary system pipes external to' containment is calculated to be a very low frequency. In addition, at Limerick the isolation valves are specifically designed.to close in the event of. such a rupture.
Therefore, the combined frequency of such -a combination of failures (rupture plus a double isolation valve failure) is calculated to bei negligible relative to other potential core-damage contributors and is not explicitly included in the Limerick'model. The-frequency for interfacing LOCA is far below the more dominant-core damage. contributors. This judgement is based on evaluation of historical incidents. A number of incidents have occurred in BWR nuclear. operating experience in which operator error, use of testable check ~ valves, and on-line surveillance testing of low pressure injection valves have exposed low pressure ECCS. piping to high pressure and high temperature water. These incidents have demonstrated that the real capability of low'. pressure systems is 'not exceeded. .Because ~ Limerick Technical Specifications do not require this on-line testing of .the interfacing val.ves these incidents are considered very unlikely and is.not exnlicitly included in the Limerick model. 3.4.3.3.4 . Summary [
-The potential initiating frequency of a LOCA outside containment due to the rupture of a high energy line or an interfacing LOCA is found to be negligible (less than 10-7 per year).
3-110
3 , gt SPEC;AL lhlTIATORS Core Damage
!nitiator Frecuency internal Flood 8E-8 Reactor Water Level Reference 6.66E-8 l Leg Leak or break 1.8E-8 O Loss- of Service Water Loss of 1 DC Bus 2.74E-9 High Drywell Temperature 1E-8 .
Loss of Instrument Air 1E-8 to 1E-9 Loss of a Single AC Bus 1E-8 to 1E-9 O
---._._____._____.____.____.___j
O . CORE DAMAGE FREQUENCY BY INITIATOR (INTERNAL INITIATORS ONLY) CORE DAMAGE INITIATOR FREOUENCY % CONTRIBUTION-Transients 2.16E-06 36.5
. Loss of Condenser Vacuum (cdf=1.03E-06)
Turbine Trip (cdf=2.81E-07) MSIV Closure (cdf=4.74E-07) Manual Shutdown (cdf=1.95E-07) Loss of Feedwater (cdf=1.50E-07) IORV Event ( cd.f = 2 . 4 8 E-0 8 ) TE-Loss of Offsite Power 2.32E-06 39.3 Station Blackout (cdf=1.42E-06) Common cause Failure of (cdf=3.790-07) Batteries Support State TE1 (cdf=2.71E-07) Support State TE4 (cdf=8.87E-08) Support State TE2 (cdf=8.21E-08)
/~) Support State TE3 (cdf=7.60E-08)
(./ AI'S d Secuences 1.17E-06 19.8 l Turbine Trip (cdf=3.77E-07) Loss of Condenser Vacuum (cdf=3.75E-07)
'MSIV Closure (cdf=2.40E-07) l IORV (cdf=8.56E-08)
Loss of Offsite Power (cdf=7.83E-08) Loss of Feedwater (cdf=1.75E-08) LQC&s 1.58E-07 2.7 Medium LOCA (cdf=1.09E-07) Large LOCA (cdf=4.45E-08) Small LOCA (cdf=4.45E-09) Random Vessel Rupture 1.0E-07 1.7 5.91E-06 100.0 l
/ /-2 l
r
j uU Y COMPARISON OF CLASS FREQUENCIES INTERNAL INITIATORS . l FREOUENCY'(PER YEAR) GLASA PRA/ SARA CURRENT . REASONS FOR CHANGES 1 1.2E-05 4.44E-06 .EOPs, Trraining, LOOP. Modeling, Initiator Frequency, . ADS Modifica-tion. 2 9.6E-07 1.42E-07 Plant Performance and Data,- Initiator' Frequency, Venting 3~ 1.1E-06 2.73E-07 Initiator Frequency, EOPs, Revised Modeling,' Lowering 4 1.3E-07 1.05E-07 MSIV Closure' Set Point S 2.7E-08 1.0E-08 Mean/ Median, Not Included in NUREG-1150. l l 1 l 1-3
H
' COMPARISON OF CLASS FREQUENCIES FIRE INITIATORS FREOUENCY (PER YEAR)'
CLASS PRA/ SARA EPRRENT REASON FOR CHANGES 1 2.5E-06 4.2E-06 Plant Design, New Initiator and Suppression Data, New Plant Model-2 9.3E-07 3.3E-08 As Class 1 Plus Venting 3' - - 4 - - S _ _ OL 1-4 I l l l 1 .
- ..j_ .
, NJ . COMPARISON OF CLASS FREQUENCIES. FLOOD AND OTHER INITIATORS FREOUENCY (PER YEAR) 9.LMLE -PRA/ SARA GEERENT* REASON FOR CHANGE FLOOD OTHER 1 <5E-07 BE 9E-08 Elimination of ' Conservatism 2- <7E-08
.3 4
S l 1 l
- 2E-07 Used I
j I 1-5 O l
13 V . UNCERTAINTY o NOT DONE FOR UPDATE o WAS DONE IN SARA o ESTIMATED BY BNL IN NUREG/CR-3028 o NUREG-1150 PEACH BOTTOM ANALYSIS (3 V l l l 2-1 I l 1 l l l l l 1
,7 .: 'lO < __._.-- -
me; -. 3.- f; $'-
.fl,,;l, _ }
j;:' i.l M , UNCERTAINTY ESTIMATES
.NUREG-1150 SARA NUREG/CR-3028. 2ND DRAFT-a <.
INTERNAL: 6.'5/3'.8 8.9/5.6 6.8/5.4 25 50 37-L. IFIRE B.6/8.2, - 5.3/11 70- 58 r ,- I' p p s . i KEY:- h- -Ratio 95% to-Median / Ratio Median to 5%' Ratio 95% to 5% 2-2 O. m _._______________1.._
1 a l
)
2 f~s f (( )'. ' A.6 . C0%PLETE LOSS OF OFFSITE POWER I Complete' loss of offsite power to a generating station is an event wh2ch as influenced by local factors such as type of , weather exposure, transmission system design, and operatang procedures. Therefore, a local or regional data base .is more suitable than a national data base for predicting the frequency and duration of such events at a specific plant. Limerick ' Generating Station is connected to the Pennsylvania-Sew Jersey-Maryland Interconnection (PJM) System and the remainder of the -PECo System via five transmission l i ne's . Section A.6.1 reviews the PJM/PECO data base and analytical techniques used in this study to determine (1) the frequency of complete loss of offsite power and (2) the probability of recovery of offsite power as a function of time from interruption. The analyses show a relatively high reliability for the PJM/PECo plants. Even so, the use of these levels of reliability in this study is probably conservative since the five transmission line d e.< 2 pn' a t Limerick exceeds the average level of r edundar.ey for the plan:s included in the data base. r~m ; (_) Section A.6.2 discusses the specific case of Loss of Offsite l Power caused by trip of the Limerick turbine generator. A.6.1 PJM/PECo Exnerience Complete loss of offsite power experience for PJM nuclear plants is sum.marined in Tables A.6.1. In total, these plants have an experience of four occurrences in S3.71 plant years. The exposure for each site is calculated as the amount of time at least one unit at a site is operating at or near full power. Time ir which all units a: a site are shutdown is not included because the recovery time is so long that recosery of offsite power is essentially assured before core damage occurs. Additionally, the configuration of offsite connections for plants shutdown are sonetines signif icantly altered to the point where that configuration would be prchibited during power operation. For these reasons loss of offsite power occurrences at hope Creek and Sale while sha:down for outages have not been included. l O A-S3
e*' j% lu) . Table A.6.1 COMPLETE LOSS.OF OFFS TE' POWER Pennsylvania-New Jersey-Maryland Interconnection (PJ") Nuclear Plant Experience from Commerical Operation through December 19S7= Average Plant Exnosure Occurrences Dutntion (Plant-Tears) Olinutes) Calvert Cliffs 11.23 1 350 Calvert Cliifs 1 90
.. J Oyster Creek 11.16 1 Ils Hope Creek 0.80 0 -
Peach Bottom !!.51 0 - i
< Limerick 1.49 0 -
Salem S.S 0 -
,-} \ Susquehant.a 3.02 1 11 Three Mile Island 5.7 O_
Total 53.71 4 142.25
*E::posure time for 1987 was estimated i
The annual frequency of complete loss of offsite power is 4/53.71 = i 0.074.
)
l A-S4 i
w- , . e 1 i ai s I k g [ ' Another:fimportant'. factor' is. the probability ~ of Erecovery of
'offsite' power within. specific times. The PJM/PECo data base was- ^ ' "
i .
, , . g' again. used 'in'this' assessment. The recovery times for.the'four pI : occurrences actual 1y experienced were used to determine:the mean
- ~ ,' Jrecovery time .and..the
- variance of , recovery time. A gamma.
b.6 distribution was then constructed to fit th'e mean- and var 2ance. This, distribution is n's'hown'in: Table A 6.2. TABLE A'.6.2-4 PROBABILITY DISTRIBUTION OF REC 0VERV-TIME l, . .- L. ' Recovery Cumulative L ' . Time _ Density Density-
, 4 (Min.) . Function Function 0.12: 0.00932 0.001 0.57' O.00E57 0.005- ' I .16 = 0.00823 'O.010 ; 6.38 . 0.00730 0.050 23.55 0.00G70 -- 0.100 '
29.64 0.00578 0.200 48.28 0.0049S 0.300 70.07 0.00422 0.400
-96.10 '0.00349 0.500-128.IS- 0.00277 0.600
[/ }:. w '169.79 0.00207 0.700 22S.75 0.00137 0.800 330.09 0.00068 0.900 431.84, 0.00034 0.950 66S.99- 0.00007 0.990 Means = 142.25 STD.DEV = 145.7
-y -alpha'= 0.95 Beta =:149 4 i L.
f L .. O: A-S5 i8
. ; t .. I
_ . _ . _ - _ _ - _ - _ . _ . - - - _ _ - - _ . _ . _ . i
---7.-_
j k) The probability that recovery takes more thtn a given number of hours can be found from this distribution. Spe c 212 ca l ] ) , P(Recovery.of offsite power >- I hour = 0.65 P(Recovery of offsite power > 2 hours) = 0.423 P(Recovery of offsite power > 5 hours) = 0.15 P(Recovery of offsite power > 10 hours) = 0.01596 P(Recovery of offsite power > 20 hours) = 2.76 E-4
.A '.N )
i t f I l l l l l
, A-S6 l
l L-_.__
[ fa
.6 A A. 6.2 . l.o s s of' ~ 0f f site Power Resulting f rom Turbine / Generator Tr2E s
A sudden loss'_of.. a. significant portion of grid generating-capacity due to the lack of. grid stability may result from in-plant' transient esents that cause a-turbine or generator trip.
' .If the sudden-loss'of genetator exceeds the transiens' stability limit of .the f. oca l or regional grid system, then all'.offsite E power to the plant could be lost. Based upon information developed for. VASH-1400, th'e probability for' complete . loss of-offsite power following a turbine or generator trip was estimated ~
at approximately 1 x 10-2.per demand. This failure probability for any- .particular plant. could be lower depending on the transmission systems, the- transient stability limit resulting from high installed capacity, the extent of grid connections with other- large utilities, and the number of transmission lines connecting the plant -to the grid. It i's judged that the conditional probability of such a scenario is substantially less than that assumed in WASH-1400. In order'to support the judgment that a value of 10-3 per reactor year is.a conservative estimate, an evaluation is performed-using the nuclear plant experience data base. Two cases need to be evaluated and summed to calculate a best estimate:
- 1. Offsite pow'er loss due to load rejection at time of transient (Contribution 1).
- 2. Offsite power loss during the time immediately following a shutdown - any shutdown (Contribution 2).
Contribution 1 The loss of offsite power frequency initiated by a transient within the plant can be developed from data which were not available during the WASH-1400 investigation. Using only the nuclear operating experience data, it is found that in more than 700 reactor years of nuclear experience there are no recorded cases of a loss of offsite power being induced by a nuclear plant trip. Based upon these data, an estimate can be made of the frequency of such postulated occurrences: 1 Probability (700 Ex years) (9 Transients per Reactor year) Probability 1.6 x 10-* per Reactor year t- Contribution 2
- h. The loss of offsite power may also occur as a random independent failure at anytime during the year. If it occurs during the 10 t hours immediately following a reacter shutdown, the result may be a test of the plant systems similar to a loss of offsite po"er I A-S7
n- . - _ _ _ _ _ _ . .
,~\ \' J (LOSP). Therefore, the contribution from such 2nstances is calcula'ted below, based upon PJM grid data.
LOSP f requency (per Rx year) = 0.G74 per Reactor Year
= S'4E-0 per Hour Thus, the conditional probability of the loss of offsite power due to randot independent causes during the reactor safe shutdown 2s estimated usin; the failure frequency of .074/ year and a mission time of 10 hours following a shutdown:
S.4E-6/Hr x 10 Hr = E.4E-5/ shutdown . Therefore, the total conditional probabilities of the loss of offsite power during, or as a direct result of, a transient or a manual shutdown are as follows: 2.4E-4 per transient (Contributions I and 2) S.4 x 10 5 per manual shutdown (Contribution 2 only) !
/D N.)
i O (~) 1 A-ss ' L
BNL AREAS OF CONCERN IN 7h. ACCIDENT SEQUENCE QUANTIFICATION l
- 1. Deficiencies in Incorporation of Dependencies in the Various Types of Logic Trees s
- 2. Disagreement With Some System Unavailabilities and Other Event Tree Values i
- 3. Differences in Frequencies of Initiating Events O-
\
- i
1 INCORPORATION OF DEPENDENCIES 1
- Impact of Dependencies Introduced by Support Systems i Servicing Multiple Frontline Systems
- Impact of Dependencies Introduced by Hardware Shared Among Frontline Systems
- Dependence Between Q and W Functions 1 - Dependence Between Q Function and MSIV Closure Initiator ' %/
19
- Dependence Between U and W Functions - Vapor Suppression Function O
0
FREQUENCY OF TRANSIENT INITIATING EVENTS I") (s ORIGINAL HL{L UPDATED Turbine Trip 3.98 8.17 5. 6
- MSIV Closure 1.78 1.23 0.23 Loss of Offsite Power 0.05 0.17 0.074 IORV 0.07 0.25 0.07 f'
I' Manual' Shutdowns 3.2 3.2 3.2 Loss of Feedwater 0.19 Loss of Condenser Vacuum O.38 O 9.08 13.02 9.75 TOTAL
$ gevised June, M61 += 2.55 4. ref les.+ L Gs exprience . i j
O I
. . . . _ _ _ _ _ _ _}}