ML20247M909
| ML20247M909 | |
| Person / Time | |
|---|---|
| Issue date: | 03/29/1989 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1725, NUDOCS 8904060150 | |
| Download: ML20247M909 (109) | |
Text
--- ,
ACAS71- l7aS~
l .
OgI O UNITED STATES
~~
S'/4 4 NUCLEAR REGULATORY COMMISSIOd )
4 i
.____________________________________ l ADVISORY COMMITTEE dN REACTOR SAFEGUARDS In the Matter of: )
)
SUBCOMMITTEE ON )
INSTRUMENTATION AND )
CONTROL SYSTEMS ) ,
w-> g. t ;.
Pages: 1 through 80 fj"6 {
O. _,,,,,, ww% j f , ~ 9{
'y nj u 1,1 < ' J -
Place: Bethesda r Maryland l(\s,} L ,
, , g ,
,;g i J,1 , . , . , ,
, - - m i n ,w, s Date: March 29, 1989 Q y { d kj j Lj t ,f w griA( k ' ' Pd .
_________________________________________________________j i
l HERITAGE REPORTING CORPORATION O ormn 1220 L Street, N.W., Suke 600 i j WasMeston, D,C. 20005 kf't$ -
(202) 628 48tA 8904060150 890329 PDR ACRS T
___-1725 T'DC
4 1
(} UNITED STATES NUCLEAR REGULATORY COMMISSION 1
j ADVISORY COMMITTEE ON REACTOR SAFEGUARDS i I I
In the Matter of: )
)
Subcommittee Meeting on )
Instrumentation and )
Control Systeme )
Wednesday, i March 29, 1989 Room P-422, Phillips Building 7920 Norfolk Avenue Bethesda, Maryland The meeting conynned, pursuant to notice, at 1:00 p.m.
EEFORE: DR. WILLIAM KERR Chairman Professor of Nuclear Engineering Director, Office of Energy Research University of Michigan O Ann Arbor, ilichigan ciUBCOMMITTEE MEMBERS . PRESEtLT:
MR. CARLYLE MICHELSON Retired Principal Nuclear Engineer '
Ter.nessee Valley Authority Knoxville, Tennessee, and Retired Direc tor, Of fice for Analysis and Evaluation of Operational Data U.S. Nuclear Regu'atory Commission Washington, D.C.
(
(Continued on next page. )
(
() Heritage Reporting Corporation (202) 628-4888 l
L_ _ _ _ _ _ _ _ _ - _ _ _ . .
2 L) SUBCOMMITTEE MEMBERS PM 3ENI:
MR. CHARLES J. WYI.IE Retired Chief Engineer Electrical Division Duke Power Company Charlot.te, North Carolina DR. HAROLD W. LEWIS Professor of Physjes l Department of Physics University of California i Santa Barbe.ra, California AC_BS STAFF MEMBERS PRESXNJ_;. .
Paul Davis W. Lipinski G.QGNIZANT STAFF MEMBEk PRESENI:
Medhat El-Zeftawy j
I 1
1 l
l l
1 IT
\#
Heritage Repcrting Corporation )
(202) 628-4888 )
l L
l i
L
1
({} a 1 EB2 GEED 1EHE y i
2 DR. KERR: The meeting will come to order. 1 3 This is a meeting of the Advisory Committee on !
4 Reactor Safeguards,-the Subcommittee on ' Instrumentation and 5 Control Systems.
l 6 My name is Kerr. I'm the Subcommittee Chairman. '
.7 Other committee members here today are Mr. Michelson, Mr.
8 Wylie, Mr. Lewis.
9 Do we still expect Mr. Carroll? !
10 MR. EL-ZEFTAWY: He's supposed to be here. j I
11 DR. KERR: And as consultants, Mr. Davis and Mr. 1 12 Lipinski.
e~g 15 We are meeting to review the proposed resolution
,b 14 of Generic Issue 101, Boiling Water Reactor Water Level 15 Redundancy.
16 Mr. E -Zeftany is the Cognizant ACRS Staff Member 17 for the meeting.
18 Rules for participation were announced as part of 19 the Notice of the meeting published in the Federal Register 20 of Tuesday, March 7th 1989.
21 A transcript of the meeting is being kept and will 22 be made available as stated in the Federal Register Notice.
23 1 ask that each speaker identify himself or I
24 herself and use a microphone. l 25 I presume we have had no requests for any oral or Heritage Reporting Corporation (202) 628-4888 u________________.___ _ . _ _
i 1 written presentations nor have we received any written l 2 input.
3 I would ask at this point if any metabers of the i 4 Subcommittee or our consultants have any comments or 5 specific things that they would like to see covered in the 6 course of the Subcommittee meeting. i l
7 I would be interested in the course of the 8 Subcommittee meeting to learn what is the criterion or what 9 is the set of criteria that determines when an issue has 10 been settled or what determines a fix.
11 I am also--well, my impression is that the staff 12 probably reached the right decision in this case, although I 13 have some difficulty understanding how they reached it. And 14 so I may ask some questions.
15 For example, it appears that one started out to '
i 16 use a single failure criterion in determining what to look 17 at and then somewhere in the course of the analysis decided 18 to go to reliability. And I may be misinterpreting 19 something so perhaps your presentation will indicate it.
20 From the report from INEL I couldn't tell what the logic 21 was.
22 Then I'm also curious about what seems to have 22 been a decision to ignore common mode failures in the risk 24 analysis. Because in moat risk analyses that are now being 25 done. I think it is the common mode failures that frequently O
Heritage Reporting Corporation (202) 628-4888
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ i
/~)
-(/
5 1 provide significant contributions, so somewhere I would be 2 interested in more than just a comment that says "We decided 3 to-ignore them."
4 I also wonder if in the course of this analysis or 5 in further work there is any consensus or if it makes sense 6 to talk about the availability of the level system. That 7 is, an appropriate value for that availability.
8 You recall, we have four other systems, the 9 auxiliary feedwater system, the PWR, sort of an informal 10 target or goal, and I wondered if this study indicated any 11 such goal or even if it makes sense to talk about this.
12 This is more a question or corroborative answer 13 maybe. It's my impression that one did not, in the course 14 of their study, look at what I would call severe accident 15 sequences. You might comment on whether that's the case or 16 not.
17 And I ask this partly because in the IPE's that 18 are scheduled, I presume that one might go into some of the 19 severe accident sequences, at least part way.
20 MR. MINNER: What's your definition of a severe 21 accident?
22 DR. KERR: Something beyond the design basis.
23 And when utilities do the IPE, will it be assumed 24 that this issue has been resolved or will they also be asked 25 to look at sequences that might involve the water level.
O Heritage Reporting Corporation (202) 628-4888
l.
l-f%
s_/
e 1 I raise this question because some of the core 2 melt sequences apparently give core melt probabilities in 3 the neighborhood of 10 -6 per year and it's my impression l 4 that the criteria for looking at sequences in the IPE would 5 include sequences with that sort of frequency.
6 And then I'm interested in whether or not the 1150 7 plants looked at this issue. Is there a risk contribution 8 from malfunction or inaccuracy or whatever of the water 9 level indicator that one can identify in the 1150 plants?
10 That's probably enough at this point.
11 I will at this point turn things over to Mr. Baer, 12 who I believe is going to open the session.
13 I'm just going to make some (q/
f HR. BAER:
14 introductory remarks, and sort of give an overall summary.
13 I'm not going to be using any slides. The major 16 portion of the presentation is going to be given by Andy 17 Szukiewicz, the Task Manager, and Brent Collins, of INEL, 18 who is a contractor who did most of this work.
19 I'm not sure we can answer--at least I can't 20 answer many of the questions that you raised. I'll try to 21 answer one or two as part of this overall summary.
22 Just to make sure everyone understands, the 23 Generic Issue had a postulated event of a failure of an PWR 24 water level instrument line. And as a result of this 25 postulated failure there's a number of potential O
Heritage Reporting Corporation (202) 628-4888
7-L 1 consequences. The instrument-line itself is a'small LOCA.
2 And the failed instrument line can provide a'falselhigh-3 . level signal that results in a reduction or termination of:
4 main feedwater flow and also for certain designs can defeat 5 the automatic signal that starts the RCIC or the HPCI, 6 depending on which particular level line is broken..
7 And then furthermore, if the failure is outside of 8 containment, there would be no increase in containment 9 pressure and for certain plant designs, there would not be 10 an automatic depressurization system signal.
11 So those were the potential concerns that were 12 looked at. It was not a single failure analysis. People.in 13 INEL, in order to identify those sequences that might have 14 some potential' consequences only went as far as postulating 15 instrument line in a single failure, and since all of those 16 ended up with relatively low core melt frequencies, they 17 didn't go beyond that to look at multiple failures. ,
1P I'll have to ask'them about the common mode 19 aspect. But the intent was a PRA-type analysis, not a :
20 deterministic single failure analysis.
21 The contractor made calculations of core melt l 22 frequencies, and the conclusions were that core melt i
23 frequencies were relatively low. They were less than about 2
-6 for all situations, and there were Leo groups of plants 24 10
-6 25 where they were barely in the range of 10 core melt O
Heritage Reporting Corporation (202) 628-4888 Y _ _ __ _ _____ _ _ ___ _
O 8
(.)
1 frequency.
2 And our conclusion was that regarding backfit 3 criteria, 10CFR 51.09, there's two basic criteria. One, 4 there is a substantial increase in overall protection of 5 public health and safety. We thought that 10 -6 at most 6 barely met that criteria.
7 The second criteria is that to take action you 8 need to have a favorable cost benefit ratio. When we looked l 9 at those groups of plants that had core melt frequencies in
-6 and looked at the cost of making a fix 10 the range of 10 11 even using them--probably optimistic costs estimates, that l 12 is, low cost estimates. Neither of those groups of plants
,f' T 13 made our guidelines of $1,000 a man rem. And therefore ;
(./ j 14 those are the bases on which we concluded that the issue'was 15 resolved and no action can be justified under our 16 regulations.
17 That's basically the criteria that we use. We try 18 to look at whether the risk is significant. And if it 19 isn't, or if it's marginal, to look at the cost benefit and 20 make a decision on those two ratios.
21 We try to use some judgment along with it but in 22 this case the number didn't seem to support any action.
23 DR. LEWIS: When using the $1000 a man rem, that 24 means you have a Level 3 PRA on each plant? Otherwise how 25 can you carry it down to $1000 a man rem?
i Heritage Reporting Corporation (202) 628-4888 j
9 1 MR. MINNER: We do have a Lerel 3 margin--
2 DR. LEWIS: But you do have to carry it all-the l 3 way.
4 MR. MINNER: Yes..
5 DR. LEWIS: Well, I'm just curious, because if you 6 are using that as a criteria for ignoring things--when you 7 say a partial, that means that it's still sort of--
8 MR. MINNER: We only looked at Level 3. We didn't 9 'look at all of them.
10 DR. KERR: Hal, what they did is sort of a generic 11 Level 3 for what they considered to be a typical plant.
12 They did do that. ;
r" 13 bR. LEWIS: I see. At least the report says they
(]/ .
14 did.
i 15 Mr, BAER: All right. With that much of an j 16 introduction and overall summar# 3et me turn it over to the 17 Task Manager, Andy Szukiewicz, who will go into somewhat 2
18 more detail of what we did. And as I said, we also have l
19 Brent Collins here from INEL.
20 DR. KERR: From what you said--I want to try to l 21 understand. This was not really a look at the desirable !
1 I
22 reliability of a water level system. I w.. a very 1
23 specialized look at a particular kind of failure of this )
24 system. Is that--
25 MR. MINNER: It's actually, if 1 recall correctly, O)
Heritage Reporting Corporation (202) 628-4888
/'\- 10 V
1 this particular issue was a subset of a broader issue that 2 was identified as something that was going to be looked at-3 MR. SZUKIEWICZ: It was a broader issue identified 4 in the early 1980s which was worked upon and that was 5 Generic Issue 50 which was called Level of Instrumentation 6 in BWR Plants.
7 My name is Andrew Szukiewicz. I'm the Task 8 Manager for this Issue.
9 As Bob was saying that Generic Issue 101 is really 10 a sub concern of an overall issue that was worked upon in 11 the early 1980s and was resolved I think between 1984 and 12 1985. And one of the areas that was opened at that time was fg 13 to go back and reinvestigate the' effects of instrument line V
14 breaks on these level sensors.
15 And so the issue was prioritized in the mid-16 eighties and identified as a higher priority issue to be 17 looked at and with a very specific charter of evaluating the 18 efforts of instrument line breaks and their effects on safe 19 shutdown.
20 And in that review we evaluated a break plus an !
l 21 additional single failure of both protection systems and 22 control systems.
23 Brent has a detailed description of the 24 methodology and the scope of the review, so I just wanted to l
25 brush on it now and we can get into the details. We O
Heritage Reporting Corporation (202) 628-4888
i tO 11 )
\J 1 basically look at all the plants, so it's more than just a 2 generic study because we looked at all plants, categorized 3 them into five different groups, depending on the 4 characteristics of the plants and then did very plant-5 specific reviews from one of each of the groups. But Brent 6 will get into it in a little bit more detail.
7 DR. KERR: In all cases, did you look at the 8 sensing line break as the initiating event?
9 MR. SZUKIEWICZ: Yes, we did. Yes.
10 DR. KERR: Never in conjunction with some other 11 event?
12 MR. SZUKIEWICZ: No, that's correct.
13 For my presentation--
14 DR. KERR: In a sense, if you take this single 15 failure as an addition, they looked at it in conjunction 16 .with whatever single failure was considered.
17 MR. SZUKIEWICZ: Right. That's right.
18 DR. KERR: Those were considered--
19 MR, SZUKIEWICZ: They were currently two 20 independent failures. And an additional independent single 21 failure.
22 DR. KERR: You didn't consider the cases wherein 23 some other initiating event caused the sensing line to 24 break--
25 MR. SZUKIEWICZ: That's correct.
O O
Heritage Reporting Corporation (202) 628-4888
_ - _ _ _ _ - _ - _ _ _ _ _ _ _ - - - _ - _ - _ _ = _ - - - _ _ _ - - _ _ _ _ _ - _ . - _ _ _ _ - _ - - _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ - _ . __
12 L
L 1. DR. KERR: You needed two things to consider.
2 MR. SZUKIEWICZ: That's correct'.
3 DR.-KERR: Along with a single failure?
4 MR. SZUKIEWICZ: That's correct. Although we did-5 go back and look at the--we did a literature search on the 6 history of past events as far as breaks go.- And although we 7 identified that there were some significant leaks, we've 8 never really identified a break. But we took the worst-case 9 condition where you would have a break of a line.
10 For my presentation I would just like to summarize 11 the technical findings and the proposed resolution and then 12 turn it over to Brent where we could talk about the scope 13 and the details of the review.
14 As Bob mentioned, based on our studies, we are not 15 proposing any new requirements or backfits. During our 16 review we did identify that there were some early vintage 17- plants, primarily five of them, that automatic protection 18 systems can be inhibited by an instrument line break. These 19- particular ones were the old plants, Oyster Creek, Nine 20 Mile, Big Rock, Dresden 1, and Millstone. !
i 21 For those particular plants, although automatic i 22 protection systems primarily negating the automatic ADS l 23 actuation could be inhibited. Operator actions would be 24 needed to achieve safe shutdown. j 25 Based on these breaks, we did some studies of the O
Heritage Reporting Corporation (202) 628-4888
___ _ _ _ ___ _ _ _ _ __ __ __ _ __ _ _ _ J
1 1
l l
1 l
J 13 f}
\s
}
1 time it would take for core uncovery. And basically we came )
1 2 up to the conclusion that for those particular plants we had 3 a substantial amount of time for the operator to take 4 relatively simple action to actuate depressurization to make 5 sure if you needed to actuate your production systems to ,
6 assure that the core would not be uncovered.
7 DR. LEWIS: Just so I understand--the fact that 8 the protection system can be inhibited by a break, does that '
9 mean that there are no' cases in which it is erroneously 10 activated by a break or that that doesn't lead to any 11 additional riek?
12 MR. SZUKIEWICZ: That activated other than gg 13 challenges to the protection system. We did not identify Q. 14 any that--
1 15 DR. LEWIS: There were no cases in which it could l 16 cause an activation?
17 MR. SZUKIEWICZ: That's correct.
18 All the other plants--and again, we looked at all 19 of the plants. We believe that they provide sufficient 20 redundancy and independence to achieve safe shutdown 21 automatically. And this is with the addition of an l 22 additional single failure.
23 Some of these plants have already initiated or had 24 designs from the start that provided these features because 25 of their redundancy and independence. And other systems O
Heritage Reporting Corporation (202) 628-4888
() 14 1 satisfied these requirements because of some THI action 2 plans that required certala modification to their systems.
3 And the bottom line is that-- ,
4 DR. LEWIS: I hate to interrupt you but I'm really '
5 trying to educate myself. I 6 I think Mr. Baer said that a break could cause a ?
i I
7 false overfill indication. And so in answer to the previous 8 question, that means that the false overfill indication will I
9 not initiate any--
10 MR. SZUKIEWICZ: You would need a break--
11 DR. LEWIS: No. Very specifically, would a false 12 overfill indication indicate any active action?
13 MR. SZUKIEWICZ: Yes, it could cause a scram.
14 DR. LEWIS: Okay.
15 MR. SZUKIEWICZ: It'll also trip feedwater and 16 then you are going to have to do something to get wate .
17 MR. BAER: The real potential problem is that the 18 operator is faced with the a high level signal on one 19 channel potentially and a normal level on the other.
20 DR. LEWIS: Not only the operator. That's what 21 I'm getting at.
22 MR. BAER: What I'm concerned that we look at was 23 a level of actually decreasing and on some plants the broken 24 line may be providing the protection signals to the RCIC or 25 HPCI and you would not get the automatic initiation of that O
v Heritage Reporting Corporation (202) 628-4888
I I
J O system.
l 1 l j
2 DR. LEWIS: But I'm just ac concerned about--
j 3 MR. BAER: We used conservative assumptions on how -j 4 long can the line be played. There are plenty of other .l l
5 initiations. I think it would be a minor-- l i
6 DR. LEWIS: But that's a different subject. i j
7 DR. KERR: Mr. Lewis, the report makes a statement j i
8 that'although there were cases in which protective action (
9 was initiated, the investigators assumed that initiation of
{
10 protective action before it is needed is conservative.
11 DR. LEWIS: Yeah.
12 MR. DAVIS: Now, there's another issue, A47, which 13 addressed the overfill phenomenon.
l
}
l 14 MR. SZUKIEWICZ: And in summary, on the technical 1
15 findings, we believe that the public risk associated with 16 instrument line breaks be sufficiently small so that plant 17 upgrades are not warranted.
18 For the generic resolution, we have proposed to 19 issue a Generic Letter for information to all BWRs. In that 20 eletter we will be issuing our analysis and the technical l 21 findings report which is embodied in NUREG/CR 5112 which Dr.
22 Kerr just referenced in one of his statements.
23 In the Generic Letter the licensees are expected 24 to review and verify that the plant designs, the plant-25 specific designs, are correctly represented by the staff Heritage Reporting Corporation (202) 628-4888 l
. ,c') 16 b
1 analysis. But we feel that based on our review and Mr.
2 Collins will talk about it a little bit, we have done an 3 extensive review on all the designs to make sure that it is 4 correctly represented. But just to make sure that they 5 didn't make any changes that we did not know about. We are 6 requesting that they do this, i
7 The licensee is also requested to maintain ;
i
/
8 appropriate procedures in operator training in order to 9 readily identify and mitigate water level instrumentation )
10 line leaks and breaks.
11 During our review, we looked at emergency
.. procedures and we evaluated the operator training programs g 13 and assured ourselves that indeed they do have a pretty good (3J 14 program and procedures to identify and mitigate these kinds 15 of situations.
16 MR. DAVIS: May I ask a question on that last l 17 point. It's my recollection that, and please correct me if 18 I'm wrong, that if the opere..or receives a signal of high 19 dry wall temperature, which would be the case for an 20 instrument liae break in the dry wall, he is instructed to 21 marmally ini?sinte ADS. And if he does that, then of course 22 he has no longer the option of core cooling with the RCIC or ;
23 the HPCI.
24 And it wasn't clear to me in the EGIG Report that 25 that action was accounted for in the success paths that were Heritage Reporting Corporation (202) 628-4888
'(T
% ,1 17 1 uaed to estiinate coro damage frequency.
2 Is it in fact true that the current BWR system- 1 1
3 oriented. procedures do call for the operator to go to ADS on l 4 high dry well temperature? Do you recall that?
i 5 MR. SZUK1EWICZ: I don't recall--
6 MR. BAER: I don't know, but I hope not, just on 7 that alone.
8 DR. LEWIS: Yeah.
l 9 MR. SZUKIEWICZ: There's a number of plants that 10 have the isolation condensers and there the procedures are 11 very specific to make sure --
12 HR. DAVIS: I'm not talking abaut those. Those-gg 13 are somu of the older ones like Oyste r Creek. I'm talking V
14 about the more recent designs. Now, this came up sorae t 21- 3 15 ago and I questioned that logic and it,*s my recollection 16 that that is, in fact, the current operating procedures for 17 high dry well temperature.
18 DR. KERR: Mr. Devis, do they manually initiate 19 dry well spray?
20 MR. DAVIS: No, Manuall'y initiate automatic 21 depressurization.
3 22 DR. KERRi ch, oh.
23 MR. DAVIS: Upon a high dry well temperature 24 tignal, independent of the reactor vesse.1 water level. I 25 coul.d he wrong. But if in fact these procedures still Heritage Reporting Corporation (202) 628-4888
)
'{
1
/3' 18 v
1 exist, it seems like some of the logic may be in question.
2 MR. M1NNER: Why would that be di f ferent than the 3 automatic--
4 HR. DAVIS: Well, the automatic won't go unless 5 you get both high dry well temperature and low reactor 6 vessel water level. Here we are talking about only having 7 one of those signals. Because the reactor vessel water I
8 level will go false high. And so you wouldn't get the 9 entonatAc--
10 It might be worth looking at that. As I say, I's 11 not absolutely sure, but this is what was my recollection 12 when we talked about this some time ago.
p+ 13 And the other ing tl. . ' "*s curious about, is t
14 it possible that the high dry wall temperature will initiate 15 a containment iaolation signal which wil.1 disable RCTC and 16 HPCI because it closes the isolation valves in the 17 steamlines that penetrate the containment.
18 MR. BAER: Not supposed to. Not unless it's a 19 misdesign. Those lines don't isolate. Cercainly not the 20 HPCI because that's the accident indication. It's not 21 suppose to isolate in a high dry well temperature or 22 pressure.
23 You mentioned temperature. Pressure also or 24 <either or?
25 MR. DAVIS: As I recall, it's temperature. This l
l Heritage Reporting Corporation (202) 628-4888
! O '*
I came up because if you lose dry well coolers, the 2 temperature goes up and the operator initiates ADS unde? the 3 symptom oriented proceduees. That's my recollection. I 4 don't want to belabor it.
5 MR. BAER: You've peaked my curiosity. I'll find 6 out but I don't kr.ow.
7- MR. DAVIS: Well, I alse think this isolation 8 program should be looked at. Because in reading the llatch 9 FSAR,.for example, it looks like there are cases where these 10 systems would be isolated due to a high containment 11 temperature signal.
12 MR. SZUKIEWICZ: I'm not sure about your question r~g 13 If you initiate the--if you start depressurizing, then
(~/
14 there's a number of backups that can be used to make sure 15 that you don't have core uncovery. The problem was was that 16 you could have failures that would not depressurize you and 17 to a low enough level to make certain systems--
18 MR. DAVIS: No. No, I'm not suggesting that if he 19 does go to ADS thah he gets into trouble. He still has low 20 pressure coolant injection ar.d low pressure core spray in 21 some designa. But what I am suggesting is that he no longer j 22 has the high pressure injection option. And in the EGIG 23 Report it looked like that opticn was retained in the logic 24 structure. J 25 And maybe we'll hear more about it when we get O
Heritage Reporting Corporation (202) 628-4888 J
rm 20 b
1 into it.
2 MR. SZUKIEWICZ: Okay. But on some plants he may 3 not have the high pressure injection option, but he might 4 have it if they have RCIC, which would be the other option.
5 MR. DAVIS: Right.
6 MR. SZUKIEWICZ: Which would be off of different 7 level sensors.
8 MR. DAVIS: Yes.
9 MR. SZURIEWICZ: And we could go into that in more 10 detail.
f 11 MR. DAVIS: Yen, but he won't have that if he goes 12 to ADS, because he's lost his steam. He won't have either es 13 RCIC or HPCl u
! ") 14 MR. SZUKIEWICZ: Okay.
15 DR. LEWIS: You say you brought this up some time 16 ago. Do you remember who had the action of following it up?
i 17 MR. DAVIS: No, I den't.
18 DR. LEWIb: It might be worth checking. ,
i 19 MR. DAVIS: My memory is not as good as it used to 20 be. I mean I'd have to check my notes and see. l l
21 DR. LEWIS: Just curious.
22 MR. SZUKIEWICZ: We can look into that. 1 23 DR. LEWIS: But do you also think that initiating 24 ADS manually may automatice.lly isolate containment? I 25 couldn't understand for sure the question you were raising Heritage Reporting Corporation (202) 628-4888
)
i 1
l fh. 31 l
%-)' '
1 in thst--
2 MR. DAVIS; No, it was a se parate issue, the 1
3 containment isolation. Which I am suggesting also occurs on i i
4 high dry well temperature. )
j
^
5 DR. LEWIS: Automatically or manually?
6 MR. DAVIS: Automatically. The objective is to j i
7 avoid any release and the high dry well temperature suggests !
8 a break that will diccharge material to the suppression 9 pool.
10' Now, the isolation may not affect the two high l 11 pressure systems. That's the question that I'm not clear i
12 about. l 13 MR. SZUKIEWIC7: I don't know specifically what 14 those procedures are but we can take a look at it.
15 Okay. So the' resolution is fairly 16 straightforward, and what I'd like to do now is discuss the 17 scope and the methodology on how we conducted our review.
18 And I'd like to present Mr. Brent Collina from INEL who did 19 the review.
20 MR. COLLINS: My name is Brent Collins and I'm an 21 engineer with the INEL. And I was involved in quite a bit 22 of the work we did on this Generic Issue.
23 l'd like to kind of go over a brief summary of how 24 we analyzed this problek from the initiation of the question 25 to the resolutions. l Heritage Reporting Corporation
)
(202) 628-4888
- r. ..
j 22
\_) i 1 The first thing we did was we went through a date 1
2 gathering operation in which we got our hands on as much l l
3 information on all of the BWR plants that im could. We went !
I 4 to_the PM offices. We went to FSARs. We went to syutem ;
5 descriptions for BWRs.
6 DR. KERR: Excuse me. What is a PM office?
7 MR. COLLINS: The program Manager, NRC. project -
8 Manager.
4 S DR.'KERR: Thank you.
16 MR. COLLINS: Supposedly we thought that was our 11 best route for getting the most up-to-date information in 12 the plants. And at this point it was a plant-specific 13 endeavor. We locked at each pinnt specific and tried to get u 14 as much data on each plant as we could. That data came in 15 the fore. of pNIDa, system descriptions, logic diagrams, l 16 anything about the plants we could actually just throw in a 17 box and save.
l 18 Once we had alJ that data then we compiled it for 19 a comparison. We tried to look at each plant specific and :
10 say, what systems do they have? What things are available?
l l
21 What is the configuration o"I the plant? Se that we could 22 start to take different plants that looked alika and start l
23 lumping them together.
24 Based on that grouping, we came up with five L 26 different gecups + hat we categorized these plants into. And i r
1 m b
%]
Heritage Reporting Corporation i (202) 62&a4888 i
l g i thene were baned on different' systems, different 2 characteristics of the plant.
3 Once we had these five groups, we went back and 4 pulled out a single plant from each group. Typically a 5 plant we had the most informatier. on, the plant that we felt
!- 6 most con.fortable with. The information that we had. And we 7 did a. cense line break single failure analysis for that b specific plant. So we actually did this analysis for five 9 specific plants, one from each group.
10 As a result of this analysis--
11 DR. KERE: There was a single active component 32 failurey right?
l r~g 13 MR. COLLINS: Yes. We limited it to single active i V
14 failures. Wc didn't want to take credit for, like, two pipe i
.i 15 breaks. I 16 DR. KERR; Thank you.
17 Just out of curiosity, why did you use that i 18 approach rather than simply using that line break as an 19 accident initiator which yon then follow through as one 20 would in the case of some other accident initiator?
21 MR. COLLINS: pardon? I didn't hear the--
22 DR. KERR: I naid why did you use that approach 23 rather than simply letting the line break be an accident 24 initiater as part of a Level 1 pRA which you then followed j 25 through to see what consequences occurred?
)
Heritage Rep.orting Corporation (202) 628-4888
.N __ = ____._____ __ _ ___
i i
l i
- 24 1 MR.-COLLINS: I guess there's no good argument why 1 1
2 we didn't take that approach. )
3 DR. KERR: Okay. .I 4 MR. COLh1NS: What,ue did do then was from the 1
, 5 line break in single failure analysis, we identified somo 6 sequences that we felt were potentially significant and f' rom !
7 those sequences then we calculated a probability of that 3 8 particular sequence failing. That probability included 9 considerations for no operator action. L'e did not give the 10 operator any credit for any action, be it good or bad, for 11 two hours. We assumed he just sat back and did not know 12 anything was occurring to him, which we felt was 13 conservative.
14 MR. MICHELSON: Excuse me. When the sensing line 15 broke, what assumptions did you make concerning potential 16 consequential damage to surrounding devices to the break?
17 MR. COLLINS: Nothing.
18 MR. MICHELSON: You assumed that nothing in the 19 way of that one-inch line break would be affected to it 20 adversely?
21 MR. COLLINS: No.
22 MR. HICHELSON: Is that a reasonable assumption?
23 MR. SZUKIEWICZ: You kind of look at steam 24 impingement or changes in the environment.
25 MR. MICHELSON: Is that a reasonable assumption?
l Heritage Reporting Corporation (202) 628-4888
I
/~ 25
\_T/
1 That 's a f airly confined territory. A lot of things in that 2 area. And ncne of them, you are saying, would be affected 3 by the line break? I guess no environmental effects either.
4 MR. SZUKIEWICZ: As far as the environmental 5 effects in those areas, you would have worst-case conditions S under, you know, LOCA conditions, which the equipment has to 7 address. Those kind of environments. That we didn't think 8 that the environmental effects of an instrument break would .
9 be worse than a pipe break.
10 But there have been studies in the past looking at 11 steam impingement and, in fact, there was a bulletin that 12 requested all plants to look at pipe breaks. Now, that may s 13 have not included specific instrument pipe breaks, but--
rA 14 DR. KERR: It didn't include this, no.
15 MR. SZUKIEWICZ: They did look at some sort of 16 things. And the effects, steam impingement, effects on 17 them.
18 MR. DAVIS: These are one-inch lines, is that 19 right?
20 MR. SZUKIEWICZ: Typically, yes.
21 DR. LEWIS: Why did they have them that big?
22 MR. MICHELSON: So people can stand on them.
23 Without breaking them. If you made them smaller, people 24 would stand on them and they would break.
25 DR. LEWIS: Well, I could stand on a half-inch Heritage Reporting Corporation (202) 628-4888
(
l r^s 2e 1_f ;
1 pipe right now. )
I 2 MR. MICHELSON: Well, it depends on how yo9 do it.
t 3 How big a mechanic you are. I 1
4 DR. LEWIS: But it sort of poses the question--
5 MB. MICHELSON: I think it was physical integrity 6 mainly. I 7 DR. LEWIS: You know, the smaller pipe poses less 8 of a--
{
9 MR. MICHELSON: Sura. j 1
10 DR. LEWIS: Right?
11 MR. MICHELSON: Yeah. Well, you can orifice the 12 line too which in some cases they do if they worry about ;
13 certain kinds of effects.
~; y 14 MR. COLLINS: Once we had those probability 60 of 1
f 15 the failures calculated, we then went on to determine 16 release categories and calculated the public risk from the 17 individual potential scenarios that we'd identified.
18 We then looked at typical fixes that looked like 19 the best fixes for the particular scenarios that we 20 identified. We determined approximately what it would cost 21 to install that fix int 6 the plant. And based on those 22 calculations then we came up with a cost benefit of plant 23 risk related to a dollar factor.
24 That was the overall general epproach that we 25 took. A little .more specifically what we did--on our plant Heritage Reporting F rooration (202) 628-48bu l
(
l
t 27
)'
1 classification, we ended up basically breaking the groups or 2 all the plants up into five groups, 3 The primary things that we used at the time we did 1
4 it to make those differentiations were the reactor level 5 sensing line configuration. Over the geneiation of the 6 BWRs, they evolved from having different configurations 7 where they have, in some cases, a total of eight 8 penetrations--four for the measurements, two for the f
9 reference legs, and they had also conf.igurations that had 10 four measurements and four individual reference leg returns.
11 So becausa of the way the instruments are hung on 12 those various lines, thet was one thing we used for a break 13 in t he groups.
f
(:) 14 The other thing we used was the BWR model type.
l 15 De it a Model 2, a Model 3, Model 4. Those types of i l
16 categorizations actually brought into account these other 17 two we've listedy containment type and the system 18 configuration. That is, these things like the syetem I 19 configuration where we had some plants with an HCIC, some 20 with just the isolLtion condeneer, that had us looking at 21 the BWR Lype as a way to possibly break up the different 22 groups.
1 23 Sci ne of the older plants--I said earlier that we 24 addressed all of the plants. We did throw out some of the
{ 25 older plants. Lacrosse, ilumbolt Day, Dresden 1, plants that O
11e rit,sge Seporting Corporation ;
(202) 628-4888 l
3 28 (V
1 were no longer active we didn't look at.
2 We did call Group number 1 the BWR 2s. 'I he 3 primary thing here is they do have fouf vessel penetrations -
1 4 and when I talk about vessel penetrations now I'm.enly i
5 talking about the reference legs and I'll explain why in a )
1 6 couple of minutes. l l
7 These plants do have an isolation condenser and 3 8 they are only the two older plants. Well, three. older 9 plants within this group. We have Oyster Creek, Nine-Mile 10 point and I believe Big Rock also fell into that group.
11 primarily it's the vessel penetrations that turns ;
12 out to be the difference in the plants. We look at a
. 13 typical arrangeme... for these plants. We see that we do 14 have the four measurement legs coming through some 15 instrumentation for .tndividual condensate pots and the four 16 individual reference legs going back into the vessel.
17 The problem with the Group 1 plants is although 18 they have this four-legged arrangement, they take an l 19 individual leg and they put multiple channels on that leg.
20 So a break in a single leg potentially takes out two of your 21 channels, which makes your ringle failure to something in 22 the other side susceptible to some sort of consequence.
! 23 Likewise, the reason they have the other two legs 24 then is they do put some specialized instrumentation on 25 there two legs, but again it's doubled up, o
l %
l Heritage Reporting Corporation (202) 628-4888 ~
. - - _ _ . _ _ _ . b
l i
1
- r%
29
{}
1 And the feedwater for this particular plant
(
2 grouping, the feedwt i^* control comes'off of this other leg 3 also.
4 MR3 DAVIS: Excuse me. Are'those lines separated?
5 MD. COLLINS: Yes.
6 HR. DAVIS: So that a break in one wouldn't cause 7 a break in another one from pipe whip or impingement?
8 MR. COLI, INS: We didn't look at that type of a 9 ' relationship. 1 10 MR. MICHELSON: Where is the containment boundary 11 en your drawing? In other words, is it all inside a ;
i 12 containment? {
13 MR. COLLINS: Not always, no.
14 MR. MICHELSON: I taought not. And I think it's 15 important to understand which part is inside and which part 16 is outside and how it's isolated and what effects you can 17 get by that separation.
18 MR. COLLINS: What we found out as far as another 19 important factor of where the break was was whether it was L 20 inside or outside the dry well. Being inside the dry well, 21 you get automatic actuations that you don't normally get 22 outside the dry well.
23 As far as outside the dr well, typically there's 24 a very short section of pipe before there's a check valve to 25 help prevent any leakage from a break outside of that check 1
CE)
Heritage Reporting Corporation l (202) 628-4888
k
(:)
1 valve. And we've pretty much limited ourselves to saying 2 that that check valve was not going to fail. Therefore, the 3 only significant problems we would get into is if the break 4 was between the dry well and the check valve outside of the 5 dry well.
6 MR. MICHELSON: How often are those' check valve ;
7 ever verified after they are installed?
8 MR. COLLINS: As far as that sort of testing and 9 stuff, I'm not familiar.
10 MR. SZUKIEWICZ: They are actually flow valves.
11 MR. MICHELSON: They are excess flow checks I 12 would expect.
13 MR. COLLINS: Right.
e
%- 14 MR. BAER: The point I'd like to make'is if you 15 examine the frequency of pipe break used in the analysis, 16 it's much more conservative than actual operating 17 experience. So that I think when you've looked at the risk I
i 18 perspective and the bias is that rask is over estimated, not 19 under estimated.
I 20 DR. LEWIS: That's never an excuse to do a risk--
l f 21 DR. KERR: Mr. Lewis, the microphone please.
22 please use it. Those important words may never get 23 recorded.
24 DR. LEWIS: Oh, well. It may be a reason for 25 analysis, but once having decided to do the analysis, it's O Corporation Heritage Reporting (202) 628-4888
_ ___---_ __ _ - - i
J. < l
]
II o -(~T 31 i
\_)
l 1 not an excuse to do it badly.
2 MR. DAER: Just a point. It is only a short pipe' )
3 and then there's an excess flow check valve. And so we i 1
4 don't even worry about that little piece of pipe. And the-5 point I'm making is that the first frequency that was used ,
6 is large enough to encompass the situation of whether the 7 excess flow valve works or not.
8 D R -. LEWIS: I may have misunderstood you.
9 MR. BAER: I'm saying I think you use a number 10 high enough to encompass the inadequacy of the valves.
11 DR. LEWIS: So we should really use a different 12 break.
13 MR. DAVIS: I'm a little bit confused. Your
()]
14 report for the Group l's assumed a 100-foot length of pipe 15 outside the dry well.
16 MR. COLLINS: Inside the dry well.. We used twelve 17 foot outside.
18 MR. DAVIS: Oh, I'm sorry. You are right.
I 19 Twelve foot outside, okay. So you are assuming it is twelve 20 feet to that restriction or check valve--
l 21 MR. COLLINS: We took one standard section of pipe 22 and used that standard section of pipe. I think if you look 23 at the probability numbers associated with the failures, 24 you'll find that the wales on the ends are probably major 25 contributors.
1 Heritage Reporting Corporation (202) 628-4888 L______-----_______--
i
('}
v 32 1 MR. DAVIS: Okay.
2 MR. COLLINS: The numbers are quite high for 3 probability failures.
4 DR. KERR: Now, the arrows point to places.at 5 which a break was postulated?
6 MR. COLLINS: Yes. That was my next item to talk 7 about.
8 DR. KERR: Okay.
9 MR. COLLINS: We postulated that the breaks were 10 in these areas being on this side of the instruments. We 11 looked at them for.down here on the measurement legs, but if 12 we break them down here, the instruments basically go down 13 scale with a false low reading. Ideally, it goes to zero.
14 We felt if it goes to zero and other things happen, that's 15 at least a safe direction for the instruments to fail. We 16 didn't .look at any problems associated with inadvertent 17 actuations or false actuations. We felt ae Jong as they 18 came on, that that was better than not coming on at all. So 19 we felt that that was a conservative safe direction to fail.
20 All of our analysis then looked at the reference 21 legs. Where if we break these reference legs, the 22 indication goes high upscale and there's not as many 23 automatic actuations off of a high liquid level sensing than 24 there is off of a low liquid level sensing in the vessel.
25 Basically it would be reading atmospheric with a vessel n
U Heritage Reporting Corporation (202) 628-4888
33 1 pressure for a reference.
2- DR. . LEWIS: That's right. /.nd therefore it would 3 read negative, but it will pin its-1f.
~4- MR. COLLINS: Yes.
5 DR. KERR: And you postulated one break at a time 6 I assume?
7 MR. COLLINS: Only one break at a time, yes.
8 We.actually went through and we looked at a break 9 right here and then we said, what does that effect? And 10 then we broke this one and we said, how does it change 11 individually, independent of what this break was? And we 12 went through each of the individual legs and said, what is 13 affected and what systems are still available.?
i O 14 It turns out actually up here I believe the most l important problem that we get into is not necessarily the 15 16 water leaking because of the break in the sense li'ne. It is 17 if we've got feedwater control, assuming worst-case single 18 element control off of that one transmitter. That one 19 transmitter fails in a high direction. Say it pegs high.
20 The feedwater control system automatically ramps down. And 21 you start losing your vessel inventory because your 22 feedwater has just basically shut itself down. And that's 23 probably the most severe consequence from the break versus 24 the break itself.
25 DR. KERR: Is that just a one out of two control?
O Heritage Reporting Corporation l (202) 628-4888
/~T 34
(_/ ,
1 MR. COLLINS: What?
2 DR. KERR: Is it just a one out of two control? ]
3 MR. COLLINS: We assumed worst case, yes. That it 4 was just--
5 DR. KERR: Well, that's a design, not a worst 6 case. Is the design for two-channel control? Or single-7 channel control?
8 MR. COLLINS: It's one to the other.
9 DR. KERR: Is it auctioneer them you mean?
10 MR. SZUKIEWICZ: I don't believe. I think there's 11 just a selection--
12 DR. KERR: I see. You can select A or B to
('T 13 control them but having selected it, you've got one channel?
J 14 MR. COLLINS: Yes.
15 MR. SZUKIEWICZ: B or D or one of those.
16 DR. KERR: I understand.
17 MR. COLLINS: That's the worst case. There may be 18 other--
19 DR. KERR: It's a non-safety system.
20 MR. COLLINS: Non-safety.
21 DR. KERR: Okay.
22 MR. COLLINS: Once we identified these breaks, 23 then we used what we called a success matrix. If you look 24 at this kind of like a flow chart, if we have the initiating 25 event being down the left-hand side and a mitigation of the Heritage Reporting Corporation (202) 628-4888 a_________ __ _ _ _ . . _ _ _
1 35 1 ' event down the right-hand side, we need any single path
-2~ throughLthis matrix we felt would give us some' successful 3 . mitigation of the event.
4 When you'look at a chart like this, the first 5 thing.you start' picking out are places where there are 6 single non-redundant paths. Those are the ones that would 7 start to be suspect. We found no cases where we-could 8 interrupt a scram to the system. Although the break in 9 single failure may take out the level actuation to the RPS..
10 We'always found some other-way to get a turbine trip causing 11 the scram and MSIV closure causing the scram. Whatever the 12 scenario we came up with, we always got a reactor shutdown 13 from the protection system.
14 ' DR . KERR: The RPS on level isn't one out of two 15 twice?
16 MR. COLLINS: Yes. Not always on a high level.
17 DR. KERR: Well, that's what I'm talking about, a 18 high level. Just one out of two?
19 MR. COLLINS: Only the newer plants have a direct 20 scram off the high level. The older plants do not.
21 DR. KERR: Do not. Okay.
22 MR. COLLINS: Within the decay heat removal we l23 found that the ADS was our only suspect. We never did find 24 any way that we could successfully take out all three paths 25 of the isolation condenser, the steam system or the release O Heritage Reporting Corporation (202) 628-4888 I
I I.
L_ ____-___ -___________ _
CE) 1 safety valves.
2 With the ADS the biggest problem was the break 3 outside the dry well ,and not getting the automatic 4 actuation. Sometimes when we found systems didn't operate 5 in our study, it wasn't necessarily because the system 6 itself was broken., it was still'available for operator 7 action. But the automatic actuation did not take place.
8 We did find cases where we did get the ADS not to 9 automatically actuate, therefore with no ADS we never get 10 any low pressure core spray. Therefore this path of 11 mitigation was taken away from us at some point.
12 Likewise, these older plants only have feedwater 13 injection and therefore that's prone to single failures plus 14 it's within the one break if we break the instrument that 15 it's controlling off of. This guy ramps down to zero. And 16 he's in effective for us.
17 So we did find our scenarios that would take into 18 account those two paths on our flow.
19 Our potential fixes would then address aspects of 20 ADS and some other sort of injection down here such as the 21 high pressure injection system.
22 DR. KERR: Let's see. If you postulate a 23 feedwater injection as your sing failure and then locked 24 at a common mode failure of core sprays, you might get a 25 somewhat different result than apparently you got.
Heritage Reporting Corporation (202) 628-4888 1
1 Am I correct that you did not take into 2 consideration any--
3 MR. COLLINS: We did not look at any common model 4 failures, no.
5 DR. KFRR: Okay. And I think Mr. Baer said he was 6 going to ask you why you didn't. Could I ask you why you 7 didn't?
8 MR. COLLINS: Why we did not look at any common 9 mode failures?
10 DR. KERR: Yes, sir.
11 MR. COLLINS: Because basically the way the 12 problem was described to us was that it was the break of a 13 sense line and a single active failure. And therefore, we 14 only looked at potentially si:.gle active failures that would 15 take out safety--
16 DR. KERR: I guess I don't understand that 17 response because once you postulated the single active 18 failure, you then did a reliability analysis for the rest of 10 the system.
20 Now, most of the experience up to now with the 21 reliability analyses, if people are doing common mode 22 failure analyses correctly, is that a significant 23 contributor is always a common mode failure. j 24 So I'm puzzled by your comment.
26 MR. BAER: Can I clarify a point'? I'm not sure if O
Heritage Reporting Corporation I (202) 628-4838 l
C_______________________ __
38
]
1 Brent--you didn't look at common mode failures, but I think 2 for this group of reactors for the postulated events and the '
3 ningle failur9, you had no success paths, other than manual 4 operation of the ADS sometime after two hours, right?
5 MR. COLLINS: That's right.
6 MR. BAER: So whether you postulated common mode 1 7 failures, you can't be worse off. You've got no feedwater l l
8 from the initiating event. And you have no ADS because the 9 break is outside of the containment--
10 DR. KERR: You did a reliability analysis for the 11 rest of the system. I haven't seen the details of that and 12 I might not understand it if I did. But presumably the
{
s 13 reliability analysis was one which ignored single common 14 mode failures. Now, either that makes sense because there 1
15 are none possible. I don't know what the other alternative 16 is. But surely--
17 MR. BAER: I don't believe we did a reliability 18 anal).is--
19 DR. KERR: Well, if you look at page 20, for 20 example, and page 21 in the report, there is something which 21 I would call a reliability analysis which gives the 22 likelihood of instrument line break and no core reflood, for 23 example, f l-l 24 MR. BAER: Okay. Ir that what you mean by--
25 DR. KERR: I mean I wouldn't know what else to
, (:)
Heritage Reporting Corporation (202) 628-4888
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - l
39 (1) 1 call that. You could call it a fall tree or something, but 2 it's in effect a reliability analysis.
3 MR. COLLINS: No, we didn't include any common 4 failures into their--
5 MR. DAVIS: Well, you must have included' common 6 mode failures in, fo.c example, failure of the feed pump 7 control circuit. You didn't assume that the five control 8 devices failed independently.
9 MR. COLLINS: No. You are right.
10 MR. DAVIS: You did, in fact, consider common 11 mode feilures within a single system.
12 MR. COLLINS: Yes.
13 Within our Group 2 plants these were the older--
14 MR. MINNER: I'm confused. I don't see what you 15 would think the common mode was--looking at Figure 11.
16 DR. KERR: I don't claim that I know what you 17 would get a common mode failure of either. I simply 18 observed the statement in the report that common mode 19 failures were not treated.
20 Now, if you--
21 MR. MINNER: What I am trying to understand is I 22 don't think--
23 MR. COLLINS: The fall tree one? I don't have <
24 that one.
25 DR. KERR: If there is no possibility of common O
Heritage Reporting Corporation (202) 628-4888
() 40 1 mode failure in any of the systems treated, then I agree.
2 They shouldn't be treated. That was not the statement--
3 MR. MINNER: I mean the answer is correct, but I 4 think the answer is--I guess I'm asking, is common mode 5 failure applicable to this particular analysis? I don't 6 think it is.
I 7 DR. KERR: I don't understand how you could have--
8 MR. MINNER: Well, look at Figure 11 and I think 9 you'll understand it.
10 DR. KERR: Figure 11 is a fall tree. It does not 11 lay out the equipment configuration. And I don't know what 12 the equipment configuration is--
13 MR. MINNER: Well, the configuration has nothing
)
14 to do with it.
15 DR. KERR: Well, equipment configuration has a f
16 great deal to do with common mode failures.
17 MR. MINNER: It can. But in this case it does 18 not.
19 DR. KERR: Well, if you--this is the first 20 question I asked Mr. Collins. I said, did you conclude that 21 common mode failures were impossible? And he said, no, we 22 just didn't consider them. Now, I may be misinterpreting 23 him or maybe I misunderstood him.
24 MR. MINNE3: I think he's given you the wrong 25 answer.
( l Heritage Reporting Corporation (202) 628-4888 1
41 1 DR. LEWIS: You say they are impossible.
2 MR. MINNER: In this particular case I think he's 3 wrong.
4 DR. KERR: They don't come in with a calculation.
5 That's the truth.
6 DR. LEWIS: That's what he's saying. But the 7 calculation is the equipment.
8 HR. MINNER: Can I try one more with the 9 operator's--I don't really know the answer and I'm not 10 asserting that I do. I'm quite sure that he has not 11 considered common mode failure.
12 I guess you are going to assume that they want 13 some equipment in there to help that--
O~~ 14 MR. LIPINSKI: Let me make a statement.
15 If there is a common mode failure that knocks out 16 whatever the operator can manipulate, it doesn't matter how 17 much time he has. He just won't be able to take manual 18 action. If there's the right kind of common mode failure 19 that knocks out whatever he can manipulate. And your tree 20 doesn't show a thing. You've got to kn9w what the equipment 21 is. .
22 The same thing with your feedwater. You took a 23 number here for twenty-four hours, but, again, given what j f'
24 the common mode failure is--
25 MR. SZUKIEWICZ: We assumed that it wouldn't bc I i
($)
Heritage Reporting Corporation j (202) 628-4888 l l
1
() 42 1 1 available at all and we looked at other plants--
2 If we didn't look at common mode failures in the 3 sense that we postulated a single failure that would knock 4 out the high pressure--
5 HR. LIPINSKI: Well, the general philosophy of 6 common mode failures, you have to look at the equipment 7 relationship and say I can discount them based upon a 8 particular failure that I postulate.
9 DR. KERR: Mr. Saukiewicz, I could use the 10 argument you are using to demonstrate that there are never 11 common mode failures for diesel generators. Because diesel 12 generators in the general design criteria are required to be
/~ 13 separate and independent. Therefore I can't have common
- Qj) 14 mode failures.
[
l 15 But if you'll look at PRA's that deal with diesel 16 generators, you'll find that common mode failure is a 17 principal contributor to diesel unreliability.
l l 18 So I don't--I mean it may well be that common mode l
i j 19 failures do not contribute anything to this problem. I l
20 don't know. But I don't believe anybody else knows either 21 at this point. I should put it another way. I don't have 22 any evidence that anybody else knows at this point.
, 23 MR. SZUKIEWICZ: I just wanted to clear up that we z 24 did not look at common mode failures.
l 25 DR. KERR: Okay. Well, now that we've settled l
Heritage Reporting Corporation I (202) 628-4888 1
l
O x).
43 I that issue, let's continue.
2 MR. COLLINS: Within our Group 2 plants, again a 3 fairly small group. We only found three plants that we put 4 into thia grouping. They were the older BWR 3's.
5 A significant point here is that they had two 6 vessel penetrations, those two being the reference legs.
7 And it still had the eld isolation condensers.
8 Typically configuration for the Group 2 plants t 9 had, for the reference legs, Jtst two penetrations coming 10 out and T'ing off into four individual condensate pots 12 through the channel instruments into the measure legs.
12 This arrangement is very similar to the Group 1 r" 13 arrangement although the Group 1 had four_ penetration for i v' these reference legs. Again, since they had the multiple l
14 15 channels hung on to an individual leg, you have the 16 potential of being able to break this leg someplace up here 17 and taking out two channels, leaving single failure to 18 effect something from the other side.
19 The success matri:: that we found for Group 2 20 starts showing some redundancy within inventory control. We l
l 21 no longer just have the feedwater system, but indeed we do 22 have some high pressure injection systems for this grouping 23 of plants.
24 Again, with the reactor shutdown s we didn't find j 25 any RPS means of stopping the RPS from scramming the reactor O
Heritage Reporting Corporation j (202) 628-4888
') 44
~J 1 one way or the other.
2 ADS again, single element within our flow path, 3 break outside the dry well, no automatic actuation. 1 4 Therefore, if we can take him cut we . inhibit any core spray I
5 or low pressure systems.
6 Even though we did start to get some redundancy I 7 here we did find scenarios because of if we took the break 8 off the partAcular leg, again, assuming worst-case single 9 element controller for the feedwater so that the break gave
]
10 the feedwater erreneous signals, the HPCI system has several 11 eingle failures in it. Therefore, we were able to find some 12 break icenarios that would take out this path through the 13 ADS or our lower path, our inventory control through the
l l 15 Group 3 was almost identical to the Group 2 16 plants, except for the RCIC. Within our Group 3 we started 17 getting the RCIC., So now we started getting a system s
18 capable of replacing inventory versus the isolation 19 condenser approach of removing decay heat.
20 This group was the newer BWR 3's, the older ,
21 vintage 4's, but again they still only had two vessel ,
l 22 penetrations. So they looked very much like the Group 2 23 plants and are susceptible to some of the same scenarios of q 24 the Group 2 plants.
I 25 This is our largest single group out of our five p
(_) i Heritage Reporting Corporation (202) 628-4888 l
(( l 1
45
/^)
- s. l 1 groups of plants. This greup has more plants than any other i
2 grouping. l l
3 A Jook at a typical arrangement for the Group 3 j
)
4 and it's identical to Group 2 with the exception that we 5 chose to Just shcw this picture with the single condensate )
I 6 pots. Some plants had this arrangement. Some had an 7 arrangement just like Group 2 where the condensate pots were 8 out here. But the important thing still being a common tie 9 here where the break in this line takes out both a Channel A 10 and a Channel C, for example.
11 DR. KERR: When you say it takes you out, you mean 12 it really causes it to read high?
13 MR. COLLINS: Right.
O 14 DR. KERR: Thank you.
15 MR. COLLINS: Some sort of erroneous signal.
s 16 DR. KERR: Uh-huh.
17 MR. COLLINS: Our success matrix--let me, while 18 I've got a typical one here, give you a lead in into the 19 cuccess matrix for this group.
20 It turns out that when we look at the success 21 matrix, Group 3 now has not only a high pressure coolant 22 injection system, but it also has the RCIC system. But the 23 thing that they did in their design was typically the RCIC 24 instruments are sll in these two channels, for example. And 25 the HpCI instrumentation is all in these two channels. And
' Heritage Reporting Corporation I (202) 628-4888 )
)
l
-_-----------_- ---_ _ i
46
-(V~Y 1 therefore, if I break this leg, I take the RCIC instruments 2 and peg them all high, baeically t9ke RCIC out of the 3 picture. And that makes it susceptible to the KPCI single 4 failure on the other side.
5 When I look at my Group 3 success matrix, I still 6 have the RPS as a single path but no way to completely take 7 it out of the picture.
8 ADS, although there are two trains that still 9 basically actuates on the same principles. So a break 10 outside the dry well will take out both of these paths.
11 Assuming worst-case feedwater control, we make the 12 feedwater control system erroneously respond.
13 We can take out, depending on which leg the break 14 is in, we can take out the RCIC instrumentation which leaves 15 HPCI susceptible to a single failure. If we go to the other 1
16 side, the B side of our typical arrangement, the break, 17 assuming again that the feedwater control is switched over 18 to that leg worst case.
19 Where feedwater control comes out, the HPCI 20 instrumentation is now affected by the break so the RCIC may 21 be susceptible to the single failure.
22 MR. MICHELSON: Aren't those one out of two twice 23 on HPCI and RCIC?
24 MR. COLLINS: But they are all on one side.
25 MR. MICHELSON: Oh, I thought it was one on one Corporation r' '
Heritage Reporting )
(202) 628-4888
- _ _ _ _ _ - _ _ _ _ _ 1
47 I) s.
1 side, two on the other.
2 MR. COLLINS: For this group of plants--
3 MR. MICHELSON: It is?
4 MR. COLLINS: There's enough instrumentation on 5 the one side or the other that it can affect them. If you 6 are on the channel that has--or on the side that has two 7 instruments, you've got two of them pegged high at minimum.
8 MR. MICHELSON: Yeah, but I thought it looked j 9 twice. One on two--twice means one on each side will fix, 10 at least something on each side.
11 MR. COLLINS: That seems to be the problem is with j 12 the HPCI and RCIC on the typical arrangement for the Group
/~'N 13 3, they don't put--for example, there is not a HPCI and a i V
14 RCIC Channel A coming off of this instrument. And there's !
15 not a 11PCI and a RCIC Channel C coming off of this 16 instrument. They may have a HPCI/RCIC combination to where f
17 there are enough instruments within these two channels that 18 it doesn't matter what I have on the other side. I'd really 19 need a plant specific like a P&ID type deal to actually show l
20 you how they're hung off at the legs, but it is an unusual 21 arrangement. It's not as clean as saying that there's an 22 individual instrument for those two systems in each channel 23 that's shown on that picture. I 24 Because of that arrangement we get the scenario 4 25 where the instrument break will inhibit enough instruments Heritage Reporting Corporation (202) 628-4888 l \
48
[)i .
.r.
1 on one of those systems, or the other one, depending on 2 which side the break is on. And then the single failure can 3 get to the other one. So we did come up with a scenario in 4 which we did not have-a path of flow through the ADS and at 5 the same time through the HPCI/RCIC high pressure. system.
6 Group 4 turned out to be the newer BWR 4's and the 7 BWR 5's. The important thing here is they've gone back to
- 8. the four vessel penetrations for the reference legs. Newer 9 plants have the RCIC, the listing of the plantu within this 10 group.
11 The advantage of going back to the four vessel 12 penetrations, which was similar to Group 1, but a better 13 design in the new plants shows four individual legs and an i .
l 14 individual channel associated with each individual leg. So 15 everything is separated at this point on these newer plants, 16 We found no--because of this arrangement, we found 17 no scenarios of a break in single failure that would affect 18 the Group 4 plants.
19 MR. KERR: When you refer to four vessel 20 penetrations I could get the impression from that picture I
21 that there were eight.
22 MR. COLLINS: That's what I said at the first, 23 that there are two types of configurations that we saw.
l 24 There are some with six penetrations, and some with eight i d
l 25 penetrations. The ones with the six penetrations have only 2
Heritage Reporting Corporation (202) 628-4888 1
n I
49 1 the two reference lines that are common to two other 2 measurement lines. And then there were the ones like this 3 configuration that has individual reference lines for each--
4 DR. KERR: Okay. I misunderstood you. I thought 5 you said when you put that slide on that the Group 4 went 6 back to four penetrations.
7 MR. COLLINS: Reference-like penetrations. I said 8 penetrations. I'm talking reference-like penetrations. As 9 opposed to Group 3 where I only had two reference-like 10 penetrations.
11 DR. KERR: Okay. Thank you.
12 MR. COLLINS: But with this individual leg versue fT 13 channel breakup, to break in one of these individual legs G
14 only takes out one channel of instruments causing it to read 15 high. Single failure le not capable because of the logic 16 and the additional instr uments available to completely take 17 out enougn of our systems.
18 Group 4 success matrix looks similar to our Group l l
19 3, but now because of that leg configuration for the l i
20 measurements and the way the level instruments are hung on {
21 those various legs with the four reference penetrations I
22 versus two, and the way the RCIC and HPCI instruments are 1 i
23 now attached to these legs, we never found any kind of l 24 scenario that kept us from getting through this path with j i
25 one approach or another. )
'1 1 (~b Heritage Reporting Corporation l (202) 628-4888 i
L___-._---__-- - _ _ - - _ _ _ _ _ _ . _ - .
\_/- l 1 Likewise, Group 5 we could have probably included ;
2 into Group 4 with hindsight. Groyp 5 is. BWR 6's. They 3 still have four vessel penetrations for the reference legs.
4 like the Group l's. And they are RCIC. And there are just 5 the four plants within this group.
6 The typical arrangement looks very similar to the 7 Group 4 plants with an individual channel associated with an l 8 individual log so the break will only affect one channel, f 9 leaving the single failure needing to get three other 10 channels and it can't Le done under our assumptions.
11 The Group 5 success patrix, agcin very similar to
~
12 the Group 4 success matrix and because of the four reference
/~S 13 legs in the configuration of the instruments, We'always had (d
j 14 some path through this chart where we alnays got a reactor lii shutdewn to decay heat removai or sono form of inventory 16 control.
17 Now, from this we arrived at six potential 18 scenarios leading to the core celt. Within Group 1 we i 19 identified three scenarios One of them'ceing on the A or 20 the C leg outside the dry well with the main feedwater 21 control arstem failure. The break outside the dry well took 22 out our ADS path and cur success matrix. Tt4 single failure 23 took out any kind of high pressure injection that we had for 1
24 these older plants. 1 l
25 The other two we had were in the B or the D leg. l O
Heritage Reporting Corporation
(?.02) 628-4888
l
,Q 51 A/
1 These were the legs that had the ADS and the feedwater 2 ' control instruments on them. So the break in the B or the D 3 leg inside~ dry well, a single failure of the ADS level 4 transmitter would give us a failure of three of the four 5 level transmitters,'so we wouldn't get any ADS actuation S And because the feedwater is controlling c"f of one of these 7 two signals, the feedwater is' controlling erroneously.
8 Therefore, the single failure of the ADS level transmitter 9 and locatior. of the break gives us a potential scenario.
10 The other scenario being the B or the D leg 11 outside the dry well. Because it's outside the dry well, we 42 get ADS autouatic actuation. It's in the instrument line 13 that the feedwater control is on. Therefore we don't need
)
14 any other single failure for that scenario.
15 DR. KERR: Ncw, I believe you said earlier that 16 you reached these conclusions assuming that the operator 17 taxes no action.
18 MR. COLLINS: For two hours, that's right.
19 DR. KERR: Thank you.
20 MR. COLLINS: Within Group 2 we identified two 21 scenarios. Agnin, this was a two-reference line design.
22 Therefore a break in one of thase legs took out half cf the 23 instrumentation. Therefore, because of the arrangement, a 24 single fa!1ure in the HPCI system or a single failure in one 25 of the other level instruments give us these two potential
/
Heritage Reporting Corporation
!202) 628-49s8
)- '
1 scenarios.
2 The Group 3, we had a two-legged machine again. j l
3 The break outsi6e the dry well again inhibits our automatie i 4 ADS actuation, and because of the way the instrumentation is 5 hung on the legs, unbalanced on the sides, either the HPCI 6 or a RCIC single failure, but not both would prevent us from !
7 getting a successful route through those flow paths.
8 And again, Group 4 and 5, we had no corabinations 9 of a break or a single failure that were significant. -
10 MR. MICHELSON: Let me make sure I understand 11 Group 3. You are saying that if you get an A or a B leg 12 break, one or the other, and you just taka out HPCI 3 you are g- 13 on your way to a core melt?
14 MR. COLLINS: What it gets is--
15 MR. MICHELSON: Or if you just take out RCIC, you 16 are on the way?
17 MR. COLLINS: If I break the A leg--
18 MR. MICHELSON: Then both--
19 MR. COLLINS: If I break the A leg, then I take 20 out the--I believe the A leg is the RCIC instruments.
21 MR. MICHELSON: It's go t one of them on it, yes.
22 MR. COLLIHS: Yes.
4 23 MR. BAER: --consequential failure. If the A 24 instrument line breaks, you're saying the automatic signal 25 on one of the other fails and there's a consequential
/ ~';
- ~
Q/
Heritage Reporting Corporation (202) 628-4888 I
1 - - - - - - - - - --
1
[
y.
53 h.
1 failure--
2 MR. COLLINS: Right. If it's one or!other of the' 3 legs, the instruments for one~of these systems.is affected.
4 MR. MICHELSON: That's true.
5 MR. COLLINS: And therefore the single failure 6 will--
7 MR. MICHELSON: Has to be the other one.
3; 8 MR. COT. LINS : --be able to get the.other one.
D' MR. MICHELSON: I guess format.
10 MR. COLLINS: For these potential scenarios we did 11 a risk calculation. These numbers we felt were somewhat 12 conservative. Again, we assumed the operator took no action 13 for two hours. He did nothing good. He did nothing. bad.
O 14 He just did nothing.
15 MR. DAVIS: Yes, but you assign that a probability
~ 16 .of 10'4 , right?
17 MR. COLLINS: Yes.
18 MP. DAVIS: So it wasn't an implicit assumption?
19 MR. COLLINS: True.
20 MR. DAVIS: Okay.
~4 21 MR. MICHELSON: How does 10 come into this?
~4 22 What w6s the sign of probability to 10 ?
23 MR. DAVID: An operator error.
24 DR. LEWIS: No. An operator failure to act. They 25 did not take into account operator errors for measurement.
o O Heritage Reporting Corporation (202) 628-4888 y
l .e
G k) 04 1 MR. COLLINS: Yeah. That's operator. error, i
2 DR. KERR: .I thought it was assumed that the 3 operator didn't do anything for two hours with the 4 probability 1.
5 MR. DAVIS: No. That's what he implied before, G but that isn't the case.
7 DR. LEWIS.: No, they are built in.
8 DR. KLRR: Wait a minute. There's a probability 9 of one in 10,000 that he does nothing? Or that he does 10 something?
11 DR. LEWIS: There's a probability of zero that he
~4 12 does something wrong. A probability of 10 that he does e'% 13 nothing. Therefore there's a probability of 9999 out of V
14 10,000 that he does good--that's the assumption. Have I 15 stated it wrong?
16 MR. COLLINS: We put the operator error into the 17 probability. When I said that We treated it as though the 18 operator did nothing for two hours, we did that, for a 19 example, on the break scenario where it affects the 20 feedwater control. If he reaches up immediately, recognizes 21 i t _. reaches up and switches channels, he'n all of a sudden 22 got appropriate control and all of the scenarios change.
23 Therefore, we took the assumption that he did nothing for 24 that period of time to see what systems on their own would 25 come into play and then we kind of backed up frpm that O
Heritage Reporting Corporation (202) 628-4888
l' 1
i 55 1 approach and put the numbers in. l 2 DH, KERR: I'm glad that's clear to you.
~4 3 MR. SZUKIEWICZ: 10 ctso was only used on the 4 very old plants where you had more time before you had core 5- uncovery.
6 DR. KERR: I see.
L 7 MR. SZUKIEWICZ: The other plants, the newer 8 plants, we used a figure of 10 -2 I believe.
9 DR. KERR: I see.
10 MR. COLLINS: Yes, Group 3 we used a different 11 number. j 12 DR. KERR: Please proceed, Mr, Collins. We won't l 1
13 interrupt you.
14 MR. COLLINS: Within these core melt frequencies, 15 we ended up with values anywhere from 10-6 up or down to
-11 16 10 for these various scenarios.
17 Ye took these core melt frequencies and we 18 calculated a risk for the plant., a risk with no modification 19 in man rem for the plant. This column of numbers represents 20 the remaining life of the plant. For the older plants, we 21 looked at a thirteen, fourteen-year remaining lifetime.
22 Some of the Group 3, we used like thirty-year potential 23 lifetime of ^che plant.
24 And we calculated the risk in man rem per plant.
25 We took the fixe.s that we determined to be the most probable 26 realistic-type best fixes for the scenarios. One was the
.O Heritage Reporting Corporation ,
(202) 628-4888 l
e: , ,
.~
1 .
i 56
. ' .u-(')
1 uddition of the high-pressure. injection system >, which the l
l 2 older plants di.dn't have specifically. The other one was 3 some sort of a modification within the logic, the 4 transmitters, the ADS, some other system of breaking up that i 5 instrumentation being on two different reference legs versus G four individual ones.
7 So if we take these scenarios with the fixes, the 8 risks assbeiated, we icyok at a change in risk, this being 9 the first column from the previous slide, subtracting each 10 of the other two columns, we end up combining the scenarios 11 and we end up with a total of about 105 reduction in risk 12 due to the addition of a high pressure injection system for 13 Group 1 3 down around 1 for Group 2, quite high for Group 3.
14 Numbers not differing very much for the change in the ADS 15 modification for an instrument for configuration of the 16 transmitters.
17 We then took these different fixes and we did our 18 best engineering analysis and calculations to come up with a 19 potential cost to install these systems. We took one case 20 where we assumed that the fix could be done with no outage 21 required. Therefore it would be done during a normally 22 scheduled refueling outage or some other activity in the 23 plant that the loss of power production and other things (
24 that came into account were not affected by this work being 25 done.
Heritage Reporting Corporation (202) 628-4888
l 1
j.
57 1
L 1 The other approach was we took an outage-required I L
- l. 2 worst case that you had to shut the plant down only for this 3 fix and therefore all of that power production and other l 4 costs associated with the outage would be related only to i 5 this fix. We end up with some fairly large numbers for some l 1
6 of those. Again, the ADS modification being the cheaper.
7 MR. LIPINSKI: These numbers are per plant?
8 MR. COLLINS: Yes.
9 MR. LIPINSKI: Somewhere your arithmetic lost me 10 here as to when you are talking about totals and per plant 11 values.
12 MR. COLLINS: Everything from the risk 13 calculations, these are per plant.
dgs 14 MR. LIPINSKI: Thank you.
15 MR. COLLINS: We then took the reduction in risk 16 and combined it with these costs. And we ended up with a 17 cost benefit in dollars per man rem reduction.
18 These numbers assume no outage is required. That 19 the work could be done in relation to previously scheduled 20 outage. We end up with some numbers as high as three 21 million for a high pressure injection system to Group 2. We 22 end up with some down $1100 ADS modification for Group 3.
23 In looking at these various numbers, we said, are 24 there any alternatives that are cost effective? And our 25 conclusion was basically, no, we can't justify any of these O
Heritage Reporting Corporation (202) 628-4888
'( Se 1 . fixes to the plant primarily because this fix here has 2 somewhat been addressed already by the plants.
3 We performed a sensitivity study--
4 DR. LEWIS: Sensitivity is different from 5 uncertainty.
6 MR. COLLINS: We didn't do an uncertainty analysis l
7 per se. We did it within our sensitivity scoping.
8 DR. LEWIS: Well, sensitivity has nothing to do 9 with uncertainty. So in a sense, I really don't know how 10 weJ1 I think you've done with this. And you don't either.
11 MR. COLLINS: Probably not. However, I do feel 12 that the numbers we have used are quite conservative for the )
13 most part.
14 DR. LEWIS: Conservatism has, again, nothing to do 15 with uncertainty.
16 DR. KERR: Let us continue.
17 MR. COLLINS: That's all I've got to say.
18 DR. KERR: Questions? Comments?
I l 19 Mr. Collins, on page--I think it's 30, but in any 20 event, someone who wrote the report or some group recommends 21 that all BWR licensees should recognize the potential 22 problems associated with instrument line breaks, and have 23 procedures that assist the operator in recognizing the 24 symptoms of a level sensing line break and define the 25 appropriate operator action to be taken.
O Heritage Reporting Corporation (202) 628-4888
/~T 59
(.) l 1 Is that wishful thinking or is there some reason 2 to believe that such procedure either exist or could be 3 developed?
4 MR. BAER: Well, let me respond to.that because 1 5 looked at the study as an intricate prneess. We took the 6 numbers and the sensitivities--
7 One that we were using, the results are very .
1 P dependent on what we were assuming for operator error 9 failure. Go not only do these procedures exist, but this is 10 part of the normal training and periodic retraining of PWR 11 operators be aware of the problems they.could get with the 12 failure of an i.nstrument line or that they could get f- 13 potentially--
U 14 DR. KERR: Did Andy talk to any operators? Or did j
1 15 he just talk to people in NRR and--
l 16 MR. SZUKIEWICZ: We talked to people in NRR. We 17 talked to resident inspectors that were involved.
18 MR. COLLINS: We did check with some of our 19 licensing operator examiners.
l 20 DR. KERR: The reason I ask is because the 21 question I was going to ask, for my own education, is how 22 does an operator know that a sensing line is broken? Or 23 does he not need to know that in the symptom-oriented world?
24 MR. SZUKIEWICZ: There are a number of level 25 indications which say he would see an indication.
.C )
Heritage Reporting Corporation i
(202) 628-4888 l
l
60 1 DR. KERR: And you'd have some reason to know
]
2 which one is correct? l 3 MR. SZUKIEWICZ: Not just to do that. But then 4 there is some other indication, indications like radiation 5 level, broken line, outside containment.
6 DR. KERR: Okay.
7 MR. SZUKIEWICZ: Those are the only two--
8 DR. KERR: Okay. So you are convinced that from a 9 detailed look that indeed a well-trained operator could 10 recognize within maybe fifteen minutes that the thing was 11 broken or badly leaking and that people have thought enough 12 about this to know what to do?
I wouldn't say fifteen minutes--
{} 13 14 MR. SZUKIEWICZ:
DR. KERR: Thirty minutes?
15 MR. COLLINS: This scenario was one of the 16 favorite scenarios of our licensing examiner people from the 17 INEL.
18 DR. KERR: That's great. I'm trying to find out 19 whether anybody has looked at it from an operator's point of l 20 view who knows something about plants. I don't. So I don't 21 know whether that statement makes any sense or not. If it 22 does make some sense, it's reassuring. Because it says in 23 spite of all this stuff, you know pretty well when the thing 24 occurs and you know what to do.
25 But there's nothing in the report that leads me to O
Heritage Reporting Corporation (202) 628-4888
J
("'
O 61 1 believe that that is indeed the case. So I'm asking for the 2 wisdom of people who have looked at this in more detail than 1 l
3 I have.
4 MR. SZUKIEWICZ: What's your question again, Dr.
5 Kerr?
I 6 DR. KERR: Well, given a line break of this kind, 7 a well-trained operator would be able to recognize it within 8 fifteen or twenty minutes and would know what to do about 9 it so that it would be a rather innocuous incident.
10 MR. SZUKIEWICZ: The people that I talked to in 11 the region mentioned that there are, specially in BWR's in 12 the level sensing--with their level transmitters, entered 13 procedures on what to look for at different conditions, what iO 14 to look for to make sure that a lot of cases that you don't 15 actuate an ADS. And they all felt that there are existing 16 procedures that it appears to be--
17 DR. LEWIS: I thought it was said earlier that all 18 these procedures had been reviewed here and found adequate.
19 Did I misunderstand that?
20 I can give a datum that has nothing to do with l 21 nuclear power but it may have some relevance. And that is l
22 in 747s that don't go overseas that are two independent 23 inertials. And I was once sitting up front in the cockpit 24 with a senior American--
25 MR. DAVIS: Mr. Chairman, I have a question. ,
10
%)
Heritage Reporting Corporation (202) 628-4888 L - __ _ _ __ __-___ ____ - _
i
/" 62 C')
1 DR. KERR: Yes, sir. Mr. Davis.
2 MR. DAVIS: Your table 2-6 on page 34 doesn't 3 agree with your last slide. Is there a difference that I'm 4 missing or, if not, which one is correct?
5 MR. COLLINS: The slides are correct.
6 MR. DAVIS: As a matter of fact, there's a 7 difference in the numbers.
8 MR. COLLINS: That's correct.
9 MR. DAVIS: Table 2-6 would lead you to more 10 comfort than the last slide.
11 DR. KERR: You put this in a Generic Letter and it 12 doesn't mean much to me. You may be telling people to do f- 13 something that is impossible. Now, I rather doubt that L
14 that is the case. But I'd feel more comfortable if I knew 15 that we weren't giving an injunction that was meaningless.
16 MR. BAER: Well, based on what Andy said, the 17 operator examiner is putting this on a simulator, means that 18 when they examine an operator it's just one of the things 19 unknown to the operator that they say, oh, here's the risk.
20 what do you do? And they are tested on this. In a 21 simulator rather than a cockpit.
22 MR. DAVIS: It would be interesting to know if he 23 also is given a signal of high dry well temperature at the 24 same time that these level changes are showing up. And if 25 that's the case it's still my recollection that he'll go to O
Heritage Reporting Corporation (202) 628-4888
() 1 ADS.
I 2 MR. BAER: I've got that question down. I'm 3 concerned--I'm not sure if the concern is that he will go to 4 ADS or he won't. But my concern is that he will. But I'm 5 going to check on it. ;
6 MR. DAVIS: My concern is he will also.
7 MR. LIpINSKI: I don't think you'll find that 8 simulators are designed to take that line break and raise 9 the dry well pressure. Unless it was part of the initial l 10 spec to build a simulator. After the fact, it doesn't 11 happen. Unless the guy who's at the console knows it.
12 MR.'BAER: $2000 a man rem versus $1100 a man rem f~ 13 is close enough to our guidelines that we went back and D 14 looked at that. It turned out that although I don't think 15 we could justify posing a requirement--all the plants in 16 Group 3 having either now made that modification which is an 17 ADS modification--
18 DR. KERR: Well, see,--
19 MR. BAER: The one group of plants that look like 20 they were reasonably close to the marginal--
21 DR. KE1;R: Well, considering all that we're asking 22 operating plants to do these days or people that operate 23 plant to do these days, I certainly don't want to put an 24 additional burden unless it's necessary.
25 But I keep remembering that at TMI 2, if people i
Heritage Reporting Corporation (202) 628-4888
l
() 64 1 had had some way of knowing what the water level was in that 2 plant, we might not have had the problems we had, and I'd 3 sure hate to see us overlook something that would get us in 4 that bind again.
5 I mean if we could--
6 MR. SZUKIEWICZ: I agree with you when it comes to 7 BWRs, the procedures focusing on the levels.
8 DR. KERR: May I raise a completely different 9 issue? I don't know that there's an answer to it now, but 10 it seems to me somebody needs to be working on it. If there 11 is not, that is.
12 How do the people who do IPE's and at the came s 13 time attempt to settle generic issues deal with this one?
('J L
14 And I just use that as an example. Now, are they free now 15 to assume that this generic issue has been dealt with or do 16 they have to go through something that is in effect a 17 reanalysis similar to what Mr. Collins has described?
18 MR. MINNER: They have to go through a reanalysis.
19 DR. KERR. Well, then, what is the significance 20 then of what we are doing here today?
21 MR. SZUKIEWICZ: It's on the record.
22 DR. KERR: No, I'm sure it must have some. And I 23 am trying to--
24 MR. SZUKIEWICZ: The significance is that we are 25 not requiring a licensee to do anything. The IPE program is 1 (
l Heritage Reporting Corporation ;
(202) 628-4888
/' 65
() -'
1 partially voluntary.
2 DR. LEWIS: I wonder if I could respond to that. }
3 Something that concerns me is in terms of sending out a '
4 Generic Letter and calling attention to this. It's not that 5 it will hurt incorrect instructions in it, although that's a
-4 6 finite probability, probably greater than 10 . But that if 7 you are really convinced as you seem to be that this is a 8 negative risk, that is it kind of a non-issue, but burdening 9 pecple with things that are not issues sort of takes away l
10 their attention from things that are real issues. And this 11 comes up in training of all kinds of people. To prepare 12 them for every possible emergency, as you said you do. l o 13 Emphasize on the simulator things that are not going to 14 happen. It takes away brain space from things that will j 15 happen.
16 So I think that it has apparently zero value. It 17 actually has a negative. That's one thing I am a little bit 18 concerned about.
19 The second thing I'm concerned about--let me be 20 really a bad fellow here. You are dealing with things which 21 in some cases barely fail the mcrgins for requirement i
22 change. They are little ways off and you are reasonably 1
23 comfortable with that because they are conservative j 24 assumptions.
25 MR. BAER: One point being that the one fix that O
Heritage Reporting Corporation (202) 628-4888 l
n
/"N .
66
()
1 looked pretty effective had been made.
2 DR. LEWIS: Yes. that's right.
3 The' numbers we saw I was told were assessments. I 4 don't know what the best estimate is. But a best estimate 5 is usually interpreted around here as something like--if 6 there are real uncertainties in an analysis and if one found 7 that in fact there is something like a factor of 10, which 8 is not at all unreasonable, and if you deem it's the best S estimate, because after all, most people do what they are 10 supposed tu ao, then with a normal factor of 10 uncertainty, 11 the fee is a factor of 14 larger than the medium of the best 12 estimate, and poof, all these things that are now looking as 13 if t hey were cost ineffective suddenly become cost effective 14 by factors of 10. And it just means that you've done the 15 analysis more carefully than requiring all kinds of change.
16 That troubles me. ;
17 MR. MINNER: Well, we pointed out that the 18 uncertainty is a problem. This issue is sensitive. So I 19 guess that's why we thought--and despite the numbers--if we 20 didn't look at common mode failures in that area to see if 21 there was some mechanism that could knock that out.
22 MR. LIpINSKI: The general mode failure--I can 23 discount them based on particular fare--
24 DR. LEWIS: I would regard the probability of the 1
25 operator not doing something wrong as zero. And if I had my Heritage Reporting Corporation 1 (202) 628-4888 I
_ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . - _ _ _ - - _ - _ _ l
($)
I druthers I wouldn't consider it.
2 MR. SZUKIEWICZ: On the analysis. But it is 3 considered in the Generic Letter.
4 DR. LEWIS: But the other one, if you had done a-5 better analysis, you would have--
6 DR. KERR: Let's continue.
7 DR. LEWIS: What I'm saying is that it's not the 8 uncertainty that's causing it, it's that uncertainty--would 9 be that much larger than your best estimate, and then you 10 would be forced by your reading, not by your uncertainty.
11 MR. MINNER: You are saying we may have a bias in 12 the analysis.
13 DR. LEWIS: I'm saying you do have a bias of the t-14 analysis. That if you take into account the uncertainty--it I 15 is always larger than the medium, and the amount larger 16 depends on the uncertainty. That's just a mathematical 17 statement. You can't quarrel with that. You can quarrel 18 with it, but--I happen to remember the number. If 19 uncertainty is a factor of 10 and a distribution is larger 1 20 than a medium by a factor of 14. It happens to be 14 for 21 that particular case. It's not a trivial fact.
22 MR. BAER: I guess it's a judgment call. But in 23 one area the bias is clearly in the opposite direction.
-2 24 What was used was a failure probability of 10 for a break 25 outside of containment. So we should have seen roughly l
O Heritage Reporting Corporation (202) 628-4888 l
I __
i l
i 1 three of them. Inside of the containment, a factor of 10.
2 We should have seen three pipe breaks and what we have seen 1
3 is eight or ten leaks inside containment. !
i 4 So I think if someone applies the uncertainty ]
S analysis, that would push it into the opposite direction.
6 Furthermore, the type of analysis done by .INEL to 7 identify potentially severe sequences, is that every one of l 8 these instrument line breaks were in the worst location 9 where it defeated the system. The uncertainty analysis 10 would have to say, hey, that isn't true in all cases, where 11 fifty / fifty could get the break where it would give you 12 maybe other concerns in operation.
/~s 13 DR. LEWIS: You know, I don't want to prolong this i b but I think it's insane to do worst-case analysis. It's a 14 15 way of lying to yourself, but I don't think it gives you 16 much confidence that it gives you the right answer.
17 DR. KERR: Any further questions?
18 MR. LIPINSKI: I have a question.
19 DR. KERR: Go ahead.
20 MR. LIPINSKI: You said the slides are correct.
21 On high pressure injection you've got 4,810,000. Your Table 22 C-1 has 8,810,000. You trimmed it down by 4 million?
23 MR. BAER: There's a factor of error to allow, I 24 believe--
25 MR. LIPINSKI: Yeah, but on the cost estimates? !
)
Heritage Reporting Corporation (202) 628-4888 l l
1
0 1 MR. BAER: On the cost without outage--
2 MR. LIPINSKI: Option 1 in the report says i 3 8,810,000. Your slide has 4,810,000, 4 MR. BAER: The cost doesn't change.
5- MR. LIPINSKI: That's what I figured. So is the 6 VuGraph a typographical error? The report is on C-3.
7 MR. DAVIS: While he's looking--
8 DR. KERR: Go ahead, Pete.
9 MR.. DAVIS: I wanted to go back to Dr. Kerr's ,
10 earlier point on the IPE's and argue with you a little bit, 11 Warren.
12 I think--it's my recollection that BWR PRA's do 13 not generally examine instrument line breaks as an 14 initiating event. They use small break and large break 15 LOCA's only. And I don't believe there's a location 16 dependency.
17 In v.iew of that, I don't think they are going to 18 pick this up in the IPE's. The station blackout they will 19 because that's always been an important initiating event in 20 all PRA's and will be examined, but unless they are 21 specifically told for the IPE's to look at this one, I don't 22 think they'll do it.
23 DR. KERR: Well, Pete, what I was asking was, one 24 of the options I think in an IPE is to clean up all the GI's 25 and unresolved safety issues at the same time. And my O
lieritage Reporting Corporation
[ (202) 628-4888
70 b s_)
1 question was if they say--would it be acceptable for them to 2 say that that issue has been resolved, we don't have to deal 3 with it, and I think Warren's answer was, no, that wouldn't 4
4 be enough, that they would still have to--if they wanted to 1 5 use the IpE process to also deal with the GI's, they would 6 have to do it plant specifically.
7 Did I misinterpret your answer, Warren?
8 MR. MINNER: Well--
9 DR. KERR: I wasn't asking ebout the process
{
10 alone, but the special case, and I think this is given as an 11 option--
12 MR. MINNER: I think your definition has confused
('}
V 13 me. But what I'm concerned with requires to report the 14 information, okay. That's what it requires. Let's just 15 say if you are doing a good job, they would have looked at 16 the sequence and picked it up.
17 Now, if the next question is what do you do about 18 those things, and that is the question of where 'che genetic j 19 Assues come in. Now, if they had picked this up and then 20 looked back at our records, there's nothing on it, so we 21 don't have to do anything. I'd say that would be okay.
22 We'd prabably accept that.
l 23 But if you had another issue in which you had made 1
i 24 sorne requirements and they had picked it up as I
i 25 vulnerability, I would say, hey, if they didn't come in with Heritage Deporting Corporation (202) 628-4888
1 come proposal to fix it, I would say no.
2 DR. KERR: So it is possible, as you understand 3 it, that if they said we dealt with this generic issue and 4 we agree with the staff's evaluation and it's no longer an 5 issue, that might fly?
6 MR. MINNER: That might fly.
l 7 DR. KERR: One additional question. Do you in 8 the course of studies of this sort make any recommendations 9 to the people who are involved with new plants, the people 10 that are involved with reviewing new plants--let's suppose 11 that your current analysis is based on backfitting and going 12 in and working on existing plants, but suppose you had to 13 make a recommendation to people who are designing new O
\/ 14 plants, would you tell them to do something different on the 15 basis of this study?
16 MR. BAER: This particular case?
17 DR. KERR: Yes.
l l 18 UR. BAER; In this particular, probably not.
19 DR. KERR: Okay.
20 MR. MINNER: But the new plants have agreed that 21 they will resolve all generic issues.
22 DR. KERR: You know, you might learn something in 23 a study like this. I rather imagine somebody does. And--
24 MR. MINNER: Well, they had to do a pRA.
25 DR. KERR: Look, there is a the6 rem called the Heritage Reporting Corporation j (202) 628-4888
() 1 cease theorem that says PRA's can demonstrate that a plant 2 is unsafe. They can never demonstrate that it is safe.
3 PRA's only reveal what you already know about the plant but 4 maybe have tiot analyzed.
5 MR. MINNER: I've done a lot of programs.
6 DR. KERR: So, I just thought, you know, if you 7 really learn something important that you wouldn't want to 8 put on an exieting plant because it's too much trouble, too I
9 expensive, and all, but if you thought of a design change a 10 that might be very straightforward when you are designing a j 11 plant. It would be unfortunate not to get this information 12 to plant designers.
13 MR. MINNER: I think we are a little lax in--but ,
( 14 if you did your cost impact analysis--
15 DR. KERR.: But this cost impact analysis is for an 16 existing operating plant. It ought to be different for a 17 plant that's being designed.
18 MR. MINNER: I guess.
19 DR. KERR: Markedly different I would think.
L 20 MR. MINNER: Yes, markedly different, that's 21 right.
22 DR. KERR: Any further questions? !
23 MR. MINNER: But I guess the new plants are all 24 probably--
25 DR. KERR: Mr. Lewis.
O Heritage Reporting Corporation (202) 628-4888 h _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ . _ ._ l
O 1 DR. LEWIS: I wish I were less confused than I was 2 a half hour ago, but I regret to report that it isn't true.
3 I had had the impression that no uncertainty 6 4 analysis was done, but Appendix C, which is labeled 5 " Sensitivity Analysis," actually describes an uncertainty.
6 It just so happens. And it's a very interesting one because 7 the way it was done is a little bit--I've got to say, on the 8 strange side, because error factors were chosen from various 9 data bases and reports and the error factor on operator 10 error is plus or minus a factor of 10 for some plants and 11 plus or minus a factor of 30 for other plants. plus or 12 minus a factor of 30 is fairly big, you know.
13 Further, it says that the way the analysis was O 14 done was to take samples from between the bounds, evenly 15 distributed, for which there is no theoretical basis 1B whatever. And to run 3,000 points in each case, and then 17 the uncertainties and the cost benefit ratios are given on 18 Table C-3. And the famous $1100 per man rem shows up as a 19 fifth percentile, not as a best estimate.
l 20 So in fact, the numbers that you said were best :
21 estimates are really lower bounds on this completely screwy 22 uncertainty--forgive me.
23 DR. KERR: Somewhat unconventional.
24 DR. LEWIS: Unconventional. Thank you. That's 25 the word I was groping for. Uncertainty analysis. That Heritage Reporting Corporation (202) 628-4888
'( 74 1 leaves me in worse shape than I was a little while ago. But 2 .I'll let it go at that. That's not a question.
3 MR. BAER: There is a factor of two hours.
4 DR. LEWIS: How can that be with errors of plus or 5 m .s factors of 10?
6 MR. BAER: I'm ,iust telling you that at the time 7 that this was run, that the best estimate was a factor of 8 two errors and the best estimate was $2200.
9 DR. LEWIS: I'm sure you won't object if I go away 10 confused.
11 DR. KERR: Any other comments or questions?
12 MR. LIPINSKI: What was the answer to the question 13 on the cost estimates?
14 MR. COLLINS: On the cost estimates on Table C-1?
15 Page C-3 should be $4 million. That's a typo. Yes. It 10 should be the same as one of the previous tables. I think 17 it's Table 5.
18 DR. LEWIS: All these things on C-1 are fifth 19 percentile now?
20 MR. SZUKIEWICZ: Table 5 is the one that table was 21 taken from, I believe.
22 DR. LEWIS: Table C-1, should I reed that as fifth 23 percentile numbers?
24 MR. SZUKIEWICZ: No, not those.
25 DR. LEWIS: No, I understand the error on that Heritage Reporting Corporation (202) 628-4888 r
J
. 75.
1 one. I just wonder what the numbers mean. These numbers 2 are meant to be so-called end dollars. But then C-3'is--and 3 it's just a coincidence that they are the same.
4 DR. KERR: Any further questions?
5 Andy, if you would use that mike--
6 MR. SZUKIEWICZ: The sensitivity study that we did 7 was to evaluate the numbers that we showed in that graph and 8 we evaluated the $1100 per man rem figure, and then when we 9 did the sensitivity study, it showed us that that value was 10 on the low side. In practicality you'd have--it would be a 11 lot higher value. And that's all this is trying to say.
12 D T. . LEWIS: No, but it actually says in Appendix 13 C--but that depends very much on how you weight the value.
O 14 DR. KERR: Are there further questions or l
15 comments?
16 MR. MINNER: Did we answer all your questions? ,
i 17 DR. KERR: No. But I don't think I'll raise them 1
18 again in the light of our discussion. I have answers--
19 partial answers to some of them that I did not have before. ,
l 20 I believe that we need to write a letter at this l
21 meeting to someone, perhaps at EDO, saying that we--I'm not l
22 sure, that we approve the resolution but here are some I 23 comments, or that we don't approve the resolution and say 24 why.
l 25 My own recommendation would be approval but with O
Heritage Reporting Corporation (202) 628-4888 1
(
'8
(:)
1 some comments. But I'm open to suggestions. 3 2 DR. LEWIS: Can I ask a question for 3 clarification? l 4 The fact that this particular study--and I'm sure ,
f 5 it was answered nnd I missed it--that this particular study l l
6 c concentrated on a tensing line failure, means that other 7 means of water level sensing failure were dealt with 8 somewhere else?
9 DR. KERR: I asked a question earlier and it is my 10 interpretation of the answer that the iscue of water level 11 indication in BWR's was a generic issue. Generic Issue 50, 12 if I remember correctly. It was dealt with except this 13 particular facet of the problem was left as an open issue 14 and was identified as a Generic Issue 101.
15 DR. LEWIS: I see. Okay. !
16 MR. MINNER: That's correct.
17 DR. LEWIS: This is sort of the dregs from that 18 earlier--it's a little late to get into that.
19 MR. MINNER: Well, the industry volunteered to do 20 the other fixes.
21 DR. KERR: But did not feel that this was a 22 serious problem.
23 MR. MINNER: That's right.
24 DR. KERR: Okay.
25 MR. LIPINSKI: When it comes to the Generic Letter i
Heritage Reporting Corporation (202) 628-4888
_d
l I
)
1 and the question of common mode failure, you really have to 2 address that issue for every plant individually, And just 3- takc a look at one plant and say I-can assume this for every 4 plant.
5 It seems the letter should'have a paragraph !
i S telling plant managers to ask the individual plant owners to 7 lock at their plant relative to possible common mode 8 failure. To point out what your analysis involved. the 9 assumptions tha.t went into i t .. Call it to thel. attention 10 to see whether this is indeed another factor that should be 11 ir.cluded for a particular planc.
12 MR. MINNEP.: Well, we used to do that for all fg 13 generic issues. We ussd to send a letter to each utility V 14 that says, . hey, here's to the generic issues which we l
j 15 prioritized--I forget how we phrased it, should we look at' 16 this to ensure that there isn't some quirk--
17 MR. LIPINSKI: Yes, but in this particular case 18 where there's a question of the completeness of the analysis 19 based on that issue, your letter probably should raise that.
20 MR. WYLIE: The earlier studies that were done on 21 the water icvel indications and effects on control f
22 production systems, did it not look at common pode failures?
23 MR. MINNER: 1 can't remember what the earlier ;
E4 studies were about, so I can't answer that.
I ?5 MR. WYLIE: That's where it should have been done, Heritage Reporting Corporation (202) 628-4888 l
l l
78 c.
1 isn't it? Since this is limited to a very narrow scope and 2 just the sensor lines.
3 MR. DAVIS: it's unique though, Charlie, in that 4 the break in the line can be a common cause initiator 5 itself.
6 MR. WYLIE: I understand that.
7 MR. DAVIS: The other failures that were examinea 8 in the previous issue did not involve these environinental 9 concerns that this one is uniquely a problem with.
10 MR. S2UKIEWICZ: Well, they did address the ;
i 11 environmental concerns because one of the fixes they !
12 recommended was to put some insulation around the reference 1
(y 13 legs so the accuracy of the level sensors won't change but L/
14 the--
15 MR. DAVIS: That's right. But that was only 16 because of the problem if you have high dry well teinperature 17 yoe can get erroneous readings in the R way system. But 18 that was not a rupture in the line that causes the--
19 particularly outside the containment, I think is the concern 20 here, where you can possibly have an interaction with the 21 feedwater control.
l 22 MR. WYLIE: But I would assume that all the i 23 transmitters were identical, weren't they?
24 MR. DAVIS: You mean in the same plant?
25 MR. WYLIE: Yes.
O Heritage Reporting Corporation (202) 628-4888 i
O 79 {
V 1 1 'HR, WYLIE: So I mean that's e. source of common 4 i: mode failure right there. And I assume that was looked at.
(
3 I don't know.
4 MR. DAVIS: Yes, 1 thi.nk it was. but I'd have to 5 go back.
6 DR. KERR: This is all very interesting but 1 need 7 some guidance on what to da about-- 4 s MR. WYLIE: Oh, I agree with you what you 9 proposed.
10 DR. LEWIS: I would agree with that. l 11 DR. KERR: Okay. 3 12 DR. LEWIS: But I must say my comfort with it q 13 3 3esn't come--
h 14 DR. KERR: I wish you would think about possible 15 comments and give me some to--I will draft a letter.
16 MR. DAVIS: I agree.
17 MR. MI CllELSON : The comfort should come with the 18 narrowness of the scope.
19 DR. LEWIS: Yes. Hy discomfort comes from two 20 things. One is the narrowness of the scope and it just l
21 shifts my discomfort to what came before.
22 MR. MICHELSON: But now you wonder, well, how did i I
23 the stuff before ever get handled. I f you are curious, 1
24 you'll go back. '
25 DR. LEW1Sr If the big issue was handled at this O I He ri taste Reporting Corporation (202) 628-4888 )
i I
l
- ---- _ _ -------- --- - --- 1
80 1 level, I'm concerned. My other comfort from this comes from 2 the fact that I keep getting reassured.that really it isn't I
3 going to happen. i 4 MR. MICHELSON: Have we actually read a letter-- 1 5 DR. KERR: We'll have a full Committee meeting to 6 provide a set of information.
7 The next question is, do we need to take up staff 8 time with a presentation to the full Committee or will a 9 Subcommittee report suffice in your view?
10 DR. LEWIS: My advice is that you report it.
11 MR. MICHELSON: Unless there's some really 12 argumentative problems that come up.
13 DR. KERR: Does the staff want to make a 14 presentation? I don't want to--
15 MR. MINNER: Both Bob and I are going to be there 16 on the afternoon of the--well, let me take that back. What 17 day is this scheduled for?
18 DR, KERR: The 6th, Thursday. It might be worth a 19 short presentation. Prepare about a twenty-minute 20 presentation? A twenty-five minute presentation? What do 31 we have, a half an hour?
22 MR MINNER: A half an hour.
23 DR, KERR: A half-hour presentation?
24 We thank you and that ends the recorded session.
25 (Whereupon, the proceedings were concluded.)
f Heritage Reporting Corporation (202) G28-4888 l
1
) 1 CERTIFICATE 2
3 This is to certify that the attached proceedings before the 4 United States Nuclear Regulatory Commission in the-matter 5 of:
6 Name: ACRS SUBCOMMITTEE ON INSTRUMENTATION, & CONTROL SYSTEMS 7
8 Docket Numbert 9 Place: Bethesda, Maryland 10 Date: March 29, 1989 11 were held as herein appears, and that this is the original 12 transcript thereo! for the file of the United States Nuclear 13 Regulatory Commission taken stenographically by me and,
() 14 thereafter reduced to typewriting by me or under the 15 directiun of the court reporting company, and that the I
16 transcript is a true and accurate record of the foregoing {
l 17 proceedings.
18 Ja/ ( tc h 6 x V/'* f f 3 N /
~
19 (Signature typed) : l l
! 20 Official Reporter 21 Heritage Reporting Corporation 22 23 24 25 O Heritage Reporting Corporation (202) 628-4888
~
L E
N BI l ,
" Er o Y r
,t C S ega M gi N E at s A T n e D N Y S av OS Mn N
I kl U E L E O 9 Bap sa D T R T T 8 l E Ti c E M N I
9 ,zi
,n O
R N O MC O 1 f
i eciP r
h w, 1
0 L I T
A O
CD 9 Ces n 1 E T N
B N U A 2 r i
,ki l I
V E S N H e a zo ul G E S S O B.S C L E R R TI C RAB P CA R R A T E N A s r
E H E M t o
T T M n
e A OR U
s W T T S
P e
r N
R 1 W
B 0
- Y S T Y R DA N
DL L
E A E P T
L E L
I BE M YT A E I
G E I S
C M R E HA V N B D M NT I
N E
I H
M R
AO D T ER A
U EI V B C A
O NU UA NG I
S N Y AR L O F
R E
DN E
L P T
W NT U
CA T P R O EN S NE OE D
E O T
T D NT M LA G S T
I TM CO D
E E E U I
CH U P R
TT N I F AS E
N L
B AN F S I
SA NH OI K MN E LW FE iT D C UF '
A EO B I SA H O T AO T S E S B
N I
L U
VD AT EU DE I
WO O V V I -
- F S N R O
_ T OB TL N
TE A ON U RF I
A M E
EI T L TW I
CO OS TO NT OS S Y D C R I
OT RN AD HR A L E S LE U RU TE P C A TT RO I
PE F Q K NN
- N E R
CM U O
T HOS ASMNO RE ED N
S EAR I H
- I I T RC R
_ A R HN W M TS R I
E PA FOTC OP I N ON LU OS2A L E E BS T
- Ti O UY - - L D U O AN PSN O T N AB I
I
- l L l l
~
N I E s D T R A O E I W T YT RT B R FN P R N L L
O P
I RE E S O E R DL O
I A
O E
R S
VE DP R
P P OV A
RE E
T T N
G N
NE AR NIN L I
U O I
AGR I D WY L T NE NI TS T N L A I E T I
VC NA K O F AI MAWA O M R L A
E E RR OTT R E E R
S O F C I
OO R
T R AB GR E N I
N H
~
D'E DOI ETTO R R O
C E ER T
tai SR MS F T CA EE D K
A 1 R D ES UP NE E N PNI S QO AL 0 T T
A XGS EISL Y E
R ND FN YE 1 E S L E2 EEA EAI TL I I
S1 1 RDN R C A T A ASN EE T G I R
E Y5 L
AR SNF EAF SRD N E UI E N NC ELA M E A/ SPT E DY S E LU G
N T S N CD R I
G E E E U R E AE E OA T U S S
U C HH C RE S I I N
O S S
I N LTT LPRI I
r o
f
~
f e o 1 d
e l
i r
u r a d 0 f
e 1 d e o i s
n l
g hI o c
i n
s tG e t s s 3
c n
u Mr n a n o
o r
a O
ff o l a i s
r g <
a R a 5 e o W a o b r
s B m o
t n e p yi s I c i i
n u r A s l o
ay l g f
r o t n
a c
n r
g man i
n a l
a sh n c i
r t e a e ea se mA u
h t
a g l c
e i
z r
o en z i S
i p g yt a e l a a n t m t a o a nl O D C C Ap 1 2 3 4
' l l
d f
o i l
c t b n t y u p e o i l
i e
t a
(
c b a t a
l u
b l c r
o u l a
c i
- s. p l a
c c
s t e d n
y l a d n
a l u a s a l c n n a c
s e i o
O A s i
r o
t a
c e g i f c e i o n e
t a
c c
o u
c m
d e e s e o s a
e l
i a
h t n l e
s st t a e i f
c r e
oi p f e
M i n s .
i n e n e t
ge m b 1
i s u r r a/
u e<
t 0 l t as ei s ri 1 oa r voc Ff J E O I 5 6
7 G
j ;
i n
w l
o o
f n e i o
n h t
t a
r
)
i o n o i u
g C
I t s n
a p u
o c
i c o r
- e C
I C
g n f i R i [
o o s t n
g s i i
n s
i l
a d i
e n e e e
p t
C f i
s s
l s
t 3
y t y
i e t f
t a v l n
l e e n c e
l c
o e
m l
a r e t r
o m i n
a c
P w a e
R W
t n
o t
s B C S R R W
B:
1 2
3 4
s O es i .
h a Tb
s 1 t
n r n o s e t e
i t t s k n
1 m a r
o n e i
o n t p e e P i
e d r p a n e n C e O t t o l n e a i
u o c
3 s
c r e M o
r 2 I l
e s d n
e i
t n
o t
s y i e
n s n a O N G R W
r a
e v
o c
l o
s B M 4 4 I O ,
g n n n L e a a h
a d e
t k a a C
CC u e l
r t t sB g n P o L e
e /
D e
m t a -
e n s nt e P oC O ) I go d n
S s at a D er A u s
r s r c e
r
. P o AoL t r
o e c
l s
a e s l
ReV ak ca )
B iep )
C (A r & s yB t e A u e
r T
(
a S se s
s d
nt e P n
oCQ D AP r
1 o C
\g p L e
u d e B o
r t k l
t a a u er o G P sB o
L g
e l sB e
A nd n n a a h
CA
- - , l l
ll)\\1{
u o Ml TI Ti0N ,
u x dJ h Da)ni o
y rl o ro i
Wtc t n
en t
r S dj e o t e vC a [ck( F I
n M /eQ s
s I !l ' ' ,! lllllllllll e
o c c Dfjst y
t u D/r e s hye e
ae Hv al ea S hAD in t
l a de t at es Sv
/l f a a ym o
- A c
n o Sg
/
i l
e e
c Re e
1 Mc\ [RQ D P
U I!lI! !j,! ' lllllllll l O
R r ow n
G Ob "(
t o cd at e u R Sh o '
I Nl T TING EVENT ll
s t
n r n i o s e e t t s 2 ma n t r
o p
n e 1 i
e d 2 3 a e n e O p t n
n e t a o n n n u o c
p s n
c d
e d
e t
o o
r S I e
s e
c i n
o s
e r
s e
r l
i s
s n t a D D M G R Wa r
e v
o c
l o
s B M2 4 I
- * = *
- O '
o d D 5 1
d e t k a a l
e 0
1 0
B n l t
u e r n
n a
L L
a P sB o
f h C
t n _
e t
e a
m i
s _
vO n
e n d e
n B
l go ni C
o g
e h e
n n
a L C at a B r
r c r o
l o e -
AoL t c
a e s
s _
R eV l
ak ca g
C e -
i per e L
A l
e n
n t
a yB a s h C
T d n
e n
vb i o
C 2
p u d o e t k a a A
O r l l t
u e r . n e
G P sB o
h C
n a
h i
x r en t a au r wc t
a ei n ei F V M@Ob@gde _
s -
s! !!lll e
, l l Illl
_ o c c /,
u Di ae s
S bT t
Sv
/l l
a f a o
s 2 / Q P
U ! !!l! ll , j!!lll O
R G hj(
o "
1~
NI TI AT N E ENT
2
/
1 2
/
1 a n
k n 2
/
k c a ci s ii r l t
n s c t r eh wt s a e
u mc i
e n a p c n p u tz L mo e ot o a r i r
e s
ui 3 R i
ni t a a a t CHBF FSS r - - - - - -
pB Wt nt o e s
n e
O n u nIc c
e 3
c n 3
/
r o aI
/
I l
e o
c 2
2
/
1 e
e 3 s4 /
s k G Wa ev oI CC R k r
r 1
s e
ynd r
r eY a o l
n oi r B M- 2 2 R l l
t i
F tn A eC s
- *
- e c mn o e wmna i
i t d r n a g o r o u l i r e u MQP BVD3 O .
d t k a a L e n
n l t
u e/
r B
a n
a P sB o
' h C
t n
e t e
a ,
m s nt e n e d P n
o nV B l
e go ni C
o h
n n
a C
at ra rc r o
l o e AoL t c
a s eV s
e R
l ak c a C i
per t e
l e
n n
a a
yB s nt h
C T d Po e
n nU ,
o C
3 p
u d e
o t k a a A
o g l
u e l
e r t sB r
L e n n
G P o
A h C
a
x he i
t r
a
/D Te M T V s MDbegq n D @
\
s e
o c! l i: ! iil !!l ll c
u 39 S %j
' a 3 eQ 0 eD P
U ! l :: l l l !!l l i O
R G Q,V o .
I Tl A T 46 Ev T
2 t
i n
o P
t l e
n s i
e n 2 M mo ni s P N
e t n 4 i t
a a r t
0 3 Wl i
\
nt o e o p c n t e - -
u e p a
_ I s o
r 5
/
L I
/
I l
e e s c n
s n G R k r
e oC v cI k 2 Wa BM44R C e e
r k
/
1 C i c l e
- * * *
- r l e e a 3
0 i mSa F L L O -
d n n L e
- C h
a a
d C e -
t k l
a a
.t u er g
t sB o e n P L e e B
m t a
e n s
oO O t
n B e P g d l ni o n e o n C n a
h at C ra rc o AoL t r
o e c
a l
s s
e l R eV ak c-a i
per D l
e y e n n
T B t
a a s h d
nt e P n
oC Q C o
4 C g
u e ~-
p d e
L D
u t k a a o u e l
t r sB o r P o g e A G L A
l e
n n
a h
C
I x
D i
r t
a M
b D 7
N sM s
e h Q O C
o c I c ll ll ; ';l i 'j l lll u -
S A B ] h*
4 P
S D
A d
S D
h Ya a
U ll l !; l jl!!! llll O!
R G p'Q 7
o -
Nl T'ATI N0 Ev NT lllI
d n
e B
y r r e t
s r
ei v
n n P R e o mis t t n a o 5 i t
a t r p n e e p o n :
o e -
c a u 3 s
o r 6 I
I I
l e
s n
e c
f G
u s n GR <
r e
v o
c C
I n
od n
W B
a M4 L C
R t
i l
n a r
CG o '
i ! l 4
d e a n d L
C h
C a l e
t k a a u e r t
sB g t o e n P L D
e t e
j/
a m s e n d nt e PoCC D l
go ni C
n o
h e
n n
a C
ata r -
rc r o AoL l
t o e s c
a se l ReV ak c a C i
per e l
e n
y t n T B a a s t h n
e d P n
oC Q -
C o
C 5 \g e
L p d e C u t k l
a a u er o t sB g
o r P o
L e A '
G A l
e n
n a
h C
x D
i r
t a
- bV D MeD' v C
I N
s C R
O C
s e
o c c lll l l !l lll: l l lll l u ]fjs t
y e
S A B ae e
Sv S S /l f a D D 5 A A i l
eV e
P U i:ll!ll !l'l l l lll Ol R
G QV o -
I Nl T ATI NG Ev T
s S y
i o e l
o r
r '
)
h t
o r r t e b u n t a l i
a C o
t i
m t
o n F e
r e
s n
a t
n
(
n et l g a n
t T
r d e m e
m C
I e cl w r e u C r SeS a i i d
l u t s
r R e e q t v e y s r e e
L M F n
L e R e
S I
l n
I o
l i
r u
a l
i S n C e C Ae r v
. i f f a D o P e P I
M A N H L H l e l g
To l l
e l
l l
l e
l l
e l
l e
i n
s i
NC w y
l l
e w
e w
y w
y w
y wd y n d
oE D r
y r D r
D r
D r
D r a To t n o d e D e d e
d e
d e
d e
k a pu e o k
C' r i i i i i
i s d s s s s r t t i t t t t b g b a u s u u u u Pg c o O i n O O O O f
o s i f
n L , , , , , ,
h g g g g g g st s f k e e e e e e n n i
a L L L L L L o ro od e r "
C B
B B
i t f a
i t
a aB D D i nt n i n
ye "
r o
r o
r o
r o
r o
r o
b a mc of i
b m
o rL a " " " " " "
ci o g n c o
A B B A A A i
" " " " " " N s N m i o
r m a n
u S e
c a b c a b oS .p u
o 4 r 1 2 3 G
s y r r r r r r n c n
/
y
/
y
/
y y
/ /
y
/
y i
o et r
o CMq l
e u e 0 1
0 1
6 8
1 1
6 e E E E E E E t r 4 0 0 5 8 5 l
a F 2 4
2 2
4 2
o u l
c a
C i o a b
c a
b r 1 1 1 2 2 3 a
k n p p p p p p i
s S e
c u
o r
u o
r u
o r
u o
r u
o r
u o
r G G G G G R G o ,
mD n san n ol a e e e e e e s ar rOa/cm iP t l b
l b
l b
l l l n T i
g i
g i
g b
i g
b i
g b
i g
o nf e i i i i i i i l l l l l l h oiR g g g g g g t i d - e e e e e e i i t o n N N N N N N WdMa i
t a kA d
M i
c i R
s f
i d n o re o i t u i
d s t
a n
M l
d snP Arieo/ 4 4 2 5 h
Pt cm E E
3 E E
3 e
ot n ih jR e t
Wign i n
6 1
7 2
3 1
8 1
4 3
8 3
H l
a k sf M a
iO P R r nt t
e ht t aP i
o a l
n i
c/ 2 2 3 f
Wfd m i
- - 1 0 A i E E E e 6 3 0 k
s oR 3 i
. 0 4 7 0 iM R n 1 2 1 1 2 3 k NM o a i
s . .
R a b c a b o 2 2 3 o
i 1 1 1 r
a p p p p p p
. n u u u u u u e o o o o c o c r r r r r r
- S G G G G G G
e n S el va D L P/
A l s o n am t
o e r e et i
u D ui dn d a 22 3 D n AM -
EE
- 1 .
1 3E 3
0 o 31 6 6 6
) r n 0 0 47 4 0 n sn i . .
t oi 1 21 1 1 2 1 3 c s o
io u n dei o e r
ti t t ct R ai t ua k i cm dc i si f s R d n i
ef i o a r MrT d
ko O i sM r t (d n e n a ol ks e iP t
u i
Rp soDf c/
ojm e s
n n e oo r
nI o r
n i
t e - 22 3 iP ct i
r n - - 8 8 - 2 ui du a EE 4 4 1 E 1 2
d ssM d 31 47 4 0 0 0 eoe gt R A e r i n 1 21 1 1 2 1 3 n k s
t oP
- m a i hgt e h R i
s C H y S
O t p 1
l 2 ,
l 3
n u p . .
a p . .
a a or t t uab c o uab o l
o T o T P G r r G G G F
0 0, Sa 0, 5 Dcf i
5 1 Ai 1 8
d 3 ,
o 1 M
r t
e t
l di enm sAs 0 s n 0 e n a 0 0 Vor 0, 0, s rtT oar i
0 0' 3
t e t ct a
e n en l
i a
o 3
8, 2
8 1
1 et a R Pid d
A_
O i m
- t m s ee rt us 0 0 E sy sS e
0 0,
0 0,
0 r 0 1 Pn o 1 8,
- t i 8,
s ht i
gc e
4 3 1
- o Hj I
n C
)
- d e
- r i
u e g
q e a) r t d
- o ue
- ne or i
- g u t
i nat hq t
ie
- w uo wr O
- ( {
t l i
- a a t t o o T T
~
i
s v ei y
ni v t c
t e A af f o o o e nE r
N N N r e t-n) At l A o s h od C t
e i
i t w ci r y
- l yuu n o
p m
rd q i t
0 0
- o S a aee D c 0 0, 0 c
- mRr Ai f i d
o 0
0, 3
0 2
2 0
1 1
t o
s m mg e M n o
u i t
SeRt a g e s i
f a
c O t - u l a L t
r e 0 0
i d
o ino f
n o ec h t i 0 0, m ea i t ni mt 0 0 0 r nmon
~e ie d
d e Af rws e T n
a r
7, 6
2 0
0, 2
0 2
9 l
i i
a m
s R
Br
/eg e) d 8 t pn a1 s .i e r 0 m (3 e s m l a s n
u s on m 0 s K.
C' r u 0 0, 0 t nI o et e i 0 0 0 aI as i t
d P- e i
r c ts 9, 0 4,
9, l pm l
l s d h nS j y 5 4 3 5
1 3 te oA A g i
H i
p u n i
D( oi o rt Gc t
A 1 2 3 s!
O t n u p p u u p p u MT oM l
a or o r
or r o
P G G G G
- l,
-