ML20137W698
| ML20137W698 | |
| Person / Time | |
|---|---|
| Issue date: | 02/12/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1487, NUDOCS 8602200310 | |
| Download: ML20137W698 (237) | |
Text
.
Ef/f87 O UhITED STATES NUCLEAR REGULATORY COMMISSION
__ _ ORLGJNAL IN THE MATTER OF: DOCKET NO:
ADVISORY CO2D1ITTEE ON REACTOR SAFEGUARDS SUBC0bO!ITTEE ON T!iE WESTINGliOUSE WATER REACTORS O
C) ,
LOCATION: WASilINGTON, D. C. PAGES: 1- 182 DATE: WEDNESDAY, FEBRUARY 12,.1986 22 0 060212 T-1487 PDR wgg -
U L h
)0NOK; lam 0y0f70mhC[]3D}l$
\
p' ACE-FEDERAL REPORTERS, INC.
qk) OfficalReporters 444 North Capitol Street Washington, D.C. 20001 0g (202m7 3700 NATIONWIDE COVEPAGk L
i f
I CR25726.0 BRT/sjg <; 1 1 UNITED STATES OF AMERICA C) 2 NUCLEAR REGULATORY COMMISSION I
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 3
SUBCOMMITTEE ON l 4 4
! THE WESTINGHOUSE WATER REACTORS 5
l 6
Nuclear Regulatory Commission Room 1046 7
1717 H Street, N.W.
I Wa sh ingtor. , D. C.
8 9 Wednesday, February 12, 1986 1
l 10 l The meeting of the ACRS subcommittee convened at 11 l 8:30 a.m., Mr. Jesse C. Ebersole, chairman, presiding.
l 12 l ACRS MEMBERS PRESENT:
E 14 MR. JESSE C. EDERSOLE MR. IIAROLD ETriERINGTON 15 MR. CARLYLE MICHELSON DR. CIIESTER P. SIESS i MR. GLENN A. REED l 16 ' MR. DAVID A. WARD I
. MR. CIIARLES J. WYLIE f 17: DR. WILLIAM KERR l l l 10 F p DR. IVAN CATTON, Consultant !
19 i
!i DR. RICilARD SAVIO, ACRS Staff Member i h I 20 D 1"
il !
21 'j !
I l 22 J i
23 h i 2
M Forfe,cl Retorters, Inc. I 25 l
I !
'I N I l
PUBLIC NOTICE BY THE UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS WEDNESDAY, FEBRUARY 12, 1986 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
No member of the ACRS Staff and no participant at this mee, ting accepts any responsibility for errors or inaccuracies of statement or data contained in this transcript.
I O
l l 25726.0 2 l BRT
('N i #
1 PROCEE D I NGS 2 MR. EBERSOLE: The meeting will now come to 3 order. This is a meeting of the Advisory Committee on 4 Reactor Safeguards Subcommittee on the Westinghouse Water I 5 Reactors.
l 6 I am Jesse Ebersole, the subcommittee Chairman.
7 The other ACRS members are: Mr. Etherington, 8 Mr. Michelson, Dr. Siess, Mr. Reed, Mr. Hard, and 9 Mr. Hylie.
I 10 Our consultant, Dr. Catton, is here assisting us 11 i' our review.
l 12 The purpose of this meeting is to discuss the
(~l v 13 November 21, 1985 water hammer event at SONGS-1.
14 Copies of the agenda for this meeting are available and 15 will be passed out to anyone who has not yet received a 16 copy. The topic is also scheduled for discussion by the 17 ACRS tomorrow afternoon.
18 Dr. Savio is the cognizant ACRS Staff member fo..
19 this meeting. The rules for participation in this meeting 20 have been announced as part of the notice of this meeting 21 previously published in the Federal Register on January 31, 22 1986. A transcript of the meeting is being kept and will 23 be made available as stated in the Federal Register notice.
24 It is requested that each speaker first identify himself or 25 herself and speak with sufficient clarity and volume so O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 3 BRT
,m k_ 1 that he or she can be readily heard. He have received no ,
2 written comments or requests for time to make oral 3 statements from members of the public.
4 I just want to make a few observations. It 5 seems to me, this is another case where NRC is -- intends l
t
! 6 to wait until some sort of a physical event occurs to 7 attempt the fix of a problem.
i 8 For instance, we had Salem for the scram case, l
1 i 9 we had the Brown's Ferry fire for the fire case, we had 1
10 Davis-Besse for the aux feedwater case, and in each of i 11 these cases, the problems were reasonably well known in 1
12 advance, that there was intrinsic potential for damage to
'T 13 t
(G the station by malfunction of a piece of equipment.
14 In the check valve case it's a very old issue, 15 indeed. I recall many years ago a discussion with one of 16 the consultants who said, we have been asked to investigate 17 the dynamics of check valves and the associated 18 motor-driven valves and pipes in certain designs, and when 19 he got to a certain point in the investigation his customer 20 found things weren't going so well so he turned off the 21 investigation and that was the conclusion of that study.
22 These check valves that failed here will be a 23 catalyst to have the check valve problems at large 24 investigated, I hope. Certainly, on a more general basis I 25 hope we can take this as the incentive or the catalyst to fm
\-]
ACE FEDERAL REPORTERS, INC.
202-347-3h Nationwide Coverage 800-336-6M6
25726.0 4 BRT r3
\ l 1 start a broader or more general study of the check valve 2 and associated valve and pipe dynamics at large.
3 Again, this seems to be the catalyst event to 4 take this on since we have been unable to get it going 5 before this time.
6 I believe our first speaker is " r. Richard 7 Dudley. Mr. Dudley?
3 fir. REED: Jesse, I would like to make a comment.
9 It's my recollection that San Onofre is, very early, the 10 second commercial reactor, utility reactor, Monroe being 11 the first, San Onofre being the second. And it's very much 12 a one-of-a-kind design that -- never was the safety
() 13 injection or feed systems done subsequently that same way.
14 I would like to have some indication, based on what you 15 just said, how that relates to whether we have a generic 16 issue or whether we have a one-of-a-kind issue.
17 P1R . EBERSOLE: Hell, I believe it's true that 18 today if you look at check valves at large you'll find some 19 that are simple, old-type flapper-type valves that crank 20 into their hardened seats with great gusto and have the 21 potential of carrying the parts backwards into the backup 22 valves, thus causing them to cascade into failure. We have 23 others that have hydraulic piston cylinder arrangements to 24 gradually let the thing come down on its seat, consistent 25 with the admission of high-pressure fluids to the back end.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationaide Coverage 800-336-6M6
I 25726.0 5 BRT
/m
(I 1 So there's little, if any, real standardization or common 2 design at present in the industry and I think we need to 3 look at that.
4 MR. SIESS: Is it good or bad that they are not l 5 standard?
6 MR. EBERSOLE: Oh, I think it's bad that we 7 don't have a common standard for design of these things.
8 MR. SIESS: If we had them all standard then 9 they'd all be failing at once.
10 MR. EBERSOLE: That's always the perennial 11 argument.
12 MR. SIESS: I wonder where we should be going.
() 13 MR. EBERSOLE: If we have one bad accident it 14 colors the whole industry.
15 MR. SIESS: Especially if all the plants are 16 alike.
17 MR. EBERSOLE: I don't know how to put that to 18 rest. Well, go ahead.
19 Well, in this this case, before we get started, 20 had there ever been a study of check valve dynamics at this 21 plant?
22 MR. DUDLEY: I am not aware of one.
23 My name is Dick Dudley, I'm the project manager 24 for San Onofre Unit 1. I have had that responsibility 25 since July 11, 1975.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 6 BRT l
1 MR. MICHELSON: Can you move the microphone a 2 little closer to your mouth? It's not going over tc3 well.
3 flaybe I better hold it.
4 MR. f1ICHE LSON : It's now getting better.
5 f1R. DUDLEY: What I would like to talk about 6 today is a brief rehash of certain licensing activities 7 that have taken place and are currently taking place at San 8 Onofre Unit 1, in order that the committee gets a better 9 idea of activities at the site in addition to the 10 activities that are ongoing now with relationship to the 11 water hammer and the resolution of those concerns.
12 (Slide) 13 San Onofre Unit I was shut down on February 27,
{ })
14 1982, for a steam generator inspection. I apologize for 15 the error in the slide, but the steam generator inspection 16 was the primary reason for shutting the plant down. During 17 that shutdown, during the SEP review, concerns were raised 18 regarding the ability of the plant to meet its original 19 seismic design basis of 0.5 G.
20 Because of those concerns the utility at that 21 point committed to upgrade the point to a .67 G level, 22 which is consistent with the seismic design of San Onofre 23 Units 2 and 3 also located on the site. The G level is the 24 same bu' the response factors are somewhat different.
25 The NRC, on August 11, issued a confirmatory O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
25726.0 7 BRT b,m 1 order which confirmed the Licensee's commitments and 2 precluded restart of the plant until the seismic concerns 3 were resolved. During this period of time, restart of the 4 plant was somewhat uncertain. Because of that, the utility 5 deferred quite a bit of licensing work that was -- would 6 have been done on Unit 1, except for -- they did proceed 7 with the SEP review activities.
8 Because of deferriin this work, a backlog of 9 licensing actions began to accumulate, or certainly if 10 there was a backlog it wasn't reduced.
11 December 12, 1983, the utility submitted a 12 return-to-service plan for San Onofre Unit 1, and this plan
() 13 included the seismic upgrading of safety-related structures:
14 electrical systems and plant equipment that was needed to 15 achieve hot shutdown.
16 In addition to this, the Licensee proposed an 17 integrated living schedule methodology that would allow the 18 utility greater flexibility in implementing many of the 19 capital projects that would be ultimately coming out of the 20 backlog of licensing actions that had accumulated.
21 In November of 1984, the NRC allowed the 22 facility to restart after they completed the seismic upgrades 23 that are shown here.
24 (Slide.)
25 The systematic evaluation program was still p
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6M6
25726.0 8 BRT k
1 ongoing and the draft integrated plant safety assessment 2 report was issued in April 1985. Briefings were provided 3 to the ACRS on the staff's SEP work on June 19, 1985 to the 4 subcommittee; on August 9 to the full committee. On Auoust 5 13 the ACRS issued a letter which supported the Staff's SEP 6 conclusions.
7 The utility had originally planned to operate 8 Unit 1 from November 1984 until November 30, 1985, when 9 they planned to shut down the plant and proceed in 10 accordance with the integrated living schedule to complete 11 a significant number of plant modifications. This would 12 include completion of the seismic modifications to the
() 13 plant for the remainder of the systems necessary to 14 mitigate accidents and to achieve the cold shutdown. This 15 would include completion of the equipment qualification 16 program.
17 We had issued an exemption to the utility that 18 allowed them to operate until November 30, with 70-some 19 pieces of equipment that did not fully meet the EQ 20 guidelines. Those will all be replaced during the outage.
21 The Licensee also is completing fire protection 22 modifications. This would include the installation of a 23 dedicated shutdown system for fires in certain areas. The 24 dedicated shutdown system will have its own diesel. This 25 will be a nonsafety-related diesel.
O V
1 ACE FEDERAL REPORTERS, INC.
l 202-347-3700 Nationmide Coserage 800-336-6646
25726.0 9 BRT Y l Licensee also intends to install a third 2 auxiliary feedwater pump, and for the next operating cycle 3 this auxiliary feedwater pump will be powered by the 4 nonsafety-related diesel from the dedicated shutdown system.
5 The licensee's plans then are to upgrade this third 6 auxiliary feedwater pump to a third train of auxiliary 7 feedwater in the next refueling outage. However, on 8 November 21, 1985, the water hammer event occurred. Due to 9 the significance of the event, the NRC accident 10 investigation team was dispatched to the site.
11 The Licensee then chose to begin their refueling 12 outage, which was scheduled for November 30. They entered
() 13 that outage nine days early.
14 MR. REED: Could you just clarify that dedicated 15 shutdown system? What do you mean by a " dedicated shutdown 16 system"? Does that have to do with decay heat removal or 17 does it have to do with something else?
18 MR. DUDLEY: Yes. It would allow the Licensee, 19 for example, in the event of a fire in the control room, it 20 would allow the Licensee to open certain breakers and take 21 control of the plant from this remote system and allow them 22 to go all the way to cold shutdown condition.
23 MR. REED: That's something that has been in the 24 cards for five years or so?
25 MR. DUDLEY: Yes, it is. I just wanted the O>
l l
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6 4 6
25726.0 10 BRT U 1 committee to know the licensing status on the 2 implementation of some of these items.
3 MR. EBERSOLE: Really that's too broad a term 4 for it. It's really an auxiliary control system; alternate 5 control system. It's no alternate power, just an alternate 6 control for the loss of control to the control room.
7 MR. MICHELSON: This is just an Appendix R 8 modificatir.n?
9 MR. DUDLEY: Yes, it is. These are all the 10 remarks that I have prepared. If you have any questions 11 I'd be glad to go over them. Okay.
12 (Slide.)
I~'; 13 MR. MARTIN: What follows is the report of the
\~/
14 accident investigation team for the San Onofre event.
15 (Slide.)
16 The team was composed of myself, Thomas T.
17 Martin from Region 1; Matt Chiramal, who is in the back --
18 would you raise your hand? Bill Kennedy, Wayne Lanning, Al 19 Serkiz, and Steve Showe who is not here today.
20 (Slide.)
21 One of the things we found early on in the 22 investigation was we needed to understand the plant, 23 because of its unique design. It's operated by Southern 24 California Edison, located south of Los Angeles near San 25 Clemente, Westinghouse three-loop PURs. It was licensed in
(.,
'V)
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
25726.0 11 BRT
(.
1 1967. It's a spherical containment with concrete enclosure 2 building, has electric main feedwater pumps that also 3 function as safety injection pumps.
4 (Slide.)
5 Quick overview. Here's the containment, the 6 turbine -- this flash evaporator feedwater heating will 7 play a role in the event. There are heaters on both sides.
8 The auxiliary transformer, which is -- acts like a startup 9 transformer, will play a role. Unit 2 and Unit 3 are back 10 over here. The circ water pumps are back down here and the 11 Pacific Ocean is there. Diesels are here, the control room 12 is in this area. The thing called the feedwater mezzanine
(~ ; 13 is underneath this area, and this is where we'll actually V
14 have the major steam leak develop.
15 MR. EBERSOLE: With the feedwater pumps acting 16 as safety iniection, would you clarify how they handle the 17 fluids to get borated water into it?
18 MR. MARTIN: Actually pump trips, valve shifts.
19 It then takes a suction from a thing called a safety 20 injection pump, wnich is like a booster pump, comes right 21 out of the RUST, into the safety injection pump, into the 22 feedwater pump and then the reactor coolant loops.
23 The feedwater pump is a booster pump. I forget 24 what the shutoff head is. It's basically an intermediate 25 head system.
(';
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 12 BRT
/~~T 1 MR. EBERSOLE: And then it takes auxiliary 2 suction then from the borated water source which is also 3 pumped? -
4 MR. MARTIN: Yes, sir.
5 MR. EBERSOLE: Thank you.
6 (Slide.)
7 It also has one turbine and one electric 8 auxiliary feedwater pump. Dick Dudley spoke about a second 9 electric pump that will be installed. It was one immediate 10 and one delayed-access off-site power source. We'll talk 11 about that in a second.
12 It has diesels that do start but do not
() 13 automatically load on loss of powe only. It takes a 14 safety injection in conjunction with a loss of power for 15 them to automatically load.
16 MR. EBERSOLE: May I ask on that particular, 17 that third bullet, where was the automatic loading denied?
18 It would have appeared, had it been present, that some of 19 these things wouldn't have happened?
20 MR. MARTIN: I can't answer that. Dick, can you?
21 MR. DUDLEY: No. The diesel was installed in 22 1976. I can't really say what occurred during that review.
23 MR. EBERSOLE: Perhaps it couldn't carry the 24 load. I don't know. We'll see.
25 MR. MARTIN: They have some new diesels there.
(y V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 13 BRT
~'
1 They are capable carrying the safety-related loads on the 2 buses, but the buses have other than safety-related loads 3 on them, and we'll show that when we go through this little 4 educational slide.
5 The first one here is the condensate system.
6 You have four condensate pumps with discharge check valves.
7 These are made by the same manufacturers as the ones that 8 failed. And then a common header. Nothing prevents the 9 flow from coming like this, basically it has a common 10 header.
11 Here the flash evaporators -- this is the one 12 that's going to fail and the condensate train goes up and em
() 13 meets the main feed system.
14 Here's the alternate suction from the safety 15 injection pump feeding into the pump. These are 16 hydraulic-operated valves here. Basically this is what 17 shifts around, and you start having boric acid -- borate 18 water going to the reactor coolant loops during the safety 19 injection. The first point, feedwater feeders, these are 20 the high pressure ones, up around 1300 pounds. The system 21 back behind here in the condensate system is a 160-pound 22 system. Two pumps, again. Nothing to prevent flow back 23 through here except these check valves. These check valves 24 are going to become important. These check valves are 25 going to fail and all of these check valves are going to
()
~
m ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33 6 6646
25726.0 14 BRT O"'
1 fail up here.
2 There's a common header, then a feed to each 3 steam generator, feed control valve, a bypass control valve 4 for low flow, check valves -- these are, like 10-inch.
5 This is a four-inch check valve then into the steam 6 generator.
7 The AFU comes in ahead of these check valves 8 just before it enters into containment, so these need to 9 seat to make sure the AFU goes to the steam generators.
10 This valve and this valve, the disc will be 11 completely disconnected from its hinge arm. This valve 12 here will be loose and actually will stick when it tries to 13 shut.
14 These valves, which are the four-inch, seem to 15 seal fine, although the B one receives the major water 16 hammer damage and its bonnet is the one that is lifted and 17 you have the major steam water leak.
18 (Slide.)
19 f1R. MICHELSON: Excuse me, before you leave that 20 slide, there was something in the report that referred to 21 the fact that on the B line the valves were throttled to 50 22 percent? Or they were each 50 percent flow or something?
23 MR. MARTIN: That's the AFU system, sir.
24 There's a single control valve that feeds this. To make 25 sure that they have the redundancy, there are two 50 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-316-6646
25726.0 15 BRT
\ _)
1 percent valves which feed this, one from each one of the 2 AFW trains.
3 MR. MICHELSON: On the 50 percent valves were 4 they set to throttle at 50 percent?
5 MR. MARTIN: They were set to control 50 percent 6 flow, so together they deliver the 100 percent.
7 MR. MICHELSON: They were the same size as the 8 others but just throttled back?
9 MR. MARTIN: Yes, sir.
10 MR. REED: The penetrations thro'igh containment, 11 are you going to show those? I guess if you talk about 12 containment penetrations, the rules for this class would be
() 13 probably one check valve outside containment and one inside 14 containment; is that correct?
15 MR. MARTIN: No. I don't believe that's correct.
16 This is not considered -- this is a closed system, 17 basically in accordance with Appendix J.
18 MR. REED: A closed system?
19 F1R. f1ARTIN: Yes. It goes right to the steam 20 generator --
21 F1R . MICHELSON: They aren't using the good rules 22 at all.
23 MR. REED: Are there more check valves between 24 that point and the steam generator?
25 f tR. MARTIN: No, sir.
A G
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-66:6
25726.0 16 BRT s
1 MR. REED: None inside containment?
2 MR. MARTIN: No, sir. This is the main steam 3 system. Now there's a common generator for all steam 4 generators to feed, therefore a problem in any one is 5 communicated to all three. The feed lines feed three, and 6 then here's the steam lines coming out. These manual 7 valves, and they are manual, not motor-operated isolation 8 valves -- the operator has to go down and put a ring on the 9 valve to shut it. This is how they stop the cooldown that 10 might result from leakages down that way or from problems 11 with the turbine stop valves or governor valves.
12 MR. EBERSOLE: Let me ask, in a shutdown mode
() 13 where you are going to go to evaporator coolant to the 14 steam generators, this says it must be able to do that at 15 virtually atmospheric pressure because you have a potential 16 for opening all steam generators to atmosphere.
17 MR. MARTIN: Yes, sir.
18 MR. EBERSOLE: Is the emergency cooling mode 19 using the steam generators hypothetically done at very low 20 secondary pressures?
21 MR. MARTIN: Whether it's designed to do that I 22 can't answer. Dick, can you answer that?
23 MR. DUDLEY: No.
24 MR. MARTIN: He have a presentation by the 1 25 Licensee afterwards. He can probably answer. l l
ACE-FEDERAL REPORTERS, lNC. I 202-347-3700 Nationwide Coserage 804 336-6646 l
25726.0 17 BRT 1 MR. EBERSOLE: It's a neat trick to do it 2 without overcooling.
3 MR. MARTIN: True. They reduce the flow to the 4 steam generators to the lower rate they use, 25 --
5 MR. EBERSOLE: Do they lower the water level?
6 MR. MARTIN: They let it go all the way off 7 scale low; basically they protected the primary to the 8 secondary.
9 MR. MICHELSON: Do you know if the Licensee ever 10 did a three steam generator blowdown analysis upon the 11 effect of primary reactivity, et cetera?
12 MR. MARTIN: I know there was analysis performed
() 13 for main steam line and feedline breaks. Since the design 14 hasn't changed in that respect it would have to have 15 assumed all three blew down.
16 MR. MICHELSON: So this is a three steam 17 generator blowdown case.
18 MR. MARTIN: That's correct. I know the 19 analysis has been done. I don't know the details of it.
20 MR. EBERSOLE: I would suspect, then, the 21 containment would have to tolerate, of course, secondary 22 blowdown into the containment, the contents of all steam 23 generators.
24 MR. MARTIN: I know in designing the AFW, that 25 was one of the things they looked at.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33& 6646
25726.0 18 BRT O 1 MR. EBERSOLE: So we still have to backtrack 2 into history a little bit to see. Thank you.
3 MR. MARTIN: Yes, sir?
4 MR. REED: Follow that steam line out and you 5 come down to the first valve which you said they had to 6 crank closed?
7 MR. MARTIN: Yes, sir.
8 MR. REED: Is that a stop check valve or just a 9 manual stop valve?
10 MR. MARTIN: It's my understanding it's just a 11 manual stop valve. I don't know the full details of the 12 valve internals, but I understand it's just a manual valve.
(x
) 13 MR. EBERSOLE: This is the one that's 14 air-operated?
15 MR. MARTIN: No. This one is basically a very 16 large hand wheel with the ability --
17 MR. EBERSOLE: You have to take a wrench up to 18 it too big for a man to use and have a drive motor to do it; 19 is that correct?
20 MR. MARTIN: It doesn't have a drive motor on it.
21 You use an air-operated wrench, but a man could do it if 22 you put a couple of men on it. It's much easier, though, 23 to put an air wrer.ch on it.
24 MR. REED: I have to ask this question at this 25 time: Was California a power-boiler-code state at the time L'O J
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-(M6 l
l
25726.0 19 BRT O 1 this plant was built?
2 MR. MARTIN: I can't answer that. Ken, can you 3 answer that?
4 MR. BASKIN: Yes, it was.
5 MR. MARTIN: That's Ken Baskins, the Licensee.
6 MR. EBERSOLE: How does a man go up to the valve 7 if the hypothetical break is at the valve proper?
8 MR. MARTIN: Very difficult. You'll find in 9 this particular event when it lets go behind him he 10 evacuates the area very quickly.
11 MR. EBERSOLE: Well, I guess we'd have some very 12 interesting root observations to make about this design
() 13 before we get done.
14 (Slide.)
15 MR. MARTIN: This is the electrical system and 16 it will play a role in it. There are two diesel generators 17 feeding their safety-related buses. But you'll notice that 18 there are other things on here: some station service 19 transformers, some lighting transformers; normally in 20 normal operation the startup transformer they call the 21 auxiliary transformer C feeds the two safety-related buses.
22 And it gets its power from the switchyard.
23 The diesels sit here, they are not operating.
24 The generator feeds out through a motor-operated 25 disconnect to a common bus here which then feeds out to the ACE-FEDERAL REPORTERS, INC.
202-347-3X0 Nationwide Cmerage 800-33MM6
1 l
i 25726.0 20 BRT !
I switchyard and also feeds auxiliary power back to the two 2 nonsafety-related buses, which have the reactor coolant 3 pumps, circ water pumps and things like that on it.
4 MR. EBERSOLE: So the safety-related power 5 source is in fact from the switchyard, not necessarily from 6 the output of the turbine generator?
7 f1R . MARTIN: That's correct. And the switchyard 8 is considered the preferred source, even over the diesels.
9 That's the training that is given the operators.
10 Uhat we'll find is although the normal lineup 11 with these powered directly off the generator and these 12 powered directly off the off-site source, in this event
() 13 they had an unusual alignment.
14 They had this bus powered from this bus because 15 it had a ground on this system here. So the off-site 16 source was carrying this safety-related bus, which happens 17 to carry one of the feedwater pumps, and all the other 18 three buses were being carried on the output of the 19 generator. Okay?
20 MR. MICHELSON: Are the diesel generators 21 capable of starting the feedwater pumps without first 22 clearing the board and keeping it clear?
23 MR. MARTIN: This is a load stripping -- but 24 they are able to run the feedwater pumps off the diesels.
25 MR. MICHELSON: That's not the answer to my O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4 646
25726.0 21 BRT b' 1 question. Can I start the feedwater pump off the diesels 2 without first stripping the board and keeping it stripped?
3 MR. MARTIN: Ken, can you answer that?
4 MR. EBERSOLE: You are talking about the 5 starting load?
6 MR. MICHELSON: Starting load off a 3600 7 horsepower diesel.
8 MR. MARTIN: My understanding is yes.
9 MR. MICHELSON: I don't want your understanding.
10 I want the answer.
11 MR. MARTIN: If you have a blackout and 12 everything is dead, the diesels go on and load everything
~^
(d 13 onto the bus and then they start the loads, so then it is 14 capable of carrying the start.
15 MR. MICHELSON: The question is the sequence in 16 which you start the loads.
17 MR. MARTIN: There's a sequencer that loads 18 these things on.
19 MR. MICHELSON: You have to start the board 20 again to get them started?
21 MR. EBERSOLE: What's the size of the diesel 22 generators?
23 MR. BASKIN: I'm Ken Baskin. The diesel 24 generators are about 5-1/2, 6 megawatts; they are rated at 25 8800 --
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 22 BRT Q l MR. EBERSOLE: Can they take an across-the-line 2 start of the main feedwater pumps?
3 MR. BASKIN: Yes.
4 MR. MICHELSON: Can they take a fully loaded 5 board? I mean all the other loads can come on and they 6 come on line and they'll still start it?
7 f1R. RAINSBERRY: Yes.
8 MR. MARTIN: We have a loss of in-plant AC power.
9 All the power in-plant is gone, all the power is in the 10 switchyard. He have an inoperable feed pump check valve 11 which leads to a rupture of the flash evaporator, this is 12 because the running field water pump overpressurizes the (v w) 13 opposite header. He then have a loss of feedwater after 14 the reactor is tripped and the turbine is tripped. He then 15 have multiple inoperable feedwater check valves, a total of 16 five, which then allow backblow from all the steam 17 generators so all steam generators are losing inventory 18 through the feedwater and condensate system back through 19 the rupture and the isolation flash evaporator.
20 The water hammer then occurs in the B feedline, 21 causing the piping, piping supports and component damage.
22 He have a damaged feedwater check valves that develops a 23 significant steam water leak. This is in the area of the 24 main steam isolation valve, that the man was shutting.
25 He have a plant that is then shut down and O
V .
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage MWA3W6M4
25726.0 23 BRT
,o i )
I cooled down safely by the operators.
2 (Slide.)
3 MR. MICHELSON: The flash evaporator, of course, 4 ruptured. The shell ruptured. How many tubes were 5 actually ruptured in the evaporator?
6 MR. MARTIN: When we completed our investigation 7 we knew of one tube that had ruptured. The tube is about 8 350 pounds. The shell is about 15 pounds.
9 MR. MICHELSON: What size is it?
10 MR. MARTIN: It's about an inch.
11 MR. MICHELSON: So that was really preventing 12 the steam generators from blowing down at a much more rapid
()
s-13 rate?
14 MR. MARTIN: That's correct, sir. But if there 15 had not been sufficient opening there and sufficient 16 resistance in the line, others would have ruptured.
17 Basically there was nothing to stop the flow there except 18 350-pound piping, and the steam generators were at 19 significantly higher pressures than that.
20 This is the charter of the team: to determine 21 what happened, identify probable causes of what happened 22 and make appropriate findings and conclusions to form a 23 basis for possible follow-up actions.
24 (Slide.)
25 MR. MARTIN: I should point out we were not
- O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336- % 46
25726.0 24 BRT 1 asked to review the design. In fact, we were specifically 2 told not to.
3 Fact-finding methodology, we conducted 4 interviews, meetings with the Licensee, we reviewed plant 5 data, which was extremely limited -- one because of the 6 loss of electric power most of the recording instruments 7 are lost. the computer that they have that would have 8 captured data, that was inoperable. The personnel logs we 9 reviewed, the quarantined equipment we utilized to 10 basically say, these are the pieces of equipment we think 11 were involved in the event. Let's not do anything with 12 them until we develop very systematic approaches to 13 investigating what went wrong with them.
( He found that our 14 initial envelope of quarantine equipment was actually too 15 small. We had to expand it slightly when we found 16 additional check valve problems. ,
17 MR. MICHELSON: On the question of computer 18 inoperability, why was it inoperable? Was there a wrong 19 switch in the wrong position or was the power disconnected?
20 MR. MARTIN: Due to earlier troubleshooting 21 activities, trying to find the ground on the safety-related 22 buses they had turned off loads, turned them back on. One 23 of the loads they turned off was this computer. When you 24 turned it back on the disk drives come back up to speed but 25 you need to push a reset to reestablish the software O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage MXK336-6646
25726.0 25 BRT O 1 initiation point. That was not done.
2 MR. MICHELSON: There is no requirement in the 3 procedures to do this on power interruption?
4 MR. MARTIN: Not on the procedures they were 5 following.
6 MR. MICHELSON: If they had a power interruption 7 during another kind of accident then I guess you are 8 assured the computer won't track the accident; is that 9 correct?
10 MR. MARTIN: It will track the accident because 11 the -- if you don't have this complete loss of in-plant 12 power, then this computer stays on the line and keeps the 13 data.
14 MR. MICHELSON: Are you saying there's an 15 automatic power transfer to an alternate source and if it's 16 not dead you are okay, but if it's dead too, only then do 17 you have to reset the computer?
18 MR. MARTIN: Can you describe the actual lineup 19 of that? Matt?
20 MR. CHIRAMAL: I didn't look at the computer.
21 MR. BASKINS: I don't believe it automatically 22 transfers.
23 MR. MICHELSON: So if you lose your power supply 24 you really have lost it. This says during accidents of 25 other kinds you are also going to lose the computer needed ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage m)-336-6M6
25726.0 26 BRT p
\_ 1 to track it?
2 MR. MARTIN: Recognize these would only be fed 3 by one of the safety related buses and this is the first 4 time --
5 f1R . MICHELSON: Well, safety-related buses have 6 been known to lose their power, of course.
7 t1R . t1ARTIN: Since 1967, in this plant, that's 8 the -- this is the first time.
9 MR. MICHELSON: That's not the point.
10 MR. EBERSOLE: Let me ask this question. By and 11 large, computers are, at least I thought, always r.o be fed 12 by inverter sources driven by batteries for the explicit
() 13 purpose of riding through the transients that they need to 14 record. Is it that this design simply didn't do that?
15 MR. MARTIN: That's my understanding.
16 MR. EBERSOLE: So they go blind in the presence 17 of --
18 MR. REED: I thought what Jesse said was correct.
19 Early computers and up until the coming of SPDS, the 20 computers were 95, 98 percent reliable and not necessarily 21 supplied by vital power; is that correct?
22 MR. EBERSOLE: How do you mean? Oh, no, no.
23 May back for 25 years the computers have been 24 inverter-driven by battery sources.
25 f1R. REED: There weren't computers.
O ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-33MM6
25726.0 27 BRT 1 MR. EBERSOLE: I'm talking about transient 2 recorders and recorders for recording events like this type 3 of cascade. They always go back to batteries, don't they?
4 MR. REED: Knowing a lot about the Yankee Rowe 5 and Haddam design, I don't think you would find that kind 6 of power reliability in that vintage.
7 f1R . EBERSOLE: Does this mean automatically that 8 the records of the transient are lost?
9 MR. REED: Data logs and computers in that 10 vicinity are supposed to be only 98, 99 percent reliable.
11 If they are gone, you have indicators and other 12 instrumentation recorders to take care of it.
(} 13 MR. ETHERINGTON: 25 years isn't very much 14 different from the time this plant was being designed.
15 MR. MICHELSON: A lot of people in the recent 16 past have been upgrading their plant computers so they are 17 a little more reliable in terms of, at least, power 18 supplies.
19 MR. EBERSOLE: I'm hearing here that the 20 transient event are incapable of being supported if they 21 are associated with a power outage?
22 MR. MARTIN: In this particular case almost all 23 are lost plus this computer.
24 MR. EBERSOLE: There's no record of events, no 25 battery source?
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80-3 W6646
25726.0 28 BRT f n
f~') 1 MR. MARTIN: One of the things the Licensee is 2 doing is looking at his ability to reconstruct events from 3 data available. One of the things they are looking at is 4 this system.
5 MR. EBERSOLE: How generic is this problem, the 6 absence of recording data in the event of cascading events 7 causing a loss of power, except for the batteries?
8 MR. MARTIN: Because of the unique'ess n of this 9 design I can't answer th a t .'
10 MR. MEDFORD: This is Mark Medford, licensing 11 manager for the Licensee. I would like to interject one 12 thing. I believe that the computer is on our vital bus 4.
() 13 We have four vital buses. Three of them have backup power.
14 Unfortunately vital bus 4 at the time of the event did not.
15 MR. EBERSOLE: Okay.
16 MR. MEDFORD: One of the design changes which we 17 are going to implement is to provide a backup source of 18 power to vital bulls 4. s
\
19 MR. EBERSOLE: Okay. '
20 MR. MICHELSON: Does that'mean that the computer, 21 then, resets itself after the power -- the backup power 22 comes onto the vital bus?
23 MR. MEDFORD: I'm glad you asked that. A second 24 change we are considering, and in fact will implement -- we 25 haven't designed the change yet -- is to altet Mhe software rs -
b ACE-FEDERAL REPORTERS, INC. 4 202-347-3700 Nationwide Coverage 86336-646 J l
25726.0 29 BRT
[)
~#
1 in such a way that you don't require a resetting.
l l 2 MR. MICHELSON: Okay. l
)
3 MR. EBERSOLE: What is a " vital" definition? I 4 thought it automatically connoted the presence of a battery l 5 source and an inverter which was free of switching l 6 transients. You know, the bus was free of switching 7 transients. What is a vital bus?
l 8 MR. MEDFORD: That's not true in this case. 1 9 This is Jack Rainsberry, he's a licensing 10 supervisor for Unit 1. Can you help?
i 11 MR. EBERSOLE: By what right do you call it 12 vital when it's subject to switching transients?
[^') 13 MR. RAINSBERRY: It does have safety-related V
14 controls on it and it automatically switches to --
l 15 MR. EBERSOLE: It's called vital by virtue of 16 the circuits connected to it, not by what it supplies.
l 17 MR. RAINSBERRY: It will switch, too, with an 18 automatic source. It's not a UPS --
19 MR. EBERSOLE: Thank you.
20 (Slide.)
21 MR. MARTIN: In the sequence of events they had 22 a small saltwater leak into the main condenser causing a 23 blowdown of about 10 GPM into the steam generator. This 24 will be important. The status of steam generator blowdown 25 is not indicated in the control room. Unless you remember A
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 30 BRT 1 that it is on, you will not recognize that this is a loss 2 of inventory.
3 The unit was operating at 60 percent power. It I
4 was down at that power because, one, they had one of the t
5 circ water pumps off to take care of the main condenser and 6 they had been down even to lower levels when they were 7 trying to hunt for a ground.
l 8 The critical function monitor system, that was
- 9 the computer, was disabled. It was disabled because of l
l 10 earlier ground troubleshooting activity. The electrical
{
j 11 ground troubleshooting was in progress. They had finally l
12 isolated it to the C auxiliary transformer, their start-up
! f') 13 transformer. And the electric plant was in unusual v
14 alignment.
I 15 MR. EBERSOLE: That blowdown, that's common --
h 16 that's 100 GPN for each steam generator? l 17 MR. MARTIN: That's correct.
18 MR. EBERSOLE: What's the aux feed input?
19 MR. MARTIN: The aux feed comes on and will not 20 exceed 150 GPM to prevent water hammer, but could go up to 21 as high as, I think it's 50 or 300.
22 MR. EBERSOLE: So I have a potential leak in the 23 secondary system that equals to or maybe exceeds the aux 24 feed flow, but I don'* L /e any indication of its presence?
25 MR. MAiT ', >u don't have indication of its O)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 31 BRT (y
1 presence in the control room.
2 MR. EBERSOLE: Where do you have it?
3 MR. MARTIN: Out in the plant. Bill, can you 4 comment on where that is located?
5 MR. KENNEDY: No. I'm afraid I do not know the 6 locations.
7 MR. MARTIN: This will cause a problem to them 8 because they won't recognize it for about 45 minutes into 9 the transient.
10 MR. EBERSOLE: Surely that is going to be fixed, 11 MR. MICHELSON: If the computer is not reset how 12 do you know that from the control room?
(v ) 13 MR. MARTIN: You don't know it from the control 14 room. That's the whole point.
l 15 MR. MICHELSON: You have to go to the computer 16 itself?
17 MR. MARTIN: The computer will not indicate 18 blowdown status. The computer is located in the tech l
19 support center.
20 MR. MICHELSON: On the computer itself, it losas 21 power, power comes back on, there's no automatic reset; how 22 do you know that the computer is not now on standby?
23 MR. MARTIN: Licensee?
24 MR. MICHELSON: Do you have to walk up to the 25 computer and see a little light on? Is it indicated or O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 32 BRT x
)
1 alarmed in the control room?
2 MR. MEDFORD: There's no alarm that would 3 indicate that.
4 MR. MICHELSON: Where is the computer relative 5 to the control room?
6 MR. MEDFORD: In the tech support center 7 immediately adjacent.
8 MR. MICHELSON: So you have to walk over there 9 if you wonder at any time if your computer is on standby 10 for a transient?
11 MR. MARTIN: There's one monitor facing into the 12 control room so you can see it from the control room.
() 13 MR. MICHELSON: You can see the fact that the 14 indicating light is on indicating that it isn't reset? Or 15 is there even an indicating light?
16 MR. KENNEDY: There is a CRT that faces the 17 control room through a window. The shift technical adviser 18 commented that he knew the machine was not useful because 19 the screen was displaying a pattern which wasn't indicative 20 of the normal behavior of the system.
21 MR. MICHELSON: You are operating along under 22 normal conditions and you glance over there; is there a 23 different kind of pattern on it if the computer hasn't been 24 reset?
25 MR. KENNEDY: As I understand it, yes. You can O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-336-6M6
l 25726.0 33 BRT )
/'T
's',/ i 1 request specific patterns, specific data to be provided and j 2 it is normally, apparently it indicates that it is l l
3 operating, functioning normally. I 4 MR. MICHELSON: I'm not sure that has yet 5 answered my question. It's always nice to know that your 6 computer is not on-line when you thought it might be -- was 7 supposed to be on-line, for instance. How do I know that 8 the computer is not -- not in an operable condition?
9 MR. KENNEDY: They aia not know that. In fact, 10 when the trip of the plant occurred the computer had 11 recovered enough to attempt to go through an immediate trip 12 report. The immediate trip report produced data that was
() 13 approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> old and looked very strange. The 14 power was still up after the trip.
15 That took some time to sort out, reset the 16 system and the n it started to provide useful data.
17 MR. MICHELSON: That was the first clear 18 indication that the thing had not been reset?
19 MR. KENNEDY: Yes, sir.
20 MR. EBERSOLE: Do you happen to know whether the 21 condition of the blowdown system is commonly not advertised 22 in the main control room?
23 MR. MARTIN: I do not know.
24 MR. EBERSOLE: You don't know. Had this control 25 room been subjected to one of these human factors analyses?
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage fo)-33MM6
25726.0 34 BRT 1 MR. MARTIN: Bill?
2 MR. KENNEDY: Sir, they have not completed a 3 control room design review. That's pending and will be 4 completed.
5 MR. EBERSOLE: If they did a control room design 6 review, would they analyze informational outputs which 7 weren't even in the control room, like this one?
8 MR. MEDFORD: Yes, we would. And will.
9 MR. EBERSOLE: Thank you.
10 MR. MEDFORD: Let me add, in this specific case 11 another of the design modifications we are incorporating is 12 to both provide status indication of the blowdown in
('] 13 control room and to provide automatic isolation of blowdown
%./
14 on initiation of aux feed.
15 MR. EBERSOLE: Thank you, thank you. Go ahead.
16 (Slide.)
17 MR. MARTIN: At time zero, because of a fault 18 that occurred on the line feeding the 1C bus, at this point 19 that line is energized because the transformer is energized 20 feeding the 2C bus, but the braker is open over in IC. But 21 that line is still energized. It fails in two places, 22 causing a differential trip which takes out this auxiliary 23 transformer which takes out the 2C bus and takes out the 24 east feedwater pump.
25 The NS phone rings immediately. This is going (j
s ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 35 BRT O
V 1 to cause problems. The shift supervisor will answer this, 2 wondering how the NRC is so perceptive, they just had a 3 scram and the NRC knows about it. This was probably caused 4 by the transient. We have not been able to duplicate this; 5 the feedwater pump check valve fails, the flash evaporator 6 tube ruptures -- this is the result of the one feedwater 7 pump that is running overpressurizing the other side, since 8 this discharge is up around 1100 pounds and the tubes on 9 the flash evaporator are about 350 pounds, we basically 10 have overpressurized and we rupture here.
11 Diesel generator number 2 starts but by design 12 does not load.
f~)
(> 13 MR. EBERSOLE: Let me ask a question at this 14 point. I understand this is the first failure, electrical 15 failure that occurred like this; correct?
16 MR. MARTIN: Yes, sir.
17 MR. EBERSOLE: Had there been other events in 18 which one of the C buses, the preferred buses were lost and 19 the system did work and the check valves did function 20 according to their design? Or have we been protected from 21 that by the simultaneous -- rather the fact that we had a 22 single alternate power source?
23 MR. MARTIN: In fact, there's even tests that 24 the Licensee runs to test these check valves where they run 25 one pump with the other one shut down to test the check O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 804336-6M6
25726.0 36 BRT
- 1 valve.
2 The last check for this check valve was in the 3 November time frame, if I remember right; is that true, 4 Wayne?
5 MR. LANNING: That's right.
6 MR. MARTIN: There's a test part of the IST 7 program which could be conducted as frequently as once a 8 year.
9 MR. EBERSOLE: In this case the aux feed pump --
10 rather the main feed pumps are pressurized and drive 11 towards the other side?
12 MR. HARTIN: One is operating. The other is
,P) 13 shut down.
J 14 MR. EBERSOLE: What are the transient times for 15 l that test? Do they trip one pump and quickly -- they just 16 trip one pump? That's all?
17 MR. MARTIN: They start one pump, verify they 18 have flow to the steam generator so they know the pump is 19 working, then they open the other pump's discharge valve 20 and verify that there's no increase in pressure on the idle 21 pump.
22 MR. EBERSOLE: Okay. Thank you.
23 MR. MICHELSON: The flash evaporator has a 24 relief valve on it?
25 MR. MARTIN: 15 pounds. That is on the steam I)
(_/ l I
l i
l l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 37 BRT (D
\/
I side.
2 MR. MICHELSON: I would have thought that the 3 flash evaporator would have been -- the safety would have 4 been sized at least for one tube rupture; just by code, you 5 plan for an accident by the code and decide how big a 6 relief valve you need. Wasn't the safety sized for a 7 rupture in the tube in the evaporator?
8 f1R . MEDFORD: I'll speculate it was structured 9 for a tube with a 360 psi in the tube.
10 MR. F1ICHE LSON : That's probably the explanation.
11 It was sized too small because although it's sized for one 12 tube it wasn't sized for that much pressure inside the tube.
13 That's probably what happened. Go ahead.
[( )
14 MR. fiARTIN : Okay. Within 19 seconds the 15 operators recognized that they have a condition which their 16 procedures require to trip, so the operators trip the 17 reactor and they trip the unit generator which then results 18 in a total loss of in-plant AC power. Remember there's 19 power still in the yard.
20 The containment isolates automatically; the 21 diesel generator -- the other diesel generator starts but 22 again by design does not load.
23 f1R. EBERSOLE: The containment isolates blowdown; 24 is that correct?
25 f1R . MARTIN: That's correct. At this point it n)
\_
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 8%336-6M6
25726.0 38 BRT q\ /
1 has gone shut. The turbine-driven auxiliary feedwater pump 2 receives a start signal. Because it is steam driven it 3 will start but by design it takes about 3.5 minutes for 4 this to come up to speed and start delivery.
5 MR. MICHELSON: Is that a worthington turbine?
6 MR. MARTIN: It's not a cherry turbine. It's a 7 standard reaction turbine.
8 MR. MICHELSON: Who is the manufacturer of the 9 turbine for the auxiliary feedwater pumps?
10 MR. MEDFORD: I'm not sure.
11 MR. MICHELSON: Is that a full-pressure rated 12 pump? Does it use 1000 pound or does it throttle 600 13 pounds or do you know?
( ')
14 MR. RAINSBERRY: I don' t recall. I don't know.
15 MR. MICHELSON: Thank you.
16 MR. EBERSOLE: On the decay heat rejection, you 17 had a motor driven backup --
18 MR. MARTIN: But it has lost power at this point.
19 MR. EBERSOLE: But the hypothetical complete 20 loss of steam pressure, that kills the steam supply to the 21 pump -- doesn't it? -- like the Davis-Besse case if you 22 lose the main steam line anywhere?
23 MR. MARTIN: Yes, sir.
24 MR. EBERSOLE: Where is the single failure 25 criterion for aux feed?
O ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 804336-6M6
25726.0 39 BRT 1 MR. MARTIN: Licensee?
2 MR. MEDFORD: We are adding a second 3 motor-driven aux feed pump.
4 MR. EBERSOLE: I'm talking about where is it now?
5 I'm taking this as a classic example. In fact, do we have 6 the single failure criteria intrinsic?
7 MR. KERR: When this plant was built, single 8 factor was not a safety --
9 MR. MEDFORD: It's a generic problem for plants 10 of that vintage and later.
11 MR. EBERSOLE: We don't have the single failure 12 criteria in place.
/~T 13 MR. MEDFORD: Yes.
'w]
14 MR. EBERSOLE: We probably need a spread sheet 15 for the plants that don't have this.
16 MR. MICHELSON: Are you presently contemplating 17 automatic loading of the bus?
18 MR. MEDFORD: Yes, we are.
19 MR. MARTIN: Also at this point we had four 20 additional feedwater check valves that failed as a result 21 of the blowdown of all three steam water generators.
22 MR. EBERSOLE: Wait a minute. TMI II happened a 23 few years ago. I thought this brought attention to this 24 drastic void in the rationale and we did, I thought, 25 implement actions to get the single criteria applied to the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 40 BRT r~]s
\
'~
l secondary supply side; am I wrong? Or have we got all 2 sorts of things in the field out there that are just kind 3 of a morass of --
4 MR. MEDFORD: Increased attention was paid to 5 the aux feed system.
6 MR. EBERSOLE: I remember 10 of them were found 7 to be nonseismic.
8 MR. MEDFORD: That's true. And there's been 9 substantial upgrat ng, including on this plant, for example, 10 automatic initiation of the aux feed system was added to 11 this plant as a result of TMI. I don't believe it's fair 12 to say that there's an absolute NRC requirement that aux
() 13 feed systems meet single failure criteria. I believe you 14 would find other units, I think you can find other units 15 that are not committed to the addition of the second 16 motor-driven pump.
17 MR. EBERSOLE: Thank you.
18 MR. f1ARTIN: Okay. The feedwater lines begin to 19 empty as a result of the low level in the steam generators.
20 (Slide.)
21 Now you have steam qoing into the feed rings and 22 actually going back through the feedwater lines. The 23 feedwater lines are now emptying. The steam generators 24 lose inventory via the failed check valves and ruptured 25 tube in the flash evaporator.
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 8%336-M46
25726.0 41 BRT
/')
V 1 The loss-of-voltage automatic transfer scheme is 2 initiated. This is a scheme which sets up conditions for 3 repowering from the yard. It basically opens up a number 4 of circuit breakers. Then, once it verifies that the 5 voltage on the discharge of the generator is below a 6 certain level, it then opens the motor-operated disconnect 7 which then allows you to now back feed from the switchyard 8 through the main transformer to the auxiliary transformer.
9 f1R. EBERSOLE: Is this the GDC 17 --
10 MR. f1 ARTIN: This is delayed access -- they 11 receive a number of alarms indicating the safety injection 12 actuation at this point.
() 13 Fire truck arrives on the site. This is as a 14 result of a fire watch down in the 4 kV room reporting what 15 he thinks is smoke, which is really steam all over the 16 place right outside his door.
17 (slide.)
18 The operators are now attempting to restore 19 power because the automatic sequencer has failed to 20 function. It has not reclosed all the breakers it was 21 supposed to and the turbine driven aux feedwater pump is 22 now delivering approximately 130 GPM to the steam 23 generators, but it's really not going to the steam 24 generators, it's going right back out the check valves to 25 the flar'. evaporator.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-33MM6
25726.0 42 BRT 1 MR. EBERSOLE: Could you throw that other slide 2 back? I want to ask you a question about the main feed 3 pumps acting as safety injection pumps. These are 4 exclusive actions? If it's going to be doing one it can't 5 do the other?
6 MR. MARTIN: One or the other, that's correct.
7 (Slide.)
8 f1R. EBERSOLE: How does the rationale cope with 9 the fact that I have to have secondary injection as well as 10 injection to the primary load, 11 f1R. MARTIN: The auxiliary feedwater.
12 f1R . EBERSOLE: So you throw away the pump and
(~) 13 then use aux feed for secondary supply?
s_-
14 MR. MARTIN: And again, as an historical note 15 this plant has never experienced a complete loss of main 16 feedwater and aux feedwater prior to this time.
17 MR. EBERSOLE: Well, I should hope not.
18 (Slide.)
19 MR. MARTIN: Okay. The operators finally 20 restore power on the fifth attempt. There's a number of 21 problems closing the breakers. These are 250 kV breakers; 22 they are not familiar with them. They are not frequently 23 cycled, there's a number of interlocks and the operators 24 are not knowledgeable and as a result it takes five times 25 for them to succeed. Once they do succeed, though, they O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6M6
25726.0 43 BRT 1 now repowered from off-site and the motor-driven auxiliary 2 feeduater pump then starts because of a low level in the 3 steam generators. It flows to 155 GPM per steam generator, 4 but again we were still going out toward the break.
5 The atmospheric steam dumps open temporarily 6 because their controls were set such that the temperature 7 would cause them to open. The operators now following 8 their procedures, close the feedwater isolation valves.
9 These are big motor-operated isolation valves, just 10 upstream of the feed control valves. And this action 11 terminates the blowdown of the steam generators through 12 these feed lines.
I~ 13 V) At this point the feed lines begin to refill 14 because the auxiliary feedwater is still going to all the 15 steam generators.
16 MR. EBERSOLE: If the check valves which failed 17 had disintegrated in pieces, this wouldn't have been 18 j possible, would it?
19 MR. MARTIN: If the pieces had gotten under the 20 valve and somehow jammed it open, that's possible.
21 Although they would have had to go through the feed reg 22 valve, which is a caged assembly. It's probably unlikely 23 that pieces could have gotten through the feed reg valve to 24 this motor-operated val"e which is upstream of it.
25 MR. EBERSOLE: tkay.
O
(-)
ACE FEDERAL, REPORTERS, INC.
202-347-3700 Nationwide Coverage Mn33MM6
25726.0 44 BRT
' 1 MR. MARTIN: The radiation monitors -- by the l l
1 2 way, now we have power, so the containment isolation valves 3 could be reopened.
4 The radiation monitors are in alarm because they 5 had lost power and they have an alarm there. When the 6 radiation monitor alarm is reset, steam generator blowdown 7 reopens to its open condition, because that was one of the 8 close signals for the steam generator blowdown. So we now 9 have reestablished 1030 GPM blowdown from each of the steam 10 generators. We are feeding at about 155 right now, so wo 11 should be catching up but then you are steaming still 12 through the main steam isolation valves.
O 13 (Slide.)
V 14 MR. MICHELSON: And there was no containment 15 isolation signal then existing at this time?
16 MR. MARTIN: Because of the loss of power.
17 There was a very slight increase in pressure in the 18 containment. He have a lot of air-operated valves in 19 containment. They normally operate with a mini purge valve 20 open just to prevent the slow build-up of pressure in there.
21 When they had a loss of power, the valve had shut. There 22 had been a slight increase in prem- -a.
23 MR. MICHELSON: Thank you.
24 MR. MARTIN: Okay. The RCS pressure --
25 temperature and pressurizer level indicates to the O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Cmerage m)-336-u46
25726.0 45 BRT t
'~'
1 operators that they have a rapid cooldown.
2 They have one operator that's taking care of the 3 reactor plant, another taking care of the secondary plant 4 and the auxiliary feedwater; the operator taking care of 5 the power plant starts a charging pump. The second pump 6 will then auto-start.
7 He later starts the reactor coolant pump B to 8 reestablish force convection. He has had natural 9 circulation and they had verified that earlier.
10 He then directs the operator at the AFW system 11 to terminate the auxiliary feedwater because he thinks that 12 this must be the source of the excess cooldown of the
[)
\~/
13 primary system.
14 Yes, sir?
15 MR. EBERSOLE: You say reactor coolant pump 16 started. I don't remember when it stopped. Because it 17 gets its power supply out of the yard?
18 MR. MARTIN: Once you brought power out of the 19 main transformer, that has to come to the auxiliary buses 20 and then feed the safety-related buses, because that's the 21 way they are connected. Can we have that drawing again 22 real quick?
23 MR. EBERSOLE: I thought the main coolant pumps 24 along with the other pumps were on a separate bus --
25 MR. MARTIN: Circ water pump also.
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-3364M6
25726.0 46 BRT 1 MR. EBERSOLE: And those had power available all 2 through this thing?
3 MR. MARTIN: No, they did not. The -- basically 4 when this was lost these went dead, these three went dead 5 when this was tripped.
6 MR. EBERSOLE: Okay, the connection down there 7 then was open? Prom the unit transfctmer? The bus lA and 8 IB?
9 MR. MARTIN: Talking about these right here?
10 MR. EBERSOLE: They did not remain hot?
11 MR. MARTIN: When the generator was tripped as 12 part of the reactor trip procedure, this then deenergized
('}
(-
13 this portion because there's also breakers out here that 14 open up. You do not want to backfeed power into a 15 generator that's --
16 MR. EBERSOLE: Until you open the disconnect.
17 MR. MARTIN: That's right. So they open. Then, 18 after a low voltage is sensed on here, indicating that you --
19 MR. EBERSOLE: Okay.
20 MR. MARTIN: Then you open this.
21 MR. EBERSOLE: I got you. So the pump stopped 22 at that point. They were not deliberately stopped?
23 MR. MARTIN: No. They stopped as soon as you 24 tripped this generator right here.
25 MR. EBERSOLE: When you reenergized did they O
ACE-FEDERAL REPORTERS, INC.
202-347 37(X) Nationwide Cmerage 801)-33MM6
25726.0 47 BRT O 1 automatically come back on?
2 MR. MARTIN: No. They had to be turned back on.
3 So these are powered from the now off-site power. This is 4 powered from off-site power. And this is powered from 5 off-site power coming from the backfeed system.
6 MR. EBERSOLE: Why should they have come back 7 and and compounded -- and there's an argument whether they 8 did -- compounded the overchilling event?
9 MR. MARTIN: The operators wanted to get back to 10 a forced circulation mode.
11 MR. EBERSOLE: Doesn't that enhance overcooling?
12 MR. MARTIN: They needed a charging pump to get 13 pressure control because that gives them the spray in the f( )
14 pressurizer. They chose the B pump because that gives them 15 the most reliable and the most forceful spray. They knew 16 that with the starting of the charging pumps they were 17 going to raise the level in the pressurizer and they wanted 18 to have the ability to control pressure so they needed one 19 pump. They wanted one pump on the line.
20 MR. MICHELSON: Are buses lA and IB class lE?
21 MR. MARTIN: No, sir. They are not 22 safety-related.
23 MR. MICHELSON: Thank you.
24 (Slide.)
25 MR. MARTIN: At this point he directs the O
ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-(M6
25726.0 48 BRT
' 1 feedwater. The supervisor watching this hears this and 2 directs they be reestablished. The auxiliary feedwater 3 goes to zero and is brought back to about 40 GPM, it's 4 really 25 GPM indicated. That's what he shot for. The 5 5 GPM per steam generator is significant in that a steam 6 generator, regardless of level indication, that is 7 receiving at least 25 GPM, is not considered a dry steam 8 generator.
9 So, he got back to what he considered was the 10 minimum flow allowable to the steam generators.
11 MR. EBERSOLE: What sort of secondary pressure 12 do we have now? Is it relieving?
cN 13 MR. MARTIN: They shifted the controls on the V
14 steam dumps, the steam dumps have shut. At this point wo 15 should be holding pressure but it's causing cooldown. I 16 don't know what the pressure was at this point. Do you 17 have the number? Wayne do you know?
18 MR. LANNING: Certainly below load steam 19 pressure because of the flow down through the check valve.
20 MR. MARTIN: One of our problems is most of the 21 recorders are dead at this point. The timing of where the 22 chart is, it has been dead for a period of time and we 23 couldn't time them very well. As a result, much of the 24 trend data is unavailable to the operators. It doesn't 25 mean much to them.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage Nn336-(M6
I l
25726.0 49 i BRT i
(~'\ '
1 Okay. So flow is reestablished at 40 GPM, but 2 remember we have a 100 GPM blowdown going, so as expected, 3 the steam generator levels are going to be dropping fairly 4 quickly and we also have some steam coming from the steam 5 generator. As a result of this the control room directs 6 the nuclear plant operators to go out into the field to 7 shut these main steam isolation valves because they think 8 they have excessive cooldown as a result of too much steam 9 coming out of the steam generators.
10 At this point an unusual event is declared, 11 although NRC is never notified of that fact.
12 MR. MICHELSON: Before you leave this point you
(} 13 say that the operators believe that if there's 25 GPM on 14 the steam generator, then it's not a dry one. Does that 15 belief appear in procedures?
16 MR. MARTIN: Yes, sir, it does. It gives three 17 criteria and that's one of them.
18 MR. MICHELSON: When the procedures were 19 formulated, does it caution that this is only true if 20 there's no blowdown from that same steam generator?
21 MR. MARTIN: I don' t remember that from the 22 procedures.
23 MR. MICHELSON: You just fail to recognize that 24 you might have the normal blowdown of 100 GMM and only 25 25 going in and this is a flaw in the procedures.
O ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage *)n-33MM6
25726.0 50 BRT O 1 MR. MARTIN: This will be caught later when they 2 are going through the safety trees.
3 (Slide.)
4 Now for a period we have had the feedwater lines 5 refilling because of the auxiliary feedwater. The blowdown 6 is still going on on the steam generator.
7 At some point here, what, about nine minutes 8 after the isolation valve had been shut, a water hammer 9 occurs. The water hammer causes the -- a small, four-inch 10 check valve bonnet to be lifted stretching the bonnet bolts 11 by about a half an inch, blowing out the gasket and 12 resulting in a significant steam water leak right behind
() 13 the guy shutting the main steam isolation valves. He hears 14 it and evacuates the area promptly. His normal route would 15 be in stairs that would go above this leak. Ile concludes 16 not to go that route but takes a longer route. He'll 17 arrive down here some three minutes later, in the control 18 room, described as " dripping wet," and reports the steam 19 line leak.
20 MR. EBERSOLE: When he was closing this valve 21 with his pneumatic operator, where was the steam going that 22 he was supposedly going to shut off?
23 MR. MARTIN: To traps and things like that that 24 had not been isolated.
25 MR. EBERSOLE: Minor leaks that were routine?
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationmide Cmerage 800-3%ua6
25726.0 51 BRT 1 MR. MARTIN: That's correct, sir.
2 MR. EBERSOLE: When he started to close it, 3 coincidental with that he had this valve failure that you 4 are talking about?
5 MR. MARTIN: In fact, he believes it to be a 6 steam line break and reports it as such.
7 MR. EBERSOLE: Was that incidental to his 8 closing this process down?
9 MR. MARTIN: It occurred at the same time, but 4
10 does not seem to be related at all but just happens to be 11 in the same area.
12 MR. EBERSOLE: We seem to have just a generation
() 13 of check valve failures in the same time span.
14 MR. MARTIN: But this particular check valve 15 failed as a result of a water hammer and basically the 16 pressure pulse comes down and then is redirected up toward 17 the bonnet and the calculated pressure, I believe, was 18 18,000 pounds or something like that? What was it?
19 FROM THE FLOOR: On the order of 5- to 6000 psi.
20 MR. ETHERINGTON: That was calculated from what?
21 MR. MARTIN: Basically the stretch of the bolts.
22 MR. EBERSOLE: These pipes and tracks and so 23 forth that were leaking, that's a routine state of 24 equipment?
25 MR. MARTIN: That's correct.
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 33MM6
25726.0 52 BRT 1 MR. EBERSOLE: Every time you shut down you have 2 these steam leaks occurring out of the commonly connected 3 boilers.
4 f1R . MARTIN: And their normal procedures would 5 send somebody out there to shut these valves. This is the 6 way they reduce the cooldown and put it under control.
7 MR. EBERSOLE: What do they do, close off each 8 such leaking valve?
9 flR. MARTIN: They shut the big main valves.
10 Then they only have to worry about leakage through the 11 steam dumps.
12 MR. EBERSOLE: Okay. All right.
() 13 MR. MARTIN: At about this time they receive a 14 thrust-bearing high-temperature alarm on the B reactor 15 coolpnt pump. The temperature has been plodding along and 16 suddenly just goes off scale. They believe it's probably a 17 faulty indicator, because it didn't trend up but, being 18 prudent, they decide to go ahead and start the A and C 19 pumps and shut down the B, although that will take them 20 some time to do.
21 They had some concerns about containment cooling, 22 that slight increase in pressure, and there has been no 23 containment fan coolers in operation.
24 There's a lash-up that they have to work out to 25 reestablish that cooling, so it takes them a while to get O
ACE. FEDERAL. REPORTERS, INC.
202 347 3700 Nationwide Coverage sk336-6646
25726.0 53 BRT O 1 the containment cooling reestablished.
2 The operators at this point see that the diesels 3 have been operating for some time without any load. They 4 go ahead and secure the diesel generators. The fire 5 protection foam system in the area of the lube oil 6 reservoirs had actuated early on because of the steam that 7 was dumping up there from the seals on the main feedwater 8 pumps. At this point they see no fire, no necessity for 9 this foam system to be operated, so they secure the foam 10 system.
11 fir. EBERSOL~a Did the operator succeed in 12 closing this valve in his haste to get away?
(} 13 MR. f tARTIN : No. lie just got started. That's 14 all.
15 MR. EBERSOLE: The overcooling event persisted?
16 MR. MARTIN: That was a large event of the steam 17 generators blowing down to the feedline and what the --
18 again, they had recorder problems so they couldn't see a 19 trend. They didn't know what, you know, when the cooling 20 occurred.
21 MR. EBERSOLE: Are the feed lines high in the 22 boiler, high enough for them to have accepted steam?
23 fir . fiARTIN : Definitely.
24 MR. EBERSOLE: So now they are steam lines.
25 MR. f1ARTIN: Very quickly, the water level on a O
ACE FEDERAL, REPORTERS, INC.
202 347 3700 Nationwide Coverage M n336 M46
25726.0 54 BRT 1 trip drops below those feed rings.
2 MR. EBERSOLE: They have become discharge points 3 for the steam system proper.
4 MR. MARTIN: All the steam generators are able 5 to feed the 13 feed ring, which is feeding the leak, which 6 is now not isolated.
7 flR. EBERSOLE: Now I'm losing secondary steam 8 through the feed lines.
9 f!R. MARTIN: Through all three steam generators.
10 MR. EBERSOLE: They are having an overcooling 11 event and now I'll ask the question. There's some internal 12 arguments about when they turned on the main coolant pumps.
13 Did that enhance or did it retard overcooling? And was 14 that -- is that in the instrinction to prevent overcoolinq, 15 to start or stop them? The issue being, if you start them 16 up you have very good coupling to the secondary side. On 17 the other hand you have very poor mixing or local cooling 18 ,)roblems in the vessel.
19 Is that all rationalized and on paper?
20 MR. MARTIN: Wayne, can you answer that?
21 f1R. EBERSOLE: Do you or do you not turn on the 22 main coolant pumps to diminish an overcooling event?
23 MR. LANNING: 1 think the purpose of restarting 24 the pumps was to achieve pressure control.
25 f1R . CBERSOLE: Ilu t that comes at a price, O
Acti Frior RAL Rf!PORTIIRS, INC.
202 347 1700 Nationwitte Coverage Mn)4(M4
25726.0 55 BRT O 1 possibly.
2 MR. LANNING: That could come at a prico, 3 possibly. But in this caso they were cooling with the B --
4 with the A and C steam generators, reduced steam water flow 5 to those steam generators, so I don't think in this caso it 6 exacerbated the overcooling.
7 MR. EBERSOLE: Okay.
8 f1R . MARTIN: Maybe the Licensoo can answer that?
9 Is that one of your considerations?
10 f1R. MEDFORD: I concur with Wayno's assessment.
11 Wo basically got control of the overcooling by two sets of 12 actions, one was isolating blowdown and the other was
() 13 isolating field lines. Once that had been done we had 14 pretty good control of the overcooling.
15 MR. MARTIN: Bill?
16 MR. KENNEDY: I would like to comment on a term 17 they were using. They did not have an overcooling, meaning 18 greater than 100 degroos por hour. They had greater than 19 their normal shutdown cooling rate which is in the 20 neighborhood of 50 degrees por hour.
21 They had a controllable cooldown, moaning it was 22 loss than 100 degroes por hour and it wasn't that they 23 could adjust it very much, but it wasn't an overcooldown 24 ovent, rapid cooling nituation.
25 MR. EllC RSOLE : Okay. Thank you.
O Acti FliolinAI. RI!PonTiins, INC.
202 34717(x) Nationside Cmerage No-11MM6
25726.0 56 BRT
-s i ?
T 1 (Slide.)
\
2 MR. MARTIN: At this point they start the A pump, 3 then they start the C pump. This is,in preparation to shut 4 down the B pump. Steam generator blowdown still on, the
\ l 5 generator levels are all off scale low -- they are all 6 steaming to the break in the feedwa d system. Reactor 7 coolant B is stopped. The AFW flow to A and C generators 8 is taken up to 70 GPM, not because they know there's 9 soraething wrong with the B steam generator but just because 10 thar.'s where the reactor coolant pumps are operating and 11 their desire is to establish a fairly rapid cooldown sa 12 they can get to the point where they can isolate the steam IT 13 leak. They still think it's a steam line leak. But.they
(.)
14 do not want to go above 100 GPM.
15 MR. EBERSOLE: I'll ask you this, when they i
16 drive these generators to very low levels, virtually empty 17 you tell me, they become B and W super heater designs, s
18 don't they? ,They get very high-temperature steam? !!ow hot 19 do they get and are they prepared for it?
t 20 f1R. MARTIN: I don't have any infot nation on 21 Ken, can you answer that; t h <7 *. . s 22 f1R. EBERSOLE: Are you rigged for that sort of 23 super heat?
24 ,
MR. MEDFORD: I don' t have an answer either.
25 ,
MR. MARTIN: At this point the STA has gone O
ACE-FEDERAL REPORTERS, INC.
202 347 37(O Nationwide Coverage 800-33MM6 ,
25726.0 57 BRT (D
- '/
1 through the safety trees and asks a question: Is steam 2 generator blowdown still going on? Lights go on and they 3 recognize they forgot to deal with it, so they send a man 4 to the monitor panel, he cranks the set-point on the 5 radiation monitor down to a low level. That looks like a 6 high radiation level on that line. It then isolates 7 automatically. They then send an operator out to terminate l 8 the individual steam generator blowdowns by shutting the l 9 micro valves.
10 Steam generator A and C levels then start to 11 come back on-scale and they attempt to identify where the 12 steam leak is coming from.
/' 13
} (Slide.)
14 MR. EBERSOLE: I'm a little concerned about, the 15 operator apparently has to go out to isolate containment, 16 which is in essence what he did when he closed the blowdown
(
l 17 valves.
l 18 MR. MARTIN: The way he did that was he used the 19 automatic feature of the radiation monitor. That monitor 20 is in the control room.
21 MR. MICHELSON: That's a funny way to isolate 22 containment, isn't it?
23 MR. MARTIN: He wanted only to isolate this one 24 line.
25 MR. MICHELSON: Why didn't he want to isolate O
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage *4336-tM6
25726.0 58 BRT s
\& 1 containment?
2 MR. MARTIN: He had no need to.
3 MR. MICHELSON: He had nothing oise he needed.
4 MR. MARTIN: lie did need containment cooling.
5 MR. MICHELSON: He does have the option at any 6 time of walking up and isolating containment in total, from 7 the control room?
8 MR. fiARTIN : I assume he does, yes.
9 MR. MEDFORD: That's correct.
10 MR. MICHELSON: But he just wanted to get this 11 one particular valve. Apparently it is not operable from 12 the control room.
[^')
V 13 MR. MEDFORD: In effect he has simulated a 14 containment operation.
15 MR. MICHELSON: Not all containment valves are 16 operable except unless you want to do a complete 17 containment isolation; operable from the control room, that I 18 is?
19 MR. MEDFORD: That's correct.
20 MR. MARTIN: At this point the operators are 21 unable to start the cire water pumps. That's important 22 because they provide the turbine plant cooling water, which 23 provides cooling water for the containment. The saltwater 24 cooling system, all these are intertied and capable of 25 multi-modes of operation but they are going to need -- if O
ACE-FEDERAI REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-66:6
25726.0 59 BRT
(~)
'~'
1 they can't have the circ water pumps, they are going to 2 have to use the screen wash pumps and the saltwater pumps 3 to provide the cooling for the RHR system and for the 4 turbine plant cooling water system so that they can 5 maintain the containment cool, so there's some lash-ups 6 that they go through here which are designed -- the lines 7 are there, the valves are there -- but it's an abnormal 8 operation for them to go into this particular mode to 9 establish the heat sink for all these auxiliary systems.
10 MR. MICHELSON: Just in the way of a more 11 generic question: If this had been a loss of coolant 12 accident in which RHR was needed, has some study been done
(} 13 for all loss of coolant accidents? Can you still get 14 outside of containment and run around and do these lash-ups 15 necessary to make RHR work?
16 MR. MEDFORD: RHR is not one of the systems 17 required.
18 MR. MICHELSON: You don't need it for l
19 post-accident at all?
l 20 MR. RAINSBERRY: We have a recirculation system 21 that is separate from the RHR system.
22 MR. MICHELSON: So, it is a safety-related 23 system, though? Isn't it?
l 24 MR. MEDFORD: It is the normal system used to 25 get to cold shutdown. It is not used for --
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
b 46 I- 25726.0 60 BRT
+
(~)r
(_ 1 MR. MARTIN: At this point, because of the 4 2 temperature they entered mode 4. They try to establish RHR. f 3 They think the interlock for low pressure has been l 4 satisfied. They try to operate the valves that open the 5 RHR inlets to the RHR system. They don't respond. They 6 assume that there's a problem with the interlocks. They go
.7 back into the panels and actually override the relays, 8 causing the establishment of conditions necessary to open i
9 RHR.
10 It turns out that the set points on the 11 interlock were not what the operators assumed they were and j 12 there was a need to change the procedures there and they
() 13 are being modified.
14 MR. EBERSOLE: Let me ask you about the 15 standardization of this mode system that you refer to. You 16 throw up there, you know, mode 4, or whatever.
17 Uhen you talk about mode for one plant is it
- 18 standard for mode for another plant?
19 MR. MARTIN: Pretty much standard. That part of 20 the tech specs is pretty well adopted.
21 MR. EBERSOLE: So I don't need to worry about 22 this being a specific mode for San Onofre?
i 23 MR. MICHELSON: This isn't a particular tech 24 opec --
25 MR. MARTIN: Yes, sir?
! C:) i i
i i
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
25726.0 61 BRT (3
V 1 MR. REED: A significant happening and operator 2 discovery caused the blowdown taking place. I believe that 3 was discovered, as I saw in the literature, by an STA; is 4 that correct?
5 MR. MARTI#: It was discovered by the control 6 room staff as a result of the STA going through his safety 7 trees and asking questions. He has no indication in the 8 control room of the status, so his question was -- he got 9 down there: Is blowdown isolated? Is blowdown isolated?
10 They recognize, then, wait, we reestablished the radiation 11 monitors that would have caused them to open. The light 12 came on. People recognized what they needed to do.
() 13 MR. REED: I like to bring that up because STAS 14 are around -- it's a divided position with respect to STAS 15 and whether they should be as they should be and I think 16 this proves that one form of STA is useful.
17 MR. MARTIN: The team's assessment in this case 18 was the control room staff worked very well together. The 19 STA's role of going through the trees and reminding the 20 shift supervisor what was important and what things needed 21 to be considered worked very well. They worked as a very 22 good team in this case.
23 MR. MARTIN: Okay. About five hours later the 24 unusual event has been terminated. Six hours later the 25 feedwater leak has been isolated and finally they entered O
I I
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804 336-6646
25726.0 62 BRT
\'_) I the mode 5.
2 (Slide.)
3 The next evening, that night, shortly after 4 midnight they make a containment entry and they find the 5 water hammer damage on the B feedwater lines.
6 (Slide.)
7 MR. REED: That water hammer, it would be hard 8 to say when it occurred. If it occurred from refilling of 9 the steam generators, I assume these steam generators do 10 not have U tube injections on the feed ring?
11 MR. MARTIN: They are normal feed rings with 12 bottom holes.
() 13 MR. REED: So it would be hard to say when that 14 damage occurred, whether it occurred on refilling or when 15 it occurred, early on or when.
16 MR. MARTIN: But they heard the water hammer.
17 MR. REED: They heard the water hammer certainly 17 out in the steam plant.
19 MR. MARTIN: That's correct.
20 MR. REED: There could have been more than one 21 water hammer?
22 MR. MARTIN: We don't believe so.
23 MR. REED: I see a problem here in the feed line 24 check valve arrangement. I just don't think you should 25 have a plant with a normal feed ring that doesn't have a D
b ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 8fn336-6646
25726.0 63 BRT 1 check valve right close to the steam generator --
2 MR. CATTON: That works.
3 MR. REED: -- that works. That's what I mean.
4 ,
MR. MARTIN: Okay. This particular viewgraph is 5 rather negative. I think there are some positive points I 6 need to make here.
7 The operators discovered very quickly that they 8 needed to scram the plant and within 19 seconds they 9 scrammed the plant. They worked very well as a team in the 10 control room. The operators did place higher priorities on 11 protecting primary system than the secondary system, and 12 the plant was safely shut down.
13 Given those points, though, there were some
(}
14 operator errors. They failed to follow the appropriate 15 procedure when troubleshooting electrical grounds.
16 It appeared to us that they avoided, wherever 17 they could, entry into an action statement when there were 18 still other things that they could do.
19 If they had followed the procedures verbatim 20 they would have entered into an action statement, found 21 that the problems were not where they were looking, and 22 come right back out of the action statement. There would 23 have been no liability in doing so. But they seemed -- if 24 there was anything else they could do before entering that 25 action statement, they would try to do it.
,~
O ACE-FEDERAL REPORT Ek3, INC.
202-347-3700 Nationside Coverage 800-336-6M6
25726.0 64 BRT U 1 There was some difficulty in reestablishing 2 plant power, the five times trying to shut the breakers to 3 provide off-site power into the plant.
4 There was the problem of --
5 MR. MICHELSON: Excuse me, on that particular 6 case was that a problem of -- do you think -- of the 7 inadequacy of the instruction or the inadequacy of the 8 training or the operator just forgot what he had been told 9 and trained to do?
10 MR. MARTIN: A combination of all of them. Had 11 they followed their procedure and broken out the procedure 12 and followed the procedure they would have probably gotten I~T 13 it right.
\-]
14 The first time they did it he forgot to use a 15 sync bypass switch. That's necessary because one side of 16 the line is dead. The other is not.
17 MR. MICHELSON: You are saying the procedure is 18 fully adequate?
19 MR. MARTIN: If followed, it would have been 20 successful.
21 MR. MICHELSON: Had he ever been trained on that 22 procedure, ever gone through it on either simulation or 23 actual performance?
24 MR. MARTIN: Bill, can you comment on that?
25 MR. KENNEDY: I can't say directly he has. I
/N i (m) '
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6 j
25726.0 65 BRT O 1 say they decided intentionally not to follow the procedure
' i 2 because they knew the plant was in an unusual electric l 3 plant lineup and that they thought the procedure would not 4 address that successfully. So they decided to bring power
- 5 back in one bus at a time rather than lash all the buses 6 together, then bring power in.
7 flR. MICHELSON: The procedure was only for i
- 8 lashing together? A lashing together process?
9 MR. KENNEDY: That's correct.
10 MR. MICHELSON: This was an unusual procedure 11 and they thought the procedure wouldn't pertain?
12 MR. KENNEDY: Right. What they missed was a
() 13 piece of information that I would call emb- ded in the 14 procedure, having to do with the sequencing of some steps, l 15 which they missed, which caused the first three attempts to 16 be unsuccessful.
) 17 MR. MICHELSON: I think what you told me is they l 18 didn't really have a procedure for precisely what they j 19 wanted to do?
20 MR. KENNEDY: Yes.
21 t1R. KERR: Were there any serious consequences 22 of.their failing to establish in-plant power on the first 23 try?
l
, 24 MR. MARTIN: It basically delayed the time of i
25 the auxiliary feedwater. It's kind of an aside. It's
[ ()
ACE FEDERAL REPORTERS, INC. "
202-347-3700 Nationwide Coverage 800-336-6646
-.~. ...., ,,. , ~ . . , . - . . - , . , . . . . . . , _ . _ . . . . . . . . _ . , - - _ , - , - , . _ . _ , - - . - - - . , , . - - . . - . - - .
25726.0 66 BRT ,
q 1 another complicating factor in this event. That's all it 2 is. It delayed the time it took them to get power back.
3 MR. KERR: By about how much?
4 MR. MARTIN: By about 30 seconds.
5 MR. CATTON: Did the lack of the computer affect 6 it?
7 MR. KENNEDY: I would not say so.
8 MR. CATTON: So the information they didn't have 9 they didn't need?
10 MR. KENNEDY: It may have helped explain the 11 transient early on in the event, to see if things behaved 12 as expected or not expected.
(} 13 MR. MARTIN: And the other one was their failure to -- well, the inadvertent reestablishment of steam 14 15 generator blowdown was important and their failure to reset 16 the Fox III computer, which was the critical parameters 17 computer.
18 The STA performance was good. There were some 19 problems with the emergency coordinator and his statements 20 to the NRC through his misunderstanding of what had 21 occurred didn't -- certainly caused problems in 22 communications between NRC and the Licensee, which is 23 really the last point there.
24 MR. REED: Who was the emergency coordinator?
25 What was he, a member of the guard force?
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-3346M6
25726.0 67 BRT n
'^] l MR. MARTIN: No, he was a member of the plant 2 management. He was a licensed SRO. He had come on-sito 3 early that morning to help in the troubleshooting 4 activities for the ground. He had left the control room 5 after he was confident they knew approximately where it was.
6 He had gone to his trailer and he heard the noise of the 7 trip. He returnvd to the control room within minutes and 8 because he wasn't there early on didn't know the full 9 sequence of how they got to where they were. He released 10 the shift supervisor of his control of emergency 11 coordinator and tried to carry this burden of communication 12 with everybody in the outside world.
() 13 MR. REED: It sounds like he was a quality 14 person, training, background, SRO, but he comes in to the 15 scene. That would beef the case for any emergency 16 coordinator. So they don't get up to speed -- in fact, you 17 can be on the scene right through the whole thing and not 18 know what's going on.
19 MR. MARTIN: In his particular case, he went the 20 extra step of hypothesizing and communicating as fact 21 certain things that didn't occur.
22 MR. REED: Isn't that what the NRC red phone guy 23 wants him to do? Speed up --
24 MR. MARTIN: That's where you have to refuse to 25 answer that kind of question, say, I don't know.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 68 BRT
(~\
U 1 MR. REED: Then they'll say: Well, why don't 2 you know? Just like a news reporter. You have to know.
3 Give me the answer.
4 MR. MARTIN: There is some training necessary 5 for NRC in this also.
6 MR. REED: I would say so.
7 MR. KENNEDY: I would add, sir, he was not --
8 the management person who took this role would not normally 9 take that role. He was acting in another person's position 10 because of an illness, and therefore was empowered to take 11 this role and took it. When he came in and did not know 12 what was the current status, he reported that fact to the
()
sJ 13 NRC when they wanted to know what the condition was. He 14 said, give me some time and I'll go find out and get back 15 to you and they continued to ask questions and he answered 16 them to his understanding of how he expected it would have 17 occurred.
18 MR. KERR: Are you criticizing him for that?
19 MR. KENNEDY: He could have continued to stress 20 that he did not know.
21 MR. KERR: I'm sorry, that wasn't the question I 22 asked. You are criticizing him for answering questions 23 which you insisted that he answer? That's what you just i
24 told me.
25 MR. MARTIN: The insistence is improper. The 1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M646
25726.0 69 BRT
'i 1 question was asked again.
2 MR. KERR: Look, he said, I don't know, I'll
.3 find out.
4 He was asked, what do you think? He told.you 5 what he thought. Now you are saying he really should have 6 known? That's what it sounds to me like.
7 MR. KENNEDY: It was not couched as a: What do 8 you think? It was asked as an explain --
9 MR. KERR: Come on. If his first answer was "I 10 don't know, let me find out," and you keep asking the same 11 question you may not put it in those words, but you are, in 12 effect, asking him to hypothesize. I mean, he said he
(). 13 doesn't know. What else can he do?
14 MR. KENNEDY: I agree. I would criticize him 15 for not continuing to stress that he could not answer the 16 questions.
17 MR. KERR: Well, I sure wouldn't.
18 MR. CATTON: Obviously you have never been 19 questioned by NRC.
20 MR. KENNEDY: Oh, yes, I have.
21 MR. KERR: Did his poor performance or his 22 performance errors contribute to the seriousness of the 23 consequences?
24 MR. KENNEDY: At the plant, no, sir. The NRC's 25 understanding, yes, sir.
O ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
25726.0 70 BRT
(~\/
\
1 MR. CATTON: I'm not sure that matters.
2 MR. ETHERINGTON: Would you say the performance 3 was good, of the STA?
4 MR. MARTIN: Yes, he quickly assumed a role 5 supportive of the shift supervisor, keeping him apprised of 6 what things were important from a safety standpoint. He 7 was reviewing what was on the safety trees to verify that 8 things had been done.
9 When he f>und something that did not -- he did 10 not know the answer, he asked questions. When he found 11 things were in an alert level or important, he notified the 12 shift supervisor: Have you considered such and such? Have 13 you looked at such and such?
/~)'s
\-
14 So, he was like a conscience. He'was reminding 15 the shift supervisor of important issues that the safety 16 trees had identified for him.
17 MR. PEED: Harold, you may not have caught the 18 fact that he was instrumental in finding where the water 19 was going, going out of the steam generators, dry out.
20 MR. ETHERINGTON: I had missed that.
21 MR. EBERSOLE: This red telephone, isn't this 22 really just an added burden on the staff and will tend to 23 compound the difficulties of the operators to recover, to 1
24 have a continuous, on-stream flow of information to some ,
1 1
25 Washington bureaucrat?
() I 1
1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 71 BRT
('~') 1 MR. MARTIN: What we found was the way it was 2 used this time didn't add to the appropriateness of the 3 response.
1 4 MR. EBERSOLE: There has to be a man dedicated l 5 to do this and he's not likely to be the most valuable one 6 who is going to save the plant.
7 MR. MARTIN: Our findings here were that both 8 NRC and the Licensee needed to reexamine this function and 9 how they were going to use it.
10 MR. EBERSOLE: Isn't it likely he's going to get 11 lousy information first on because it's going to be 12 conjecture?
(~ 13 MR. MARTIN:
V) That's one of the things that needs 14 to be avoided. Two, the types of questions. You don't ask --
15 they were leading questions that were asked by NRC and 16 those leading questions led to responses which were 17 appropriate to this plant. But because the questioners had 18 the perception of a standard Westinghouse four-loop plant, 19 they misinterpreted the response.
20 There were a number of examples in the report 21 that point out that because we didn't understand the plant 22 because the questions being asked were not symptom -- or, 23 what is the pressure in the plant; what are the l l
24 thermocouples reading? We were trying to understand the 1 25 event, making sure the plant was safe.
O
'O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
1 25726.0 72 BRT
\- 1 We feel there was a significant training burden 2 that NRC has got to enter into in training their people on 3 how to ask the right questions to be supportive of their 4 role to be sure the Licensee is doing the right things to 5 protect the plant and at the same time the Licensee has to 6 be cautioned in not speculating or trying to put together a 7 story which seems logical to them if they don't have any of 8 the facts.
9 In fact, in this particular case, what was 10 communicated looked like an ATWAS, it sounded like an ATUAS.
11 MR. REED: I would like to make a point here.
12 When I was on the receiving end of the red phone
/~ 13 installation at its outset -- and the rules of the red L)'
14 phone, I never liked this reporting as soon as possible or 15 right close behind an event.
16 Hell, in the shuttle incident nobody is 17 reporting; in FAA incidents nobody is '3 ally reporting, and 18 I objected to that "as soon as possible" phrase. I still 19 object to it. I think it diverts people from their safety 20 activities and I really don't think the NRC red phone 21 contributes one hell of a lot.
22 If there is a need, will be a need, in that 23 there will be anv contributory significance from the NRC 24 being told immediately or as soon as possible, fine and 25 dandy. But I think we ought to find a way to differentiate O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 800-336-y46
25726.0 73 BRT 1 so we don't divert these people from real safety activities.
2 I think what you have here is a very good 3 example that should be studied in depth for its impact, 4 whether or not the phraseology and the red phone should be 5 changed.
6 MR. MARTIN: We would not disagree with you in 7 the way it was used in this case. It did not contribute 8 positively to the response. It did distract the people.
9 In fact, the ringing of the NS phone right at time zero, 10 the shift supervisor properly made sure his plant got 11 scrammed and the people responded before he picked it up.
12 But his information even then was not firm and so, his
() 13 response at that point, he couldn't really give a good 14 story. It would have been much better if he had 15 or 20 15 minutes to digest it and had somebody else make that 16 communication.
17 MR. EBERSOLE: You really need another bullet up 18 there about the NRC. Well, you have, " communications not 19 effective."
20 MR. MARTIN: There are findings which we'll get 21 to. The finding will bring this out.
22 MR. MICHELSON: The phone wasn't supposed to be 23 used to begin with. At that point in time there was some 24 kind of glitch that caused it to ring, so I think this was 25 an exception. You wouldn't want to throw the red phone out n
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-336-6646
25726.0 74 BRT
(~)
{
/
1 just because it happened in this case to ring.
2 MR. CATTON: Are there any examples of where it 3 was helpful?
4 MR. MICHELSON: I'm sure it got all screwed up 5 in the beginning. The first thing that happens when you 6 scram your reactor, for the red phone to ring, it gets 7 everybody off on the wrong foot.
8 MR. REED: I'll have to look for a bypass.
9 MR. KERR: Let me ask. You said, I believe, the 10 NRC needed to understand what is going on in order to try 11 to do it, whether the operator was doing the right thing.
12 Suppose you decided the operator was not doing
() 13 the right thing. What would NRC do?
14 MR. MARTIN: Make recommendations. We do not 15 have authority at that point to do anything else.
16 MR. KERR: What does making recommendations 17 imply to you?
18 MR. MARTIN: Have you considered the steam 19 generator blowdown? Has it been shut?
20 MR. REED: That would be an obvious 21 recommendation from 1000 miles away?
22 MR. MARTIN: That's true.
23 ,
MR. REED: Frankly, if I was 1000 miles away and 24 I was the dog of the NRC, I wouldn't dare to make any 25 recommendations to the operating staff with respect to n
U ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
25726.0 75 BRT
,rx
\ )
1 anything concrete to do, except, maybe, evacuate the 2 population.
3 MR. MICHELSON: I thought that was what the red 4 phone was for, so the NRC would have some idea.
5 MR. MARTIN: I happen to agree with you, it's 6 better to have NRC on-site. That's why they would move 7 them to the site if the plant was in an unsafe condition.
8 How do you know it's in an unsafe condition --
9 MR. KERR: It's really better to have the NRC 10 people there where they can tell the operational crew in 11 person what they should be doing?
12 MR. MARTIN: Again, that's recommendations. No, 28 13 sir, correct.
(}
14 (Discussion off the record.)
15 MR. MARTIN: Let's go forward now. The 16 equipment problems, the power supply cable appears to be a 17 ground fault. There was two places where the cable was 18 ruptured. They are located in cable trays under a 19 feedwater line with' a flange. The flange shows evidence of 20 long-term water leakage in that area. It is that there may 21 have been some connection between this water leakage and 22 the ground fault. The analysis is ongoing. We don't know 23 any more about it.
24 The flash evaporator failure we have already 25 talked about.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
. - ~ .-.. . _ . . .- . _ - -- - _. _
I 25726.0 76 BRT 1 The' safety injection annunciator, it was the 4
.2 first spurious alarm. We found now it will always alarm on 3 loss of power whether you have a safety injection or not 4 and we deem that a design deficiency.
F 5 The safeguard load sequencing system also 1
- 6 indicated a safety injection actuation. To date we have
- 7 not been able to determine why that occurred and the ;
, 8 analysis is ongoing.
9 So you had two indications of safety injection, f
10 the operator reviewed those two, reviewed plant conditions, i
11 reviewed containment conditions, determined that a safety 4
12 injection was not warranted and verified that the safety 1
(}
13 injection loads had not started.
. 14 MR. MICHELSON: By that last statement do you 15 mean they verified that the valve realignments hadn't 16 occurred and so forth?
i 17 MR. MARTIN: And the pump hadn't started in that 18 mode.
19 MR. MICHELSON: These would have been the 20 feedwater pumps, wouldn't they?
21 MR., MARTIN: That's correct. But even later on, 22 if he thought there was a safety injection, his diesels 23 were operating. They were sitting there ready to be loaded. ,
24 All they had to do was to cause the bus to be stripped, to 25 load those buses, and then to load these things on.
4 O
j 3
/\CE-FEDERAL REPORTERS, INC.
l 202 347 3700 Nationwide Coverage 800-336-6646
- - , _ . . . . . , . _ _ _ . . , _ _ . - __ _ . . _ _ . . _ . . _ _ _ _ . . _ _ , _ . . . _ . _ . _ . . . . . _ . , _ . . ~ _ _ . - . - - - . - - . -
25726.0 77 BRT r3 t,')
1 MR. MICHELSON: The bus was already stripped, 2 wasn't it?
3 MR. MARTIN: No. Just deenergized at this point.
4 MR. MICHELSON: Didn't that strip the bus?
5 MR. CHIRAMAL: Yes. It would strip the bus.
6 MR. MARTIN: Okay. But all he had to do was 7 close in the diesels and start these, but he saw there was 8 no need to do that at this point.
9 Okay, the loss of voltage --
10 MR. MICHELSON: He was not yet reconnected?
11 MR. MART!N: No. This is in the four-minute 12 period.
() 13 MR. MICHELSON: This was the time he was in the 14 total blackout, diesels were running but he decided not to 15 even reconnect his emergency -- normal lighting?
16 MR. MARTIN: He had emergency lighting, but 17 normal lighting, you are correct. He didn't even do that 18 because their procedures and training emphasizes that 19 off-site is the preferred source.
20 MR. MICHELSON: He had off-site. He knew that, 21 I hope?
22 MR. MARTIN: And therefore they were waiting for 23 the auto transfer scheme to complete, which sets up the 24 conditions necessary to repower the plant, shutting one 25 breaker in the switchyard.
i l
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3346686
25726.0 78 BRT
~'
1 MR. MICHELSON: That is all automatic, isn't it?
2 MR. MARTIN: It takes approximately two to four 3 minutes for that scheme to complete. But then there is one 4 manual operation at the end.
5 MR. MICHELSON: Let me make sure. Automatically 6 it takes three to four minutes for the breakers to realign 7 themselves?
8 MR. MARTIN: Because what you have to do is 9 decay the voltage on the bus at the discharge of the 10 generator to the point that you can open that no load 11 disconnect.
12 MR. MICHELSON: Okay. He was waiting for his (D 13 break shift to disconnect. Thank you.
G' 14 MR. EBERSOLE: The coastdown must take two or 15 three seconds.
16 MR. MICHELSON: He said he can't reconnect them 17 until he --
18 MR. MARTIN: If you can't feed anything the 19 voltage will stay up.
20 MR. EBERSOLE: Doesn't it go to a very low level 21 in a few seconds?
22 MR. MARTIN: No, the generator is still rotating 23 at 3600 rpms. There's always some residual and the voltage 24 will stay up.
25 MR. EBERSOLE: That high?
P G
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 33&6M6 a
25726.0 79 BRT l O 1 MR. MARTIN: Yes.
2 MR. EBERSOLE: That last manual action, you are 3 telling me the diesel does not automatically connect?
4 HMR. MARTIN: Unless you have a safety injection 5 in coincidence with the loss of power.
6 MR. EBERSOLE: Why didn't it come on, then, 7 within three and four?
8 MR. MARTIN: Because there was no safety 9 injection. These were erroneous indications.
10 MR. EBERSOLE: Oh, they.weren't authentic.
11 MR. MICHELSON: People that don't have generator t 12 breakers, then, can never connect their diesels onto
(} 13 emergency buses for three or four minutes? Is that right?
14 MR. MARTIN: No, no, no. Let's get the 15 electrical schematic back on here.
16 MR. MICHELSON: This is not the only plant in 17 the country. i 18 MR. EBERSOLE: I find it hard to believe the 19 residual can be so high.
l 20 MR. MARTIN: You have a generator coasting down, 21 you have residual magnetism and no loads, no current that 4
22 gives you the IR drop, to give you the lower voltages, so 23 this voltage coasts down very slowly. You are sensing the j
- 24 voltage on this bus.
i 25 Now, this breaker is open, that breaker is open.
!O l'
i ACE-FEDERAL REPORTERS, INC.
! 202 347-3700 Nationwide Coverage 800-336-6M6
. - - - - - . , . . , . . , - - - . - - . ~ . - . - - - - - . . , . . , - . . . , - - , - . . - - , _ - - - - - - _ . - -..
25726.0 80 BRT
(~')
\- So, it's not feeding anything.
1 The diesels could be closed 2 onto this because it is open and this is open. (Indicating.)
3 So this is basically a dead bull with a diesel 4 sitting here operating, ready to be loaded. All you have 5 to do is close that breaker. Over here, closing the 6 breaker and you have reenergized the buses.
7 f1R. MICHELSON: They just don't have that 8 feature?
9 MR. MARTIN: They don't have it today.
10 MR. EBERSOLE: The MOV that senses the slowly 11 decaying voltage and doesn't clear until it's visually zero?
12 MR. MARTIN: What is the voltage?
13 MR. CHIRAMAL: 40 percent. Then it will open.
(v~]
14 MR. EBERSOLE: Then it will open.
15 MR. MARTIN: But it takes time. This is a 16 motor-operated thing, so it drives open and how long it 17 takes to drive open, we have not tested that because with --
18 we had one diesel down during the refuelin7 outage that was 19 being disassembled and worked on, we had an auxiliary 20 transformer that was out because of this current fault, so 21 this was the only reliable off-site source. He didn't want 22 to play around with this and try to find out what happened 23 until they reestablished these additional power sources, so 24 we basically said: Let's leave thdt for future analysis.
25 MR. MICHELSON: The operator, then, could have O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6M6
25726.0 81 BRT f~)/
\
1 closed the diesel if he wanted to. He just elected not to?
2 MR. MARTIN: That's correct.
3 MR. MICHELSON: Some reason why?
4 MR. MARTIN: The training and procedures 5 emphasize that off-site is the preferred source if it's 6 available?
7 MR. MICHELSON: Even if you are sitting in a 8 blackout. How black is the control room on emergency 9 lighting?
10 MR. KENNEDY: Their emergency lighting is nearly 11 the same as their normal lighting.
12 MR. MICHELSON: So they weren't really in the
() 13 dark then. Thank you.
14 MR. MARTIN: Let's go on.
15 (Slide.)
16 Ue talked about the computer, the turbine 17 rupture discs; four of the eight failed. That's expected 18 on a loss of power event.
19 Emergency notification system, the spurious 20 rings, we talked about that. So far we cannot reproduce 21 that.
22 RCP thrust bearing, there was a high temperature 23 alarm. It has now been determined it is a failed detector.
24 Then we started getting into the check valve. The studies 25 were stretched and you had the body to bonnet leak that was O
(/
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 8(Xh336-6M6
25726.0 82 BRT 1 as a result of the water hammer. l I
2 (Slide.)
3 These check valves failed at the beginning of i 1
l 4 the event. This first one, the nut was missing that 5 connects the disc to the hinge arm and the disc was 6 separated from the hinge arm and in fact laying in the 7 bottom of the valve.
8 The second one, same condition.
9 The third one, the nut was loose and stuck open.
10 As a result, when the ditch approached the seating surface 11 it gouged into it and was held in the open position.
12 This one is in the same condition.
()
13 MR. EBERSOLE: Let me ask you a question about 14 the check valve. There's a general design rationale that 15 if you put one or two or three or n devices in a series 16 which are presumably redundant with each other, you don't 17 know what you have until the last one fails, unless you 18 monitor the progress failures of the others? Are there any 19 actions to do that?
20 MR. MARTIN: They are part of the IST program.
21 All these valves are. But the failure mechanism -- one of 22 the points we are going to make, these valves will fail 23 within less than a year's time. All of them.
24 MR. EBERSOLE: You told me they had a check on 25 these in November. Was that the last valve that was O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33M646
25726.0 83 BRT
'-) I working that caused it to pass successfully?
2 MR. MARTIN: Was that the last --
3 MR. EBERSOLE: Or would you know? Suppose the 4 last check valve in the line worked and you never did know 5 what the rest of them were doing?
6 MR. MARTIN: Each check valve is tested 7 individually.
8 MR. EBERSOLE: Oh, is that right?
9 MR. MARTIN: And some of them were tested as 10 late as February of '85.
11 MR. EBERSOLE: How did they check it 12 individually? Are they artificially blocked open?
() 13 MR. MARTIN: The ones that go to the steam 14 generator, what they do is they use the head of water in 15 the steam generator, which only is about 10 feet, about 5 16 psi. They shut the upstream motor-operated valve, they 17 open a drain just downstream of that motor-operated valve, 18 let the line drain and then if the leakage goes to zero or 19 very low leakage, they say that check valve is performing 20 its safety function.
21 MR. EBERSOLE: That's not a dynamic test.
22 That's just a static test to see that it seats.
23 MR. MARTIN: Demonstrates it can perform its 24 safety function, that being, it can throttle flow. Not 25 necessarily leak tight. The C test for the IST requires
('1 V
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
25726.0 84 BRT 1 for pheck valves that they be able to throttle and 2 esser.tially provide a differential pressure.
3 MR. MICHELSON: These are isolation valves in 4 some cases, of course.
5 MR. MARTIN: They are not containment isolation
' 1 6 valves. ', '
7 MR. MICHELSON: The feedwater checks -- if they 8 aren't then which ones are they? '
9 MR. MARTIN: Appendix J, they are not required 11 0 to be containment isolation valves.
Il MR. MICHELSON: Which ones are?
> jl 2s MR. MARTIN: Feedwater does not require it. It
,5 13 is seismic. Up to those check valves it's a seismic system
(} ,
14 because of the AFW system over there.
15 MR. MICHELSON: That's fine. But upstream of 16 the check valves it's not seismic. I was just inquiring
. r 17 where are the isolation valves for the line, because that's 18 the first valve coming out of the system?
19 MR. MARTIN: The safety-related piping goes all 20 the way up to those MOVs, the matter operated valve that 21 was shut terminated this particular transient. All those
, s '
22 are safety-related valves. All of those are subject ,tc the 23 IST program. \
s,, .
24 MR. MIti!9LSON : You are avoiding my question.
25 My question is which, by identity, are the so-called s
ACE-FEDERAL REPORTERS, INC.
202 ?47-3700 Nationwide Coverage 800 336-6M6 ,
L
25726.0 85 BRT
\
I containment isolation valves?
2 MR. MARTIN: There are no containment isolation 3 valves on feedwater. This is a PliR. By appendix J it is 4 not required.
5 f1R. MICHELSON: Okay.
6 MR. REED: Back to the check valves. Just a 7 minute. Are you telling me that those check valves were 8 either tested or maintained within the last year?
9 f1R. MARTIN: Yes, sir.
10 MR. REED: And this event of failure of these 11 check valves took place in the last year? This is an old 12 plant that has been running for 20 years. Are these the n
13 same check valves that were there originally?
(v) 14 MR. MARTIN: Some of them have failed before, 15 similar failure, and have been repaired.
16 MR. REED: And now all of a sudden five have 17 failed. It will be interesting to look at the maintenance 18 history on these valves and find out whether the people 19 maintaining them were aware and were alert to a generic 20 design deficiency?
21 MR. MARTIN: I think we are going to get to that.
22 That's our major finding, by the way. But let us talk 23 about that when we arrive at it and I know the Licensee has 24 a specific presentation just about those check valves.
25 Okay. The flow control valve, because the O
V ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80fb3346646 n _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . -
25726.0 86 BRT.
~
1 upstream or downstream check valve was laying in the bottom
- 2. of its body -- the flow control valve took a shot and 3 because the whole line moved laterally -- the actuator sits 4 on a very tall yoke and is a very massive unit; as a result, t
5 the inertia and bending stress caused the yoke to break and l 6 the stem to be bent.
7 MR. MICHELSON: Was the stem bent inside the 8 valve or outside the valve?
9 MR. MARTIN: I don't remember. Licensee, can 10 you answer that? Where was the bond?
11 MR. RAINSBERRY: I don't remember either.
12 MR. MICHELSON: What type of valve is it?
() 13 MR. SERKIZ: Carl, as I recall, when I looked at 14 the components, if there was a bend it would be external to 15 the body because it goes up and then your unit is up above.
16 MR. MICHELSON: This is a hydraulic-operated 17 valve?
18 MR. MARTIN: No. This one is air-operated.
19 MR. MICHELSON: What kind of valve? Is it gate?
20 Y?
21 MR. MARTIN: It has the cage in it. It has two 22 discs. They would be globe valves. This is a flow control 23 valve. It's double disc --
24 MR. SERKIZ: A fisher valve and then it's a 25 different machine.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
25726.0 . 87 BRT
! ./"Y
~
l MR. MICHELSON: You mean it's not a plug valve?
2 MR. SERKIZ: They have the two plugs. They come 3 in place. It's the flow control valve.
4 MR. MARTIN: It looks like a dumbbell.
5 MR. MICHELSON: What position was it in for the 6 water hammer?
7 MR. SERKIZ: Full closed position, because they 8- would have, at that time, through procedures, closed the 9 motor-operated and all the other valves.
10 MR. MICHELSON: So this was thought at that time 11 to be in the fully closed position?
12 MR. SERKIZ: If you ask now what fully closed
(} 13 means in terms of actual closure -- it's fully closed.
14 MP. MICHELSON: There is no damage to that valve 15 other than bending the stem? i l
16 MR. SERKIZ: Yes. The stem was bent when it 17 came up through the packing and it would come out through I 18 the pneumatic driver.
19 MR. MICHELSON: Broke the yoke, but no damage to 20 the plug itself or gate or whatever you want to call it?
21 MR. SERKIZ: If you bear with me a second, we 22 have a picture in here. That picture is in your report, 23 Carl.
1 24 MR. MICHELSON: It is? I must have missed it. l l
25 Do you recall what figures? This must be the photograph, '
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
. _. ..2 m m .u -m 2 .,a __.m_a
'25726.0 88 BRT i.
1 then. I thought it was a diagram. You couldn't tell 2 anything with these.
3 MR. SERKIZ: If you look at figure 6.38 and 6.39.
4 On page 6.45. 6.38.
5 MR. MICHELSON: That's one of the photographs.
6' You can't tell a thing from that.
7 Oh. Yes. There's a photograph. Go ahead.
8 That takes care of my question. Thank you.
9 MR. MARTIN: B steam generator feedwater line 10 had a crack in it. It is bent, dented. We believe that is 11 .all associated with the water hammer. And licensee was 12 continuing his evaluation there.
13 The feedwater line and supports and snubbers, a
}'
14 number of them were damaged again, water hammer, and 15 licensee's evaluation is ongoing.
16 The auxiliary feedwater line supports, there was 17 some displacement of the line but we found no damage there.
18 Containment sphere, there were some small, 19 crack-like indications in a saddle that went around both 20 the feedwater line and one of the safety injection lines.
21 Since then the Licensee has done further evaluation there 22 and I'm sure will be able to comment on those. And there 23 was also a problem with the security access control system, 24 wnere it had a failure, mal-operation, and the operatorr.
25 had procedures and keys and things like that that they were
- /~T'
\_/
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
_ . - _ _ , - _ - _ . , _ _ . . - . . . _ , _ . _ - . . , _ _ _ = _ ___, _ -__ -_ .- - . _ _ _ _ _ - -_.-. -. . . -
__, . - . ~ _ - -.-- . -- --- .-------
1 25726.0 89 BRT 1 ready to respond to such an event, and there was really no 2 safety or safeguards interface problem. The operators, as 3 a result.of our previous actions and their previous actions, 4 were ready for this event.
5 (Slide.)
6 MR. MICHELSON: One of the things I didn't find 7 in the report, unless I missed it, and that is the piping 8 configuration in the immediate vicinity of the feedwater 9 check valves.
10 In other words, hou close was the control valve 11 to the check valve in terms of potential turbulence-12 generation and also in terms of the water hammer; I think
() 13 it's a very important matter to know exactly where these 14 busted valves are relative to the other obstructions in-the 15 system.
16 MR. MARTIN: They are very close to each other.
17 MR. MICHELSON: Did I miss it? Is it in the 18 report?
19 MR.. MARTIN: If it's not, we have it as backup.
20 MR. MICHELSON: I would like to receive a copy --
21 MR. SERKIZ: 6.30 in there will give you some 22 idea. But the dimensions you are asking for, Carl, would 23 have to come off a piping diagram.
24 MR. MICHELSON: Figure 6.30?
25 MR. SERKIZ: 6.30. But there's no dimensions.
O ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
4
.25726.0 90 BRT 1 If you are asking for the dimensions, I can get them to you.
i 2 MR. MICHELSON: Yes. That one didn't give mo 3 quite the help I needed, because I didn't know whether it 4 was one foot or 10 feet or 100 feet or what?
S MR. SERKIZ: In round terms the stations, if we 4
. 6 are referring to figure 6.30, would extend about 15 feet.
'7 MR. MICHELSON: There's 15 feet between the
! 8 check valve and the flow control valve?
i 9 MR. SERKIZ: No, no. All the valves in 6.30 l 10 would be upstream from where this point stops from, FWS3.42,
! .11 going back up to the motor-operated valve 20.
j 12 MR. KENNEDY: You may be able to see it by
(} 13 factoring in the identified valves in 6.32 and figure 6.33.
14 You see how close they are.
15 MR. f1ARTIN: Do you have the drawing that shows
=
16 the main steam isolation valve?
17 MR. MICHELSON: See, we didn't have photographs. -
l l 18 MR. SERKIZ: Carl, look on figure 6.33. That 19 will give you a quick indication. 6.33.
20 MR. MICHELSON: Oh, yes, that's it. See, our I 21 photographs are all black. They Xerox.
- 22 MR. SERKIZ
- You have a Xerox of a Xerox.
] 23 MR. MICHELSON: When can we get a copy of the ;
I
- 24 NUREG, since that's really what he needed.
25 MR. LANNING: We had your name, I thought, I
J
()
t ACE-FEDERAL REPORTERS, INC.
1 202-347-3700 Nationwide Coverage 800-336-6M6
_ . . _ - . _ - . _ - - . . . . . . - - - - _ . - - _ _ , _ , _ . - _ ~ . - - - , . - - . - - _ . - , . - . . . . . - _ - .
25726.0 91 BRT l
O~' 1 according to the standard distribution.
2 MR. MICHELSON: We got Xeroxes, that's all I got.
3 The photographs were totally black in the Xerox. This is 4 what I needed.
5 MR. SERKIZ: If you need the dimensions --
.6 MR. MICHELSON: No. This is great.
7 MR. SERKIZ: As I was saying, 15, 10 feet.
8 MR. SAVIO: They are on their way. They haven't 9 arrived yet.
10 MR. MICHELSON: This is what I needed. This 11 would have done it. Thank you.
12 MR. EBERSOLE: Before you go into this, let me
() 13 ask you if you can throw back 4.3, the main feed system 14 'which also shows the interfacing aux feed system? All I 15 really want -- I thought it would be worthwhile -- is to 16 look at the totality of check valve failures.
17 MR. MARTIN: Let's go back to 52?
18 MR. EBERSOLE: The main feed system which shows 19 the point of insertion of aux feed. And then ask the 20 question how many more check valves failures would have put 21 us in real duress.
22 MR. MARTIN: You lost most all of them at this 23 point.
24 MR. EBERSOLE: Let's go back to the nest of 25 checks on the right there.
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-3366M6
25726.0 92 ,
BRT 1 MR. MARTIN: This is a six-inch, a 10-inch and a 2 12-inch. This would have been sitting there shut anyway.
3 This is the valve that they use, and these two are the ones 4 they shut per procedure that isolated this.
5 MR. EBERSOLE: Hypothetically, this failure 6l might involve fragments coming down into those valves and 7 ruining them, okay?
8 MR. MARTIN: That's correct. But once --
9 MR. EBERSOLE: If I lost that whole set, lost 10 that boiler.
11 MR. MARTIN: You have lost it as a unit.
12 MR. EBERSOLE: I'm talking about as a heat sink
() -13 to get rid of decay heat. The other boilers are still 14 functioning. I'm trying to get a notion of how many valve --
15 MR. MARTIN: Recognize we had all of them 16 blowing down. You are right. If for some reason -- this 17 is sitting shut, so even if a part comes back here, I won't 18 bother this one.
19 MR. EBERSOLE: Until you open it. Well, it 20 won't even come back.
21 MR. MARTIN: These two, the thing that started 22 the refill and failed, it would be the shutting these two --
23~ this check valve had already failed though, and was blowing 24 back this way. The mechanism to cause the shattering of 25 this did not occur until after these were shut.
O l
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
25726.0 93 BRT
/~T
\'") 1 MR. EBERSOLE: Yes. That was in fact what 2 caused them to do it.
3 MR. MARTIN: Because it started the refilling 4 and led ultimately to the water hammer.
5 MR. EBERSOLE: I'm just trying to find the 6 residual competence to redecay heat via the secondaries.
7 How much residual competence did you have? Ilow many more 8 check valve failures would have put me in real trouble?
9 MR. MARTIN: You basically have failed all the 10 check valves except these little four-inch, which are in 11 shut-off lines already. All the other check valves have 12 failed.
13 MR. CATTON: All six of them?
b' T 14 MR. MARTIN: Only five. 1, 2, 3, 4, 5.
15 MR. EBERSOLE: So the whole active s9t of check 16 valves went?
17 MR. MARTIN: Yes, sir.
18 MR. EBERSOLE: I'm having real tr.suble believing 19 that this was a point in time where we found that they had 20 all failed progressively prior to this time, but rather it 21 was a point in time where we had some physical event.
22 MR. MARTIN: There's a unique thing that we 23 haven't gotten to there. The plant, because of the two 24 steam generator tube sleeving and plugging, et cetera, no 25 longer -- can't go to full power any more. It's full power ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 94 BRT
('T
~'
1 is around 90 percent power.
2 It looks like at this point that these check 3 valves are improperly sized, such that at the lower flow 4 rates associated with its lower power, they sit down into 5 the flow stream and licensee -- I don't want to get his 6 thunder, but he's going to tell you some information why 7 these failed very rapidly under this particular situation.
8 So he thinks he has been able to demonstrate why, within 9 less than one year, all these seemed to fail.
10 MR. CATTON: Flow vibration caused the nuts to 11 loosen up?
12 MR. MARTIN: Yes, fv ') 13 MR. EBERSOLE: Suppose these valves remained 14 open, and I was still able to -- not the steam driven but 15 through the motor-driven pumps -- if I could drive primary 16 water in through the -- sorry -- secondary water through 17 the aux feed pumps, the motor-driven pumps, would it have 18 gone to the boiler by virtue of gravity or whatever?
19 MR. MARTIN: No.
20 MR. EBERSOLE: It would have drained out to the 21 low point?
22 MR. MARTIN: This is about 10 feet below the 23 feed ring.
24 MR. EBERSOLE: So I have a dried up secondary if 25 I had been unable to cope with these?
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6M6 j 1
25726.0 95 BRT
(~h
~
l MR. MARTIN: If there was no way to shut off 2 these paths, you are correct.
3 MR. EBERSOLE: This gets back then to the 4 critical nature of the check valves and their dynamic 5 performance, whether they shatter. I would think it might 6 be very prudent to have motor-operated or other kinds of 7 valves in front of these damn things to preclude the l
8 potential for -- you know, they can come close. Especially 9 if you have a broken pipe down here, they can come shut 10 with a crash.
11 MR. MARTIN: If you take that hypothesis, then 12 these valves, which are manual valves which will ultimately
() 13 be used in the event --
14 MR. EBERSOLE: If you can get this without 15 having to kill yourself.
16 MR. MARTIN: But if these had not shut off 17 successfully, then you would not have had the refilling and 18 the water hammer. The fact that they shut off and are now 19 in place as barriers now can't be compromised by pieces 20 coming back, because there's no flow path here.
21 MR. EBERSOLE: Sure. If they were able to shut.
22 MR. MARTIN: That's right. So any scenario you 23 try to use to get there, if you compromise these, then the 24 water hammer will not destroy this station because you 25 certainly have a tremendous attenuation of the pressure if O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
i i
25726.0 96 l BRT
(~')
'# 1 there's a free path through there.
2 As a result, then this line would probably be 3 available, this manual valve would probably be available to 4 shut.
5 MR. EBERSOLE: If you can physically get near it?
6 MR. MARTIN: Yes, sir.
7 MR. REED: I noticed we used the word shatter 8 here, you used it, Jesse used it. These valves came apart.
9 I am not sure how they were tank-welded or what. In my 10 experience, I don't think I have ever seen a shattered 11 check valve. I have seen a lot of them come apart.
12 MR. MARTIN: Although we don't know it's
/~)
(_/
13 connected with the event, this valve, the four-inch, which 14 took the brunt of the shock wave, its seat is just cracked 15 like spiders. It's spiders all through the stellite seat 16 there.
17 MR. REED: Did it shatter?
18 MR. MARTIN: In this case it didn't.
19 MR. EBERSOLE: Isn't it a fact of life that the 20 dynamic duty on a check valve is rarely if ever, certainly 21 sporadically, checked by opening the pipe below it? Nobody 22 looks at that dynamic need performance as a routine design 23 consideration, do they?
24 MR. MARTIN: I can't answer that.
25 MR. EBERSOLE: Only a few people do. It's not a O
v ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 8(XF336-6M6
25726.0 97 BRT
(
1 generic requirement that you look at the dynamics of 2 closure against an upstream failure of the pipe.
3 MR. fiARTIN: Can you answer that?
4 MR. LANNING: That's true. There's no such test 5 that includes consideration of a pipe rupture upstream.
6 MR. EBERSOLE: Even for the boiler that handles 7 primary fluid?
8 MR. LANNING: Yes.
9 MR. EBERSOLE: I mentioned studies to a certain 10 consultant and they begin to look dismal and the 11 unfortunate result of that was to cut off viewing down the 12 dismal road. I think that's a sorry state of affairs; when
() 13 you start getting poor results in your analysis, you turn 14 off your analysis, but nevertheless, that's what sometimes 15 happens. I think now we have the trigger to do some of 16 this work.
17 Go ahead.
18 MR. MARTIN: We believe the event was 19 significant because all in-plant AC power was lost for four 20 minutes, all steam generator feedwater was lost for three 21 minutes, a severe water hammer was experienced in the 22 feedwater system that caused a leak, damaged the plant 23 equipment and challenged the integrity of the ultimate heat 24 sink, all indicated steam generator water levels dropped 25 and the reactor coolant system experienced an acceptable O
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-(M6
25726.0 98 BRT 1 but unnecessary cooldown transient.
2 (Slide.)
3 From here on we go into our findings. The 4 primary cause of the water hammer was the failure of the 5 multiple check valves. The failure of the five check 6 valves in the feedwater system provided a mechanism for a 7 potential common mode failure of the heat sink, that the 8 long horizontal runs of the feodwater piping with the 9 potential for voiding are particularly susceptible to 10 destructive steam concentration-induced water hammers.
11 It turns out 9B feedwater line is the longest 12 one. It runs about 202 feet -- 220, I guess, and it's
(~T 13 perfectly horizontal the whole length.
x-)
14 (Slide.)
15 MR. MICHELSON: Have you thought about why the 16 check valve in the bypass line was the one to have lifted 17 its bonnet instead of the flow control valve in the main 18 line? Since the check valve in the main line was already 19 open and apparently laying in the bottom or something, why 20 didn't the bonnet on the big line lift instead of on the 21 little line?
22 f1R. MARTIN: Carl, it would be difficult to say 23 because we don't know for a fact that the upstream valves 24 were fully shut. If they are not fully shut you attenuate 25 the wave.
O ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80rk336-6646
a m34 gJ.- '4 4 m+ 2'.- A-.J*-J**.6 m4 A4.8
-s -+ _ ._.4 S JL c_, +A..L3-l
,. 25726.0 ~
99 i BRT O. 1 MR. MICHELSON: That's right. You know if the
[ 2 bypass was fully shut, I guess, but you are not sure if the I 3 main flow control valve was? I 4 MR. MARTIN: As a matter of fact, we don't know i
! 5 if the-main flow was except for the check valve. The j 6 indication says they were shut, but there's too many 7 parameters we don't have knowledge of.
8 MR. MICHELSON: A small leak won't make any
- 9 difference. The check valves themselves I think have 10 little holes in them purposely.
I 11 MR. LANNING: Not these.
4 12 MR. MICHELSON: Not these?
13 MR. MARTIN: The ones down by the feed pumps do
(
14 but these up here do not.
f 15 MR. MICHELSON: But in this case you knew the 16 check valve probably had already lost its disc; is that c
i 17 right?
)
l 18 MR. MARTIN: Yes.
I i 19 MR. MICHELSON: On the main line?
f I 20 MR. MARTIN: Because of the denting.
l 21 MR. MICHELSON: And yet the only damage to the I
] 22 flow control valve seemed to be operator damage due to a 23 rapid acceleration, I guess, bending the stem and breaking I
24 the yoke.
l 25 MR. MARTIN: That's what it appears to be.
- ()
i ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 804336-6M6
_- _ _ _ - - - . _ . _ _ . . ~ . . _ _ _. . _ _ _ _ . , , _ _, -
25726.0 100 BRT O 1 MR. MICHELSON: Yet no damage to the gate. Of 2 course it depends on the design of the plug or gate or 3 whatever it has got as to whether it would have -- but the 4 bonnet certainly saw a large pressure wave I would have 5 thought, since it's seeing the reflected more so than the 6 little bypass where it has to go through a little elbow and 7 so forth to get to it. It doesn't make sense that the main 8 line didn't see the big damage instead of the bypass line.
9 flR . EBERSOLE: May I ask, what capability if any 10 has this plant t '. > feed coolant --
11 MR. MICHELSON: Let me get an answer to my 12 comment or question first. Did you think about it? Has
()
13 anybody looked at why the bonnet on the bypass lifted 14 instead of the bonnet on the main line?
15 MR. SERKIZ: Carl, I can't give you an exact 16 answer because we don't have a detailed stress analysis on 17 all of them. Structurally, the bigger valves are 18 structurally more capable of taking a certain psi loading.
19 MR. MICHELSON: There's a lot bigger areas, too.
20 MR. SERKIZ: All that we know after the fact is 21 that the four-inch valve, the bolts plastically yielded.
22 As soon as that opened and started to reduce the water 23 hammer, that acts as a safety, in effect.
24 MR. MICHELSON: You can see my question, though.
25 My question is why did the little one on the bypass lift A
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage M0-336-6M6
i I
}
25726.0 101 BRT "A )
1 instead of the bonnet on the main line?
2 MR. SERKIZ: On the bigger one the bolts are 3 bigger.
4 MR. MICHELSON: But it didn't seem --
5 theoretically, why the shock would have propagated itself 6 at that point --
7 MR. SERKIZ: It's at the same point for all 8 practical purposes.
9 MR. MARTIN: Carl's point is when you go around 10 that 90 degrees you lose pressure.
11 MR. MICHELSON: It's 90 degrees away from where 12 we think the water hammer occurred. There must have been a
() 13 great deal of attenuation through the elbow, but that's 14 where the shock seemed to have found a weak point?
15 MR. CATTON: Was it on the big pipe or the 16 little one?
17 MR. MICHELSON: On the big one.
18 MR. MARTIN: It propagates all through.
19 MR. CATTON: I understand it propagates, but 20 where did it initially take place?
21 MR. MARTIN: Near the steam generator in the 22 horizontal line.
23 MR. CATTON: That's in the big one?
24 MR. MARTIN: Yes.
25 MR. SERKIZ: In the last horizontal run before O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage MXb336-6646
25726.0 102 BRT I)
l taking the vertical upturn into the steam generator.
2 MR. MICHELSON: A long way from this little 3 bypass line which takes of f from the bypass line back to 4 the valve.
5 MR. EBERSOLE: I would like to get to a real 6 root issue here. This is an unprecedented failure as far 7 as I know, all these coincident failures. I would like to 8 loft us off to the Palo Verde case, and I ask the question 9 for this reason. This plant, I think, at least it has some 10 PORVs, and presumably may have some potential for feed / bleed 11 if Palo Verde or its kind do not. Are the/ standing in the 12 presence of this sort of check valve failure and therefore
(~T 13 up the creek for sure? That's why I asked the question, V
14 can this plant feed / bleed, because I know the Palo Verde 15 plants cannot.
16 MR. KENNEDY: The procedures do have 17 instructions for feed / bleed.
18 MR. EBERSOLE: They can feed and bleed. On the 19 other hand, I must now say in the absence of another 20 situation, Palo Verde may have a network of valves 21 precisely like this, and that has not been introduced into 22 the PRA studies or aux feed failures. So let's inject into 23 those studies the result of this plant performance.
24 MR. MARTIN: Do you have any comment on that, 25 Mayne?
O ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-3366M6
- . - . .- ~- .
1 1
25726.0 103 BRT O-1 MR. LANNING: No.
4 2 MR. KERR: Let's not be guilty of asking for
! 3 information which the staff doesn't have.
l 4 MR. EBERSOLE: I'm saying let's inject the i
j 5 findings here into the Palo Verde study. We've got to take 6 this result and inject it into the current -- were is it, 7 10 to the minus 5 failure rate of aux feed?
8 MR. KERR: That's a goal.
9 MR. CATTON: This one has two CE plants.
10 Wouldn't they be the first to carry this over to their f 11 Units 2 and 3?
i 12 MR. EBERSOLE: Here we have a new input to aux i
() 13 feed failure, which, as far I know, is not present in tho j
14 plants that have defended the plants.without feed / bleed.
15 MR. KERR: Water hammer is not unknown
] 16 previously.
1 17 MR. EBERSOLE: But its consequences in this t 18 failure haven't been defined.
19 MR. KERR: I have heard discussions about the 20 consequences of water hammer since I have been on this 21 committee. I 22 MR. EBERSOLE: They have not been inserted into 4
23 Palo Verde.
24 MR. CATTON: They have been considered resolved?
i 25 MR. MICHELSON: Yes. They have been resolved.
k C:)
i 1
ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6M6
_ . . _ _ _ _ . _ - - - _ . ~ . _ - . - , ,
25726.0 104 BRT
('N
%) 1 MR. MARTIN: The timing of the five check valve 2 failures could not be ascertained with certainty; however, 3 the team concluded that all check valves had failed prior 4 to the event because the missing parts of the valves were 5 not found in the inspected feedwater piping after the event, 6 which would have been upstream of the valves.
7 MR. MICHELSON: In that regard, has the utility 8 looked at other check valves, you know, that were not 9 thought to have failed? Did they look at some of them 10 uc.rway to see their condition?
11 MR. MAR 1IN: In this plant there were 13 check 12 valves of the same design. They looked at all but one
() 13 before we had left and that one that was left was in the 14 turbine plant cooling water system.
15 My understanding is they are looking at other 16 check valves, but I don't know the details of their 17 examination.
18 MR. MICHELSON: Have they found any indications 19 of problems with these valves other than the ones that 20 failed?
21 MR. MARTIN: My understanding is they have not, 22 but we'll be having specification on the check valves.
23 MR. KERR: Number 6, the second sentence seems 24 like something of a non sequitur. The first sentence 25 refers to the testing not being able to tell whether check O(>
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8M336-6M6
25726.0 105 BRT
('"'
1 valves are closed. The second sentence, although it has a 2 "thus," seems to refer to the inability of testing to 3 determine whether the valve is damaged or not, not whether 4 it's closed.
5 Is there a sentence left out or something?
6 MR. MARTIN: Well, let me see.
7 Uhat it's basically saying is we don't have 8 confidence in the surveillance procedure as it is written 9 to tell you that your check valves are failing in this mode.
10 MR. KERR: But the first sentence says, 11 " determining whether check valves are closed." Now, if it 12 meant to say " determine whether check valves are
(} 13 deteriorating," then I can understand the two sentences.
14 MR. MARTIN: Okay. I understand what you are 15 saying. We don't even believe it could demonstrate 16 adequately that the check valves are closed.
17 MR. KERR: So the "thus" ought to be out of 18 there and that's a second thought not related to the first, 19 I assume?
20 MR. MARTIN: Okay. That's part of it. The 21 trouble is the results of the test as designed are very 22 subjective. Basically you are looking at a one-inch line 23 and looking at the flow coming out of that line that has 24 maybe a 10-pound differential pressure --
10 pound?
25 Pive-pound differential pressure driving it through --
O l
l l
ACE-FEDERA1. REPORTERS, INC.
202-347 3700 Nationwide Cmerage 80tk336-6646
. - - . _ - _ - _ - . . = - ~ = - -. _- _.
N 25726.0 106 BRT I
4 1 MR. KERR: I don't disagree with that. It just 2 seems to me those are two separate thoughts. I was trying i 3 to understand whether I.was missing something. I think you
, 4 have now convinced mere that there are really two separate
! S thoughts. One has to do with testing, find out whether the e
6 check valves are closed; the other has to do with testing, I 7 find out whether they are deteriorating.
8 MR. MARTIN: That's the last thing here, I think, 9 what we are talking about.
t
- 10 The first is the test as designed looks for flow 1 11 coming out of a little hose, and saying that that flow is
- 12 representative of throttling or not, and you only have 5 psi 4
() 13 in the water through th'is network of pipes, network of j 14 valves, the little drain valves, the long holes. And you 1
15 look at that and say, aha, that means it's closed.
i
] 16 Is it 1 DP34679? 10 GPM? No mesh.
17 MR. KERR: I'm convinced.
j 18 MR. MARTIN: Last sentence here, further, even I 19 if they demonstrate that they are closed, it doesn't 20 demonstrate it's going to be closed the next time because 21 you may just be lucky this time, the seat hits and bounces I
22 off and then shuts. Even though the nut is loose, maybe J 23 even the nut is off. It doesn't really matter.
! 24 (Slide.)
J I
L 25 At the time we completed our investigation the i
()
l
- ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336 6M6 f
25726.0 107 BRT (l
1 in-service testings program had still not been approved by 2 the NRC. It had been revised in its entirety in 1984 and 3 part of the problem with resolving that was both SCE and 4 NRC's responsibility.
5 We then looked at the resolution of the water 6 hammer unresolved safety issue. We found that it did not 7 specifically address the prevention and mitigation of the 8 consequences of condensation-induced water hammers in 9 feedwater piping upstream of the feed ring. In our 10 interviews of NRC Staff we determined that the resolution 11 of the water hammer issue failed to develop a citable 12 reference, a decision or discussion that provided a basis
(} 13 for excluding further consideration of feedwater piping 14 water hammer. However, we did note in the regulatory 15 analysis that the Staff acknowledged that the elimination 16 of water hammer:o was not feasible; that the frequency of 17 water hammers had been substantially reduced by changes in 18 design and operation; and that studies of water hammer had 19 revealed a significant lesser safety concern than 20 previously hypothesized.
21 It appeared to us that further consideration of 22 water hammer due to main feedwater line voiding was not 23 pursued due to a lack of reported occurrence in NRC U.S.
24 plants.
25 MR. EBERSOLE: That's right. He always wait for O
ACE-FEDERAL IAEPORTERS, INC.
202 347-37(o Nationwide Coverage NH-33M/46
25726.0 108 BRT O
\' 1 something to happen.
2 (Slide.)
3 MR. CATTON: What is done differently in the 4 check valve testing for the primary system to RHR7 5 MR. MARTIN: We did not look at those.
6 MR. CATTON: It seems to me that you would test 7 check valves the same way throughout a plant. If those 8 check valves are tested similarly to these you would come 9 to the same conclusions if they are not tested?
10 MR. MARTIN: Licensee even recognized some of 11 the deficiencies in their test procedure and had revised 12 test procedure even before this event so that the February
() 13 tests that were performed on the check valves near the 14 steam generators used the plant in mode 3 with a 700-pound 15 differential. Now when you've got an open hose and 700 16 pounds, if you can't tell the difference whether it's open 17 or shut -- that's no longer subjective.
18 MR. KERR: Does one have to use check valves?
19 MR. MARTIN: Instead of a stop valve of some 20 type?
21 MR. KERR: A positive action valvo.
22 MR. EBERSOLE: If you don't the reverse pressure 23 will get to the low-pressure components so quick.
24 MR. KERR: You just use an ordinary valve in the 25 thing.
O ACE. FEDERAL REPORTERS, INC.
202-347 3700 Nation *ide Coserase *n33MM6
25726.0 109 BRT 1 MR. MICHELSON: Not fast enough.
2 MR. KERR: In the first place, some of the 3 fastness that we require comes out of these design basis 4 accident analyses that we do which are unrealistic.
5 MR. MICHELSON: No, no, for the overpressure 6 protection they are code requirements.
7 MR. KERR: If you are telling me that the things 8 don't work --
9 MR. MICHELSON: Then you've got to fix them.
10 MR. KERR: But can you?
11 MR. EBERSOLE: Suppose you don't have a check 12 valve and the pump trips?
() 13 MR. KERR: You mean somehow the nuclear industry 14 has gotten a lot of defective check valves and really if 15 they got the right valves we wouldn't have this problem?
16 MR. MICHELSON: No, if you have misapplications 17 such as downstream turbulence -- in other words, by putting 18 a check valvo very close to a flow control valve with a lot 19 of downstream turbulence you are going to expect the check 20 valve to flutter, and if it flutters it's going to fret and 21 wear. So you have to be careful to put the valves in that 22 will take that kind of condition.
23 There have been cases where the stake nuts have 24 oven become unstaked, in sufficiently violent conditions.
25 MR. KERR The answer to my question is you need O
ACE.FEDERAt. REPORTERS, INC.
202 347 3700 Nationwide Coverage Nn11M6:4
25726.0 110 BRT 1 to have et.'ck valves?
2 MR. EBERSOLE: The alternative would be an 3 extremely fact valve driven by a hydraulic tube or cylinder, 4 to reduce the closing times. Check valves which are faster 1 5 than any others.
6 MR. REED: If you'll think back to the status of 7 your subcommittee, I filed five to 10 problem areas and one 8 of them was valves. We have a very active committee under 9 Mr. Michelson, who is looking at the fragmented industry of 10 valve business and to the -- not upgraded to nuclear grado, 11 really.
12 MR. KERR: The experts tell me we have to have
() 13 check valves. That's why I asked.
14 MR. MICHELSON: In these particular cases the 15 check valve is the logical choice.
16 MR. EBERSOLE: Is there any particular reason?
17 If you go, theoretically there's no particular reason that 18 you couldn't drive a valve, disc, closed by means other 19 than reverse flow, but it would have to be an extremely 20 high capacity, high-torque machine.
21 MR. KERR: Main stream isolation valves are 22 extremely rapid, aren't they?
23 MR. EBERSOLE: They close by virtue of reverse 24 flow.
25 MR. MICHELSON: There are other types that are a O
i ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage Mk336 6644
25726.0 111 BRT 1 little less susceptible to this problem, but not --
2 MR. KERR: In any event, these valves aren't 3 going to solve these problems.
4 MR. REED: He's going to solve it right away.
5 f1R. EBERSOLE: The consequences of not having 6 rapid closure would be pressurization of the downstream 7 components quickly. You'd blow the pump casing, low 8 pressure -- it gets back into the suction side so fast. So 9 you have to move fast.
10 MR. MARTIN: Okay. One of our concerns was when 11 we looked at the J tube fix for the feedwater ring water 12 hammer, we noted that, whether it said it therein or not,
() 13 it relied heavily on the fact that the check valves held 14 because the J tubes were really designed to prevent 15 draindown of the feed ring, and if you got the check valves 16 on the inlet to those feed rings unseated, then the J tubes 17 are not going to do anything for you.
18 We looked at the root cause of the phase-to-phase 19 fault. The loss of power was due to the phase-to-phase 20 fault. I don't think any more needs to be said there.
21 We also noted that the plant was designed to 22 experience an extended loss of in-plant AC power, based 23 upon the lack of an automatic loading of the diesels.
24 We further noted that the procedures on loss of 25 AC power lacked guidance on how long the operators could O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage fun 33MM6
A. ,
25726.0 112 BRT 4 m
'-] 1 attempt to restore power from an of f-site source before the 2 diesel generator should be loaded. And, further, how long -
3 the diesel generators can run unloaded without overheating 4 if their AC-powered fans remain deenergized. This caution 5 had been in previous procedures but it had somehow gotten 6 dropped.
7 (Slide.)
8 He noted the ntation loss of voltage auto 9 transfer scheme for establishing the delayed access did not 10 function as designed and they are evaluating that.
1 11 We noted that there were a number of spurious 12 indications of safety injection, that the operators did the
() 13 right thing. However, we felt that annunciation of safety 14 injection when it is really not warranted challenges 4-the 15 operators, causes additional concerns, and we didn't feel 16 that was appropriate. The operating staff, with the 17 concurrence of management, did follow appropriate 18 procedures when troubleshooting the electric ground. ,
19 However, their actions were unnecessarily -- they ,
20 unnecessarily delayed entry into the technical 21 specification action statement requirements that cod 1d 22 require plant shutdown.
- 23 HR. KERR: Excuse me. You read "did." That 24 says "did not." Which is it, "did not follow appropriate 25 procedures"?
O ACE-FEDERAL REPORTERS, INC.
202 347-3701) Nationwide Cmcrage 800-336-6646 r
25726.0 113 BRT 1 MR. MARTIN: You are correct. Did not. They 2 had a procedure, and if they had had followed it step by 3 step it would have taken them into action statement. Every 4 time they got to a statement that would have taken them to 5 action statement they looked at what else could we do 6 before we enter that action statement to remove other 7 causes? And they did that three times during this event.
8 MR. KERR: Is that illegal or improper?
9 MR. MARTIN: It's not illegal. It's a concern 10 to us in that had they done those things they would have 11 very quickly defined where the problem was and, we hope, 12 they would have carried through by deenergizing the line I
() 13 which ultimately failed. They had time to do it. Because
. 14 of the way they did their testing, they actually introduced i 15 additional stress on the line that failed because it was -- {
16 we had a delta with no ground on it except this ground 17 which had developed on the cable.
18 They paralleled that system with a fault on it l 19 to a grounded wire. That generates circulating currents 20 through the degrading resistance of the cable and may have 21 contributed to early failure. The things they did were not 22 discussed in the procedure. And, as we went through the 23 procedure we said: Why did you stop here?
24 The next thing, you had done everything 25 necessary, you reduced power, you had taken off components.
O
%/
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage sn336-646
25726.0 114 BRT
[
1 All you had to do was turn the little switch. Why did you 2 not go the next step?
3 And the answer we got almost unilaterally was 4 well, that would have taken us into an action statement and 5 there were other things we could do.
6 MR. KERR: Why didn't they want to go into an j 7 action statement?
l l 8 MR. EBERSOLE: Shut them down.
l l 9 MR. MARTIN: That's subjective. I can't answer l
{ 10 the question. They never answered that.
I I
I 11 MR. KERR: It seems to me you ought to look into l l
12 that, because apparently that was a contributor to the l
l (~)
w_/
13 situation, and I would think one would want to know why I
14 they were reluctant to go into a action statement.
15 MR. EBERSOLE: Wasn't it that they had a short 16 time before they'd have to shut down?
17 MR. MARTIN: That's right. But -- this was a 18 troubleshooting activity. We allow people to enter that i
19 for surveillance, for troubleshooting -- these are allowed 20 and reasonable means for going into action statements. Our 21 concern was, why didn't you take that? Your logic was 22 wrong, is what we were communicating to them. If the 23 concern is the ground is there you are going to have to 24 shut down, you are going to have to shut down anyway if the 25 ground is there. Therefore, it's unreasonable to go these (v'h ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationside Coverage 800-336-6M6
25726.0 115 BRT (m
\,
1 extra steps.
2 MR. KERR: I don't disagree. At least I don't 3 hear anything with which I would disagree. But there seems l
4 to be a reluctance on the part of this group to go into an '
5 action statement.
6 You have pointed out one situation in which they 7 would have been much better off had they gone into an 1
8 action statement, presumably. From a number of viewpoints.
9 If I were you I would want to find out if this l 10 was just typical of this one licensee or is there some 11 widespread reluctance to go into action statements?
12 MR. MARTIN: You are asking for a generic
() 13 response and I can't give that. Jack, can you?
14 MR. HELTMAS: That's part of the sequence of 15 training procedures that they have --
1 l
16 MR. KERR: I'm not talking about training and 17 procedures. I'm talking about an attitude. By all means, 18 we want to avoid going into an action statement. Why?
l l 19 Because it, apparently, conu ibuted, at least in your view, l 20 to the situation being much worse than it might otherwise 1
21 have been.
22 MR. EBERSOLE: It sets them to a wind down for l 23 shutdown; isn't it?
24 MR. MARTIN: If they found a problem there. But 25 if they found a problem there they'd have to shut down anyway, ew ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Co erage 800-336-6646
25726.0 116 BRT
/~T
\'/ 1 I agree that the procedure could have been 2 written better so as to have allowed them to do other 3 things, safe things to check that the procedure didn't 4 address. But there was no liability in following the 5 procedure and that was our concern. Why did you take these 6 unsafe steps when you didn't have to?
7 MR. REED: I'm concerned about this action 8 statement. In fact, the way I visualize it, if these 9 people had have gone into an action statement, this 10 incident might not have happened and you might never have 11 discovered the five check valves that were in pieces. This 12 is an attitude that bothers me because I hear all the time 13 that we want to prevent scrams or trips on reactors. Many (v')
14 times trips on reactors disclose problems, and I don't 15 think we have evolved in the nuclear business yet where we 16 want to avoid things happening. We should still be 17 discovering. I'm very happy that the guy did what he did 18 because that brought about an electrical collapse which 19 brought about a disclosure of the real root issue.
20 MR. MARTIN: Mr. Reed, they would have found 21 these problems in their later in-service testing, which 22 supposedly is done on a quarterly basis, if they get around i
23 to it. )
l 24 MR. REED: Is it a disassembly, in-service 25 inspection?
O V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
,, . . . - - - . . - -~ ._- . . _ . .
.25726.0 117 BRT I I .
, 1 MR. MARTIN: When the disc is laying in the 2 bottom of the check valve it is impossible for it to
- 3 develop any DP.
4 MR. REED: You are saying they would have found a
5 it some day?
i
- 6 MR. MARTIN
- They would have found all five of
- s 7 them? 4 l 8 MR. REED
- On a 10-year program?
9 MR. MARTIN: They are on a quarter, every three
-l
- 10 months.
l 11 MR. REED: These valves are tested every three 12 months?
()
13 MR. MARTIN: That's the description, if they get 14 around to it.
15 MR. REED: You just said "if."
16 MR. MARTIN: They would get around to it in a 17 refueling outage because there's enough time during a 18 refueling outage to do these things. The refueling outage 19 was scheduled November 30. Nine days later they would have 20 gotten into a condition to find these.
21 MR. REED: Chronological things may have stacked 22 up in favor of what you are saying, but quite frankly, I'm 23 happy that the chasing of the ground was the way that it 24 was and that we disclosed another problem in valves, and I 25 hope that our valve committee will cause a great deal of C:1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 118 BRT f
x b 1 pressure to do something about the industry.
2 MR. EBERSOLE: Let me ask you how much longer 3 your presentation is? I have another meeting here.
4 MR. MICHELSON: Could you clarify a point on the 5 diesel engines? You made a point, is it correct to assume 6 that these are air-cooled diesel engines?
7 MR. MARTIN: They are water-cooled, which goes 8 to a .adiator. The radiator is in the overhead. You can 9 have convection fueling that lasts about 39 minutes.
10 (Slide.)
11 MR. MICHELSON: How long can they run without 12 building or room ventilation?
() 13 MR. MARTIN: Basically this is open screen to 14 let air come in and go up through the radiators. They have 15 fans that are AC-operated. You can operate unloaded for 16 about 39 minutes before you need to have these fans turned 17 off.
18 MR. MICHELSON: That's the only dependency on 19 off-site AC power?
20 MR. EBERSOLE: When that temperature is exceeded 21 are there loud alarms and annunciators?
22 MR. MARTIN: There are alarms. What the actual 23 set one is I don't know, but all you have to do is put the 24 diesels on the buses and they would have come on 25 automatically.
O ACE FEDERAL REPORTERS, INC. ;
202-347-3700 Nationwide Coverage 800 336-6M6
~--
25726.0 119 BRT t'~h V 1 MR. EBERSOLE: You can kill the diesels by 2 ignoring them?
3 MR. MARTIN: Yes. Once the electrical ground 4 was locked on feeder from the auxiliary transformer C to 5 bus IC, the operators did not aggressively pursue isolating 6 the auxillary transformer. Instead they opted to leave the 7 transformer energized while technicians performed 8 inspections that did not require the transformer to be 9 energized. Our concern is they left themselves vulnerable 10 while they performed inspections that didn't require that 11 vulneraoility.
12 (Slide.)
(} 13 The operator actions after the transformer trip 14 were consistent with their training. However, in the team's 15 judgment, some operators lacked detailed plant knowledge 16 and we list the areas.
17 (Slide.)
18 On occasion, some site personnel who generally 19 evaluate plant data lack sufficiently inquiring attitude.
20 As a result, certain significant indications of underlying 21 reasons for system response or component performance were 22 not detected until brought to the attention of SCE by the 23 team. It appears that SCE's process for evaluating and 24 following up events eay not be sufficiently thorough and 25 systematic to assure that failed components are detected k_s)
ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
1 25726.0 120 1 BRT s
s
) 1 and adequately explained.
2 I have to caveat that somewhat. We were pushing 3 them. Our only job was the investigation. They had 4 preparation for the refueling outage and things of that 5 nature. We may have been just a couple of steps ahead of 6 them, but our perception was that they needed to beef up 7 there event evaluation program, including, somehow, making 8 sure that the people they select for that function are very 9 inquiring and won't accept the easy answer.
10 The status of steam generator blowdown systems 11 is not indicated in the control room. We talked about that.
12 During the loss of all in-plant AC power, 13 sufficient information was available in the control room to
(~)h 14 enable the operators to follow their procedures and ensure 15 plant safety. However, control room operators had failed 16 to have the technical support center reset following the 17 electrical ground troubleshooting activities. This failure 18 disabled the computer's ability to record new plant data 19 and thereby denied the operators access to pretrip and 20 post-trip trends that would have assisted real time and 21 post-event analysis and evaluation. flad the station 22 blackout been of longer duration or involved additional 23 complications, operator responses and the functions 24 provided by the technical support center could have been 25 hampered by the lack of trend data.
O a ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
25726.0 121 BRT
-O 1 Not only did we lose the computer but most of 2 the recorders.
3 (Slide.)
4 Maintenance records are incomplete, difficult to 5 locate and, when available, lack sufficient detail to 4
6 determine what was done.
7 In particular, when we tried to follow what was I
8 done on the check valves in the past, the records were very 9 bad.
, 10 The spurious ringing of the NRC red phone did 11 cause some distraction of people and some confusion.
12 Here most of this is NRC problem.
4
(} 13 14 Communications between NRC and SCE were not effective because: the NRC duty officer was not knowledgeable about 15 the unique design of the plant and therefore misinterpreted 16 operator responses to questions; communications with the 17 plant were initially limited because statements by plant j 18 operators incorrectly implied that sufficient personnel
-19 were not available to support the establishment of an open 20 line; NRC asked leading questions and operators sometimes 21 did not correct, and in some cases appeared to confirm, 22 inaccurate information; NRC questions characteristically l 23 focused on detail rather than on the big picture; NRC 24 cluttered the communications with repetitive discussions
, 25 about the sequence of events, and we go on and on. The
- ()
ACE-FEDERAL REPORTERS, INC.
, 202-347-3700 Nationwide Coverage 800-336-6646
,.-. - .- - - - -- - . .- = _ , - - . . - . . _ . - - _
i l
25726.0 122 BRT 1 problem is, we need training. We need to do a better job 4
l 2 in our communications and these were points that we were
- 3 making.
4 (Slide.) ,
- 5 There were two malfunctions of the automated
. 6 security system; however, site personnel implemented 7 appropriate procedures and it precluded a safety-safeguards 8 interface problem and there was no significant release of 9 radioactivity.
i 10 MR. MICHELSON: Before you get to the conclusion,
)
- 11 let me ask a question on observations. Have you made any 12 inspection of the steam generator itself?
I'
() 13 MR. MARTIN: The Licensee has. They looked at 14 the feed ring and found no problem there. I do not know i
15 the results down by the tube sheet. That was being done 16 after our investigation. I'm sure the Licensee can speak 17 to that.
18 MR. MICHELSON: Thank you.
19 MR. MARTIN: The most significant aspect of the 20 event was the five safety-related feedwater system check 21 valves degraded to the point of inoperability during a
! 22 period of less than a year without detection and that their V
} 23 failure jeopardized the integrity of safety-related 24 feedwater piping. The root cause of the failures have not j 25 been determined and are still being reviewed by SCE and its
(:)
1 i
l
~ I ACE-FEDERAL REPORTERS, INC. l
. I 202-347-3700 Nationwide Coverage 800-336-6646 l
25726.0 123 BRT
-h
~
l contractors. Potential contributors include inadequate 2 maintenance, inadequate in-service testing, inadequate 3 design and-inadequate consideration of effects of reduced 4 power operations. -
5 MR. MICHELSON: When did the reduced power 6 operations start? What date?
7 MR. MARTIN: November of '84.
8 MR. CATTON: What was the percentage reduction 9 in the flow-through of these check valves?
10 MR. MARTIN: Down to 90 percent, approximately.
11 MR. CATTON: They reduced it 10 percent and that 12 10 percent, they feel --
13 MR. MARTIN: It looks like also the selection of
{}
14 the valves was inappropriate. The Licensee will speak to
- 15. that.
16 MR. MICHELSON: They had to throttle to reduce 17 the 10 percent; the turbulence downstream of the throttle 18 perhaps is the problem.
19 MR. CATTON: It depends on how far away the 20 throttle is. I have trouble with that.
21 MR. MICHELSON: It isn't very far.
22 MR. MEDFORD: We believe the design of check 23 valves was marginal. The operation at 90 percent would 24 yield the inadequacy of the design. He have performed 25 analyses, we have brought in experts and can show the O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646
i i
1 25726.0 124 I BRT I
/~. 1
( )
'~'
1 quantitative results in calculations.
2 MR. MARTIN: Talking about the contributors, we 3 could not preclude maintenance because of the type of 4 raaintenance that was done and the records that were kept.
5 In some cases, records that we did have didn't tell us what 6 was done; in other cases, we just couldn't find the records 7 of what was done.
8 The in service testing records for these valves 9 were inconsistent. There were several records and they 10 didn't agree with each other. We think that was due to 11 administrative errors. The testing procedure was not 12 rigorous. We believe that it was very subjective, in some IT 13 cases. The test acceptance criteria was subjective. The
%-)
14 testing frequency was open-ended and the test did not 15 assure detection of the failure found.
16 Our point here on the IST is that if -- it's 17 important to do these tests on a quarterly basis and if 18 these things are subject to these types of failures then 19 you are going to have to go out of your way to perform 20 these tests at that frequency.
21 In fact, even though they are quarterly tests, 22 some of these valves have not been tested for over -- just 23 about a year.
24 The check valves and valves of similar design 25 have a history of like failures. You can find that in the
<Vn ACE FEDERAL REPORTF.RS, INC.
202-347-3700 Nationside Coverage 800-336-6M6
l 1
I' 1 25726.0- 125 BRT 1 literature. You can also find these valves, some of them 2 had failed in a similar manner before at this plant. And f 3 reduced power operations at Unit 1 are now routine, and we 4 talked about the thermohydraulic -- hydraulic-induced 5 vibrations there.
6 MR. EBERSOLE: Is there any way to look through 7 the valve body and see if it's fluttering? Like X-ray?
8 MR. MARTIN: The Licensee has used X-ray. It 9 hasn't been very much use to him and I can't explain why.
10 Possibly just because it's moving it doesn't give a good 11 image.
12 MR. EBERSOLE: They have no need to go to a
(} 13 hospital. They have instruments there that will do it.
14 MR. CATTON: Why not just vibration measurements?
15 MR. EBERSOLE: Let's have a 10-minute break 16 until 20 after and then we'll nove on to the remainder of 17 the presentation.
18 MR. KERR: Do we have to discover root causes or 19 are we going to hear something that says here are the root 20 causes, two or three?
21 MR. MICHELSON: What are the real root causes?
22 MR. KERR: Does that presentation indicate there 23 were 20 root causes or the root causes are still 24 undetermined?
25 MR. MARTIN: The root cause we believe is the O
ACE-FEDERAL REPORTERS, INC.
N)2-347-3700 Nationwide Coverage 800 336-6646
25726.0 126 BRT
(";
'/ 1 failure of the check valves. Why it failed, I'll leave 2 that to Licensee because he's done more evaluation than I ,
3 have. _
4 MR. KERR: I thought the event was the failure 5 of the check valves. If one was looking for root causes 6 for that failure --
7 MR. MARTIN: What I'm telling you today, I 8 cannot tell you what the failure was. The Licensee has 9 more information.
10 MR. EBERSOLE: This is kind of an interim 11 statement. This not the conclusion of the meeting. Let's 12 have 10 minutes.
()
%J 13 (Recess.)
14 MR. MEDFORD: I'm Mark Medford, the licensing 15 manager for Southern California Edison. On my extreme 16 right is Ken Baskin, the executive for licensing; between 17 us is Mr. Rainsberry; and making our presentation today is 18 Bruce Duncil, who works in licensing. I would like to 19 first comment briefly on the agenda. Looking at the second 20 purpose, it says that NRC is to be briefed by the NRC Staff 21 and SCE utility as to the NRC utility actions and 22 recommendations. ,
l 23 Jesse, as you indicated earlier, we are in i l
24 mid-program. This is by no means the end of our evaluation.
25 We are going to present to you today some of the actions
^\
[G ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
5
^
25726.0 127 l BRT
(]) I that we have identified would be appropriate as a result of 2 the event, but don't take this to be the final result of 4 3 our evaluation of the event.
1 4 Having said that, I would like to turn it over 5 briefly to Ken Baskin, who is going to address our attitude 6 in dealing with action statements. i 7 MR. BASKIN: And I will be brief. I think it.'s 8 important to at least clarify from our perspective this 9 discussion of a few minutes earlier about action statements.
10 There was no intent, to the best of our ability, 11 to ascertain on the part of the operators or certainly no '
12 intent in their training on the part of management to have
() 13 them resist shutting down the plant. As I think Tim 14 indicated, entering the action statement was really --
t 15 would have been relatively insignificant because if you 16 found the problem you would have had to shut down the plant 4
17 anyway. If you didn't find the problem you could' have i 18 exited anyway, so it really had no bearing in that context.
19 In our training of operators, we encourage them 4 -
20 not to enter action statements lightly because an action 21 statement at least infers that some portion of plant i
22 equipment or systems is not operating properly that would 23 normally be available and therefore could normally be 24 required.
25 So it's a matter of how one looks at action
()
t 4
i ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 800-336-6646
25726.0 128 BRT 1 statements. We train our people not to enter them lightly 2 and to do whatever else they can before they go into an 3 action statement. The normal situation would be that the 4 things you are doing without entering an action statement 5 would have less impact on potential safety.
6 It turns out in this case that, had we entered 7 the action statement earlier we might have discovered the 8 situation earlier. But obviously we didn't know that at 9 the time. So, our philosophy is not to look at these 10 lightly and not to enter into them lightly, and that's the 11 reason the operators reacted the way they did.
12 MR. REED: You would have discovered the
(~'i 13 electrical situation earlier, not necessarily the check V
14 valves.
15 MR. BASKIN: That's correct. As you indicated, 16 we may or may not have discovered the check valves.
17 MR. EBERSOLE: Thank you.
18 (Slide.)
19 MR. DUNCIL: Our presentation hora is geared 20 toward the water hammer event specifically. We'll go 21 briefly over the sequence of events, but concentrating 22 mostly on the feedwater check valves themselves, on the 23 design aspects and to follow up some of the conversation 24 earlier, and then to go into a discussion of some of the 25 corrective actions that we plan to take.
,0 V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationmide Coverage 80lk336-6M6
t 25726.0 129 BRT
- 1 As Mark indicated, our evaluations are still 1
2 ongoing in many respects and we'll touch on that as we go 3 through.
a 4 Relative to the event, we agree that the event
] 5 is very significant, from the standpoint that we had,
! 6 essentially, a common mode failure of five safety-related
~
l 7 check valves. Additionally, the most significant aspect of j 8 the event is the failure of the check valves.
! 9 (Slide.)
l 10 The root cause of the water hammer event is the 11 failure of those five check valves, and we'll get into the 12 root cause of the failure of those check valves as we get l.
j
() 13 into the presentation.
14 (Slide.)
< 15 Again, briefly, one-line diagram for the 16 feedwater system, the feedwater lines to the generators, 17 the 10-inch valves. I think we pretty well hit the 18 highlights on those. The east evaporator condenser was the 19 one that ruptured when the east feedwater pump was
{ 20 deenergized, and that's what initially started the sequence 21 of events.
22 MR. MICHELSON: What is the design pressure
^
23 rating of the piping upstream of the main feedwater pump?
i 24 MR. DUNCIL: Basically what you have is a 25 nonsafety-related/ safety-related break at the upstream i
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
. __ . . _ _ ~ _
25726.0 130 BRT O -1 flange of this valve. Upstream is 360 pound piping; from 2 this point to the generators you have 1350 pound piping.
3 It's all ANSI B131.
4 MR. MICHELSON: Have you reviewed the heat 5 exchanger dasigns to see what kind of rating might have 6- been on the tubing? In other words, why did the tubing 7 rupture in the evaporator inst'ead of one of the heat 8 exchangers? ,
9 MR. DUNCIL: We have done heat testing and other 10 tests on the feedwater heaters. Obviously the tube here 11 was the weak point in the system.
12 MR. MICHELSON: Did you look at the wall O() 13 thickness of the tubing and so forth in relation to the 14 diameters and determine if this is where the rupture should 15 have occurred or have you gotten to that fine structure?
16 MR. DUNCIL: I can't speak to that. I know this 17 being 360 pound piping, this being essentially over 18 1000-pound pressure from the west feedwater pump and this 19 pump already failed and this pump secured, this piping saw 20 the full discharge pressure from this feedwater pump.
21 MR. MICHELSON: Are you also saying the inlet ,
22 nozzle for the feedwater pump was rated at the same 23 pressure as the exit nozzle?
24 MR. DUNCIL: Yes. The code break is right here.
25 MR. MICHELSON: It's not entirely unusual, but
.O ,
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 131 BRT t%
1 somewhat unusual, isn't it, to rate the suction of the pump 2 at the same pressure as the discharge?
3 MR. DUNCIL: I'm not an expert in the design, 4 but usually what you do is take a code break at a flange.
5 MR. MICHELSON: It could be because they use the 6 safety injection pumps that you have to do this?
7 MR. DUNCIL: That's absolutely right.
8 MR. EBERSOLE: Let me examine the 9 safety-injection approach to that. Presumably you are in 10 an approach to a hazardous state to a core, you may or may 11 not succeed in injecting, so you are injecting water to the 12 primary loop with potentiality, in this case if something f')
a 13 happens -- and I see two check valves in parallel which 14 would pecmit reverse flow from a primary loop. Am I 15 correct?
16 MR. DUNCIL: I do know that has been looked at.
17 MR. EBERSOLE: Is that considered an adequate 18 isolation barrier from the primary loop when that function 19 is in place and you are trying to operate as an injection 20 system? It's the -- the pump casing can't take primary 21 pressure, obviously. That's 2200 pounds.
22 MR. RAINSBERRY: There's additional pumping, and 23 at least another check valve and a MOV.
24 MR. EBERSOLE: So there's other ways to 25 intercede.
("%
G ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
l 1
1 l
25726.0 132 l BRT O 1 MR. DUNCIL: The purpose of this slide is to 2 show the major milestones in the sequence of events for the 3 main feedwater hammer event itself.
4 (Slide.)
5 We went over that pretty well in the previous 6 presentation. The times are designed to show the time 7 differences between events. There is one typographical 8 error -- I apologize for it -- here. The combined aux 9 feedwater flow was at time 4 instead of time 8. Basically 10 the steam-driven pump completed its 3-1/2 minutes' warm-up 11 cycle at about the same time the electric power was brought 12 on and the electric-driven pump was started, and of course,
(} 13 that led to the water hammer event roughly 16 or so minutes 14 after aux feedwater flow was introduced.
15 (Slide.)
16 This slide depicts the -- an isometric of the 17 feed flow lines inside containment. As Mr. Martin pointed 18 out, this B feedwater line is approximately 205 feet in 19 length. It's a horizontal runout to containment 20 penetration. A and C are on the order of 103 and 107 feet 21 respectively. Also horizontal runs.
22 The water hammer event was caused by slug 23 formation. We'll get into that in the next slide.
24 We based that on our evaluations that the slug 25 formed in this vicinity here and then traveled down the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
_. _. m _ _ _ . . . . . _. . . _ - -_._ _ _ ~ -_ _ _ .
25726.0 133
.BRT CE) 1 length of the pipe toward the feedwater mezzanine area.
2 MR. MICHELSON: Where is the location of the 3 penetration itself on your isometric?
4 MR. DUNCIL: Right here.
5 MR. MICHELSON: Is that an anchor point? Not a t
6 bellows, a true anchor; is that correct?
7 MR. DUNCIL: I believe it is.
8 MR. MICHELSON: Have you inspected the 9 penetration itself, verified no damage?
10 MR. DUNCIL: The inspections have been done.
11 MR. MICHELSON: Can you get through all the 12 piping? Have you actually inspected it?
() 13 MR. MEDFORD: I know the penetrations were
- 14 inspected. There was no damage.
15 MR. MICHELSON: You are able to do an inspection 16 of that penetration all the way up to the anchor point?
17 MR. MEDFORD: That's correct.
18 (Slide.)
^
19 MR. DUNCIL: I'might preface this slide. I am j 20 not a water hammer expert; however, Dr. Chong Chui has been 21 working directly with Dr. Paul Rothe in CREARE -- in fact, 22 CREARE was brought on board about a day after the event to l' 23 work with us in developing an understanding of the I 24 feedwater event.
, 25 From that evaluation, the simplified
)
i i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
.. .,.,-.-..- _ ..- .,, . _ . - . . _ - , , . , _ . . . - . . . , , - - ~ . . _ . _ . ,
.. - - .. - ~ . . . . . - - - __ _ _ _ _ _ - _ - _ ___ ._ _
F
,- 25726.0 134 BRT
-O 1 understanding of it is that as the B feedwater line filled 2 with feedwater, obviously the steam being on top of the 3 feedwater as the pipe fills up, we did have some steam flow 4 that was not only due to condensation but we still have the 5 steam flow due to condensation on top of this, creating j 6 turbulence in the water here and generated a slug which 7 developed in the pipe, as I mentioned, up toward the steam j_ 8 generator, traveled down the pipe.
i I
9 At the time this occurred A and C feedwater 4
10 lines had already been refilled. You will recall they are
! 11 approximately half the length of the B feedwater line.
12 At the time the B feedwater line was filling the 13 auxiliary feedwater flow to this pipe was on the order of
(}
< 14 about 25 to 40 GPM. So it was filling rather slowly.
i 15 The other thing I might point out is, based on i
16 our calculations, the actual void fraction in the pipe was i
- 17 between zero and 15 percent at the time that the slug
] 18 formed. So, as you can see here, we attempt to depict that 19 there was very little area, steam area in the pipe.
! 20 MR. MICHELSON: Where are the traps along the I
! l 21 pipe? In other words, in what direction is the pipe 22 pitched for drainage? l
! i 23 MR. DUNCIL: There is no real pitch in this pipe.
I 24 MR. MICHELSON: There's always a little pitch in 25 a pipe. Do you know it was truly level to begin with?
t
,i l i !
ACE-FEDERAL REPORTERS, INC. j 202 347-3700 Nationwide Coverage 800-3366646 l 1
r,
....,,_,,,_.,--..._-,--,,..___.__...,_...,_.-,,_-_.,,,.m., , , , , . . _ , _ . , - _ _ _ , _ , , . , , . , .... ,__,,_,_.,_m-..,,. -., , _ _ _ , _ , . . _ , -_,,,)
. . _ . ~ .. .- . . . . . - - _ _ .
i 25726.0 135 BRT 1 MR. MEDFORD: The answer is we don't know. Of 2 course, the as-found condition after the event didn't lend
- 3 itself to that and we don't know from previously existing +
i 4 in fo rma t ion .
l 5 MR. MICHELSON: There was a possibility that I
6 there was a small pitch in this pipe even from growth, 7 expansion, deflection during the life of the plant such 8 that now there was a pool of water that went along the pipe.
9 The pipe might have been completely filled at one point 10 which greatly helps the generation of a water slug.
lT I
11 MR. CATTON: Here is the flow. They talk about 12 water flowing in.
() 13 MR. MICHELSON: What bothers me is the water 14 coming in is the warmest water in the pipe. It's cold when 15 it comes in, but it's coming in way back upstream.
i 16 MR. CATTON: The steam is being condensed where 17 the water is cold.
18 MR. MICHELSON: The warmest water, though, is ,
19 right next to the steam generator and it gets progressively 20 colder way back.
21 MR. CATTON: And you don't need much cooling to 22 lead to the water hammer.
23 Is it important that you know that? See, that 24 15 percent, is it important that you know that? Or is it f
25 just important that you know whether or not a water hammer O
ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cos erage 80 4 336-6646
25726.0 136 BRT will occur under those circumstances?
I 2 MR. DUNCIL: It was important relative to our 3 understanding how that slug was formed.
4 MR. EBERSOLE: That slug is flowing toward an 5 open pipe and also into a region where further condensation 6 will take place even if the pipe is closed; right? There 7 are two exit paths for steam, one to condense and one to go 8 out of a hole.
9 MR. DUNCIL: At this point in time the MOVs and 10 feed control valves were shut so there was no real path for 11 that until the water hammer itself opened up the four-inch 12 bypass check valve. So really, the only flow that was at'
() 13 this point in time when the water slug was formed was due 14 to condensation.
15 MR. EBERSOLE: Flow into condensation.
16 MR. DUNCIL: And this, again, I might point out 17 there was only about a four-minute period of time involved --
18 well, 1 guess it was more like five minutes involved, so 19 there wasn't a lot of time for this water to heat up, so it 20 was still relatively cold water.
21 MR. REED: Would you go back to the preceding 22 slide a minute? What I would like to ask is where does the 23 auxiliary feed come in? You may have said. I would like 24 to know how close to the steam generators does it come in?
25 MR. DUNCIL: The aux feedwater piping is down O
l l
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336 6646
25726.0 137
.BRT
_4 v
/ '~ l near the feedwater mezzanine, which, as you can see, it's 2 upstream of this point. I'm not sure exactly how -- just 3 outside --
4 MR. REED: One way to prevent water hammer, and 5 of course it was resolved by the NRC before, they talked 6 about J tubes. But there were a couple of plants that 7 decided that wasn't the proper cure. The proper cure was 8 to put check valves right adjacent to the steam generators, 9 quality check valves and that would take care of the water 10 hammer thing.
11 But you have to get your auxiliary feed on the 12 downstream side of that check valve so you've got your
( ') 13 auxiliary feed outside containment rather than bringing it 14 in to right close to the steam generator and then going 15 into the feed path. I'm just looking at your cures. It 16 looks to me like you have to put in a quality check valve 17 right up next to the steam generator but you have to get 18 your auxiliary feed downstream of that.
19 MR. EBERSOLE: Next to the boiler. Which means 20 long aux feed lines.
21 MR. DUNCIL: It would in this case. '
l 22 (Slide.)
23 As I mentioned earlier, not only did we bring 24 CREARE on line to help us in our evaluation of the feedwater 25 even, we also brought on board Mettek, Incorporated, which
/~N V
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336- % 46
25726.0 138 BRT l
/
k_')s 1 is a company which assisted us in doing metallurgical l
2 assessments of the valve damage as they were released from 3 the quarantine list. We also hired Dr. Kalsi to assist in 4 helping us analyze the valve design once that became 5 available to us.
6 We opened and inspected the feedwater pump check 7 valves and also the 10-inch valves. Basically what we 8 performed was a valve leak checks, blue checks, flapper 9 inspections, and also we took a look at the gaps between 10 the disc and seat.
11 The blue check is taking a look at the valve, 12 put some bluing compound on it --
13 MR. EBERSOLE: I know.
f) 14 MR. DUNCIL: The valve flapper is several pieces.
15 You have the main hinge arm, the disc attached to it with 16 the stud, the nut. There is a locking pin on these nuts, a 17 washer underneath that. And these two lugs here are the 18 anti-rotation devices which essentially prevent the disc 19 from rotating about that hinge arm.
20 MR. EBERSOLE: When you write the specs for that 21 did you define the potential closing velocity that that 22 valve will see on its seat in a hypothetical case of a pipe 23 stream, upstream of the valve proper and full pressure 24 applied to the seat?
25 MR. DUNCIL: I can't speak to that.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
. . _ _ _~ . ~ . - . .. _ _- . . . _ . _ _ - _ _ _ _ _
- i 1 25726.0 139 BRT 1
O 1 MR. EBERSOLE: Is there a dynamic specification 2 in procurement that says this valve may face a potential l
3 upstream pipe failure, abrupt, in which case it must 4 survive in rapidity of closure the presence of full 5 upstream water and one, thus, must calculate the closing i 6 seat velocity and determine that the resilience or 1 7 inelastic energy absorption process will be sufficient for
)
8 it to survive? Did you do that?
i 9 MR. DUNCIL: I can't speak to that. I don't 10 know.
1 11 MR. REED: On the five valves that failed, were l
- 12 these locking pins put in?
13 MR. MICHELSON: On this slide, am I looking at a
[}
i 14 10-inch disc there?
15 MR. DUNCIL: This is the standard for 10-inch, l_ 16 4-inch and 12-inch. All are identical, manufactured inr i
17 Pacific valve.
- 18 MR. EBERSOLE
- I have to say that these valves -
19 are designed against an ancient belief that there will i 20 never be any pipes ruptured. The only thing that will 4 -
21 happen is you turn the motor off on a pump and it will 22 decelerate, UR squared; it will make a gradual closure and i 23 the valve, historically, all it has to see is a nice t 24 closing, and never has to see the pipe break. We live in --
i 25 MR. KERR: You can tell that by looking at that?
()
)
t j ACE-FEDERAL REPORTERS, INC.
t 202-347-3700 Nationwide Coverage 80 4 336-6646
, . - - . . . , - , - . . . _ . _ _ _ - . - , , - _ - - . - . .- ~
-25726.0 , 140,
.BRT #
(
'V)-
1 MR. EBERSOLE: Of course. I don't see anything 2 in that that keeps it from being under severe duress if it 3 has to close under those conditions.
4 MR. DUNCIL: That's a possibility. I know that 5 did not enter into what we found based on our evaluations.
6- MR. EBERSOLE: Of course,that's not what really 7 happened.
8 MR. DUNCIL: Just to review briefly what our 9 findings were relative to the five valves that had failed, 10 in a number of cases the discs had become loosened. In a 11 couple of cases, the A and B feedwater lines the discs were 12 actually disconnected from the hinge arms. And we did find
() 13 indications here, like, for example, on the C feedwater 14 line, the disc rotated and cocked such that the valve was 15 opened. Basically the valves are not functioning.
16 MR. REED: I don't find the locking pins 17 mentioned.
18 MR. DUNCIL: In most cases, as I recall, those 19 were gone.
20 MR. REED: Here they ever put in?
21 MR. DUNCIL: To the best of my knowledge they 22 were.
23 MR. REED: In other words, you have no evidenca 24 that they were?
25 MR. MEDFORD: We do have evidence. Some records O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
25726.0 141 BRT q
1 that they were put in.
2 MR. REED: You do have evidence, fragments?
3 MR. MEDFORD: Some records, not all of them.
4 MR. MICHELSON: You have evidence of corroding 5 or what? As long as I got my question in, talked about the 6 pins, were they corroded or fretted or --
7 MR. MEDFORD: They showed signs of fretting, but 8 the failures tended to be in the root of the bolt.
9 MR. REED: You have to worry about chicken or
! 10 egg here. What came first, chicken or egg? The nut may l 11 have failed subsequent to it breaking loose.
l 1
12 I don't like that pin. It looks like a pretty
(} 13 flimsy way to lock a nut on check valve which is 14 flipflopping around. It's a pretty poor device. But if l 15 you found evidence that they were put in, that the 1
16 maintenance gang didn't foul up and leave them out -- and 17 we had another case of a safety valve where some sort of 18 pin or something was not put in, that was clearly 19 established.
20 MR. MARTIN: Before you go on, some of those did 21 not have pins. In fact, the way the nut was rotated there 22 was no hole through the nut and the staking; the nut was 23 turned over, and so the last time it was put back together 24 it was not put back in the same way it was before and was 25 not restaked.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-u>46
r i
l 25726.0 142 BRT r~n
() 1 But the failure of many of these valves -- you 2 see the piece extending down from the bonnet?
3 (Slide.)
4 Basically that nut and stud rammed against that, 5 and in some cases it looks like it shattered that l 6 attachment.
7 MR. DUNCIL: ,
I'd also point out on one of the 8 valves, for example, the washer underneath the nut was 9 missing. We did find evidence in a couple of cases where j 10 the stud that the nut was on had been badly damaged.
11 In another case the nut was still attached to 1
12 the stud; however, it was cold formed on the bottom where I
(')
v 13 the disc had vibrated against it.
14 So, the level of damage I think -- this kind of 15 addresses that -- will indicate whether that pin was 16 present or not may have not had a significant contribution 17 to the valve failing.
18 MR. SERKIZ: Mr. Chairman, in the investigative 19 team's report there are photographs. The 12-inch check 20 valves that were used upstream of the pumps, main feedwater 21 pumps, had no hole, no pin, and these are shown as an 22 example of one of them, in figure 6.61 and 6.60.
23 MR. EBERSOLE: The nut was just torque-turned 24 and that was it. Was it staked?
25 MR. SERKIZ: I personally inspected this one or O
\J ACE FEDERAI, REPORTERS, INC.
202-3 * ?W' Nationwide Coverage 800-336-6646
25726.0 143 BRT b)
\~# 1 was there --
2 MR. EBERSOLE: Was there an instructional 3 procedure to reassemble the pump and that was checked out?
4 MR. SERKIZ: I'm not aware of any. I want to 5 make the clarification, not all had pins or provisions.
6 The 12-inch main ones did not. Your others did have the 7 holes. Those that had failed, the discs were off in the 8 bottom we found -- at least on one aspect, we found neither 9 parts of the nuts or the pin. Maybe the utility can 10 comment whether the other parts were found at some time.
11 MR. EBERSOLE: Could you comment on that? About 12 the absence of evidence that they were, in fact, secured
() 13 and staked in place?
14 MR. DUNCIL: No.
15 MR. EBERSOLE: Thank you.
16 MR. DUNCIL: Basically the root cause of the 17 check valve failure was that the valves were oversized.
18 Ubat that means is that when -- at the time that the valves 19 were purchased, check valves were sized by pressure drop, 20 head loss through the valve.
21 Since 1976 the crane has been revised and one 22 now sizes check valves due to minimum velocity.
23 These valves were the same size as the pipe, and 24 the minimum velocity required for this check valve to be 25 fully opened is slightly above 100 percent flow velocity on O
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationside Coverage 800-336-6M6
'25726.0 144 BRT' O 1 the line. It's on the order of 19 feet per second. The 2 minimum velocity to keep the valve open is like 18.5 feet a 3 second.
4 MR. EBERSOLE: When you say there's a resonant 5 cycle, what is the amplitude and does-it include physical 6 impacting?
7 MR. DUNCIL: Let me get to that. ,
8 MR. EBERSOLE: Go ahead.
9 MR. DUNCIL: At 90 percent flow we have the 10 minimum velocity required to keep this valve fully open is 11 less than the flow velocity, so there is approximately a 12 0.2-inch gap between this nut and the nut stop here, such
(} 13 that the valve disc dangles in the flow stream.
14 At 90 percent power, 90 percent flow conditions 15 we have about a 12.5 percent fluid velocity reduction and a 16 24 percent reduction in the force lifting this check valve.
17 So there's about a 2 degree from full open. That's, like I 18 said, about 0.2 of an inch. It resulted in that resonance 19- of 1.3 hertz where that nut was actually impacting the stop 20 inside the valve body.
21 MR. EBERSOLE: Can't that be heard by some audio 22 detector put on the shell of the valve? Can't you hear 23 this impact?
24 MR. DUNCIL: I can't speak to that. I don't 25 know how audible that would be. With full flow conditions l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwid: Coverage 800-336-6646 )
_ _ _ , , , I
r.
25726.0 145 BRT m
(' ') 1 and in the location it is, I'm not sure.
2 MR. EBERSOLE: Maybe too much background noises?
3 MR. DUNCIL: That's a possibility. Given the 4 location. In addition, when we are talking about 0.2 inch 5 amplitude. However, because of the resonance and frequency, 6 we were getting about 1 million cycles every 12 days. So 7 in six months we are talking about 15 million impacts for 8 this nut on this nut stop.
9 MR. MARTIN: Very quickly, in June of 1985 it 10 was noted that there was e noise heard downstream of this 11 valve in the feedwater line. As a result they took 12 stethoscopes and tried to localize this noise in the B 13
(~}
v feedwater line.
14 It sounded, as it was coming from the valve 15 downstream of this check valve, and the -- their safety 16 analysis determined that certain things -- it could be a 17 loose check valve, could be loose internals in the stop 18 valve, but they made the assessment that even if it did 19 fail there would be no safety consequence because it 20 couldn't get to the steam generator.
21 All their monitoring at this time, they looked 22 at the other loops and did not detect this anomaly. So my 23 assumption is that if the failure was occurring during that 24 period of time they were not able to hear it. It looks 25 like it was this check valve disc actually laying in the (3
/
._)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nauonwide Coverage MK)-336-6646
25726.0 146 BRT
(~'\
U 1 bottom of the valve because there are indentations, 2 downstream, on the downstream side, that shows where major 3 pieces of metal have been chinked out as a result of this 4 valve hitting there.
5 MR. EBERSOLE: As a result of this stethoscope --
6 what you are saying is the valve had already come off.
7 MR. MARTIN: Probably had already come off.
8 MR. EBERSOLE: Prior to it coming off there was 9 no incentive to listen to this valve to see whether this 10 function was actually taking place and you were actually 11 beating it to death?
12 MR. MARTIN: In fact, the detection of this 13 noise in June was subjective. A number of people said they
(
14 couldn't hear it. Only after certain people said yes, I 15 hear something, they tried the stethoscope technique. They 16 also tried radiographic and couldn't get anything 17 additional.
18 MR. EBERSOLE: It's a very small movement. But 19 what I'm asking is this, this is a resonance performance 20 that was analytically known prior to this failure; am I 21 correct?
22 MR. MARTIN: No.
23 MR. EBERSOLE: It was not. So there was no real 24 incentive to listen, to hear if this impact was taking 25 place on the stop; is that correct?
O ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 i
. . ~ - -
25726.0 147 BRT 1 MR. MARTIN: Although they had had a previous 2 failure of this valve where they also found the disc 3 sitting in the bottom.
4 MR. EBERSOLE: Okay. Thank you.
5 MR. MICHELSON: The previous failure of the 6 valve you are talking about, how previous was it?
7 MR. SERKIZ: It's detailed in the team's report.
8 There's a section in there starting on page 6.13, it talks 9 about that, valve failure-related findings.
10 Prior to the upgrade to the seismic requirements, 11 there was a period of time that the utility did open up 12 these valves and actually inspect them during the fueling 13 outages.
( Early on, in the first several of them they found 14 there was wear and tear and replaced the pins and the discs 15 and so on. And that, again, as I say, without going into 16 this, is summarized in here.
17 I can't give you a reason on why, after a period 18 of time we shifted into another mode and relied totally on 19 IST. But we found in the maintenance history that they had 20 inspected them, they had replaced the innards. And for a 21 while at least -- two refueling outages -- became a way of 22 life arid then it sort of slipped into the background.
23 MR. MICHELSON: I guess it's correct to assume 24 that the normal flow velocity at least used to be on the 25 order of 20 feet per second; is that right?
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
,; . 125726.0 148 BRT 1 MR. DUNCIL: On that order. 18 to 20.
2 MR. MICHELSON: What is it now with the reduced 3 power requirements?
4 MR. DUNCIL: 12-1/2 percent reduction.
5 MR. MICHELSON: Still approaches 20 feet per t
6 second?
7 MR. DUNCIL: It was 18 to 20. It's still 17 to 8 18 feet per second.
,t
- 9 MR. MICHELSON
- The vendor, I gal.her, told you 10 about the criticality of flow?
11 MR. DUNCIL: As I pointed out, these valves were 12 oversized.
13 MR. MICHELSON: But how did you know you had a
-(
14 problem with your critical lower velocity?
{ 15 MR. DUNCIL: We found it out in our subsequent j 16 lower evaluations.
17 MR. MICHELSON: Did you do a test at which you
! 18 saw the valve fluttering and determined at what point it 19 ceased to flutter?
i 20 MR. DUNCIL: We took a look at the valve design.
21 Essentially what -- how they sized these valves are with l 22 minimum velocity. There are equations with coefficients to
! 23 determine what the minimum velocity is --
24 MR. MICHELPON: You can calculate this but you
! 25 can't easily calculate the effects of downstream O
V i
! ACE-FEDERAL REPORTERS, INC.
I Nationwide Coverage 800-336-6646 202-347-3700 ,.
- ,~ -
25726.0 149 BRT 1 fluctuations. The contribution of downstream turbulence 2 could far outweigh other contributions from normal 3 turbulent flow.
t 4 MR. CATTON: That depends on what the number is.
5 MR. MICHELSON: It-was running 20 feet per 6 second. I suspect it's quite high.
7 MR. CATTON: He'd have to have the throttling 8 valve pretty damn close to the check valve.
9 MR. MICHELSON: It is pretty damn close.
10 MR. CATTON: How close?
11 MR. MICHELSON: Three to four feet. That's
'12 close. That's a couple of pipe diameters -- well, three 13 pipe diameters away. And generally the rule of thumb is
( }.
14 it's got to go out at least five or six before you can 15 start to get rid of the turbulent flow. He can't calculate I
16 this. I think he ham to test it, and that's why I wondered 17 if he was actually watching it or just calculating it. I 18 don't think you can calculate it.
1 19 MR. CATTON: I don't think they know what the lift 20 coefficient is, either.
i 21 Does that nut bang up against the stop all the 22 time?
- j. 23 MR. DUNCIL: At 100 percent flow, this valve --
24 this nut should be -- should impact -- should be resting on 25 that stop since the minimum velocity required to keep the i
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 l
m 25726.0 150 BRT O 1. disc full open is lower than the stream velocity. At 90 2 percent power that's not the case. At 90 percent power, 3 it's going to be rapidly vibrating against that.
4 MR. CATTON: Doesn't that bang up the threads on 5 the nut? That seems to be poor design.
6 MR. EBERSOLE: What's the technique that tells 7 you that you are against the stop and steady on course?
8 There's no way to tell that external to the valve, is there?
9 Not by listening? Not by anything?
10 MR. DUNCIL: I can't speak to that. I'd say 11 probably not.
12 MR. MICHELSON: Every time you start up you are
(} 13 going go through a period of time where you have lower 14 velocities and you are going to be going through this mode.
15 MR. CATTON: It's just a flaky design.
16 MR. EBERSOLE: Some do a better anvil.
17 MR. CATTON: Those things get washed out because 18 they don't work.
19 MR. DUNCIL: As you do startup and shutdown 20 there will be some impact, but, again, we are talking about 21 very limited periods of time compared to normal full power 22 operation for 180 days or so.
23 I might point out, based on this evaluation, we 24 have concluded that the valve design needs to be replaced.
25 The next slide will depict what design we are O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationside Coverage 800-336-6M6
25726.0 151 BRT
- q
\ /
1 going to. As you recall, the primary failure mechanism, 2 the root cause of the failure was that the valves were over 3 sized.
4 The Atwood-Morrill design that we are look ing at-5 here is a one-piece disc assembly so there is not a hingo 6 pin or any kind of assembly that would come apart that 7 would cause th' disc to separate from the t ,ge. In 8 addition to tha , the valves are going to be sized such 9 that they will be full open at down to 70 percent flow.
10 MR. EBERSOLE: Before you leave that, are you 11 going to have that little lip at the bottom? Are you going 12 to validate that these can intercept a full reverse flow f'N x-)
13 from a hypothetical reverse flow upstream or would it just 14 peel that lip off?
15 MR. DUNCIL: I can't speak to that.
16 MR. EBERSOLE: Somebody has to write the specs.
17 MR. MEDFORD: Jesse, we don't know what the spec 18 says on this valve, but one of the things we want to take 19 away from this meeting is to check and see what the dynamic 20 requirement is.
21 MR. EBERSOLE: That's all.
22 MR. REED: I don't think that drawing is very 23 representative of the way the disc --
24 MR. EBERSOLE: I'm looking at it in profile. I 25 understand.
O V
ACE FEDERAL REPORTERS, INC. i 202-347-3700 Nationwide Coverage Mk336-6646
l l
l l
25726.0 152 i BRT ;
e
\~'
1 (Slide.)
2 MR. DUNCIL: As a result of our evaluations, the 3 following modifications are intended. We want to replace 4 the eight Pacific check valves with Atwood-Morrill designs, 5 the two 12-inches, the three 10-inch check valves and the 6 four-inch check valves.
7 We are going to provide in-service test 8 capability for the replacement valves.
9 We are going to slope the B feedwater pipe 10 inside containment. Basically what this will do is make 11 the B feedwater line improve its -- should I say decrease 12 its susceptibility to water hammer event to the level that (3
~. )
13 ANC are susceptible.
14 MR. MICHELSON: Which level are you sloping it?
15 MR. DUNCIL: I believe the direction is toward 16 the generator.
17 MR. MEDFORD: Away from the steam generator.
18 MR. DUNCIL: We have the valves with IST 19 provision in each feedwater line inside containment.
20 MR. EBERSOLE: Are these the ones close to the 21 steam generators?
22 MR. DUNCIL: Yes.
23 MR. EBERSOLE: Does that mean you are obligated 24 now to take long aux feed lines and.go upstream of those?
25 MR. BASKIN: That's something we have to look at ACE FEDERAL REPORTERS, INC.
201-A7-3700 Nationwide Coverage 800 336-6M6
i i
25726.0 153 BRT
/^)
1 further.
2 MR. REED: In making that statement, in saying 3 you probably should put the aux feed lines downstream of 4 the closeup -- on the main steam line -- that's the way I 5 have seen it done quite successfully. You'll have to look 6 at whether you can leave them outside containment and 7 injecting into the common line, whether or not that gives 8 you the same protection against water hammers.
9 MR. DUNCIL: We were going to provide for 10 automatic closure of the main feedwater following unit trip 11 and loss of NFW pumps. That would not have precluded water 12 hammer but it would have allowed the lines to fill more f')
v 13 quickly.
14 We are going to evaluate the addition of 15 void-sensing device with a control room alarm. What that's 16 going to do basically is give us an early indication of 17 void formation in the line.
18 MR. MICHELSON: What kind of device are you 19 going to use for that?
21 MR. MICHELSON: For void formation in the pipe?
22 Is that what it is?
23 MR. DUNCIL: Steam.
24 MR. CATTON: Actually not an RTD.
25 MR. MICHELSON: What are you going to do with ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
J 25726.0 154 BRT
. 1 temperature --
2 MR. CATTON: It measures heat transfer between 3 heat and water.
1 4 MR. MICHELSON: Okay, he's going to use a heated 5 therm -- he's going to use a thermocouple and can tell
+
6 whether the top part of the pipe is voided? That could be l 7 potentially helpful. I'm not sure. He's also pitching his 8 pipe away from the steam generator now --
- 9 MR. CATTON
- Do the pipes come in horizontal 10 right to the steam generator?
11 MR. MICHELSON: No. They come up.
- 12 MR. CATTON
- So if they put the sensor in that
() 13 line --
14 MR. MICHELSON: Near the steam generator. That 2 15 will be helpful.
16 MR. CATTON: Where are you going to locate these t
17 sensors?
18 MR. DUNCIL: That's still under evaluation.
19 Both the type and location are under evaluation.
20 MR. RAINSBERRY: It's fair to say we don't have 21 a design in place so we can say something is going to be 22 done now.
23 MR. REED: I'd be more interested in putting my 24 money in prevention rather than monitoring in this case.
25 Because water hammer is a very sudden thing and fleeting O
ACE. FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
-. . - . , . = .- - . - - ~ . - - . - .-
N d
25726.0 155
- BRT.
j-1 thing and transient thing and I don't know that you could 2 get anything -- if you got anything that could reliably i 3 tell you it happened you could do anything about it.
i
! 4 MR. CATTON: You monitor whether or not to see 5 if you have voids.
6 MR. REED: You can have voids like that --
{
7 (snapping fingers) -- on a feed line break? L 8 MR. CATTON: Oh, feed line break, that's a
- 9 different story. But here they are concerned about a 10 draindown through the lines that will take place slowly.
11 And then just want to know not to refill them.
12 MR. REED: How many sensors are you going to put 4
1
() 13 on and where?
14 MR. CATTON: I don't know, but I suspect it 15 wouldn't be too difficult to figure out. I'd put one right b 16 up near the steam generator, and if it shows vapor in that 17 vertical run, I would assume I would have a water hammer l 18 unless I did something to prevent it.
1 j 19 MR. REED: I'd put my money in good design.
l 20 MR. CATTON: I would not refill that tube other
- 21 than very, very carefully.
1 22 MR. EBERSOLE: Go ahead.
{
y 23 (Slide.)
24 f!R. DUNCIL: The next slido depicts the general 25 modifications that we are in the process of pursuing; we i
ACE-FEDERAL REPORTERS, INC. !
202-347 3700 Nationwide Coverage 800-336-6646
--,, ,-.--~.- ,, - ..--, _ . - . ,., - -,.- , _ - , .,..-.- - _ ,.~ - .. _ --. _ ,.. - _ _, . _,--. ~ ,-- - - -. -
25726.0 156 BRT
- O
\'.) 1 are going to install control room indication of blowdown 2 status. I might also point out -- the asterisk and double 3 asterisk are on the next page, but basically the double 4 asterisks are things we are doing during the current outage; 5 double asterisks are things that don't have a schedule 6 associated with them yet.
7 He are going to provide all the blowdown 8 isolation on AFU actuation.
9 He are evaluating the possibility of providing 10 for automatic loading of the diesel generators on loss of 11 power. Again, a schedule hasn't been identified for that 12 proposed modification. We are going to evaluate
() 13 alternatives for providing an immediately available backup 14 source of off-site power.
15 He are going to insta11 additional protection 16 for the auxiliary transformer feeder cabling. Again, that 17 goes back to the prevention of water damage and mechanical 18 damage to the cabling that could lead to grounds.
19 Modifications to preclude spurious actuation of 20 safety injection will be implemented.
21 We are going to modify the tech support computer 22 for automatic restart following restoration.
23 MR. EBERSOLE: Let me ask you about that. Arc 24 you going to augment and improve power supply to the 25 computers so you don't experience an outage?
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cmerage 800 33M646
25726.0 157 BRT
('h 1 MR. DUNCIL: That's where this is; it's for both 2 the title tech support computer as well as vital bus 4.
3 MR. EBERSOLE: That's right. Otherwise you 4 shouldn't call it a vital bus.
5 (Slide.)
6 MR. DUNCIL: We are going to determine and 7 correct the cause of the spurious red-phone ringing; going 8 to install a control room clock not dependent on AC power; 9 and we are going to review the control room indication of 10 vital bus status as part of our design review process.
11 We are going to modify the current limiting 12 bypass reactor breaker design such that the diesel
() 13 generator breakers can be closed. This was something that 14 was identified during the event. It wasn't involved 15 directly with the event, but it was something we identified 16 in the course of our evaluations.
17 MR. EBERSOLE: What about the matter of having 18 to manually close main-stream isolation valves? Are you 19 going to look into the advantage or need for having that 20 remote manual?
21 MR. DUNCIL: I know it has been looked at.
22 MR. EBERSOLE: I recall the fellow that had to 23 run when he was down there, facing the closure of it.
24 Right now it's driven by an air wrench, isn't it?
25 MR. BASKIN: It's, in effect, a manual valve.
O ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-MM
25726.0 158 BRT p) 1 MR. EBERSOLE: He has a socket wrench of some 2 sort driven by air and he has to go up and hold it against 3 it and it doesn't sound like a monumental X-million-dollar 4 expenditure to that wrench from someplr.ee besides right 5 there.
6 MR. BASKIN: What it amounts to as soon as we --
7 we put some sort of motor operator on the valve.
8 MR. EBERSOLE: Right. Analogous to the one he 9 holds in his hand.
10 MR. BASKIN: Probably a little more expensive 11 than the one he holds in his hand. We have not discussed 12 that. It's on our list and there's lots of things we are
(~S 13 considering. We haven't discussed it, but it's certainly w) 14 not in the area of something we decided not to do, either.
15 MR. EBERSOLE: This is one of the very few 16 plants that has no ability to isolate steam flow from a 17 common agent path and thus faces, well, the loss of steam 18 supply. One of the only two aux feed pumps you have.
19 MR. BASKIN: That's correct. That's correct.
20 MR. EBERSOLE: That can occur very quickly and 21 that leaves you on one pump, doesn't it?
22 MR. BASKIN: After this outage we will have 23 another pump, so in effect we have two --
l 24 MR. REED: In your review of that I asked the 25 question: Did you follow the power boiler code. You said ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3346M6
25726.0 159 BRT
/~N, I
' 'i 1 yes, in California they have it. I have to think that 2 maybe California was different from Massachusetts at the 3 time. Because at the minimum you would have to have a stop 4 check there in the steam line because you have multiple 5 steam lines, and therefore, for industrial safety and 6 maintenance and operations, you have to have a stop check, 7 a nonreturn.
8 I think you should look closely at whether or 9 not you, perhaps, shouldn't go a little further in your 10 thinking and having nonreturn valves at each of those 11 individual steam generator lines as well as a stop valve.
12 The trend, of course, has gotten even further than that.
13 Now they have a quick-closing outgoing valve and a i
14 nonreturn in each line. You better look a little more 15 closely.
16 MR. EBERSOLE: How does this plant stand it when 17 it experiences -- I forget which one -- the massive steam 18 tube generators, where it's very nice to be able to isolate 19 the bad steam generators and operate on the others. When 20 you have multiple steam tube generator failures on one, 21 evidently they feed to the secondaries on all sides; 22 there's no way to isolate them; right?
23 MR. MEDFORD: That's correct.
24 MR. EBERSOLE: How do you get out of that? The 25 other plants have a neat way to do that, they shut off that ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
25726.0 160 BRT
(
2 I presume this has been analyzed for the worst 3 case of steam generator tube failure and the result to the 4 public?
5 MR. RAINSBERRY: It has been, yes.
6 MR. EBERSOLE: That was done in the presence of 7 no isolation on the secondary? And it came out all right?
8 Well, it must have.
9 MR. RAINSBERRY: That was reanalyzed as part of 10 SAT, wasn't it? It was reanalyzed since the 1980s time 11 frame.
12 MR. EBERSOLE: Was it reanalyzed for the
() 13 presence of AC power where you don't have the condenser 14 into which to feed the primary loop?
15 MR. RAINSBERRY: Yes. All our accidents have to 16 consider loss of off-site power.
17 MR. EBERSOLE: So you dump the primary loop into 18 outer space through the dump valves?
19 MR. RAINSBERRY: That particular procedure has 20 the cooldown trying the get the pressure equalized on the 21 two sides as fast as you can.
22 MR. EBERSOLE: Has to be on all steam generators?
23 MR. RAINSBERRY: They are all connected together.
24 MR. EBERSOLE: All right. Go ahead.
25 MR. MICHELSON: Do you use a feed and bleed O
l i
ACE-FEDERAL REPORTERS, INC. l 202 347 3700 Nationwide Coverage 8(r).336-(646
25726.0 161 BRT
/"N V 1 system to replenish primary-side water?
2 MR. RAINSBERRY: Safety-injection system and 3 then charging is used. The secondary side is used to -- in 4 the early stages.
5 MR. MICHELSON: But in the steam tube rupture 6 cases you eventually have to cease putting primary side 7 water in?
8 MR. RAINSBERRY: Correct.
9 MR. MICHELSON: Hopefully you'd isolate the 10 ruptured generatet. but in this case, you may not be able 11 to isolate the ruptured generator.
12 MR. RAINSBERRY: You have the capability, with
(~}
R./
13 the safety-injection system, in combination with charging, 14 to make up however long you need to based on the fact that 15 you can isolate that generator. And then you cool down 16 using the secondary side.
17 MR. EBERSOLE: Just discharging into atmosphere?
18 MR. RAINSBERRY: Right. And then the doses have 19 been evaluated.
20 MR. EBERSOLE: Okay. Go ahead.
21 (Slide.)
22 MR. DUNCIL: The conclusion is the most 23 significant aspect of this event is the failure of five 24 feedwater check valves. The check valve failure was a 25 result of inadequate design and that is the root cause of O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6M6 l
25726.0 162 BRT O 1 the valve failure.
2 Implementing the modifications we outlined --
3 MR. REED: I like to have operating companies 4 blame designers for inadequate design. They haven't dcne 5 it enough and the NRC hasn't done it enough, either. But 6 in what I've heard here I think, sure, it was inadequate 7 and lousy design. But I think that the maintenance folks 8 here could have been a little bit more alert to the fact 9 that they had lousir design, based on what I hear about the 10 maintenance histori and these pins and the way these things 11 work.
12 The way you find most of the time inadequate 13 design is for the maintenance folks and the operating folks
(}
14 to see that it's no damn good and do something about it.
15 f1R. EBERSOLE: That's a common problem.
16 Operating people tend to operate in their own sphere of 17 activity and just fight the plant and be proud of the fact 18 that they can run a booby trap, if they can do it.
19 Do you have some sort of backflow from operating 20 and maintenance people into the design defect or shortfall 21 area so you really get, you know, a documented, stated 22 requirement that they examine each failure in the context 23 of whether they got a good design or not? It's not a 24 common thing to have that.
25 MR. BASKIN: he don't have a routine backflow of ACE-FEDERAL REPORTERS, INC.
202-347 37(X) Nationwide Coveras- 8(Xb33MM6
25726.0 163 BRT O
J 1 information. We do have systems through our nuclear safety 2 group, and our so-called ISAG get into some of these things.
3 And it's, you know, by definition the more important things 4 are where somebody has to define what's important and what 5 isn't. In this case, it wasn't -- back in the '70s and 6 carly '80s, we didn't have that when we had the first check 7 valve failure.
8 But I think it's important to recognize here 9 that what Bruce said was correct. The root cause of the 10 water hammer was the failure of the check valves. The 11 check valves were poorly designed, and I agree 12 wholeheartedly with what Glen said, that we had numerous
(} 13 opportunities to pick it up and didn't.
14 That's different than us describing the root 15 cause of the event, and I want to make that distinction 16 because I don't think we were at that point yet.
17 MR. EBERSOLE: At some point in the maintenance 18 process you think the check valve design should have been 19 challenged and it was not?
20 f1R. BASKIN: I'm sorry?
21 MR. EBERSOLE: At some point in the main.enance 22 and operation of the plant the check valve design should 23 have been challenged and it was not?
24 f1R. 13ASKIN : Sitting here today it's easy to 25 agree to that statement. The question is why wasn't it?
U(~T ACE-FEDERAL REPORTERS, INC.
202 347 37(X) Nationwide Coverage 8(K)-33MM6
25726.0 164 BRT i )
1 And prevent such things in the future.
2 MR. EBERSOLE: Isn't this part of the mindset on 3 the part of operations people, and maintenance people, that 4 "we are going to run this booby trap until hell freezes 5 over and that's the pride of our occupation."
6 MR. KERR: There's also the maxim that says 7 hindsight is a lot better than foresight.
8 MR. EBERSOLE: Sure. You have to wait until it 9 really happens.
10 MR. REED: I have always found a healthy 11 attitude that the operators and maintenance people always 12 want to find the design wrong, absolutely no good and
() 13 that's why they have to do some work here today. And, on 14 the other hand, the designer wants to say that he is sacred 15 and never did anything wrong. There's a healthy 16 competition but there isn't any referee yet with a big club.
17 MR. MICHELSON: Could you clarify for me what 18 check valve failures were found by inspection or other 19 reasonable means prior to this event, other than the one 20 back in '74 or '76 or whenever? Here there other cases 21 where you had taken check valves apart or other similar 22 manufacture and found the damage and just fixed it and put 23 it back in?
24 MR. DUNCIL: I think based on what Mr. Martin 25 had discussed earlier, one of the early maintenar.co records O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 33MM6
25726.0 165 BRT O 1 in '75 indicated that the valve was opened and the entry 2 was basically just repaired.
3 MR. f1ICHELSON: That one we've heard about. Are 4 there any others? Or is that the only time you ever saw 5 damage to these check valves until this event?
6 MR. MEDFORD: That's the most significant, but 7 there are --
8 MR. MICHELSON: Wait a minute. Don't tell me 9 most significant. Were there --
10 f1R. KERR: Let him finish.
11 MR. MEDFORD: There were others.
12 MR. MICHELSON: Now that the event has occurred
[')
V 13 you must have gone back and checked all your records to see 14 what has happened. How many lesser significant 15 observations were there?
16 MR. MEDFORD: I haven't seen that quantified.
17 MR. MARTIN: I might respond. There was a l
18 period there where almost overy refueling outage they went 19 in and refurbished the valves and in many cases replaced 20 the disc. You wouldn't go through the process of replacing 21 the disc unless you found some major damage that simple 22 grinding would have fixed.
l 23 Our licensee is faced with the same problem.
24 The records don't tell you much more than " replaced the 25 disc." Doesn't tell you why or what the problem is, and O
\_/
i ace-FEDERAL REPORTERS, INC.
202-347 3700 Nation aide Coverage 80fk3364M6 L r
l 25726.0 166 BRT O
d 1 there's nobody around that seems to remember those details.
2 MR. MICHELSON: But there must have been a 3 problem or they wouldn't have replaced the dise?
4 MR. MARTIN: Every review and outage.
5 MR. MICHELSON: This is not an isolated 6 occurrence in '76, but rather there was a lesser degree of 7 problem. That's what I was trying to get a feel for.
8 Every outage?
9 MR. MARTIN: There was a period there, from the 10 information we had, from about '76 to '81, I guess it is, 11 every refueling outage we found some information that said 12 discs were replaced or nuts were replaced. Things like
() 13 that.
14 MR. MICHELSON: Once a year then?
15 MR. RAINSBERRY: 15-n onth cycles .
16 MR. BASKIN: Two, possibly three refuelings.
17 MR. MICHELSON: So it would not have been a 18 total surprise to have found this happen. At least it was 19 happening to a lesser extent in the past?
20 MR. BASKIN: Yes. I think in hindsight I don't 21 know that replacing the disc indicates necessarily severo 22 distress to the valve. You know?
23 MR. MICHELSON: It indicates enough that you 24 went to the expense of replacing it.
25 MR. BASKIN: And it certainly indicates it ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336- % :6
25726.0 167 BRT O
1 doesn't seat tight enough to pass the tests.
2 MR. EBERSOLE: Ever if you found they were 3 working fine that by no means says in the aspect of an 4 upstream pipe break that they would have intercepted flow.
5 They might have crashed right through even though they had 6 been operating up to that time with no problem at all.
7 I'm saying long-time experience with valves is 8 no proof that when called upon in duress to do what they 9 have to do they will do it at all; do you follow me?
10 MR. BASKIN: Yes.
11 MR. EBERSOLE: There has to be in the specs or 12 analysis or unique one-time test or something, and it's not
(} 13 something you develop over X years, 20 years, whatever --
14 it stands in the background and waits for it to happen.
15 MR. MICHELSON: Hell, there was at least a 16 one-time test of a severe water hammer against the disc and 17 it didn't integrate, but that's about the worst case I 18 guess they've ever ever seen. A pretty good case.
19 MR. DUNCIL: Relative to that conclusion, I 20 would like to point out in the past years we have operated 21 at 90 percent power. The valves have seen more damage than j 22 would normally be seen through several cycles. There 23 really isn't any criteria that I am aware of as to what the 24 requirements are replacing the disc. What we said earlier, i
25 following on what Tim was saying, the maintenance records O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationside Coverage M10-336-fM6
. . . . - _ _ , . . - . ..- - ._ , - , - , . _ . , _ - _ , ~ _ ,
25726.0 168 BRT 1 were in existence, 10 or 15 years ago, nowhere near come to 2 the capability of being able to predict detailed failure 3 mechanisms in retrospect.
4 MR. EBERSOLE: This wear you are talking about, 5 I can only see two wear patterns. It would be on the pivot 6 shaft where it works or impacting on the nut and its bolt.
7 That's the only wear there is, isn't it?
8 '
ftR. DUNCIL: As far as I'm concerned.
9 MR. EBERSOLE: Of those you don't mention any 10 wear on the pivot pin at all.
11 f tR. DUNCIL: As I recall, there was very little 12 wear on the pin.
(} 13 MR. EBERSOLE: So it focuses on the impact on 14 the stop. That's the only wear there is.
15 MR. DUNCIL: Essentially.
16 MR. MICHELSON: On your check control valves, 17 since you indicated it was oversized for various reasons, 18 was the flow control valve also oversized so they had to do 19 a considerable amount of throttling with it even at 100 20 percent power? The only 100 percent power?
21 f1R . DUNCIL: I can't speak to that.
22 MR. MICHELSON: In other words, unfortunately 23 you never supplied us a drawing of the trim internals of 24 that flow control valve, which I think would be quite 25 important to have.
l CE) l l
ACE-FEDERAL REPORTERS, INC.
I 202 347 3700 Nationwide Coverage 800-336-M:6 s
l h
t 25726.0 169 BRT 1 MR. CATTON: That would tell you about the 2 downstream turbulence --
3 MR. MICHELSON: Ilow it might have compared if 4 they reduced it another 10 percent full. How much they 5 were throttling. I don't even know what kind of a valve it 6 is in terms of the trim.
7 You don't really know the kind of trim in that 8 flow control valve, do you?
9 MR. MEDFORD: No.
10 MR. MICHELSON: Do you have an outline drawing 11 of the internals?
12 MR. MEDFORD: No, we don't.
() 13 MR. MICilELSON: Could you arrange to send one to 14 our federal representative, an outline of just that flow 15 control valve? l l
16 MR. MEDFORD: Certainly. l 1
17 MR. EBERSOLE: In inspecting the valve do you l 18 usually find any degree of turbulence upstream of the valve 19 in addition to the mass flow rate?
20 MR. MEDFORD: No.
21 MR. EBERSOLE: Just a statement about so many 22 feet of pipe and that doesn't say what you are doing 23 upstream of that point?
24 MR. MEDFORD: That's correct.
25 MR. EBERSOLE: So the specs tend to be a little O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage Mk3346646
7 25726.0 170 BRT f's
- 1 on the loose side in this context. Well, they always have 2 been.
3 MR. MICHELSON: There are certain rules of thumb 4 you use in designing to keep your flow control valve far 5 enough removed from the check valve. Those rules of thumb 6 are well known by designers but not always abided by.
7 MR. CATTON: You have to worry about how far it 8 is from a bond, too. Because a bend creates a secondary 9 flow that will cause an oscillation.
10 MR. EBERSOLE: Let's go. Carry on.
11 (Slide.)
12 MR. DUNCIL: Our last conclusion: implementing 13 the modifications outlined above will eliminate the
(~~J) 14 identified check valve failure mechanism and in general 15 improve plant performance.
16 That concludes my presentation.
l' MR. MICHELSON: The only comment on your last 18 item I would have is that I think you really haven't yet 19 addressed the reason why the valve, the previous valve was 20 failing. And it may be that that reason still exists and 21 that your new valve will not necessarily fare much better, 22 although it might be designed to take the beating a little 23 better.
24 MR. DUNCIL: Again, we believe that the failure 25 mechanism was based on, first of all, the valve being O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 171 BRT O 1 oversized such that it was not open in the flow steam fully.
2 The new valve design will allow the disc to be fully opened 3 down to at least 70 percent flow.
4 In addition to that, we believe the failure 5 mechanism of the valve, because of the impacts, was due to 6 the multi-piece construction of the hinge and disc assembly.
7 With the new valve design we have a one-piece assembly and 8 it will not be subject to the same mechanism.
9 MR. MICHELSON: My comment relative to your 70 10 percent flow is that depends upon whether you have well 11 established flow or whether you have a highly turbulent 12 condition out of your flow control valve. If you have a
() 13 highly turbulent condition it may still not be stable and 14 you may find a surprise.
15 MR. MEDFORD: We share your concern about the
- 16 flow regime in that area. That's one of the reasons why wo 17 are adding the check valves.
18 MR. MICHELSON: I don't know how that helps you 19 much, though. How do you figure that helps you? It's a 20 little more impedence in the system, but --
21 MR. DUNCIL: The check valves inside containment 22 would preclude, should those valves ever fail again, it 23 would be a backup. It doesn't reduce the turbulence in the 24 line.
25 MR. CATTON: It's quite some distance, though --
O s_/
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6M6
_ - - - - - 1
I 25726.0 172 BRT O 1 f1R. MICHELSON: It's about three, four feet.
2 He's going to have the same check valve right next to i. t .
3 MR. CATTON: lle's going to move the set inside?
4 MR. MICilELSON: lle's going to put an additional 5 one inside was my understanding.
6 MR. RAINSBERRY: Yes.
7 MR. MICHELSON: And replace the one right next 8 to the flow control valvo, and I just wondered why the 9 world changed already?
10 MR. RAINSBERRY: He are also adding a provision 11 for another flow control valve.
12 MR. EBERSOLE: That it? Thank you.
() 13 14 The remainder of today, we are of course out of time already, just a general discussion and summary 15 conclusions and future ACRS actions, and I have discussion 16 here between what was thought to be said tomorrow -- it's 17 an hour and a half -- at the end of which where we have a 18 summary of the review, NRC utility reaction and 19 recommendation for specific plana, specific and generic 20 actions, summary conclusions -- I think this is premature 21 at this time to say this.
22 What we need to identify now is to what degree 23 we can get simply support for informational discussion 24 tomorrow, late in the afternoon to the full committee?
25 Just to tell us where we are in this investigative effort O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
25726.0 173 BRT
/~T C 1 and what we can say at this time with no inference that we 2 are by any means done.
3 MR. KERR: Do we need an informational meeting?
4 MR.
- EERSOLE: I think the committee members are 5 much interested in what's going on. I wouldn't spend an 6 hour6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and a half on it Bill, but I would like to spend some 7 time. To what extent can we get some support for an 8 abbreviated statement of where we stand in this 9 investigation?
10 MR. MEDFORD: A problem is we just learned of 11 the meeting this week.
12 MR. KERR: They say we don't know quite why this
() 13 happened and what we are doing about it --
14 MR. EBERSOLE: It's premature? That's one 15 conclusion we could draw. I have no way of feeling -- gee, 16 let's have a discussion.
17 MR. MICHELSON: The committee may want to hear 18 about the event in general, but the Staff could make such a 19 presentation. Nothing controversial, I don't believe, 20 about what happer,d. But rather some of these other 21 requirements.
22 MR. EBERSOLE: That's all I'm after, a news 23 report of where we are and what went on.
24 MR. MICHELSON: But no conclusion.
25 MR. EBERSOLE: No conclusion reaching whatever.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
. . . .=. _ . -.._ . . . _ _ -. . . -_ _- . - . _ . ._ .. -
25726.0 174
, BRT
() .
1 We are still in the middle of this and maybe a projection 2 of where we will be when. That's about all I can say.
j' 3 They have had the report already and they are ;
4 j 4 certainly interested in it.
i 5 MR. REED: I think if the Staff is going to make
- 6 a less than one-hour presentation to the full committee, 7 they should try to get away from all these extraneous 8 things, 18 or 20 issues.
9 I sort of boiled it down real quickly to only i.
10 about three or four significant aspects with respect to 11 this event. One is the design thing. I see another that 12' there wasn't enough alertness on the part of the ;
13 maintenance people, the operators -- get it down to that.
(
- 14 I think a lot of this other electrical stuff doesn't go i
15 along the major path.
- 16 MR. EBERSOLE
- It doesn't get to the core by any
- 17 means. We could intercept the preliminary.
18 MR. HERNAN: Jesse, Ron Hernan with the. Staff.
{
19 We have been discussing with the ACRS Staff for 20 at least 10-days to two weeks, full committee. And we, 4
21 because of scheduling conflicts and so forth, are not in a 22 position to support a presentation at the full committee 23 meeting. It's my understanding that you were going to i
- 24 summarize. I t
25 MR.'EBERSOLE: Certainly we could summarize O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
25726.0 175 BRT 1 ourselves.
2 MR. HERNAN: We also had not invited the 3 Licensee to attend the full committee meeting.
4 MR. MICHELSON: Had not?
5 MR. HELTMAS: Had not. I guess our feeling, if 6 the team was unable to be there to make a presentation it 7 would not be appropriate to have the Licensee.
8 MR. SAVIO: You would have some scheduling 9 inf orma tion?
10 MR. HELTMAS: Yes. We can discuss the overall 11 schedule.
12 I did want to mention, because I don't think I
() 13 heard it mentioned this morning, Mr. Stello has issued a 14 letter as of February 4th, which I believe you have copies 15 of, to Mr. Dent, Mr. Taylor and Mr. Martin, which lays out 16 fairly specifically what the issues are the Staff is going 17 to work on and to some extent, a schedule, although he's 18 asked these people to respond with their detailed schedule 19 of the action items scheduled. Also with the idea that the 20 actions -- that they would report back in three to six 21 months.
22 But I can discuss with the full committee where 23 we are going from here.
24 MR. EBERSOLE: The general aspects of where we 25 are.
.o O
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationside Co ferage 800-336-6646
. . . ~ _ .. . . . _ . . _ ._
25726.0 176
! BRT
(:)
1 MR. HELTMAS: And what we conclude, what the 2 team concluded were the main issues.
f 3 MR. EBERSOLE: We can reduce this down to a nalf
} 4 hour or something and go that way with just you and
! 5 ourselves to comment on what we heard. And it certainly l 6 focuses down to the check valves in the final analysis.
f, 7 MR. MICHELSON: To be sure I understood what you 8 said, Ron, are you saying it will be three to six months !
] 9 before the plant starts up again?
10 MR. HELTMAS: No. No. Absolutely not.
11 MR. MICHELSON: Are they on any kind of hold ,
4 12 presently? What does the Licensee have to do before they l
(} 13 24 are allowed to start up again?
MR. HELTMAS: I think I'd rather have Mr. Dudley 15 address that.
16 MR. DUDLEY: I have one additional slide.
17 MR. MICHELSON: Okay. I thought that we were 18 going to hear all of that in the full committee and 19 therefore I didn't raise the question.
1 20 MR. DUDLEY: This slide is just an elaboration l
21 of the actions in particular the Staff is performing in 22 order to respond to the memo giving us direction.
23 (Slide.)
24 Region 5 has been given the lead for facility 25 restart for San Onofre. NRR the be providing information ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
25726.0 177 3RT O 1 to Region 5 and NRR will be providing the review of the 2 technical design questions that are mentioned and come up 3 in the EDO memo and the IIT. report.
4 INE was given the generic lead regarding the 5 possibility or probability of generic problems with check 6 valves, and INE is also going to look at the generic 7 adequacy of the in-service testing program and our 8 requirements related to check valve problems.
9 NRR to also assess the need of whether it is 10 required to reevaluate the water hammer USI, to 11 specifically look at the time of condensation-induced water 12 hammers, this type of water hammer that occurred at San 13 Onofre.
({
14 MR. EBERSOLE: Is there an act program of some 15 sort looking at other plants that might be on the 16 borderline of this same condition? Old plants?
17 MR. DUDLEY: Well, INE -- well, let's see. On 18 the water hammer portion or the check valve?
19 MR. EBERSOLE: With relation to check valves 20 which invites water hammer.
21 MR. DUDLEY: A1, can you speak to that?
22 MR. SERKIZ: To pick up the point that was just 23 made,_I think the check valves are the key generic question.
24 This is the first time that I am aware of in this country, 25 or anywhere, had such a gross failure of check valves. One O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-6646
25726.0 178 BRT O 1 of the points made by the IIT I think is correctly made, 2 resolution of weter hammer USI A-1 was based fundamentally 3 on looking at where water hammers had occurred. If we have 4 a situation that is as bad as this check valve failure with 5 10, then I think that is the issue to be looked at and 6 cured. We will -- and I don't have the exact plans right 7 now because one of the items that came back in a Stello 8 memorandum was to look and see if we want to reopen the 9 safety issue.
10 MR. EBERSOLE: You mean the water hammer?
11 MR. SERKIZ: Water hammer as a safety issue. We 12 did look at condensation-induced water hammer. We do know -- l
(} 13 this is well known -- this type of water hammer will give 14 you the largest structural loadings. To reopen the issue 15 on the basis of steam condensation water hammer, I'm giving 16 you my personal view, I don't think this is the way to go.
17 I think we want to work in concert with the people that are 18 looking at the check valve failure and see if that, in 19 conjunction with the fact in a check valve failure or 20 leakage would lead to this situation, that would be a 21 better way of looking at whether we want to reopen the 22 water hammer issue.
23 MR. EBERSOLE: Is there contemplated some sort 24 of issuance to other owner / operators about this matter here 25 and a requirement that they review their own check valve O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6 ,
l
~ -=. .. - .-- .- .-
1 l
l 25726.0 179 4
BRT 1 operational practice?
2 MR. SERKIZ: I.believe INE has already sent out f
3 an information notice.
4 MR. DUDLEY: Right. Early January.
- 5 MR. EBERSOLE: No requirements specific action 6 be taken or any backflow?
7 MR. DUDLEY: Information notice.
8 MR. EBERSOLE: When did that go out?
9 MR. DUDLEY: January 6, something like that.
10 MR. MARTIN: That's correct.
! 11 MR. EBERSOLE: Ron, we can report that to the 12 committee, the essence of this slide here.
(} 13 MR. MICHELSON: You might want to add that there 4 14 have been several water hammers in the HPSI return line 15 feedwater on boiling water reactors in the last two years, 16 some of which have been dislocated pipes and so forth, and
-17 they are caused by partial voiding of the feedwater line 4
18 and then subsequent injection of feedwater and then water i 19 hammer and that has led to damage of pipe but not check i
20 valve, as far as I know. But it was caused by leaky check 3
21 valves, so you might want to look at HPSI feedwater return.
22 MR. SERKIZ: Just to answer that very directly, 23 I think we are going to raise the question again on 24 reopening. We'd want to look at the occurrences, water i
25 hammer occurrences that have been reported over the last C:)
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
t i
25726.0 180 BRT 1 three years and derive on that an insight on whether we 2 should reopen the issue.
3 All I was trying to say was in this case where 4 you had such massive voiding of feedwater lines, it was a 5 very gross type of check valve failure, first time, first 6 experience.
7 MR. EBERSOLE: Go to the front end -- okay.
8 Thank you.
9 MR. DUDLEY: I know you probably want to finish.
. 10 The NRC will meet with the utility on February 19 at the 11 Region 5 office to discuss items required for restart. The 12 utility will meet with the Commission sometime in March l
(} 13 14 1986.
The originally scheduled restart date prior to 15 the water hammer was May 20. Due to the extensive 4
16 modification I don't know what the current restart date is 17 or if that is not the current restart date.
18 MR. MICHELSON: One observation, your INE is i 19 going to look at the check valve problem. Traditionally 20 the valve investigative expertise hasn't been in INE, but 21 rather in NRR or research, where the work was being done.
22 Is there some rationale for why INE is looking at this 23 problem instead of NRR? ,
24 MR. DUDLEY: I can't say why EDO made that 25 assignment but I know they are aware that EDO has had t
O ACE-FEDERAL REPORTERS, INC.
j 202 347 3700 Nationwide Coverage 800 336-6646
25726.0 181 BRT 1 experience.
2 MR. MICHELSON: If you want to go back and look 3 at past operating experience and so forth, generally AEOB 4 has been doing that and not INE. They only looked at the 5 short-term, next-30-day kind of things. Is there some 6 racionale again for the selection of INE to do this job?
7 MR. DUDLEY: I can't speak for Mr. Stello.
8 MR. MICHELSON: I wouldn't think you'd want to.
9 MR. BASKIN: If I could add two quick points.
10 Number one, recognizing it's premature and things may 11 change, but as of this moment in time we hope and expect to 12 maintain in a May 20 restart date. Again, that may change I)
v 13 based on other things.
14 MR. MICHELSON: Is there an actual holding order 15 right now?
16 MR. BASKIN: Yes.
17 MR. MICHELSON: You have a 50.54?
18 MR. BASKIN: We have a confirmatory action 19 letter from the Region.
20 MR. MICHELSON: That says until such time as the 21 Region says "go" you have to hold fast? Thank you.
22 MR. EBERSOLE: I'll be depending to some degree 23 on the subcommittee here to make comments when we get into 24 this 30-minute session.
25 MR. BASKIN: Jesse, I said I had two quick O
\_/
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6616
. .. . . . . . ~~ ~ . . . .
25726.0 182 BRT
, 1 comments. I have one other. The other -- and I think it's 2 important for the subcommittee's understanding, but not l 3 shown on any of the slides -- but one of the th'ngs we are
} 4 doing as a result of this effort -- it was started earlier i
5 but we I re expanded it -- is to do what we call a complete
] 6 review of the material condition of the plant. Recognizing i
- 7 it's an older facility, this event started as a result of i ,
8 failure of a cable, there will be a complete review of the i 9 overall conditions to make sure the plant is up to the 10 standards, as far as material conditions, that we believe 11 it is. But we're going to make sure.
12 MR. EBERSOLE: You mean in the context of aging?
}
(} 13 t1R. BASKIN: Aging and wear and all that sort of
- 14 stuff.
l 15 MR. EBERSOLE: Any further comments? We have l 16 another meeting at 1:30 here, so we will conclude this
- 17 meeting.
b l 18 (Whereupon, at 12:30 p.m., the meeting was i
i 19 adjourned.)
} 20
- 21
}
22 1'
23 1
] 24
, 25 i
4 l i l 1
t ACE-FEDERAL REPORTERS, INC.
( 202-347-3700 Nationwide Coverage 800-336-6646 i
CERTIFICATE OF OFFICIAL REPORTER
,s N
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 THE WESTINGHOUSE WATER REACTORS DOCKET NO.:
PLACE: WASHINGTON, D. C.
DATE: WEDNESDAY, FEBRUARY 12, 1986 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
(sigt)
(TYPED JOEL BREITNER Official Reporter ACE-FEDERAL REPORTERS INC.
Reporter's Affiliation,
NRR STAFF PRESENTATION TO THE O ACRS
SUBJECT:
RECENT SAN OU0FRE 1 LICENSI!!G ACTIVITIES DATE;. FEBRUARY 12, 1906
,- r PRESENTER: RICHARD DUDLEY l
O . .
j PRESENTER'S TITLE / BRANCH /DIV: SAN Of10FRE 1 PROJECT MANAGER PAD #1:PWR-A 7 -
PRESENTER'S NRC TEL. NO.: 49-27218 SUBCOMMITTEE: uESTIliGHOUSE HATER REACTORS 6
O I
~
,_s RECENT LICENSING ACTIVITIES U SAN ON0FRE, UNIT 1 o SHUT DOWN FEBRUARY 27, 1982 F09 REFUELING o SEP REVIEW RAISED CONCERNS RE: ABILITY OF PLANT TO MEET ORIGINAL SEISMIC DESIGN OF 0.50G o SCE COMMITTED TO UPGRADE TO 0.67G o NRC AUGUST 11, 1982 ORDER PRECLUDED RESTART UNT-IL SEISMIC CONCERNS RESOLVED o PLANT RESTART WAS UNCERTAIN es o SCE DEFERRED UNIT 1 LICENSING WORK, EXCEPT U FOR SEP ACTIVITIES o LICENSING ACTION BACKLOG ACCUMULATED
~ ~
o SCE SUBMITTED RETURN-TO-SERVICE PLAN ON DECEMBER 12, 1983 o SEISMIC UPGRADE OF STRUCTURES, ELECTRICAL SYSTEMS AND PLANT EQUIPMENT NEEDED TO
- ACHIEVE HOT SHUTDOWN o PROPOSED INTEGRATED LIVING SCHEDULE METHODOLOGY o NRC ALLOWED RESTART IN NOVEMBER 1984 AFTER, SEISMIC UPGRADE .
O PAD #1:PWR-A:NRR
/1$86
\_)
o ONG0ING SEP PROGRAM ISSUED DRAFT IPSAR (NUREG-0829) IN APRIL 1985 0 STAFF BRIEFINGS TO ACRS SUBCOMMITTEE - JUNE 19, 1985
- FULL COMMITTEE - AUGUST 9, 1985 o ACRS LETTER ISSUED AUGUST 13, 1985 SUPPORTING STAFF'S CONCLUSIONS o SCE PLANNED TO OPERATE SONGS 1 UNTIL NOVEMBER 30, 1985 AND SHUTDOWN FOR UPGRADES IN ACCORDANCE WITH (1)
COMPLETE SEISMIC MODIFICATIONS COMPLETE EQUIPMENT QUALIFICATION COMPLETE FIRE PROTECTION MODS _ _
INSTALL DEDICATED SHUTDOWN SYSTEM INSTALL THIRD AFW PUMP (UPGRADE TO THIRD SAFETY-GRADE TRAIN IN NEXT OUTAGE) o NOVEMBER 21, 1985 WATER HAMMER EVENT o INCIDENT INVESTIGATION TEAM DISPATCHED TO SITE o LICENSEE BEGAN REFUELING OUTAGE EARLY ,
O PAD #1:PWR-A:NRR RICHARD DUDLEY 2/12/86
~'
NRC STAFF ACTIONS T0. RESOLVE
() LOSS OF POWER / WATER HAMMER CONCERNS IIT REPORT JANUARY 20, 1986
~
o o IIT PRESENTATION TO COMMISSION; JANUARY 22, 1986 o ED0 MEM0 - FEBRUARY 4, 1986:
REGION V LEAD FOR FACILITY RESTART NRR TO REVIEW TECHNICAL DESIGN QUESTIONS IE GIVEN GENERIC LEAD RE: CHECK VALVE y.
- PROBLEMS AND GENERIC ADEQUACY OF IST NRR TO ASSESS NEED TO REEVALUATE WATER HAMMER USI o NRC TO MEET WITH SCE ON FEBRUARY 19, 1986, T0
(-) DISCUSS ITEMS REQUIRED FOR RESTART o SCE TO MEET WITH COMMISSION IN MARCH 1986
~
o PLANT RESTART ORIGINALLY SCHEDULED FOR MAY 20, 1986 O
PAD #1:PWR-A:NRR
/12)86
. - - . _ - - - - - - . .. , . _ . - . . - .. _ - - .-.-- - - - - _ .~. - . - , - . . . . -
.Q @.
I 4
i . !-
! I 1
COMMISSION BRIEFING l
1 i
LOSS OF POWEP AND WATER HAMMER EVENT !
! SONGS 1 NOVEMBER 21, 1985 l
i T. T. MARTIN 488 1280
,. JANUARY 22, 1986 e
/!
O O O 1
l SAN ON0FRE, llNIT 1 (SONGS-1)
- OPERATED BY SOUTHERN CALIFORNIA EDIS0N LOCATED SOUTH OF LOS ANGELES.NEAR SAN CLEMENTE,
, CALIF 0PNIA 1
i l WESTINGHOUSE 3-LOOP PWR
{ LICENSED 1967 l 1337 MWT / 450 MWE SPHERICAL STEEL CONTAINMENT WITH CONCRETE l 1 ENCLOSilRE BilILDING i
I
- ELECTRIC MAIN FEEDWATER PUMPS ALSO FilNCTION AS SAFETY INJECTION PUMPS T. T. MARTIN, 488 1280 j .. JANUARY 22, 1986 i
- a,
O O O SAN ON0FPE, UNIT 1 CONT'D 1 TilRBINF. AND 1 ELECTRIC AUXILIARY FEEDWATER PUMP 1 IMMEDIATE AND 1 DELAYED ACCESS OFFSITE POWER SOURCE DIESELS START, BUT DO NOT AllT0MATICALLY LOAD ON LOSS OF POWER n
s T. T. MAPTIN, 488 I?80
- i. JANUARY 22, 1986 R
O O O i
k j EVENT OVERVIEW LOSS OF IN-PLANT AC POWER i INOPERABLE FEED PUMP CHECK VALVE LEADS TO RUPTilRE OF CONDENSATE SYSTEM COMPONENT i
- LOSS OF FEEDWATER MllLTIPLE INOPERABLE FEEDWATER CHECK VALVES ALLOW i
BACI' FLOW FPOM All STEAM GENERATOPS WATER HAMMEP IN B FEED LINE CAllSES PIPING, PIPING SilPPORTS AND COMPONENT DAMAGE i DAMAGED FEEDWATER CHECK VALVE DEVELOPS SIGNIFICANT i j STEAM-WATER LEAK Pl. ANT SHUTDOWN AND C00LDOWN COMPLETED SAFELY 1
- i. T. T. MARTIN, 488 1280 JANUARY 22, 1986 4
O O O 4
i 4
i TEAM ESTABLISHED BY ED0 ON NOVEMBER 22, 1985 l 1 l IN C0FFORMANCE WITH COMMISSION APPROVED
- IFCIDEPT INVESTIGATION PROGRAM CHARTER
- j *
- DETERMINE WHAT HAPPENED i
) IDENTIFY PROBABl.E CAllSES OF WHAT HAPPENED
. MAKE APPROPRI ATE FINDINGS AND CONCLIISIONS T0 i FORM BASIS F0P POSSIBLE FOLLOW-0N ACTIONS l
1 j ,
~
i T. T. MAP. TIN, 488 1280 !
'- JANilAPY 27, 1986 t
1 -
D
l .
o o o i
l l .
j !
i i
i i
i TEAM MEMBEPS a
THOMAS T. MARTIF, TEAM LEADER MATTHEW CHIRAMAL t i
WILLIAM G. KENNEDY WAYNE D. l.ANNING AI.ECK W. SERKIZ l
I
! STEVEN K. SHOPE 1(
- T. T. MARTIN, 488 1280 JANIJARY 22, 1986 s.
}
1 _ _ _ _ . - - - - - _ _ _-. . . . - - - - .
O O O 1
i i
i 4
. FACT FINDING METH0D0L0GY I
i l INTERVIEWS AND MEETINGS
- PLANT DATA '
l I
i PEPSONNEL LOGS I
, QUARANTINED E0llIPMENT i
l OBSERVATIONS
! STATUS REPORTS s
1 ? -
j -
- T, T. MARTIN, 488 1280 t
.. JANilARY 22, 1986 4
l 1
pP
-- =. .
T O O O
)
i 1
SE0llENCE OF EVENTS INITIAL CONDITIONS SMALL SALTWATER LEAK INTO MAIN CONDENSER UNIT DPERATING AT 60 PERCENT POWER STEAM GENEPATOR BLOWDOWN ABOUT 100 GPM/ STEAM i GENERATOR CRITICAL FUNCTION MONITOR SYSTEM DISABLED ELECTRICAL GROUND TR0llBLESH00 TING IN PROGRESS ELECTRIC PLANT IN UNIISUAL ALIGNMENT T. T. MARTIN, 488 1280 JANUARY 22, 1986 i
1 f,
O O O SE0llENCE OF EVENTS T=0 (MIN) AUXILIARY TRANSFORMER C ISOLATES (PARTIAL. LOSS OF AC POWER) j T=0+ ENS PHONE RINGS FEEDWATER PUMP CHECK VALVE FAILS TO CLOSE FLASH EVAPOPATOR TilBE RUPTilRES DIESEL GENERATOR 2 STARTS, BilT ;
BY DESIGN DOES NOT LOAD T=1/3 OPERATORS TRIP PEACTOR AND llNIT GENERATOR (TOTAL LOSS OF INPLANT A/C POWER)
CONTAINMENT ISOLATES DIESEL GENERATOR 1 STARTS, BUT BY i DESIGN DOES NOT LOAD
! TllRBIFE-DRIVEN AUXILIARY FEEDWATEP PUMP RECEIVES START SIGNAL F0llR ADDITIONAL FEEDWATER CHECK VALVES IN0PEPAPLE T. T. MAPTIN, 488 1280 JANilARY 22, 1986 i
?
o O O SEQUENCE OF EVENTS T=1/3+ FEEDWATER LINES PEGIN TO EMPTY STEAM GENERATORS BEGIN TO LOSE INVENTORY ,
VIA FAILED CHECK VALVES AND RUPTilRED TilBE IN FLASH EVAPORATOR ,
l l LOSS OF VOLTAGE A11TOMATIC TRANSFER SCHFME INITIATED
- ALARMS INDICATE SAFETY INJECTION ACTl!ATION T=3 FIPE TRUCK ARRIVES ON SITE i
/0
O .
O O SEQUENCE OF EVENTS T=ll DPERATORS ATTEMPT TO RESTORE POWEP TIIRBINE-DRIVEN AFW FLOW AF0llT 130 GPM/
STEAM GENERATOR T = 12 + OPERATORS RESTORE POWER ON FIFTH ATTEMPT MOTOR-DRIVEN AFW PilMP STARTS INCREASING FLOW TO 155 GPM/SG ATMOSPHERIC STEAM DUMPS OPEN OPERATORS CLOSE FEEDWATEP ISOLATION VALVES PADIATION MONITOR ALARMS RESET (S/G ELOWDOWN RE-ESTABLISHED) 5 kNkR ? 98
//
O O O SE0llENCE OF EVENTS T=6 RCS PRESSUPE, TEMPERATURE AND PPESSilRIZER LEVEL INDICATE RAPID C00LDOWN T=7 OPERATORS START CHAPGING PUMP SECOND PUMP AllT0 STARTS LATER T=10 PEACTOR COOLANT PUMP B STARTED T=11 OPERATORS TERMINATE (10 SECONDS) AND RE-ESTABLISH AUXILIARY FEEDWATER FLOW AT 40 GPM/SG T=15 UNUSUAL EVENT DECLARED ONSITE/NOT TO NRC
\
T. T. MARTIN, 488 1280 JANllARY 22, 1986
/A
O O O 1
SE0llENCE OF EVENTS T=16 WATER HAMMER OCCURRED T=18 RCP B THRIJST BEARING HIGH TEMPERATURE ALARM T=19 PLANT EQUIPMENT OPERATOR INFORMS CONTROL ROOM 0F "STEAMLINE BPEAK"
- T=25 CONTAINMENT COOLING PE-ESTABLISHED 4
T=29 OPERATORS SECURE DIESEL GENEPATORS
- OPERATORS SECilPE LilBE OIL RESERVOIR
- F0AM SYSTEM T. T. MARTIN, 1188 1280 JANI!AP.Y 22, 1986 6
O O O i
i SE0llENCE OF EVENTS T=33 PCP A STARTED j T=36 RCP C STARTED j STEAM GENERATOR LEVELS OFF-SCALE LOW i
T=37 RCP B STOPPED i AFW FLOW TO A AND C INCREASED TO 70 GPM/SG TO ESTABLISH RAPID Bi1T ACCEPTABLE C00LDOWN T=39 STEAM GENERATOR BLOWDOWN DISCOVERED AND TERMINATED i
T=40 STEAM GENERATOP A AND C LEVEL.S ONSCALE ATTEMPTS BEGIN TO IDENTIFY STEAM LEAK 5
! T. T. MAPTIN, 488 1280
, JANilARY 22, 1986 e.
M
O O O
! SE0llENCE OF EVENTS ,
T=90 (MIN) OPERATORS UNABLE TO START i CIRCULATING WATER PilMP 4
T=3 (HR) ENTERED MODE 4 T=4 OPERATOR OVERRODE PHR INTERLOCK I
- T=5 llNUSUAL EVENT TERMINATED i
T=6 FEEDWATER LEAKAGE ISOLATED 4
! T=10 ENTERED MODF. 5 1
\
l T. T MARTIN, 488 1280 JANilARY 22, 1986 n.
/S'
. _ _ . . . _ _ _ _ _ _ . _ . _ . . . _ . . _ . . . _ _ . - _ . . . _ . . _ _ . . . _ . _ _ . _ _ _ _ . _ . _ _ . _ . _ . _ . _ ~ . ~ .
1 l
O O O
! t i
i
,- i SEQllFNCE OF Et'ENTS E
! NOVEMP>ER 22 CONTAINMENT ENTRY FOUND EVIDENCE OF a
l WATER 11AMMER DAMAGE i
i d
t i
1 i
l l
j . .
i T. T. MARTIN, 488 1280 l
4 JANilARY 22, 1986 l
4 1
4
O O O PERSONNEL PERFORMANCE EVALUATIONS OPERATOP EPR0RS FAILED TO FOLLOW APPROPRI ATE PP0CEDl!RES WHEN TP0llBLESH00 TING ELECTRICAL GROUND DIFFICULTY IN RE-ESTABLISHING INPLANT POWER INADVERTENTLY PE-ESTABLISHED STEAM GENERATOP BLOWDOWN DID NOT RESET FOX III COMPUTER STA PERFORMANCE GOOD EMERGENCY COORDINATOR PERFORMANCE ERRORS FRC-LICENSEE COMMilNICATIONS WERE F0T EFFECTIVE T. T. MARTIN, 488 1280 JANilAPY 22, 1986
/7 ,
O O O F0lIIPMENT PROBLEMS PP0BABLE ITEM NATURE OF FAILURE ROOT CAUSE COMMENTS
- 1. POWER SUPPLY CABLE GROUND FAlll_T WATER INLEAKAGE ANALYSES ONG0lFG
- 2. FLASH EVAPORATOR l' NIT TilBE/SHELL RllPTURED OVERPRESSURIZED DifE NO LONGER USED TO FAILED CHECK VALVE 3.
SAFETY INJECTION SPURIOUS ALARM LOFS OF POWER DESIGN INADE011ACY
, ANNIINCI ATOP
- 4. SAFEGUARD LOAD INDICATED SAFETY NOT IDENTIFIED ANALYSES ONG0ING SEQUENCING SYSTEM INJECTION ACTUATION i
- 5. LOSS OF VOLTAGE AUTO FAILED TO REAl.IGN NOT IDENTIFIED ANALYSES ONG0ING
, TRANSFER SCHEME CIRCUIT BREAKERS TO i RESTORE POWER T. T. MARTIN, 488 1280
, JANilARY 22, 1986 i
i /8"
O O O E0lilPMENT PROBLEMS CONT'D PROBABLE ITEM NATllRE OF FAllVRE P00T CJttlSE COMMENTS
- 6. FOX III COMPUTEP NO RECORDED DATA POWEf( INTERPllPTION RESET RE0llIRED BEFORE/AFTEP TRIP
- 7. TilRBINE RUPTURE RUPTIJPED OVERPRESSilRIZED DUE EXPECTED DISKS (4 0F 8) TO LOSS OF POWER
- 8. EMERGENCY NOTIFICA- SPURIOUS RINGS NOT IDENTIFIED CANNOT REPRODilCE TION SYSTEM
- 9. RCP THRUST BEARING HIGH TEMPERATURE FAILED DETECT 0P ALARM 10., CHECK VALVE STUDS STRETCHED WATER HAMMER
' FWS-378
- BODY TO BONNET LEAK T. T. MARTIN, 488 1280 JANUARY 22, 1986
/7
.O o o EQUIPMENT PROBLEMS CONT'D PROBABLE
,I ITEM NATURE OF FAILURE ROOT CAllSE COMMENTS
)
i 11. CHECK VALVE NUT MISSING NOT IDENTIFIED EVALUATION ONG0ING
. FWS-345 DISC SEPARATED FP0M HINGE ARM
- 12. CHECK VALVE NilT MISSING , NOT IDENTIFIED EVAlllATION ONG0ING FWS-346 DISC SEPARATED FROM HINGE APM ,
j l 13. CHECK VALVE NUT LOOSE NOT IDENTIFIED EVALUATION ONG0ING FWS-398 STUCK-0 PEN 14, CHECK VALVE NUT LOOSE NOT IDENTIFIED NOT PINNED FWS-438 STUCK-0 PEN EVALUATION ONG0ING T. T. MARTIN, 488 1280 JANUARY 22, 1986 i
2 D'
O O O E0lllPMENT PROBLEMS CONT'D PROBABLE ITEM NATURE OF FAllllRE P00T CAllSE COMMENTS
]5. CHECK VALVE NUT LOOSE NOT IDENTIFIED NOT PINNED FWS-439 STilCK-OPEN EVALUATION ONG0ING
- 16. FLOW CONTROL VALVE BROKEN YOKE WATER HAMMER INERTIA FCV-457 BENT STEM
- 17. B STEAM GEFERATOR CRACKED WATER HAMMER EVAlllATION FEEDWATER LINE BENT ONGOING DENTED BEING PEPOVED MOST SilSCEPTIBLE PIPING
- 18. FEEDWATEP LINE DAMAGED WATEP HAMMER EVALUATION
, SllPPORTS 8 SNUBBERS ONGOING T. T. MAPTIN, 488 I?80 JANilARY ??, 1986 a?/.
O O O i
E0lllPMENT PP0BLEMS CONT'D PP0BAFLE ITEM NATURE OF FAllllRE P00T CAllSE CnMMENTS 4
- 19. AUXILIARY FEEDWATER DISPLACEMENT WATER HAMMER LINE SUPPORTS 4
- 20. CONTAINMENT SPHERE SMALL CPACK-LIKE NOT IDENTIFIED EVALUATION INDICATIONS ONG0ING j 21. SECURITY ACCESS SECURE SAFEGIIAPDS NO SAFETY / SAFE-CONTROLS INFORMATION GUARDS INTERFACE PROBLEM i T. T. MARTIN, 488 1780 JANt!ARY 22, 1986
- i.
O O O i
l
- PRINCIPAL FINDINGS AND CONCLilSIONS i
! Tile EVENT WAS SIGNIFICANT ALL INPLANT AC POWER WAS l.0ST FOR FOUR MINUTES
- ALL STEAM GENERATOR FEEDWATER WAS LOST FOR TilREE MINIITES i
A SEVERE WATER HAMMER WAS EXPERIENCED IN THE FEEDWATER SYSTEM CAllSED A LEAK i
l DAMAGED PLANT E0llIPMENT t
j -
CHALLEP!GED THE It!TEGRITY OF Tile Ill.TIMATE HEAT SINK i ,
l ALL INDICATED STEAM GENERATOR WATER I.EVELS DROPPED BELOW SCALE i
i
- THE REACTOR COOLANT SYSTEM EXPERIENCED AF ACCEPTABLE BitT !!NNECESSARY C00LDOWN TRANSIEtlT
, t T. T. MARTIN, 1488 1280 JANilARY 22, 1986 j ni
g q v
g Lj a
SIGNIFICANT FINDINGS (NOT NECESSAPILY IN ORDFR OF THEIR SIGNIFICANCE)
- 1. THE PPIMAPY CAllSE FOR THE WATEP HAMMER IN THE FEEDWATER PIPING WAS THE FAILUPE OF MULTIPLE CHECK VALVES IN THE FEEDWATER SYSTEM. THESE FAILilPES PERMITTED THE PIP!NG TO EMPTY AND FILL WITH STEAM BEFORE THE MOTOP-0PERATED FEEDWATER IS01.ATION VALVES WEPE CLOSED. ALTil0 UGH THE STEAM CONDENSATION-INDUCED WATER HAMMER OCCURRED !N ONLY ONE FEEDWATER LINE, THE POTINTIAL EXISTED FOR WATER HAMMEP TO OCCUR THROUGH0llT THE SAFETY-RELATED PORTIONS OF THE FEEDWATER SYSTEM.
- 2. THE FAILURES OF THE FIVE CHECK VALVES IN THE FEEDWATER SYSTEM PROVIDED A MECHANISM FOR POTENTIAL COMMON MODE FAILURE OF THE HEAT SINK PROVIDED BY THE THREE STEAM GENERATORS. THE FAILED CHECK VALVES PERMITTED HIGH PRESSURE STEAM AND WATER FROM THE STEAM GENERATOPS TO FLOW BACK TO THE LOW PRESSURE CONDENSATE SYSTEM; THE BACKFLOW CARRIED WITH IT THE AUXILIARY FEEDWATER FLOW NECESSARY TO MAINTAIN THE HEAT SINK PROVIDED BY THE STEAM GENERATORS. OPERATOR ACTIONS WERE NECESSARY TO STOP THE BACKLEAKAGE AND PREVEtlT A MORE SEPIOUS SEQUENCE OF EVENTS.
- 3. LONG HORI70NTAL RUNS OF FEEDWATER PIPING WITH THE POTENTIAL FOR VOIDING ARE PARTICULAPLY SUSCEPTIBLE 7 TO DESTRUCTIVE STEAM CONDENSATION-INDilCED WATER HAMMERS.
i FlfRTHER, DPERATORS ARE NOT PROVIDED THE MEANS F0P DFTECTING THE VOIDING OF THESE LINES OR GIVEN GilIDANCE ON APPROPRIATE WAYS TO DEAL WITH THE SITllATION. DESIGN OR PROCEDl' PAL CHAf'GES MAY BE WAPPANTED.
T. T. MARTIN, 488 1280 JANUARY 22, 1986 2V
O O O
~
SIGNIFICANT FINDINGS CONT'D
- 4. THE FLASH EVAPORATOR FAILED WHEN OVERPRESSL7IZED BY Tile DISCl1ARGE FLOW 0F AN OPERATING FEEDWATER PI'MP DUE TO THE PARTI AL LOSS OF POWER AND A STIICK OPEN FEEDWATER PUMP DISCHARGE CHECK VALVE THAT SHOULD HAVE PREVENTED THE BACKFLOW.
- 5. Tile TIMING OF THE FIVE CHECK VALVE FAILURES COULD NOT BE ASCERTAINED WITH CERTAINTY, THE TEAM CONCLUDED THAT ALL CHECK VALVES HAD FAILED PRIOR TO THE EVENT RECAllSE THE MISSING PAPTS TO THE VALVES WERE NOT F0lIND IN THE INSPECTED FEEDWATEP PIPING AFTER THE EVENT, NOISE FROM Tile B STEAM GENERATOR FEEDWATEP PIPING, EVIDENT TO PLANT PERSONNEL SINCE JUNE 24, 1985, SIIPPOPTS THE CONCLilSION TilAT THE FEEDWATEP CONTP01. STATION CHECK val.VE IN THE B FEEDWATER I.INF. HAD FAILED EARLIER. THE INSPECTION OF THE STEAM GENER-ATOPS HAS NOT YET BEEN COMPLETED BY SCE.
- 6. THE SURVEILI.ANCE PROCEDURE FOR TESTING THF. CHECK VALVES I.. THE INSF.PVICE TESTING (IST) PROGRAM LACKED ADE0llATE METHODS AND OBJECTIVE ACCEPfANCE CRITEPIA FOR DETER-MINING WHETHER CHECK VALVES ARE CLOSED. THilS, ALTHOUGH THE CHECK VAI.VES HAD- BEEN TESTED WITilIN THE PAST YEAR, OPERATORS MAY HAVE MISINTERPRETED THE TEST RESilLTS, FURTHERM0RE, THE IST IS NOT DESIGNED TO DETECT DEVELOPING CONDITIONS THAT MAY LEAD
'TO THE FAILURE OF THE CH'ECK VALVES; E.G., LOOSE DISKS AND STilD N!!TS.
T. T. MARTIN, 488 1280 JANUARY 22, 1986
.2S
O O O SIGNIFICANT FINDINGS CONT'D
- 7. THE NRC HAD NOT COMPLETED ITS REVIEW 0F SCE'S INSERVICE TESTING PROGRAM. THE INITIAL PROGRAM WAS SUBMITTED IN SEPTEMBER 1977 AND REVISED IN ITS ENTIRETY ON JANUARY 24, 1984. DISAGREEMENT BETWEEN SCE AND NRC ON RESOLUTION OF CERTAIN OPEN ISSUES AND SCHEDULING PROBLEMS WITil NRC'S REVIEW HAVE SUP.STANTIVELY CONTRIBilTED TO THIS DELAY.
- 8. THE RESOLUTION OF THE UNRESOLVED SAFETY ISSUE, llSI A-1, " WATER HAMMER," DID NOT SPECIFICALLY ADDRESS THE PREVENTION AND MITIGATION OF THE CONSEQUENCES OF CONDENSATION- INDUCED WATER HAMMERS IN FEEDWATER PIPING llPSTREAM 0F THE FEEDRING.
INTERVIEWS OF NRC STAFF INVOLVED IN RES0Ltl TION OF WATEP HAMMER ISSUES FAILED TO DEVELOP CITABLE REFERENCES, DECISIONS, OR DISCUSSIONS TilAT PROVIDED A BASIS FOR EXCLUDING FURTHER CONSIDERATION OF FEEDWATER PIPING WATER HAMMER. HOWEVER, IN THE REGULATORY ANALYSIS OF THE RESOLUTION OF USI A-1, THE STAFF ACKNOWLEDGED THAT ELIMINATION OF WATER HAMMFRS IS NOT FEASIBLE, THAT THE FPEQUENCY OF WATER HAMMERS HAD BEEN SilBSTAf!TTAl.LY PEDUCED BY CHANGES IN DESIGN AND OPERATIONS, AND THAT STilDIES OF WATER HAMMER HAD REVEALED A SIGNIFICANTLY LESSER SAFETY CONCERN THAN PREVI0llSLY HYP0THESIZED. IT APPEARS THAT FilRTHER CONSIDERATION OF WATER HAMMERS DllE TO MAIN FEEDWATER LINE VOIDING WAS NOT PURSilED DilE TO A LACK OF REPORTED OCCURRENCES IN ll.S.
'RLANTS.
T. T. MAPTIN, 488 1280 JANilARY 22, 1986 cWo
O O O SIGNIFICANT FINDINGS CONT'D
, 9. NPC'S RELIANCE ON "J" TUBES TO DELAY THE DEVELOPMENT OF CONDITIONS NECESSARY TO SUPPORT STEAM GENEPATOR WATER HAMMER IMPLICITLY ASSUMES THAT FEEDWATER CHECK VALVE INTEGRITY k'0VLD BE MAINTAINED TO PREVENT STEAM GENERATOP FEEDRING V0IDING. HOWEVEP, CORRESPONDING REGULATORY PE0lilREMENTS TO ENSUPE THAT THESE CHECK VALVES PERFORMED THIS SAFETY FUNCTION WERE NOT PART OF THE RES0LilTION OF THE WATER llAMMER ISSUE.
- 10. THE POOT CAUSF FOR THE LOSS OF POWER WAS A PHASE-TO-PHASE FAULT OF AN ELECTRICAL CABLE
! FROM AUXILIARY TRANSFORMER C TO BUS IC. THE l'NDERLYING REASON FOR THE CABLE FAILURE
- HAS NOT YET BEEN DETERMINED
- HOWEVER, IT APPEAPS THAT THE CABLE MAY HAVE BECOME WETTED j BY A LONG-TERM FLANGE LEAK FROM THE FEEDWATER SYSTEM, RUNNING AB0VE Tile CABLE TRAY,
- 11. Tile PLANT IS DESIGNED TO EXPERIENCE AN EXTENDED LOSS OF INPLANT AC POWER ON LOSS OF
- 0FFSITE POWER WITHOUT SAFETY INJECTION. OPERATORS ARE REQUIRED TO RESTORE POWER FROM THE SWITCHYARD OR TO LOAD THE DIESEL GENERATORS TO RESTORE INPLANT POWER. SCE'S EMERGENCY OPERATING INSTRUCTIONS ON LOSS OF AC POWER LACK GUIDANCE ON HOW LONG OPERATORS CAN ATTEMPT TO RESIDRE POWER FROM 0FFSITE SOUPCES BEFORE THE DIESEL
, GENERATOPS SHOULD BE LOADED FOLLOWING A LOSS OF INPLANT AC POWER, OR HOW LONG THE
. DIESEL GENEPATORS CAN RUN UNLOADED WITH0llT OVEPHEATING, IF THEIR AC-POWERED RADIATOR FANS REMAIU DE-ENERGIZED.
T. T. MARTIN, 488 1280 JANUARY 22, 1986 N
O O O SIGNIFICANT FINDINGS CONT'D
- 12. THE STATI0tl LOSS OF VOLTAGE AUTO TRANSFER SCHEME FOR ESTABLISHING THE DELAYED ACCESS TO 0FFSITE POWER MAY NOT HAVE FUNCTIONED AS DESIGNED. SCE EVAllIATI0f!S ARE CONTINilING.
- 13. THE MULTIPLE SPURI0llS INDICATIONS EARLY IN THE EVEFT THAT A SAFETY INJECTION ACTUATION HAD OCCURRED, ADDED TO THE CONFUSION OF THE SITUATION AND UNNECESSARILY INCREASED THE BilRDEN ON THE OPERATORS. OPEPATORS DIAGNOSED PLANT CONDITIONS AND APPROPRIATELY DIS-REGARDED THESE INDICATIONS. THE SAFETY INJECTION ANNUNCIATOR wit.L AI. WAYS INCORRECTLY ALARM OF A LOSS OF AC POWER. THIS IS A DESIGN DEFICIENCY. TFE CAllSE OF THE SPURI0liS INDICATION ON BOTH SAFEGUARD LOAD SEQUENCEP SYSTEM PANELS IS STILL (JNKNOWN.
- 14. THE OPERATIt!G STAFF, WITH THE CONCllRRENCE OF MANAGEMENT, DID NOT F0LLOW APPROPRIATE PROCEDl!RES WHEN TROUBLESHOOTING THE ELECTRICAL GR0llND. THEIP ACTIONS tlNNECESSARILY DELAYED ENTRY INTO TECHNICAL SPECIFICAlION ACTION STATEMENT RE0llIREMENTS THAT C0llLD PE0llIRE Pl. ANT SHilTDOWN.
- 15. ONCE THE ELECTRICAL GR0l!ND WAS LOCATED ON THE FEEDER FROM AlIXILI ARY TRANSFORMER C TO BilS JC, THE OPERATORS DID NOT AGGRESSIVELY PURSilE ISOLATING THE AlIXILIAPY
\ TRANSFORMER. INSTEAD,'THEY OPTED TO LEAVE THE TRANSFORMER ENERGIZED WHILE TECHNICIANS PERF0PMED INSPECTIONS THAT DID NOT RE0llIRE THE TRANSFORMER TO BE ENERGI7ED.
T. T. MARTIN, 488 1780 JANUARY 22, 1986 2.Y
,7 ,
C,, U V SIGNIFICANT FINDINGS CONT'D
- 16. THE OPEPATOPS' ACTIONS AFTER THE TRANSFOPMER TRIP WERE CONSISTENT WITH THEIR TPAINING.
HOWEVER, IN THE TEAM'S JilDGMENT, SOME OPERATOPS, LACKED DETAILED PLANT KNOWLEDGE IN THE FOLLOWING AREAS:
CAUTIONS ASSOCIATED WITH PARAI.LELING TRANSFDPMERS PE0llIPEMENTS FOR RESETTING llNIT GENERATOR TRIPS THE PROCESS F0P OPERATING 220KV CIRCIIIT BPEAKEPS EXPECTED INDICATIONS 4!D TIMING OF THE LOSS OF VOLTAGE AUTOMATIC TRANSFER SCHEME SETPOINTS FOR RESIDUAL. HEAT REMOVAL SYSTEM PRESSURE INTERLOCK EXPECTED INDICATION AND MEANING OF LItHTS ON SLSS SE0DENCEP PANEL.S OPERAPILITY OF DIESEL GENERATORS WITH AUXILIARY TRANSFORMER C REACTOR C0ll BYPASS BREAKERS REMOVED.
THESE DEFICIENCIES MAY BE D'UE TO INADE0llATE OPERATOR TRAINING AND/0P PROCEDilRES.
T. T. MARTIN, 488 1780 JANUARY 22, 1986 2
O O O SIGNIFICAFT FINDINGS CONT'D
- 17. ON OCCASI0fl, SOME SITE PEPSONNEL WHO GENERALLY EVALUATE PLANT DATA LACKED A SUFFICIENTLY IN0UIRING ATTITUDE. AS A RESULT, CERTAIN SIGNIFICANT INDICATIONS OF UNDERLYING REASONS FOR SYSTEM RFSPONSE OR COMPONENT PEPFORMANCE WERE NOT DETECTED UNTIL BROUGHT TO THE ATTENTION OF SCE BY THE TEAM. IT APPEAPS THAT SCE'S PROCESS FOR EVALUATING AND FOLLOWING UP EVENTS MAY NOT BE SUFFICIENTLY THOPOUGH AND SYSTEMATIC TO ASSURE THAT FAILED COMPONENTS ARE DETECTED AND ADE0llATELY EXPLAINED.
]8. THE STATUS OF THE STEAM GENERATOR Bl.0VDOWN SYSTEM IS NOT INDICATED IN THE CONTROL P00M. THE PEESTABLISHMENT OF BLOWDOWN WHEN THE PADIATION MONITORS WERE RESFT WAS NOT RECOGNIZED AND ADVERSELY CONTRIBilTED TO THE C00LDOWN OF THE PEACTOR COOLANT SYSTEM AND TO THE del.AY IN REC 0VERING THE STEAM GENERATOR LEVELS.
- 19. DilRING THE LOSS OF ALL INPLANT AC POWER, SilFFICIENT INFORMATION WAS AVAILABLE IN THE CONTPOL ROOM TO ENABLE THE OPEPATOPS TO FOLLOW THEIR PROCEDllRES AND ENSUPE PLANT SAFETY. HOWEVER, CONTROL ROOM GPEPATOPS HAD FAILED TO HAVE THE TECHNICAL SUPPORT CENTEP COMPIITER RESET FOLLOWING ELECTPICAL GR0llND TP0llBLESH00 TING ACTIVITIES. THIS FAILURE DISABLED THE COMPUTER'S ABILITY TO RECORD NEW PLANT DATA AND THEREBY DENIED THE OPEPATORS ACCESS TO PRE-TRIP AND POST-TRIP TRENDS THAT W0llLD PAVE ASSISTED PEAL TIME AND POST-EVENT ANALYSIS AND EVALUATION. HAD THE STATION BLACK 0llT BEEN OF LONGEP
\DlIPATION, OR INVOLVED ADDITIONAL COMPLICATIONS, OPERATOR RESPONSES AND THE FUNCTIONS PROVIDED BY THE TECHNICAL SilPPORT CENTER COULD HAVE BEEN HAMPERED BY THE LACK OF TREND DATA.
T. T. MAPTIN, 1280 JANIIAPY 22, 1986 30
_ _ _ _ _ _ _ ~ - _ - - .. . _ _ . _ _
l "O O O
- SIGNIFICANT FINDINGS CONT'D
- 20. STATION MAINTENANCE REC 0PDS APE INCOMPl.ETE, DIFFICULT TO LOCATE AND, WHEN AVAILABLE, LACK SUFFICIENT DETAIL TO DETERMINE WHAT WAS DONE.
- 21. THE SPURIOUS RINGING OF THE NRC RED PHONE AT THE BEGINNING OF THE EVENT HAS NOT BEEN EXPl.AINED, BilT IT DISTRACTED CONTROL ROOM PEPSONNEL AND CONTRIBUTED TO THE CONFUSION IN THE COMMUNICATIONS BETWEEN SCE AND NRC,
! 22. ENS COMMllNICATIONS BETWEEN NRC AND SCE WERE NOT EFFECTIVE BECAUSE: (1) THE NRC DllTY OFFICER WAS NOT KNOWLEDGEABLE ABOUT THE UNI 0llE DESIGN OF THE PLANT AND, THEREF0PE, MISINTERPPETED OPERATOR RESPONSES TO QUESTIONS; (2) COMMUNICATIONS WITH THE PLANT WEPE INITI ALLY LIMITED BECAllSE STATEMENTS BY PLANT OPERATORS INC0FRECTLY IMPLIED THAT I
SUFFICIENT PERSONNEL WERE NOT AVAILABLE TO SlIPP0PT THE ESTABLISHMENT OF AN OPEN LINE; (3) NPC ASKED LEADING QUESTIONS AND OPERATORS SOMETIMES DID NOT CORRECT, AND IN SOME CASES APPEARED TO CONFIRM, INACCURATE INFORMATION; (4) NRC QUESTIONS CHARACTER-ISTICALLY FOCllSED ON DETAILS PATHER THAN ON THE " BIG PICTURE"; (5) NPC CLUTTERED THE i COMMllNICATIONS CHANNEL WITH REPF_TITIVE DISCUSSIONS ABOUT THE SE0llENCE OF EVENTS AS ADDITIONAL NPC PERSONNEL CAME ON THE LINE TO THE EXCLllSION OF OBTAINING NEW PLANT INFORMATION; (6) NRC RESIDENT INSPECTORS RELIEVED MORE KNOWLEDGEABLE PLANT OPEPATORS i AS ENS COPMUNICATOPS AND REESTABLISHED THE COMMllNICATIONS AT A LOCATION REMOTE FROM
\REAL TIME PLANT INF0PMA' TION: AND, (7) PLANT OPERATORS FAILED TO INFORM THE NRC OF THE DECLARATION OF AN UNilSUAL EVENT.
T. T. MARTIN, 488 1280 JANUAPY 22, 1986 I
i .7 /
O O O SIGNIFICANT FINDINGS CONT'D
- 23. THERE WERE TWO MALFUNCTIONS OF THE AUTOMATED SECURITY ACCESS CONTROL EQUIPMENT; HOWEVER, SITF_ PERSONNEL IMPLEMENTED APPPOPRIATE PLANNED COMPEFSATORY MEASURES, TilEREBY PRECI.UDING A SAFETY-SAFEGilARDS INTERFACE PROBLEM,
- 24. THERE WAS NO SIGNIFICANT RELEASE OF RADI0 ACTIVITY,
O O O CONCLilSION THE MOST SIGNIFICANT ASPECT OF THE EVENT FAS THAT FIVE SAFETY-RELATED FEEDWATER SYSTEM CHECK VALVES DEGPADED TO THE POINT OF INOPERABILITY DUPING A PEPIOD OF LESS TilAN A YEAR, WITH0llT DETECTION, AND THAT THEIR FAILURE JEOPAPDIZED THE INTEGRITY OF SAFETY-RELATED FEEDWATER PIPING.
TFE POOT CAllSES OF THE CHECK VALVE FAILURES HAVE NOT BEEN DETERMINED AND ARE STILL UNDER REVIEW BY SCE AND ITS CONTRACT 0PS POTENTI AL CONTRIDllTOPS TO THIS PROBLEM INCLl'DE INADEQUATE MAINTENANCE, INADEQUATE INSERVICE TESTING, INADE0llATE DESIGN, AND INADE0llATE COPSIDERATION OF THE EFFECTS OF PEDUCED POWER OPERATIONS.
T. T. MARTIN, 488 1280 JAMilARY 22, 1986 i.
- -O O O l PDTENTIAL CONTRIBilTORS TO FAlll!RES MAINTENANCE RECORDS F0P THESE VALVES WEPE EITHER MISSING OR LACKED SPECIFICITY ON WHAT WAS DONE INSERVICE TESTIFG PECORDS FOR THESE VALVFS WERE INCONSISTENT; THE TESTING PROCEDURE WAS NOT RIG 0ROUS; THE TEST ACCEPTANCE CRITERIA WERE SUBJECTIVE; THE TESTING FREQUENCY WAS OPEN-ENDED
- AND, THE TEST DID NOT ASSURE DETECTION OF THE FAILURES F0llND.
TilESE CHECK VALVES AND VALVES OF SIMILAR DESIGN HAVE A HISTORY OF LIKE FAlll]RES REDUCED POWER OPERATIONS AT UNIT 1 ARE NOW ROUTINE BECAllSE OF STEAM GENERATOR TllBE PLilGGIt!G AND SLEEVING, AND THE REDilCED FEEDWATER FLOW MAY HAVE INCPEASED THE SUSCEPTIBILITY OF CHECK VALVE COMPONENTS TO HYDRAllLIC-INDUCED VIBRATION, T. T. MARTIN, 488 ]280 JAFilARY 72, 1986 24
l l
. i O l AGENDA MEETING WITH ACRS SAN ONOFRE UNIT 1 WATER HAMMER EVENT OF NOVEMBER 21, 1985 FEBRUARY 12, 1986 INTRODUCTION i
WATER HAMMER EVENT O I FEEDWATER CHECK VALVES CORRECTIVE ACTIONS CONCLUSIONS O
1 i
O INTRODUCTION o EVENT WAS VERY SIGNIFICANT o MOST SIGNIFICANT ASPECT OF EVENT IS FAILURE OF FIVE FEE 0 WATER CHECK VALVES O
O
O O O MAIN AND AUXILIARY FEEDWATER SAN ONOFRE UNIT 1 L V AP COND P7 i / Il ?!L 7'7 v^y^
,,3 g 4 , ,,
- - - ~ ".
' ~~~~1HxHh -
(( _
j llV <
tWS 438 tlV .- - ._
852A ISI PI llP lilli ll.ll! . 38Ill 311 0 PI liill 2 Nil PI lilli Bb4A
. ./
{ p bI CitMBil ggg gggt g gg illHM S Af i IY 10 SAF L TY INJL Cil0N INJLCilON ,
i
! e f LV AP CUNO gg
,P / W. Mf W PUMP F li -
S Til 4 Til
^ ~~ '
- E i bi Muv ""
M r
, p, p, N -
IIV tWS 8528 ISI PI flP llill Q ny gg nyn . -- --
439 F L ASit
-\ bill
-.3110 PI Hill 2ND PI liill g54g PI '~_.
4'd 4 T"^
CllMBH Dil AIN COULLil M
468
[A1 INj
, - <. CV-143 IWS iWS OI ("3 < *ii 4ib =-- ' " "
dj ;M N N MOV ICV IWS IWS 22 458 398 396 3;
[' 'HNj M CV-144 IW5 IWS g
( (") a 3i8 3ib < >
e tu d: ;k N M MOV ICV IWS IWS 20 451 346 342 II 456
( M Nj [' 'h Cv-i42 eWS #ws
( (~) <- 3is ali = 80 "
o ,
~
c MAIN FEE 0 WATER WATER HAMMER EVENT o EAST FEE 0 WATER PUMP LOSES POWER (0000) o EAST FEE 0 WATER PUMP DISCHARGE CHECK VALVE (FWS-438) ALLOWS REVERSE FLOW o LOUD NOISE HEARD IN AREA 0F EAST FEE 0 WATER EVAPORATOR CONDENSER o WEST FEE 0 WATER PUMP LOSES POWER (0001) o FEE 0 WATER CHECK VALVES (FWS-398,-345 & -346) A[LOWLINESTOVOID O
o AUXILIARY FEE 0 WATER FLOW INITIATED AT 100 GPM (0004) o COMBINED AFW FLOW CONTROLLEO TO 135-150 GPM (0008) o MAIN FEE 0 WATER FLOW CONTROL VALVES AND MOTOR OPERATED BLOCK VALVES SHUT IN ACCORDANCE WITH EMERGENCY OPERATING INSTRUCTION (0008) 1
'A' o FEE 0 WATER LINES AND 'C' REFILL o AUXILIARY FEE 0 WATER FLOW REDUCED TO ZERO THEN INCREASED TO 25-40 GPM (0009)
O o WATER HAMMER EVENT EXPERIENCE 0 (0020)
l
+ -
O
~
~ 1 2<0
< ~
t
,h UU
~
N F <
o f
=
M
~\
(YT sa ,! ( V 3 are s,!s UKy 4
are 8 UKAg g .
TYPICAL WATER SLUG MODEL IN A HORIZONTAL PIPE STEAM FEEDWATER PUMP GENERATOR
/
q j I SUBC00 LED WATER STEAM STEAM FLOW STEAM 7 3 FEE 0 WATER
' N # PUMP GENERATOR O <
lV AUXILIARY FEEDWATER FLOW SLUG FRONT
- FEEDWATER STEAM GENERATOR lt PUMP V0ID FRACTION APPR0XIMATELY ZERO TO 15 PERCENT O
A HIGH PRESSURE FEE 0LINE CHECK VALVES DAMAGE ASSESSMENTS o OPENED AND INSPECTED MFW PUMP CHECK VALVES o OPENED AND INSPECTED MFW 10" CHECK VALVES o PERFORMED
- VALVE LEAK CHECK
- BLUE CHECKS
- FLAPPER INSPECTIONS
- DISC / SEAT GAP MEASUREMENTS l
O l
e G
- O
. i;;E j
-::si s,
)
. AM. . .. .
4 .
- 3
- ij q
es _ . .
- z. .. '
- .;l :!: ..
- !. J.k:;
- f ".::q:;: ,
O
- i:. : .
F l:i L "%s):E' g i gy
. .':.i%
- -. ?W '. ....
' ..- :!ji;!?
"; i..J; 3 r "i' ?. ,: ::'
i[ .' -
l I'
- a
.fij 1
4.:
- , y.
- .
F :
2
.:) , ,!! j Gl. '.j.!; ,
m.
.e 99 O
l
HIGH PRESSURE FEE 0LINE CHECK VALVES OAMAGE FINDINGS O
o E. FEE 0 WATER PUMP CHECK VALVE, FWS-438 DISC NOT FULLY SHUT (5-10% OPEN)
DISC NUT LOOSE <
DISC ROTATED COUNTER-CLOCKWISE l o W. FEE 0 WATER PUMP CHECK VALVE, FWS-439 DISC NUT LOOSE DISC PARTIALLY OPEN (ANT! ROTATION LUG LODGED-4 UNDER HINGE ARM) o "A" FEE 0 WATER 10-INCH CHECK VALVE, FWS-345 a
DISC MOUNTING STUD BROKEN OFF AT BOTTOM THREAD DISC EDGES DAMAGED FROM IMPROPER FIT I HINGE BAOLY WORN BY DISC BOLTS
- o "B" FEE 0 WATER 10-INCH CHECK VALVE, FWS-346 DISC FOUND AT BOTTOM 0F PIPE THREADED CONNECTOR ON DISC SHEARED 4
HINGE PIN ARM ELONGATED BETWEEN PIN AND ARM DISC SEATING SURFACE HAS TWO DENTS BONNET STOP SHOWS IMPACT DAMAGE o "C" FEE 0 WATER 10-INCH CHECK VALVE, FWS-398 VALVE DISC NUT BACKED OFF AND WORN
,i DISC ROTATED AND C0CKED IN SEAT (OPEN 20 DEGREES)
HINGE PIN BAOLY WORN DISC E0GES DAMAGED
()
i 4
i
-c<vw--, _.,.-,-,c-.. ..-.__.4--._m.. . . , - _ , , . , , ,,,,, .- ,., ,, ,..,_ - -,.,-_,, ._.-e,e_- - , - . - - .,,,,.m_,.,.c . _ , , , - - , , , , ,. - _ , , 9.--r--
O PACIFIC CHECK VALVE DESIGN o VALVES ARE OVERSIZE 0 o FULL FLOW VELOCITY IS MARGINAL o 90 PERCENT REACTOR POWER OPERATION 12.5 PERCENT FLUID VELOCITY REOUCTION 24 PERCENT DISC LIFTING FORCE REDUCTION 2 OEGREES FROM FULL OPEN 1 TO 3 HZ RESONANCE; 106 CYCLES IN 12 DAYS O
. rrn rrn rrnrrn
,_ i
> l k _
s
> i xUf '-
4 PACIFIC VALVE '
DESIGN 7" ",
( ,
1 1
o ATWOOD-MORRILL CHECK VALVE DESIGN
! o VALVE PROPERLY SIZED TO RE' MAIN FULL OPEN 00WN TO 70% FLOW o ONE PIECE DISC / HINGE ASSEMBLY PRECLUDES OBSERVED FAILURE MECHANISM i
i
. r O rtifhr C r1 rFh_nCrufh
- hMN m\ 3 Awl hush 4wussssi
/ 7
/
- wn
[ .mwi ,m I hANNNNN4% j lh%.
ffa N #
/
V, w=r sure ATWOOD-MORRILL VALVE DESIGN O
/* g
\_/
MODIFICATIONS ADDRESSING FEE 0 WATER WATER HAMMER o REPLACE 8 PACIFIC CHECK VALVES WITH ATWOOO-MORRILL VALVE DESIGN
, o PROVIDE IMPROVED IN-SERVICE TEST CAPABILITY FOR REPLACEMENT VALVES (t; ) o SLOPE "B" FEE 0 WATER PIPE INSIDE CONTAINMENT o A00 CHECK VALVES. WITH IST PROVISIONS, IN EACH FEE 0 WATER LINE INSIDE CONTAINMENT 4
o PROVIDE FOR AUTOMATIC CLOSURE Of MAIN FEE 0 WATER FCV'S FOLLOWING UNIT TRIP AND LOSS OF MFW PUMPS o EVALUATE ADDITION Of V010 SENSING DEVICE WITH CONTROL ROOM ALARM
( )
s_-
l
. .O l GENERAL M00!FICATIONS l
l 0 INSTALL CONTROL ROOM INDICATION OF BLOWOOWN STATUS
- o PROVIDE AUTOMATIC BLOWODWN ISOLATION ON AFW ACTUATION
- i ,
l o EVALUATE POSSIBILITY OF PROVIDING FOR AUTOMATIC LOADING OF DIESEL GENERATORS ON LOSS OF POWER **
o EVALUATE ALTERNATIVES FOR PROVIDING AN IMME0!ATELY AVAILABLE BACKUP SOURCE OF 0FFSITE POWER **
!O o INSTALL ADDITIONAL PROTECTION FOR AUXILIARY TRANSFORMER FEE 0ER CABLING
- l o MODIFICATIONS TO PRECLUDE SPURIOUS ANNUNCIATION OF SAFETY INJECTION
- l o M00!FY TSC COMPUTER FOR AUTOMATIC RESTART FOLLOWING POWER RESTORATION
l o EVALUATE PROVISION OF UPS FOR BUSSES SUCH AS VITAL BUS NO. 4** ,
l
...-sn-
r-n.
'-/ GENERAL MODIFICATIONS (CONTINUED) o OETERMINE AND CORRECT CAUSE OF SPURIOUS RED PHONE RINGING **
o INSTALL CONTROL ROOM CLOCK NOT DEPENDENT ON AC POWER
- o REVIEW CONTROL ROOM INDICATION OF VITAL BUS STATUS (CROR)*
o MODIFY CURRENT LIMITING BYPASS REACTOR BREAKER DESIGN TO PERMIT OG TO CLOSE ONTO 4 KV BUS WHEN BREAKER IS RACKED OUT*
ry
)
NOTES:
- TENTATIVELY PLANNE0 FOR CURRENT OUTAGE
- SCHEDULE NOT YET ESTABLISHED O
O \
CONCLUSIONS o THE MOST SIGNIFICANT ASPECT OF THIS EVENT IS FAILURE Of f!VE FEE 0 WATER CHECK VALVES o CHECK VALVE FAILURE WAS THE RESULT OF INADEQUATE DESIGN o IMPLEMENTING THE MODIFICATIONS OUTLINED ABOVE WILL ELIMINATE THE IDENTIFIED CHECK VALVE FAILURE MECHANISM AND IMPROVE PLANT
( ) PERFORMANCE O
,