ML20210B132
| ML20210B132 | |
| Person / Time | |
|---|---|
| Issue date: | 09/09/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1546, NUDOCS 8609170456 | |
| Download: ML20210B132 (220) | |
Text
. .- -
O M \fL RCRSF /M6 o UN11El> STATES V
NUCLEAR REGULATORY COMMISSION IN THE MA' ITER OF: DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON DECAY HEAT REMOVAL SYSTEMS c
LOCATION: WASHINGTON, D. C. PAGES: 1- 179 DATE: TUESDAY, SEPTEMBER 9, 1986 2"' .""?'??(
h lkf, j ,
b&___umoempsone ACE-FEDERAL REPORTERS, INC.
O .
'D a w-444 North CapitolStreet l
\ Washington, D.C. 20001 l g\ (202)347-3700 ge0917.>4Se 960 W NATIONWIDE COVERAGE ,
c, 4 P liH
PUBLIC NOTICE BY THE ,
'- t UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS
. TUESDAY, SEPTEMBER 9, 1986 i'
, The contents of this. stenographic transcript of the l
proceedings of the United States Nuclear. Regulatory Commission's Advisory Committee, on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of ,
1 the discussions recorded at the meeting held on the above date. '
~~
No member of the ACRS Staff and no parti'cipant at
() this meeting accepts any responsibility for errors or inaccuracies of statement or data contained-in this l- transcript.
1 t E
1
$h N[-
l Vg, l '
(-h -06000101 1 marysimons 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION '
3 1..
-4 ADVISORY COMMITTEE ON REACTOR SAFEiUARDS 9:
5
'3UB. COMMITTEE ON. DECAY HEAT REMOVAL SYSTEMS 6
7
( -
a3 8 Nuclear Regulatory Commission
- Q T ,' . . 9 Room 1.046 10 1717 H Street, N.W. r s .
11- Wshington, D.C.
12 Tuesday, September 9, 1986
~13 The subcommittee convened, pursuant to notice, 14 at 1:00 p.m., David Ward, Chairman _of the Subc'ommittee, 15 presiding.
16 ACRS MEMBERS PRESENT:
'Y 17 D. WARD, Chairman
._) 18 J. EBERSOLE
's , . 19 C. MICHELSON 20 ACRS CONSULTANTS PRESENT:
y 21 I. CATTON 22 P. DAVIS v ..
~
t 23 COGNIZANT ACRS STAFF ENGINEER:
24 P. BOEHNERT
)
8 ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-33G6646
.. -. -.., . . -...- - - .... . . = . -. - .. ..~.. . . . . - . . . . . - . .-.
i' 2
O 06000101 v1e-
.1 NRC STAFF AND INDUSTRY PRESENTERS:
3 ..
i 2 W. MINNERS 3 S. DIAB
) '4 B. DENNING 5 M. WILLIAMS 6 F. MANNING
.' 7 J. CALLAN l 8 R. HERNAN '
i l'
.' 9' J. ENOS i
l 10 1
l 11 12 13
-14 t 15 *
! 16 i
l 17 i 18
!. ~19 f'
20 t
i i 21 i
, 22 23 l 24' i
l 25
\
i ACE-FEDERAL REPORTERS, INC, 202-347-3700 Nationwide Coverage , 800-336-6646
- ....,..-._-,.-...----_.,._.____,__....._._._...__..~.-.__<_~,..._._._.u__ _---._c
f'-
k .
06000101. 3
] marysimons j 1 PROCEEDILNGS
.2 -MR. WARD: The. meeting will come to order.
l 3 This is a meeting of the Advisory Committee on Reactor 4 Safeguards Subcommittee on Decay Heat Removal Systems. I 5 am David' Ward, the subcommittee chairman. Other-ACRS 6 members in attendance are Mr. EbersoleJand Mr. Michelson.
7 Also in attendance are ACRS consultants Mr. Catton and 4
8 Mr.-Davis.
9 -The purpose of thh meeting is to review NRR's 10 Action Plan to address concerns with the reliability of 11 certain plants' auxiliary feedwater systems. Paul-Boehnert 12 -is the cognizant ACRS staff member for this meeting.
.- -13 The' rules for participation in today's meeting
.4 1 have been announced as part of the notice of the meeting 15 previously published in the Federal Register on August 22 16 and September 4, 1986. A-transcript of.the. meeting-is 17 being-kept and will be made available, as stated in,the 18 Federal Register notice. I request that each speaker first 19 identify herself or himself and speak with sufficient 20 clarity and volume so that she or he can be readily heard.
21 We have received no written comments from 22 members of the public,.nor have we received requests for 23 time to make oral statements from members of the public.
24 Let me remind the subcommittee of what we are 25 about. Late last year the committee focused some of its I
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
_, . , . _ _ _ . . . _ , . _ , , ~ _ , _ . _ . . - _ . - . . _ . - . . , . . - . _ _ - , . , _ _ , , _ .
ps, 06000101 4 j j marysimons v
1 concern about auxiliary feedwater systems, and of course 2 the staff had been concerned about the reliability of these 3 systems, particularly at certain plants, for some time.
4 In December the committee wrote a letter to 5 Mr. Dircks, expressing our concern that there had not yet 6 been a firm plan and schedule developed for resolution of 7 the issue. We also expressed the concern at that time that 8 auxiliary feedwater system operational experience was not 9 being adequately analyzed, that there was a large body of 10 experience but that it wasn't being mined for useful 11 information.
12 In March we had a brief. meeting and report from
/~
\_3/ 13 the staff, and NRR has indeed seemed to be focusing in on a 14 more explicit program. Since then, however, we understand 15 that there have been some changes in the approach they plan 16 to take, and that is what we are going to hear about 17 today. I don't know how well developed the present 18 approach is. We haven't received any written material-on 19 it, but we will hear about it today.
l 20 If what we see or hear gives us cause for 21 concern, we may want to ask the staff, or whomever, to come
- 22 in and talk to the full committee later in the week and ask i
23 the full committee to make any suitable recommendations as-24 a result of that.
l 25 Carl or Jesse, do you have anything you would O
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6 l
l
-l 06000101 5 '
(Jl'marysimons u
1 .like to comment on at this. time? J 2 MR. EBERSOLE: I would just like t'o make an
- 3. observation. From deep and dark history I always~ carry the 4 remembrance that the feedwater system in a PWR~was not 5 recognized until perhaps early 1970 as a critical saf'ety 6 system. There were no answers to the questions of what 7 happened when you lose force convection insofar as
.8 feedwater to the PWR, what happens to the heat, where.does 9 it get out?
10 The industry should carry this as a burden and 11 never forget it. They didn't recognize the critical nature ;
12 of this system even though it is disadvantageous 1y coupled 13 to a primary loop which then must transfer the heat to this 14 secondary system, which was ill provided in history and:not.
15 even recognize it as anything other than a commercial set 16 of equipment in items until it was forced to do so with the 17 advent of TMI-2. We mast all carry that as a burden and do-j' 18 'whatever it takes to correct it, if we can, with this kind 19 of system. That is all I have to say.
4 20 MR. WARD: Thank you, Jesse.
21 Mr. Diab, I guess you are going to start off l 22 with the staff presentation?
i-23 MR. DIAB: Yes, sir.
24 MR. WARD: Let me ask you, Sammy, would you 25 remind us -- if you are not going to -- I ask if you will L
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 8%33&6M6
- , - - - . . - . - -.., ---- . . - - - - - . , , - . - - ~ . - , , - - - . - . - - - - - , , , , .
06000101 6 (s/;- marysimons 1 remind us of how this issue evolved and particularly how it 2 focused, on the how and why it's focused on the particular 3 plant.
4 MR. DIAB: My name is Sammy Diab. I work for 5 the~ Division of. Safety Review and Oversight, and I am Desk 6 Manager for the Generic Issue 124, which is auxiliary 7 feedwater systems reliability .
8 If I can start by responding to your 9 observation about how this issue keyed down on those seven 10 plants that we have on the list. Since right after TMI, 11 the staff went out and asked that. reliability analyses be
! '12 conducted for'the auxiliary feedwater systems for all the
('h
(/ 13 PWRs. The reliability analyses were done on a simplistic 14 approach, basically to have a set of numbers that can be 15 compared against each other for the whole population of-16 plants.
17 The reliability estimates for the auxiliary 18 feedwater' systems were then compiled and compared in a 19 couple of NUREGs, NUREG-0611 and NUREG-0635, for i
20 Westinghouse and CE plants.
l 21 For B&W plants they were done a little bit 22 differently. Like the individual licensees, they had their 23 analyses, reliability analyses, submitted to the staff, 24 which the staff, in turn, reviewed and issued the CRs, 25 also, keeping in mind the same general requirements that i
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
06000101 7 marysimons
~(
'I were observed in 0611 and 0635.
l-l 2- Those reliability analyses were basically
~
3 hardware-oriented, meaning design-oriented. ~ They didn't i
4- really'get-very deep into how different plants were run and-5 how some plants had more failure rates than others. As I 6 said earlier, the~ objective is to be able to. compare the' l
7 plants, the hardware.
8 As a r'esult of these reliability analyses,
~
j 9 several plants.were identified as falling short. There
- 10 were basically three reliability regions, . ranging from 11 ten to the minus two to ten to the minus three, then from 12 ten to the minus -hree to ten to the minus fo'ur and then 13 from ten to the minus four to ten to the minus five i
14 unavailability per demand for an aux- feedwater system .to --
15 16 MR. EBERSOLE: Was that associated with'a 17 challenge frequency?
18 MR. DIAB: No. - That is not associated with the l
19 challenge frequency.
20 MR. EBERSOLE: It doesn't mean much, unless you 21 have a challenge frequency, to' talk about reliability.
22 MR. DIAB: Once it's called upon, the i
23 probability or the chance that an aux feedwater system'will i
24 fail, given certain failure rates for pumps and valves, et
! 25 cetera, that was assumed as an aux feedwater system
+
l ACE-FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coverage 800-33G6646
4 06000101 8
' ()'marysimons 1 reliability number, given that you have a demand, whether 2 that demand be a loss of main feed.or loss of offsite power i
3 or anything else. -So it's just a' number assigned or a 4
i 4 number estimated for the aux feedwater system to function 5 properly once it's called upon to do so.
i 6- Some plants made an effort at the time to 7 upgrade their' systems by adding different pieces of 8 hardware -- pumps, valves -- changing piping, things like 9 that.
f 10 MR. EBERSOLE: As'I recall, there were ten of 11 .them that had no seismic competence. Is that correct?
12 MR. DIAB: I am sorry. "No seismic" what?
13 MR. EBERSOLE: No evident seismic competence, 14 although they had the intrinsic built-in strength of piping 15 systems and. equipment.
16 MR. DIAB: I don't really know.
MR. EBERSOLE: It was ten.
17 l -
18 MR. DIAB: Okay.
l I
19 MR. EBERSOLE: I think none of them were QA,
(
i 20' were given the privilege of tests, pedigree, et cetera, of 21 a safety-grade system. ,
22 MR. DIAB: Well, at the time, I don't think 23 many plants considered their aux feedwater systems to be 24 safety-grade.
l 25 MR..EBERSOLE: That is precisely what I am ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-33 4 6646
06000101 9 marysimons
{
1 saying.
2 MR. DIAB: Right. That was brought in focus
~
3- later on. People considered that the aux feeds were just
~
~
4 that: auxiliary systems that convenienced them.
~
5 Well, that list was narrowed down eventually to 6 the seven'that we have today. Now,.those seven plants, . in
- 7. the last several years since TMI, have done modifications 8 and upgrades to various degrees in response to staff 9 initiatives. So today if we were to go out and see a 10 plant, it's going to be different than that plant that was 11 analyzed at the TMI period.
12 MR. EBERSOLE: your statement that those were 13 considered as systems of convenience leads again to-the 14 question which follows: Where does the heat go from the 15 PWR when they don't have a secondary system to take it 16 away? At that time, there was no consideration at large 17 for feed grade or whatever, so we had an astonishingly 18 invisible problem, which was astonishingly neglected for 19 years and years and years. We can't escape that, and we 20 must be careful'in what we do now and keep that in mind.
21 MR. DIAB: Oh, of course.
22 MR. WARD: Sam, were these seven plants
'3 2 selected because they fell in this lowest range?
24 MR. DIAB: That's correct.
25 MR. WARD: They are also the only plants that ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
l 06000101 10
- t()marysimons i have only two aux feed pumps except for, I guess, three 2 multiunit plants where there are cross-connections 3 available. So was the decision made on the latter or on 4 the analysis that showed they really had low reliability, 5 or are those the'same thing?
6 MR. DIAB: It turns out, as I mentioned 7 earlier, those reliability analyses were hardware-oriented, 8 meaning that they looked at the number of pieces of
. 9 equipment that you have and how it's tied together. Once j
10 you assigned the same failure rates to a pump, having two 11 pumps or having three pumps will give you better 12 availability than having two pumps. That was a contributor 4
13 to singling out the two units or the two feedwater pump, 14 aux feedwater pump plants to be of lower reliability.
15 The plants that go out to the high-reliability 16 range were the ones that were later identified as deficient 17 as far as the reliability goes.
18 MR. EBERSOLE: Did you find a variability 19 between the turbine-driven versus the electrics?
20 MR. DIAB: I am sorry, I didn't hear that, i 21 MR. EBERSOLE: For a two-pump configuration, 22 did you find that the turbine-driven pumps were better or 23 worse than the electrics?
24 MR. DIAB: I think this will be addressed later 4
25 by the AEOD staff.
1 ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
7, 06000101 11 i i marysimons a
1 MR. EBERSOLE: Well, you know, Davis-Besse is a 2 case in point.
3 MR. DIAB: The turbine-driven pumps have always 4 had more problems than motor-driven ~ pumps.
5 MR. EBERSOLE: Yes. My impression is that, 6 .yes.
7 MR. DIAB: I would like to make a comment at 8 this point. I am here to introduce the issue a little bit, 9 and then we will be'able to get AEOD to make their 10 presentations, followed by I&E's presentation, and then we 11 will come later to answer detailed questions and to also 12 discuss the rest of the program. But I hope I answered n} - 13 your question.
14 MR. WARD: Yes. I think that's fine.
15 MR. DIAB: To give a historical perspective.
16 MR. WARD: There is one other type of similar 17 question I would like to have you clarify for us before wo 18 start of here, something which Mr. Ebersole alluded to.
19 For some reason, we have never been able to get a good -- I 20 haven't, anyway -- been able to get a good, solid answer or 21 understanding of this. That is that the dependency on aux 22 feed, whether there is the same dependency on aux feed 23 given, let's say, a reactor turbine trip, in all plants.
24 Now, maybe you can just disabuse me of this
_ 25 right off and we can forget about it. But my perception is V
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
f 12 7 . 06000101 marysimons
-(
1 that there is a type of plant -- let's say, Type A -- where 2 there is a reactor trip and a turbine trip, the feedwater 3 for removal of decay heat in these, you can go to hot 4 shutdown by running down on the main feedwater pumps and 5 operate stably at hot shutdown on the main feedwater 6 pumps. There is another type of plant we will call Type B 7 where that can't be done and where to go to stable hot 8 shutdown, the auxiliary feed pumps have to pick up the 9 load.
10 Now, it would seem then that a lesser 11 reliability might be necessary for Type A plants than for 12 Type B~ plants. Could you tell me whether there really is a 13 Type A and a Type B plant; and if there is, do you anywhere 14 in your program recognize a difference here insofar as 15 requirements for aux feed?
16 MR. DIAB: I think you are alluding to,-let's 17 say, the number of demands on a system.
18 MR. WARD: Right.
19 MR. DIAB: In one plant it may not be called 20 upon as often as in the other.
21 MR. WARD: Right.
22 MR. DIAB: As far as Type A and Type B, I think 23 all plants eventually will have -- let me say most plants 24 eventually -- will want to go to a smaller source of water 25 than the main feed pumps because they are just too big and ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 804336-6M6
06000101 13 f ) marysimons 1 you can run them on recirculation, partial recirculation
-2 for so long, but after the reactor is tripped and the decay 3 heat is reduced, you don't want to overcool by overheating 4 and also you don't really need to remove as much heat as 5 the main feed pumps are capable of, even on the bypass 6 control valves, which are smaller valves.
7 MR. WARD: No. I understand that. But I am 8 just trying to differentiate whether there are some plant 9 units that could get by for 30 minutes or for one hour or 10 for eight hours without having aux feed available at all.
11 MR. DIAB: yes. Some plants will have so-12 called starter pumps. They rely on the starter pump for 13 startup and shutdown as far as small amount of water 14 supply.
15 MR. WARD: Well, are there any plants that
- 16 could rely on their main feed pumps for an extended period l
17 after a reactor trip?
18 MR. DIAB: I am not' really sure, but I doubt it 19 because, as I say, very big pumps -- and you don't really 20 need all the water that they supply, so you have to 21 recirculate most of the water and only take the little bit 22 that you need. That sounds to be an inefficient way of 23 running those big, expensive pumps.
21 MR. WARD: yes. I am not talking about 25 effjciency.
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
_ . . . _ . _ ~ . _ . _ _ _ . - _ _ . - - , _ _ _ _ _ _ _ _ - _ _ . _ _ _ _ _ - . . - . _ _
06000101 14
'[v) marysimons 1 MR. EBERSOLE: No. That's right. I think that 2 I have.the same question. Are there any systems that are 3 capable of being run back to the level of flow that's 4 required just to make up the decay heat loss or decay heat 5 removal? Can you run back a main feedwater that far?
6 MR. DIAB: I don't see why not.
7 MR. EBERSOLE: Yes, I can see a lot of 8 reasons. It depends on the kind of throttling valve you're 9 using and whether or not they.can be throttled to that 10 extent.
11 MR. DIAB: That's correct.
12 MR. EBERSOLE: If they have a bypass 13 arrangement, whether it's big enough, because you know 14 there's a minimum flow requirement on the pump where it 15 can't operate at all.
16 MR. DIAB: That's right.
17 MR. EBERSOLE: These are very large pumps and f
18 it's a very large flow.
19 MR. DIAB: That's right. The minimum flow, you 20 can get that through the recirculation as far as the pump 21 is concerned.
22 MR. EBERSOLE: But what we're asking is is that 23 the way it's arranged? Is it arranged such that a plant i
24 can operate off of its main feedwater system if it so 25 wishes even when it's only removing decay heat?
ACE. FEDERAL REPORTERS, INC.
202-347 370) Nationwide Coverage 8W) 3346M6
06000101 15
/
\.)1 marysimons 1 MR. DIAB: Some plants are designed that way 2 and some plants are designed differently. Some plants have 3 the. bypass control valves which will control -- you will be 4 at throttle down and control the flow of the main feed to 5 the steam generators to a certain number of g.p.m.,
6 hundreds, couple of hundred, something like that.
7 MR. WARD: But, you see, Sammy, if that's so, 8 it would seem to me that that's a big safety plus for that.
9 10 MR. DIAB: Yes, it is.
11 MR. WARD: I don't understand why that isn't a 12 conspicuous part of the analysis that's being made about n
k-) 13 aux feed concerns.
14 MR. DIAB: I guess what we're trying to focus 15 on in this study is we picked a couple of events, okay, a 16 couple or three cases, that we used and then we tried to 17 measure all the plants at least to consolidate the problem c
18 a little bit. We talk about the loss of main feed. We 19 don't talk about the partial loss of main feed; we talk i
- 20 about the total loss of main feed. So if you have no main 21 feed, that's not going to help you if you throttle it, f 22 MR. WARD
- That's right. But that's only one 23 class of accident.
24 MR. DIAB: Yes. The loss of offsite power --
l . 25 MR. WARD: Another challenge.
l ACE-FEDERAL REPORTERS, INC.
! 202 347 3700 Nationwide Coverage 800 336 6646 i
06000101 16
( j. marysimons v
1 MR. DIAB: Right. The loss of offsite power is 2 also going to lead to the loss of main feed.
3 MR. EBERSOLE: May I ask a question? We have 4 not gotten into two other things I think we should 5 consider: The main drives: there is a bewildering 6 variety, determined usually by the mogul in charge of 7 design, I guess --
8 MR. WARD: Wait, Jesse, is this on the same 9 question?
10 MR. EBERSOLE: Same question.
11 MR. WARD: Okay.
12 MR. EBERSOLE: And we haven't mentioned the O
k../
13 characteristic of the drive system. It can be an electric 14 motor, straight drive, induction, full flow.
15 Mh. DIAB: That's correct.
16 MR. EBERSOLE: It can be one where they slip 17 the hydraulic fluids from coupling, in which you have 18 infinite variability down to whatever you want, to zero 19 coupling. In the case of a turbine, certainly the steam 20 emission system to the turbine can be modulated to whatever 21 output you want on the discharge to a zero bypass.
22 There is a host of things on the drive side 23 that needs to be looked at, and all of these things, I 24 think, need to be focused on what is the best
,- 25 configuration, if we can do it in this bewildering system V
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationside coserage 800-334 % *6
06000101 17
()marysimons
-1 that we have where everybody has-his personal choice.
2 MR. DIAB: Okay. Assuming that you have a loss 3 of main feed function, or assuming that you have a loss of l 4 offsite power and then you have a steam-driven main feed 5 pumps --
6 MR. EBERSOLE: Then you can modulate steam flow 7 to it.
8 MR. DIAB: Okay. You know, you're going to run 9 out of steam --
10 MR. EBERSOLE: Unfortunately, however, the main 11 steam systems, unless they can exhaust to atmosphere off 12 the turbine side, on the exhaust side of the turbine-driven 13 pumps, if they depended on the condensers, then those pumps 14 are dead, too --
15 MR. DIAB: That is correct.
16 MR. EBERSOLE: -- if you have loss of AC 17 power. But all these things, you see, they just get put 18 together like Mulligan stew, and this little integral 19 consideration of how they ought to be packaged, that's 20 characteristic of our evolution of this business.
21 MR. DIAB: That's correct.
22 MR. EBERSOLE: I am sympathetic to your 23 difficulties. We are trying to unsort the Mulligan stew.
24 MR. DIAB: I think, in order to get over some l 25 of these difficulties, you know, we tend to try to define
(~)
i
! ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-334 6646
06000101 18
,,.) marysimons (G
1 the problem or at least limit the problem. The problem is 2 large and if we can get into these details, I am sure we 3 will do very well. But it will take too many resources at 4 this point.
$ MR. EBERSOLE: It is a global observation.that 6 you make. It's. complicated: it's difficult. It shouldn't 7 be that way, but it is.
8 MR. WARD: Hold it. Let me go back. I don't 9 feel that -- maybe you gave me an answer on this, and the 10 reason I am pressing it, I guess, is that I have been II unable to get this answered to my satisfaction earlier. I 12 guess what you have said, you've acknowledged there there 13 may indeed be a Type A plant out there or some units may be 14 Type A, but when you look at all the chc11enges, the i
15 l demands there might be on an auxiliary feedwater systems, I 16 mean, you've got loss -- main feed pump trip, loss of AC 17 l power, and then reactor trip, let's say. I guess what 1
18 l you're saying is that that latter category represents a i
19 relatively small fraction of the total populaticn and so 20 you are, in essence, kind of 10noring the Type A/ Type B 21 difference in this analysis?
22 MR. DIAB: No, I am not really ignoring it.
23 Okay. The sert of Type A and Type B, let's talk about the 24 Type B that has more challenges or you need the aux feed.
, 25 If the auxiliary feedwater systems or the emergency ACE-FEDERAL REPORTERS, INC.
202- m.nm Nationwide Cmcrage f4MM &4
06000101 19 i( ) marysimons 1 feedwater systems are the only small feed pumps that are 2 available there, of course,.you're going to need them to 3 startup and shutdown, and in any emergency, of course, on 4 top of that.
5 Yes, I mean a system that's called upon so many 6 times, of course, has to be just as reliable as the other 7 systems that are not called upon so many times except that 8 when it's needed in an emergency, it should be available, 9 The fact that it's called upon so many times, I guess, can 10 be an asset as well as a liability, an asset in the sense 11 that it's called upon and you have a lot of information or 12 data that you will get any signal. If something has gone O
\-) 13 wrong, you will know it.
14 MR. EBERSOLE: Pardon me. That makes me think 15 of testing a parachute by jumping out with it yourself 16 instead of throwing a weight out with it attached. If 17 those situations where you have to have it and you can't 18 afford to lose it are your examples for testing it and 19 getting experience on it, then that's the wrong way to go.
20 MR. DIAB: I mean, this is not a single 21 system. We're talking about a redundant and a reverse 22 system that you have two or three things --
23 MR. EBERSOLE: Oh, I know. I am talking about 24 it as an integral system, however. What I am saying, by 25 the way, I think there are incidents which we might call ACE FEDERA1. REPORTERS, INC.
202-347 3700 Nationside Coverage M O MW46
06000101 20
()marysimons 1 benign, that cal) for aux feedwater for a short period of 2 time which then can be picked-up after that short interval 3 of exposure and need by the main feeds which have been 4 adjusted for the new low flows. I think that may be the 5 case in the Westinghouse plants which always have to go 6 back to aux feed; they never throttle down to take away 7' decay heat.
8 On the other hand, GE plants, I believe, 9 currently are designed to ramp down, but they don't have l 10 the benefit of exhaust steam to atmosphere or the other I
11 things you might do to make the turbine-driven pumps more
{
- 12 reliable than they could be. So to a great extent again, 13 you have to take each plant as a plant in point.
14 MR. MINNERS: Well, I think as you see, when 15 Sammy gets into his presentation, that's exactly the 16 approach it will take. I think we've recognized that a
! 17 generic universally applicable approach doesn't work; just i
18 putting a third train on all plants is not the answer to l 19 the problem.
i j 20 MR. DIAB: Right.
21 MR. MINNERS: So as Sammy will go on, I think l 22 he will show you that we are probably thinking more along 23 the same trains.
24 MR. DIAB: I think Warren put it very well.
l l 25 (Slide]
l I
' ACE. FEDERAL RiiPORTERS, INC.
202-347 37t0 Nationwide Cmerage AOLD6646
06000101 21 (v_,) marysimons 1 Well, this is apparently a status progress 2 report that we are trying to present to you here today.
3 Basically, we are presenting to you a modified resolution 4 approach. That modified resolution approach is different 5 than what we presented to you here in March. We are going 6 to discuss this later on.
7 We have a modified resolution plan: How do we 8 plan to go about resolving this issue for t,he population of 9 plants?
10 MR. EBERSOLE: One more minor question. There 11 g is a particular class of plants. They are the Combustion 12 Engineering /Palo Verde type in which there is a totality of 13 dependence on aux feedwater to take away heat. you know, 14 they don't have any PORVs. And a zero plane to heat 15 removal include the direct process. In your orientation 16 toward reliability, are you looking at that as the worst 17 case?
18 MR. MINNERS: No. That issue will be addressed 19 again. The effort of resolution of A-45, that is a 20 separate issue which is going to be dealt with after A-45 21 is decided.
22 MR. EBERSOLE: But inevitably, that's the 23 highest demand problem for aux feedwater where there is no 24 other recourse.
25 MR. DIAB: yes.
ACE-FEDERAL REPORTERS. INC.
202 347.);m Nanon.iae cmerage un3356646
06000101 22
( ) marysimons 1 MR. EBERSOLE: Whatever you say.
2 MR. DIAB: That's correct. The auxiliary 3 feedwater issue, we take it as just the auxiliary feedwater 4 issue for the time being. I know the Standard Review Plan.
5 allown for some credit for the feed-and-bleed. We are 6 trying to ensure ourselves that the auxiliary feedwater 7 systems are sufficiently reliable; therefore, if a feed-an:'
a bleed exists, of course, that is an addition.
9 After that, I'will get into the program, too, 10 and then at the end will present to you the schedule for t
11 resolution we are trying to work towards.
12 (Slide) 13 Okay, the first question: Why not stick with
. 14 the aid proposed resolution method approach? Well, after i
- 15 we presented our program here to you earlier this year, we i
l 16 went out for comment with the staff, and we received l
17 several negative comments, meaning that we didn't have a 18 universal agreement in NRR'on the approach we had taken or 19 we planned to take.
j 20 One other reason for not sticking to the old 21 proposed resolution, we think it is a very lengthy process, 22 "very lengthy" meaning that we first have to go through the 23 . process of issuing the generic letter to the affected 24 plants, where the licensees after that will have to conduct i 25 their aux.feedwater system for aux feedwater reliability l
I I
i ACE. FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Cmerage M10-33MM4 t
l 06000101 23
( ) marysimons 1 analysis, staff will, in turn, have to review it, and 2 followed by an SER.
3 Overall, it's a very lengthy process, and we thought that we might be able to cut it much shorter than 4l 5 that, while achieving very similar results.
6 [ Slide]
7, Okay. Now, the next question is: why the 8 modified resolution approach? The modified resolution 9l approach is a short-term concentrated review effort that we I
lo g planned to take for each one of the affected plants and we 3
11 3 will pull that off by a report, a findings report, specific 5
12 6 to every one of the plants.
rm k- 13 The review effort will benefit from an ongoing E
14 inspection program undertaken by the Office of Inspection 5
and Enforcement, called SSFI, the safety system functionni 15 [
16 inspection program. I believe the committee has heard 17 h something about this, and you are going to hear some more i
18 about it today.
19 It will also benefit from the Prairie Island 20 aux foodwater reliability analysis. The Prairie Island t
21 plants have submitted to us reliability analysis for their 22 lauxiliaryfeedwatersystem, and it is still under review.
! 23 It looks very promising, though.
24 ,
It will also benefit from other industry and 25 i staff efforts that are currently underway, i
ACE lEDERAL REPORTERS, INC.
202 1U-1*m Nationwide Coserage Mn34tM6
_ . _ - . _ . - - , _ .._ ~_ _ , _ _ . . _ . . _ . . _ _ _ _ _ . . . _ _ _ _ _ _ . _ . - . , _ _ _ _ . , ~ _ , . - _ _ _ , . . . _ _ , ,
e 4-r 06000101' 24 '
, ( ) marysimons 1 1 MR. EBERSOLE: In the SSFI, the word i
a 1 2 " inspection" always causes a hangup for me because it
- 3 implies a comparative analysis of the equipment to the 4 drawings
- Does it follow the logic of the drawings' the ,
5 intent, the operational program and so forth; in short, 6 it's a comparative analysis against a preestablished notion 7 that the design was correct. I can't get rid of that i
- a connotation of inspection. Tell me it's larger than that 9 here.
10 MR. DIAB: As far as I understand it --
11 MR. EBERSOLE: Did'I&E tell you that?
12 MR. DIAB: Yes.
( 13 MR. EBERSOLE: Well, you are telling me in i
i 14 their behalf, I guess.
I 15 MR. MINNERS: I do not think we are. I think i 16 I&E ought to speak for themselves.
17 MR. DIAB: We will benefit from whatever their i
4 18 findings are.
19 MR. EBERSOLE: Do you agree with that 20 connotation?
l 21 MR. MINNERS: No.
22 MR. DIAB I am not that knowledgeable with the I 23 I&E operations. That may be correct. I am not really 1
24 sure.
l i 25 MR. EBERSOLE: It's like comparing a heat pump s
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Cmcrage 85336-6M6
- _ _ . . - - - ~ . . , _ . _ _ _ . _ . _ _ _ . _ , - _ _ . _ . _ . . _ , _ - . _ . _ . - _ , _ . , , _ _ _ _ _ . . _ _ _ - - _ _ - . _ . _
06000101 25 (fmarysimons 1 to a --
2 MR. DIAB: Well, they will address the 3 committee and the subcommittee in another half an hour or 4 so. I understand it's a little bit more than that.
5 MR. EBERSOLE: I should hope so.
6 MR. DIAB: Yes. But we will find out.
7 [ Slide]
8 This is the modified resolution plan, rather 9 than approach. We are really talking about a two-tiered 10 approach here. We are going to address the seven plants 11 first. We're going after the st.en plants, and then after 12 we're done with the seven plants, after we write our
/ 13 reports about the seven plants, we will be in a much better 14 position to judge what next. But this is, apparently, on 15 our agenda.
16 MR. WARD: So there isn't any definite plan 17 right now to look at the rest of the plants; you're going l
, 18 to look at the seven and come up with some program?
! 19 MR. DIAB: Yes'. Right. Not right now. We
. 20 have, as I will discuss later on today, a program that i
21 covers approximately about seven months.
22 MR. WARD: Pardon?
i l 23 MR. DIAB: A seven-month period that we we plan 24 to spend on the seven plants. At the conclusion of that I
i 25 period of time, we will be able to judge better what it is i
I ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 80)-33M6:6
- = ..
26 marysimons
()06000101 1 we need to do next, whether it's a generic letter for the 2 rest of the plants, whether it's some IEE information 3 notice, whether it's some other means.
4 MR. WARD: Well, surely, the resolution or l
5 pending resolution of A-45 is going to influence what you 6 do there; right?
l 7 MR. DIAB: I an-sure it will to some degree,
~
8 yes, 9 MR. WARD: Well, I should think that they would-10 be very much intertwined.
i i
11 MR. DIAB: Well, you can say that our efforts 12 as far as auxiliary feedwater systems are concerned can be j v 13 used as a subset of A-45. In other words, if we can assure 14 ourselves and the staff that the auxiliary feedwater 15 systems after certain things are said and done are i
16 sufficiently reliabic, of course, that's one brick that we 17 can use for A-45.
! 18 MR. WARD: Yes. Let's see now, where are you i
l 19 organizationally on 124 and A-45.
20 MR. MINNERS: Same branch.
21 MR. WARD: So we can assume there would be a 22 certain amount of coordination?
23 MR. MINNERS: Of course. Our usual good i
24 coordination.
'l 25 [ Laughter]
! (1) i ACE. FEDERAL REPORTERS, INC.
! 202 147 3700 Nationside Cmerage MX) 3%6646
_ , _ . - - _ ., _ _ _ _ . . _ . . . - , _ . , _ _ _ _ . . _ , _ _ . . _ _ _ _ _ _ . _ _ ~ . _ , . . _ . , _ , , . _ _ _ , . , - . , , , . . _ _ . , . _ _ _ . _ . - - _
06000101 27 marysimons
[: 1 MR. DIAB: If there are no more questions at
! 2 this point, I will give the floor to the next speaker.
i 3 4
- 5 i
1 6 1.
i 7
(
(
k
! 9 i
i 10 1
l j 11 l 12 13 14 I
15
)
i 16 1 17 1
la
[
! 19 i
20 f
l 21 l 22 i
! 23 1
24 I
l O
Am FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 33M646
4
\
[, t s 06000202 , 28
()marysimons ,
1 MR. DENNING: Good afternoon. My name is Bob 2 Denning. I am Section Chief of the Program Development 3 Section, AEOD. I just wanted to make a few brief commento 4 to put our presentation into a context.
5 We came dosn, as Mr. Ward mentioned, back in 6 December of '85 and gave some operational experience based 7 on 1984 data on auxiliary feedwater systems and t'iat was .
8 across all plants within '84.
9 Then following that presentation we continued 10 to work in coordination with NRR in support their effort on 11 Issue 124. Through mutual discussion our scope was 12 narrowed from all plants, all PWRs,-to the two-train PWRs, n<
(- 13 and that is what we will be addressing in our discussion 14 today.
1 \
15 ' Our scope also extended to looking into the 16 existing data sources available to the staff which shakes 17 out to be licensee event reporting and NPRDS.
18 By way of preliminary products in support of 19 NRR we produced a report in March of '86 that was based on 20 a look at the two-train plants using LER data and then just 21 recently we produced a more extensive report integrating 22 NPRDS data.
23 With that, I would like to turn the floor over
\
24 to Fred Manning who is the engineer who performed the work 25 on this task.
ACE-FEDERAL. REPORTERS, INC.
202-34747(N) Nationaide Cos erage 8trK3W6M6
' I 00000202 29
(. -_') marysimons I c:v' MR. MANNING: As introduced, my name is Fred 1
2 '
Manning and I am with-AEOD. I am a systems engineer with 3 AEOD.
4 s (Slide.)
\' 5 I am going to describe the work on the aux 6 feedwater system that was done by AEOD at the request of i 7 NRR in support of the generic issue 124.
8 The plan or the effort that was done by AEOD <
9 involves, as was just pointed out, the analysis of aux 10 feedwater component and train level failure data to t
11 determine potentially significant incidents and elements 12 and events and then to try to make inferences from these O
(__/ 13 incidents regarding auxiliary feedwater system 14 reliability. That was our plan then.
15 The scope of this effort was to use LERs and 1
16 NPRDS data and use it for the time interval from 1981 17 through 1985.
s 18 I would like to emphasize though that we didn't 19 go any further than what was available on the record for 20 NPRDS and LERs. In other words, we didn't trace further 21 back and make sure it was a complete source of data, nor in 22 fact even verify that it was a complete set for each plant.
23 (Slide.)
24 This analysis ---
f
, 25 MR. MICHELSON: Excuse me. I assune that you v
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage Rn33M646
,~ 06000202 30
( ) marysimons v
1 are certain that this is.not a complete set.
2 MR. MANNING: I think that is right what you 3 just said.
4 MR. MANNING: Now the question is how 5 incomplete a set is it. Do you have any feel about what 6 fraction of the likely number of failures and so forth we 7 are really looking at when we look at NPRDS and LER?
8 MR. MANNING: I have no feel for that, no.
9 MR. MINNERS: I have seen reports somewhere 10 where people compared LERs with NPRDS, but I don't remember 11 the ---
12 MR. MICHELSON: What I am trying to get a feel
(~'\
(-) 13 for though is out of the total population of failures how 14 many do we ever likely pick up from LER plus NPRDS, 20 15 percent or 50 percent?
16 MR. DENNING: I think the answer to your 17 question.on where the stuff is coming from and what 18 percentage depends on the time frame. In '81 through '83 19 we are talking about the bulk of our information coming 20 from LER information, and at that point we got reports out 21 on training level dysfunctions and faults for detecting 22 specifications before occurrences.
23 In '84 we shifted gears and removed the train 24 level items from the LERs and set the LER reporting Isvel 25 at the system level, the system loss of 7_
V ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336 6646
06000202 31
- qp_j marysimons 1 function level and the burden for the-train level and 2 component level then is carried by NPRDS.
3 So the issue then becomes the completeness of 4 NPRDS training and AEOD does a semi-annual evaluation of 5 NPRDS. Our latest assessments show that for comparing 6 failures that we find in LERs against what is found in 7 NPRDS after a reasonable amount of time, three or four B months, we see perhaps 75 percent matching, 75 to 80 9 percent versus an audit sample that we pulled from LERs.
10 MR. MICHELSON: That means that you are seeing 4
11 in NPRDS most of what is reported in.LERs at a system-4 12 level. Now how about failures that were not reported as 13 .LERs, do you have any feeling for how many of:those are 14J also escaping into NPRDS7 Since the other way I know words 15 pretty well because they kind of ask on the LER form what.
16 is reported in the NPRDS. -
17 MR. DENNING: The view in that sampling has 18 heen that there'is a presumption that there is nothing 19 special about the report that I find in an LER. That is l 20 'j ust a source of. knowledge about a failure at the plant 21 that ought to be in the NPRDS.
. 22 Now if I get rid of.that assumption and say 23 well the plant is going to put in those failures and put
, 24 their effort there and give that special treatment as
'25 opposed to it being a random kind of thing, then my 4
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
- 06000202 32 (s)marysimons 1 yardstick is not as good.
2 MR. MICHELSON: In reality isn't that exactly 3 what is happening? Isn't INPO checking.to see that the 4 NPRDS reported failures are showing up in their NPRDS 5 system?
6 MR. DENNING: That's true.
7 MR. MICHELSON: So I expect that that is a very 8 good measure at all, but I don't know, you know, I don't 9 have any feeling for what we are still missing. Maybe it 1'O isn't very much, but I certainly have got the feeling that 11 you can give me.
12 MR. DENNING: Well, I think it is a concern A
> (_) 13 that is a concern that is shared by the people that are 14 doing this evaluation.
15 MR. MICHELSON: Have you ever looked back at 16 the '81 '83 frame to see how much was reported then and 17 then try to somehow assume that about the same failure.
i l 18 rates are continuing to occur and see how the NPRDS data i~
l 19 today compares with what was reported back then. Is there 1
! 20 some way of bncking into it?
21- MR. DENNING: To-look at LER data from the '81 l
l' 22 time frame?
23 MR. MICHELSON: Sure, yes, and assume that is 24 typical of what would be happening.today and see what t 25 fraction of that sort of failure rate that is occurring as
("h s_)
l ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6M6
06000202 33
( ) marysimons 1 reported by NPRDS.
2 MR. DENNING: I would no, we haven't done that.
3 MR. MICHELSON: That would be one way of 4 backing into the question. I don't know how else you would 5 back into it.
6 MR. DENNING: Well, we are searching for a new 7 yardstick of information about failures at the plants-that 8 the staff has access to and there is not a long list of 9 such things outside of the LER system.
10' MR. MICHELSON: The other way of backing into 11 it of course is to take a particular plant as an example 12 and look in depth at the actual failures by looking at 13 maintenance records and so forth on that system'for a 14 period of one year and then comparing with what was 15 reported in the NPRDS that year.
16 MR. DENNING: Yes, we could do that.
17 MR. MICHELSON: LAs I understand it, some plants 18' are only giving NPRDS a fairly small percentage of the 19 failures now and some plants are giving a high percentage 20 of their failures to NPRDS, and I don't know how these 21 plants stack up. You could back into it a wee bit that way 22 by looking at the general record on NPRDS for these 23 facilities. Are these big ~ contributors or little 24 contributors?
25 MR. . DENNING: The particular plants we are ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
. - . - . ._. .- . _ _ . - , _ - _ _ _ - , - - ~ , - - - . , . . - - . - -
06000202 34
[d marysimons I talking about?
2 MR. MICHELSON: Yes, the particular ones we are 3 talking about.
4 MR. DENNING: I think generally the rule of 5 thumb from what we.have seen is that the smaller older 6 plants, just like in most other things, are on the low end 7 of the number of reports going into the system.
8 MR. MICHELSON: ~That is generally the case, 9 yes.
10 MR. DENNING: And the four-loop newer plants 11 are on the other end of the spectrum.
12 MR. WARD: As far as the number of failures.
1
-* 13 being reported or the fraction of total failures being 14 reported?
15 MR. DENNING: That is the unknown. All we.know 16 is the numerator again or the denominator.
i 17 MR. WARD: Bob, let me ask you, what would a 18 statistician say about ---
19 MR. DENNING: Something that.couldn't be
- 20 understood ---
l 21 (Laughter.)
22 MR. WARD: No, about the importance of this. I.
t 23 mean obviously in most activities in technology we don't 24 have a hundred percent samples, and we are happy with
. 25 sampling as long as we can have some faith that it is i
i i
ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage SJ0-336-6646 l
06000202 3'5
(-s.}marysimons 1 accurate or random and that we aren't missing classes of 2 things. So what is your pos-ition on that?
3 MR. DENNING: Our position is that at: this 4 point in the life of NPRDS -- when we started out 5 monitoring it, things were so poor and so many of the 6- yields were set on zero that just getting the stuff that 7 would be captured in LERs was a big change in our 8 improvement. Now we have reached the point where they have 9 achieved that, they have met that. test, whether by in the 10 course of doing things per NPRDS reporting or because of 11 some direction they have gotten from somewhere. At any 12 rate, things are to the point.where we see'most of what we s-
/ 13 pick up from LERs is in NPRDS.
14 Now the question is, as I said a while ago, are 15 those LER reported failures-a special class and thereby no 16 longer have something that is representative of NPRDS-17 reporting in general, and our suspicion is that that is 18 probably true in a number of cases where we shouldn't 19 consider that as -- or apply that fraction across the 20 board, across the NPRDS scope.
21 So we have that concern and we are trying to 22 find a more broad based sample to use to make these 23 checks. That is where we are and so far we haven't 24 identified any good, solid alternatives to such things as 25 plant 'Jisits and grabbing maintenance records and so on and ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 36
( marysimons I so forth.
2 MR. MICHELSON: Well,.have you ever tried 3 'something simpler perhaps like just taking a sample of 4 maybe six or seven of these utilities that you believe gave 5 essentially comparable auxiliary feedwater systems, or at 6 least somewhat' comparable, and then compare the NPRDS 7 records of how has utility "A" given and utility "B" and 8 "C" when the assumption is they ought to be all about the 9 same?
10 MR. DENNING: Well, essentially that is what 11 Fred has done I think.
~
12 MR. MICHELSON: Well, no, he has restricted it 13 to things reported in LERs I thought.
14 MR. DENNING: Half of his presentation or most 15 of his presentation includes NPRDS data. So he has looked 16 at a set of plants with roughly the same design.
17 MR. MICHELSON: Well, what I was trying to get 18 a feel for was what fraction of the population do you.think
! 19 you are looking at and it is about the same for all these 20 utilities and is it-representative, and I am not sure that 21 I got any real answer'on that.
22 MR. EBERSOLE: Let me ask,'in the previous 15 23 years, a span of time three times as long as '80 to '85, 24 what, if anything, is know about the performance of.these 25 standard grade commercial systems in that interval?
ACE-FEDERAL REPORTERS, INC.
202-347-3700 . Nationwide Coverage 800-336-6646
06000202' 37 (a,') marysimons 1 MR. MANNING: What standard grade commercial 2 system do you mean, the aux feedwater systems and such?
3 MR. EBERSOLE: Yes.
4 MR. MANNING: I don't know. I haven't looked 5 at that. I have strictly looked at the ---
6 MR. EBERSOLE: You don't know what happened.
7 MR. MANNING: Right, I don't.
8 MR. EBERSOLE: We made it evidently until TMI-9 2.
10 MR. MANNING: I am not sure that.the old data 11 would necessarily be that valuable now if you looked it up 12 and checked it out because of the changes that have been
(~h
(/ 13 made.
14 MR. EBERSOLE: That is about all I can say is 15 we made it.
16 MR. MANNING: That.is true, but even if you
~
17 went back and seriously looked at that, I-am not sure what i 18 you would gain.
19 MR. EBERSOLE: We don't now many times we got 20 close to not making it.
21 MR. MANNING: Probably we did close a lot of
. 22 times, but how you would project that forward, I am not 23 sure.
24 (Slide.)-
25 In doing this study, just using the LERs and ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
r 06000202 38
] ).marysimons 1 the NPRDSs is sort.of difficult in.itself. They are not 2 really directly usable, these sources of data. You have'to 3 massage them and put them into a usable fashion.
!- 4 For each LER and NPRDS event we really had to
~
5 evaluate,'for example, a component or a function that was 6 involved, and we also had to maybe in some cases verify the 7 component that was described in the LER against the system i 8 drawing. In many cases we had to do this to ensure really 9 what it was that was being' talked about. - In other words,
! 10. the event identifier in the event report may not have
.11 clearly defined what it was.
12 Then, lastly, as part of this preliminary work
!O 13 we summarized each of these events into a concise, one-line 14 type description both describing the event and indicating 15 the' component that had failed and the way it had failed.
16 Then we just sorted this data by plants and 17 subsorted it by component and time of failure. That was-18 just to get the data ready.
19 (Slide.)
20 There really were two phases to this-study. In
! 21 phase one we used the LER data alone, just looking at in 22 its context to try to determine some qualitative insights.
23 In other words, repetitive failures, we tried to see what t-24 was failure most often, for example, and if they are 25 bunched or whatever.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationyide Coserage 800-336-6M6
. .- - - _ . _ . _ . . , _ . _ . _ __ _ _ . _ .-- . - . _ _ , _ . - - _ ~ . _ _ _ , . _ _ _ . . _,_ _.~,_- _ ._._-._ __ _ -_
06000202 39
()marysimons 1 In the phase two of this we extended the scope 2 slightly. We tried.to -- well, we picked up and used NPRDS 3 records and we tried to get some quantitative insights into 4 what the data was telling us. We tried to use some of' 5 those crude statistical tests to try to get some inferences 6 of what those data could mean in relation to the 7 reliability of the system.
8 (Slide.)
9 In the phase one results, in other words just 10 this qualitative look at the answers, and only of the LERs, 11 it became apparent that each plant really has in the 12 auxiliary feedwater system, each plant really has its own 13 peculiar or particular problems that affect that plant.
14 Another thing that showed up from this listing 15 of data was that in many instances the problems can go
] 16 uncorrected for long periods of tjae. For example, I 17 recall ANO 1-2, Unit 2, had problems with their feedwater 18 regulating valves and they had considerable failures. I 19 think it is listed in these slides as may be 37 or so. It 20 was a case of just repetitively failing.
21 Those type of valves were an electrohydraulic 22 type of valve, and eventually they were replaced by a 23 different type of valve and the problem was corrected.
24 That was an extreme case I think of repetitive failing.
25 So you can have problems going uncorrected for ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
l 06000202 40
) marysimons g
1 long time as was indicated for Arkansas Nuclear One, Unit 2 2.
4 3 ':MR. EBERSOLE: Is that in the presence of 4 appropriate regulatory pressures?
5 MR. MANNING: I am not sure how to answer-it 6 other than to say that the problem apparently wasn't really 7 fixed. In other words, we got these repeated ---
8 MR. EBERSOLE: I guess Davis-Besse was the 9 classic example. I didn't know about ANO-1.
10 'MR. MANNING: I suspect it wasn't highlighted 11 enough maybe.
12 MR. EBERSOLE: How do you highlight to the
- 3 13 proper degree?
14 MR. MANNINGi Pardon?
i 15 MR. EBERSOLE: How do you go about determining 16 the degree of pressure.to force these things when you have 17 an unresponsive utility?
18 MR. MANNING: I don't know the answer to that.
39 MR. EBERSOLE: Do you have some criteria?-
20 MR. MANNING: No, I' don't. There may be, but I 21 don't know.
22 MR. EBERSOLE: Many of the items-that are 23 subject to repetitive difficulty are not reported as LER 24 problems because they weren't system related. They are i
25 reported to NPRDS, which to my knowledge the agency doesn't ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 41
(,. )marysimons 1 really monitor in any systematic regular fashion looking 2 for regulatory difficulties. I think if they did they 3 probably wouldn't have access to NPRDS much longer.
4 So how do you really think that -- I looked 5 through'the reports you wrote and I_ find that indeed many 6 interesting patterns being reported to NPRDS which don't 7 show up in the LER system. Obviously they were having 8 trouble at Crystal. River with their pump governor 9 -arrangement and it just keeps coming on down the line 10 several times over a period of time, but there is no-11 regulatory monitoring of that other than at strictly the 12 local inspector level. He is probably the only one that is (h
\_ / 13 aware of this.
14 MR. WILLIAMS: We do have a program to monitor 15 the NPRDS to some degree.
16 MR. MICHELSON: Which program is that?
17 MR. WILLIAMS: This is the NPRDS methodology
! 18 program.
19 MR. MICHELSON: Oh, yes, but that is for a 20 totally different purpose though.
21 MR. WILLIAMS: The feed reg valve problem is 22 one of the issues in the methodology review that we have 23 done. In that particular area we found some trends on feed 24 reg valves. In particular, I think we identified one plant 25 as an outlier plant,.which was the No. 2 outlier plant, ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 42
['l marysimons v
1 for feed reg failures. There were 80 valves.
2 MR. MICHELSON: Was that a program of routinely 3 . monitoring to look for trouble or was that where you had a 4 specific question that was raised and you went in and did 5 an NPRDS type search? !
6' MR. WILLIAMS: It was really a specific 7 question, and that was ---
8 MR. MICHELSON: And that is'a different issue.
9 I think Jesse ~was raising the question of, you know, the 10 regulators routinely are seeing these things happen without 11 any apparent action, and the point I was raising is whether t
12 or not the-regulators even monitor for this kind of a 13 problem as opposed to going in and looking at a specific 14 issue and then searching NPRDS. I noticed that, but I 15 don't think you monitor the data base looking for trouble, 16 a regulatory type monitoring.
17 MR. DENNING: The program that Mark referred to 18 is kind of midway between monitoring the information on a 19 teletype kind of a program and looking on a specific 20 issue. What it consists of is a list of major components
( 21 in the plants ranked by a general risk significance, and 22 then we march down that list.
23 For example, the first area that we treated was 24 PWR primary he at removal. So we picked on main feed reg
_ 25 valves as the first component, and the components that are ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3346M6
06000202 43
[ D marysimons
\~J 1 being done now are main feed pumps and their drives. On a 2 BWR list heated generators are'the No. 1. topic un the 3 list.
4 So there is a program where we.will go look at 5 a component based on risk perspectives, but lacking any 6 recent driving issue. Now that is not the same treatment 7' that LERs get, and by volume we couldn't treat NPRDS data 8 the way the LERs are treated. There are too many reports.
9 INPO has gone off on another tact by way of 10 contrast with what we are doing of focusing on major 11 components, and the reason we did that was because we hoped 12 that the data would be better and that it would be easier
[
s/ 13 to demonstrate the significance of the results by doing 14 that.
15 INPO's program has a large volume data 16 crunching aspect to it with a number of different tests for 17 frequency of occurrence across many different classes of 18 components and types of things going on. So they will have 19 perhaps 10 alarms, and when five of these alarms go off in 20 some specific area, then they decide to go take a look at 21 what is going on with the data. But it is largely an U T 22 automated type of thing.
23 MR. MICHELSON: And they don't inform you of 24 these studies necessarily of what they are looking at 25 specifically.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6M6
06'000202 44
()marysimons 1 MR. DENNING: That's true.
2 MR. MICHELSON: And'you may or may not be aware 3 -of where they know the industry has problems until much i 4 later in the game. Eventually you become aware of them if 5 they become sufficiently serious, but I think in the i 6 regulatory sense the NPRDS is not used as a screening tool 7 to look for potential areas of difficult except in the few 8 instances.you have cited.
9 I think the local resident inspection is really 10 the only-one that is aware of the breakdown of components 11 on auxiliary feedwater when they don't prompt LERs. Isn't 12 that essentially it? And if he-becomes aware and he 13 worries about it, maybe he will call Atlanta or wherever 14 and get them appraised and get them worried a little bit.
15 It is kind.of a bottoms up feedback I think unless somebody-16 monitors NPRDS to try to get ---
17 MR. MANNING: That one of the conclusions we 18 make here.
.19 MR. EBERSOLE: Well, I well recall several 20 years ago that transition process and I bitterly opposed it
~
[ 21 where you lost monitoring and any regulatory control over {
22 single train failures. And it~was a gesture that industry 23 will arise to its level of competence, or whatever that may
! 24 be, and this is going to be an opportunity for them to show ,
! 25 how really good they are. Has that in fact materialized?
l (:)
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6646
06000202 45 f..s
()marysimons 1 I gather from that second sentence that it hasn't.
2 MR. MANNING: I think in some ways it has. We 3 get more data, more pertinent' data, and yet we lack ---
4 MR. EBERSOLE: It's a spectrum.
5 MR. MANNING: Yes. --- we do lack other 6 things.
7 I would like to look at the MR. EBERSOLE:
8 bottom 10 percent.
9 MR. MICHELSON: In terms of access to NPRDS, 10 and by that I mean terminal access so they can work the 11 database; is that correct?
12 MR. DENNING: Yes. I think the major O
\ 13 headquarters offices all have at least one individual ---
14 MR. MICHELSON: You do have access at the 15 regional level? I know you do at headquarters, but ---
16 MR. DENNING: I am not entirely certain.
17 MR. MICHELSON: I am just trying to get a feel 18 if the person in the region wanted to become aware of what 19 was in NPRDS could he even do so?
20 MR. WARD: We don't have an answer to that I 21 guess. Let's go ahead.
22 MR. MANNING: The last item there sort of 23 follows on the others. These failures when they do occur, 24 they can sometimes occur sort of bunched up and it would 25 appear that you would have fairly significant variations of ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
, .. - - ~-_ .
06000202 46
()marysimons I unavailability of the aux feedwater system as a function of 2 time. Now we don't have a quantitative indicator of that, 3 but there was a report put out and sent to NRR in which sne 4 listed all these various failures. A lot of them do occur 5 in short periods of time.
6 So those are the three conclusions then that we 7 make. Qualitative findings are.results that we have made 8 from this look at the LERs alone.
9 MR. MICHELSON: Could I ask you, I have a 10 letter dated July 25th from Hebdon Speis. pertaining to 11 review of an auxiliary feedwater related LER and NPRDS 12 data. Is that the actual report that I am looking at or a 13 synopsis of it or what?
14 MR. MANNING: That is the actual report.
t 15 MR. MICHELSON: This is the report you are 16 referring to that you wrote?
17 MR. MANNING: Right.
18 MR. MICHELSON: Okay. Thank you.
19 MR. MANNING: There are actually two reports.
20 There is a report prior to that, a shorter one in which 21 only the LERs were listed.
22 MR. MICHELSON: This was the proprietary and i
23 the other was non-proprietary.
24 MR. MANNING: The second one is proprietary and 25 the first one wasn't.
ACE-FEDERAL. REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
. , , 06000202 47
( ) marysimons 1 MR. MICHELSON: But that is the total extent of 2 the material?
3- MR. MANNING: Right.
4 MR. MICHELSON: Thank you. Are you going to 5 talk about that report any further?
6 MR. MANNING: Only very slightly because it 7 involves a lot of.NPRDS items which are proprietary. So it 8 would have to be a different type of meeting.
9 MR. MICHELSON: I always got the feeling though 10 that the claim was on NPRDS that as.long as you didn't 11 mention the utility you could talk about the failure. -Is 12 that correct? As far.as being proprietary, it was relating
/~~g
(_/ 13 the failure to the utility that they didn't want to make 14 public.
15 MR. WILLIAMS: I think that is the intent, but 16 we haven't drawn clear lines for briefing for that.
17 MR. MICHELSON: In other words, we can't then 18 ask the questions on this report unless we close the 19 session even if we don't mention names?
20 MR. WARD: We do have the report. If you have 21 got a question why don't you try to couch it so you don't 22 mention names.
23 MR. MICHELSON: I noticed, for instance, in the 24 NPRDS system you pick up a number of failures of s' ingle l 25- components and so forth. Some of these failures I would
/_s1 V
1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 48 ,
, (s)marysimons 1 have seriously thought would have led to LERs. Have you 2 ever checked a few of these in the NPRDS to see why they 3 didn't show up as LERs?
4 MR. MANNING: As I said at the start, we really 5 didn't do anything by use that file was it was. We didn't 6 quality check.it or anything.
7 MR. MICHELSON: Okay. Is the work being done 8 in Idaho still try to monitor the NPRDS database relative 9 to stuff showing up there and why it didn't'show up as 10 LERs? I know they did it the other way around. They were 11 checking to see if something in an LER showed up in an i
12 NPRDS, but did they do the reverse check as well?
\/ 13 .MR. DENNING: No, we don't go back and do that.
14 MR. MICHELSON: It would be hard to say, but 15 there were a number of these that I was surprised would not
{
16 have led to a system level failure.
i 17 Also, I noticed that one utility had trouble
{
18 with their check valves of the swing disk type. So they 19 changed to piston type, and this was a couple or three i
i 20 years ago. Yet that kind of a lesson learned after they
! 21 had a couple of failures that they showed earlier, that i 22 kind of a lesson learned somehow never propagated into the l 23 business until we got to San Onofre and then we started I
24 finally getting the word. But I was surprised because they i
l 25 apparently learned to get rid of their swing checks in the l
l l
l ACE-FEDERAL REPORTERS, lNC.
202-347 3700 Nationside Coverage 800-336-6646 i
06000202 49
, ()marysimons 1 auxiliary feedwater system in this case.
2 MR. MANNING: Was that on the steam inlet valve a 3 portion or the steam supply?
4 MR. MICHELSON: I imagine it was. It was on a 5 steam line, yes. They had two or three recordings of a 6 check valve loose and then finally a recording that says we 7 have replaced them noN with lift checks.
i 8 MR. MANNING: So that wouldn't necessarily
, 9 directly relate though I guess to like San Onofre where I
10 they had all those items.
11 MR. MICHELSON: Well, it relates to check 12- valves giving difficulty and finally realizing that in some
)f 13 applications you just have to get rid of them.
14 MR. EBERSOLE: Carl, you said check valves in I
15 the aux feedwater system. In general check valves in the i ;
16 main feedwater system stop the reverse flow of aux feed.
17 MR. MICHELSON: Well, these are in the steam j 18 lines.
j 19 MR. EBERSOLE: Oh, okay.
20 MR. MICHELSON: At least that is what it says -
I
{ 21 here.
22 MR. WARD: Anything else on.that, Carl?
23 MR. MICHELSON: No.
24 MR. WARD: Shall we go ahead.
! 25 MR. MANNING: The phase two effort then was
.( )
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80M36-6646
_ , _ - _ _ . _ . _ _ _ _ . _ . - _ _ , . . _ . . ~ . _ . . . _ , _ . . . , , , _ . - - . , _ . _
06000202 50
[v.3] marysimons 1 just an expanded effort from the phase one in which we 2 included the NPRDS records to expand the analysis 3 somewhat.
4 Then we tried to do two things. We tried to 5 get some quantitative estimates of the significance of some
~
6 of these failures and tried to relate them reliability-wise 7 to what we might expect to see in the -- related to 8 quantities we do see to what might be predicted by PRAs or 9 reliability studies.
10 (Slide.)
11 So we used some generic failure rates to try to 12 estimate the failures of some of the major components A
k-) 13 within the auxiliary feedwater system, for example, the
~14 pumps and the feedwater regulating valves.
15 We tried to estimate demands, for example, that 16 each one would be subject to in this five-year interval 17 that we are talking about and, finally, the unavailability 18 for demand.
19 We multiplied all those together and tried to 20 get an estimate of how many failures we should expect to 21 see.
22 (Slide.)
23 Here then is sort of an sample of what was 24 done, and this is just for pumps, turbine driven pumps in 25 particular. We used for the turbines ten to the minus two o
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
,_.s 06000202 51
(}marysimons 1 for demand unavailability, and we made a rough assumption 2 here or estimation of how many times per year they may be 3 commanded on excluding down time of the plant and 4 everything. They are not considering that. They assumed 5 they would be tested monthly and that there be on the 6 average maybe a half a dozen scrams per year-of which 50 7 percent of those you may need the aux feedwater system.
8 MR. HARD: What are the other 50 percent?
9 MR. MANNING: That you may in fact not need 10 this, that it is conceivable to use the main feedwater 11 system to bring the plant down and to go into the closed 12 loop RHR.
O
(_) 13 MR. EBERSOLE: If I can jump-back now to ATWS, 14 as you will recall. It is required that we scram the 15 reactor or we have big trouble. What are the comparative 16 numbers there? It is a little less than 15 times a year, 17 or about 8 or 10, but the required reliability is 18 phenomenally higher. But is the need any greater?
19 MR. MINNERS: I don't think the reliability is
\
20 higher. As I remember it, the reactor trip system 21 reliability was about ten to the minus four.
22 MR. EDERSOLE: But this is ten to the minus 23 two.
24 MR. MINNERS: Are you talking about systems or i
_ 25 are you talking about equipment?
G 1
I i
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
=
- 4 "
f_s 06000202 52
( ) marysimons 1 MR. EBERSOLE: I thought he was talking about 2 systems.
3 MR. MANNING: No, Jesse. This is just a 4 turbine pump.
5' MR. EBERSOLE: Oh, that is just a turbine pump. >
6 i MR. MANNING: We are trying to find out how 7 many failures we.should expect to see if we go look at the 8 $ data file.
MR. EBERSOLE:
9 I see. That is an individual j 10 and turbine driven pump.
l 11 MR. MANNING: Right.
MR. EBERSOLE:
12 . Okay. I'll withdraw.
O 13 l
l MR. MANNING: So we know how many we should 14 expect to find when we go look at it, and you can see it 15 says that we should have maybe a little less than one 16 failure. Maybe one failure is what we should expect if our 17 estimations are right or reasonably right.
18 (Slide.)
19 Before getting into showing you how we use 20 that, here is a sliding showing those major types of 21 failures that we have seen in the operational data for the f
3 22 aux feedwater system. -
23 For example, we see quite a few pump failures, I 24 as you might expect. We have seen some flow 25 instrumentation failures and, finally, some of the flow ACE FEDERAL REPORTERS, INC. !
102347 3RU Nationwide Coverage 800-33MM6
- . . _ . . . . . - - - - _ _ _ _. _..._ . _ . _ -_ ~ ,. _ _, . _ _ _ . . _ _ _ _ _ . - _ _
_ ~ _ _ _ . _ .
06000202 53
( marysimons 1 control valve failures.
2 Now here is just a count of those. failures that 3 we have seen.
4 MR. MINNERS: This is over five years?
5 MR. MANNING: Right, this is.from the five-year 6 observation period. We see about nine failures to start of 7 turbine driven type pumps for Arkansas Nuclear One and zero 8 failures for the motor driven pumps.
9 Now the failure to run failures, that was sort 10 of a judgment thing. Those mainly consist of, for example, 11 bearing failures or bearing degrades. I don't recall 12- seeing any pumps that, for example; just bound up or froze 13 in their tracks. But these were instances where they-had 14 performed maintenance on the bearings and found them to be
! 15 galled or found water in the bearings and such as that.
j 16 Those types of things involving bearings would mainly be
- 17 what affects the failure to run.
18 MR. MICHELSON: Is failure to run defined in 19 terms of time, two days or two hours?
i-20 MR. MANNING: Right. For this measure we l
[ 21 determined I think that you might need this auxiliary I
- 22 feedwu*er system for quite a big difference of periods.
I \
23 They cot,1d run anywhere f rom two hours to say three days 24 that you may need aux feedwater.
25 MR. MICHELSON: How did you define failure to ACE-FEDERAL REPORTERS, lNC.
202-347-3700 Nationwide Coverage 800-336-6 4 6
.06000202 54
. (.s)marysimor.s 1 run then?
2 MR. MANNING: We defined failure to run as the 3 whole interval.
4 MR. MICHELSON: How long of an interval?
5 MR. MAWNING: Let me back off. If we found a i
6 degraded component bearing-wise in the data file-we assumed 7 that that was like an incipient failure and we could expect 8 it maybe would failure shortly thereafter. We might be 9 wrong and it could run for a long time. Conservatively 10 then we took ---
11 MR. WARD: You are talking about inference 12 only?
/
13 MR. MANNING: We took these galled bearings and i 14 type of' thing and made the assumption that we were going to 15 count that as a failure.
16 MR. WARD: Are any of those actual failures to 17 run? I'mean are any of those events where ---
i
! 18 MR. MANNING: As I said, I don't think I have 19 seen an instance where it just came to a stop, but they l 20 . were the type of failures where you had water in the 21 bearings and leakage in there or it was disassembled and 1
22 the bearings were found galled and stuff like that.
23 MR. WARD: So those would typically be found 24 during routine preventive maintenance?
l 25 MR. MANNING: During some maintenance time, 1
I l
ACE-FEDERAL REPORTERS, INC.
,, 202 347-3700 Nationwide Coverage 800-33(>M46
y
, 06000202 55
( ) marysimons ws i either they suspected it and tore it apart or it was during 2 a routine maintenance or whatever.
3 MR. EBERSOLE: May I ask this. You say 4 feedwater system failures in the title, and then you say 5 pump failures. Are those individual pumps?
6 MR. MANNING: Right. Maybe it is a little 7 misleading. We should have said major categories of 8 components within the aux feedwater system.
9 MR. EBERSOLE: Components within, okay. That 10 makes it -- that fits.
11 MR. MANNING: This slide shows, too, by the way 12 -- remember we said each plant sort of its own peculiar
(~h k- 13 propensities to fail. Flow instrumentation, for example, 14 affected mainly two plants here from our data sources. We 15 have seen these two. The auxiliary feedwater flow control 16 valve failures, you can see that it just affects certain 17 plants. And, likewise, these pump failures.
l 18 MR. EBERSOLE: How many pump control valves are 19 there? There is one per pump, isn't there?
20 MR. MANNING: Now the flow control valves, 21 there are really about eight total per plant. There are 22 two series valves from each pump to each steam generator 23 and there are four sets of those. I think that is l
24 generally how it is.
1
, 25 MR. EBERSOLE: Are we getting by then with l (/
l 1
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cos erage 800-3346M6 l
A
i.
\
\
06000202 4- 56 I marysimons '
1- redundancy? C+
2 MR. MANNING: I thiak because.we do have ,
3 various'different paths ~the flow control valves aren't 4 really the more important type of thing'in there. It can ,
5 fail in fact and you can live with.a fail flow control x s 6' valve and use' alternate paths or,alternateisteam generators
~__.
7 even to cool down.
8 The major thing I'think is the auxiliary 9 feedwater pumps themselves because they are really the j 10 least redundant of all. Even flow instrumentation is'in a !
11 way is somewhat redundant. Well, for one thing yo'u have ,.
12 redundancy there I think on the flow contiol 13 instrumentation, but you also have level instrumentation in -
14 steam generators. So if you fail it you could conceivably
,i 15 maybe get by with just --- ,
16 MR. EBERSOLE: .In general if,you fail an aux 17 feedwater flow control valve can you manually provide flow?,
18 MR. MANNING: ,Through a'different path and I
- l 19 suspect you could;go out possibly at least on some of the
- 20 valves and maybe open it. I don't know.
21 MR. EBERSOLE: Open a bypaso.
22 MR. MANNING: For the auxiliary feedwater, I am ,
23 not that familiar, but I don't think they have bypasses.
24 That is comparatively small compared to the main feedwater.
25 MR. WARD: From reading the report I gather l 8 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage < 800-336-6646
, _ _ _ _ - ~ .
i s
^
4 t
- 06003202 57
(~)
~_
mary'simons
)
1 that you are convinced that Prairie Island really does have 1' i
2 better equipment performance than ANO and it is not just a s
3 reporting deficiency of the sort we were worrying about a 4 few minutes ago?
5 MR. MANNING: The data just says that.
6 MR. WARD: Okay, and you don't have any 7 judgment or you don't have any opinion on ---
'\ ' 8 MR. MANNING: I called each of the various 9 residents at each of these seven plants to try to in fact 10 see how they classified the auxiliary feedwater system, be 11 they classified as an engineering safety feature system, in 12 other words, it required full reporting or not. And in (y
k J' 13 talking to them I remember talking to the Prairie Island i
14 guy and he seemed to feel generally that they did have a 15 good plant.
16 MR. WARD: But does he think they are reporting a \
N 17 all of the individual item failures that they get?
18 MR. MANNING: I don't think they do. I don't 19 think they have to. There are still ways around it.
20 MR. WARD: As I recall, ANO, one of their units 21 was yes and the other no as far as whether it was an 22 engineering safety feature, but they seemed to be reporting 23 for both units. Now is that because they are having a lot 24 of failures or are they just doing a good job of reporting?
25 MR. MANNING: I'm afraid I don't know.
ACE-FEDERAL REPORTERS. INC, 202 347 3700 Nationwide Courage 8M336-6646
I 58
(~)
v 06000202 marysimons 1 MR. MICHELSON: That is the difficulty I had 2 also.
3 MR. WARD: It'is so marked in here.
4 MR. MANNING: However, I think that NRR is 5 planning some plant visits, and I presume one of the things 6 they would look at would be the relevancy and consistency 7 of' reporting.
8 MR. DIAB: If I could make.a comment here.
9 MR. WARD: All right.
10 HR. DIAB: The Prairie Island plant, they have 11 submitted to us a reliability analysis, and as part of that 12 they-discuss how they would cut their plant specific
> 13 aspects and by discussions with them a lot of that will be 14 available for us to audit to verify that that data does 15 indeed reflect their operations.
16 MR. WARD: So you are saying you have some 17 reason to believe that these numbers for Prairie Island 18 really reflect good equipment performance and not just 19 under-reporting?
20 MR. DIAB: We still have to review that and 21 verify it.
22 MR. WARD: All right.
23 MR. MICHELSON: Has a functional' inspection 24 been done on Prairie Island auxiliary feedwater? There is 25 going to be one, isn't there?
l ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 804 336-6646
06000202 59
()~marysimons 1 MR. DIAB: I am not really sure there will be 2 done.
3 MR. MANNING: No,.I think you can say it the 4' other way. It is unlikely that there will be one. It
'S would be against the region's policy to do it.
4 6 MR. MICHELSON: Why is that?
7 MR. DIAB: I&E is here.
8- MR.. CALLAN: I am-Joe Callan. I used to be the 1
9- resident inspector at ANO from 1979 to 1984. So I am going 10 to talk in that capacity, although that is not why I am 11 here. l 12 I am a little concerned about some of the m
'13 comments'that have been made regarding ANO's emergency 14 feedwater systems for Unit I and Unit 2. Both units have 15 safety grade or engineering safeguards feature BFW systems 16 now for both Unit 1 and Unit 2. That didn't used to be the-17 case, but that is the case now.
18 Additionally, regarding Unit 2 which heads up Iq that list there as having the most troublesome data, it is 20 important to keep in mind that Unit 2 went commercial in 21 1980, and most of those problems are heavily skewed towards 22 the first two years of operation and probably more towards 23 the first year of operation.
24 I recall vividly the problems that Unit 2 went
_ 25 through in 1981 regarding the flow control valves in ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 800-336-6646 '
4 06000202 60
()marysimons 1 particular and to a lesser extent the pumps. The nine 2 failures to start were attributed to design problems with a 3 Woodward governor that remedied I think in 1982, but I am 4 not sure about that date.
5 We subsequently went back-and did an SSFI in 6 ANO in January of this year, although.it was Unit 1. I
~
7 went back and took a look at Unit 2 data and I think it is 8 safe to say that in the last two or three years there has 9 been a very dramatic difference in failure rates of pumps 10 and valves for both units.
11 I am not sure how representative those comments 12 are for the other units there, but certainly for the case
/~}
(- 13 of both units of ANO I think'those failure rates are 14 heavily skewed towards the first part of the time period 15 being. looked at.
16 MR. MICHELSON: Well, did you look at the AEOD 17 letter in which they studied the NPRDS database as well as 18 the LER database?
19 MR. CALLAN: yes, I looked at that.
20 MR. MICHELSON: Well, I don't get a feeling 21 that there is a skewing because it seems to be well 22 distributed over the entire time frame. you can look at 23 the dates of the LERs and you can look at the dates on ---
24 now the NPRDS, I can never be quite sure whether this is 25 chronological or not of course. I had assumed that this ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationaide Coverage 800 336-6646
l l
I
- 06000202 61
( ) marysimons 1 was kind of chronological since LER fit that way, but I 2 don't know for sure. Maybe you could verify whether these 3 are chronological.
4 MR. CALLAN: I will-check on that.
5 MR. MICHELSON: And if it is chronological, 6 then it appears to be a distribution extending well_over ---
7 8 MR. CALLAN: I don't know in detail what has 9 happened since 1984, but I can say that from 1980 through 10 1984 there was a significant drop-off in problems.
11 MR. MANNING: That's right, and I should have 12 mentioned that. It is mentioned in the report that the O. 13 flow control valves really appear to have been fixed 14 because they changed out the valves. I should have 15 mentioned that here. So that is really not.a problem any 16 more we don't think.
17 MR. MICHELSON: These are apparently hydraulic 18 flow control valves and they were having lots of leakage 19 problems of various sorts. At least that is what the NPRDS 20 said.
21 MR. EBERSOLE: As a quite related matter on ANO-22 2, that is a Combustion Engineering plant, one of the-later 23 ones, and somebody in management there had what I regard as 24 prudence to put in I think a four-inch mainline discharge 25 to the primnry loop which is manually operable. Could you O comment on the use of that and the care taken to prevent ACE FEDERAL REPORTERS, INC.
202-347-3700 Nation ideCoverage 800 336-6M6
06000202 62 (faarysimons 1 inadvertent release,-et cetera, et cetera?
2 MR. MANNING: I am sorry, I couldn't hear you.
3 MR. EBERSOLE: 'I understand that there is a 4 primary relief valve-in ANO-2 which is the equivalent.of 5 you might say a motor operated remote manual ---
6 MR. CALLAN: Equivalent to a PORV?
7 MR. EBERSOLE: No, no, better:than a PORV in 8 that it doesn't have automatic functions which are ---
9 MR. CALLAN: They have what they-call an;ECCS 10 valve, an emergency core cooling system valve. It is 11 basically a'four-inch gate valve that makes the-whatever 12 hole size bigger in order to facilitate HPI cooling or what
.13 .they call HPCI cooling.
14 MR. EBERSOLE: Was there a statement of logic 15 in the. basis, the argumentative basis, et cetera, ever made 16 for the presence of that valve by the company and accepted 17 b'y NRC7 18 MR. CALLAN: I really am not qualified to talk 19 about that. I know as a resident I looked into that and:
20 satisfied myself that it served a useful purpose. My 21 concern was always that if they ever opened it that they !
22 would shut it and you are talking tremendous DPs across 23 that valve. It was a gate valve.
24 MR. EBERSOLE: Of course. It is a single 25 valve?
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
l 06000202 63
(_)marysimons 1 MR. CALLAN: A single valve.
2 MR. EBERSOLE: It is not in duplicate?
3 MR. CALLAN: Well, no, but ---
4 MR. EBERSOLE: I thought it was a paired valve.
5 MR. CALLAN: You are testing my memory here.
6 As part of some of the action team or action plan 7 modifications done along with the LTOP, the low 8 temperature pressure reliefs and the high-point vents, the 9 ECCS vent was somehow linked with that. ~ So I think they 10 are a downstream isolation valve.
11 MR. EBERSOLE: I think'what we need to get is a 12 statement from whatever source and in whatever form we can I
\_ 13 through NRC, the staff validating the need for and prese,nce 14 of that valve there and the pros and cons of its presence.
15 Can we get that?
16 MR.-MANNING: What does that have to do with 17 auxiliary feedwater reliability?
18 MR. EBERSOLE: Because that is a backup to it.
19 That is an expression of the fact that they don't have that i
20 much confidence in aux feedwater.
21 MR. MINNERS: Well, I would guess that when we 22 did our review of ANO-2 that those kind of features would
- 23 be included in our assessment.
24 MR. EBERSOLE: I never saw that.
25 MR. MINNERS: We haven't done the assessment ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6M6 t
l
~
l i"
- 06000202. 64 '
()'marysimons 1 yet. That is why you haven't seen it. The purpose of_this.
2 program is to go out and look at those plant specific
- 31 things.
! 4- MR. EBERSOLE: Okay. Well then we can include t
5 it now.
6 MR. MINNERS: Pardon?
f 7. MR. EBERSOLE: We can include its presence now 8 as a consideration which reflects a degree of confidence or 9 lack of it in the aux feedwater system. you know ofl course 10 that Palo Verde doesn't have such a thing, and it would be 1
11 interesting to_see the difference and-by what depreciative 12 factor did ANO-2 judge that that was necessary where-of ,
i.
}' 13 course Palo' Verde does not, when I find in fact that the 14 ANO-2 aux feedwater system is probably better than that.at-15 Palo-Verde, p 16- MR. MICHELSON: I think this chart is a good i ~
! 17 example now of the problem of whether or not we are
- 18 counting from equal populations. It is entirely possible 19 that ANO is1just doing a very fine job of reporting all of 20 its failures to NPRDS because I think all you did was 4
i 21 counted noses as near as I could tell from your report.
i 22 This is just a count of the number of times it showed up in 23' the NPRDS database or the LER.
. 24' MR. WILLIAMS: This question of a complete data
}- 25 set was one of the first things that came up when we i
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 1----.---.-._.. . _ . . . _ _ . , . ~ , . . . . _ _ _ . . , _ _ _ , _ _ , _ _ _ . _ _ _ _
06000202 65 i7-)marysimons
~
1 started to tilink about that, it is still ---
2 MR. MICHELSON: I can't believe that Prairie 3 Island didn't have any valve problems whatsoever in a five-
.4 year period.
5 MR. MANNING: You have a question mark as to 6 whether these in fact are true.
7 MR. MICHELSON: Yes, and the low ones may 8- indeed not be low. It is just not reported.
9 MR.' WARD: I think we heard that the NRR had a 10 means for getting a handle on that.
11 MR. DIAB: Yes. I am not really sure how they 12 are affected here. Of course over the five-year period 13 Prairie Island has a combined -- about 23 years of 14 combined operation and they have reported some failure 15 rates.
16 MR. MICHELSON: But in this five years we are 17' talking about they reported none.
i 18 MR. DIAB: As far as LERs.
19 MR. MICHELSON: Which means either they had 20 none or they aren't reporting what they had.
21 MR. MINNERS: They do not use their auxiliary 22 feedwater system very much, too, because they don't have l 23 main feedwater failure very often.
l 24 MR. EBERSOLE: And it can ramp down on a 25 turbine trip?
MR. MICHELSON: You would have to use it for l
l l ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
- t. . _ _ - . . . , _ . __.
, . , - _ . - - - - . . . - . . - , - . _ _ _ - . - . . . . - _ , , , , ~ , . _ - .
o I
,. 06000202 66
()marysimons 1 startup and shutdown if it can't ramp down. Now that.is a 2 very important point, and that is-you really need to 3 understand which one of these plants is using it for 4 u
' start'p and shutdown and which ones aren't.
I 5 MR. DIAB: That's true. I think we will cover J
- 6 Prairie Island to some degree later on today and we will 1-
, 7 talk about their philosophy of upgrading their aux 8 feedwater system.
9 MR. WARD: Okay. They leave most of the valves 10 that are needed for operation in position either open or f' -11 closed, but they have very few valves that they have to 12 close.
} A)
(- 13 MR. WARD: Okay. Fred, how about you trying to 14 go ahead now.
15 (Laughter.)
i h 16 MR. MANNING: Okay. The most important thing
- 1 17 about this chart maybe is remember we tried to as part of 18 this preliminary work to arrive at an estimate based .cn1 the
! 19 number of demands that we felt that there would be for like i
l 20 the flow instrumentation and the number of hours involved 21 and the failure rate per hour. We tried to determine or 22 estimate how many failures we should have expected to see 23 in this five-year period for.each plant, and this is this 24 bottom line.
25 you can see there is some agreement with some
.(:)
. ACE FEDERAL REPORTERS, lNC. l 202-347-3700 Nationwide Coserage 800-336-6646 I
. . ~ - - - - - - . - - , , - . . ,
, 06000202 67
()marysimons
~
1 of the things, but.in other cases there is a fair 1y 2 . substantial amount of disagreement.
3 MR. WARD: Im sorry,. Fred, how did you get the 4 expected number?
5 MR. MANNING: For example, like that pump 6 example we showed, we estimated the number of times we 7 would call upon the pump per year and we knew the generic 8 failure rate for pumps in those five years, and multiple it 9 all together and you get the_---
10 MR. WARD: I've got it.
11 MR. MANNING: So we tried to show the 12 quantitative thing,-that maybe the results you would get
\- 13 from using the PRA results maybe wouldn't exactly jibe with 14 actual operation data and there was some discrepancy there.
15 (Slide.)
16 Here is a further breakdown of the pump 17 failures and subcategorizing the pump failures by plant.
18 Again, these plants, you could generally consider them as-19 sort of a fail-to-start type failures or a fail-to-run. We 20 classified as failed to start like the governor controls, 21 .the inlet valve for a turbine driven pump and maybe 22 misalignment. The fail to run could be like the bearing 23 type failures or where we had several of the pump internal 24 drum misalignment failures which in some of the LERs they 25 noted that that resulted in a subsequent high bearing ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 68
[()marysimons 1 temperature. So those types of failures could be fail to 2 run also.
3 MR. EBERSOLE: .I wonder if you could just throw 4 back that earlier chart. I am completely at sea as to what 5 that last line meant.
6 (Slide.)
7 What does that last line mean, expecting 8 8/10th? What is 8/10ths? What are the units of 8/10ths?
9 MR. MANNING: Okay. If you~ multiply.those 10 numbers out you get a number. It doesn't have to be a 11 round number.
12 MR. EBERSOLE: Well,-just translate to me what 13 8/10ths means.
14 MR. MANNING: We should say we should expect 15 one or less failures.
I 16 MR. EBERSOLE: For what?
17 MR. MANNING: Over that five-year interval.
l 18 'MR. EBERSOLE: Over five years?
i 19 MR. MANNING: Right, and if we went to the data I 20 file we should have found less than one failure.
21 MR. EBERSOLE: On the average?
l 22 MR. MANNING: Right.
23 MR. EBERSOLE: Got it. Thank you.
i l 24 MR. MANNING: Likewise, if you don't like i 25 decibels, we should have found less than one failure then, t
ACE FEDERAL REPORTERS, INC.
I 202 347 1700 Nationwide Coserage 800-336-6646
06000202 69
()marysimons 1 MR. EBERSOLE: For the total five years.
2 MR. MANNING: Right.
3 MR. EBERSOLE: Thank you.
4 (Slide.)
5 This again shows that the pumps are dissimilar 6 enough at the different plants or the things afflicting 7 them are different at different plants. For example, a 8 pump drum alignment and adjustment problem just showed up 9 at I guess just one plant in particular, but it showed up 10 sort of repetitively though.
11 In fact, two of those problems, the 12 instrumentation and the flow control valve problems appear 13 to be fixed. They made changes or did whatever.
14 MR. EBERSOLE: I take it we are homing in on l 15 the. systemic failures piece by piece here.
16 MR. MANNING: Yes. So that is what this thing 17 shows.
.18 (Slide.)
19 MR. MICHELSON: One thing that bothers me just 20 a wee bit is that over the years I am sure I have heard a 21 number of lectures and perhaps have given a few myself 22 about the danger of counting noses, and you know the danger 4 23 of counting LERs and so forth, but that is precisely what 24 you are doing here and I just am still at a loss as to 25 whether it is meaningless or meaningful.
ACE-FEDERAL REPORTERS. INC.
202-347 3700 Nationwide Cmerage 800-336 4 46
. . ~ . ._
b 6
7s 06000202 70
-t
. )marysimons
.1 MR. MANNING: Well, we could be way off on-2 these numbers, but ---
3 MR. MICHELSON: The number of the count, I 4 don't doubt you are correct on, but you don't know what 5 population you should have had.
6 MR. MANNING: But remember'that is that one-
- 7. sidedness of this data source that we know.it can't be no j 8 worse than the worse ones, but it could be a lot better 9 than -- well, I guess I said it ---
t i 10 MR. MICHELSON: Well, I think the whole thing c 11 is really -- I think it is meaningless without a better 12 handle on the database. In other words, you have got to 13 have somehow an idea that from plant to plant you are 14 looking at a comparable situation.
i 15 MR. MANNING: For some of these plants that 16 have the high failure rate I think it gives you an 17 indication that that is something you should look at.
18 MR. MICHELSON: I don't know if you do or not.
- 19 Plants with.high LERs may just mean they are very honest 20 people.
21 MR. MANNING: But that still shows that using 22 what we think is contemporary PRA measures, those doesn't t
23 match up with PRA.
. 24 MR. MICHELSON: Well, I don't want to prolong 25 the argument. I have some trouble with your PRA approach l ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nation *ide Coverage 800-3366M6
._ _ _ . . _ _ , . . , _ . , _ . . _ _ _ . . . _ _ _ _ _ . . _ . _ . ~ . . - . _ . _ _ , - , _ _ , , _ _ . _ . _ , _ , . - . _ _ . , _ - - -
06000202 71
()marysimons 1 also, br.t that is another matter.
2 MR. MANNING: Generally then.our conclusions 3 would be in addition from phase one we had those 4 qualitative conclusions. Here are some maybe a little bit 5 more and some of them are quantitative. Some of those 6- components, major components did seeA to have a higher than 7 nominal failure rate, for example, some of the pumps and 8 the valves.
9 MR. WARD: Some of them, they all did.
10 MR. MANNING: No, they didn't all. Remember, 11 some of those were zeros on that chart.
12 (Slide.)
O)
\- 13 They weren't all out of the thing. All these 14 where we see a zero up on top, that is less than what we 15 expect.
16 MR. WARD: Well, the turbine, yes.
17 MR. MANNING: All these pretty much are all 18 good. So then this thing was fixed. After you threw away 19 that problem area that would be down to zero.
20 MR. WARD: I thought you said pump failures.
21 If I just look at failure to start ---
22 MR. MANNING: General 2V that is ---
23 MR. WARD: --- the turbine pumps are all worse 24 than -- I mean, first of all, thoce plants have been 25 singled out as the worse from the line diagram standpoint.
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationside Coverage 8 % 336-6686
-s 06000202 72
( ) marysimons 1 They have only got two trains and a simple minded PRA 2 expresses doubt about those. Then if you try to put in 3 some sort of equipment performance numbers into a PRA, you 4 say they are even worse because they are performing worse 5 than generic pumps perform and you have got this incomplete 6 indication which can only be worse.
7 MR. MANNING: One thing that is misleading here 8 is the fail to start and a lot of these would be 9 recoverable and I should have mentioned that.
10 MR. WARD: yes.
11 MR. MANNING: The point I wanted to get at 12 maybe though with the next slide is that these fail-to-runs kJ 13 would not be recoverable. These are all without recover, 14 but they in fact probably couldn't be recovered. You can't 15 go out and quickly replace the bearing, for example.
16 The point we are making is maybe these types of 17 things were the more important failures than the fail to 19 start, and yet our testing at the plants really emphasized 19 the fail to start. We really tests the heck out of that 20 monthly and everything and for 15 minutes to two hours.
21 MR. WARD: But they may be contributing to the 22 fail to run.
23 MR. MANNING: yes, but maybe in fact it is 24 really the fail-to-runs that are more important we really l 25 don't stress that in our test.
O x_J l
t ACE-FEDERAL REPORTERS. INC.
l 202-347 37(x) Nationwide Coverage W-3346M6 I _
. . _ , _ . _ . . _ _ _ __ ~
. _- ~ . . -
f~ 06000202 73
( j marysimons-1 MR. EBERSOLE: Well, the motor driven ANO-1 2 types, why is it that they had such a high set of failures?
3 MR. MANNING: You're right, it does look a 4 little -- I could check'the figures and see and make sure.
5 MR. EBERSOLE: Was it in the switchgear maybe?
6 MR. MANNING: No. I should maybe check that.
7 MR. MICHELSON: The governors were apparently 8 quite troublesome according to your subsequent data.
9 MR. MANNING: As I recall, this column is 10 pretty much right, and I am not sure about this ---
11 MR. EBERSOLE: There is no governor on a motor 12 driven machine.
n/x- 13 MR. MANNING: Right, right.
14 MR. MINNERS: It looks like the pump drum liner 15 was the problem here.
16 MR. MANNING: That would be the fail to run 17 type problem giving maybe a high bearing temperature since 18 some of those did result in high bearing temperatures.
19 MR. EBERSOLE: In general the motor driven i
20 pumps simply locked solid to the impellers or do they have f
21 a speaking total between.them?
22 MR. MANNING: I'm not sure, but I think you 23 have maybe got a point that that figure should be t
24 rechecked.
25 (Slide.)
i i
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nanonwide Cmcrage 8 5 336-6646 l
06000202 74
.(s)marysimons 1 Item 2 then, this maybe is one of the more 2 important findings I think. Really our testing does 3 emphasize start, but really the running time may be the 4 more important type thing and we really don't emphasize 5 that. ,
6 Then a third are these failures that maybe are 7 more significant and they are the non-recoverable type 8 failures, the bearing type failures, for example. They 9 seem to be increasingly more reported by-NPRDS. Now the 10 NPRDS isn't as visible a data source within NRC as the 11 LERs. The LERs everybody seems to look at, but NPRDS is --
12 I guess maybe several reasons, it is proprietary in some b
NJ 13 instances and maybe doesn't' get the visibility.
14 MR. MICHELSON: An uncontrolled database, the 15 problem is you have no requirement to report it at all.
16 you report as little or as much as you wish. When you 17 start to use that kind of data and then compare one utility 18 with another is a very dangerous grounds.
19 MR. MANNING: But even, in spite of those 20 problems, it is just not that accessible to people to use 21 it to find these types of problems that may be more 22 serious.
23 MR. MINNERS: What is the alternative, Carl?
24 Tell us what not to do.
4
- 25 MR. MICHELSON: The alternative was that if the U
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 86336-M46
I 06000202 75
()marysimons 1 NPRDS system wasn't reporting properly the single component 2 failures in a two-year period you were to go back to the 3 Commission and tell them so and start revising it with rule-4 making to get the reports mandatory one way or another. We 5 gave the industry their chance. Now the question is is the 6 industry really developing a database which you can use for 7 regulation. If it isn't you will have to develop your own, 8 which means they will have to send the reports to you 9 instead of NPRDS.
10 I think there is nothing wrong with the NPRDS 11 system, the database development. The problem is that not 12 everybody is reporting to it, and then to use that kind of 13 unknown population in drawing these conclusions I think is 14 dangerous. The ANO conclusion may not be valid at all. I 15 just don't know. It depends on how honest they were in i
16 reporting comparing with Prairie Island, for instance.
17 MR. MINNERS: you might doubt the Prairie 18 Island conclusion, but how can you doubt the ANO 19 conclusion?
i 20 MR. MICl!ELSON: Well, Prairie Island may be l
21 just as bad off, you know. I don't know.
22 MR. MANNIN0: That's right, Prairie Island may
! 23 be as bad off, but ANO certainly is probably ---
l 24 MR. MICHELSON: Oh, yes, clearly ANO has got l
l 25 things to thing about, there is no doubt, but to say that i
l ACE FEDERAt. REPORTERS, INC.
202 347 3700 Natierinide Coverage N0fk34-M
1 i
i e 00000202 76 -
l ()marysimons 1 Prairie Island looks good, I don't that is true at all. It l 2 just depends on whether -- if they are reporting everything 4
l 3 on an equal basis to ANO, then indeed they look good, but I l
l 4 don't know that that is the case.
f-i 5 MR. WILLIAMS: INPO has done a pretty big job I
l 6 because NPRDS is a terribly significant effort and they a
j 7 have made pretty big improvements.
I l
- 8 MR. MICHELSON
- Oh, yes, but you know yourself ,
i j 9 what the scatter is in the reporting of percentages. Some I
j 10 utilities are reporting very high percentages and some are 11 still reporting very low ones. So it is dangerous to start f
12 counting noses with the low counter and the high ecunter.
13 It is just, you know, not even safe to talk about.
{
i 14 MR. WILLIAMS: And we still have to get the 15 right yardstick for reporting it.
l 16 MR. WARD: Let's take a short break, i
17 (Recess taken.)
18 i
l 19 i
j 20 21 i 22 l 23 l 24 25
!(2)
L ACE. FEDERAL REPORTERS. INC.
202,347 3700 Nationwide Coverage 80k13MM6
. . - ._ _ =
,2 I
, i i
06000101 77
(),marysimons l, 1 MR. WARD: Mr. Callan, I think you're next. r I
i 2 MR. CALLAN: I am Joe Callan. I am the Section 3 Chief of the Performance Appraisal Section. ,
f' '4 MR. WARD: Gentlemen, please come to order.
- +
! 5 MR. CALLAN: I as the Chief of the Performance l 6 Appraisal Section. That is-the section that has been i 7 performing the safety system functional inspections over .
l B the last year.
I 9 We performed an SSFI at Arkansas Nuclear 1, f l
10 Unit 1, in January of this year. It was the third SSFI 11 that we have performed. Since then we have done two t
i 12 additional ones.
! i 4
13 ; MR. MICHELSON: Which ones have you done since i
', then?
t 14 lI i 15 ,
MR. CALLAN: We did TMI-1 and Oconee. In i
i 16 r addition to those two that were performed by our section, I l 17 recently assisted Region V in performing a Region V SSFI at l i
l 18 Trojan.
19 MR. MICHELSON: Is it true that these were l
20 being selected on the basis of possible difficulty, or are 21 you just randomly selecting them? .
l 22 i MR. CALLAN: The answer.is probably somewhere I
i 23 in between. As it will become-evident when I get into my ;
! 24 presentation, we are looking primarily for piants that have
- 25 a significant design change history. The inspection !
\
l ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80M346M6 .
06000101 78 4 I~T marysimons V
1 methodology focuses on a safety system that has been 2 heavily modified over the years, with the intent of 3 determining or assessing whether or not the system, as it 4 exists at the time of the inspection, reflects properly the 5 design basis.
6 MR. MICHELSON: But that is only one aspect 7 that you're looking at.
8 MR. CALLAN: Well, that's right, but we look.at 9 that from many perspectives, not only from engineering but 10 also from maintenance, training, operations, testing, and 4 11 its integrated tests, trying to get at the same bottom J
12 line.
. 13 MR. MICHELSON: Well, my concern is that the 14 agency seems to be kind of bearing down on those it thinks .
15 are having difficulty, I guess on the rationale that the 16 good actor ought not to be inspected so much. But it might i 17 be very enlightening on occasion to take what you think is 18 a good actor to get at the other end, to get a benchmark on 19 that end to see if indeed when you start looking at basic 20 design requirements and all the things which are specified ,
21 and see if that train looks so much cleaner in the good 22 actor and what is apparently is a good actor, just to keep 23 the logic straight. '
i 24 That is why I asked about Prairie Island, for 25 instance, which is purported to be a good actor. I would ACE-FEDERAL. REPORTERS, INC.
202-347-37(X) Narionwide Coverage 800-3%6M6 .
_ __ - . _ . _ _ . _ . . _ _ _ . _ . _ . - . . _ . _ . , _ _ _ _ _ . . _ _ , _ _ _ _ . - . - _ _ , . _ . _ _ . . _ . . . _ . . _ _ _ . , _ _ ~ . _ . . , _ . .
i l-79 marysimons l ()06000101 1 like to see a safety inspection, functional inspection done 2 on that just to get a benchmark for a purportedly good
}
3 actor, just don't do it many times but at least do it
! 4 once. .It'might be very enlightening.
5 MR. CALLAN: My intent is to briefly go over 6 some of the significant findings as a result of the ANO-1 7 SSFI. I think it would be instructive in doing that to 8 review briefly the' modification history of the ANO Unit 1 9 EFW system. I think it may be fairly representative of 10 auxiliary feedwater or emergency feedwater systems of PWRs.
11 One of the reasons, incidentally, that we have 12 been looking at EFW or AFW systems on PWRs is because it 13 has been our experience that they are the most heavily 14 modified, or one of the most heavily modified safety 15 systems.
16 [ Slide) 17 This is where ANO-1 began. What you see in 18 this slide is a nonsafety-grade emergency feedwater 19 system. The motor-driven pump was powered off a nonsafety-20 grade power supply. It was nonseismic, and in most 21 respects, nonsafety-grade. As a result of the TMI accident-22 in 1979, ANO-1 was shut down for some quick fixes. Keep in 23 mind this is somewhat simplistic, but just in order to save 24 time I am trying to summarize here.
ace FEDERAL REPORTERS, INC.
202447 1700 Nation *ide Coverage 800436-6M6
7s 06000101 80
- marysimons s
1 These were come immediate quick fixes that were 2 performed on the system. Then the unit was allowed to 3 restart.
4 [ Slide]
5 During the next outage the following year, some 6 more significant modifications were made to the system --
7 quite significant changes. As a result of those 8 modifications, the ANO-1 EFW system looked like that and 9 operated in that configuration for approximately the next 10 three or four yeary, sort of a hybrid safety system:
11 portions of it were safety-grade; portions were still not 12 safety-grade.
13 (Slide) 14 MR. WARD: Joe, earlier we had a comment. I 15 think it was your comment that there were a large number of 16 failures reported or problems reported at ANO-2 because it
- 17 was in this period that AEOD was looking at the this. It 18 was in the startup phase. There was also a large number 19 reported for ANO-1. It wasn't in its startup phase, but it 20 was in a sense, with all these modifications. Do you
(
l 21 attribute the large number for 22 ANO-1 to this bunch of modifications that was made?
( 23 MR. CALLAN: Obviously, somewhat. For example, 24 just to name one LER that sticks in my mind, one of these 25 two pumps -- I think it was the motor-driven pump -- was l
l l
ACE-FEDERAL REPORTERS, INC.
! 202-3 0 -3700 Nationwide Coserage 804336-646 L
e l
r i
r 81 !
()06000101 marysimons 1 rendered inoperable because installing safety-grade suction 2 piping, a weld 4ng glove was left in one of the pipes and-it ,
3 ended up clogging the inlet strainer to the pump, rendering
'4 the pump -- that's one of those cases where the pump would
'S not have started.
6 But in general, though, my recollection of Unit l
7 1's problems is it focused mainly on the motor-driven pumps !
8 problems with the balance, it turns out, a relatively !
9 sophisticated adjustment that had to be made to these pumps .
10 to remove the axial thrust. The suction comes in, and then :
1 8 11 it's diverted so that you neutralize the actual thrust. If ,
12 1 yt : a Jon't make that adjustment right, you can wipe out your i
Os 13 thrust bearing. That is, in fact, what happened on a !
14 number of occasions until they figured out what they were 15 doing wrong.
16 In addition, they were plagued by, I think, the ,
!- 17 same flow control valve kind of problems that Unit 2 was [
- i l 18 plagued with. They had this sophisticated electrohydraulic ;
. 19' design. j
- )
20 (Slide)
I 21 To finish up the 1982 modifications, these ?
l 22 obviously were not trivial modifications. For example, ;
r 23 they installed a new turbine driver on the turbine-drive l I
24 pump. ,
! I 25 (Slide]
.i I 1 f l
ACE. FEDERAL REPORTERS. INC. !
, 202-347-3700 Nationwide Coserage 800-33MM6
_ _ ~ - . -
I
'l 06000101 82 marysimons ,
1 MR. MICHELSON: Mbat turbine were they using?
i 2 MR. CALLAN: The Terry turbine with the 3 Moodward governor.
4 MR, MICHELSON: It's still Terry?
5 MR. CALLAN: Terry turbine. !
6 MR. MICHELSON: For both Units 1 and 2?
7 MR. CALLAN: For both Units 1 and 2. Right.
6 MR. MICHELSON: Unit 2 was not a Morthington?
f 9 MR. CALLANs No. Both Terry turbines, unless 10 they've been replaced. They have been replaced -- no, il still Terry.
j 12 MR. MICHELSON: .Both Terrys.
() 13 MR. CALLAN: Right.
14 MR. EBERSOLE: They have an electronic speed 15 control it says here.
16 MR. CALLAN: They do now. They had an 17 electrosechanical earlier, an earlier design that both 18 units had considerable problems with.
19 MR, EBERSOLE: There have been some very 20 interesting lightning-related accidents there recently, 21 which are fascinating in their degree of penetration into 22 the lower levels of the system. Can I spike this while I 23 am killing the electric pump?
24 MR. CALLAN: I am sorry?
25 MR. EBERSOLE: Can I electrically spike and O
ace FEDERAL REPORTERS, INC.
2n2 K n o NanonmiJe Cmerage 8(X436W> 6 j
83
~
marysimons
([06000101 1 ruin this speed control at the same time I have lost the 2 aux feed pump?
3 MR. CALLAN: That's an interesting question, 4 and I don't have any notion whether or not you can.
5 MR. EBERSOLE: Well, it's an electric 6 dependency, which I don't understand at the moment. It may !
7 be a battery-driven thing, I am not sure.
8 MR. CALLAN: Well, the philosophy of the 9 turbine-driven pump right now in 1986 is to make it 10 dependent on DC power only, remove the dependency on AC.
11 MR. EBERSOLE: As an ancient question, do they 12 remove the waste heat with an electric fan for it O- 13 "
-r someplace?
14 MR. CALLAN: I don't recall seeing that, but 15 it's been a couple of years since I have looked at it.
16 MR. EBERSOLE: It was a common dependency which 17 stood for years in many plants.
18 MR. CALLAN: As we will see, the system has i 19 undergone so many changes that right now, I rtnink, tho l 20 turbine-driven system is totally independent of AC and 1
j 21 dependent on DC only,
( 22 MR. EBERSOLE: One of the more typical things
- 23 was there was a steam line break detection system dependent l
24 on temperature rise, a thing that could be synthesized by i 25 Joss of room ventilation which effectively l l
(:)
l I l
l ACE-FEDERAL REPORil:RS, INC 202 147-1700 NationwiJe roserac N0 H63/46
.J
h 06000101 84
( } marysimons 1 meant when you lost AC power, you lost the turbine, too.
2 MR. CALLAN: I wasn't aware that.
3 MR. EBERSOLE: It was common.
4 MR. CALLAN: I wasn't aware of that.
s
~
5 The system was further modified during the 1983-6 1984 time frame. Then finally, during the 1984 outage, 7 very extensive modifications were made. Which brings us up 8 to the system that we inspected in January of 1986. That's 9 what we inspected.
[)
10 MR. WARD: Let me ask you a ques' tion. The s x
- 11 series of modifications, were they planned in that wayato 12 be carried out over a period of three or four years?
/)
_/ 13 MR. CALLAN: That's correct. They constituted, 14 , by and large, the TMI Action Plan required upgrades that 15 were built into a living schedule that was negotiated 16 between the licensee and NRR, with negotiated milestone '
s 17 dates, installation dates.
18 MR. WARD: Okay. So it was all part of the a
19 ;; same planned program --
k 20 ! MR. CALLAN: Right.
I 21 MR. WARD: -- that was carried out.
{
22 MR. CALLAN: In 1981 I knew that in 1984 23 certain things would be done at certain outages. yes.
24 ! Again, just to reemphasize, this is what we 25 inspected. Without walking you it, but you can just take e
q i
l I ace-FFOLRAL REPORTERS, INC.
i :o: m.pm i. ira m ic roverage m33wM6
06000101 85
()marysimons s 1 a mental picture of that, and that's what the licensee had 2 six years perviously.
\
3 [ Slide]
4 That means that when we were out there, that 3
5 they had a safety seismic-grade condensate storage tank
\(
6 here being installed. It was not hooked up, but at some.
[ 7 phase. When they get this installed, this condensate 8 storage tank installed, that will remove much of the 9 dependency that currently exists for the service water 10 suction. It's alternate suction.
s 11 MR. EBERSOLE: Is the service water system a 12 seismic system?
, 13 MR. CALLAN: It's seismic, yes, and constitutes 14 the ultimate heat sink. Obviously, you don't want to use s 15 that unless you really have to to avoid the dependency on 16 condensate storage tanks. It's superior.
17 MR. EBERSOLE: But now that's what's used is 18 the service water system?
19 MR. CALLAN: Yes. Unless they hooked up that 20 condensate storate tank in the last few months.
21 MR. EBERSOLE: Which is, in' turn, driven by 22 safety-grade motors and pumps, et cetera?
23 MR. CALLAN: Exactly.
24 MR. EBERSOLE: It has access to the diesels?
s s S
25 MR. CALLAN: Yes. It's the same heat sink that
),
s ACE. FEDERAL REPORTERS. INC.
s 202 347-3700 Nationwide Coverage 800-336-6 4 6
. - - . . _ .. . =- - . ... - -
l l
06000101 86 '
( ) marysimons 1 cools the decay heat removal heat exchanger. l 2 MR. EBERSOLE: What pressure does that develop 3 at the suction point on the pumps?
4 MR. CALLAN: I don't recall.
5 MR. EBERSOLE: Can the steam generators be 6 deliberately blown down to receive service water without 7 the booster pumps, which is, in effect, what the> turbine-8 driven pumps are -- I mean, the aux feed pumps are?
9 Suppose that I just blown down. Suppose that I have lost 10 them both. Do I have adequate service water pressure and a 11 capability to blow down the boilers to receive water 12 directly from r 'rvice water?
4
- 13 MR. CALLAN: I really can't answer the 14 question.
15 MR. EBERSOLE: It would enhance the safety.
16 MR. CALLAN: My notion is that you can't. But I
17 I don't know that for sure. I know that the licensee does 18 not take credit for that.
19 MR. EBERSOLE: He says the boilers are too 20 high, 21 MR. MICHELSON: you can use just hdrostatic 22 pressure.
23 MR. CALLAN: Right.
24 MR. MICHELSON: I think the boilers are too 25 high for --
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000101 87
() marysimons 1 MR. EBERSOLE: I am not talking about 2 hydrostatic; I am talking about 200 psi or thereabouts.
3 MR. MICHELSON: Which pumps do you want to use?
4 MR. EBERSOLE: Service water.
5 MR. MICHELSON: For what? ,
6 MR. EBERSOLE: Right through the pumps. Right 7 through the pumps as they are there. You see, I don't know 8 what the suction pressure is when I stop the main pumps. I 1.
< 9 presume you could deliver water right to the pumps.
10 MR. MICHELSON: Yes. It looks like there is a 11 possible path --
12 MR. CALLAN: There is a pathway, but the issue 13 is: Can they; is there enough head? And I don't know the 14 answer to that exactly.
15 MR. EBERSOLE: With the secondary blown down.
16 MR. CALLAN: That's right. I don't know.
17 That's a lot of piping, and there is a head difference.
18 Keep in mind also that in the scenario, the limiting 19 scenario here, you need your containment coolers, which are 20 also cooled by service water.- You have other heat loads on l 21 service water that would compete.
22 MR. EBERSOLE: 'Well, you're using service water l
l 23 anyway, i
24 MR. CALLAN: Yes.
.(:).
i I
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8043 4 6646
4.
t 06000101 88
()marysimons 1 MR. EBERSOLE: you're just boosting the 2 pressure up because you've got high boiler pressura.
3 MR. CALLAN: I don't know. Again, the licensee 4 does not take credit for that mechanism in designing the 5 system. I have seen designs where that is explicitly taken 6 credit for, but that is not the case ~in this system. I am 7 not sure if that's been studied.
8 Now, so this is the system-that we inspected.
9 .It should be clear we had'a lot of design changes to mull 10 .over, to review. In general, we found the system to be
'l 11 functional, a-thoroughly.well-designed system. The 12 operators seemed to like it. They seemed to have a lot of 1
v 13 confidence in the system. They seemed to have.a lot of 14 confidence-in the control system associated with it.
~
L 15 We identified a concern regarding a testing of 16 some of these check valves, and those concerns, of course, i 17 arose out of the San Onofre event, which had happened just
! 18 not too long before we'went to Arkansas. But beyond that I
19 He did not find any significant design problems with this l
20 portion of the system. We identified some concerns with 4
21 the way the limit switches and torque switches were set on 22 some of these motor-operated valves. That happens to be 23 generic issue in the industry. There is an IE Bulletin 24 85-03 that addresses that concern. The licensee has a 25 program to resolve those issues.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-336-6646
I I
i 1
l 06000101 89 !
( ) marysimons l 1 MR. EBERSOLE: Let me ask you about the San 2 Onofre incident in this context. I think one of the main I 3 contibutors to that was they thought they were being 4 conservative in reducing flow by about 10 or 15 percent, 5 and they went into a mode of operation where the valve just 6 didn't hang up in plane and hold steady on its stop but it 7 started flapping. In an aux feedwater control valve 8 system, inevitably there has to be a. flow variability as 9 you go down in power. Therefore, I need a valve that can 10 operate steadily without banging around at a variety-of 11 flow rates. Have I got those?
12 MR. CALLAN: I am not sure I understand. You 13 mean a valve that will give you linear flow?
14 MR. EBERSOLE: That will give me flow without 15 the valve just banging around because of the instabilities 16 in the flow pattern.
- 17 MR. CALLAN
- Okay.
18 MR. EBERSOLE: Do you recall San Onofre got in 19 trouble because they went down about 10 or 15 percent.
20 MR. CALLAN: You're asking me whether or not l
21 these valves --
22 MR. EBERSOLE: They face variability in flow 23 rates. The sink jets.
24 MR. CALLAN: These valves right here?
25 MR. EBERSOLE: Yes.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
e 06000101 90 marysimons
{}
1 MR. CALLAN: .Okay.
2 MR. EBERSOLE: They by necessity have to deal 3 with a range of flows whereas main feedwater does not. Are 4 they designed to accommodate variable flow rates without 5 banging around?
6 MR. CALLAN: Well, again I guess I can't answer 7 -that question explicitly. These valves here modulate.
8 They will modulate.
9 MR. EBERSOLE: Of course. That's the ones that 10 do the modulation.
11 MR. CALLAN: I am sorry, these are the 12 modulators.
13 MR. EBERSOLE: Whatever. The f act is, . aus I 14 say, there is a variable flow rate on those valves.
15 MR. CALLAN: Okay. But you're talking in terms 16 of them banging. ,
17 MR. EBERSOLE: yes. Whatever.-
18 MR. WARD: We're going to have a gentleman from 19 ANO speak next.
l 20 MR. EBERSOLE: All right.
I 21 MR. WARD: I think he can address these last 22 questions.
l 23 MR. CALLAN: All I can say is that in our l
l 24 review of the performance history of the system we'didn't 25 come across any evidence of those valves banging.
1.
ACE-FEDERAL REPORTERS, INC.
! 202-347-3700 Nationwide Coverage 800-336-6 4 6 l , . , _ - - . - , . . _ . , - - - , - - .
. , _ . . . , . - . . _ . , - _ - _ . . _ . - . - . _ . . . , . . , . . _ . ~ . , . . ~ . ~ . . . - . _ - , . - .
06000101 91
-(f_marysimons 1 MR. EBERSOLE: Or being damaged?
2 MR. CALLAN: Yes. But it's.an interesting 3 question because I am about to talk abou't some check valves 4 where that wasn't the case and resulted_-- and the biggest 5 problem we found has to be some check valves in the steam
- 6. supply system.
7 This is a portion of the Unit 1 EFW system that 8 was not covered on the previous schematics. It shows the 9 steam emission system into the turbine-driven pump, and 10 this is the system as'we found it. Now, this was supposed 11 to be single-failure-proof, fully safety-grade. Jhe 12 identified a design flaw in the system such that a single 13 failure would have taken out the turbine-driven pump and 14 the otor-driven pump and resulted in all loss of. emergency 15 feedwater.
i 16 Now, going back to something I said earlier, 17 the system is not dependent on AC power during operation..
18 These AC valves are normally open'so.that the only valves i
l 19 that have to stroke for initiation are.DC valves. These 20 valves stroke open on initiation and emit steam into the 21 driver.
22 The concern is: if you have a faulty steam 23 generator, a steam break, say, where my pointer is, flowing l
24 down A steam geneerator, the need to develop, to isolate l 25 the A steam generator from-the good. steam ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage _ 800-336-6M6
.__ - _ _ - . _ _ . - . . . , , _ . - - . . _ . . . _ _ . - . . _ _ _ . . . _ , _ _ . - _ - ._---~._ _ __
-C6000101 92 marysimons 1 generator, the B one, to ensure adequqate steam supply to 2 your driver. In order to isolate the A steam generator, 3 this AC valve has to shut.
4 Now,' it turns out that this AC valve is powered 5 off the same safety bus as the motor-driven pump. So if 6 you hypothesize a loss'of that bus, that safety bus, you 7 would lose the ability to shut that valve and also the 8 motor-driven pump.
9 So if the initiating event was a steam line 10 rupture on the A steam generator and the single failure was 11 the loss of that safety bus, under this design you would 12 challenge or degrade your ability to provide steam to the
() 13 turbine driver.
14 It sounds pretty straightforward, I guess, when 15 it's explained that way, but again I think it's instructive 16 to just briefly walk through how that happened.
17 [ Slide) 18 This is the original design back in 1979. This 19 is how steam got to the driver before TMI. Normally shut 20 AC valves on initiation, this valve would open, that valve 21 would open, allowing steam.
- Obviously, it is not safety-22 grade, not single-failure-proof.
23 (Slide]
24 That, very simply, is the system that we 25 O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646
I l
l 06000101- 93 i
( ) marysimons 1 inspected. Those modifications were put in' partially in l
.'2 1982 and finished up in 1984. Before we left the site, 3 this is the system the licensee had. The design flaw was 4 remedied by putting in check valves.
5 Now, getting back to the question of check 6 valves slamming shut. The reason, it turns out,.that the 7 licensee left those check valves out, because those check 8 valves existed in the proposed design, the original design 9 of this system, because of check valve failure problems on 10 Unit 2, ANO-2, where these valves would flutter during 11 operation and self-destruct and result in the internals 12 being distributed in downstream piping, the licensee made a 13 judgement that rather-than subject Unit 1 to the same kind 14 of problems, to leave the check valves out.
15 MR. MICHELSON': Why didn't they'go to. lift 4 16 checks, which they did on Unit 27 17 MR. CALLAN: I think they did. I think -- did 18 you go to those?
19 MR. EBERSOLE: Instead of leaving them out --
20 MR. CALLAN: yes, I don't recall the solution, 21 but the replacement valves were adequate to remove the 22 problem, so they don't destruct anymore.
23 But that was the single biggest design flaw we 24 found in the system. But I want to be careful to put that 25 in context of how the system was transformed over the ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6M6
06000101 94
) marysimons 1 . years, to show you how design flaws can be introduced into 2 the system.
3 MR. MICHELSON: Excuse me. In your system 4 inspection did you actually go back and find the record 5 where the check valve was removed and the safety analysis 6 justification on why there was no unreviewed safety 7- question and so forth?
8 MR. CALLAN: yes, we looked at that. The check 9 valve was never installed. The design was modified before 10 the hardware was --
11 MR. MICHELSON: Before it was installed. I 12 see.
13 MR. CALLAN: Okay. But nevertheless, we did 14 look back --
15 MR. MICHELSON: I am surprised. I am a little 16 confused. The problem showed up on Unit 2, so you left 17 these off on Unit 1. The Unit 1 was built before Unit 2.
18 I have lost the chronology here.
19 MR. CALLAN: Okay. Unit 1 originally did not 20 have any check valves. Unit 1 did not have a safety-grade 21 FEW system. Unit 2 did. Unit 2 was built with a safety-22 grade system.
23 MR. MICHELSON: Well, then the Unit 2 they 24 first put the check valves in and --
25 MR. CALLAN: yes, and had --
ACE-FEDERAL REPORTERS, INC. f 202-347-3700 Nationwide Coserage 8011336-6646
{
06000101 95
( ) marysimons 1- MR. MICHELSON: -- then decided not to put them 2 in on Unit 1.
3 MR. CALLAN: Because of lessons learned from 4 Unit 2.
5 MR. MICHELSON: So the Unit I was always 6 deficient as far as this particular single-failure is 7 concerned.
8 MR. CALLAN: Yes.
9 MR. MICHELSON: Okay. Thank you.
10 MR. CALLAN: I guess one could a_gue -- and I 11 think it's safe to say -- that even with the flawed design, 12 this design here, it was still superior to the original O
(/ 13 design, even though it didn't have the check valves. The 14 final design, of course, is fully safety-grade. This is 15 now single-failure-proof so that'a faulted red bus, a red 16 safety bus which would prevent you from shutting that 17 valve, would now not blow down both steam generators but it 18 would limit the damage to only one steam generator.
l 19 MR. EBERSOLE: Tell me, did they consider and 20 reject having the motor-driven pump being given access to 21 two power sources, the swing transfer, which would be two 22 diesels? That's usual'ly the most. direct correction of that 23 problem.
24 MR. CALLAN: I don't recall if they went
()
I.
i
< ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
06000101 96
()_marysimons 4
1 through that analysis.
2 MR. EBERSOLE: A very important load like that, 3 you need a potential factor of two in reliability. That's 4 about the limit.
5 MR. CALLAN: Yes.
6 MR. EBERSOLE: But that can be comforting.
7 MR. CALLAN: Yes. In this case, that would 8 have mitigated this problem.
9 MR. DAVIS: I have a question. You are i
10 postulating here a steam line at a location that is not 11 isolable, I guess.
12 MR. CALLAN: Right.
13 MR. DAVIS: Then, following that, you have a 14 loss of offsite power?
15 MR. CALLAN: Right. That's assumed for a 16 design-basis accident.
17 MR. DAVIS: Then, following that, you have a 18 loss of red bus?
19 MR. CALLAN: Yes.
20 MR. DAVIS: I think if you were to assign 21 probabilities to these events, you would find that that 22 sequence has exceedingly low probability and would not be a 23 very risk-significant scenario. I guess my concern is you 24 seem to be looking at things on the basis of design-basis 25 accidents coupled with single-failure criteria ACE-FEDERAL REPORTERS. INC.
202-347-3XO Nationwide Coverage 800-336-6646
l 06000101 97
()marysimons 1 rather than at some risk-based methodology that I think.
2 would probably find more important problems.
3 MR. CALLAN: Well, we do. We do look at it 4 primarily from a design-basis assessment, but it's 5 important-to emphasize that before we got into that design-6 basis scenario problem, we went through the system fairly 7 thoroughly, and as I indicated earlier, we concluded that 8 the system was basically sound. It seemed to be a fairly 9 reliable system as evidenced by its performance history; it 10 had certainly earned the confidence of the operators, which-i 11 we feel is significant.
12 So from a reliability standpoint, under normal l 13 conditions and the report reflects it; the insp< action 14 report makes that statement -- that under normal conditions l 15 the system appeared to be adequate. It was only in the 16 limited design-basis condition that this flaw would have 1
17 made any difference.
l 18 MR. EBERSOLE: On the other hand, this wasn't --
19 MR. WARD: I don't think you heard what he 20 said. He says you're calling it a flaw.- you're saying if l
- 21. you look at it from a risk standpoint, it may not be a l
22 significant flaw. You didn't really respond to that.
23 MR. CALLAN: Well, my response is that the 24 system was supposed to be designed to handle that 25 situation, and it would not have. I know that begs the ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6646 l - _ . , _ _ _ . . _ _ _ . _ - _ _ . . _ _ . _ _ . _ _ _ _ _ _ _ . _ . _ _ _ , _ _ _ , , _ _ , , _ _ _ , _ . , _ _ , ,__ , _ _ _
m 06000101 98
( ) marysimons 1 question, but that's the only. response I can give you. It
. 2 was not designed -- it was not installed in accordance with 3 its design basis. That's what we were looking for.
4 MR. EBERSOLE: On the other hand, you did not 5 need a total AC power failure. All you needed was a branch 6 system. failure due to the transient loading for the one' 7 that carried the electric feed pump because it carried the 8 valve as well.
9 MR. DAVIS: I know, but you have to postulate 10 loss of offsite power to ---
11 MR. EBERSOLE: No; all you need to do is ,
12 postulate a source failure at the bus level because of bus 13 overloading, you don't need to lose the rest of the power; 14 just that part of it.
15 MR. MICHELSON: The electric is on that bus.
16 It would have to be one or the other anyhow.
17 MR. EBERSOLE: you don't need to lose all of l
- 18 everything; just that bus due to the transient overloading.
i 19 l
20 MR. DAVIS: Can't the motor-driven pump be 21 operated from offsite power?
l l
22 MR. EBERSOLE: Oh, yes, and from the diesels.
l l 23 MR. MINNERS: It operates from a bus, not from l
l 24 offsite power.
25 MR. EBERSOLE: It operates from a bus, which ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coserage 800-336-6646
-- 06000101 99
( ) marysimons 1 carries a whole lot of loads which only undergo transients 2 in transient loadings.
3 MR. MINNERS: That's what a fault is.
4 MR. EBERSOLE: So if you have a transient 5 loading condition --
6 MR. MINNERS: Could you put that slide up?
7 [ Slide]
8 MR. EBERSOLE: If you have a transient lohding, 9 then it kicks that bus out.
10 MR. DAVIS: But the report said the loss of the 11 -diesel would be the most likely loss of a bus, as I 12' recall. That bus can be energized by offsite power also.
13 MR. MINNERS: I agree with you. It is probably 14 a low-probability event.
15 MR. DAVIS: You start off with about a 10-5 for 16 a steam line break and go down from there.
17 MR. WARD: Okay. I think the point is made.
l 18 MR. MICHELSON: I don't see the main steam l
i 19 circuit, of course, but you know, without.the check valves 20 there that break is going to blow down both steam l
l 21 generators back through this auxiliary steam line l 22 connection unless you have check valves in it. Was that of l 23 any concern? Or do you blow down both steam generatore for l
l 24 any steam line break? It just depends on how the steam 25 line system is arranged.
l O
l ACE FEDERAL REPORTERS, INC.
202-347 3700 Naticawide Coverage 800-336-6M6
( - - - -.
i 06000101 100 marysimons 1 MR. CALLAN: Well, this schematic shows the 2 steam line. Here is the --
3 MR. MICHELSON: It's a cross connection with 4 the two steam generators, and if there is no check valve in 5 it and you don't shut the valves, you lose the -- you blow 6 down both steam generators.
7 MR. EBERSOLE: In that connection, let me ask 8 you if --
9 MR. CALLAN: You have isolation valves.
10 MR. MICHELSON: So then you lose the auxiliary 11 feedwater turbines. But your cross connection, where the 12 two feeder valves, the 2667 and the 2617.
13 MR. CALLAN: These are the steam inlet valves.
14 MR. EBERSOLE: you blow down B generator to A 15 if the break is on the A side.
16 MR. CALLAN: That's the point. Right.
17 Exactly. That's the point. That was our concern.
l 18 MR. EBERSOLE: That's why they put them in. It
! 19 doesn't necessarily have to do with the auxiliary feedwater 20 turbine reliability either. You have to that there anyway.
21 MR. CALLAN: Well, the point we're trying to
! 22 make is that with that one failure, the failure of that red 23 bus, you'd lose your motor-driven pump because that's 24 powered off that bus, and you'd also blow down both steam
(
l 25 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6M6
06000101 101
(_,) marysimons 1 generators, which takes out this pump, which means you 2 would lose both pumps and lose all EFW. That was the 3 point.
4 MR. EBERSOLE: Now let's imagine the check 5 valves are there, which I understand they are.
6 MR. CALLAN: Yes.
7 MR. EBERSOLE: And the others are there. And 8 ask, I think, the critical question: Has there been an 9 examination of the dynamic competence of these valves to 10 withstand full open-ended blowdown flow and still close, 11 the check valve being the first to attempt to respond with, 12 usually, a thunderous clap against its seat, which may s 13 disintegrate it and carry its parts right on into the other 14 valve, wh ich maybe can' t work anyway? Has a dynamic 15 analysis been made to show that each of these check valves 16 will stand open-ended blowdown flow and close? In the irst 17 case, the check valve may close too fast. The.second is 18 the other may not be able to even move at all. That 19 requires circuit breakers, all sorts of things.
20 MR. CALLAN: Sounds like a generic check valve
- 21 issue. Is that sort of thing looked at by --
22 MR. EBERSOLE: It's here.
23 MR. CALLAN: Yes.
24 MR. EBERSOLE: It's purported to be a part of f 25 your safety system functional inspection. In fact, one
- O l ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4334 6646
t
'I 06000101 102
()marysimons 1 part of it says you will look at each component to see if 2 it will perform the service for which it was-included. And 3 you've got to look at the check valves and see if they will
- 4 work; you've got to look at the other valves to see if
-5 they'll work.
6 MR. CALLAN: That's why, if you recall, I 7 expressed the concern that we identified regarding failure 8 to test check valves. Now, what we will do is verify -- we i 9 will check to see if the licensee is periodically or at.any 10 time verifying the check valves are still there, that they 11 work. Now, that doesn't answer your concern, because you 4 12 could test the check valves once a month and they still 13 wouldn't perform in the way --
14 MR. MICHELSON: Testing valves is normally done 15 by unloading, swinging them back and forth and putting a 16 number on their vital statistics which goes with the PRAs 17 and deluded you even worse.
18 MR. CALLAN: Well, that's something -- the 19 specific concern that you have raised is something we have 20 not looked at-in SSFIs.
21 MR. MICHELSON: My point is that it's supposed 22 to have been raised because I have read your module 23 concerning how to do SSFIs, and it says that it will be 24 done, and I assume that you are indeed checking each 25 component to see if it can perform the function for which ACE-FEDERAL REPORTERS, INC.
202-317-3700 Nationside Covesage 800-336-6646
. . , . ~ _ _ _ _ .
,n 06000101 103 s J marysimons 1 it was intended. If it was intended to isolate a break, 2 then you check to see that it would perform that.
3 MR. CALLAN: yes.
4 MR. MICHELSON: Now, you do that by looking at 5 the design specification to see if the manufacturer was 6 told that was a required function and so forth. That's 7 what a functional inspection is supposed to do.
8 MR. CALLAN: That's right. We in the past, i
9 say, at previous inspections, have looked at lift pressure 10 for check valves, et cetera. What I am responding to is
, 11 the specific concern that you raised. We have not looked l 12 at that specific issue in previous SSFIs. That is one 13 specific one.
14 MR. DAVIS: During normal operation are those .-
15 check valves open or closed?
{
16 MR. CALLAN: They are normally shut. There is i
17 no flow to those.
18 MR. DAVIS: Okay. They don't have to stand the l
19 dynamic closing?
l 20 MR. EBERSOLE: Oh, that's right. You're right.
21 MR. DAVIS: The other problem is that you have 22 to look at the possibility that those could be stuck closed L
23 and then you can't use a turbine-driven pump at any time.
f 24 MR. CALLAN: But that would become evident
()
1 ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
06000101- 104
(}'marysimons I during the monthly surveillance test.
2 MR. DAVIS: Well, all I am saying.is.that you 3 have to offset that against the advantshe.of putting then 4 in there in the first place.
5 MR. CALLAN: Right. Well, it was that thought 6 process that resulted, I think, in the licensee leaving 7 them out.
8 MR. DAVIS: On a risk-basis evaluatior., it 9 might be difficult to justify.
10 MR. MICHELSON: I believe, indeed, the check 11 valve does have to close when the full differential, if you 12 break A steam line and the cross lines are open all the p 13 time, that check valve has to snap closed if --
i
. 14 MR. EBERSOLE: But it's already closed.
15 MR. MICHELSON: If they're already closed in i 16 the fully closed position.
17 MR. DAVIS: That is what'he said.
l 18 MR. MICHELSON: But they are not necessarily. ,
19 We just don't know.
- 20 MR. DAVIS
- You don't know where they are, f
21 MR. MICHELSON: They are now,'but they weren't 22- before.
23 MR. EBERSOLE: There is no external indication 24 of where they are. They're lift checks.
i 25 MR. MINNERS: Joe, can I find that
! ACE-FEDERAL REPORTERS, INC.
l' 202-347 3700 Nationwide Coverage 800-336-6M6
- - . , - - .- .. . - . - , . , - , , , - - - - , . . - - . - - - . . - - - - . ~ - , - , - , . . -. - - -- - .- - - .-
. 06000101 105
(} marysimons; t 1 configuration in the FSAR?
2 MR. CALLAN: This one right here?
3 MR. MICHELSON: Did you get that out of the 4 FSAR?
5 MR. CALLAN: The updated FSAR should reflect 6 that. No, we didn't get that from --
7 MR. MICHELSON: You didn't get it from them, so it might be in there and it might not.
8 ll I
9I MR. CALLAN: It's required to be. It's 10 required to be since -- annually that has to be updated.
l 11 i MR. EBERSOLE: What sort of periodic tests are 12 performed on the checks?
O- 13 MR. CALLAN: Pardon me?
14 MR. EBERSOLE: What sort of periodic checks --
15 MR. CALLAN: Well, it depends on how the 16 application of the check valve,.the range of tests would go
- 17 from Type A valve, which is the kind of valve that has to 18 be leaktight, and that would require the most elaborate
! 19 form of tests. You have to measure the leakage past the 20 check valve.
21 MR. EBERSOLE: I am not interested in that.
22 MR. CALLAN: Okay. The simplest kind of a test 23 would be the kind associated with the Type C application, 24 which is a check valve that you simply have to demonstrate
. 25 that it seats.
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6M6
06000101 106 marysimons
(
1 MR. EBERSOLE: How frequent is that?
2 MR. CALLAN: It depends on the licensee's 3 commitment to the ASME section 11 testing requirement, but 4 typically for check valves like that, Type C check valves 5 would be once per refueling outage.
6 MR. EBERSOLE: Once a fuel outage. For a
- 7 section 11?
8 MR. CALLAN: For that kind of a check valve 9 test, or sometimes it wjl1 be'--
10 MR. EBERSOLE: Every three years.
11 MR. MICHELSON: There is a certain sample of 12 them every three years, and you get around to all of them -
13 every -11 years or something like .that, is it? These check 14 valves perform no isolation. ,
15 MR. EBERSOLE: That is plenty of time for them 16 to rust solid.
17 MR. CALLAN: That's why I threw in the caveat 18 about whatever the licensee is committed to, oftentimes 19 each licensee has a customized tech spec commitment to the '
i 20 code.
21 l MR. EBERSOLE: We will find out when the 22 licensee speaks later.
23 MR. WARD: Any other questions for Mr. Callan? ,
24 [No response.] '
25 Thank you very much.
ACE-FEDERAL REPORTERS, INC.
M-347-3XU Nationwide Coverage 800-336-6M6
1 06000101 h 107 marysimons l( ):
1 MR. CALLAN: Sure.
2 MR. WARD: Mr. Diab, before we go back to you, i
3 I would like~to give Mr. Enos of Arkansas Power and Light' 4 a chance to say a few words.
5 Mr. Enos? l l 6 MR. ENOS: Yes.
i 7 MR. WARD: Ted, we had a couple of questions
- i. O that are hanging, too, which we would like to address to i 9 you after you make the comments you would like to make.
10 MR. ENOS: My name is Ted Enos. I.am the >
- 11 Manager of Nuclear Engineering and Licensing for AP&L. I- ;
12 Would make just a few comments, and I do appreciate the i
13 opportunity to speak, particularly since I wasn't on the !
r .
14 agenda. But I will kind of work backwards in the few
, 15 comments I would like to make and some of the things that ,
16 We have gone through here.
17 I think, with regard to the check valve issue, -
l l 18 l' you gentlemen have just gone through sort of the dilemma t .
19 that we went through in developing that issue, and where 20 that came from. Essentially, what we came down to was i k
- 21 competing issues, you have, as you appropriately noted, a l 22 very low probability event, a steam line break followed by
- I L '
23 a loss of offsite power and failure of a particular bus or L 24 a diesel. It'is important to note that the loss of offsite ;
I l 25 power must also be in there.
ACE-FEDERAL REPORTERS, INC.
l 202 347-3700 Nationwide Coverage 600-336-6M6 L, --,......_.-.___._,_._._._.__._.,..,,__..__.,_.,,._,....__-..._,-..,._._..._.,_,.._.__.j
06000101 108
(~_)smarysimons 1 Yes, sir?
2 MR. EBERSOLE: Only the failure of a bus.
3 MR. ENOS: Failure of a bus and also loss of 4 offsite power.
5 MR. EBERSOLE: How so?
6 MR. ENOS: If you have offsite power available, 7 you still have main feed pump that you can possibly use.
8 MR. EBERSOLE: Oh, and you can throttle them 9 down.
10 MR. ENOS: You have condensate pumps that you 11 can put water in the generators with and also service water 12 pumps.
13 MR. EBERSOLE: I get the picture. I was not 14 clear. The Type A pumps that he talked about.
15 MR. ENOS: That's right.- That's right.
l 16 Looking specifically at emergency'feedwater, auxiliary
[ 17 feedwater, you could go to just the bus failure, but'if you l
l 18 look at the bigger picture -- and we tried to look at this l
r 19 from the standpoint of: put water in the generator.
l 20 That's the mission. So therefore, you still have the other l
j 21 systems over here that are available.
22 It's also maybe an ancillary point that ANO-1 23 is an older vintage plant than some of the bases behind 24 design basis was different for that plant. Before the 25 steam line break analysis, loss of offsite power was not l (2) l l
l l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6 4 6 t
,_ 06000101 109 -
( marysimons 1 an assumption in the original design-basis analysis for 2 ANO-1. So that has well contributed a bit to the way in 3 which we looked at this event and looked at the necessity 4 for the check valves. We looked at loss of offsite power, 5 but it wasn't necessarily needed to be considered in our 6 evaluations to be consistent with original design basis.
7 Also, again, we have the competing thing. The 8 low proability thing from the desing-basis standpoint, the ;
9 experience that we had with the check valves on our other ,
10 units, which was a real-world, everyday reliabillty 11 probelm, and over the period of approximately four years, 12 and I think Mr. Callan went through a pretty good 13 discussion of the evolution of this system, and this is 14 about $20.million worth of modifications that we made on 15 ANO-1. I am going to step into this in a minute from the i
16 standpoint of the reliability numbers.
17 But' essentially, what we are looking at is a o
18 totally new system, talking about completely new valves, 19 new piping systems, new steam emission systems, new turbine 20 governor systems, new electrical systems. There is 21 essentially nothing of the original emergency feedwater 22 system on Unit I that still remains today. It's a brand-23 new system as of.1984. Involving such a major change in 24 the facility over a four-to-five-year period of time and 25 getting into something of a complex and competing ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coserage 800-336-6M6
I i
,- 3 06000101 110 (xs) marysimons 1 evaluation as we did with the check valves, we got into the 2 same dilemma that you did, and I thought it appropriate to 3 explain some of that that came from there.
4 There was another point, and I will accurately 5 point out that this analysis was done after the SSFI 6 inspection, but we did look at if we had this theoretical 7 break in a cross-tie between the generators, how long would 8 we have steam available to the turbine-driven pump before ,
9 we completely blew down both generators? In other words, 10 i we're not going to lose that turbine-driven pump 11 immediately; we're tying the generators together with a 12 four-inch line, so therefore the impact generator, if you O
kJ 13 ;
would, is still going to be mostly full of steam.
14 f Our analysis showed that we could probably run 15 y a the turbine-driven pump for approximately 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br />, even 16 , with those generators tied together, before we got to the k
l 17 l point where we didn't have enough steam to put enough water I
- 18 in. It's a downward spiral is what it boils down to. As l
l 19 you lose header pressure in the good generator, you can't l 20 pump in as much water, therefore you can't create quite as l
l 21 much steam and you wind yourself down. But analysis shows 22 about 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br /> under that scenario.
? I 23 l MR. EBERSOLE: What's the containment pressure l
l 24 ! at 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br />?
l i 25 MR. ENOS: This being ANO-1, which is a once- ,
- (Z) I l'
l i
I ACE-FEDERAI. REPORTERS, INC.
l l 202-347-3700 Nationside Cm erage 800-336-6646
I i
c..x 06000101 111
) marysimonsl 1 through steam generator, I don't have exactly those 2 numbers. But the steam line break really doesn't challenge 3 the containment design. We're dealing with basically ,
4 something on the order of 30-some-odd pounds against a 59-5 pound design pressure, and that's a very peaking type of 6 thing. It comes down very quickly. The additional mass 7 and energy is being released through a four-inch line and 8 is probably down in the noise level with regard to the 9 overall mass and energy that's --
10 MR. EBERSOLE: You've already built one 11 generator.
I 12 MR. ENOS: That's right.
13 MR. EBERSOLE: Through a large break.
14 MR. ENOS: That's right.
15 MR. EBERSOLE: Now you're dumping the other one 16 through a Gmall break.
17 MR. ENOS: That's riaht. You're dumping it 18 through a four-inch line. You're already past your peak 19 pressure. What you're going to be doing at that point is 20 you're going to be slowing the rate of depressurization.
-21 It should not have any impact on the peak pressure and 22 would slow your overall rate of depressurization by less 23 than --
24 MR. MICHELSON: Assuming coolers?
25 MR. ENOS: That's assuming you would have l
ACE-FEDERAL REPORTERS, INC.
i 202-347-3700 Narionwide Coverage 800-336-66:6
06000101 1 112
()merysimons 1 coolers, yes.
2 MR. EBERSOLE: Suppose you break the four-inch 2
3 line. Now you have a pressure loss in the four-inch line.
4 Can you still run the turbines on the residual pressure if' S they're ' coupled together?
6 MR. ENOS: Yes. That was the analysis that we 7 performed with if that four-inch line you had the break,
- 8 what's the best -- let me get to it here.
- ?
9 (Slide]
f 10 Okay. In the theoretical event that we had*the 11 steam line break here, wnich would leave this line open all J
12 the way back to this generator, under those circumstances I
13 the analysis would show that.we would still maintain steam 14 preassure in this generator, piping losses are such,that it
- 15 would provide sufficient head still to the turbine pump to
> 16 provide water back to this generator for about 40 hours4.62963e-4 days <br />0.0111 hours <br />6.613757e-5 weeks <br />1.522e-5 months <br />. .
s i 17 MR. EBERSOLE: Suppose I blow the four-inch 18 line after you have made the T coming down to right there.
i 19 MR. ENOS: Right here?
l 20 MR. EBERSOLE: Yes. Now tell me about the 21 physical environmental effects of blowing that down in the
- 22 context of what it does to the electrical network?
23 MR. ENOS: In the context of what it does to 24 the electrical network, that would be a high-energy line j ()
I ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 8(6336-6M6
- ~
s s
06000101 113
) marysimons
~
1 break and that point right there is outside of a 2 containment. So the effects on the plant from the high-3- energy line break and the blowdown of that line is t
4 analyzed, pipe whip and cabling, et cetera.
5 MR. EBERSOLE: Is condensation on cold contacts 6 and open electrical apparatus considered as a 7 shortcircuiting mechanism to stop electrical apparatus?
8 That's a lot of steam.
9 MR. ENOS: Yes, that's right. We look at those 10 types of things generally throughout the plant. Basically, I
11 in the areas of where these high-energy lines can impact on 12 the plant, we don't have that type of equipment. We're
('S (m/ 13 basically looking at instrument tubing and piping systems.
14 We don't have any junction boxes, et cetera. So we're more 15 concerned with jet Ampingement and pipe whip.
16 MR. MICHELSON: Where does the steam go to?
i 17 MR. EBERSOLE: No, I am talking about the l 18~ general dispersion of steam.
19 MR. ENOS: Beginning of the steam?
20 MR. EBERSOLE: yes.
21 MR. ENOS: I can't specifically address that.
l 22 The vent path of the steam is analyzed and the pressures in l
l 23 the compartments nad the euipment that are in those L
i 24 compartments.
l l
l 25 l ACE-FEDERAL REPORTERS, INC.
! 202 347-3700 Nationwide Coverage 800-3364M6
06000101 114
. ( ) marysimons l 1 MR. EBERSOLE: The submersion o,f open 2 electrical apparatus invariably leads to condensation on 3 cold contacts.
4 MR. ENOS: That's true.s The equipment, the 5 electrical equipment that falls in the area that's impacted 6 by that steam in the blowdown path, is required to be 7 environmentally qualified for that specific environment.
8 MR. EBERSOLE: For condensation?
9 MR. ENOS:
. yes, sir.
10 MR. EBERSOLE: Open electric. Is that done by 11' allowing a fraction of the operating current, leakage 12 current, to flow to ground or phase to phase, or is it done 13 by requiring you keep those things dry?
14 MR. ENOS: Now, it's done by a type-testing, an 15 environmental qualification type testing similar.to --
i 16 MR. EBERSOLE: Ch, yes. Sandia pointed out, in 17 order to pass the test for that, some of the testers 18 preheated the electrical apparatus to preclude condensation 19 on the surface contacts.
20 MR. ENOS: yes. I understand.
21 MR. EBERSOLE: Do you follow me?
22 MR. ENOS: yes, I understand what you're 23 talking about. Generally, in the LOCA-type tests, et 24 cetera, the time duration you're looking at, the blowdown 25 O
\ms ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide ( overage 8 % 336-6646
06000101 115
,( ) marysimons 1 durations, that kind of gets you beyond that point anyway.
I 2 I can't speak specifically to how that was considered, but 3 they are qualified for that environment in the same way as 4 . we environmentally qualify in-conta'inment equipment;.only J
5 two different profiles.
6 MR. EBERSOLE: Well', there was a period of time 7 when NRC -- I think,. erroneously -- permitted partial 8 leakage currents as long as it didn't defeat the functional 9 percentage of operating current doing its thing. I think 10 that's a style that has passed -- Watts Bar is a case in 11 point -- in favor of hermetically sealing and protecting 12 the splices and terminal blocks, et cetera, which will leak 13 into ground or phase to phase.
14 MR. ENOS: Right.
15 MR. EBERSOLE: It would be interesting to know 16 what your philosophy is, whether you're adopting the l
17 leakage path rationale or not. .
18 MR. ENOS: I can't really speak to that, I am 19 sorry. Withg regard to splices, of course, we're using 20 hermetically sealing splicing terminal blocks. Qualified
- 21 terminal blocks would not necessarily sealed on the 22 terminal blocks; mostly a phenolic-type terminal block.
23 MR. EBERSOLE: Which do condense.
24 MR. ENOS: Which do condense, yes.
I 25 MR. EBERSOLE: Then evidently, you're in'that
- (:)
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000101 116
( ) marysimons 1 realm of plant design that takes the leakage current 2 rationale and tries to retain a sufficient mil 11 ampere flow 3 to do whatever it has to do.
4 MR. ENOS: Will it function after the test, and
.5 then beyond that you get me off into a EQ a little further 6 than I can go.
7 -MR. EBERSOLE: All right.
8 MR. ENOS: I wanted to make that point with 9 regard to the analysis. I think, as Mr. Callan pointed 10 out, during the inspection we looked at this for the 11 plant. We had to shut the plant down for some problems we 12 were having with diesel generators during the inspection.
13 We chose to go. ahead and install the check valves.
14 The question on ISI, I am not sure of the 15 actual ISI requirements with regard to those valves. Our 16 particular program requires them to be inspected once per 17 refueling outage. So we will -- our ISI program -'- will 18 get those check valves once.per refueling outage.
19 MR. WARD: Once per shutdown?
20 MR. ENOS: I am sorry?
21 MR. WARD: Once per shutdown inspection? What 22 is the inspection?
23 MR. ENOS: It's a leak test and pressure test.
24 MR. EDERSOLE: What about the dynamic 25 competence that we talked about?
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3364/>46
06000101 117
()marysimons 1 MR. ENOS: I am going to have to check that 2 point with regard to -- actually, we plan on~ changing these 3 valves out at the next refueling outage to a different 4 valve which we think will be -- a different check valve --
5 which we think will have a higher reliability.
6 MR. WARD: It'a.the same question.
7 MR. ENOS: yes, the same question applies 8 there. The changeout is being made speficially because of' 9 this concern to a valve that we haven't seen this type of 10 problem with before. I will have to check my facts on.it, 11 but'I believe that we will actually be going in and looking 12 at these vballves each refueling outage because of the 13 problems that we have seen on ANO-2.
14 MR. EBERSOLE: What is in your procurement.
15 specifications on these valves in the context of 16 guaranteeing dynamic performance, and how are you validate 4
17 that you've got it, by test or analysis or whatever?
t 18 MR. ENOS: I can't speak to that. You know, a
-19 Q valve, with regard to that specific aspect, I can't ,
20 address. We're fully aware of the concern that you're L
l 21 talking about with regard to these and particularly with 22 regard to San Onofre and those issues.
23 We are participating -- I think you are 24 probably aware that there is a'multiowners group effort
()
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide coverage 80 4 336-6646
06000101 118
( ) marysimons 1 that's involved in working with INPO to look at the San 2 Onofre check valve problem and to come up with some 3 performance standard and evaluation criteria on how to 4 address these. One of the people at AP&L is working 5 directly on that committee and working in developing that 6 criteria. So we are very much aware of that, and that's 7 the direction we will'be going with regard to addressing 8 those in the future.
9 MR. MICHELSON: With regard to the steam line 10 on the auxiliary feedwater system, if you have a rupture of 11 the steam line in the local compartments somewhere, the 12 intent is for the steam line isolation valve, I guess, to 13 close, using temperature switches on the ceilings or 14 whatever. You detect the break, in other words, and you 1
15 close the valve. In that correct?
16 MR. ENOS: No, we don't have that design in our 17 facility.
18 MR. MICHELSON: You don't have any provision j 19 for automatic isolation?
20 MR. ENOS: No. As a matter of fact this l
! 21 particular line break is within the capacity of the normal l
l 22 feedwater water control system, so if that high-energy line t
23 break were to occur, the feedwater system would just adjust
(
24 for it and the plant would continue to runm.
25 MR. MICHELSON: How do you handle the f
I l
ACE FEDERAL REPORTERS, INC.
\_ -- - -. - . _ - . __ ._ - - - -
06000101 119
( ) marysimons 1 continuing passage of steam into the compartments?
2 MR. ENOS: We have analyzed that for a period 3 of one hour, looking at the flow of steam through the 4 compartments, impacting on the --
5 MR. MICHELSON: Unrestricted?
6 MR. ENOS: Unrestricted for one hour, and rely 7 upon operator action --
8 MR. MICHELSON: Have you traced where all the 9 steam will go to?
10 MR. ENOS: Yes, sir.
11 MR. MICHELSON: It doesn't give you any problem 12 with a safe shutdown capability?
b N/ 13 MR. ENOS: Yes, sir, that's correct.
14 MR. MICHELSON: Now, this is an interesting 15 design, indeed. Can you tell us just a wee bit more about 16 where the steam-driven turbine is, and steam lines, 17 relative to other safety-related equipment?
18 MR. ENOS: I am sorry, I can't.
19 MR. MICHELSON: Is it in its own chase and its 20 own compartment?
21 MR. ENOS: Yes.
22 MR. MICHELSON: Is it that anything that enters 23 it is vented to atmosphere and not to other parts of the 24 building?
_ 25 MR. ENOS: It has its own room. We have a V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
06000101 120
( ) marysimons I separate emergency feedwater room,-and for every place that 2 these lines can rupture within~the auxiliary building, we 3 have defined the path for that steam out as to where it's 4 going to go. As a matter of fact, in some cases we've even
, 5 made modifications to the facility to allow a steam flow j 6 path out.
7 MR. MICHELSON: ou'll have to do a little more 8 than define. You have to make sure it's blocked. This 4
9 means that there is no ventilation ducts or anything like
, 10 .that in these rooms that can vent the steam out elsewhere
> 11 in the building?
12 MR. ENOS: No. We have dampers that should 13 take care of that problem for us from the standpoint of --
14 MR. EBERSOLE: You have got ventilation dampers l
15 that --
16 MR. ENOS: Yes.
17 MR. EBERSOLE: -- are going to stop this 18 hypothetical steam flow?
19 MR. ENOS: For the most part, they should, yes.
20 MR. EBERSOLE: What compartment pressures are.
i 21 we talking about, roughly?
, 22 MR. ENOS: One, two pounds.
I 23 MR. EBERSOLE: One to two pounds?
24 MR. ENOS: Yes. If memory serves me right, 25 something on the order of 120 -- I am being told two ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646 l
06000101 121
) marysimons 1 pounds -- it's about 120 or 140 degrees, somewhere along 2 there. Closer to 200 degrees.
, 3 MR. MICHELSON: It would be all of 200 --
4 MR. ENOS: Yes.
5 MR. MICHELSON: -- because you're flashing a 6 thousand pounds --
7 MR. ENOS: Yes, that's right.
8 MR. MICHELSON: -- of steam into the room 9 through a hot line.
10 MR. ENOS: Very quickly it comes into --
11 MR. MICHELSON: It would be well over 200 12 degrees, unless you've got a lot of heat sinks for one 13 hour. That means heat sinks are pretty well gone.
14 MR. EBERSOLE: You've got big galvanized sheet 15 metal, tin ductwork, with swinging. dampers to preclude 16 backflow of steam, say, into the control room?
! 17 MR. ENOS: But there's not a path from these 38 areas to the control room.
19 MR. MICHELSON: You have no ventilation in this 20 room?
i 21 MR. ENOS: There is a room cooler in the room.
22 MR. MICHELSON: And no ventilation?
23 MR. ENOS: There is ventilation on the floor, j H24 but what we're looking at is the vent paths that we've 25 provided for the thing, and we may get some amount that's ACE-FEDERAL REPORT 8RS, INC.
h 202-347 3700 Nationwide Coverage 800-336-6646 l
_ _ _ . . . _ _ . _ . . _ _ - _ _ _ - . , , _ _. ~ _ _ _ _ , - . _ , . _ , . . _ _ _ _ _ . _ . .-. _ . _ .
S 06000101 122
( ) marysimons 1 going to get into the ductwork, et cetera, but it's going 2 to-go the path of least resistance is where the majority of i
! 3 the steam flow is going to go. It's going to be -- if the 4 ventilation systems are on -- it's going to be from feeding 5 against positive pressures there -aus opposed to a less 6 resistant path as it goes up stairways and that sort of 7 thing.
8 That.'s what we've analyzed, and what we've 9 found for the thing is: give it a low resistance path out l
10 of the building, be it up stairways or through doorways, et 11 cetera. That's the path that we look at from the 12 standpoint of qualifying the equipment.
13 MR. MICHELSON: Two pounds is a pretty 14 substantial pressure. It will drive steam in all
. 15 directions. Of course, there's no preferred direction;.any ,
- 16 path that's available to it, it will'--
17 MR. ENOS: Right.
18 MR. MICHELSON: -- be pumped out with a two-l 19 pound driving force.
l
[
20 MR. ENOS: Now, this was not -- and I want to l
21 . point out that this wasn't something we sort of glossed our 22 ' hands over. I mean, the auxiliary buildings were nodalized
~23 in a computer model and the temperature / condition profiles 24 of pressures, et cetera, were provided into that. This was 25 done by computerized l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-(M6
- , . . . - _ . , . . . . _ . . . - . - ~ . . . - . . . - . _ . - . . - _ . . . . . _ , _ . - , - . _ . . - - _ . _ . . . . - _ . . , - , - . - , . . _ . .
123 marysimons
()06000101
~1 study of looking at compartment pressures, et cetera.
1 2 MR. MICHELSON: I have seen such studies, and 1 3 the problem is that they don't generally include things
'4 like the holes where the ductwork is coming into the room 5 and so forth. They're highly idealized.
6 MR. ENOS: Appendix.R is supposed to take care 7 of those.
8 MR. MICHELSON: Appendix R has nothing to do 9 with this question. It.'s another question entirely.
10 MR. EBERSOLE: Do you keep the doors locked, 11 and which way do they swing into these rooms?
12 MR. ENOS: The rooms down are locked and r
13 figured into the analysis are which doors are locked, i'
14 required to be locked, and in some cases we have doors with 15 special latches on them which will give for certain high-16 energy line breaks, .25 pound delta Ps and that sort of 17 thing.
- 18 MR. EBERSOLE
- Which will give?
19 MR. ENOS: Yes.
20 MR. EBERSOLE: And then what results from an 21 environmental standpoint when they give?
22 MR. ENOS: Well, it's designed to do that. In 23 other words, if we have a high-energy line break in th a 24 room, we want that door to open to let the steam go out 25 that way. So we design a lock on that door that will give l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000101 124
()marysimons 1- and allow the door to open out with a .25 pressure.
2 MR. EBERSOLE: How about is there is a 3 switchboard on the other side of the door?
4 MR. ENOS: That's part of the steam path of
, 5 where it's going to go, and that's analyzed as well. We ,
6 follow it all the way out the door.
7 MR. EBERSOLE: Including the condensation
(
8 phenomenon?
9 MR. ENOS: Right.
10 MR. EBERSOLE: Okay.
11 MR. MICHELSON: Are you environmentally 12 qualifying that equipment, or how'are you treating it?
13 MR. ENOS: Environmentally qualified to those 14 service conditions.
15 MR. MICHELSON: In other words, you're not 16 allowing the steam to enter any pieces of equipment?
17 MR. ENOS: That's correct. Well, not l 18 necessarily we're not allowing it to enter those pieces of i
19 equipment, but let's say if there's a limitorqueor S&B 20 operator sitting out there on the valve --
21 MR. MICHELSON: Let's take the switchboard, 22 which is a far better -- switchgear or electrical control 23 or inverter.
24 MR. ENOS: Now, I don't believe that we have 25 any switchgear that's in a vent path. So we haven't had ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-(M6
06000101 125
( ) marysimons 1 to look at that.
2 MR. MICHELSON: It'd be an interesting study to 3 look at.
4 MR. ENOS: A couple of other comments I would -
5 make if I can move on with respect to some of the numbers 6 that were presented earlier on the reliability numbers, 7 particularly the numbers that were presnted for ANO-1 and 8 ANO-2.
I 9 We have undertaken a rather extensive study of' i
j 10 the reliability of these systems, particularly on a 11 component basis and on a challenge basis. Not meaning to 12 contradict some data that has been previously put up here -
+
13 -- I am going to do that anyway, but it's probably because 14 you all don't have the benefit of that data, and perhaps 15 that's my mistake.
16 We have looked at in our facility, each time
- 17 anything goes wrong in a plant, if a valve doesn't respond 18 properly or whatever, we generate what's called a " Report 19 of Abnormal Condition," or a RAC. We have research from 20 1980 forward, every RAC that has been generated on the 21 facility. We have also researched all of the operator logs 22 to determine every demand of the emergency feedwater 23 systems or auxiliary feedwater systems.
24 From that we have generated component 25 reliability rates which, based on information that we have ACE-FEDERAL REPORTERS, INC.
j 202 347 3700 Nationwide Coverage 800-336-6646
126 marysimons
( ) 06000101-1 found from INPO in their industry study, we find that the 2 component reliability rates of both the Unit 1 and the Unit 3' 2 system is about a standard deviation better than the j 4 industry average that's been calculated by INPO.
].
5 I think the point _that was made earlier with 6 regard to the timing of the failures that we have seen at-T 7 the units is very valid. Everything, the failures that we 8 have had at these units for 1981, 1982, and '83, looking at 9 those time frame failures on ANO-1 is very misleading 10 because the system that is there now is.a totally different 11 system.
I 12 The failure rates that we have seen on Unit 1 13 since January of 1985 is we have had no failure of a pump i 14 to start on demand; we have had no failures of components 15 in the system that would interfere with -- by the way, let l 16 me explain, we define " failure" as "a failure.of a i 17 component in such a way as it would prohibit the capability l
18 of putting water in the steam generator."
19 Now, we can take a valve failure. The valve
- 20 has to open to put water in the generator, and later on i
21 when we got ready to shut the system down, it wouldn't 22 close. That's a problem, but we don't consider that a 23 failure from an EFW reliability standpoint if that problem 1
i 24 was specific to the valve closing. We want to put water in I' 25 the generator, so we want to know what is the I i
1 ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-M*6
?
, 06000101 127 l ( ) marysimons 1 reliability of the~ system to put water in the' generator. '
2 Therefore, you have to look at those things that impact 3 that reliability.
! 4 So we find on ANO-1 from January 1985 forward 5 we have had no failures of pumps or valves or motor drivers 6 that would impact the capability to put water in the 7 generator.
8 On ANO-2, I guess, we disagree a little bit on 9 the numbers that were presented. Certainly, we did have 10 problems with those flow control valves. I had a number of 11 18 failures of flow control valves in the period '81 12 through '84 for ANO-2. We have not, in '85 and to date in
^
13 '86, we have had no failures of pumps or motor drivers or i
j 14 turbine drivers to provide water on demand. We have had 15 one failure of a steam emission valve, which we finally 16 found the problem during this most recent refueling outage i
17 that we had during some of our work on 85-03, Bulleting 85-l 18 03 that Mr. Callan talked about earlier. I believe we have i
19 that problem corrected.
20 (Slide]
I 21 So I would point out that from our viewpoint e
i 22 with regard to looking at the failures of what kept us from 23 putting water in the generator -- and I have a little graph I
24 that I think is indicative; I hope you can see it from ;
25 there -- that's the component failure rate that we v
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6M6
06000101 128
- marysimons
(
1 calculate on ANO-2, this being 1981, '82, '83, '84, and 2 this was through June '85. That number will be up just a 3 little bit because late in '85 we did have one valve
- 4 failure.
5 [ Slide]
6 That shows the component failure rate in the 7 '84 '85 '86 time frame as somethign less than .0001 per 8 demand per component in the system.
9 MR. MICHELSON: May I ask you have you reported i 10- all failures of auxiliary feedwater components to the NPRDS 11 system? In other words, is the data base complete in that ,
'12 regard?
~/ 13 MR. ENOS: Well, the data base is probably i
14 overly complete.
15 MR. MICHELSON: How could it be overly 16 complete?
17 MR. ENOS: Well, NPRDS was established to 18 provide the industry a means of knowing what problems were 19 happening at other places so they could take a look at do .
20 we potentially have that problem there? There is reporting i
! 21 criteria set up in the NPRDS, in what things you put and l 22 what things you don't. In fact, in the spirit of trying to 23 keep others informed of equipment problems that we were i 24 having in trying to build the data base with NPRDS, we have 25 reported things on that system that are not l
O ACE-FEDERAL REPORTERS, INC.
202-347-3XU Nationwide Coverage 800-33MM6
... .._-_.___..--_.__,,._____._....,m_. _ , , _ _ _ . _ , _ , , , , , . _ _ . _ _ _ , _ . . _ . _ . , , _ , , _ , _ _ _ _ _
4 06000101 129 l ()marysimons 1 required by the criteria to be reported.
2 MR. MICHELSON: Well, the answer to my question 3 -- I would like an answer to my question -- these failures 4 you're talking about since January of '85, have all of
, 5 those been reported to NPRDS, do you know?
6 MR. ENOS: Yes, they have, in answer to your 7 question.
8 MR. HICHELSON: So if I go to the NPRDS data 9 base and pull off all ANO-1 and all ANO-2 valve failures 10 for auxiliary feedwater, I have a complete set?
11 MR. ENOS: Right. You should have it. Now, 12 one would need to keep in mind that the criteria by which i
s 13 you call something a failure on NPRDS and the criteria by i
14 which we do reliability analysis and consider failures 15 there is different. So the numbers of failures may be 16 different.
17 MR. MICHELSON: In which direction?
18 MR. ENOS: NPRDS should be higher. If you say 19 that --
20 MR. MICHELSON: Okay, but all the failures 21 you're talking about then indeed are on NPRDS.
22 MR. ENOS: They should be, yes.
23 MR. MICHELSON: Yes. Okay.
24 MR. ENOS: Okay. There should be -- in our 25 reliability evaluation -- there should be failures i
' ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 85336-6M6
l l
130 marysimons
()06000101 1 reported to NPRDS that we did not consider in our 2 evaluation because they did not impact the capability of 3 putting water in the generator.
l 4 MR. MICHELSON: Yes. I can understand that. .I i
'S just wanted to get one benchmark point. At least ANO-1 and
, 6 ANO-2 is presumably a complete set on NPRDS since at least j
l 7 January of '857
, 8 MR. ENOS: That's my understanding. I will 9 have to go back and talk to the NPRDS folks and make sure 10 that's the case. But it should be.
i 11 MR. MICHELSON: At least it's one data point.
, 12 MR. ENOS: Right.
i 13 MR. MICHELSON: Now if we could just ask every 1
14 other utility the same question and get the same answer, 15 then we'd have something we could start to work with and
. 16 that would be another way of getting an idea of the l 17 completeness of that data base.
i 18 MR. ENOS: Another thing that we found in going l 19 through that reliability evaluation, again we sit people in -
I ;
- 20 a room and handed them the stacks of operator logs since t
!. 21 day one of the operation, in which we did a nice job of 22 covering up the table out here, and we found the demand 23 rate on the systems, when you consider all of the testing, i 24 surveillance testing, et cetera, and the demands on i
25 components to be a fairly substantial rate. We are i
i ACE-FEDERAL REPORTERS, INC. i 1
i 202-347-3700 Nationside Coscrage 800-3364M6
, ~ . . . _ - _ - _ _ . ., - _ _ ,,, _ _ _ _ . _ _ . _ _ _
l 06000101 131 I marysimons
-1 actually looking, all total, at something like a hundred 2 plus demands per year. If you take into consideration that ;
3- you're going to hit each pump a minimum of 12 times just 4 through monthly surveillance. So there are 24 shots at 5 your two pumps during the course of one year. You may do 6 the test several times when you're down.
7 You may do maintenance on it. Some of the 8 pumps, like in the case of ANO-2, the system is used for 9 startup and shutdown, so it has a number of demands on it-10 for startup and shutdown. You have additional surveillance 11 testing in the 18-month interval, et cetera.
12 So the numbers'of demands are fairly 13 substantial, and it's very difficult to dig that out, but 14 it took us about three weeks going through that data to 15 pull all of that out. But we find that once you have 16 actually pulled that out, that is where you get component 17 reliability numbers in the ten to the minus three to ten to 18 the minus four range from actual data.
19 Let me go to I think a real early point this 20 morning that was asked on the difference between Type A and 21 7 plants. I think Mr. Ward you asked the question 'iri the 22 operations of these systems.
23 There are some significant operational 24 differences. For example, on ANO-2 the emergency ACE. FEDERAL REPORTERS, INC.
202-347 3700 Natinnwide Coserage 800-336 6646
- - - ~
l 1
l l
06000101 132 l
()marysimons i i feedwater system is used for startup and shutdown of the 2 unit. Also from the control systems standpoint on the loss 3 of one main feedwater pump the reactor trips which is going 4 to also generate a demand on emergenc, feedwater.
5 I would contrast that with ANO-1, which has a 6 third startup pump, and that startup pump is what is used 7 for startup and shutdown of the plant and not the emergency 8 feedwater system.
9 Also on the control systems standpoint of i
i 10 things on the loss of one main feedwater pump the
! 11 integrated' control system will run the plant back to 50 12 percent and keep the plant on line using the second main 13 feedwater pump. Therefore, you don't get a challenge to
- 14 EFW on-that system where you would get a challenge to the 15 to the rest of your systems.
16 So there are significant differences with i 17 regard to what generates the challenges between the plants.
18 MR. EBERSOLE: you are the very edge now where 19 I think I need to ask that question. Do you know of a 20 package that has justified this four-inch primary loop 21 discharge evidently put there because of your very lack of 22 concern for this inflexibility on this CE design that you 23 did have on ANO-17 24 MR. ENOS: yes, j
()
ACE-17EDERAL REPORTERS, INC.
202 347 3700 Nation *;de Coscrate 8M 316W46 J
f 06000101 133 marysimons
(
1 MR. EBERSOLE: Can that piece of paper be 2 furnished to us?
'3 MR. ENOS: Do you have a book laying there that
, 4 looks like this?
. 5 MR. EBERSOLE: I sure do. Is it in there for 6 goodness sakes?
7 MR. ENOS: If you will flip over about, oh, 8 under the ANO-2 one if you would flip over about 12 pages 9 in that book, it ought to be one that.looks like this.
10 MR. EBERSOLE: But where is the narrative 11 support?
12 MR. ENOS: I will give it to you right now if x 13 you want to take just a couple of minutes, or I will give 14 it to you later.
I 15 MR. EBERSOLE: I would like to have it written.
16 MR. WARD: Well, go ahead and tell us what you 17 have to tell us.
18 MR. ENOS: Okay. On ANO-2 there is a-flange on 19 top of the pressurizer which was originally intended to be 20 for a PORV and was not installed. In the early licensing 21 days a concern arose with regard to boron precipitation in 22 the core and two valves were stuck on a stem of pipe on top 23 of the pressurizer to be opened to allow through flow 24 through the core to eliminate boron precipitation.
25 That went away. Before ANO-2 got a license we ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationside Coverage 800-336-6646
+
t t
- i 06000101 134
()marysimons ,
1 went to split' hot and cold leg injection to take care of l l
2 the boron precipitation problem. We left the ECCS vent-l 3 line there as a last ditch. It is in the procedures and it 4 is there in case primary pressure hangs up at such a point 5 that the HPSI pumps cannot inject enough water into the 6^ system that we can blow down the primary.
7 MR. EBERSOLE: That was in parallel or series?
8 MR. ENOS: They are in series.
9 MR. EBERSOLE: And they are remote motor 10 operated?
11 MR. ENOS: They are remote motor operated, 12 saf ety grarle ---
O 13 MR. EBERSOLE: Gate valves.
f 14 MR. ENOS: Gate valves and we have done 7 15 analysis on them to demonstrate that they are capable of 16 removing full decay heat. -
4 :
- 17 MR. EBERSOLE
- Well early on are those lines *
[
18 big enough to take away the heat by sensible heat -
l 19 transport, or do you not have to revert to a mixed phase !
t
- 20 flow?
l I 21 MR. ENOS: What you should be able to do if i
( 22 the anlaysis is correct is open those valves and turn on !
l f 23 HPI pump and you are done. That will take care of it. l t
24 MR. WARD: You don't have to have boiling --- ;
25 MR. ENOS: The boiler condenser mode? Well, ;
f
~
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Cmerage 800-33MM6
, . _ - - - - . , - ..,_.,.,m.-.____.---
. . . . . . . _ . ...-. .- = . . . . . _ . . . . .
135 06000101 (fmarysimons i 1 you may go to a saturated condition in the core. I am not >
4 2 sure on that point. you probably will at some. point. That 3 is a pretty good size --- ,
I 4 MR. EBERSOLE': What do you call a gpa flow i'
S rate? ,
, 6 i MR. ENOS: A gpa flow rate is about 500 gpa.
- 7 MR. EBERSOLE
- I don't think that will take 8 away in sensible heat the initial decay energy.
I 9 MR. MICHAELSON: For 500 gpm?
10 MR. ENOS: yes.
11 MR. MICHAELSON: Sure.
t 12 MR. EBERSOLE: In just sensible heat?
- 13 I MR. MICHAELSON: Do you mean in sensible heat ;
s 14 or decay heat?
]
I 15 MR. EBERSOLE: I said sensible heat. ;
16 MR. MICHAELSON: Well, what do you mean? [
I 17 MR. EBERSOLE: The warm-up of water and the 18 discharge of liquid water.
19 MR. ENOS: Right. >
f 20 MR. EBERSOLE: Three are not many BTUs in I 21 liquid water heated up from the inlet condition to the l
22 discharge. you have got to go through a phase change to :
l 23 get heat out.
i
! 24 MR. ENOS: Right. But at the time you get to ;
l 25 this point, analytically at the time you get to the point i ' '
I i l
b a
! ACE-FEDERAL REPORTERS, INC. [
j 202447 3Rx) Nationw Je Coverage 8(t)-33&&46
f, '06000101 136
( ) marysimons 1 of opening these valves you are going to be completely 2 voided in the pressurizer and probably getting down close 3 to the hot legs. So you most likely are going to be 4 discharging steam. As a matter of fact ---
5 MR. EBERSOLE: Well, that is what you didn't 6 say.
7 MR. ENOS: I'm sorry. What you have is 8 essentially a small break LOCA at the top of the 9 pressurizer and behaves like that classical type of 10 anlaysis.
11 MR. EBERSOLE: You are going to have some 12 thunderous percolator action. !
O.
13 MR. ENOS: That's right.
14 MR. WARD: And you have got a steam / water 15 interface down in the vessel then?
16 MR. ENOS: Yes. It will be somewhere in the 17 vessel and maybe even up above. '
18 MR. EBERSOLE: If you could hold it at a placid t i 19 comfortable level you could just convert it to what it l
20 should have started at ---
21 MR. MICHAELSON: you don't have any vessel i 22 level indicators, do you?
! I 23 MR. ENOS: We will have. t
~
i 24 (Laughter.)
l 25 Well, as a matter of fact, we do now.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Cowerage M10 31MM6
}
06000101 137
-[).marysimons 1 MR. MICHAELSON: Oh, you are going to have 2 them?
3 MR. ENOS: We are.just coming back up from'a 4 refueling outage. As a matter of fact, it should have gone 5 critical last night. We installed the reactor vessel level 6 indication ---
7 MR. MICHAELSON: Are.you the only B&W type 8 plant -- are you talking about ANO-17 9 MR. ENOS: ANO-2 just came up from a refueling 10 outage. ANO-1 just went down and it will have the l 11 instrumentation when it comes up, yes, sir.
12 MR. WARD: This is what, heated junction.
13 thermocouples?
14 MR. ENOS: Well, this is our own brand of --
, 15 this is the TEC system I think that you may have heard of.
16 We call it the gamma thermometer system. It is somewhat- T 17 similar to the other one that you --- ,
l 18 MR. EBERSOLE: Now it is not hot leg level. It f l 19 is vessel level.
20 MR. ENOS: It is vessel level, yes.
i 21 MR. EBERSOLE: You are in high cotton. From ,
22 Arkansas you know what that means.
1 23 MR. ENOS: Yes, talking about high cotton.
- i 24 (Laughter.)
i >
. 25 I know what you do there, too.
l I ACE FEDER.AL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336-6646
06000202 139 marysimons 1 HR. WARD: Let's see, are there any remaining-2 questions that we need to ask? It seemed that there was 3 one other thing remaining.
4 MR. MICHELSON: Yes, I want to ask, you are 5 aware of course of the staff's report on the safety system G functional inspection?
7 MR. ENOS: Yes.
8 MR. MICHELSON: And I am sure you have seen a 9 copy and reviewed it. Do you agree with the observations 10 or is there a rebuttal to that that says these observations 11 are not true as it relates to auxiliary feedwater?
12 MR. EBERSOLE: We have provided a response to i
) 13 the NRC on the three category A items, one of which is the 14 check valves. We had a meeting with the region on 15 particularly the check valve issue, which I think Mr.
16 Callan pretty well recapped what our presentation was 17 there, and I think I gave you the sense of that as well.
j 18 There was the issue on seismic category 2 over l 19 one which we have responded to the staff and they have l
20 acknowledged that that is a generic issue and have pulled
! 21 that back as a plant specific item into the generic arena.
l l 22 MR. MICHELSON: The one I have in mind just as l
23 a typical, although I have a number and I can't get into 24 them, is the need for room cooling in the auxiliary 25 feedwater room. You do not have a cooler and apparently ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nation ideCoverage Mn-33MM6
06000202 139 ;
} marysimons 1 the calculations weren't performed to show that the system 2 you have can function without cooling, and I was wondering 3 what your rebuttal to that particular problem was.
4 MR. ENOS: I.am aware of that and I do riot 5 believe th'at we have completed the evaluation on the t
6 necessity of that.
7 MR. MICHELSON: You haven't necessarily 8 rebutted to all of the observations?
9 MR. ENOS: No. We have only responded to the 10 three items in the category A. We have not responded to 11 the rest of the report.
12 MR. MICHELSON: There were some questions
() 13 raised also relative to the calculations of the environment 14 for your environmental qualifications equipment relative to ,
15 these various auxiliary steamline breaks.
l 16 MR. ENOS: Right. ,
17 MR. MICHELSON: And of course that could have 18 some bearing on the final conclusion.
- 19 MR. .ENOS
- Subsequent to that, and this seems 20 to have been our year for inspections this year, we have 21 also had our environmental qualification inspection
! 22 subsequent to the SSFI, and I think the items of 23 temperature profiles were resolved during that inspection.
24 MR. MICHELSON: But I asked you a general 25 question on the SSFI, and that is what is your feeling or i
J l
- ACE-FEDERAL REPORTERS, INC.
202 347 3701) Nationwide Coserage M)-3 %-6646
i 06000202 140 marysimons
{
1 reaction or comments concerning its goodness or whatever?
2 MR. ENOS: Well, one way that I would certainly 3 describe the SSFI is thorough. It is that. I think there 4 are definitely merits to the SSFI program, and I wish we 5 could rename it because I cannot say that acronym.
.l 6 (Laughter.)
7 But I think there definitely are merits to the
[ 8 program. Certainly-it is a look at our systems and a look 9 at our programs in a way that I don't think they may have 10 been looked at before and that has value.
11 I think on the downside of the program it was 12 extremely resource intensive from our standpoint. We
() 13 estimated somewhere between two to three man-years of 14 effort to support that inspection, and that is a lot of 15 resource to go to that, i
l 16 MR. WARD: How much did the staff put into 17 that?
i 18 MR. ENOS: I don't know. There must have been
! 19 four or five hundred of them there at the site.
t
- 20 MR. MINNERS: I forget what the numbers are,
\
21 but there are man-years.
, 22 MR. ENOS: Yes. It is intensive.
l 23 MR. WARD: I am trying to figure out what is 24 the ratio between the licensee and the staff.
. 25 MR. MANNING: I don't know whether it was or O
, \>
i ACE-FEDERAL REPORTERS, INC.
202 Nationwide Coverage 810-UMM6
_- 347 3 700 . , - . . - , . - - , - - - . - .
06000202 141 ,
~ marysimons I not. .It said tnis special announced team inspection ,
2 involved 756 inspection hours to perform the in-depth ;
3 assessment. I don't know if that is everything they did or 4 not, but their number is 756.
5 MR. ENOS: I wouldn't be surprised to see a 6 number ---
7 MR. WARD: The licensee has to spend more time, 8 more man-years than the staff does.
9 MR. ENOS: The part of it that is difficult is 10 in ANO-1 you are dealing with a plant that was designed in-11 the-late 60's. So we are retrieving information that is 15 12 and 20 years old, and you have 15 people doing an
() 13 inspection and they can request a lot of information in a 14 very short period of time. It is very difficult to 15 retrieve design information that is that old and put things 16 together. So it becomes a real effort to support that.
17 MR. DAVIS: Does to the two to three years 18 include responding to the report?
19 MR. ENOS: Two to three years is what we -- not 20 responding to it in the way of follow-up design actions or 21 implementation, but two to three man-years basically in 22 supporting that effort and follow-up with paper responses.
23 MR. DAVIS: Do you have any rough idea how much 24 money we have spent in reviewing that?
25 MR. ENOS: On ANO-1 approximately $20 million.
O ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage RM3366646
. . .- - - ..- . .. _. _= - .- - .- --
/
06000202 142 marysimons
- 1 It has been a lot of money. On ANO-2, I think we. point 2 out, as I think Mr. Callan did, that that was the first-3 system that was licensed to APCSP10-1. It was a full 4 safety grade system when it was licensed and in effect met 5 all of the post-TMI requirements as of the date of'its 6 licensing. The modifications that we have made on that 7 system had been reliability modifications, . turbine governor 8 controls, et cetera.
4 9 MR. EBERSOLE: Dave?
10 MR. WARD: Yes, Jesse.
11 MR. EBERSOLE: I would like to make an 12 observation which I hope you would consider. I think you ,
() 13 are on delicate grounds if you are environmentally
, 14 qualifying-apparatus with the permission of NRC to use the 15 thesis that it can be. exposed to condensation, leakage, et 16 cetera.
17 The reason for this is, as academia loves and 18 regulators sometimes accept, the experimenters conducting,
~
19 you know, running the tests almost invariably use something 20 like still water with absolutely clean contact surfaces 21 probably widely spaced, I'm not sure. In real life the 22 condensation process takes place on dust covered apparatus, 23 contaminated water and entirely different physical 24 conditions than the laboratory condition under which it 25 says it works great. You may find that too late.
O 8
ACE-FEDERAL REPORTERS, INC. ,
202-347 3700 Nationwide Coverage 800-33MM6
06000202 143
} marysimons 1 I think I am correct in the fact that a 2 conservative design or even a reasonable one to come down 3 on several-levels will invariably adopt the hermetically 4 sealed approach rather than the partially failed approach 5 design of operational logic.
6 To design toward a partial operational failure 7 is a fundamental flaw in logic in the first place, to allow 8 something to partially fail and invoke the residual as 9 being enough.
10 Are y'ou with me?
11 MR. ENOS: I understand, yes.
12 MR. EBERSOLE: I think maybe you might look~at
() 13 that and see a lot of money.
14 MR. ENOS: I appreciate your concern and I will 15 discuss that with our people when I return.
16 MR. WARD: Pete, you had a question?
17 MR. DAVIS: On this SSFI, and I am having with 18 ' it, too ---
19 (Laughter.)
20 --- there is more than the name that-I would 21 like to change, however. Does it bother you that the 22 framework that these investigations are undertaking is 23 entirely deterministic apparently? In other words, there 24 is a heavy flavor of design basis accidents with some 25 arbitrary failures added on to that rather than a risk base O
ACE-FEDERAL REPORTERS, INC.
202-347-3R)0 Nationside Coverage 800-336-6M6
06000202 144 marysimons 1 approach.
2 We have done a lot of work now on what causes 3 core melt accidents and we know the role that auxiliary 4 feedwater plays in t' hose dominant sequences, and yet it 5 seems to me like in this program it is more of a 6 deterministic design basic approach which can result in 7 adding things to the system that are really potentially_
8 detrimental to risk.
9 Do you agree with that assessment or do you 10 have any concerns in that area?
11 MR. ENOS: I think there certainly is some 7
12 validity in what you are saying in not pointing a finger j () 13 specifically at SSFI. 'I think the industry and the~NRC as 14 a whole is struggling with this issue.
15 I think we saw 'hings t in the SSFI of.sometimes 16 we were looking at comparisons to design. basis and
- 17 sometimes we were looking at comparisons to engineering 18 practice, but not really whether or not that was in design 4
19 basis. The classic mentality both within the NRC and us as
~
20 licensees is give me design basis. I know what ball game I l
21 am in.
l 22 I think we are trying, we Arkansas is trying 23 and I think the industry is trying and I think the NRC is I
24 trying to do some of the things that I heard Mr. Denton say 25 last year, which is let's pay attention to what-it is the ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336- 4 86
l l
I 06000202 145
( } marysimons I system is required to do to provide protection and not let 2 some low probability event get in the way of providing the 3 fundamental protection.
4 I think those are some of the things that we 5 are all beginning to move'toward, but we have had our feet 6 tied into the design basis for a long time and it is kind 7 of hard to get out of the. mud and start moving over and we 8 are getting into unknown territory.
9 So I think when I see specifically in your 10 question is, nothing intentional and I think it is real, 11 but I think it is a difficulty and a change of philosophy 12 that both the industry and the NRC are looking at at this
() 13 point from my perception.
14 MR. WARD: I think that is a good point, Pete.
15 MR. MICHELSON: I would like to make a speech 16 on that then, too.
17 (Laughter.)
18 MR. WARD: Just a short one. I mean we are on 19 the threshold of perhaps a new era in' regulation. We have 20 crossed over the border of that threshold a-little bit, and 21 if IEE is starting a brand new program, then perhaps there 22 should at least be consideration given to basing them on 23 some more modern concepts of risk.
24 MR. MICHELSON: My understanding though of the 25 SSFI is perhaps different than yours then. This was not an (v\
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
06000202 146 marysimons 1 attempt to go and redo the analysis on a probabilistic 2 basis or otherwise.- There was no attempt one way or the 3 other. It was my understanding that it was to go'in and 4 determine the functional readiness of the systems, and a 5 part of what they would do is go back and look and see if 6 the system met the design basis that the utility thought it 7 was meeting and that the NRC thought it was meeting.
8 Okay, if the utility thinks they can handle 9 single failure with loss of.offsite power, then that is the 10 way they look because that is our understanding. They are 11 just trying to verify that our understandings of that 12 system are correct. So they went through and they checked
() 13 the'FSAR and then they checked to see if the design 14 specifications got what the FSAR said.
15 Then'they went and looked at the components and 16 said will these components do what we thought that would do 17 and what basis do we have to believe that they will do it.
18 This was the approach.
19 MR. MINNERS: You are talking about two 20 different things, Carl.
21 MR. MICHELSON: I beg your pardon?
22 MR. MINNERS: You.are talking about two 23 different things. He is talking about the redesign of the 24 system and you are talking about the SSFI.
25 MR. MICHELSON: Well, that is what I thought he O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-3364 6;6
06000202 147 marysimons I was referring to.
2 MR. MINNERS: No. He is talking about.
3 redesigning the system.
4 MR. ENOS: I am not arguing that that was the 5 purpose. I agree with you. ,
6 MR. MICHELSON: Okay. .But having determined 7 now that you have got a difficulty and you have to redesign 8 the system, the SSFI doesn't do that at all. It is just a 9 straight out report.
10 Now if you want to redesign the system, and if 11 the NRC regulators say it has.to be redesigned, then there 12 are processes of going through that and maybe then you will
() 13 turn to PRA or whatever, but SSFI never ---
14 MR. MINNERS: In a way the SSFI did redesign 15 the system because it made them put check valves back.in 16 the system which their original design said ---
17 MR. MICHELSON: Well, not in'the-sense that 18 they made them put it in,- but rather they pointed out the 19 difficulty and~the utility said that this their solution.
20 MR. MINNERS: Right, in very nice, polite 21 terms.
22 MR. WARD: But I guess, Carl, you point is at 23 that point in the analysis or negotiations there could have 24 been or should have been perhaps a probabilistic risk base E 25 judgment brought to bear on the decision.
l ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 804336-6M6
06000202 148
_marysimons
- 1. MR. MICHELSON: The SSFI is just trying to 2 determine do we have what we thought we had, and.now having 3 found out that you don't, if that is the case, then what 4 .you do about it has gotten very many attacks.
5 MR. DAVIS: Another approach would be do you 6 have what you need, and not if what you have is ---
7 MR. MICHELSON: That wasn't the approach of 8 SSFI.
9 M.t . DAVIS: No, I know it isn't. I am 10 suggesting that might be a consideration.
11 MR. MICHELSON: That is an entirely different 12 kind of a problem.
() 13 MR. WARD: Okay. We had better -- thank you,
! 14 Mr. Enos.
15 MR. DAVIS: Mr. Ward, gentlemen, _I appreciate 16 the opportunity to address this.
17 MR. WARD: We have got one more speaker, and 18 this is really the heard of the subcommittee meeting which 19 we are getting to a little late but, Sam, you go ahead.
20 MR. DIAB: Well, everybody is probably. tired by 21 now and maybe I can zip through my presentation very quick.
22 (Slide.)
23 I would like to continue the presentation we 24 started earlier today.
25 This is the last slide we showed showing the ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coserage 800-3364 616
06000202 149
{ } marysimons 1 modified .esolution approach. As I said, it is a two-tier 2 approach with seven plants that I will discuss next, and 3 the rest of the PWRs, what NRR will do with those will be 4 much clearer after we finish our study with the seven 5 plants.
- 6 (Slide.)
7 MR. MICHELSON: Could you give us a list of 8 those seven plants? 'Is it the Arkansas 1 and 2, Crystal 9 River,' Fort Calhoun, Prairie Island and Rancho Seco?
10 MR. DIAB: That's it.
11 MR. MICHELSON: Apparently it is not true 12 though that there will be a functional inspection of all
) 13 seven of those because I asked specifically on Prairie 14 Island and they'said there wouldn't be one. So they are 15 not going to do functional inspections on these seven 16 plants, t
17 MR. MINNERS: They are not planned.
18 MR. MICHELSON: I had kind of gotten somehow 19 the impression that the seven plants would have functional i
20 inspections.
i 21 MR. DIAB: The functional inspection.that IEE 22 does is really we have nothing to do with that and we 23 understand they are probably not going to do functional 24 inspections on these plants.
25 I am going to go down the list here with the
- i. b ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
06000202 150
-['T
\_)
marysimons 1- seven plants to show the status of each one of the seven as 2 of today.
3 As.far as ANO-1 is concerned, the SSFI has been 4 done by IEE. The plant apparently will be reviewed-by.the 5 review group which is the subject of this presentation.
6 ANO-1 is also subject to the B&W Owners Group design 7 reassessment program.
8 MR. WARD: What review group are you talking 9 about in the second bullet?
10 MR. DIAB: The review group is a review group 11 that we are forming to resolve the generic issue 124.
12 Rancho Seco, they are here at NRR for a restart 13 review. We won't start until that review is completed.
14 That is a multidisciplinary staff review of the aux 15 feedwater system, among other system. It is really across 16 the board.
17 Design, that will include design modification 18 procedures and training, support systems, indication and 19 control. That plant is also a part of the B&W Owners Group 20 effort.
21 (Slide.)
22 Crystal River, the same thing. That first line 23 is misspelled. It is supposed to be the B&W Owners Group 24 design reassessment.
25 We understand also that the licensee has an ACE-FEDERAL REPORTERS, INC.
202-347-3700 Stionwide Coverage 800-3364M6
. . - _ . . . . . . _ . . .-= - -- . . - . . .- -. .- - - . ~
06000232 151 marysimons
-1~ availability analysis.
'2 (Slide.) .
3 . Prairie Island, we have an aux feedwater system 4 reliabil'ity analysis that is under review by us, and that 5 reliability analysis considered many of the. things that we 6 talked to you earlier about earlier in the year, things 7 that the staff, you know, would like to see in'a 8 reliability analysis like plant specific data and support 9 systems and human ~ failure, et cetera.
10 Prairie Island again,.'it is one of the best run 11 plants. The record shows that it-is a very well managed 12 plant and it has one of the highest capacity factors in the
() 13 nation. That indicates a certain amount of control.
14 MR. WARD: What record are you talking about?
4 15 you are saying is one of the best managed plants inLthe 16 country. Who says?
17 MR. DIAB: That is a report that apparently was 18 nade by the NRC.
i 19 MR. WARD: Those SALP ratings you mean?
I 20 MR. DIAB: I am not really sure. I am sure it
, 21 has something to do with the SALP, at least it is maybe a i
22 factor in it, but it has been reported about 14 other 23 plants that were identified as'the best managed plants, and 24 there is another list'also for the p'lants that have l'
l 25 problems with their management.
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
06000202 152 marysimons 1 ANO-2, that is going to be subject to the 2 review group I guess as well as others. I noted there that 3 ANO-2 has a dedicated feed and bleed or bleed and feed 4 arrangement through the valve that was discussed earlier.
5 We also understand that ANO-2 had a reliability 6 analysis done by the' licensee. That we have not seen.
7 Fort Calhoun, again they made certain 8 modifications to their systems and they were subject to 9 this review group.
10 MR. WARD: Would you remind me, is Fort Calhoun 11 -- as I recall, Fort Calhoun really has a three-train 12 system,_ don't they?
) 13 MR. DIAB: I think it is has a two-train 14 system.
15 MR. WARD: It is a two train. The information 16 I have is in error.
17 (Slide.)
18 The B&W Owners Group effort that is underway 19 right now includes all the B&W plants on our list.
20 Basically it is a reassessment to the design basis for the 21 aux feedwater system. The B&W Owners Group reviews more 22 than just the aux feedwater system. It is a significant 23 portion of it.
24 Some of the things they are looking for are 25 improving the reliability of the main feed, improving the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80433MM6
i l
l 06000202 153
} marysimons J
1 reliability of the aux feed and limiting the number of 2 challenges.
3 I believe the full committee will hear about 4 this program on Friday.
5 MR. MICHELSON: Which program are you referring
~6 to?
7 MR. DIAB: The B&W Owners Group effort.
8 (Slide.)
9 Now the' review program that we'are undertaking 10 now, first of all, it will start by forming the review 11' team. We are talking about a five-person team plus a team 12 leader. We have already conducted preliminary discussions
() 13 with the perspective team members.
14 I think we are shooting for and hopefully will 15 achieve -- the team should start to operate within a week's 16 time from now.
17 MR. WARD: 'Are these all NRR people? Are you
~
18 telling us what the team is made up of later somewhere?
, 19 MR. DIAB: No. <
l 20 MR. WARD: Would you tell us now then.
l 21 MR.' DIAB: Basically it is'made up of four l
l 22 individuals from NRR and we are trying to recruit one
- 23 person from I&E, and the functions of everybody will be l
24 split according to expertise so that we can cover ---
[
l 25 MR. WARD: And are they going to be dedicated i
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6 4 6
06000202 154
} marysimons 1 to this task or are they going to be spending part of their 2 time?
3 MR. DIAB: Most of their time for the next 4 soveral months, and we will-get into that a little bit 5 later.
6 MR. EBERSOLE: May I ask you a question since 7 you are looking at the individual plants now. One of the 8 weaknesses of the main feedwater pumps to provide mini-9 flows or low flows, five or six hundred gpms, is that they 10 are dependent on exhaust discharge in the main condensers 11 which are electrically cooled with circ-water.
12 Do any plants incorporate the advantage of open 13 cycle discharge to the roof for the main feed pumps as an 14 auxiliary _means of allowing them'to operate?
15 MR. DIAB: I'm sorry?
16 MR. EBERSOLE: Are they permissive to exhaust 17 to atmosphere?
l-18 MR. DIAB: From the condenser?
19 MR. EBERSOLE: No, from another exhaust train.
20 MR. DIAB: All of these plants have atmospheric 21 dump valves.
22 MR. EBERSOLE: On the discharge of the main 23 feed pumps?
24 MR. DIAB: No, that is on the steam line.
25 MR. EBERSOLE: No, no, no.
O
. ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804 336-6616
06000202 155
' marysimons 1 MR. WARD: He means can the aux feed turbine ---
2 MR. EBERSOLE: No, no, the main feed-turbine.
3 MR. WARD: Oh, the main feed turbines discharge 4 directly_to atmosphere or do they have to use the 5 condenser?
6 MR. EBERSOLE: Under low flow conditions.
7 MR. DIAB: Probably not, and I am not really S sure, but ---
9 MR. EBERSOLE: So then that makes them 10 automatically dependent on the electric system and the 11 condenser circ water and all that?
12 MR. DIAB: As far as the main feed pumps are
) 13 concerned.
14 MR. EBERSOLE: So they don't have the 15 prerogative given to the relief valves to open to i
16 atmosphere?
17 MR. DIAB: Not to keep the main feed pumps
! 18 running.
19 MR. EBERSOLE: Even at low volumetric flows.
20 Okay. I just wanted to know whether you even took that-i
~ 21 course.
f L 22 MR. DIAB: The review scope that we are ,
23 planning to take, we are going to show that on Table 1. We
( 24 will cover the auxiliary feedwater system and support-l-
l 25 systems. We will audit selected sections of the preventive
()
i ACE FEDERAL REPORTERS, INC 202-347-3700 Nationwide Coverage 800-336-66:6 l_
06000202 156
(~N marysimons 1 and corrective maintenance activities during the last 12 2 months for a particular plant. We will audit selected 3 sections of the surveillance testing procedures and post-4 maintenance testing.
5 (Slide.)
6 MR. WARD: Now this is a different team for 7 each of the seven plants?
8 MR. DIAB: No, it is the same team.
9 MR. WARD: Okay.
10 MR. DIAB: The review program -- well, this is 11 a continuation of the review program and some'of the 12 possible outcomes that we are going to try to reach will be
() 13 either the review team is reasonably assured that the aux 14 , feedwater system is adequate and sufficiently reliable, or 9
15 with certain provisions we feel that the aux feed will be 16 adequately and sufficiently reliable or a negative finding 17 meaning that we think that t'he aux feedwater system is not 18 very adequate.
19 As far as the implementation of these findings, 20 first of all, any finding that we reach, the review group 21 reaches will be discussed thoroughly with the licensee just 22 to make sure that we both are talking about the same 23 perceived deficiency.
24 Any future or potential plant backfits will be 25 handled through the normal channels and the normal O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 8(xt3364M6
f 06000202 157
/~h marysimons
()
1 procedures.
2 (Slide.)
3 Next I would like to ---
4 MR. WARD: So that means any recommendation 5 from the review group is going to be subjected to the 6 backfit rule.
7 MR. DIAB: Yes. Next I would like to show you 8 Table 1 with a few things on it. The review group will 9 study hopefully before we visit a plant. Before we visit 10 the plant we will have reviewed or studied these things and 11 outline certain areas we would like to cover once we are at 12 the plant and then we will come back and try to put our O)
(_ 13 findings in a plant specific report.
14 MR. MICHELSON: Are you looking at the 15 operating experience of these plants at all relative to 16 auxiliary feedwater?
17 MR. DIAB: Yes. We are looking at the 18 operating experience in the limited period of time which is 19 the last 12 months.
20 MR. MICHELSON: I didn't find it as a line item 21 anywhere and I just wondered ---
22 MR. DIAB: Maintenance records and surveillance 23 testing, that ---
24 MR. MICHELSON: Surveillance testing is not 25 quite the same thing.
ACE FEDERAL REPORTERS, INC.
202 347-3700 Nationside Coverage 8%336-6M6
j 06000202 158 marysimons 1 MR. MINNERS: What do you mean by operating 2 experience?
- 3 MR. MICHELSON: I mean you are going to read 4 all the LERs from that particular plant as it relates to 5 auxiliary feedwater for the period of time of interest as a 6 part of the background preparation which I assume Table 1 7 is the background preparation.
8 MR. DIAB: If you look at Item No. 8 on the 9 table there you will see ---
10 MR. MICHELSON: Let me ask a different l
11 question. What are you.trying to accomplish with this 12 team? What is the objective, team objective? They are
( 13 trying to determine if this thing is sufficiently reliable; 14 is that right? You tell me. What is the team trying to 15 do?
1 16 MR. MINNERS: It's the previous slide.
4 17 MR. MICHELSON: Okay.
18 .(Slide.)
19 I read that a little differently. I read that j 20 as the review outcome. What is the purpose of the team?
21 MR. MINNERS: The purpose is come to one of 22 those outcomes.
23 MR. MICHELSON: Okay. It is going to determine 24 that it is sufficiently reliable. Doesn't it look at 5
25 operating experience as a part of that process?
4 2
Au:E-FEDERAL REPORTERS, lNC.
202 347-3700 Nationside Coverage 8043364M6
i 0FGOO202 159 marysimons (v ')
1 MR. DIAB: Yes, and that is shown on the table.
2 MR. MICHELSON: Does it look at NPRDS as a part 3 of that process to see what kind of failure experience they 4 would have and so forth? Now it says in here, and maybe I 5 misinterpreted it, it says " Post-maintenance testing LERs,"
6 and I thought that meant only those relating to post-7 maintenance testing.
8 MR. DIAB: No. Post-maintenance testing ---
9 MR. MICHELSON: Maybe I read that wrong.
10 MR. DIAB: --- and I think we should have a 11 comma there, there is a comma missing --- LERs, NPRDS ---
12 MR. MICHELSON: Okay, all right.
,q
(_) 13 MR. DIAB: --- significant operating events, 14 reportable occurrences, et cetera.
15 MR. MICHELSON: Are you looking at that more 16 than from the viewpoint of surveillance testing?
17 MR. DIAB: I'm sorry, what is the question 18 again?
19 MR. MICHELSON: Well, if your equipment fails, f
20 and it may not have failed during an surveillance test, it 21 might have failed during a demand ---
22 MR. DIAB: We are not looking at just the 23 surveillance testing. We are looking at LERs and the 24 NPRDS.
25 MR. MICHELSON: Okay. I'm not going to -- Item
(~)
%_)
{
i ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationmide Cmerage 800-336-6M6 l
06000202 160
/' marysimons 1 8 is surveillance testing. That is what Item 8 is.
2 MR. DIAB: That was not che intention.
3 Surveillance testing is an item or No. 8.
4 MR. MICHELSON: WelJ then there is more than i 5 comma problems. There are parentheses problems, too.
6 MR. WARD: Don't take that one off yet, Sam, 7 because I think as I read the rest of your charts -- you 8 know, this is really quite important. you have backed away I i 9 from the position you had in March where you were looking 10 at some quantitative goals for reliability and now I am 11 seeing here that you want to be reasonably assured, you 12 want to make sure the system le adequate and sufficiently (n_) 13 reliable. Now what do those mean?
14 I Here is an error. Just recently the 15 Commissioners have laid a safety goal on. Are you going to 16 use that? Is that going to be part of yout consideration?
17 MR. DIAB: What we are trying to do right now 18 is we are trying to understand or use some of the 19 experience that we have had, have accumulated from the 20 industry efforts, from the staff efforts over the last few 21 years as far as reviewing and re-reviewing the aux 22 feedwater system.
23 Now we have in house here the auxiliary 24 feedwater system reliability analysis for Prairie Island 25 which is a substantial amount of analysis. We also have A
N ACE-FEDERAL REPORTERS, INC.
20:-347 3700 mrion. iae cos erag.e Ko34ua6
}
06000202 161
()
%J marysimons 1 the changes or modifications that Prairie Island has done 2 on their plant since the TMI event or accident.
3 MR. WARD: Well, what is your standard going to 4 be, that the other plants have to do things as good as 5 Prairie Island did or you have got to have a third train or 6 you.have got to meet the safety goal or you have got to 7 meet some reliability number? I mean what are you going to 8 use as the basis for your judgment?
9 MR. DIAB: Okay. It will have to be somewhat 10 subjective and we are going to.try to sort of measure what 11 we see in these plants relative to other plants that we 12 have seen.
13 : The first plant we are going to is Prairie 14 g Island and we have the benefit of the reliability 15 h1 analysis. So we are going to review that first and we are 0
16 l going to measure other plants ---
0 17 ) MR. WARD: By other plants you mean what, other u
18 PWRs?
19 ? MR. DIAB: The other PWRs on the list, the rest 9
Il 20 1 of the seven plants or the other five plants.
21 MR. WARD: You want to insist that they are all 22 as good as the average one or the best one on that list or 23 what? I am trying to understand what your definition of 24 reasonable assurance is.
25 f MR. DIAB: Reasonable assurance is going to
- (1) l l
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 801)-336&M
r-06000202 162 '
marysimons ,
1 have to be, like I said, subjective in our minds in the 2 s'ense that we have --- i 3 MR. WARD: I think I used the wrong term, and .
4 I'm sorry to interrupt you. What do you mean by 5 sufficiently reliable? That is the key. What do you mean 6 by that?
7 MR. DIAB: That is not a very easy question to I
8 answer because when you say sufficiently reliable it has to 9 be compared to what, compared to a reasonably good plant. .
10 Now we have, like I mentioned earlier, and I think we are 11 going to show a couple of slides on that, the Prairie 12 Island reliability analysis. They give us some performance
/~T
(_j 13 data and they also give us some reliability numbers. Now l'
14 that reliability analysis is being reviewed in the staff 15 right now. We show reliabilities in the order of ten to 16 the minus five per demand, which is in the. higher range of 17 reliability as defined earlier by the staff.
18 Given that that is correct, and we are going to 19 try to find out in the next few weeks by reviewing the 20 reliability analysis and then finally going out and 21 visiting the plant and auditing their data base that they 22 use.
23 So like I say, we are measuring plants against 24 each other. We don't have an absolute ---
25 MR. WARD: So you are saying that you think (n
s_/
)
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationaide Cos crage 800-33MM6
i 06000202 .
163 (G
w]
marysimons 1 even with a two-train system that Prairie Island can get by 1
2 l having good maintenance and operating practices I guess, 3 that Prairie Island can achieve a reliability of ten to the 4 minus fifth?
5 MR. DIAB: Yes.
6 MR. WARD: With a two-train system.
7 MR. DIAB: Yes, that is what ---
8 MR. WARD: And you are saying if they can prove 9 to you that that is it, then you are going to tell the 10 other plants why don't you do this, too, and you won't have 11 l to add another train? Is that what you are saying?
12 ! MR. DIAB: I am not coming with a mind-set that O h t ) they have three trains or anybody should have three trains.
13 l ,
14 MR. WARD: But you seem to have a mind-set that 15 if they can come up with a ten to the minus fifth on demand 16 number and prove it you will accept that.
17 MR. DIAS: Yes. I mean if everybody can show, 18 I which was our position, you know, the staff's position 19 earlier that utilities should show that they have aux 20 ! feedwater systems that are as good as ten to the minus five 21 or less-failures per demand.
22 MR. WARD: That is pretty good, and I don't 23 know whether they can prove it or not. But if that is your l 24 criteria, why don't you say that.
25 '
MR. DIAB: What, ten to the minus fSve?
l f))
f ACE FEDERAL REPORTERS, INC.
202 347-3700 Nuionmide Coserage 800-3 E6(4
06000202 164
(~'i marysimons V
1 MR. WARD: Yes.
2 MR. DIAS: Well, if you put that criterion as 3 ten to the minus five, you can't really go and make a plant 4 visit and come back and determine that they have met that 5 criteria. Short of doing an analysis you can't quantiff.
6 What we are proposing here and what we are trying to do is 7 to go out and visit the plant and review the things that in 8 the past have proven to be the reason or the root causes 9 for problems in many plants like the turbine driven pumps, 10 like the motor operat6d valves, like the problems in 11 diesels and that sort of thing, maintenance, incorrect 12 testing or improper testing, that kind of thing.
/^
(_)\ 13 ;l If we are successful in that, I think we will i
14 r do good in assuring the reliability of these systems.
15 MR. MICHELSON: What do you see as the 16 difference between this team inspection and the functional inspection performed by I&E in their process as it applies 17 l.
20 to auxiliary feedwater?
19 i MR. DIAB: Our task is somewhat limited in what 20 11 the SSFI does.
21 MR. MICHELSON: The items that you are looking 22 at seem to be reasonably comparable to what I&E looks at.
23 Is it the amount of time that you spend at it, is that the 24 difference or do you think their scope is somewhat broader or less broad?
/~N 25 l, N- \,
ACE FEDERAL REPORTERS, INC.
202-347-3AO Nationwide Coserage K43364646
1 06000202 165 marysimons 1 MR. DIAB: I think like they said earlier, they 2 are trying.to look and make.sure that the design basis is 3 adequately reflected in the plant. l 5
4 MR. MICHELSON: Well, their literature says it p
5 differently. They are trying to verify that the system is 6 functionally ready. What are you trying to verify? L 7 MR. DIAB: That the system is functionally 8 available and ready.
9 MR. MICHELSON: I thought that essentially you 10 are trying to verify the same thing I believe.
11 MR. D1AB: If we are successful in coming close 12 to what they do I think that will be good also. I don't '
tO
(_) 13 think.we are going to go as far as going back to the design -
14 basis of the plant and trying to make sure that the plant i i 15 is designed properly according to the FSAR when the SAR was I 16 written.
- 17 We are trying to look at the aspects of the 18 plant operation and procedures that produce.= problems that 19 makes the system unavailable. And if we can identify most
20 of these and be successful in having them correct it, I -
, 21 think that will add to the availability and'the reliability -
22 of the system. l 23 MR. WARD: It seems then that your program is 24 to go out, look at the plants and identify problems that 25 you can identify and ask them, ask or demand or something ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-336M46
- _ . . _ . ~ . _-. .. . - - , - - - - . . _ _ _ _ _ - -,
B 06000202 o~
166
- marysimons -
1 that the licensee get rid of.those problems. That is 2 subjective, I will grant you that.
4 j 3 MR. DIAB: Well, if we identify a problem, and 4 ,
there are many problems that have been discussed I
5 extensively with I&E bulletins and INPO reports and things l
i l 6 l of that sort. But if we can still go out and identify some i
7 deficiencies like that and come back and report it and 8 bring it to their attention, I think that will be a gcod l 9 achievanent.
I 10 l MR. WARD: Wnat are you going to do if you run 11 up again -- what if the licensee says I don't think I need 12 to fix this -and you run up against a backfit rule? What j 13 Sappens then?
l 14 i MR. DIAB: I think we will have to -- at that 1C time when we get ready to go into the backfit rule, we will L
16 have to go by the backfit rules, meaning we will have to 17 , assign probabilities and run the numbers like you normally 18 do when you want to make a backfit.
l MR. EBERSOLE: Isn't that no more then a )
19 l L 20 numeralization of what shou, be competent and unbiased i
21 1 judgment?
! 4 22 MR, DIAB: Except at that point in time you f
23 will have concrete things that are applicable.to a certain 24 plant. Right now, or the previous approach that we 25 l discussed here before was to come back with a generic set ace FEDERAL REPORTERS, INC.
l l 202-347-3700 Nationside Coserage 800-33MM6 i
06000202 167
/"N marysimons
(_)
1 of circumstances that apply to everybody and we go out and i
2 run the reliability analysis and risk based on those-to
!- 3 make it globally applicable to every plant.
4 MR. EBERSOLE: My impression is though that the 5 numeralization process is generally an escape route against C having to do-anything.
. 7 MR. DIAB: I think that is mostly what the 8 backfit rule requires.
, n -
9 MR. EBERSOLE: Well, that is because that is
- 10 what it is. It is an escape route. .
11 MR. DIAB: Do you want to. comment on that, 12 Warren?
13 MR. EBERSOLE: It'makes possible an escape 14 otherwise not judgmentally possible.
15 MR. MINNERS: Oh, I have seen people escape on
. 16 ! judgment calls and I have seen people escape on numbers and
- 17 I have seen people made to do things on design basis event l 18 . judgments, and I have seen-people made to do things on the l 19 basis of numbers. ,
20 MR. EBERSOLE: Numeralization, however, is ---
21 MR. MINNERS: Any combination or permutation is i
! 22 possible and has occurred.
- 23 MR. EBERSOLE
- yes. I am not at all assured l
l 24 that the numeralization route is always the way.
l l
25 MR. MINNERS: Well, I think you have to go to L
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nmen*ide Coserage 800-336-Ma6
06000202 168
- marysimons
'- 1 the quantification, Jesse, when you get into these 2 balancing situations. If you are making a -judgment of i
3 whether check valves ~should be put'in the system.or not put 4 i
~
4 in the system, I don't know how you decide that unless you
! 5 probably do a quantitative analysis when.you-have got to do 6 balancing. If you don't have to do balancing, then you can 7 just do good engineering practice. -
8 MR. EBERSOLE: If you have the statistics with ,
9 which to do it.
- 10 MR. MINNERS
- I think you have to do it even if 11 you don't have the statistics. That.is the'only way-you 12 can balance it.
_ () 13 MR. EBERSOLE: Then it become synthetic.
14 (Slide.)
- 15 MR. DIAB
- Next I would like to show you'the _
)
16 schedules for doing these efforts on individual plants.
j
[ 17 For Prairie Island we are shooting for a visit 18 to the plant on the week of September 29th, and that is not 19 very far from now, with a report written around the 3rd of 20 November. And you go down the list.
21 ANO-1, Rancho Seco and Crystal River, we have l
22 to coordinate our efforts as far as potential or, you know, 23 whether we really need to go and visit the plants or not, 24 and also on the report date with the other efforts that we 25 are doing right now like the Rancho Seco restart effort and l
l l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
06000202 169
/^ marysimons ls)3 1 .the B&W Owners Group effort.
2 MR. MICHELSON: Is there some reason why you 3 would go to ANO-1 since the SSFI was already performed on 4 that plant for auxiliary feedwater?
5 MR. MINNERS: I think the indication is we are -
6 not going to.
7 MR. MICHELSON: Well, yes. Of all the choices, 8 I hope that one is the last choice of a plant to go to 9 since it has already been through the wringer once already.
10 MR. DIAB: The next couple of slides are about 11 Prairie Island. Again, like I say, at Prairie Island we 12 have an opportunity to review this plant extensively
) 13 because they offered their help as far as making our visit 2 14 there as successful as possible and making every piece of 15 information available to us for auditing and also having 16 this reliability analysis in house that the staff is 17 reviewing right now.
l 18 (Slide.)
i 19 This is just a list of things that they have 20 done. That is before the reliability analysis. In other 21 words, this is the plant as it exists today without some 22 further things that I will discuss.
23 (Slide.)
24 After the reliability analysis is concluded l 25 they listed certain things that became improved, and l
()
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
06000202 170
(~T marysimons
(-)
1 improving these things will increase the reliability of the 2 aux feedwater system.
3 (Slide.)
4 Now this is a table that shows some of the 5 reliability estimates that they have arrived ut.
6 You see the first colurm on this table, this 7 was the number that was calculated in 0611 for Prairie 8 Island. Now they made a benchmark analysis, making a model 9 that we will try to reproduce the numbers that were 10 estimated in the 0611 days, and I think they were very 11 successful in that.
~12 you can see that in the pre-TMI benchmark they
() 13 show like 3.4, ten to the minus three as opposed to one 14 times ten to the minus three. So they were trying to show 15 that they understood how-the model was made and how the 16 numbers were generated.
17 MR. EBERSOLE: What is this, per challenge?
t 18 MR. L'AB: This is per demand, that is correct, 19 on the loss of main feed event and the loss of offsite 20 power.
21 If you look at the last column there you will l 22' see that the best estimate for the plant as exists today, 23 two times ten to the minus five and 8.7 times ten to the j 24 minus five respectively.
25 MR. EBERSOLE: Does loss of offsite power mean n
(
L
! ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-(M6
._ - , , = _ . . . _ . . _ _ _ - _ . . _ . -__ . - _ . _ _ _ _ . . . - . ~ _ __ - _.
. . . - ..~ .. .- ~. - - . . . , .
' ., 06000202 171 l marysimons 1 per demand. basis?
2 MR. DIAB: Well, once there is a loss of h 3 offsite power you. trip the reactor-and you trip the 4 ' turbine.
5 MR. EBERSOLE: What does 1.---
6 MR.. DIAB: Ten to the minus three?
7 MR. EBERSOLE: Yes, loss of offsite power.
i j 8 MR. MINNERS: One in every thousand loss of 4 '9 offsite power events the auxiliary feedwater system won't-10 .run.
11 MR. EBERSOLE: Oh, I've got it.
i.
12 MR. DIAB: I'put a note at the bottom of this
( 13 table that these numbers have not been verified.yet. So 14 they are provided by the licensee and our review'is-still s
15 underway.
- liS MR. WARD
- Sam, there is not a great deal of-
'17 difference in numbers, but I really don't understand the i 18 difference between t'he licensee's benchmark post-TMI-and-19 best estimate post-TMI. -Would you explain that again.
20 MR. DIAB: Okay. The benchmark announced as 21 post-TMI is taking the model that they made to. reproduce 22 the NUREG 0611 numbers,'the pre-TMI-numbers and factor in 23' all the changes that were implemented in the plant using 24 the generic data that was used in the 0611 or 0635 and come 25 out to this number. 5.4 to the minus five.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33MM6
06000202 172 marysimons 1 They went and used their .cnni plant specific 2 data, and they.also went and used some other things that I
3 were not adequately covered by the 0611 like the support 4 systems, for instance. .They went-into detail on that. So 5 these numbers, what.you are see as best estimates,_they are 6 the best estimate in the sense that if they had used some 7 of the assumptions that were used in 0611, they would 8 probably have come with much-better numbers than what you 9 see there. The reason that they are not very different is 10 that they~took away from the model that was'used in 0611.
11 I don't know, did I answer your question?
12 MR. WARD: I think I have got a rough idea.
i 13 MR. DIAB: And I think that is basically it.
j 14 What I show after that is the Prairie Island PNIDs.
15 That concludes my presentation.
16 MR. EBERSOLE: Would you go back to that little 17 table.
18 (Slide.)
19 Why did you invoke that once in every thousand 20 loss of main feed or loss of offsite power rather than-just 21 flat out saying what the per demand failure rate Was?
22 MR. DIAB: This is it. This is the per 23- demand. Every time you have a demand on a loss of main 24 feed event, you have a failure rate of ten the minus three.
25 MR. MANNING: Systems have different a
ace-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M446
. _ . . _ _ _ _ _ . _ _ , . _ . _ _ _ . . . _ _ , _ _ . . _ _ _ _ . . , _ _ , _ _ _ _ . . _ _ . ~ . , _ . . _ . _ . . -
4 ~l
. i
- l l
06000202 173 marysimons ]
1 availability depending on what kind of events' challenge 2 them.
3 MR. EBERSOLE: A failure rate of ten.'to the 4 minus three which is the same as expressed there for main 4
5 feed or offsite power' failure.
G MR. DAVIS: That surprises mue a little ' bit 7 because in the Prairie Island design the two aux feed 8 systems are' cross-connected, and I would presume if you 9 lose offsite power that means that each plant needs 10 auxiliary feed at the same' time. Whereas if you lose main 11 feed in one plant you have the other system's aux feed i
12 available.
() 13 MR. DIAB: That is correct.
14 MR. DAVIS: Which would make the reliability =
15 greater for the loss of main feed occurring.
16 MR. DIAB: That is correct.
17 MR. DAVIS: Why~isn't that reflected in these 18 numbers then?
19 MR. DIAB: That is because of the assumptions 20 that went into the 0611.
21 MR. WARD: You have to look at-the best 22 estimate one.
23 MR. DIAB: Yes. You can look at the best i
24 estimate and it will give you some of that. Now those 25 valves are manua'l valves that again you have to send
- O A
i ACE-FEDERAL REPORTERS, INC.
202 Nationside Coverage 800-336-6646
. . -- .,_ ., _ . , ._ _ _ _,, _ . .., _,_ _ _ _. _ -347-3 700._
~
06000202 174 marysimons 1 somebody down there to open the valve and switch over the 2 pump. ~But those are the numbers that were found in 0611 3 and they were'trying to reproduce.those numbers and all of 4 the assumptions were not available. Right now if somebody 5 wants to reconstruct 0611 they wouldn't be able to do it
~6 easily.
1
! 7 MR. DAVIS: Just -as a comment, I find those 8 numbers very low for a two train system.
9 MR.. DIAB: Which numbers, the best estimate?
10 MR. DAVIS: All of them, except for the 0611'.
11 Well, all of them starting with post-TMI.
12 MR. WARD: What is the problem?
13 MR. DAVIS: Pardon me?
14 MR. WARD: I mean if we believe those numbers, 15 what is'the problem?
16 MR. DAVIS: There isn't any problem if you i 17 believe the numbers. That is what I am saying. I am 18 having a little trouble believing them.
19 MR. EBERSOLE: That's right.
t l 20 MR. DAVIS: Why would he account for common
< 21 cause failures on B-57 i
j 22 MR. EBERSOLE: What about for tube rupture?
l 23 MR. DIAB: I'm sorry?
[-
l 24 MR. EBERSOLE: Which is more likely than any of I
25 the events you have got up there, f
ACE-FEDERAL REPORTERS, INC.
202-347-3700 . Nationwide Coverage 8 % 336 # 46 L
060002O2 175
/ marysimons V) 1 MR. DIAB: The steam generator tube rupture?
2 MR. EBERSOLE: Yes.
3 MR. DIAB: What about it?
4 MR. EBERSOLE: So what do I do then with my 5 system?
6 MR. DIAB: Well, if you have a steam generator 7 tube rupture -- again, I mean these two events were 8 picked. I mean, those are not the only events that 9 challenged aux feed. You challenge aux feed every time you
, 10 have a small break LOCA, like you are suggesting that you 11 challenge the aux feed every time you have a tube rupture.
12 MR. EBERSOLE: This time I would have-to p)
(_ 13 depressurize one boilder and live on the other one with the 14 feed system. I don't know that it would look any worse or 15 better.
16 MR. HERNAN: Sam, maybe you can clear up 17 something in my mind if not for the record. You mentioned 18 several times using a criteria of then to the minus five.
19 Now our standard review plan talks about a ban of ten to-20 the minus four to ten to the minus fifth unavailability per 21 demand. These plants that we have singled out are plants 22 that to the best of my knowledge cannot meet the ten to the
- 23 minus four. What am I missing?
24 Do any of these plants as defined in NUREG 0611 25 meet a ten to the minus four utavailabilty per demand at (G>
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
+
06000202 176 marysimons
, 1 the present time?
2 MR. DIAB: All of the other plants ---
3 MR. HERNAN: The seven other plants.
1 4 MR. DIAB: Excluding the seven plants. All of 5 the other plants excluding ---
6 MR. HERNAN: I am talking about'the seven- . How
{
7 many of those meet ten to the minus four?
8 1Ut. DIAB: Well,'as the analysis was done back 9 in.1980, none of them meets that ten to the minus four.
.10 They don't meet the high reliability range. As-of today it I- 11 is very conceivable that they meet it because'-- let me 12 just give you one example here.
)
~
- i. 13 For the Prairie Island analysis the one failure.
a 14 or the one problem that made this plant in the lower 15 reliability region was these valves here,.these manual 16 valves inside the containment.for each unit. If both of 17 those valves were to be left in the closed position after l
i l 18 maintenance, the common cause failure probability of that t
19 was.such-that this was -- as a matter of fact, this was the i 20 ten to the minus three, which is the probability of failure l
21 of the aux feedwater system ~for the plant, and that is what 22 got them into the lower reliability.
23 So once they fixed that, the-next cut-set is so 24 low that they will' be'in the higher range, and this is one l 25 of.the modifications that they made.
1
() -
l i
j ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-fM6
<06000202 '177
-sarysimons 1 MR. HERNAN: If we take a plant though that-has 2 got a two pump system and previously said was ten to the 3 minus~ third and they made all the post-TMI improvements and j 4 they redo the calculation and come out to -two times ten to 5 the minus four. What would this study'then do with'that 6 plant?
1
! 7 MR. DIAB: That study that we are undertaking?
8 MR. HERNAN: Yes.
i
!- 9 'MR. DIAB: It depends on their procedures and 10 how they calibrate their setpo'ints, how they. maintain their
- 11 valves,.how they maintain their turbine drives and how they 12 maintain their diesel -- some of'the plants have diesel
( 13 operating pumps -- how they are testing program saw it, and .
14 it is very conceivable that we will find it acceptable.
15 This is different than what you have heard l 16 before. This is based on good engineering judgment and 17 based on going out and doing a review the old fashioned 18 way, you know doing an engineering review and coming back 19 with a list of things that a good engineer or a good i
20 operator wouldn't do.
l L 21 MR. WARD: Let's see, is there an owners group l
22 that is looking at this consolidated issue?
23 MR. DIAB: Not as far as the seven ---
l 24 MR. WARD: The Seven Sisters Owners Group or l-l something like that.
E .
25 L
' ACE-FEDERAL REPORTERS, INC.
l '202-347-3700 Nationwid: Coverage 800 336-6M6
06000202 178
(") marysimons V
1 (Laughter.)
2 MR. DIAB: Not as far as the seven plants are 3 concerned, but we are going to plan to discuss every one of 4 these plants with the licensee extensively just to make 5 sure that we are not missing anything.
~
6 MR. WARD: It certainly is an old fashioned 7 approach. I mean you have ifcensed plants for the last 10
, 8 years or something based on either the ten to the minus 9 fourth or ten to the minus fifth or maybe really based on 10 three trains. Maybe that is what has given the reviewers 11 an adequate, warm feeling is really three trains.
12 MR. DIAB: See, the thing is when you get into 13 the three trains there is nothing that says that when a 14 plant has three trains that it shouldn't run bad. do you 15 can have three trains and still maintain them in an 16 inadequate manner that you end up with a lousy system, t
17 MR. EBERSOLE: Tell me, there was no pre-18 existing statistical analysis of Davis-Besse, was there?
19 MR. DIAB: Not that I know of.
20 MR. MINNERS: Pre-existing to what?
21 MR. EBERSOLE: To its failure.
22 MR. MINNERS: Yes, there was.
23 MR. WARD: Oh, sure.
24 MR. EBERSOLE: And what reliability of the aux 25 feed did that show?
O
%J ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33MM6
marys ns
}
1 MR. HERNAN: Between ten to the minus third to 2 ten to the minus fourth.
3 MR. EBERSOLE: But it was showing that high a 4 degree of reliability, and I am talking.about ten to the i
5 minus three ---
. 6 MR. HERNAN: Ten to the minus three is high.
7 MR. EBERSOLE: Well, I am just saying ---
8 MR. WARD: It was on the poor side.
i 9 MR. EBERSOLE: --- that under those numerical 10 conditions, however, it in fact has happened, although they 11 picked it up.
12 MR. WARD: Well, something that has a 13 probability of ten to'the minus sixth ---
14 MR. EBERSOLE: I am just questioning what 15 ' happened in this instance.
16 MR. WARD: It just did.
17 (Laughter.)
18 MR. EBERSOLE: So at ten to the minus three it 19 did it.
20 MR. WARD: Sam, do you have anything else?
21 MR. DIAB: Not really. That's it.
22 MR. WARD: Okay. Let me talk to the 23 subcommittee now for a minute.
24 (Whereupon,'at 5:10 p.m., the reported 25 proceedings concluded.)
O ACE-FEDERAL REPORTERS, INC.
- 202-347-3700 Nationwide Coverage 800 336-6M6
V CERTIFICATE OF OFFICIAL REPORTER A
U This is to certify that the attached proceedings before the UNITED STATES NUCLEAR REGULATORY COMMISSION -~ in the matter of:
NAME OF PROCEEDING: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS SUBCOMMITTEE ON DECAY HEAT REMOVAL SYSTEMS DOCKET NO.:
PLACE: WisSHINGTON , D. C.
DATE: TUESDAY, SEPTEMBER 9, 1986 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
4 (sigt) A N*f N (TYPED) .O MARY C. SIMONS Official Reporter ACE-FEDERAL REPORTERS,.INC.
- Reporter's Affiliation l
l O
v l
j
O NRR STAFF PRESENTATION TO THE ACRS
(
SUBJECT:
AUXIi.IARY FEEDWATER SYSTEM RELIABILITY,
- t. GENE 81C ISSUE N0. 124 i
DATE:
SEPTEMBER 9, 1986 l
O PRESENTER: SAMMY S. DIAB PRESENTER'S TITLE / BRANCH /DIV: TASK ftANAGER, REACTOR SAFETY ISSUES BRANCH / DIVISION OF SAFETY REVIEW & OVERSIGHT PRESENTER'S NRC TEL. NO.: 492-4083 SUBCOMMITTEE: DHRS O
O SEPTEMBER 9, 1986 j
i- PRESENTATION TO THE ACRS DHRS ,
SUBCOMMillth ,
~
AUXILIARY FEEDWATER SYSTEM (AFWS) RELIABILITY ~~
4 GENERIC ISSUE No. 124 i
THIS IS A STATUS PROGRESS REPORT l
MODIFIED RESOLUTION APPROACH
, MODIFIED RESOLUTION PLAN g
PROGRAM SCHEDULE FOR RESOLUTION O
. ,. _ _ - ,__ ~ _ . . _ . - .._ _ -..
O GI 12f4 MODIFIED RESOLUTION APPROACH WHY N0T STICK WITH "0LD" PROPOSED RESOLUTION?
NEGATIVE COMMENTS FROM NRR REVIEW 0F BACKFIT ANALYSIS LENGTHY PROCESS (ISSUANCE OF GL,' LICENSEES TO CONDUCT AFWS ras,-STAFF REVIEW.0F ras FOLLOWED BY ISSUANCE OF
! SERs) i f
i l
O WHY THE " MODIFIED" RESOLUTION APPROACH?
SHORT TERM CONCENTRATE.D REVIEW EFFORT FOLLOWED BY AN AFWS REVIEW FINDINGS REPORT FOR EACH OF THE SEVEN PLANTS
' REVIEW EFFORT WILL BENEFIT FROM THE OIE SSFI PROGRAM THE PRAIRIE ISLAND AFWS RELIABILITY ANALYSIS
-- OTHER INDUSTRY AND STAFF EFFORTS I
i j
i io i
O GI 124 MODIFIED RESOLUTION APPROACH TWO TIER APPROACH THE SEVEN PLANTS '
HE REST OF PWRs O
l O
O GI 124 THE SEVEN PLANTS ANO-1
- SSFI DISCUSSED BY OIE
- WILL BE REVIEWED BY REVIEW GROUP
- B8W OG FOR' DESIGN REASSESSEMENT l
RANCHO SECO RESTART EFFORT
- RESTART EFFORT
- EXTENSIVE MULTIDECIPLINE STAFF REVIEW 0F AFWS
((() (AMONG OTHER SYSTEMS)
- DESIGN MODS
- PROCEDURES & TRAINING
- SUPPORT SYSTEMS
- INDICATION & CONTROL
I t
?
4
& CRYSTAL RIVER l'
1
- REASSESSMENT OF DESIGN REASSESSEMENT .
e i
- LICENSEE'S RELIABILITY ANALYSIS 1 4
)
!lo t +
i j
~
i- ,
{D r
F i
i f
! I i
i .
l t
?
l i
I !
i E
I l
f i
t i
(~N PRAIRIE ISLAND - 1 8 2 xJ
- AFWS RA CURRENTLY UNDER STAFF REVIEW
- CONSIDERS ELEMENTS REQUESTED BY STAFF
- BENCHMARKED AGAINST NUREG-0611
- ESTIMATED HIGH AFWS RELIABILITIES i
L i - ESTIMATED RELIABILITIES NOT YET VERIFIED
- THE HIGHEST CAPACITY FACTOR FOR 1983 THROUGH 1985
- REPORTED AS ONE OF.THE BEST MANAGED PLANTS IN THE COUNTRY ANO-2
- AFW REVIEW GROUP-
- DEDICATED BLEED 8 FEED ARRANGEMENT
- LICENSEE'S RELIABILITY ANALYSIS I
FT CAH0VN
- AFW REVIEW GROUP y I
BEW OG (INCLUDES R. SECO, C. RIVER, ANO-1)
- REASSESSMENT OF DESIGN BASIS OF AFWS
- IMPROVE RELIABILITY OF MFWS
- IMPROVE RELIABILITY OF AFWS f - LIMIT CHALLENGES I
- CURRENTLY UNDERWAY
(]
- WILL BE PRESENTED TO THE FULL COMMITTEE THIS FRIDAY o
'w)
O REVIEW PROGRAM REVIEW TEAM
- A FIVE PERSON TEAM PLUS A TEAM LEADER
- PRELIMINARY DISCUSSIONS WITH' PROSPECTIVE TEAM MEMBERS
- TEAM SHOULD START OPERATION WITHIN A WEEK L
o REVIEW SCOPE
~
- AREAS TO BE COVERED ARE SHOWN ON TABLE 1 O
- AFW AND SUPPORT SYSTEMS
- AUDIT SELECiED PREVENTIVE AND CORRECTIVE MAINTENANCE ACTIVITIES Dim,ING LAST 12 MOS,
- AUDIT-SELECTED SURVEILLANCE TESTING PROCEDURES AND POST MAINTENANCE TESTING O
m _ -- - _
((() REVIEW PROGRAM POSSIBLE REV_IEW OUTCOME
- THE REVIEW TEAM IS REASONABLY ASSURED THAT THE AFWS IS ADEQUATE AND SUFFICIENTLY RELIABLE
- PROVIDED CERTAIN PLANT-SPECIFIC EEASURES WERE TAKEN THE AFWS IS ADEQUATE AND SUFFICIENTLY RELIABLE, l
o
- WITHOUT SUBSYANTIAL CHANGES TO THE AFWS' DESIGN MAINTENANCE, TESTING AND/0R OPERATI0ii, THE REVIEW TEAM CAN NOT BE REASONABLY ASSURED THAT TiilS SAFETY SYSTEM IS SUFFICIENTLY RELIABLE
( JMPLEMENTATION OF FINDINGS
- ANY REVIEW TEAM FINDINGS KILL Bt DISCUSSED WITH LICEN.;2E
" ANY PLANT BACKFITS WILL BE HANDLED BY NORMAL PROCEDURES.
C)
i TABLE 1. AREAS COVERED BY STAFF REVIEW i 1. P& ids AFWS, ADVs 'PORVs (FEED a BLEED)
SUPPORTSYSTEMS(E.G., POWER SUPPLIES, COMPRESSED AIR OR NITROGEN i 2. FSAR SYSTEMS, LUBRICATION, COOLING)
- 3. SYSTEM DESCRIPTION i 4. TECH.41 CAL SPECIFICATION ;
( 5.
lac LOGIC DIAGRAMS L
I 6. SERs SINCE 1980 I
~
- 7. MAINTENANCE (PREVENTIVE PROGRAMS, CORRECTIVE ACTIVITIES)
LAST 12 MOS lb 8. SURVEILLANCE TESTING (FEW PROCEDURES AND RESULTS, POST MAINTENANCE TESTING LERs, NPRDS, SOEs, R0s, A0s, DURING THE LAST 12 MOS i
- 9. EMERGENCY OPERATING PROCEDURES (LOSS OF HEAT SINK) ,
i
- 10. FaBLEED PROCEDURES (SEF. 9 AB0VE), TRAINING (YES/NO) l i ,
i I
i r
O ,
O t
SCHEDULE REVIEW TEAM VISIT REPORT ;
P. ISLAND 182 SEP, 29, 1986 Nov. 3, 1986 -
ANO-2 NOV. 17, 1986 DEC. 22, 1986 ,
FT. CALHOUN JAN, 5, 1987 FEB. 9, 1987
, ANO-1 NOTE (1) NOTE (2) ;
[
- R. SEC0 NOTE (1) NOTE (2)
_ C. RIVER ~ NOTE (1) NOTE (2) iO NOTE (1) THE NEED FOR AND SCHEDULE OF VISITS TO THESE PLANTS WILL BE DETERMINED IN LIGHT OF THE ONGOING STAFF WORK ON THESE PLANTS.
I NOTE (2) REVIEW TEAM REPORTS FOR THESE PLANTS WILL BE COORDINATED WITH ONGOING STAFF EFFORTS.
1 I
i V
. O f
v- -,-,s- -y e. - - , - - - - . - - --,m , , , . ,w_ r - - .~ . - ,-- - - -
, - ,e w4-.
O PRAIRIE ISLAND UNITS - 182 POST TMI AFWS MODIFICATIONS
- II.E.1.1 AND II.E.1.2 0F NUREG-0737
- GL 81-14 SEISMIC RESISTANCE OF AFWAS
- 10 CFR 50.49 EQ OF ELEC. EQUIPMENT
- TD AFWPs ARE MADE AC-INDPENDENT t
I
- TURBINE STEAM ADMISSION. VALVE - AIR OPERATED
- AFWPs PRE-LUBRICATION
- TURBINE STEAM ADMISSION VALVE RELOCATED OUTSIDE THE PUMP ROOM AND O MODIFIED TO ALLOW LOCAL START OF THE TD AFWP
- MANUAL VALVES
- OUTSIDE CONTAINMENT: ARE LOCKED OPEN. T. SPECS - MONTHLY INSPECTION
- INSIDE CONTAINMENT:
- DOUBLE VERIFICATION OF EACH VALVE BEFORE PLANT STARTUP
- AFWPs USED AS WATER SOURCE AT STARTUP
- MAINTENANCE OF THESE VALVES IS DONE BELOW 350*F RCS TEMP.
O
O PRAIRIE ISLAND UNITS - 182 POST TMI AFWS MODIFICATIONS (CONT )
- EMERGENCY PROCEDURES HAVE BEEN REVISED TO INCLUDE:
- TRANSFER OF AFWPs SUCTION TO ALTERNATE WATER SOURCE
- T. SPECS REVISED TO LIMIT OUTAGE TIME OF AN AFWP TO 72 HR i
- CONTROL ROOM
- ANNUNCIATION IF TD AFWP OVERSPEED MECHANISM IS TRIPPED
- DUAL CST LEVEL INDICATION UNITS 1 & 2 T. SPECS REQUIRE O N E S CONNECT MANUAL VALVES TO BE TAGGED IN THE O
- STEAM BINDING 0F AFWPs (IEB 85-01)
- CHECK TEMPERATURE BY TOUCHING PIPING CLOSE TO PUMPS ONCE EVERY l 4 HRS
- AFWPs PAINTED BY TEMPERATURE-SENSITIVE COLOR PAINT
- MOVs LIMIT, TOROUE AND BYPASS SWITCH SETTINGS (IEB 85-03)
- STAFF REVIEW 0F SUBMITTAL IS UNDERWAY
- MOST MOVs ARE IN CORRECT POSITION FOR AFWS OPERATION
- OPERATOR INTERVENTION IS ONLY REQUIRED TO SWITCH AFWS SUCTION OR FOR RECOVERY OF MALFUNCTION O
__. n s . ._ ... - __ - . - - .. - .- . _ , . . . - . . .
l l
l
- ;O i
l n
PRAIRIE ISLAND AFWS RELIABILITY f
STAFF ' LICENSEE 1
- BENCH MARK
- BEST EESTIMATE*
NUREG 0611 PRE TMI Posi IMI Posi IMI '
LOMF 1.0E-3 3.4E-3 5.4E-5 2.0E-5 L
+
I: -LOOP 1.0E-3 3.4E-3 5.4E-5 8.7E-5
- O
- AFWS RELIABILITIES ESTIMATED BY LICENSEE AND HAVE NOT YET.BEEN VERIFIED.
t 4
t I
i 1
l n
N.
.IN r E
N.
E OA E NT A.
C C O E M M CC o A A N R e E E 2 A R 2u I E N%
T I T O O T A
INO I
2 I ,- E E
F 2
T 0
N &
3 A 9 r 0 2
it l 4 3 in i 2 u N n B v-u A U
, i u -uv n
H
)
H
' 8 I 4
2 .
2 i 3
W 3,
LI 4
= u ,4 3
2 Xi.- c m
3 3 -
v
. u u
u u -
l 4
b s F 9
4 u-2 0
2 v 2 2
3 u u 3 W
v l
B u-m =
3 2, H
)
e 8 3
2 e, 3 W
= J LS
, 3 g
. 2 g 3 a3g . W F M wVA AMg .
3 LY3 !.
F 6 .
2 e
3 2
A 2 0
2 X4c -
3 n v u v L - - Eu-ul ,
- - )
3 4
2 2
3 W
- 3, 6 L3
, 3 W M 3 3, 2 ir 3 -
, w u u u ;
k l .
s 2 n 5 2
4 2
- i. 0 2
v 2 3
3 u u w
l B v-M u
9
' H
)
3 2
2 3
W - 2\ ( . 3 L3
=
- - 3
- c 2 3
w
- : a r
N. N.
E E C
u C
M
- d. ... O
.2 :
- n ,A - NE o M , 5:,
a A f e
OC I N
l E
1 E
T I CA= I 't,,
$ S t l RN A N 2
Oi M Lo a t s % lT e S N
c U
- Ii 4 !
j1 i1I < ffij ij
~
o o o.
RELIABILITY ANALYSIS OF AFWS OPERATIONAL DATA '
PLAN - ANALYZE AFW COMPONENT AND TRAIN LEVEL FAILURE DATA T0 DETERMINE POTENTIALLY.SIGNIFICANT INCIDENTS OR EVENTS AND MAKE INFERENCES REGARDING AFWS RELIABILITY, SCOPE - USE LER AND NPRDS DATA FROM 1981-85 INCLUSIVE FOR:
~
ARKANSAS NUCLEAR ONE, UNIT 1 ARKANSAS NUCLEAR ONE, UNIT 2 CRYSTAL RIVER, UNIT 3 FT. CALHOUN PRAIRIE ISLAND, UNIT 1 PRAIRIE ISLAND, UNIT 2 RANCHO SECO m
AE0D 9/8/86
_ _ _ ~
i O O O- '
i j .i I
i i EFFORTS REQUIRED .
- 1. OPERATIONAL DATA NOT DIRECTLY USEABLE 1
- 2. REVIEW EACH REPORTED EVENT.FOR: l I
{
COMPONENT OR FUNCTION EFFECTED VERIFY COMP 0NENT,.ETC., AGAINST SYSTEM DRAWING SUMMARIZE EVENT INTO ONE-LINE DESCRIPTION i
i 5
I
.I
,. . , , . , - . . - , . . . , , - - . . . . _ . . T=*5 . . . , - . _ . .
--Awea we --4
--a-4. . - e-ehau e w .lu -m-anme,.... a,edumm., * , A-..A4h.'-ue.m. - --aw w w -4e - eneu. -
e.h+ha%. %4 e.w-4.Saews=h..9- ease-e-m .a.eg, a- .e- '-
O O o AFWS DATA ANALYSIS PERFORMED IN TWO PHASES
- 1. USE LER DATA FOR QUALITATIVE INSIGHTS
- 2. USE LER AND NPRDS DATA FOR QUANTITATIVE INSIGHTS .
w w
0 0 0 PHASE 1 RESULTS EACH PLANT HAS ITS OWN PARTICULAR PROBLEMS PROBLEMS CAN GO UNCORRECTED FOR LONG PERIODS AFWS RELIABILITY CAN HAVE LARGE VARIATIONS WITH TIME
. . - - - - . _ = .- - . - _ - . .- .-
b
- PHASE 2 EFFORTS i
INCLUDE NPRDS RECORDS IN ANALYSIS USE GENERIC FAILURE RATES TO ESTIMATE EXPECTED FAILURE I QUANTITIES i
i I -
COMPARE EXPECTED FAILURE QUANTITIES WITH REPORTED
! FAILURES i
1 i
I i
i l
i l
i 1
O O O SAMPLE CALCULATION OF EXPECTED TD PUMP FAILURES
- USE GENERIC UNAVAILABILITY, 0 = 10-2/ DEMAND ASSUME TD PUMP COMMANDED ON = APPR0X. 15 TIMES / YEAR PERIOD EVALUATED = APPR0X. 5 YEARS - .
.. (15)(IE-2)(5) = APPR0XIMATELY 0.8 ESTIMATED EXPECTED FAILURES PER PLANT A "
l' MAJOR CATEGORIES OF REPORTED AFWS FAILURES AFW PUMP FAILURES -
FAIL TO START FAIL To RUN AFW FL0w AFW FLOW PLANT NAME TD MD TD AND MD INSTRUMENTATION CONTROL VALVES ANO-2 9 0 1 0 37 CR-3 8 2 5 7 3 ANO-1 6 6 9 0 8 -
RANCHO SECO 5 1 3 7 2 FT. CALHOUN 3 1 1 0 4 PRAIRIE I. 1 3 0 0 0 0 PRAIRIE I. 2 3 0 0 0 0 1
EXPECTED 0.8 -
0.4 0.09 0.6 l
w .-.
O O O BREAKDOWN OF REPORTED PUMP FAILURES PUMP FAILURE TYPES ANO-2 CR-3 AND-1 R.S FT. CAL. P.-l. 1 P.I. 2 GOVERf!OR/ CONTROLS 5 4 5 3 2 0 3 -
PACKIllG/ SEALS 1 2 2 1 1 0 0 BEARINGS 0 2 4 2 0 0 0 STEAM INLET VALVE FOR TD PUMPS 4 4 1 3 1 1 0 PUMP DRUM ALIGflMENT/ ADJUSTMENT 0 0 7 0 0 0 0 MISCELLANEOUS 0 3 2 0 1 3 0
_ J
O O O CONCLUSIONS HIGHER THAN NORMAL FAILURE RATES FOR SOME AFWS COMPONENTS AFWS TESTING EMPHASIZES START - ENDURANCE ALSO IMPORTANT POTENTIALLY SIGNIFICANT AFWS PROBLEMS (E.G., BEARING FAILURE)
INCREASINGLY REPORTED IN LESS VISIBLE NPRDS l
~
j jl
. G A G G oM T S
T S
O O
'M : r O
N N M
M q4> ' 4 , ,
7 > r N
O I
M I A
T R
MU l
TG M OI EN F
N RO APC M h M
N N M
M -
M I u
)
M E AIN 8 7
7BR -
PU P M T
.. (
A' M E :
CR I
d i
VE M
/
\.
=
+ rg 9;
rL 3J m r m c :
h T R E m S C
E S
N h m
f .
E D
N O
C F
~
m . ! <
i i
- O i
ANO-1 POST TMI RESTART MODIFICATIONS (1979) i i
I o MOTOR DRIVEN PUMP ICS AUTO START o CONTROL GRAM FLOW INDICATION I
i i
\ o CONTROL GRADE ANTICIPATORY REACTOR TRIP
\
O LOSS OF BOTH MFW PUMPS j TURBINE TRIP o EFW AUTO START ANNUNCIATION IN CONTROL ROOM O ..
O 1980 EFW SYSTEM MODIFICATIONS o NNI AUTO INITIATION LOSS OF ALL RCPs LOSS OF MFW PUMPS LOW OTSG LEVEL o MOTOR ORIVEN PUMP CONNECTED TO SAFETY GRADE POWER o SAFETY GRADE FLOW INDICATION o SW SUCTION - REDUNDANT SOURCE Q
o EFW LOW SUCTION PRESSURE ALARM o SAFETY GRADE ANTICIPATORY REACTOR TRIP LOSS OF MAIN FEE 0 WATER TURBINE TRIP o UPGRADE 0 NNI POWER SUPPLIES (CR-3) o MOTOR S TURBINE ORIVEN PUMP BEARING COOLING SYSTEM A00E0 0 ..
D
O O -
O 1
1 i
i I HL PDIS PAL PS D ,,g h FiO ONNI-Y
- O-O / -
l SERVKE k ,, h "J I'!
- 4 "J l TO SG E248 WATER -
i ,, ,,,,; ,
! LOOP 21 : }
HS P7A l -
! FROMCST :
"J l" g pn TOFi ,9NNI-X
! HSh < r PS --OPDT '/ .
TO i
j SERVICE AM
_M ,,
y REclRC.p) a- (,/
- h; m.
j r TO SG E24A WATER - ,, ,,
.a FO l LOOPI _
P78 ..
~.
ma v,
W wa PDIS PAL , , "g
?[ - -CONDENSER l
r RECIRC.TO CST 1 r l TO RECIRC. ?I 0 'H 'H '6-- '
' L 74 @
! >C r CIRC. WATER FLUME l
ANO-1 i
^ mm POST-1980 CONFIGURATION l
1 l
1 -
1982 EFW SYSTEM MODIFICATIONS Q
o EFIC CABINETS INSTALLED AND POWERED o NEW C09 INSTALLED o OTSG LEVEL MEASUREMENT SYSTEM OTSG LEVEL TAPS NEW LEVEL TRANSMITTERS
, EFIC COMPENSATION INDICATION ON C09 INDICATION ON SPDS o OTSG PRESSURE MEASUREMENT SYSTEM 0 MSL TAPS PRESSURE TRANSMITTERS EFIC COMPENSATION INDICATION ON C09
- i INPUT TO STEAM LINE ISOL. SYSTEM j INDICATION ON SPDS o PROTECTION SYSTEM INPUTS TO EFIC RPS, LOSS OF RCPs, LOSS OF MFWPs ESAS LOSS RCS PZR OR HIGH RB PZR O ..
O 1982 EFW SYSTEM MODIFICATIONS (Continued) l l
. o NEW EFW TURBINE ORIVER INSTALLED ELECTRONIC SPEED CONTROL W/ RAMP EXISTING STEAM ADMISSION VALVES .
TO STEP OPEN I EFW STEAM PIPING MODIFIED
! NEW DC STEAM ADMISSION VALVES I WELDED IN - LOCKED OPEN
!O o EFW RELATED ANNUNCIATION CONSOLI0ATED ON K12
\
l 9
1 l
l
?
l; O
O 1983 - 84 EFN MSIGN ACTIVITIES o DISCHARGE PIPING MSIGN o ELECTRICAL COMPONENT INTEWACE MSIGN o ENVIRONMENTAL QUALIFICATION o INCLUSION OF SETPOINT ANAL YSIS O
'l i
I e
l
O 1984 EFW SYSTEM MODIFICATIMS i
o EFW DISCHARGE PIPING G VAL VES INSTALLED o EFIC CONNECTED TO CONTROLLED COMPONENTS o EFW AND EFIC CONTROLS INSTALLED M C09 o
EFW AND EFIC ANNUNCIATIM INSTALLED ON K12 o SAFETY GRADE DC MCCs INSTALLED o
EFW TURBINE DRIVER STEAM ADMISSION SYSTEM CONNECTED
.e O s, t
O o -
a-(====m' _
I I "o" CST I I L i -- J rT SW LOOP I T7 17 a
^ ^ '
Y CV3850 l- _I <,i
!
- YARWAY CHA CHD l l CV 2646 CV 2670 CV 2803 RECIRC
,t a M d I VALVE ]p _j A STEAM I I T RATOR i g 8
AC o' hw CV 2627 HEADER l P7B '1I h 645 l -T MOTOR FI S l ,
gJ l DRIVEN CH. A ,,
M I TO CST d- ;TO FLUME D FC H . " CV 2870 F
g TO CONDENSER 3
J g C ' .T AC YARWAY CV2869_ CHC
, g fRECIRC VALVE p, CHB S V2647 /_ V2620 t ,
l l
I aN ..
\
1 d '
]
CV 2006 P7A s'l CHA CV 648 S
CHD CV2626 TURBINE I C A ,
h- DRIVEN DC AC i 1 g CH. 8 CV 3851 ANO-l PRESENT 1 A -
o O -
ICS(LOW VACUUM) TRN A CONT TRN B MSLI TRN A MSLI CONT CV 2618 CV2619 CV 2676 CV 2668 Q FC ATMOS MC FC
?4 TRN A MSLI 8 SAFETY T B LI VALVES TRN A EFW V S O TRN SLI 6 TRN B EFW CV 2692 MN CV 2617 STEAM ik AC ' ST M LC r, w CV 2666 8
A 1 7
~
GE E TRN B EFW-- r s' -,f CV 2613:
' CV 2663 DC Mi 2 63 h DC U TRIP THROTTLE VALVE GOVERNOR VALVE I f K3
) ANO-l MAIN STE AM SYSTEM 1985 CONFIGURATION
~
O ano_1 STEAM ADMISSION TO TURBINE FOR EFW PUMP P-7A
( ORIGINAL )
O O A >4 34 a .
AC (GREEN)
O ,
O Y
O O ,
O ANo-i STEAN ADMISSION TO TURBINE FOR EFW PUMP P-7A
( iR6 )
Ac (RED) Ac (GREEN)
A B O
i i oc 1 (GREEN)g 2
% (oc RED) l t
O
.,n_-_- -... _ - - - -.,,__..,._ - - . - - - - - . _ _ - . - - - . , , _ , , - - ---a ,-
i o
ANO-1 i
STEAN ADMISSION TO .
- TURBIE FOR EFW PUNP P-7A
- ( JANUARY,1986 )
i
A Z 'A B lO I
1 i
4
)
o .
, _ _ _ _ - _ _ _ _