ML20137M463
| ML20137M463 | |
| Person / Time | |
|---|---|
| Issue date: | 09/10/1985 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1443, NUDOCS 8509130242 | |
| Download: ML20137M463 (240) | |
Text
{{#Wiki_filter:Qh?St:/443 . 1 UNITED STATES OF AMERICA l
/'] NUCLEAR REGULATORY COMMISSION
(./ l In the matter of: ADVISORY COMMITTEE ON REACTOR SAFEGUARDS Subcommittee on Reactor Operations Docket No. 1 O ef~MaFFRF a L 2 ~ eR[py e Jo )!o: Remove from ACRS 0fice iC6 OPV / Location: Washington, D. C. 1 - 201 Date: Tuesday, September 10, 1985 Pages: f0 ob ANN RILEY & ASSOCIATES (m) Court Reporters 1625 I St., N.W. Suite 921 g91 2 850910 Tashington, D.C. 20006 T-1443 PDR (202) 293-3950
.~ . - - . . .
3 1 SunW 1 UNITED STATES OF AMERICA
.;O l \_) 2 NUCLEAR REGULATORY COMMISSION y; 3 4 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS J.
6 6 Subcommittee on Reactor Operations 7 8 Nuclear' Regulatory Commission 1717 H Street, N. W. 9 Room 1046 Washington, D. C. 10 Tuesday, September 10, 1985 11 12 () 13 The meeting o# the Subcommittee on Reactor Opera-14 tions convened at 1:02 p.m., Jesse Ebersole, Chairman of 16 the Subcommittee, presiding.
! 16 1
PRESENT FOR THE ACRS SUBCOMMITTEE: 17 Jesse C. Ebersole, Chairman David A. Ward 18 Carlyle Michelson ^ Glenn A. Reed , 19 Charles J. Wylie 30 DESIGNATED FEDERAL EMPLOYEE: 21 Richard K. Major , 22
-ALSO PRESENT:
23 E. ROSSI O " G. IIAMMER C1 # F. CHERITY
t
, g.
4 :.-- f 5 - l' E. WEISS t J 2 J. HENDERSON ; f i. j- 3 .V. HODGE
.4 H.-BAILEY-- ,
f 5 T. COLBURN, i !. '6 J. BEARD. i i 1 7 B. SHEWMON 8 r 9 10 t
. 11 12 ~
13 l l. 14 15 ' I t 16 3 i-4 !- 17 i 1 1s . 19 30 ! I - 21 . l c 1: t
- 3. -
l i N
- j. 34 4
1 8 ? I c. 5' P 1
-- - ~ ..-=w---, ____.%u_.em4 .-.,,,,-n..gw, ,.wm, myn-%,m.,_y.,,,pMg.g.,m,,,.. .__77,
3'
#1-1-SueW 1
_P _R_ _O _C _E _E _D _I _N _G _S
-- 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_ Reactor Operations.
5 I am Jesse Ebersole, Chairman of the Subcommittee. 6 The other ACRS members present today are Carlyle Michelson, 7 Glenn Reed, Dave Ward and Chuck Wylie. 8 And we may have Mr. Forrest Remick later. 9 The purpose of this meeting is to discuss recent 10 plant operating experiences. 'The Offices of Inspection and 11 Enforcement and Nuclear Reactor Regulation have identified 12 a number of current events for discussion and a few items () 13 were suggested by Subcommittee members. 14 The meeting is open to the public, but should A 15 - proprietary information be discussed, the meeting would have 16 to be closed for that portion. 17 At the end of the day's discussion, we will 18 identify those items to be presented to the' full Committee 18 - on Thursday, September 12, 1985 from 8:45 to 10:45 a.m. 20 Mr. Ricnard Major, on my right, is the assigned 21 ACRS Staff member for the meeting. 22 A transcript of the meeting is being kept, and i 23 it is requested that each speaker first identity himself or 24 , (~'i. herself and speak with sufficient clarity and volume so O 25 that he or she can be readily heard. i
4
#1-2-SueW 1 We have not received any requests to make oral (3
(_) 2 statements, nor have we received any written comments from 3 members.of the public. 4 Do any members of this subcommittee have any 5 comments to make about the schedule or any general comments? 6 (No response.) 7 If not, I will make one. And that is, as we 8 go through these, help me identify those which should be 4 9 carried to the full Committee. 10 We will now proceed with the meeting, and I call 11 -upon the NRC Staff. I presume that is Mr. Rossi. And I 12 note that it says fourteen topics. And we will be monitoring () 13 the progress, it's like an average of say fifteen minutes 14 apiece or something? 15 MR. ROSSI: Okay. And we will ask each presenter 16 to also monitor his own time. , 17 MR. EBERSOLE: It's all yours, Ernie. 18 MR. ROSSI: I'm -- 19 MR. MICHELSON: Excuse me. Before we get started, 20 there was one topic. And I don't see it on here. And that ! 21 was the CO2 release at Hope Creek last week. E MR. ROSSI: No, that is indeed on here. The 23 fourth item, inadvertent actuation. 24 MR. MICHELSON: Oh, I'm sorry. Yeah, there it x 25 is. Okay. Good. Thank you.
. , _ . . _ _- _ _ - ~. _ _ _ _
5 ' #1-3-SueW 1 MR. ROSSI: I'm Ernie Rossi, Chief of the Events '
%,, 2 Analysis Branch in the Office of Inspection and Enforcement.
3 And we do have fourteen events to talk about this
-4 afternoon. And the presentations will be given by members
- 5 of both the Office of Nuclear Reactor Regulation and the 6 Office of Inspection and Enforcement.
7 In addition to the people that are making presenta-8 tions, we have a number of other people with us to take part I 9 in the meeting and answer questions that may arise. 10 We have'Alex Dromerick from the Office of 11 Inspection and Enforcement; Ron Hernan.from the Office of i 12 Nuclear Reactor Regulation; Steve Elrod from NRC, Region II;
) 13 and, Nick Chrissotimos from NRC, Region III. . 14 We can begin now with the discussion of an event is on Seabrook on main steam safety valve flow deficiency, and 16 that will be given by Gary Hammer from the Office of Nuclear 17 Reactor Regulation.;
i INDEXX 18 MR. HAMMER: Yes. I'm Gary Hammer with the 19 Mechanical Engineering Branch in NRR. 20 This item concerns a problem that was exhibited 21 at the Seabrock plant which-is presently under construction 22 and has no' license. They were doing some testing on their 23 main steam safety valves, _and they did not achieve full rated 24 flow capacity. 25 The safety significance is generic in nature i
~ . . . .. - . . . . - ..
6
#1-6-SueW t because of the similarity between the Seabrook valves and A)
(_, 2 .the other valves in other PWRs. And it involves possible 3 inadequate overpressure protection of the secondary system. 4 The tests were conducted at Wyle Laboratories last i 5 Fall. And the valves are manufactured.by Crosby Valve
- 6 Company.
7 The valves weren't actually tested in order to 1 8 determine discharge flow capacity, they were actually being 9 used in a test to determine adequate discharge piping con-10 figurations and several different discharge piping arrangements 11 were tested. And for none of those valves did they achieve i 12 any more than fifty percent of the disk lift off of the seat , () 13 of the valve with the vendor as-received ring settings, 'which 14 came right from Crosby. 15 (Slide.) 16 Here is a diagram of the valve itself, or a 17 typical valve, I should say. A Crosby valve would look 18 very similar to'this. And you can see'the guide ring and F 19 the nozzle ring in this area and that area. , 30 (Pointing.) 21 The valves sit on the main steam. header. This is 22 the inlet to the valve. As the setpoint, pressure setpoint, 23 of the valve is reached the disk in this area rises and the 24 steam discharges out the discharge outlet. 26 The rings play an important role in that they are i
7
#1-7-SueW 1 dynamic control devices which help give the disk additional b' \~ / 2 lift in order to develop the full discharge capacity of the 3 valve.
4 MR. EBERSOLE: May I ask a question? I.just 5 happened to be' reading on GESSAR II that valves like this 6 are put on the system and they can't be practically tested 7 unless you really just drive it on up and test it. 8 And so you sit there until you need them, and then 9 you find out whether they work or not. Is this true? 10 MR. HAMMER: Maybe someone could help me with that 11 on the GESSAR system. 12 MR. EBERSOLE: I'm talking about in a general b 13 context. 14 MR. HAMMER: In a general way. It's true that in 15 the plant, the valves -- you-usually can't take the system 16 pressure up high enough to reach the setpoint. 17 MR. EBERSOLE: So you don't know whether they work 18 or not until you have to have them. 18
.MR. HAMMER: Well, supposedly the valve should be 30 tested on the bench periodically.
21 MR. EBERSOLE: Yeah. 22 , MR. CHERNY: I wonder if I'could make a comment ~ 23 on that? I'm Frank Cherny from NRR. 24 ("'}
\s The valves similar to these that are used on the 26 recent BRW designs, in general these are ASME Class 1 valves
8 8-SueW 1 whereas these are Class 2. They are full flow tested either 2 at Wyle or at the valve manufacturers and the rings adjusted 3 prior to shipment to the plant site. 4 That is not the case for these Class 2 main steam 5 safeties that we are talking about here. 6 MR. EBERSOLE: Uh-huh. Why is that? 7 MR. HAMMER: The Code has no requirement -- 8 MR. CHERNY: The ASME Code has no requirement to 9 do that, to full flow test these valves on a valve-by-valve 10 basis. 11 The GESSAR Class 1 thing that I was talking about, 12 my understanding is that's a GE requirement for those plants. 13 MR. EBERSOLE: But these don't require a test? 14 MR. CHERNY: As the Code is now written, that's 15 correct. 16 MR. EBERSOLE: And why is the Code written that 17 way? 18 MR. HAMMER: Well, originally it was set up so 19 that you could extrapolate from tests on smaller valves. MR. EBERSOLE: Yeah. The type test. 21 MR. HAMMER: Right. I think that was considered 22 to be an adequate method of doing it. 23 MR. EBERSOLE: liow do we know this isn't a chronic 24 ( problem? 25 MR. IIAMMER: We think it's very -- probably more
. . . . - - .~.
i 9
#1-9-SueW 1 prevalent than just the problem at Seabrook because of the z
p ' 2 similarities with other valves. ' 3 MR. EBERSOLE: On the other side of the coin, are 4 these not rated at a hundred percent ~ power on the secondary 5 side? They are full power release? 6 MR. HAMMER: They are designed I believe to carry 7 a hundred and ten percent of full steam flow. 8 MR. EBERSOLE: Full steam flow? 8 MR. HAMMER: Right. 10 MR. EBERSOLE: Is -- 11 MR. HAMMER: If they open, you know, a hundred 12 percent." 13 , MR. EBERSOLE: Right. So they haven't been reduced 14 as have the boilers and perhaps the primary relief valve to 15 account for the fact that they think the reactor will always is scram. 17 MR. HAMMER: That's true. 18 MR. EBERSOLE: Now, in the reactor -- I remember 18 in '68, I bumped into the fact that the relief safety 20 capacity was sixty-seven percent and it has just recently 21-been a deal made with ASME in the nuclear business to say 22 that it would always scram. 23 So, this condition mixed with that thesis that it (-sj 24 will always scram is not hazardous anyway, is it? Or, is it? V 26 MR. HAMMER: I beg your pardon?
- . - ~ . -_ .- - ._. . . - - . - .
10 l
,-#1-10-SueW g- MR. EBERSOLE: I say, the fact that it can't '
,)
2 meet i hundred percent requirement if the reactor scrams 3 is not particularly significant anyway, is it? 4 MR. HAMMER: Well, the Staff is looking at the 5 ' significance of --
- 6 MR. EBERSOLE
- Yeah, I guess what I'm trying to
; 7 say, we need to throw all of these factors into_a pot and a stir them all together and see whether or not we need to, or s we can, apply the trip logic to this, including the primary 10 loop as well as this; or, we go down the less conservative 11 road of saying -- which is not, of course, always true --
12 that the reactor will always scram. 13 MR. REED: Jesse, let me go back into history just 14 a little bit of how we got to a hundred and ten percent ASME 15 for all the Code valves on the secondary side of the steam 1 ' 16 generators. 17 It happened in the creation of Yankee Rowe -- 18 MR. EBERSOLE: Uh-huh. t 19 MR. REED: -- and there were -- one of the 20 designers on Yankee Rowe happened to be a_ member of the ; 21 Power Boiler Rules Committee or something. There was a great 2 deal of argument of whether or not you could ever achieve a 23 hundred percent. steaming on the secondary side with nuclear 24 reactors. O 25 And, of course, with regular boilers, fossil fired
11
'#1-ll-SueW i boilers, you have to have that capacity.
. s_) 2 MR. EBERSOLE:' Yeah.
3 MR. REED: I think a compromise was sort of made. 4 At one time, the Boiler Rules people I think -- and 5 Massachusetts was the center of this controversy at the 6 time, and a member state, the Boiler Rules Committee was 7 talking a hundred percent relieving capacity, more than a 8 hundred percent. 9 In fact, they wanted full firing relieving capacity 10 and no one knew what full firing rate on the nuclear core 11 was. a
. 12 MR. EBERSOLE: Uh-huh.
() 13 MR. REED: It might be a thousand percent. They 14 wanted it both on the primary side and the secondary side. 15 l So, after a lot of -- I think it was a case and 16 it's probably back in the record, a case that relates to 17 Yankee Rowe where they decided: Okay, we will go with the 18 determined transient need on the primary side as the 19 engineers come out with it,.which is something on the order 30 of maybe ten or twenty percent power on the pressurizer. And 21 we will go with a hundred and ten percent on the secondary 22 side. 23 It was sort of an arbitrary picked-out-of-the-air 24 decision. It could be looked at. {-
%/
26
.MR. EBERSOLE: I remember, this really is what
12
- 1-12-SueW l triggered the ATWS problem, because Oyster Creek I think O 2 was a predecessor and it had a hundred percent. Browns 3 Ferry showed on the scene with sixty-sevan. And then one 4 began to look at the curves and progressively delay the 5 scram until you began to realize that the boiler was going 6
to come apart if you didn't get the rods in. 7 And I'think it's probably appropriate to take 8 this as a trigger event and look at the whole mess and come 8 to, you know, a modern rationalization of where these things 10 should be. Maybe this is not as much -- as significant as 11 we think it we guarantee that we will always trip. 12 On the other hand, suppose it's on a -- what's 3 18 x_[' the worst of our PWRs, the ones where we relieve through I4 the bolts coming loose on the head. Is that the CE or B&W? MR. WARD: Wait a minute, Jesse. Isn't the 16 most severe challenge the ATWS? MR. EBERSOLE: Well, that was -- it never gets 18 into the secondary. I'm wrong. This would be a primary I' relief problem. 20 It's too fast to get into secondary I think. 21 But the relief logic may be on the secondary, and this thing 22 sort of -- 23 MR. REED: Trying to recall back in that case,
^
24 I have the feeling that a hundred and ten percent capacity 25 on the seconoary side of a PWR is overdone quite a bit.
13 4
#1-13-SueW g MR. EBERSOLE: Yeah. I would have guessed that.
2 It probably -- 3 MR. REED: It was a compromised decision. 4 MR. EBERSOLE: It should be more tolerant than 5 a boiler I should think. 6 MR. REED: With the negative coefficients. 7 MR. EBERSOLE: On the other hand, it would be a critical on the primary side if we had the same untested g condition and we have the worst combination of, you know, 10 moderate coefficient. 11 MR. REED: I think that's what ATWS is calculating. 12 MR. EBERSOLE: But ATWS always figures on the l () 13 14 Primary side a hundred percent release, doesn't it, or more? 15 MR. WARD: No. 16 MR. EBERSOLE: Whatever is available. 17 h9. CHERNY: Could I comment on that, too? 18 MR. WARD: Sure. 19 MR. CdSRN'4 : The ring settings on the pressurizer to safeties on the PWRs have been adjusted, or will be adjusted 21 imminently, as a result of the post-TMI 07372D1 item. EPRI 22 did a whole bunch of testing on valves like that and got a 23 real good understanding on how to set those kind of valves. 24 And we are reviewing that in great detail right 26 now on all the PWRs.
14 61-14-SueW 1 MR. EBERSOLE: I see. Thank you. 2 MR. WARD: .Let's see, for the secondary side? 3 MR..CHERNY: Primary. So, what I'm saying is, 4 4 you --
! 6 MR. WARD: Well, what does the ATWS rule assume?
6 MR. CHERNY: I guess I'm not the best qualified l 7 to talk about the ATWS rule. A lot of the ATWS calculations 8 that I have'seen assumed a hundred percent available reliev-9 ing capacity in the sense that the rings would be set cor-10 rectly on the valves. ' 4 11 MR. EBERSOLE: And one of them, I think it's 12 combustion, isn't it? l 13 (} MR. CHERNY: Combustion is the one you mentioned 14 with the head lifting and all of that stuff, yes. 16 MR. EBERSOLE: This factor here of non-testing 16 and not real validation would lead to substantially more 17 stretching, wouldn't it? 18 MR. CHERNY: Afraid so. 19 MR. EBERSOLE: Glory. 20 MR. CHERNY: Like I say, that's being addressed { 21 now. An'd those valves should be set okay, 22 MR. EBERSOLE: Thank you. 23 MR. WARD: Let me ask Frank -- I'm getting off 24 the track a little bit, but you mentioned that for BWRs the C-- 26 requirement for full flow testing is actually a vendor i
15 fl-15-SueW 1 requirement and not ASME Code requirement. O(j 2 Well, what's the difference between Class 1 and 3 Class 27 4 MR. CHERNY: Well, speaking of -- when I say 5 Class 1 and Class 2, I'm speaking of ASME Code Class. The 6 primary systems are Class 1 requirements. 7 The most stringent design, QA and so on require-8 ments. And -- 8 MR. WARD: Okay. But there isn't any difference ; i 10 in capacity or testing for capacity? 11 Id. CHERNY: Well, until very recently, that's 12 true. But as a result of the EPRI testing on the pressurizer 13 valves and so on, ASME has taken action to require in the 14 future, kind of as a -- call it a forward fit modification 15 to the Code. l 16 For Class 1 safety relief valves, if anybody 17 comes up with a new design and wants to get an ASME certifica-18 tion on it,,they will have to full flow test prototypes that 18 bracket all the size ranges and pressures and temperatures
# they are going to use them in.
21 And we will have to account for all of this before 22 they can sell it as an ASME certified valve. That Code 23 change should be published probably in the next addenda, 24 I think the Winter addenda, of the Code. V 26 MR. EBERSOLE: In the testing of these, if
16
#1-16-SueW 1 they do it in the shop or the fabrication plant, they have
( 2 to connect them to a fairly large boiler, don't they, to 3 get a blow down? Otherwise, it crashes on the seat and 4 tears itself up. 5 MR. CHERNY: Well, there aren't too many places a 6 you can do that. 7 MR. EBERSOLE: Uh-huh. So they do, in fact, 4 s find a place and they do that? 9 MR. CHERNY: There's.more facilities available 10 for the BWR valves with the lower pressures. 11 MR. EBERSOLE: Uh-huh. 12 MR. CHERNY: The PWR valves, the only facility in () 13 the U.S. was the one EPRI actually had built at Compression 14 Engineering in Connecticutt. 16 MR. EBERSOLE: Thank you. 16 Any other questions? I 17 MR. MICHELSON: A couple of small questions. The 18 fifty percent lift on the disk represents what fraction of 19 total flow? 1 20 MR. HAMMER: Well, it's theoretically fifty 3 21 percent. 22 MR. MICHELSON: No, I'm not sure that they are 23 proportional. 24 MP. HAMMER: T Well, my investigations show that c 25 it's proportional. I y ._,s _ _ - - , . . __, ,-- . . - , , , .,m. . ,,
1 17 ' 91-17-SueW 1 MR. MICHELSON: Okay. You are saying it's also , 2 fifty percent capacity, then, not just a guess? 3- MR. HAMMER: Yes. Right.. 4 MR. EBERSOLE: Now, that's just'a straight old 5 springloaded Casey Jones valve, isn't it? l 6 MR. HAMMER: Right. 7 MR. EBERSOLE: No pilots. All right.
! 8 MR. MICHELSON: Two questions. That was the 8 first.
l 2 10 The second' question, the periodic test required 11 by the Code for the safety valves, that's just a lift test; i 4 12 is that right? 13 MR. HAMMER: 'The periodic test? 14
'MR. MICHELSON: I think the Code requires.that is every so many years you take a certain fraction of your 16
} valves and check them. 17 MR. CHERNY: The set pressure test.
- i 18
{ MR. MICHELSON: Yeah. That was just the lift. i 18 And what is that fraction? One every three years,.isn't
# it?
i 21 MR. HAMMER: I think it's on a five year -- 22 Mh. CHEFRY: It's about-twenty percent of the j 28 valves each five years, something like that. 9 8 MR. MICHELSON: Each five years you must have 26 i done twenty percent? } 4
18
#1-18-SueW 1 MR. CHERNY: As a minimum, yeah. Most people 'J
, 2 do more. But that's the basic requirement. 3 You have to do a hundred percent of them in 4 sixty months. And you try and do them, you balance it 5 -about equal -- 6 MR. MICHELSON: Okay. A hundred percent of 7 them have to be done -- 8 MR. WARD: Twenty percent every year. 8 MR. MICHELSON: Every year. Okay. That sounds 10 reasonable. 11 MR. CHERNY: Every year, roughly every year. 12 MR. EBERSOLE: On the PWR with boron in the () 13 system on the primary side, what do you do to preclude 14 these things sticking up with the solid stack-um stylolites? 15 MR. HAMMER: Maybe someone else can help me with 16 that. But I understand that's not really been-a problem 17 in operating history. la on the primary side, a lot of valves have loop 18 seals. 20 MR. EBERSOLE: Uh-huh. 21 MR. HAMMER: And spring safety valves. I'm not 22 sure whether that helps or hurts. Of course, these are 23 directly in contact with the steam. 24 b] MR. EBERSOLE: . Yeah, sure. I MR. HAMMER: You don't have any boron on here. i
19 l-19-SueW 1 MR._EBERSOLE: Yeah, sure. Any other questions? ( (/ 2 If not, we used a little too much of your time. 3 MR. HAMMER: If I could continue with a couple 4 of other things. 5 MR. EBERSOLE: Sure. 6 (Slide.) 7 MR. HAMMEm Seabrook did, as a result of the a testing, go back and readjust their valves a substantial 8 amount. And they retested them and obtained full lift. 10 The generic implications, like I said, are that 3 11 the Seabrook valves are much like'those in other PWRs, 12 both operating and yet to receive a license.
, 13 The NRC has got several follow-up actions. We 14 are going to issue an information notice in the near future l 15 advising the industry of the problem. And we are going to
- 16 pursue'through the Division of Safety Technology in NRR'a 17 prioritization of the problem as a generic issue. And we 18 will also have some follow-up discussions with Crosby to 18 determine if there is any guidance that they --
! # MR. EBERSOLE: Was there a QA procedure that 21 would have done the thing had it been followed? 22 MR. HAMMER: I don't think it was a QA problem, 23 because Crosby made their best effort in setting the rings. l
, ]v 25 They had'two valves that they tested for these tests.
And both of them performed exactly the same. So,
, _ _ _ _ , _ . _ . ~ _ _ ._.
20 t
#1-20-SueW 1 Lit was repeated in that way. And it wasn't just like a O \s,) 2 QA type of quirk.
3 MR. EBERSOLE: I'm talking about QA in the context 4 of designing the test, not just following a ritual. 6 There must have been some QA problem in the i 6 setting. In the overall sense, they didn't achieve the 4 7 end result. They were consistently bad. 8 MR. HAMMER: Oh, I see. Yeah. I believe you 8 could call that a QA problem. 10 MR. EBERSOLE: Yeah. 11 MR. HAMMER: Okay. 12 MR. EBERSOLE: Okay. Thank you. ( 13 MR. ROSSI: Okay. Next we have an event at 14 , Fermi Unit 2, premature criticality event, that occurred 16 in July. 16 And that will be presented by Eric Weiss of 17 the Office of Inspection and Enforcement because Dave Lynch 18 from NRR couldn't be here this afternoon. INDEXX 18 MR. WEISS: Good afternoon. I'm sorry that Dave 20 is not here with us today, but I will attempt to cover this 21 as best we can with the assitance of Nick Chrissotimos from 22 , Region III. SS Fermi Unit 2 was being brought critical on July 24 (}
\_;
2nd by a relatively inexperienced operator, and he wound up
# brining it critical sooner than was called'for because of r
21 SueW 1 the deviation between the order that rods were to be pulled 2 on the pull sheet. 3 (Slide.) 4 The reactor was being started up in the source 5 range, and the licensee did have a full power license. And 6 the operator was pulling in Group 3. 7 lie was supposed to pull rods to the Notch 04 and 8 instead pulled each of eleven rods to the full out position. 9 The pull sheet was somewhat misleading in this regard, 10 although it had a column for initial position and final 11 position. It also had a column I understand that was relative Ly 12 large where one checked the full out, the rod latch. 13 And there was no indication in that last column 14 whether the rod should be checked for full out. So, as 15 he pulled them to full out he checked off that they were 16 indeed latched. 17 And he wound up getting some short period alarms 18 during this time that he was unaware of and was not advised 19 by anyone else in the control room. And as a result the 20 reactor went critical sooner than he had expected. 21 And when he recognized the criticality, he put 22 the rods full in. And that was essentially the end of the 23 event. 24 The rod worth minimizer was operating at the time; 25 however, the rod worth minimizer does not enforce pulling
,y . _
m _ L i
't 22
#1-22-SueW. 1 out to 04 versus 48. In other words, there was no safety
~ /3 k) 2 reason why.the rod worth minimizer should insist on that.
?
, < s. 3 . l .No safety limits were' violated during this
' '( $
, ,. i ,
! ,' ( event. Howeve.r, it'obviously' represents a management and
, ( ;
5 training problem. And the event rec'eived appropriate
\h 6 attention.by the region.
7 A confirmatory action letter was. issued. And 8 that's essentially it. I~ 9 MR. EBERSOLE: Leh me ask, could it have been i 10 that the procedure did not have a,n adequate statement of 4 s < 11 what not to do as you pulled them out?
.N 12 '
MR. WEISS: I think maybe Nick Chrissotimos A 13 could address that better than I. * ' 4 i 14 MR. EBERSOLE: This is a frequent omission, is 18 to tell,you what to do but it doesn't say what not to do. 16 And I'm just curious as to whether'to this might 17 be the case. 1 su 18 MR. CHRISSOTIMOS: Yeah. Nick Chrissotimos from ! 19 Region III. , I 80 The procedure which is - 'the sheet is a generic
- 21. sheet for all the rod pulls. And the right hand column 22
! which was larger than the rest, that's for)the coupling 23 check. And there was an omission there that was blank.
- 24 1
73 And I'm presuming that.the operator assumed: 26 well, I have to go to position-48 to make that coupling h
, *y
../ i s .e ,
23 fl-23-SueW 1 check and I will check it off. o kss 2 MR. EBERSOLE: Well, that's exactly what I'm 3 talking about. There.should have been a notation, do not. , 4 MR. CHRISSOTIMOS: That's correct. If there was 6 an N/A in there or it was Xd out, it would have assisted ! 6 him in probably preventing it. , 7 MR. EBERSOLE: All right. I think these things I e tend to be loaded with the absence of Do Nots. 9 Yeah. J 10 MR. REED: I notice that this one has also been 11 classified as an abnormal occurrence. Is that correct? 12 MR. CHRISSOTIMOS: I don't know that that () 13 determination -- Mr. Reed, I'm not sure that it has been 14 classified. There has been -- 15 MR. REED: Well, it's on the sheet, abnormal 16 occurrence. i d 17 MR. CHRISSOTIMOS: It's on the sheet of. people 18 looking into it as an abnormal occurrence. 19 MR. REED: -Well, one of the things that bothered to me about it is,,you know, reactors are going critical all 21 over the place and operators are doing it all over the place. 22 And all kinds of built-in safeguards exist, start-25 up rates, level, power level rates, count rates, and all , S4 kinds of other things that back up going critical, premature
- 25 criticality.
i 4
= *, e -
y .-y - -
24
#1-24-SueW 1 Now, I can see that some going critical, inadver-A i \ \ms/ 2 tent criticalities, are serious. But some I don't see as 3 sericus. And it's almost like some~of them, like starting 4 your car and so you turn the switch for the headlights 5 rather than for something else.
6 Now, I don't -- I'm worried that such an incident 7 as this is classified as an abnormal occurrence, because you 8 read the abnormal occurrence definition and I don't see that 9 this is a significant incident. So, I would like to have 10 somebody reflect on that classification. .' I 11 Now, I believe it has been classified that way, 12 because -- 13 MR. ROSSI: Well, you know, from time to time wo 14 do classify things as abnormal occurrences because they is constitute what we consider to be serious breakdowns in 16 management controls and training and that kind of thing. 17 And movements of control rods by a lot of people 18 are considered such a fundamental thing to the operation 18 of a reactor that when due care is not taken in moving them 20 that's a concern that it may be indicative of problem with 21 management controls and training. 22 J. T., it looked like you were going to add your 23 perspective to it. (g 24 MR.-BEARD: Okay. My name is J..T. Beard from V 25 NRR. f
~
&^ + -' 4 M4 m v m,
25
- 1-25-SueW 1 I think the real issue here on the abnormal (m 2 occurrence, Glenn, is not the event that occurred in the 3 sense the reactor went critical earlier ths.n maybe it was 4 planned. I think the real concern, at least in my mind, 5 is the Staff in analyzing transients and accidents is asked 6 by the applicant and various other people to give some 7 credit for the operator being trained in a certain way, 8 going through a simulator that mimics all that, and a lot 9 of casualty drills, well developed written procedures.
10 And the applicant in a lot of cases will make the 11 case: How can you postulate this guy making a mistake. 12 I mean, in view of all this training and procedures and 13 what not. 14 So, I think the real issue is, how much credit 15 can you give the operator for following procedures and is his training in other situations, not necessarily this one? 17 My personal opinion is, this event is not terribly signifi-18 cant. 18 But the underlying question of how much credit I
# can you give an operator may be significant.
21 MR. REED: 'The operator was told to pull to 22 something like two hundred and forty steps or notches or 23 something, right? He had a procedure that was incorrect. MR. BEARD: I think the procedure was a bit s.v/ 26 misleading, and it could have been improved substantially.
i l 26 (1-26-SueW 1 But I don't think that the procedure had what you call out- , l 2 and-out mistakes in it. 3 Nick can correct me if I'm wrong. 4 MR. REED: Well, I read something somewhere. 5 MR. CHRISSOTIMOS: The procedure was correct. 6 He failed to follow the procedure. 7 MR. REED: In other words, the number of notches 8 or steps was correct? i 9 I thought it said in one of the write-ups that 10 it wasn't, it was 240 versus 4 or something. 11 MR. CHRISSOTIMOS: Okay. That may be -- he was 12 predicted to go critical on like Step 16. f^
- ( ,) 13 MR. REED: Now, we are getting to the meat.
14 The prediction of criticality. We are getting to the meat. 15 MR. CHRISSOTIMOS: That was about Step 60, I 16 believe. And the predicted criticality step was 90 or 100 17 or something. -. 18 ' MR. REED: I don't think in'the' reactor business 18 of tomorrow and today.that we should be all hung up on so predicted criticality and where it is and whether we miss 21 it for some reason or other, by some number of. steps over 22 40. 23 And I believe that somebody has put in a limit 24 ' (-] V of more than 40. And that's what I'm concerned with, that 25 that kind of thing, the missing of predicted criticality is
27
#1-27-SueW 1 becoming a -- has become a significant enough issue,- because
+ (~~)
- \~ / 2 I might say it might be the physicist influence versus the 3 operator's influence that becomes significant enough to 4
4 make this an abnormal occurrence. 5 That's what worries me. I guess that's all I 6 need to say about it. Thank you. 7 MR. EBERSOLE: Any other-comments?. 8 (No response.) 9 Is that it? Thank you. 10 MR. ROSSI: Okay. Next we have Jim Henderson 11 from the Office of Inspection and Enforcement. 12 And I understand that he is going to change the (O j 13 order of -- he's giving two presentations, one on the 14 Hope C' reek CO initiation and the other one on Turkey Point 2 15 post trip loss of auxiliary feedwater. 16 He is going to interchange those two which will 17 keep all of the Turkey Point items together. So, Jim. j' ! END #1 18 Joe flws 19 20 21 4 23 4 7w 24 V m
2-1-JoeW 28 1 MR. HENDERSON: There was another reason, Mr. 2 Chairman, for reversing these.. I heard rumors you were 3 very much interested in this event, so I gave it a sequence 4 priority, although I have to say that I don't think it has 5 safety priority. l 6 The first reports were quite alarming. We talked 7 about evacuating some four thousand personnel from the site i 8 and so on. 9 It turned out that somebody must have been excited 10 because at it turned out, they evacuated the personnel from 11 the building complax, which was where the CO2 was released 12 rather than from the total site. () 13 They just moved over toward their construction 14 trailers. 15 MR. EBERSOLE : I will.tell you why I got interested e 16 in it. I found certain volumes had access -- critical-17 building volumes, like battery rooms, or whatever, they 18 - had access to CO2 which looked into very large storage, tank' 19 systems, and the concept was to cut of f a fraction - of the 20 total volume of CO2 with some sort of control valve, which-21 was not safety grade. 22 MR. HENDERSON: That is true. 28 . MR. EBERSOLE: Now, this resulted -- I understand 24 previously at Sequoia of blowing the doors off, and 25 presumably threatening the physical integrity of the whole l I i
. . , - + ,
1 29 l 2-2-JoeW 4 ;1- structure. i 2 MR. HENDERSON: I wasn' t aware of those events , 3 but there was one at Grand Gulf a while ago where a 4 repetitive control circuit fault occurred, and each 6 discharge was at the prescribed magnitude, but they kept I 6 on discharging more and more into the closed room until they 7 blew the door off there. .. 8 MR. EBERSOLE: So this comparative inocuous i g system had destructive potentials for safety systems. 10 And, therefore, the control devices on them l 11 might -- should they upgrade to safety grade levels, d 12 including siesmic. 13 .MR. HENDERSON: Well, that may possibly be. I 14 think we have to think quite a while about that,-though. us Anyway, the plant has not licensed, and had ! 16 no -- it did not claim to have its ventilation system 17 completely under control,- and many other things which might i is very well have mitigated the effect of this event. 19 The initiation was inadvertent.- We haven't 30 yet heard why,-but it-did happen, and as yhu pointed ~out,.. ! 21 - it discharged between five and six times as much CO2 as .it i 22 was supposed to, and on the order of probably sixty percent 23 or so of the total inventory of the storage tank. . 24 Discharge was through;the' normal discharge route, 25 but it was excessive, and naturally under those conditions' i
.L., , ,. , ,ua, -,
30 l '2-3-Jo W 1 if CO2 leaked out the door, the protected space was A ksl 2 sub-terranean. They have four diesel generators there, a and under each room -- diesel generator room -- they have 4 a day tank room, where the tanks, unlike most places where 5 they are buried in the ground, these tanks are mounted on 6 a box and fully available for inservice inspection. 7 Anyway, the gas got out and disbursed through a the building complex. People were evacuated, and there , 9- was an interesting complication to this, that they had 10 received their first fuel onsite the day before, and they 11 had to walk through drill related -- site drill related 12 to the various things that they might have to do, because () 13 a new presence of fuel onsite, and most all the' people -- 14 the construction forces, for instance, had been notified is that the drill was going to take place, and it shouldn't is .bodier them. 17 And when the real event requiring an evacuation 18 took place, it was a little hard to persuade them they weren't 19 the boy crying wolf. 20 And that is about the whole thing. The 23 21 personnel who were transported to_the hospital, we heard
~ 22 nothing more.
23 There seems to be a conspiracy of professional 24 silence between the individuals and their medical advisors. 25 Both the excessive discharge and' the inadvertent l ,
.2-4-JotW 31 1 actuation signal have been postulated to be problems with A 2 the control panel.
3 MR. EBERSOLE: Yes. 4 MR. HENDERSON: The control panel has been 5 removed and sent back to the manufacturer for evaluation. 6 MR. EBERSOLE : It seems to me, Mr. Henderson, 7 that the staff would be obligated to make a match between 8 the quality of the control function and the consequence of 9 its malfunctioning, expecially in the context of ruining 10 safety systems, like diesel buildings or whatever. 11 MR. HENDERSON: As far as the diesel rooms them-12 selves are concerned, we looked at that.
.( ) 13 This does not appear to be a hazard to them. In 14 fact, some' diesel rooms are -- and maybe these are too, for 15 that matter -- equipped with CARDOX systems.
16 You see, there is nothing related to the exterior 17 of the engine that is going to be affected by CO2. It would 18 be essentially inert, and the combustion here comes from 19 outside-. 20 MR. EBERSOLE: Ycah, but the cooling air for 21' the generator is inside. 22 MR. HENDERSON: That is still inert. CO2 is 23 inert in this respect. ,,
~
. 24 MR. EBERSOLE: But the cooling of the generator 25 . requires full once-through air transport to the building,
I I 2-5-JonW 32 1 and you have closed it up, presumably lets say under (~'\ seismic events when you have lost the AC System.
\ ,/ 2 !
3 You box in the generators, and CO2, simultaneously, 4 all over the place, and your entire emergency power system 5 is locked into CO2 in the context of refusing to cool the 6 generators. 7 Not the engine. 8 MR. HENDERSON: I am afraid I don't quite follow 9 that logic in this respect. 10 MR. EBERSOLE: Okay. The CO2 protected diesel 11 engines close the ventilation system to let the CO2 accumulate, 12 and then the normal cooling process of the generators which () 13 requires atmospheric flow through the room, open cycle, it 14 is forbidden, and'the generators have no air cooling. 15 MR. HENDERSON: Well, that may be true if the 16 discharge is to tha generator room. But if there is some 17 incidental ., as the19 would have been in this case, discharge 18 that leaks into the diesel generator room, it is not -- 19 MR. EBERSOLE : I am talking about it being 20 directly admitted. 21 MR. - HENDERSON : Now, I am not a fire protection 22 engineer. I don't ever want to be one, and I can't speak 4 23 of what the NRR position is on that. We can find out if it
,- g 24 is of --
25 MR. EBERSOLE : I don't know whether it is still
'2-6-JosW 33 f 1. with us or not, but I originally bumped into this finding, 2 mercoid switches used to trigger the devices that tripped )
l 4 3 the CO2 system commonl'y into the entire emergency complex. i l 4 These, of course, were notoriously non-seismic 5 and therefore they had triggered the situation where when
- 6 you were guaranteed your loss of power, you were also 7 guaranteed loss of cooling to your emergency diesel 8 generators.
9 Now, it wouldn't-have lasted long. You can 10 - run out and open the doors, but it would have. created a i 11 'little excitement. . t 12 MR. HENDERSON: You would be 'doing other things i , () 13 -to it at.the same time. Maybe you didn' t do that -- 14 MR. EBERSOLE: You would rather~not h' ave to do 2 that, and I think that led to the elimination of Marsh 16 mercoid switches from the' plant in general. l 17 But I think even today there is no seismic 18 requirements imposed on the: release of fire protection
- - 19 apparatus, and the CO2 system is just one among others, so 20 this says these are _ not safety devices , they are not i
21 coincident, redundant, or whatever, and they are likely to 22 go off at the wrong time simultaneously, all over the. place.
~
23 MR. HENDERSON: As a matter ~of fact, the issue 24 -- mid '82 information'~ notice citing a variety of instances, ! . C .. t 26 and'I mean a_ variety, where the inadventent actuation of fire l L 1.._. _ . . _ _ - . , , _,_ _ . _ _ _ _ _ . _ . . . . _. .~ _ _ ~ . - - - - -
3 2-7-JoeW- 34 1 protection systems had damaged safety-related equipment. 2 MR. EBERSOLE: Right. L 3 MR. REED: I am trying to catch up on this one 4 a little. I am surprised by-something. Are you saying that , 5 when,the CO2 system went off,-that there was closeup of the i 6 ventilation, air cooling system, t 7 MR. EBERSOLE: There is closeup of any system. t 8 which it is - attempting to flood. 9 MR. HENDERSON: I think there is a little confusion,
.10 - -Mr. Reed, in this~ case.
1 11 If the protected ' volume had been the diesel 12 generator room, then normal-ventilation to the diesel generator () 13 room would have been secured.to make sure that there was -- 14 sufficient concentrationLof CO2 is built up. 15 That is the philosophical approach. . In this case, 16 . though, the protected volume was the tank room. - The fuel . 17 tank room,_and any CO2 that got into the-diesel generator 18 room got in incidentally, and would not -- and the actuation i system would not have recognized that presence and isolated 19 30 that room. l 21 MR. REED: You make me feel better, .because I know 22 if you have' smoke.you still maintain your air flow through
~
! 23 the diesel generator rooms, and your exhaust may. switch or i ~ j 24 something like that from one direction to another, th2 l M point of exhaust, so depending on whether it-is coming.from l l L
- u. -
2-8-Jo:W 35 1 the turbine building or something like that. I s ,/ 2 But you have got to maintain your air flow, and 3 I would be surprised if a fire protection system had CO2 4 for fire protection for the diesel generator. I would think 5 that would be water. 6 MR. EBERSOLE: But they do, don't they.
~
7 MR. HENDERSON: I believe they do. I think it 8 in very common. I think there may be a couple of reasons g for it. 10 One, is that since oil is the primary hazard, 11 water may tend to spread it before it suppresses it, the 12 fi re . s' 13 The other is that CO2 is an easy thtag to recover 14 from. 15 There are, unfortunately, fairly frequent instances 16 where diesel engines develop little leaks. Their fuel lines 17 are small, and there is a high vibration in the engine, and 18 they start leaking at a threaded joint or something, and you 19 get oil on the hot manifold, and you have a fire, and the 20 CO2 is a very convenient suppressant under those circumstances, 21 and very effective. 22 MR. REED: But then you have to say that fire 23 and need for the diesels are not simultaneous. . 24 MR. WYLIE: That is right. 25 MR.: REED: Are those the laws of the land?
. =. . ~
9- - Jo;W 36 1 MR. HENDERSON: If there be a fire in a diesel 2 room, then I think it is a given that that diesel is 3 inoperable. 4 That particular one. 5 MR. EBERSOLE: A common mode influence, like 6 earthquakes, there are fires all over the place, and everything 7 goes off at once. 8 MR. HENDERSON: Well, I guess I throw my hands 9 up about that point. 10 MR. REED: I am trying to clarify something, Jim. 11 Are you the Henderson that used to work for Southern Cal 12 Edison? 13 MR. HENDERSON: Right. 14 MR. REED: I finally recognize you. Thank you. 15 MR. MICHELSON: A couple of questions. Was there 16 just one tank room involved, or did it set off in all four? 17 MR. HENDERSON: Just one. 18 MR. MICHELSON: And was the door to the tank room 19 open, or has the door been mounted yet?
# MR. HENDERSON: The door was mounted, I believe.
21 I don't know if-it was open or closed. 22 MR. MICHELSON: Those doors have been missing 23 until recently, and I just wondered -- p 24 MR. HENDERSON: That is a detail I didn't try.to 25 explore.
10-JoLW 37 1 MR. MICHELSON: So, it might be interesting to ls ) 2 find out what pressure the room reads before the CO2 spilled 3 Lout into the corridor. 4 4 MR. EBERSOLE : Yeah, I think , that is an interesting 5 -- ] 6 MR. MICHELSON: I don't understand why it spilled 7 out on the corridors. 8 MR. HENDERSON: I don' t think those doors are 9 totally -- 10 MR. MICHELSON: Oh, yeah, those doors are suppose 11 to take a few pounds pressure. 12 MR. HENDERSON: That is true. () 13 MR. MICHELSON: And also you state here that ten 14 tons instead of two tons was released. 15 MR. HENDERSON: , Well, another story said twelve.- 16 MR. MICHELSON: isn't that enough to -- 17 MR. HENDERSON: Five or six times, it doesn't i 18 make too much difference. It was an awful lot more than -- 19 MR. MICHELSON: Wait a minute now, slow down for 20 me. It makes a lot of difference to what pressure the room 21 reaches in the process of the release, and whether or not 22 the door was b]cwn open, or was there to. begin with, or 23 whatever. 24 This'is much larger than the planned release, 25 and so it becomes important for.you to ask:- What pressure did
2-ll-Jo Wal 38 ' 1 the room reach. ('T
%/ 2 MR. HENDERSON: I don't think anyone knows. I ;
3 certainly'do not. 4 MR. EBERSOLE : I think the critical thing is 5 we must either guarantee the discharge of no more than X 6 cubic feet, or we must absolutely guarantee a programmed 7 pressure relief function. 8 And that is not done. It is a very. sloppy 9 business. 10 18. HENDERSON: That doesn't exist at Oak Creek, 11 and I am not aware of any place that it does. 12 MR. MICHELSON: It sounds like they relieve the () 13 pressure up into the diesel' compartment. It bothers me 14 a little bit, because that~ is suppose to be a fire barrier 15 between -- we were supposed to be able to handle the fire. 16 Remember, this is the one I worried about the fire 17 starting down at the bottom, and going all the way to the 18 top of the building, and they gave us a big song and dance 19 about how good that fire barrier was between the tank room 30 and the diesel well. 21 MR. HENDERSON: It is nothing but a postulation 22 as f ar as I know. , whether the CO2 did or .did not get into 23 the diesel room.
/; 24 MR. MICHELSON: Where were the people that were U 25 l overcome? Where were they located?
l
2-12-Jo:Wal 39 1 MR. HENDERSON: They were scattered throughout the i N., 2 entire building complex. Some of them were affected, and some , 3 of them were not. 4 I don't know. 5 MR. MICHELSON: One might ask then during an 6 accident such as a propogated -- during an earthquake, for 7 instance, do I have to worry about people being overcome 8 by CO2? 9 MR. HENDERSON: I guess I am not the person-to 10 ask that of. 11 MR. MICHELSON: Apparently there is a lot of 12 CO2 involved, and it got to a lot of places, and you don't p) q 13 like to overcome operators at a time like that, when they 14 are in an emergency for an entirely different reason. 15 MR. HENDERSON: Well, I think the circumstances 16 of this event when the facility was still under construction, 17 and there were no licensed requirements, is entirely different 18 from the circumstances in an operating plant. 19 MR.EBERSOLE: But not in the context of having 20 controls over this inadventent excessive release of CO2. 21 I don't know of anything in place now other 22 than the single valve -- 23 MR. HENDERSON: That. bothers me. 24
' (N MR. ROSSI: I'think this event is part of a general \
25 possible problem with inadventent actuation of fire systems,
$-13-JoeW 40 1 and I think inadventent actuations of fire systems and ' 's - 2 problems that they can create have been recognized in the 3 past, and I think NRR, probably you ought to just make your
! 4 people aware of this particular event, that worry about 5 those kinds of problems. 6 I don't think it is something new that hasn' t I 7 been thought of before. 8 CO2 this time instead of water that we have had 9 in the past, but we have had other events and problems caused 10 by fire systems, and you know, that is a possible general 11 problem. 12 MR. EBERSOLE: Ernie, would you like to add to () 13 that? I would like to add to it. It is not really inadvertent , 14 but it is inadequate controller liability, or mitigative f 15 aspects to control the excessive amounts which may be 16 released.
- 17 MR. MICHELSON
- And it is non-seismically qualified 18 by the way, also.
19 MR. EBERSOLE: So, it is a compounded problem
- 20 that has been'needing attention a long time. I don't think 21 it is nice to blow doors off.
22 I don't even know that the doors were . intended 23 to las blown off, but thank heaven they did blow off instead i , / N of.the. floors blowing up. k 26 MR. HENDERSON: It is not truly clear-to me i
Sel4-Jo:W0l 41 1 whether the doors are open, closed or survived closed. (% ts j/ 2 MR. EBERSOLE: Any further questions on this. 3 All right. 4 MR. HENDERSON :' I have another one. 5 '(Slide . ) 6 It is related to Davis-Besse. . Distant relative, 7 but related. 8 This is Turkey Poibt, where they lost all their auxiliary feed pumps, all being turbine driven. g 10 MR. EBERSOLE: All being turbine driven. a 11 MR. HENDERSON: Turkey Point has three. 12 MR. EBERSOLE : We have another Davis-Besse, then. () 13 MR. HENDERSON : That is the only other site that 14 I am aware of that has only turbine drive, i 15 MR. EBERSOLE : Did it also use that unique system 16 of having a gear reduction to get the main feedwater booster i 17 Pumps? 18 MR. HENDERSON: No. 1 19 MR. EBERSOLE: They had motor driven boosters? 20 MR. HENDERSON : Main feedwater behaved normally, 21 and we didn't explore it. n~ MR. EBERSOLE: Do they have an electric booster 23 pump? 24 MR. HENDERSON: They have electric drive half-size 25 main feedpumps. Now, what they have in front of them, I am
i
-Sc15-JonW 42 1 not sure.
( ,- 2 MR. EBERSOLE: But they have electric drive main 3 feed pumps, right? 4 MR. HENDERSON: Electric drive main feed pumps. 5 MR. EBERSOLE: Okay, thank you. 6 MR. IIENDERSON: . And they were the saving element
. 7 in the transient. The safety significance, of course, is a the impairment of decay heat removal, and the -- as far as 9 the -- it is -- the similarity I think to Davis-Besse-is to relevant to this thing.
11 (Slide.) 12 Actually, the event occurred the evening of July 21, () 13 Both units were operating at full power, and at about 11:41 14 P.m., Unit-3 tripped. The licensee is determined that it is was caused by electronic transient from a lightening strike. 16 Why it affected only Unit-3 and not Unit-4 is probably 17 attributable to the variability of lightening. 18 All three auxiliary feed pumps started automatically 19 and responded normally. Everything was fine. By midnight 20 things had settled down and they were starting to 'take the 21 auxiliary feed pumps out of-service, and return them to n standby, and go over to the main feed pump and prepared it El for restart. 24 MR. EBERSOLE: Was it the main feed pump failure 25 that caused this?
}-16-Jo:W 43 1 MR. HENDERSON : No. On the hundred percent trip --
MR. EBERSOLE:
\
2 That was the lightening strike? 3 MR. HENDERSON : - Yeah, but for whatever reason 4 the feed pumps are tripped. Main feed pumps are tripped as 5 Part of the unit trip. , 4 6 MR. EBERSOLE : They are not ramped down to
.7 diminish 'the challenge frequency on the aux feeds?
-8 MR. HENDERSON: No.
9 MR. E EERSOLE : Why not? 10 MR. ROSSI: The Westinghouse plants in general 11 they isolate main feed water, essentially.
- 12 MR. EBERSOLE
- So they always challenge the o
() 13 safety system when they trip? 14 MR. ROSSI: That is -- i 15 MR. EBERSOLE: Is that common sense and loaic? 16 MR. REED:- There are many that I know of. The 4 17 boiler feeders stay going, you take them off. 1 18 MR. EBERSOLE: You ramp them down, and don't - 19 challenge the safety system. 30 MR. ROSSI: But I thought in general following
- 21 a reactor trip, as the temperature comes down and they close 22 valves.to isolate it, to limit excessive cooldowns, but --
23 MR. EBERSOLE : Oh.
,- /~' S4 MR. ROSSI: I don't think that means that they 25 can't be used on a fairly short' time -- continue to feed. ,
l
-. .. .. . . ~ - . . . . - . . .. .- . - - - ~ ~ - . - .
M17-JoeWal 44-1 It just takes some operator action.. 2 MR. EBERSOLE: .So it may be a benign challenge that
- 3 you can fall back on.
e 4 MR. ROSSI: Yeah. And the other thing is that 5 of course the plants.have considerable more inventory in 6 the secondary system than a plant like Davis-Besse, so 4 7 it is less of a concern here. 8 MR. EBERSOLE : Yeah. So, they went on -- { 4 9 MR. WARD: Wait a minute, can I get this straight? 10 ~I thought what you said was that in general, Westinghouse 11 plants main feed pumps tripped off. l 12 MR. ROSSI: They isolate the~ main feed. That
- () 13 . normally means they close valves. I believe. I don't think i 14 they trip the pumps, but- they close valves to stop the main 15 feed, to avoid an excessive.cooldown.
4 16 MR. WARD: Okay, the valves are closed. Glenn, t. 17 did you say something different? i 18 MR. REED: Well, th'at is not my experience, N- and I would think Turkey- Point would be sort of vintage l 2, Point Beach.. You get tdue trip.. 'The boiler. feeder is 1 21 immediately pulled fill up, because they ' are' running 'and
~
B- the valves haven't got cycled closed ~, -- the big main feed i 28 valves aren't down yet.. M. As soon.as the big main feed valves come down, ' l N ithe: flow into the steam generator, secondary side, is choked
Sel8-Jo Wal 45 1 off, and the operator normally takes off one pump and one 7y () 2 condensate pump. l 3 He normally will stay on his boiler feeder and l 4 one condensate pump. He normally -- he won't let the 5 auxiliary boiler feed function at all on a normal trip. 6 I don't know how these got asked to function. 7 There must have been something else. Maybe something happened 3 in the normal feed system. 9 MR. HENDERSON: It is post-TMI requirement.
- 10 MR. ROSSI: We can check on exactly what they
. 11 had, but I thought the later plants in general isolated J
12 flow from the main feedwater pumps when they -- af ter ()
, 13 reactor trip when they got~down a number of degrees about 14 the low temperature , and the reason for that was that when 15 you try to just fill the steam generator back to the normal 16 level after the shrink that you get with the trip, that 17 amount of water can cool you down below no-load, where you la don't want to go.
19 So, it is fairly common, I believe, at least on i to the later Westinghouse plants, to actuate the auxiliary 21 feed water on level. essentially overy time you trip. 22 - Now, you are right,_ Turkey Point is an older 23 plant and may be a little different, but that is my experience g-] 24 with the newer ones, and J. T., that is your recollection 26 MR. BEARD: That matches my memory. (
i
)-19-JonWal' 46 1 MR. EBERSOLE: Let me ask this. We are chewing -. (_,/
(D 2 around you know about challenge to reliability; for instance, S' Palo Verde is ten to the minus four per challenge, right, 4 and we are wondering whether that is enough without those 5 .PORVs. 8 Anyway, does one then measure the challenge rates 7 as inclusive of these transient challenges, which are 8 recoverable as a routine pattern back on the main feeds? 9 MR. ROSSI: I _ don' t know whether those are included 10 or not. 11 MR. EBERSOLE: It would be unfair if you can go 12 right back to the main feeds. It would be required if you bv 13 couldn't, wouldn't it? 14 MR. ROS3I: Well, you know if there isn't a failure 15 you can continue to use the main feeds. You can do something
' 16 manually to open up a bypass or that sort of thing, and 17 make use of the main feed.
~18 MR. WARD: Is there some place where there is a 19 comprehensive picture of what plants have what in this 20 respect that is documented?
i 21 I mean, it seems like this question -- Et MR. ROSSI: I would doubt.if there was one summary 23 of this particular function. Now,-AEOD may have one, because s 24 I know they have looked at -- _ the problem of challenging b 25 auxiliary feed on trips on Westinghouse plants.
p 2AJo Wal 47 1 MR. HENDERSON: When I started to try and find f3 2 out how the PWRs differed in auxiliary feed systems, I ran 3 into a blank wall. 4 Because the-Staff doesn't prescribe how the 5 system shall-be designed in detail. They say: How do you , 6 propose to design it? 7 And if'it is accep. table, they accept it, and 8 consequently practically everyone's is different. 8 In this case, it was a'further complication that 10 some of the details you get into are far below th'e level of 11
-detail that NRR tries to pass on.
12 MR. REED: I ..uve the feeling that this auxiliary 13 boiler feed flow over is related to whether you have steam 14 driven main feeders or electric driven main feed pumps.
- 15 I have the feeling
- hat if they are electric 16 main feed pumps, then there is no need to flop over. The 17 regulating valve cuts you back, and you stay.on one pump 18 and you stay on two pumps for a while, and you strip down.
18 EBERSOLE: MR. But these are electrics.
" MR.: REED: Yeah. That is why I say this is 21 l Turkey Point, it is Point Beach vintage.
22 MR. WARD: This is something that keeps coming 23 - up. Maybe we ought to.get a fellow to -- MR. EBERSOLE: John started out on this for 26 me, and he is --
la21-JotW21' 48
~1 MR. REED: Jesse, I was going to say, I think 4
() 2l this is-an important incident, and maybe the Staff ought i 3' to polish it for full committee. 4 I would like to see this happen, because as you 5 know I am on a campaign that auxiliary boiler feed is not 6 . good enough, that we need primary blowdown on some PWRs, so i
- 7. this only adds another arrow to my quiver; that we should a have redundancy in principle.
I
's MR. WARD: You;may be right, but there seems to 4
j 10 be a lack of a very comprehensive story.on what the plants 7 i-i 11 actually have in this regard. 12 I guess I would like to see this.
- 13 MR. EBERSOLE: I am surprised to see three turbine
_( 14 pumps. I wonder what the. rationale for that is? j N MR. MICHELSON: -You didn't finish' telling us 16 .about the event. I think you got interrupted part way 17 through. I fell asleep. 2 MR. HENDERSON: .No, I-didn't get very far though. MR. MICHELSON: I.got through the lightening trip, 20 ' and everything was going along fine -and behaving normally,
-M and then we got off on another-subject.-
lEnd 2. 22.
. SueW fols.
.23 i
. 94 O 25 t I o
, .,,,,,,-.~n - + , ,.,.,,n -
49
- 3-1-SueW 1 MR. HENDERSON: And then that's when things C\
b 2 started to come apart. 3 MR. MICHELSON: Okay. 4 MR. HENDERSON: Because they were controlling 5 on one -- on a single main feed pump and for whatever 6 reason, probably manual control, one of the steam generators 7 got a high level and that caused another trip. 8 And that dumped the main feed pump again and 9 called on the aux feed pumps to restart. And it was ncit' 10 very.long since they had been shut down and reset. 11 And there is an inherent vice characteristic 12 in the Woodward governors that'are used most generally.by gN . 13 the Terry turbines-and the a'ux feed systcm and the PWRs. When the governo is shut down, it tends to trap control 14 r 15 oil at a moderately high pressure, and unless it has at 16 least thirty minutes for that oil to leak off on a restart 17 it is already pre-cocked and will almost inevitably over-18 shoot. 19 Now there is typically on~these things -- if
# there is anything typical, and maybe I shouldn't use the 21 word -- on some anyway there is an electronic backup to 22 the governor, you might say, which is set at about a hundred --
23 - the pump normally runs a't 'a cion'stant speed. And at about_ ' M a hundred and twenty-five percent of full speed there is v 25
'an electronic senso'r that will bias'the governor $nd slow
+>
s, .
~
,,-- .-- p
50
#3-2-SueW 1 the thing back down below a hundred -percent and then let go m, , ,s i 'b 2 of it.
3 > At a hundred and fifty percent, there is a ply 4 , ball type ~ mechanical overspeed trip which upsets the linkage 5 and the stop valve goes closed and has to be manually, and to 6 a degree arduously, reset locally. 7 So, two of these turbines went to mechanical 8 (overspeed trip. The third ofte.was caught by the electronic 9 control, which incident.y was set pretty low, too low, and 10 the -- but for some reason, still not quite understood, it 11 cycled. It would go up to'the overspeed reset position and i 12 drop down below the full speed position and then come back t 13 up on about a three second cycle. 14 That;m ant it was pumping water, but it didn't 15 have much control'. LNow, also the feed water control is 4 16 normally controlled by a valve with a pneumatic actuator. s
' 17 On this third, pump, that actuntor malfunctioned i
' r 18 and the valve stuck.' 'So that they wound up with no aux 18 feed pump servic'eable, and they went back on the main feed i , , + pump with their lyw flow control and cooled down, and about 21 six days later after cleaning up the instrument air system 22 they restarted. 23 That's the history.and balance of the transient.
<s
/" MR. REED: I-would'like to point out to Jesse and 26 David that there is a memorandum from Baynard to them which
?
, e g,,,
51
#3-3-SueWi 1- attached an AEOD report which_ relates to feedwater transients.
2- It's this thick.
.3 And it does have a lot of information in it that 4- relates to this issue of feedwater. transients.
5 MR.-EBERSOLE: I Know there are conflicting 6 pressures on what you should do under this type of a 7 transient, and I think perhaps in the Turkey Point / Point 8 Beach vintage it was a variability and some people were very 9 much concerned about~ overfilling the steam generators. 10 And perhaps if their main flow control valves a 11 were a little: sluggish in response it was safer to trip 12 the' pump drive rather than to.take a chance on dumping the 13 water.down into the turbine. 14 MR. MICHELSON: -I was a little surprised to hear 15 about this problem with the Woodward governor that has just 16 been run recently. 17 Were you aware of-that problem? 18
' MR . HENDERSON: -Yes'. - That had'b'een called to the 19 attention'of the industry back in-'77.by Terry turbine
'E people.. And they'didn'.t tell us about it, and we didn't
. 21 catch up with it for another couple of'three.-years.
. 22 j But-it..was - .it'was-publicized and the operators at Turkey Point were aware'I'think. -I know ---
MR. MICHELSON: . 'In other words,.you are' telling 25 - { , , .you are fullyLaware of the problem and nave been for-some
, +- e ~ 4 , e - .l-, s-- . .w q-a, -, , , - - - . . . , , -
, e e-,g . , . -
52 .63-4-SueW 'l time?
't MR. HENDERSON: I know that the operators at 3 ~ Davis-Besse were aware of'it, and that awareness helped l
4 them out of their problems, because they didn't try.and reset
.5 the' governor.
6 When they needed the aux feed pump, they operated 7 on a local. manual control. 8 MR. EBERSOLE: Why do they carry-this -- t 9 MR. MICHELSON: Well, I wanted to finish my , 10 line of reasoning -- 1
- 11. MR .' EBERSOLE: Okay.
12 MR..MICHELSON: -- which I just got started. f() 13 If you are aware that this is a characteristic of the 14 Woodward governor, therefore, it's a characteristic of the 2 Terry turbines and not just in auxiliary feedwater systems
~
16 but in boiling water reactors as well, the same turbines, 17 same governor, I recall seeing a number of studies from i 18 - time to time which took credit for restart capability of 18 that governor, of that turbine, in case of trip. And. trips ~were postulated for various reasons,
~
8 [
'I 21 and they1 said they would manually ' restart. i 1
3E 'I don't recall in those studies ever1 reading-t
~
' 23 about-the fact that there is a problem here.of a thirty 24 minute delay and :that you 'may have to go back and ' operate
)
" the-turbine manually:and all the: rest'of-this: story.
~
. 4 o .
- -_:--_ _- --....-..-; .- - . . - - . . - . . . - - _ ~ . . - - . . . . . . .-. , - - _ _ . , , . -,
]
53
#3-5-SueW 1 MR. HENDERSON: 'You can eliminate that delay by 'fN And that was done at Turkey Point. \ms 2 local manual reset.
3 MR. MICHELSON: As I recall, most of those cases 4 of local manual reset, they said they went back and rigged 5 -it so they could now reset from the control room. 6 MR. HENDERSON: Well, now, in my -- 7 MR. MICHELSON: It was an electronic rig is what 8 they did. 9 MR. HENDERSON: -- wide range of-studies on this 10 thing, which have been -- that's about all they've been, 11 they haven't been very exhaustive -- the BWRs for-their 12 HPCI and RCIC drivers typically use these similar Terry 13 turbines.
\_-
14 However, GE bought added features which included 15 control room reset on the governor, including the bleed off 16 of any -- 17 MR. MICHELSON: Yeah. The feature they bought 18 was the ability to reset the mechanical governor from the 19 control room. 20 MR. HENDERSON: And they also bought a capability l 21 to reset by remotely reset the mechanical trip using, I 22 think it was a Limitorque actuator to do it. 23 MR. MICHELSON: Right. ,
-' w 24 Now --
MR. HENDERSON: 26 MR. MICHELSON: . But that doesn't help you out of
N 54
#3-6-SueW 1 the thirty minute problem.
p
\- 2 MR. HENDERSON: That was not NRC requirements 3 for it, I know.
4 MR. MICHELSON: Yes, it was on safety-grade 5 equipment. 6 MR. HENDERSON: Anyway , the f act -- 7 MR. MICHELSON: I thought it was. 8 MR. HENDERSON: -- is that the Bs and the Ps 8 differ oftentimes in this kind of detail. 10 1 MR. MICH3LSON: Well, I'm surprised. Was the 11 rest of the Staff fully aware that these turbines had i 12 this restart problem? [s ,) 13 Is that right? s MR. ROSSI: I don't know. 15 MR. MICHELSON: And have been for a long time 16 apparently, like years. 17 MR. HENDERSON: Sir, I would --
- MR. MICHELSON
- Before the full Committee meeting, 19 could you clarify this point of this should not have been any 20 surprise and that-we should be aware that these turbines
'21 can't be restarted for thirty minutes if they had been run 22 !
recently and so forth? i 1 23 Because I,.for one, was not aware. But maybe () 25 other people are. But I've read numerous studies on restart.
. u capability and I never recall running into this. And I'think:
55
#3-7-SueW 1 I would have remembered it, because it's a very unusual
-mq
}
2 characteristic.
.3 I thought the whole problem was that it would be 4 oversped on mechanical, then you could reset the mechanical 5 governor and they put this. operator on so they could push a 6 button and control them and reset it, and everything would 7 run fine.
8 I'm not convinced that just resetting the
'9 mechanical governor is going to make this thing restart.
10 But maybe you are, and you should tell us and also tell us 11 why this is a non-problem. 12 MR. HENDERSON: I would like to correct a mis-() 13 understanding I may have given you. 14 I was'not aware of this bleed off problem until 15' I started looking at Turkey Point, or more frankly at the 16 Davis-Besse. 17 MR. MICHELSON: That's the one C'm talking about, 18 of course, is this bleed off characteristic. II MR. HENDERSON: However, in the earlier days 20 aux feed was not a safety-related system and not classed 21 that way'at all. 22 In 1977,well before TML the vendor advised its 23 customers of this condition and these customers'did react 24
'( j to it and put -- developed, as part of their procedure, k.J compensatory actions.
56
#3-8-SueW 1 MR. MICHELSON: Did that notice deal with this be) 2 problem of pressure buildup and bleed off?
l 3 MR. HENDERSON: Yes. We didn't issue it. 4 MR. MICHELSON: Okay. 5 MR. HENDERSON: It was not under a formalized 6 Part 21. 7 MR. MICHELSON: Don't forget, many RCIC systems, 8 for instance, for many years now have been safety-grade. 9 On some plants, it said yes; some plants it said no. But i 10 on many BWRs, RCIC is safety-grade. 11 And does this problem exist with those? Same . 12 turbine,.same control. 13 MR. ROSSI: Let us take a look at what we can 14 find out between now and Thursday and if you have it on the 15 full Committee, why -- 16 MR. MICHELSON: Just tell me why it shouldn't 17 have been any surprise.
- 18 MR. ROSSI
- Okay. Let us check into it..
19 MR. EBERSOLE: Yeah. This would be one item we would discuss. 21 Go ahead, Charlie. 22 MR. WYLIE: At Davis-Besse had'they planned 23 and committed to add electric driven aux feed pumps and did they at Turkey Point? O) MR. HENDERSON: I don't know the answer to that,
'57 #3-9-SueW 1 but it's my understanding that the -- although the licensee 2 has been doing a pretty fair amount of upgrading at that 4
3 plant that they believe that their corrective action in i 4 this respect was to fine tune the procedure and to provide 5 . retraining for the auxiliary operators. 6 MR. WYLIE:- Yeah, I understand that. . 7 MR. HENDERSON: That's the total of what they l 8 proposed to do. 9 MR. WYLIE: I was just curious, because I mean 10
. it's significant'I believe, whether Davis-Besse in my mind 11 had already committed to add but just hadn't got around to 12 doing it.
() 13 MR. EBERSOLE: These main electric pumps that i 14
-you said have capacity, they are not accessible to the 15 diesel plants because of size, I guess?
16 MR. HENDERSON: That's true.. 17 MR. EBERSOLE: So we have here another rather 18 Davis-Besse type common mode failure loaded plant? 19 MR. HEMD15 EON: I think that would be a fair E statement. 21 MR. EBERSOLE: Yeah. 22 i MR. WARD: Are these three. aux feed pumps, are
" they three fifty percent capacity, or what?
24 Are they three one hundred percent" capacity? MR. HENDERSON: They are three hundred percent i .
, - , , w , ,r-- , - -- ,
- - - A - .M 58 53-10-SueW 1 capacity for two units; they are shared.
O 2 MR. WARD: Oh. 3 MR. EBERSOLE: Oh, these -- 4 MR. HENDERSON: Turkey Point has more sharing 5 probably than any other licensed site. 6 MR. EBERSOLE: What about the aux feeds? Are 7' they shared? 8 MR. WARD: That's what we are talking about. 9 MR. ROSSI: No, that's what he is speaking about, 10 the three turbine driven auxiliary feedwater pumps are 11 shared between -- 12 MR. WARD: Are shared? m
) 13 MR. ROSSI: That's what he is talking about, 14 the three, they are shared.
15 MR. WARD: Per unit, but they are shared between 16 the two units. 17 MR. EBERSOLE: What about their main feeds? 18 MR. HENDERSON: The' main feeds are dedicated, 19 half size pumps per turbine, per unit. E MR. EBERSOLE: It seems like we don't have much 21 difference here between this plant and_the Davis-Besse 22 . except their main feed pumps are turbine driven and they 23 didn't have any electric driven boosters either. 24 MR. ROSSI: Also the fact there.is considerably 26 more water inventory. 1
i 59
#3-ll-SueW 1- MR. EBERSOLE: Yeah, that helps out. Sure.
[)\ Nm 2 MR. REED: I would classify this as a low 3 level Davis-Besse. It concerns me again that redundancy 4 in auxiliary feed where it is so important to decay heat 5 removal didn't include redundancy in drivers for the 6 auxiliary feed. 7 Having all three drivers being turbine rather 8 than a mix of electric and turbine. 9 MR. WARD: Although the main feeds are electric. 10 MR. REED: Again, of course, I always go further. 11 I keep thinking that the day will come when we will have 12 redundancy in principle, primary blowdown.
- () 13 MR. EBERSOLE: I believe you mean diversity.
14 I MR. WARD: Well, that's his term. 15 MR. REED: Well, you can' call it diversity. 16 MR. WARD: Let's see, in this event, did the 17 feedwater ever drop to zero? 18 , MR. HENDERSON: No. 19 MR. WARD: It was always -- 20 MR. HENDERSON: It stayed within the normal 21 range.
- 22 MR. EBERSOLE
- Any further questions?
MR. DROMERICK: This is Alex Dromerick from I&E. 24 Jim, I would just like to remind you. I think that'you ()N
\__.
25 l t omitted saying that they did have electric startup pamp, l
60
#3-12-SueW 1 although it's not safety-related.
C\ (,,/ 2 MR. HENDERSON: Well, I thought about that and I 3 thought this thing is getting sufficiently complex and 4 confused right now, Alex, that that might not help too much. 5 But let's talk about it. Eight hundred and fifty 6 gallon per minute electric startup pumps and added a starter 7 at Turkey Point fairly recently. 8 It is not safety-grade. It is out in the yard. 9 It is manually isolated except when called upon by plant. , 10 MR. EBERSOLE: Not on the diesel either? 11 MR. HENDERSON: It's not on the diesel. And-in 12 a transient of the type we were describing, my judgment is 13 it's totally useless. (A) 14 MR. ROSSI: Jim, do you know if it.can be loaded 15 on to the diesel? Did you check whether it could or could 16 not be loaded on to the diesel? 17 It could not be, MR. HENDERSON: 18 MR. ROSSI: It cannot be loaded on to the diesel. 19 MR. HENDERSON: Well, if you-want to dump some-20 thing else it could be, but diesels do not have reserve 21 capacity to take it. E MR. ROSSI: Well, you know, the Davis-Besse 23 situation was that they had a startup feedwater pump l 24
/"'N that could indeed be put on the diesel via a procedure.
b 25 And, you know, they could get it on there. It 1 i
61
#3-13-SueW 1 didn't automatically get on there. But it could indeed 2 be put on there.
3 And I think that's an important thing to know 4 about Turkey Point, as to whether it can or cannot be. And, 5 you know, with appropriately taking off other loads if you 6 were to need it when you had a loss of offsite power. 7 And if it's sight hundred and fifty -- you say 8 it's 850 GPM? 9 MR. HENDERSON : Yes. 10 MR. ROSSI: Yes. I think we need to check and 1
- 11 find out that for sure on Thursday and we will get back to 12 you that in detail.
( 13 MR. HENDERSON: Steve, I guess there is something 14 there for you to do. 15 MR. ROSSI: I beg your pardon? 16 MR. HENDERSON: Steve-Elrod, Region II Action Chief 17 is sitting right behind you, Ernie. 18 He has been my primary source of information. 18 MR. WYLIE: Well, also I think it would be 20 important to know whether or not there'had been a previous 21 commitment to add electric driven boiler feed pumps. E MR. ELROD: This.is Steve Elrod, Region II. As 23 far as previous commitments to add electric feedpumps for 24 (-~s safety-grade service, I know of none.
\-
25 Their original aux feed. system, history tells me,
62
#3-14-SueW 1 consisted of three pumps. There were two suction tanks, one \ 2 for each unit. And they were basically lined up with a pipe 3 from tank to tank, and the three pumps sucked on this cross 4 connect pipe and discharged into a discharge pipe which went 5 from unit to unit.
6 And on each end of the pipe, went into each
~
7 generator. So, you know, three -- a crowfoot on one end 8 and a crowfoot-on the other end. 9 Somebody decided this was appropriately horrible 10 somewhere along the line, and there was a redesign. And 11 now they have basically two trains. 12 There are two discharge pipes. One of them --
/ - 7) 13 both of them -- each discharge pipo goes from unit to unit.
14 Each discharge pipe, when it gets to the unit where it's is going to, does discharge into all three generators. 16 So, each generator at the site has two trains of 17 aux feed available to it. One of the dischar;c cross 18 connect pipes has one pump, that's Train B. And one has 19 two pumps on it.
" In the case that we were talking about the other 21 night -- or, actually the other month by now -- the one 22 pump that was cycling on its electronic overspeed or 23 electrical overspeed was Train B. The two pumps that tripped s) 26
'on overspeed trip were both Train A.
Or, they call it l'and 2 I guess. as far as their
63
#3-15-SueW 1 actual numbering of them. They did a fair amount.of /' )
d; 2 investigation on why these two overspeed -- tripped on 3' mechanical overspeed. 4 And they were able to determine that the operator 5 did not follow procedure and did not secure them in the 6 proper order of shutting valves and have not been able to 7 repeat that event.using-some other means. The one that 8- was cycling on its overspeed, there is some enforcement 8 being taken on that, because it had been doing that for 10 awhile. 11 And the overspeed was set at a hundred and fifty 12 RPM too low. It's supposed to be a constant -- essentially a t 13 ( constant speed pump, which sounds great in the safety 14 arena. 15 Pump starts, pump pumps up to speed, the water 16 either goes in the generator or the water recircs. Pump 17 doesn't care what's. going on elsewhere. 18
- The overspeed was set too low and, you know, 19 the pump was allowed no overshoot. Well, that has been
" corrected also.
MR. ROSSI: Could you shed any light on the start-22 up of the electric pump, what it's capable of? 23 MR. ELROD: I can shed some light on it, but I O ~ "t* ' " vi ~ *" v' d"r i' ' ' *" procedures are there. I
h 64 i
-#3-16-SueW -1 The startup electric pumps were bought and in-
\ ~- 2 stalled, to my knowledge, to prevent having to run aux 3 feed pumps for plant startups. It was to save wear and 4 tear on these steam driven pumps.
5 MR. WARD: Are there two pumps? 6 MR. ELROD: There are two pumps. l 7 MR. WARD: It's not shared, no, no. Oh, yes,
-8 it's shared, yes. -
( 8 MR. REED: I would think that -- 1 10 MR. ELROD: There are two pumps, but they also 11 go to a cross connect. 12 MR. REED: I would think that the electrics 13 were important with respect to the fact that you don't have 14 a hell of a' lot of steam to start up with after refueling. 15 MR. ELROD: Well, that's right. In years past, 16 they used the booster pump on the main feed for part of it. 17 And they. decided while they had opportunity to buy equipment I8 from somebody, to get these. I' . They are powered from new buses that have more than one -- it's all offsite power, but more than one way 21 of getting offsite power. They also have a tie over to the five non-safety related peaking diesels over at the fossil 23 yard. l There'are procedures to use that tie. When you
- [/
L 25 start trying to apply safety-grade type thought to that area i'
,m. , ,- , y.-. - . _ . - . - , .. ,___e,_..-, , . - . .
6a ' #3-17-SueW 1 of endeavor, you know, the five peaking diesels aren't
\~,s 2 safety-related and whether they all work at once is some 3 question. But there is some diesel power available to 4 those pumps.
5 It is not the site safety-related diesels. 6 MR. EBERSOLE: Were these among the set of aux 7 pumps that were found not to be seismic or otherwise qualified , 8 the ten pumps after TMI-2? 8 Are these seism.irally qualified? 10 MR. ELROD: The aux feed pumps? 11 MR. EBERSOLE: Yes. 12 MR. ELROD: Yeah. i O) (, 13 MR. EBERSOLE: They are? There were ten that 14 were found.not to be so. 15 MR. ELROD: I don't believe these are part of 16 the ten. I do know that they had some turbine casing work 17 done on them several years ago. 18 MR. EBERSOLE: So, the aux feed pump was reverted I' as a safety function even though it was put together in
" this fashion?
MR. ELROD: Oh, absolutely. 22 MR. EBERSOLE: Yeah. 23 MR. ELROD: Or at least, later on it was. [J s_ = MR. EBERSOLE: All right. MR. HENDERSON : On that last question, there is a
66
#3-18-SueW 1 -little more something available. Just recently, Terry sent
~
4 2 us in what they call a Design Improvement Report, which I woul i 3 call'a Part 21 report, poorly disguised, that said that they i 4 had found that on certain of their turbines the casing bolts 5 and the mounting bolts were not torqued adequately and some i-6 'of.them weren't strong enough to take f.h necessary torque ! 7 - to seismically qualify them against a generic spectrum I i 8 which they tested them against; and that they sent us allist ! 8 of_the various facilities'that had th'ose turbines, as well as i 10 advising'each of their-customers. f II And Turkey Point was not on that list, which i j 12 raised a question. So I explored a'little bit further, and 13 Turkey Point has a -- did have a unique turbine, maybe not l 14 totally unique, but it was a turbine that ran at a lower to inlet steam pressure, appreciably lower in that steam pres-j- 16 sure than the steam generator pressure. II And they had an orifice and full control scheme I8 to drop the pressure. i I' MR. EBERSOLE: I see.. MR. HENDERSON: And they decided as part of their II ' overall-upgrade. awhile ago to put in on a new front-end, , 22 shall we say, and bring that up to the full. steam generator i < m pressure capability. - 1 And at that time they pr'obably might have did-
. this' seismic adequa'cy modification. That's why they'eren't w
i l 67
.#3-19-SueW 1 on the list. So, they apparently, from the standpoint of 3
(V 2 the Terry turbine, Terry did not express any concern about 3 Turkey Point turbines being able to withstand the seismic -- 4 and since it's in the quietest part of the country anyway 5 it's pretty safe. i 6 MR'.-EBERSOLE: Sure. But it's not quite in the 7 context of hurricanes. 8 MR. HENDERSON: No. 8 MR. EBERSOLE: And so this seems to me to hang 10 Turkey Point, both units on three turbine pumps across the II interval of time that a hurricane knocks down the other 12 lines; is that right? 13 j MR. ROSSI: Well, if you give no credit at all 14 , to the startup electric pumps -- 15 MR. EBERSOLE: , Which I've got sort of a dim 16 _ picture of how it would be hooked to the diesels. II But didn't you say it was outdoors anyway? 18 MR. ROSSI: It is outdoors. I' MR. EBERSOLE: So, I again see both units at Turkey Point hanging on three turbine pumps which, if failed 21 during the prolonged interval of an AC power outage caused 22 by a hurricane, don't I?
- Well, any other questions?
MR. MICHELSON: Yeah, I wanted to ask the gentle-v g . man from Region II as long as you'are here, and maybe you
68
#3-20-SueW- t won't be here for the full Committee meeting, were you aware n
1 ), m 2 Of.this problem with the Woodward governors and the bleed off 3 of the oil pressure and the thirty minute delay? 4 MR. ELROD: The answer to that question is yes. 5 Do I think about it everyday, probably not. 6 MR. MICHELSON: No, no. I didn't ask that. 7 MR. ELROD: Yes, because other plants have put 8 little signs on their Woodward governors saying that when 9 you secure this governor you have to bleed it off and be 10 sure you put the bleed off control back to the right spot 11 or the turbine won't run again. 12 MR. MICHELSON: Yeah, but these are all manual () 13 actions in the turbine room. 14 MR. ELROD: Yes. You are absolutely right. I 15 - MR. MICHELSON: And what you are saying is that 16 the operator in the control room, if he trips that turbine 17 really can't restart it until he goes downstairs and does i 18 these things. 19 MR. HENDERSON: Or sends somebody else down. I 20 MR. MICHELSON: Well, yeah, whatever. Is that 21 right? 4 22 MR. ELROD: If it's not right, it's close enough. 23 MR. EBERSOLE: Yeah. 24 ' MR. .bECEEECH: In those cases where the turbine is
.J 25 safety-related, I thought they had to be operable only 4
w _ _
, , . , . - ,y. , % - .-s
69-70
- 3-21-SueW 1 f rom the control room.
2 MR. ELROD: I can't answer that. 3 MR. EBERSOLE: No. In GESSAR II it says you ! l 4 have to go downstairs. 5 MR. MICHELSON: In GESSAR II, the RCIC is not 6 safety-related. 7 MR. EBERSOLE: Yes, it is. 8 MR. MICHELSON: It is? Are you sure? 8 MR. EBERSOLE: I just now looked at it about two 10 hours ago. 11 MR. MICHELSON: I thought it was not. If they 12 are, then -- well, of course, we will just have to ask them. ,,r~\ - 13 MR. EBERSOLE: Do you have a question? 14 MR. HERNAN: Yes. Ron Hernan with NRR. I 15 haven't heard it specifically mentioned. I thought you 16 would be interested to know that at Turkey Point there is II in progress right now, and should complete by the end of I8 this week, a team inspection which is coordinated by I&E.
' As I understand, it's a culmination of an B)I 20 review with a design inspection and a performance appraisal 21 It involves, as I understand, about eight to team review.
22 . ten inspectors. 23 Will it be a design assessment MR. EBERSOLE:
'^ s 24 effort?
25 MR. IIERNAN: It -- yes, that is my understanding.
I
. 71 (3-22-SueW 1 It will be kind of a hybrid, a combination of IDI, and 2 it will be the auxiliary feedwater system.
3' MR. EBERSOLE: Okay. Good. Good. 4 MR. HERNAN: And I'm told that it.'s both a 5 vertical and a horizontal slice. 6 MR. EBERSOLE: I'm sure we will want to hear 7 about the outcome of that later. 8 MR. HERNAN: The exit interview, I'm told, is 9 this Friday. 10 MR. EBERSOLE: Great. Any other questions? 11 (No response.) 12 We've got to move. Okay. O) \s , 13 MR. ROSSI: The next event is another one at 14 Turkey Point. This is the potential MSIV failure at low END #3 15 steam flow. Joe flws 16 17 18 j 19 j 20 21 22 23
. 36
4-1-JoeWel- 72 1 MR. REED: Jesse, a flashback. This is not on
'\ / 2 the list of incidents, but I might point out that Palo 3 Verde has' had an auxiliary boiler feed pump design inadequacy ,
4 indicent where Bechtel has reported that both auxiliary 5 boiler feed pumps at Palo Verde could have suffered to 6 inoperable status. 7 MR. MICHELSONi Both of them? 8 MR. REED: That is- what it says here in the 9 morning report. 10 MR. EBERSOLE : They are the ones that are leading
- 11. the team'on hypothetical reliability of aux feeds.
12 MR. REED: Well, I would just like to keep all , ( ) 13 these auxiliary feed things coming. 14 MR. EBERSOLE : I know what you mean. 15 MR. HODGE: Hello. I .mn Vern Hodge from IE, 16 and in July, the NRC received a Part 21 report from the , 17 Turkey Point Units, concerning a potential failure of the 18 MSIV's to close at low steam flows. 19 (slide,) 20 Two issues are involved in this problem. One 21 is that the licensee's review disclosed that there was an M unanalyzed condition for the main steam line break accident, 23 and then also inspectors found that they had been conducting 24 inadequate testing practice. 25 (Slide.) i
4-2-JoeWol 73 1 This control diagram shows how the MSIV closure IQ
' (j 2 is achieved.
3 Instrument error is fed through accumulators 4 and into an operating piston cylinder, which moves the 5 P i ston rod up or down to open or close the MSIV located 6 in the steam line down here. (Pointing.) 7 This spring is available to assist the valve 8 closure, and gravity also, but mostly steam flow assists e in closing the MSIV. 10 The problem is essentially related to these 11 accumulators being of insufficient capacity. 12 MR. REED: A question on these MSIVs. Are these () 13 sort of reverse flapper-type check valves of the Atwood 14 Moril turbine-type valve configuration? 15 MR. HODGE: Yes, they are. My understanding is 16 that they are double check valves, one group mounted in f ' ~ 17 reverse orientation. 18 MR. REED: Where they are spring kickoff, and 19 an air cylinder hold up arrangement? 20 MR. HODGE: Yes. 21 MR. REED: Thank you. I 22 MR. EBERSOLE : Are these valves used to sectionalize the main steam systems, and so preclude containment overloading
~
j 23 i
-s 24 in the case-you have a main steam line failure? \ ,/
26 MR. HODGE: Well --
- . - . . .- - . . .- . . - .. .-. -~. .=. - . . - . - . . . -- - - -
t 1 Is '4-3-Jo?.Wal ' 74 1 j ! p 1- MR. EBERSOLE: You know, they operate in conjunction , I with each other from each boiler, so that you don't get the 1 8 discharge of more than one boiler into the containment? t i' 4 MR. HODGE: I believe the answer to that question i 4
- 5. is, "Yes.'
6 MR. EBERSOLE : So, I am trying to get, you know, 1 j- 7 of the importance of their failure to close, and I suspect i 8 it may be containment overloading. To what degree I don't j 1
'8 understand.
l 10 Well, the pessimistic thing would:be to blow it 11
- l. up . - I' don't know that. Okay.
1 II l MR. HODGE: Alllright. The problem has applied 13 to both Turkey Point'and the Robinson plants.. The technical I4 -specifications require that the MSIVs close'in five seconds. j 15 At low steam flows, longer closures may be admissible, but , l
- le because the situation hasn't been analyzed, the
- requirement l
17 f applies. 18 l MR. EBERSOLE: If I could.take line 3, and ask i 18 [ you the question, why is that? And I am suspicious that E ~ it is due to the containment loading. II l MR. ROSSIf I think it is'because of that, and 3 - 8 that is what is assumed for'large steam line breaks'in l
# general'in computing your reactivity transient, and I l' N believe the five seconds is really requi.*d for primarily
! # ithe biggest breaks. i i- . _ . . _
4-4-JoeWal 75 1 MR. EBERSOLE: Ernie, is this one of the older
.\s - 2 . plants with how many -- four generators?
3 MR. ROSSI: This one has three steam generators. 4 MR. EBERSOLE: And what is the logic for contain- ! 5 ment loading from secondary side from them, one? The content 6 of one boiler? 7 MR. ROSSI: In general, the Westinghouse plants-4 8 are designed to limit the blowdown to one steam generator 9 by a combination of valves that stop the flow in the forward 10 direction, and check valves in the reverse' direction, and 11 I think this plant probably has a check valve in the MSIV 12 in series, doesn't it Ver'ne, in each line. () 13 MR. EBERSOLE: But is it a free-swinging check? 14 MR. ROSSI: And -- yeah, I think it just has a is plain check valve to stop reverse flow -- wait a minute, 16 maybe he has got it there. 17 (Slide.) 1 18 Ah, there it is. So, you should be able to take 19 -- wherever you take the break, you ought to be able to 30 assume a single failure of one valve, and still limit the 21 blowdown to one steam generator af ter the five second period. 22 MR. EBERSOLE: And we are talking about the ABC, 23 MSIVs now, and the other one is just a free-swinging check? MR. ROSSI: I think that is correct, yes. ('"
'O S4 M MR. EBERSOLE: Okay. Now, this free-swinging --
4-5-JoeW 76 1 MR.-ROSSI: That is correct. We just verified it. 2 MR. EBERSOLE: No'w, then. There is a hooker in 3 here. The free-swinging check is more often than not not 4 analyzed for impact closing velocities which shatter it 6 and blow the parts all into the other valve. Has this one 6 been done? 7 MR. ROSSI: I don't know whether that has been 8 analyzed -- i 9 MR. REED: Except that I might say they are 10 generally the same style, model, size, dimensions as the j 11 motor, or the air operated valve, generally speaking. 12 I know that they are in some plants. 13 MR. EBERSOLE: But they have no restraining 14 . devices to keep _them from clanging shut on their seats, 15 which are coming apart. 1 16 MR. REED: Which are just the same way the outgoing 17 ones are. 1 18 'MR. EBERSOLE: Aren't they retarded by the 19 mechanism, the piston?
# MR. REED: My experience, no.
21 MR. EBERSOLE: This came up, and my friend here 22 next door to me found 1*. at Sequoyah and it was found that 23 these valves would not survi,'e this flow, and various gross p 24 things had to'be done~to the valve to have some rationalization
- d 25 that you would not blow up the containment.
4-6-JotWal 77 1 Has that. been done here? O V 2- MR. IIODGE : I don't know. - 3 MR. EBERSOLE: Why don't we have a look. 4 MR. ROSSI: I don't think anyone is here today 5 that would know the answer to that. 6 MR. EBERSOLE : Well, this is a generic problem, 7 you know. Can these free moving valves sustain the rapid 8 closures associated with gross line faults. 9 We may be able to get rid of this by invalidating
- 10 the notion of having_ big pipe failures, but in the meantime i 11 we have to live with the fact that that is the way the t
l 12 design thesis is. ]() 13 MR. ROSSI: Well, that again has been a general 14~ recognized problem on the testing of valves in general, and 15 the conditions you may see in the worst accidents. 16 MR. REED: I might point out, Jesse. that these 1 17 valves are no different from the turbine stop valves. I 18 MR. EBERSOLE: No different from the turbine 19 stop valves. 20 MR. REED: They are very, very similar. The 21 dimensions -- l 22 MR. EBERSOLE: I didn't think these were nearly 23 so fast. Turbine stop valves have to operate in midi-seconds, r 24 MR. MICliFLSON: Different configurations. 25 MR. REED: These are -- I am not talking governor ( t I
Y-7-Jc Wal 78 1 valves. I am talking turbine stops, and they are reverse 7
! i x> 2 installed check valves, just like these are check valves, 3 and they have hydraulic operators arm in order to open them 4 and hold them open, and things like that.
5 But when they get fired to move, just like this 6 is going to go down very fast, those MSIVs are going to go 7 down very fast. 8 There is nothing to stop them. Once they get 9 Wiped into the steam flow, and that is what really closes to them, wiping in. 11 MR. EBERSOLE : Yeah, but those others are 12 designed and always operate under impact, and these never do. x (sj ) 13 MR. MICHELSON: They are designed for that 14 purpose, and there is a different -- slightly different 15 configuration. 16 MR. REED: There is a long story about the 17 original design, and I have seen them function in excess of 18 140 percent flow. 19 MR. EBERSOLE: Let me ask this. Is this a 20 generic deficiency? 21 MR. HODGE: We know the two plants, Turkey Point 22 and Robinson are involved. We are nct sure how many others 23 are involved. r~' i > 24 We know that about -- I think we know about 20
, J M of these valves have been used elsewhere. And so we are L
- .~ __ -_
4-8-Jo1Wal 1 preparing an information notice to cover that contingency. 2 MR. REED : Who is the maker of these? 3 MR. HODGE: Shutte Koeing. 4 MR. REED: Oh, these are Shutte Koeing. That 3 5 is a little different from Atwood Moril. 6 MR. EBERSOLE: I think the full committee would
; 7 be interested in the generic implication of this in the 1
8 context of jeopardy'to the containment. 9 MR. REED: I don't-understand Item 1. These 10 must be lashed up different from the Atwood Moril. My 11 experience is the MSIVs of this type, these flapper-type 12 check valves, all you do is -- take the air off from under is the piston, and then the spring kicks them off, and the 14 steam wipes them in. 15 Now, why does the air get involved here? 16 MR. HODGE: The air is used for both opening and 17 closing operations of the piston.
- 18 MR. REED : Ah, that is different from my 19 experience with the Atwood Moril.
20 MR. HODGE: Yes. 21 MR. MICHELSON: My understanding is this valve 22 never really will close if.there isn't enough steam flow. i 23 It will sit there part way.open. 24 MR. .HODGE: That is correct. l b 26 MR. MICHELSON: If there is not enough steam flow. l
s 4-9-JoeWal 80 1- MR. IIODGE : That is correct. 2 MR. MICHELSON: Which is really bad news if 3 you are worried about steam tube rupture incidents, and that 4 sort of thing. 5 MR. EBERSOLE:Well, ever since they started the 6 notion of sectionalizing the volumetric contents of vessels 7 to preclude containment overpressure, they have been in hot a water ever since about whether they can really sectionalize. g You know, when the primary loop became not the to primary volume of consideration, but the secondary side. It 11 has gotten even worse with the tube steam generator designs 12 where if you turn loose both of.them I don't think there is eg ( 13 much change at all that you will survive.
%)}
14 You know, the volumetric content of the tube -- 15 the volumentric content of the tube -- the new tube steam 16 generators is beyond the containment capacity as far as I 17 know, am I correct? 18 MR. IIODGE: I really don't know that. 19 MR. REED: I want to clarify one other thing yet. 20 These do not use spring kickoff for closing. They use air. 21 Is that what you told me? 22 MR. HODGE: Well, the spring is available to 23 assist in testing the valve. It closes about ten percent s 24 part way, to test the selonoid valves. 1 [
~'
! 25 11R. REED : I mean the main valve doesn't have
14-10-JoeWal' 81 1 a big, compressed spring on the operator. kj 2 You are saying there is no spring on the operator 3 to throw the valve down into the steam flow. XX INDEX 4 MR. ALLISON: Excuse me. My name is Dennis 5 Allison. The assumption here is that there is no steam 6 flow, and.you run out of spring force, and you just kind 7 of hang there, and you run out of air pressure on top, 8 and the steam pressure on the duft pulls -- 9 MR. REED-: Now, if the packings are okay, and 10 the shaft is not galled, and it is properly maintained, are l 11 yoo telling me that the spring won't throw this five hundred 12 pound thing down in the downward direction with gravity? ( )) 13 MR. ALLISON: It will throw it part way, yes. 14 MR. REED: Part way? 15 MR. ALLISON: Part way, yes. 16 MR. REED : Well, what stops it? 4 17 MR. ALLISON: Steam pressure acting on the valve 18 stem. 19 MR. REED: I don't understand the valve design 20 at all. 21 MR. EBERSOLE: I wonder if you can get your hands 22 on the procurement specifications for this valve which state 23 what it should have done. So we can pinpoint'the AE if we (g 24 can as being the culprit in this case. NJ' 26 XX INDEX MR. ELROD: This is Steve Elrod from Region II. I L
1
; 4-ll-Jo W~ 82 1 5
- j. it' There are a couple of things I can add that 4
o s
- s 2 would, I think,. help in this case.
1_ 3 These particular valves are very much like 1
- j. 4 drawn up there. The stem sticks down through the top cap i
i s of the valve, andins a little articulation linkage on it l 6 to hoop up to_ the middle of the clapper in the reverse i 7 seated check valve. ~ 8 So,:in this particular case it is true that the i I 9 disk weighs 'X' amount of pounds, and it is'like maybe a I 10 two inch stem, and you have that~ amount of pressure pulling-11 up, and the weight of the_ check valve pulling down. 3 m On most plants, we think, but I don't know'which () 13 ones, that have this same reverse seated check valve, it 14 doesn't have the stem sticking up through the top of the 2 valve. The clapper actually operates on a shaft that sticks l
- l 16 out to both sides, so you have equalized pressures there, i
l 17 and you have a full length spring. t 18 Such, that given no other functions, or things I 19 operating on it, the spring will, in fact, push the clapper SC all.the way shut. p 21 In this case it is a short spring. The spring
- It will not push the clapper all the way shut. I have looked I
- j. 28 at the tech manual myself, and got the impression that the l
M idea of the spring is to get the clapper down into the [ .
%) SS steam-flow.
4-12-JoeWal 83 1 Once the clapper hits the steam flow, the valve r~)
~- 2 will shut no matter what.
3 At another plant, I have seen the clapper come righ; 4 off the stem. 6 And the clapper still shuts. , l 6 You know, it will shut. Now, the safety 7 significance here really goes right back to the last 8 presentation. 9 The Part 21 states that if you had some sort l 10 of a. rupture downstream of this valve, not in containment , 11 necessarily, but downstream, this can bleed off steam 12 faster than the aux feed pumps put it in the generator, j () 13 which means that unless you have some other diverse type 14 aux feed pump that the Part 21 implies that you can run n 16 out, you can't feed the generator fast enough, and you could 16 run out of steam supply to run your heat removal source. 17 And when they say it applies to Robinson, we I 18 know that Robinson's valves also have the stem coming through 19 the top. We don't know, we in Region II, whatother plants ) 80' may have that particular thing. 21 And as far as the original design considerations, j 22 you got beyond me. 23 MR. EDERSOLE: This event here, if it were coupled J 84 to Turkey Point, would really put us in hot water, wouldn't f , \~ 26 it? 4
.-. . , . --. - - - . ~ _
4-13-JoeW 84 1 MR. ELROD: Where do you think these valves are? (_/ 2 MR. EBERSOLE : Wait a minute, did I miss that. 3 MR. HODGE: These are Turkey Point valves, yes. 1 4 MR. EBERSOLE : - And here, I am using my steam ) 5 from my aux feeds if I have -- 6 MR. ELROD: That is why the Part 21. 7 MR. EBERSOLE: Oh, brother. 8 MR. ELROD: This thing started off with the Region 9 being upset because they were not isolating the instrument to air when they were testing the valve. 11 Tech specs didn't really say it had to be 12 isolated, instrument air, but we were finally able to come 13 up with the fact that the ASME Code said they ought to do (V) 14 that. 15 MR. EBERSOLE : You know, this brings up the to issue, when do you think a plant is in sufficiently bad 17 shape just to shut it down and require that it be fixed. 18 I have heard some extraordinary things like over-19 turning braces onto the steam generators, where it was so really out of this world, unreasonable to shut them down, 21 but I think you are borderline, if not beyond the need to 22 do it. 23 MR. ELROD: I don' t want to conment on that one . 7
-s 24 MR. EBERSOLE: Yes.
I \' 2 MR. MICllE LSO!!: Is the system safety-grade at the
~ 4-14-JoeWal- 85 1 plant?
3 MR. ELROD: No. 3 MR. MICHELSON: The accumulators are safety-graded, j 4- at least. i. 6 MR. ELROD: _Yes. 4 MR. MICHELSON: And then don't they have -- yeah, 7 they have double check valves on the accumulators, and what a you are saying the accumulator didn't hold enough air to e . accomplish the objectives. 10 MR. ELROD: Yes. 11 MR. EBERSOLE: Ernie, when do you get an aggregate 12 of sufficient bad things like this to say, well, that is () la enough, shut it down and fix it? 14 I don't want to treat these things singularly la but you know, on block. le MR. ROSSI: I don't know the answer to that 17 question.. Obviously, what we try to do is look at the events 2 from what each one shows us, and what sequences show us, and is find the ones that are generic and do something about them 30 before we reach the point where we have a severe problem, e 21 MR. EBERSOLE: Isn't this about at the stage that at Davis-Besse is, although something that happened at Davis-BB Besse is somewhat more spectacular. 94 MR. ROSSI: We have indicated we will come down 35 on Thursday and give you more information on' it.
4-15-J eW31 86 1 MR. EBERSOLE: I am sure the full committee will A 2 want to hear it. 3 MR. ROSSI: What I will give you on Thursday is 4 more information on the design of the auxiliary feed water
- 6 system, its capabilities, and any plans for any future changes ,
4 That is what we will -- 7 MR. EBERSOLE: Good. So, that will surely be 8 one we will pick up. 8 Any other? Let's move on. j 10 MR. REED: I guess -- on Thursday, are you going i ! 11 to take up this MSIV situation. i .
- 12 MR. EBERSOLE
- Sure.
() 1 i 13 MR. REED: I would like to have flashed on the 14 board a picture of the valve so we understand the main steam 18 line valve more completely so we understand what has to fail 16 here to make these fail. 17 I, quite frankly, don't -- am not as worried about 18 the fact that they don't go all the way close. If you have 19 any large breaks, I know they are going to go close, but 30 I have to think about lesser flows of steam out through a 21 partially closed valve. st MR. EBERSOLE : Let me request you put these 23 two items together on Turkey Point, as differentiated from 84 l the fire protection system, as These are coupled problems, okay?
4-16-JoeW21 87 1 MR. ROSSI: Fine.
,fg 4
(/ 2 MR. MICIIELSON: What prevents the valve from 3 reopening. Is this the differential pressure across the 4 flapper, I guess. 5 If you were to, for one reason or another, 6 equalize the pressure, and I can think of various ways of 7 doing that, then is the operator capable of relifting the 8 disk? O YR. IIODGE : Yes. 10 MR. MICIIELSON: Because that is the way it 11 normally would work, I would think. It is the way you 12 lift the disk. iO) 13 I will think about that. Thank you. 14 MR. EBERSOLE : Any other questions? 15 (No responso.) 16 MR. EDERSOLE: Okay. 17 MR. ROSSI: Okay. Next we have another Turkey 18 Point item on recent inverter failuros, and this one will 19 be given by llenry Dailey, of the Offico of Inspection and 20 Enforcement. 21 XXXX INDEX MR. DAILEY: I am lionry Dailey from It.E, and I 22 want to talk about another Turkey Point event. 23 (Slido.) 24 Turkey Point has experienced a number of invertor f"J')
'w 25 failures over the -- in the last few months from April 4th
1 4-17-JoeW31 l 88 1 through August the 1st, there were sixteen inverter trips \/ 2 and five of these trips resulted in reactc- trips. 3 The safety significance here is that each time 4 you get a trip you get a challenge to the safety systems, 5 and also you get an interaction between the two plants, and 6 you get a potential for dual plant trips. 7 There was one occasion where an inverter from a the No. 3 plant tripped, which caused a trip of Unit-4, g and there was another case where this was suspected. 93 This interaction is caused by the way the plants 11 share these DC buses. Of course, this plant is pre-Reg Guide 12 181, which addresses the sharing of plants. r'% I believe somebody mentioned earlinr that we have (V ) 13 14 a lot of systems shared at Turkey Point. 15 (Slide.) , H5 Itere we have the two battery chargers per plant 17 with a spare for each plant. We have two batteries per is plant, and these all go down to four DC buses here, and 19 each DC bus feeds an inverter from the No. 3 unit and the so No. 4 unit also, this case here. (Pointing.) 21 And the -- then each inverter feeds one of the n four 120 volt AC vf tal buses, one of four busen for each plant u here. , x End 4. 24 ( ,) SucW fois. 25
i l 89 65-1-SueW 1 And there is a switchover here where you use the (. ( -) 2 spare bus in case one of the inverters was out. A typical 3 scenario for these inverter trips has been that the turbine 4 runs back on a rod drop signal. You get a high T ave due 5 to loss of automatic feedwater control, to loss of automatic 6 rod control. 7 And then you get a high or low steam generator 8 1cyc1 due to loss of automatic feodwater control. Then, 9 you get a reactor trip due to real and/or falso signals. to The safety injection actuates due to falso signals. 11 The pressurizer spray valvo fails as is. Dioscis a 12 start, do not load. () 13 Then, the spara inverter is put on line and return 14 to normal. The cases that I mentioned earlier about the to intoraction, there has been one case whero over hero on the to 3C, the 3C invertor was being put back on lino and the 17 procedure to put it on line wasn't followed correctly in 18 that the capacity wasn't fully chargod up before it was put 19 on the line. 20 This caused the input breaker for this inverter 21 to trip. The surge on this bus caused this input breakor 22 to this invertor to trip, which caused the trip of the 23 Number 4 plant. ry 24 (Pointing.) U # Those invertors are old Exido inverters which
t 90 i 45-2-SueW 1 are being replaced by Solid State Controls, and at tuis time 2 the job is about completed at this stage. 3 MR. EBERSOLE: Before you take that down, that ! 4 4 other one, could I ask you a question, please? I l 8 MR. BAILEY: Yes. 8 MR. EBERSOLE: The inverters lead into what is i 7 commonly called fail-free buses, don't they? f 8 MR. BAILEY: The -- l
! 8 MR. EBERSOLE: And that -- and the connotation of I .l j 10 fail-free is wrong. Really, it's free of switching transients ,
1 11 And it tends to fail more often than many other buses. It MR. BAILEY: Well, that's -- 13 MR. EBERSOLE: Is that right? f . 14 i MR. BAILEY: Well, that is what happened here in 18 this caso. The old -- I might add, the old design for tho ( ) i 18 invertorn didn't have an automatic switchover hero. It had -- l l 17 it's like, if this inverter failed an operator had to go down. l 18 I MR. EBERSOLE: It scoms to me that the switchover r j there invalidatos the inverter function. 18 j MR. BAILEY: The operator had to go down and 31 manually switch over to the spare invertor. Now, the new ! i 8 i inverters that are being put in has an automatic switchover. MR. EBERSOLE: Without losing a sixty cycle wave? i
" MR. BAILEY: It switches over -- I forgot how fast it switches over. Maybo Stovo could help me. Is ho thoro ?
l l
91
- 5-3-SueW 1 MR. EBERSOLE: Well, anyway let me ask you
() V 2 something. On the -- 3 MR. DAILEY: It is a fast switchover. 4 MR. EBERSOLE: I want to ask you, what is -- why 6 is the fail-free bus and why for an inverter? 6 If I look at all of the loads you've got up 7 there, turbine runback, high T-ave, et cetera, et cetera, 8 et cotera, I look at them and I say probably most of them 9 will stand switching transients, and cortainly with a little to help they would by putting tenderlays in. So, why the hell 11 are they all in an invertor bus instead of on an offsite 12 power stable bus, which sits there and has a reliability of O 13 stablo -- of operation which far exceeds the inverter bus? 14 Am I correct? 15 MR. DAILEY: That's a good point, but it's just 16 the design philosophy hero is to have these instrument busos 17 on those inverters. 18 MR. EDERSOLE: Yeah, but I'm trying to get to the 19 root reason. I found out long ago that the only thing that noods to bo on thoso so-called switch-froc busos is the 21 clocks and the computers. 22 And ovarything also should bo somewhero olso. 23 I would like to hear you try to challongo that. 24 (sT MR. DAILEY: Woll, I'm not going to try. V
- MR. EDERSOLE: Ernia, can you challongo that?
92 95-4-SueW 1 MR. WYLIE: I wouldn't challenge it. I agree 2 with you. But I think if you will look at the Oconee i 3 design which uses Exide inverters also, they also use a 4 regulated power supply that is -- the output is connected 8 to the output of the inverters through a Solid Stato switch. i j e MR. EBERSOLE: Well, all I'm saying is they are j 7 asking for their millions of dollars in shutdown by leaving 8 it this way. 8 MR. BAILEY: I believe the Oconee design has a l 1 1 10 design where the -- if you lose inverter it switches directly ! 11 j over to -- directly to the --
- 12 MR. WYLIE
- The offsite power.
18 MR. BAILEY: It switches directly to AC. So you 14 don't got it -- 18 MR. WYLIE: Tha,t's correct. 18 MR. BAILEY: -- from DC but yet you still have a 17 power supply. 14 MR. WYLIE: That's correct. 1 I8 { MR. EBERSOLE: lihon you do that, are you better
" off if you leave it there?
{ 21 l MR. WYLIE: Sometimes they run that way. But I 1 at j was going to ask the question, you know, what's the naturo of the failure of the inverters? [ MR. BAILEY: Woll, thoro have been a whole host , I, of differont roanons. Blown fusos sooms to have boon a lot I
I 93 4
<#5.5-SueW 1 of them. Loose connections, dirty contacts. It doesn't .t
[d \ 2 seem to be any one thing that -- 3 MR. WYLIE: Yeah. 4 MR. BAILEY: There is some thought maybe they
; 8 are just getting old.
4 MR. EBERSOLE: You know, one thing that appears 7 to have happened is a subtlety. Somebody gave them the name ) 8 fail-free, and everybody and his brother went running to it j i 8 on the connotation fail-free meant just that, not transient-j i 10 free. ] I l 11 And they hung all this junk on them. And we have 12 4 been living with the consequences which has boon operationally I8 very costly over since. I 14 ! MR. BAILEY: What you would say is to put the 18 clocks on the inverters and forget the rest? ! I8 MR. EBERSOLE: Put the clocxu and the computers l i , i 17 ! on the inverters and take the rest off and put them on off- [ te f! sito power so that they would not go out when the turbino 18 j tripped.
" MR. WYLIE: I think that if you go back about fivo !
21 years, you will find that at Oconoo they had similar problems
" as to what you are describing here.
I N i And what they found was that they had maintenanco 5
" problems. And that it took a consorted offort to retrain tho
. people in how to service those inverters.
i 3
94 i
'#5-6-SueW 1 Now, I think Oconee still has the Exide inverters.
f 4 i 2 MR. EBERSOLE: Well, is it appropriate for the 3 Staff, Ernie, in the interest of safety to ask this question? 4 Why do you have all this junk hung on the inverter
, 6 when it fails as often as it does, as compar'ed to how it a would work if it were on offsite power?
1 7 And you examined each load for its ability to 8 withstand switching transients. l 8 MR. ROSSI: Well, yeah, that may be an appropriate j 10 question. I think on the later plants, the plants are less i 11 prone to having problems with -- l 12 MR. EBERSOLE: Oh, they still hang them on in-13 herters, though. ) 14 MR. ROSSI: Yeah, they hang them on inverters but I j 18 I think they have them split up better so that -- 18 MR. EBERSOLE: We are still carrying that fail-free 1 17 ghost-like thought that it's more reliable than offsite. I 18 don't think it is, is it? I 18
- MR. ROSSI
- Experience may indicate it hasn't
" been, I guess.
21 MR. EBERSOLE: Yeah. l MR. ROSSI: J. T., do you have anything to add
" to this?
MR. BEARD: Jeano, my memory was, or is, that in some earlier years a lot of the instrument tochs and tho b
95 I.95-7-SueW 1 I&C designers that were around wanted as clean a power as 2 they could get. 3 MR. EBERSOLE: For those that needed it. 1 4 MR. BEARD: And they were not concerned so much 1 8 about loss of power as they were with what you might call 6 clean power versus dirty power so that every time somebody
; 7 flushes the john or whatever, the power trains just don't 8 go down on the btra.
9 MR. EBERSOLE: But you --
)
10 MR. BEARD: But the problem that I see that I 11 think exists is, by putting all of these things, hanging i 12 them on batteries, DC buses, inverters and switches and l 13 solar transformers and so much, I think people have got the 14 figment -- or the perception -- which may be inaccurate, l 18 that these things may never fail. j 14 And that's absurd. 17 MR. EBERSOLE: Right. 18 MR. BEARD: They will fail. And in the kinds l' of events that Henry is talking about here, what happened l 8 was the inverter died, the bus went dead. Some of thre l 81 l instruments that were hanging on the bus, a few things
- M like rod drop signals, so you end up with a runback and l l
l 8 all sorts of scenarios like Henry was describing here be-
" cause failures are'inevitabic.
" Now, on these inverters wo learned at Arkansas and i
[ i
96
- 5-8-SueW 1 other plants like Oconee had similar experiences where V 2 these buggers are very sensitive to voltage fluctuations 3 on the front end. If the DC voltage on the bus goes a 4 little high, which it will when one coupled load shifts, 8 it will blow the input fuses.
4 MR. EBERSOLE: Uh-huh. 7 MR. BEARD: And they are just extremely sensitive 8 that way. 9 MR. EBERSOLE: Well -- 10 MR. BEARD: There are some of these inverters 11 that are designed with both DC and AC inputs, and if you 12 turn one on before you turn the other one on, you blow the is inverter. , 14 And I think that the whole basis of it though was, 18 number one, the instrument guys wanted clean power as good is as they could for a lot of things that, like back in the II old days, neutron instrumentation, they didn't know how 18 well it would work on raw AC power. I' MR. EBERSOLE: Yoah.
" MR. BEARD: And they knew they had a lot of heavy 31 equipment in the plant that would be coming on and off; and,
" the second thing was, there is a period of time in the
" accident analysis between when you lose offsite power and
" you get the diosols cranked up that you need instrument
" power to do such things as turn the diesels on.
97
#5-9-SueW 1 MR. EBERSOLE: Uh-huh.
(
\ 2 MR. BEARD: So they needed -- and I don't remember '
3 the term fail-free but I remember the term uninterruptible -- 4 MR. EBERSOLE: Yes.
- 5 MR.-BEARD: -- which was an intent to say that s the bus would not ever get uninterrupted -- or get interrupted .
7 MR. EBERSOLE: Uh-huh. 8 MR. BEARD: But they had to tido you over from the 8 loss of offsite power until you got the diesels cranked up J 10 again -- 11 MR. EBERSOLE: Yeah. . 12 MR. BEARD: -- and the inverters with DC backs { () 13 were the answer at that time. 14 MR. EBERSOLE: Sure. Right. 15 MR. BEARD: I think you've got a valid concern 16 but there are a lot of other considerations that have to 17 l be thrown in the pot -- 1 18 MR. EBERSOLE: Yeah. All right. 19 MR. BEARD: -- and stirred up, too. I E MR. EBERSOLE: All right. Fine. 21 MR. ROSSI: I think some of the later plants have 88 ! equipment divided up better and use two out of four logic - 23 instead of one out of X in a lot of these places, like on 8' this -- particularly on rod drops so that the loss of a
)
88 single inverter isn't quite as much a problem. i
98
,#5-10-SueW1 MR. EBERSOLE: Yeah. Okay.
E) 2 MR. BAILEY: Well, Reg Guide 181 would not permit 3 this. , 4 MR. EBERSOLE: I see. Yeah. 5 MR. ROSSI: I mean, you know, Turkey Point again 6 is a relatively older plant. 7 MR. EBERSOLE: Yeah. 8 MR. WYLIE: One thing further about the Oconee 9 inverters, I mentioned it is a maintenance problem. And to what they found is that these inverters are quite complex 11 pieces of equipment. 12 And the Exide is not readily designed for mainte-G
) 13 nance like Solid State. I notico you are going to replace 14 with a Solid State Control Company inverters here.
15 MR. BAILEY: Yeah. 16 MR. WYLIE: And after Oconee, Duke used Solid 17 State throughout, and has as far as I know up to today, 18 on all of their plants. 19 MR. EBERSOLE: Yeah.
# MR. WYLIE: But those are different. They are 21 designed for maintenance. They are a plug-in modules. If 22 you've got problems you pul3 one out and you --
23 MR. EDERSOLE: Yeah. 24 MR. WYLIE: But you've got to know this, these 26 pieces of equipment. And I think a lot of problems with this
99
- 5-ll-SueW 1 'Exide is.that the mairtenance people are afraid of them.
b) \_,/ 2 ~They've got to be trained on them. They are going to have
'S to know those --
~
4 MR. EBERSOLE: It would seem worthwhile to at a least survey the nature of the' loads that are on the fail-6 free or uninterruptible or whatever you call it, bus, and 7 see which ones really should be there. 8 MR. BAILEY: Well, it's all your_ instrument 9 power. 10 MR. EBERSOLE: Yeah. 11 MR. BAILEY: Your vital instrument power. 12 101. EBERSOLE: Yeah. Well, certainly you've () 18 got to ride through an AC power failure. 14 MR. BAILEY: Yeah. 2 (Slide.) 16 Well, I guess the good news here is that eight 17 of these inverters have been replaced and the remaining 18 four inverters will be in place shortly. 18 And I believe that since they've got these eight 30 new ones in place, I'm not aware of any trips they have had. 81 These new inverters are designed to minimize the interaction 88 betwoon the units that we discussed before. SS MR. EBERSOLE: Uh-huh.
'M MR. BAILEY: And as a result of having all these as trips, of course, the procedures for the loss of these 120 volt
100
#5-12-SueW . 1 instrument panels has been reviewed, revised. And operators I \s,/
O
- 2 have been retrained --
4 3 MR. EBERSOLE: Yeah. 4 MR. BAILEY: -- as a result of just having to do j 5 it. ' . 6 The generic applicability here is under investiga-i 7 tion. We have not seen any other plant have this degree 8 of problems with inverters, although we -- all of them --
).
8 most of them have had some small problems but I think I J } 10 remember Oconee having some problems recently. 11 MR. WYLIE: Well, they had several years ago, like 12 I mentioned, and they had a training program for all of () la their maintenance people. 14 But also they installed a Solid State inverter 8 down there. And that has been about four or five years I is ago. And they did that to get comparable experience with
! 17 the two inverters. ! 18 MR. BAILEY: Are there any other questions on j 18 this or comments?
30 (No response.) i 21 MR. ROSSI: Okay. Henry is going to do the i 88 next one also, which is an offsite power problem on Point as Beach 1. 88 MR. BAILEY: Yes. O. 80 (Slide.)
101
#.5-13-SueW 1 The next event is an event that occurred at 2 Point Beach Unit 1 on July the 25th of this year in which 3 they had a loss of offsite power.
4 The loss of offsite power resulted from a failure 6 of this low voltage auxiliary transformer, 1X04. 6 (Slide.) 7 The safety significance here is not particularly 8 great in that this was a single failure. Systems operated {
- 8 as they were designed to operate. However, it's interesting 4
10 in that this single failure which caused the loss of offsite 11 power -- which caused the loss of offsite power -- could, 12 if the unit had tripped with it, have caused a total loss 13 of AC power except for the diesels. 14 As it turned out, the unit stayed on line. And } 16 these non-safety buses stayed energized from the -- or, j 16 could have stayed energized from it. They actually were i 17 put over here in preparations for shutting down. 18 (Pointing.) l l I' MR. EBERSOLE: It looks like the generator was 8 tied into the fault at 1XO4. How did it escape? Did it 81 just clear? 22 MR. BAILEY: How is that? 23 It looks like the generator was MR. EBERSOLE:
- 24 tied into 1XO4.
26 MR. BAILEY: No. This is open here.
102
#5-14-SueW 1 MR. EBERSOLE: Oh, I'm looking at the Unit 1
[) (/ 2 main generator. That's -- oh, I'm sorry. I'm wrong. I 3 see the break in it. 4 MR. BAILEY: This is your high voltage -- 5 MR. EBERSOLE: Okay. 6 MR. BAILEY: -- auxiliary transformer here, 1X03 -- 7 MR. EBERSOLE: Okay. 8 MR. BAILEY: -- which feeds the low voltage -- 9 MR. EBERSOLE: So it cleared up there. 10 MR. BAILEY: -- auxiliary transformer, 1XO4. This 11 feeds the 1A03 and 1A04, which -- 12 MR. EBERSOLE: Okay. 13 MR. BAILEY: -- feeds the safeguard buses. 14 MR. EBERSOLE: Right. Okay. 15 MR. BAILEY: Th9 generator feeds the unit trans-16 former which feeds the 1A01 and 1A02, which are the the non-17 safety transformers. 18 MR. EBERSOLE: Uh-huh. Yes. 18 MR. BAILEY: Of course, when this transformer here tripped and lost out on a -- actually, it turned out
~
21 to be a signal of sudden pressure rise indication which 88 turned out to be false and later on in the transient this 23 transformer was found, in fact, to be operable. 24 So, the transformer was brought back into service and the buses were energized from this transformer, which could
l 103
'#5-15-SueW 1- conclude the event.
(3 U 2 About fifteen minutes after this was lost, the 3 unusual event was declared because they were without offsite 4 power for over fifteen minutes here during this time. And 5 the licensee commenced a shutdown, because the tech specs 6 require that if either this transformer or this transformer 7 for Unit 2 are lost, then the associated unit must be in hot 8 shutdown within three hours. 9 As I said before, all systems functioned as 10 designed. And the interesting thing is that with this<de-11 sign, which we don't have any reason to believe it doesn't 12 meet GEC 17, as I'm aware of -- is that your impression, J.T.? 13 But with this design,~if the unit had tripped'at 14 the time that they lost offsite power, then you would have 16 been without -- you would have been totally without offsite 16 power in the -- or, you would have been totally without.AC 17 power except for the diesel generators, which wouldn't have 18 been too good. 19 Are there any questions or comments?
# MR. BEARD: Henry, you might want to add on that 21 one that the backup source that is referred to in CEC 17, 22 for either of the units, is through the startup transformer 23 for the other unit.
24 MR. BAILEY: Here, you mean. Yeah. 25 MR. BEARD: That's the reason why you onded up
i 104
#5-16-SueW 1 with such tech specs and things of that nature. i 2 MR. BAILEY: The backup source for this IXO4 3 transformer is from the 2X04 transformer, but it is done i
I 4 manually. It's not an automatic fast -- it's not an 5 automatic transfer. e 'And this was I understand done -- designed this t 7 way to prevent a fault over here dragging this -- both units 8 down at the same time, which has a lot of merit, too. 9 MR. EBERSOLE: Any questions? j 10 (No response.) 11 Why don't we call a ten minute break and come 4 j 12 back at 20 after? i j 13 (Whereupon, a recess is had at 3:08 p.m., to i- 14 reconvene at 3:224p.m., this same date.) 9 15 l MR. EBERSOLE: ' We will. resume the meeting. i 16 MR. ROSSI: Okay. If we are ready to go, the next 17 one we have is going to be given by Tim Colburn and it's 18 ~ a Point Beach item on safety injection recirculation line
- '19 deficiency.
.INDEXX # MR. COLBURN: My name is Tim Colburn, and I'm the II Project. Manager for Point Beach..
i . 8 l (Slide.)
'On July 23rd, 1985 while doing a post implementa-tion review of their emergency operating procedures, Wisconsin 8
f Electric Power Companyi licensee for Point Beach, discovered I f { .i m
1 e 105 45-17-SueW 1 what they considered a potential single failure which could x (-) 2 lead to disablement of both safety injection pumps. I think 3 they said it's a design deficiency. 4 What would happen would be failure of the control 5 circuitry for either safety injection recirculation line 6 isolation valve, SI 897 A or B, which are these valves here 7 and here, which come in a common line from the safety in-8 jection pumps back to the refueling water storage tank. 9 (Pointing.) 10 If the control circuitry for either of those valves 11 failed, it would cause the valve to fail close and additionall / 12 would deprive the operators of indication that the valve had
,m
( ') 13 failed close in the control room. v 14 Subsequently, if a safety injection pump were then 15 running, both safety injection pumps could overheat and be 16 damaged. And it could happen in as little as three minutes. 17 MR. EBERSOLE: This is to permit the pumps to 18 run with an open -- with a minimum recirculation? 19 MR. COLBURN : Minimum recirculation, right. 20 Seventy gallons a minute minimum recirculation flow. 21 These were what are termed intermediate head 22 safety injection pumps. 23 MR. EBERSOLE: Uh-huh.
, ' ~' 24 MR. COLBURN They get a signal about seventeen 25 hundred forty-five pounds to start and they -- their shutoff
106
#5-18-SueW 1 head is about fourteen hundred and ninety pounds, so that rx .(_,) 2 if you had, for example, a small break LOCA there might 3 be some period of time between the time they started and 4 there were actual flow into the core.
i 5 MR. EBERSOLE: Is this a rather standard design ' 6 configuration? Or, don't most of them have their own in-7 dependent recire bypass? 8 This is communized, isn't it? 9 MR. COLBURN: I don't know if this is common to 10 other plants or not. .There is an information notice being 11 prepared and it's going to be issued imminently which will 12 alert other licensees to the problem. () 13 But.it may be similar to other plants of this 14 vintage. 5 MR. EBERSOLE: Uh-huh ~. Okay. 16 MR. COLBURN: The problem would really only be 17 applicable for spurious safety injection signals and small 18 break LOCAs where the primary pressure would be maintained 19 above the safety injection pump shutoff head. 30 If you had a large break LOCA, for-example, you 21 would get flow into the core immediately. 22 MR.:EBERSOLEi Sure. , 23 MR. COLBURN: And you would not have the problem. i s 24 There is also associated with these valves an interlock with (v) 25 the containment sump isolation valves. s
107 i
#5-19-SueW 1 And what the interlock does is that if either
,q \
k ,) m 2 of these containment sump isolation valves leaves its closed 3 seat, then these valves go shut. And the reason that inter-4 lock is there is to present a pathway from the containment 5 sump back through to the refueling water storage tank, which 6 is vented to atmosphere providing a potential pathway for 7 release to the environment. 8 That also poses another failure mechanism which 9 would shut either one of the valves.- What the licensee has 10 done is a short-term fix to take care of the problem, is
. 11 they have manually opened both of these valves so that they 12 now are not connected to the operators.
.m 13 They've also locked shut both of these valves p) 14 and placed, like locking devices on the operators, both 15 at the remote station and in a control room. And at the 16 local station for these valves so that they can't be operated 17 without the locking devices being removed.
18 (Pointing.) 19 And they put plaques to inform the operators so exactly under what conditions these. valves should be it - 21 j, opened, an'd also to remind them that before these are openkdthesdvalvesshouldbe: shut.
^ .r .
22 23 x' MR.'MICHELSON: Are those the valves in the e 3 s.. ii 24
.t
('s . ye , containment sump which youfpointed to?. t 25-
N ,
! MR. COLBURN5i Yeh.'.This 851 A and B here are these i
is- g. ,
'I g < ,
( ;
^t i.,
I J. .. t y , ' t '; 63 } -{._m..- '5 g
- i.
I i 108
#5-20-SueW 1 valves from the containment sump which --
p) i' - 2 MR. MICHELSON: Well, when do those have to be 3 opened? 4 MR. COLBURN: Those are opened after a LOCA when 5 you've gotten recirculation -- 6 MR. MICHELSON: How -- < 7 MR. COLBURN: -- phase of an accident, approximately 8 maybe -- I guess if the refueling water storage tank level 9 would go down to about -- 10 MR. MICHELSON: Okay. But are those valves 11 accessible following a LOCA? 12 MR. COLBURN: Not from the manual -- from the () 13 local station but from the control room. 14 MR. MICHELSON: No. But did you lock them l 15 locally? 16 MR. COLBURN: Yes. 17 MR. MICHELSON: Well, then you have got to go 1 18 back locally to unlock them, don't you? j 19 MR. COLBURN: Only for the manual operators. You I ( 20 can still operate them remotely from the control room. 21 MR. MICHELSON: Well, what did you lock? l-22 MR. WARD: Just the local control is all that is 23 locked now. ()
\~d .
24 MR. MICHELSON: Right. You just locked out the 25 local control station,.not the power?
109 1#5-21-SueW 1 MR. COLBURN: That's correct.
/ % 's- 2 MR. MICHELSON: Okay. Thank you.
3 MR. COLBURN: In the control room -- 4 MR. MICHELSON: I misunderstood. I thought you 5 were locking out the power down there. 6 MR. COLBURN: No, no. The control room has 7 switches for these valves. Plastic covers have been put
- 8 over the switches. Those plastic covers are locked shut > 8 also.
10 And that has to be-unlocked before you can use 11 the switches in the control' room. 12 MR. EBERSOLE: Why do I -- I'm looking at the j~m. () 13 . boric acid tanks up there. 14 Do I have to admit boric acid coincident with 15 flow from rewater flowing? Or, is that for -- what is 16 that for? 17 MR. COLBURN: That's just part of the system 18
- print.
18 MR. EBERSOLE: It doesn't imply that I've got 20 to have a coincident flow from both of these to get -- 21 MR. COLBURN: N o .' , 22 MR. EBERSOLE: -- a borated -- 23 MR. COLBURN: No, no. i
/ MR. REED: It may in conflict of interest, but O)
I will try to answer -- ( l t
l l 110
#5-22-SueW e~
1 MR. WARD: Boy, I will say,
\~/ 2 MR. REED: -- a question. l 3 (Laughter.)
4 The boric acid tanks come on first -- 5 MR. COLBURN: That's right. 6 MR. REED: -- after a break to throw all the 7 concentrated boron that you can at the seam, and-then it 8 switches over to the refueling water storage tank, which 9 is also borated about two thousand. 10 MR. COLBURN: Right. These valves here are 11 interlocked with these valves here such that initially these
; 12 valves would.open.
(m) 13 (Pointing.) 14 And then when the boric acid tanks have finished 15 supplying their concentrated boric acid initially, then 16 these valves interlock to go shut and these valves would 17 open, then supply water from the refueling water storage 18 tank. 19 MR. EBERSOLE: Is the rationale here that I will 30 get either A or B boric acid tank, and that that valve down 21 the middle, that single valve, is marked M, motor, is always 22 open? 23 MR. REED: The rationale is that there are three 24 tanks. One is for Unit 1. They are A, B and C. (-s} v
" One is dedicated Unit 1, one is dedicated Unit 2,
I t-111 i , #5-23-SueW 1~ and one is a spare, because you get some problems with S 2 dilution. 3 Not, let me ask'the question. In the discussions 4 that you've had, the reason for those valves over there,
- 5 is that what we are talking about, those valves over on the 6 -extreme right?
2 7 MR. COLBURN: (Pointing.) 8 These valves, yes. t 8 MR. REED:- The reason they were put in -- and 10 it was not original design, as I recall. I believe they i 11 were put in because of dilution flow from the boric acid 12 storage tank winding its way and getting into a boric acid ( 13 mix tank, which is, what, ten times as concentrated. 14 Now, it appears that that -- did someone say that
- 15 that was a modification in,about 1973 or '
74 or something? 16 MR. COLBURN: I don't-know when the modification 17 was made. I8 MR. REED: It was a modification, however? 18 ! MR. COLBURN: I believe it was, i 20 i MR. REED: It had to do with dilution of the 21 tanks. 22 MR. MICHELSON: Well, you need them for. containment 8 l isolation anyway. They are recirculating containment fluid f through-the safety injection pumps, and you certainly can't-as be dumping it to the tank or you would have to shut those l l F
= - -
n w.. - - , w - - - -
112 j
.#5-24-SueW 1- valves if you are coming from the sump.
, rh k ,) 2 MR. REED: Well, that's a mini-flow, but -- 3 MR. MICHELSON: But if you are coming from the 4- sump, clearly you do have to isolate the refueling water 6 -storage tank I think. 6 MR. REED: Oh, yeah. 7 MR. MICHELSON: So they were needed from the 8 beginning, and maybe in '73.they discovered it. 9 MR. REED: But that's a mini-line. That's not 10 the main line. The main line is where the big black motor -- 11 MR. MICHELSON: Yeah, but even that mini-line 12 can move a lot of radioactivity out into -- () 13 MR. REED: Well, I was just searching my mind for 14 'the reason they were put in. I just draw it to your attention . l 15 I think they were put in because of dilution flow, 16 because of elevation of the tanks, relative elevations, dilu-17 tion flow somehow getting into the boric acid mix, concentrated 18 tanks. 19 MR. COLBURN: That may be correct, although the , so FSAR does not assume that you need the high concentration i 21 from the boric acid tanks to respond to an accident. 22 MR. EBERSOLE: Yeah, but you are committed in 23 your tech specs to a certain number. And if you violate
'~N 34 the eleven percent, then it's a reportable event.
25 MR. WARD: Back to Jesse's question, then. Is that l l l r L-
113
-#5 SueW g a safety-grade system?
w
.is_) 2 MR. COLBURN: Safety. injection, yes.
3 MR. WARD: Yeah. I mean, no, the boric acid 4 A and B? 5 MR. COLBURN: I would have to check on it. 6 MR. WARD: I mean, that valve down there -- 7 MR. COLBURN: (Pointing.) 8 This one? 9 MR. WARD: Down below that. The single, is that 10 valve always open or what? 11 MR. COLBURN: Yes, that's a normally open 12 valve. () 13 MR. WARD: It's a normally open valve. So, 14 that's the single failure'-- you are not vulnerable to single is failure in that system. 16 MR. EBERSOLE: This third bullet you have up here, 17 the failure of control circuitry also fails the control room 18 -annunciator. 19 MR. COLBURN: That's right. 20 MR. EBERSOLE: So, there was a built-in system 21 that masked the failure of the -- exhibit of the failure? ! 22 MR. COLBURN: That's correct. 28 MR. EBERSOLE: So, that's a design deficiency? l ( ) 24 MR. COLBURN: That's the deficiency. Not only i 25 would a single failure disable both pumps, but the operators
l 114 l 15-26-SueW 1 would be unaware that they had the failed condition on the
.\m-f)/ 2 valve. In other words, the valves could fail shut and if 3 it was a control circuitry failure that caused the valves 4 to fail shut -- there is other failure mechanisms, but if j 5 it was a control circuitry failure, the operators would be 6 unaware'of it.
7 MR. EBERSOLE: Right. Was that in contradiction 8 with the narrative that described the system that was just 4 9 found that that wasn't true? I 10 MR. COLBURN: This was discovered during a review 11 of the emergency operating procedures. And I don't -- I'm 12 pretty sure, you know, that the fact that the failure of () 13 the annunciator so that the operators would be unaware of j 14 the failure is something new that they just discovered. 15 I think they always -- or, at least somebody may 16 have known that this was a single line with two valves and 17 that both safety injection pumps tapped -- you know, the recira lo lines tapped into it. i 19 MR. EBERSOLE: Uh-huh. ! 20 MR. COLBURN: So that clearly if one of those shut, 21 you would no longer have any. safety, or any recirc flow. 22 MR. EBERSOLE: But it would be invoked that you 23 would know that was the case? I4 MR. COLBURN: That's right, because there are other --
/
( 26 besides the annunciator, there is status lights and everything L
115 95_-27-SueW 1 else in the control room for these -- ( . s 2 MR. EDERSOLE: Uh-huh. 3 MR. COLBURN: -- valves. But when the status 4 lights are failed by the failure -- 5 MR. EBERSOLE: Wouldn't the pump come up against 6 a zero flow head and that would annunciate a very high 7 pressure? 8 MR. COLBURN: But you would expect that, though. 8 If, for example, you had a small-break LOCA -- 10 MR. EBERSOLE: Yeah. 11 MR. COLBURN: -- and you knew, or even a spurious 12 start of the safety injection or pumps, some signal, you () 18 would expect that, yes, the pumps started and, yes, the 14 pressure hasn't decreased low enough yet so that I've got 15
-flow into the core.
16 You would expect that condition. 17 MR. EBERSOLE: Well, what I mean is, if these is valves were shut and there was no break and you were on 19 test, wouldn't an auxiliary indication be a red light on
" excessive pressure at zero flow?
21 END #5 Jos flws 22 23 w S4 . s_, l 26 l I L:
116 6-1-Jo2W 1 MR. COLBURN: No. 2 MR. EBERSOLE : They didn't have that? 8 MR. COLBURN: No, they would run at shutoff head 4 until such time as they actually got pressure in the primary 5 system decreased below 1490, where you actually start getting a flow into the core. 7 MR. EBERSOLE: Then they can run. 8 MR. COLBURN: You got the 70 gallon minimum 9 recire, which allows them to run at shutoff head. 10 MR. EBERSOLE : Well, that is not shutoff head, is 11 it, if it is flowing. I am talking about real shutoff head. 12 MR. COLBURN: All right. It is not quite shutoff 13 head, then. 14 That is minimum -- Mi MR. EBERSOLE: When you turned it off, what kind 16 of a pressure jump would you get? 17 MR. COLBURN! I don't know the answer to that. 18 I don't think, because of the short time frame for the pumps 18 to destroy themselves, I don't think anybody is going to test so it. 21 MR. EBERSOLE: You are right, you would have to 22 really trip them. SS MR. REED: Do you happen to know what size that N line is? My recollection is something like one one inch. (V-') 36 MR. COLBURN: Two inch line. i
i 6-2-JoeW21 117 1 MR. REED: Two inch? With an orifice down 7..
/ )
(_ / 2 to what? 3 MR. COLBURN: I don't know what the orifice size 4 is, but it restricts the flow to seven gallons per minute. 5 MR. REED: And these pumps are what, a thousand 6 gallons a minute? 7 MR. COLBURN: I thought they were 700. 8 MR. REED: Yeah, I know they test out 700 lined 9 up with all the restrictions. 10 Okay. 11 MR. WYLIE: Let me ask. I don't understand your 12 comment about the failure of the control circuitry. Also -- ( ) 13 are you saying that the signal for the valve to be shut or 14 opened is not taken of f the valve -- the switches themselves, 15 is that what you are saying? 16 MR. EBERSOLE : You mean a physical position switch. 17 MR. WYLIE: Yeah. I mean, is that what that 18 means. I am not sure what fails control room annunciator 19 means. 20 MR. COLBURN: The power supply for the control 21 circuitry is the same power supply for the annunciator and 22 the '/alve position -- 23 MR. WYLIE: That is sort of unusual, isn't it? y 24 MR. COLBURN: Yes. t i N) 26 MR. UYLIE: I mean, most annunciators have a
6-3-JoeWal 118 1 1- separate instrument power supply, and only take contact 2 outputs from limit switches and such as that. ! 3 MR. EBERSOLE: TMI threw a lot of light on this l L r 4 at the PORV level, because it didn't know where the valve i 3 was. 4 It was just an indication that you threw energy ! i 7 at it, and didn't know whether it worked or not. i 3 Ernie, is there a requirement now that if the e control circuit f ails, a separate power aspply is .used to give i 10 you fosn a physical position switch, or something, J 11 or are we still subject to a control power failure, say, i, i 12 at 'a PORV which both makes it quit, and lose its amumciatLon. i () 13 MR. ROSSI: Well, I think for critical items 14 in general that there is a requirement that they be separate,
! 15 but I don' t know that there--
I j 16 MR. EBERSOLE : Separate power supplies. 1 ! 17 MR. ROSSI: Yeah. But across the board -- I 1 f up -- J. T., do you know? j 13 MR. BEARDS. One of the things that I remember, Jesset, i 20 was that obviously everybody has looked at the PORVs. j 21 There were other plants, especially in the world i j 23 of B&W reactors, where they found things like the atmospheric j 33 steam dumps could fail. Go to the open position. Create , T j f s. 34 a significant transient, and also blind the operator in the i Q. 35 sense that he didn't know the valve was opening if he couldn't I f e . , ,- - _ _ - ,
. , , . , - _ _ . . ..,,,,y_. ,._,% y._,_ ., _ . -
4-JoeW 119 1 1 hear it, and what we did was for critical items that we could
) 2 identify, such as atmospheric steam dumps, we made certain j
i 3 that while the steam dumps weren't safety related, and 4 therefore weren't required to be single failure proof I themselves, what we did do is make the utilities change it s a so they couldn't have both the malfunction and blind the 7 operator at the same time. .i , a So, he had to at least be aware of the problem : j e so you could at least expect the operator to take some ; i c l 10 compensatory measures, whereas if he had been blinded, you . j 11 can't expect him to do anything. l 4 I l 12 . MR. EBERSOLE: What you just said was you do require ! i () 1 is two power supplies; one for' the function, and one for the ] 14 i anmnciator.. * ! i u MR. BEARD: We have not required two power supplies, I to Jesse. What we have done is just made sure that loss of l i l 17 power, loss of air, those common everyday things like blown 7 i i
- to fuses, cannot result in both loss of annunciation or loss of ;
i
- to indication, and the plant. transient together.
l 30 In some cases, it wasn't power supplies. I i' j St MR. EBERSOLE: Yeah, yeah. The function itself, l ? ) 3R sure. l 33 MR. BEARD: Yeah. One comment I was going to ' l p, as .make earlier, and maybe since I have got the horn now I will 35 make it, the design you see up here, notice there are two i i -
^
6-5-JoeW21 120 1 valves in series. k_ l 2 Now, that is a dead giveaway that somebody thought 3 that the important function is to shut off that line. 4 And given that you convince yourself that that 5 is the important function, you very rapidly convince yourself 6 that is the only important function, and then you begin to 7 convince yourself that failures that cause that function to 8 happen prematurely are no never minds, or at best, they are 9 enveloped. 10 So, that is the kind of set up that you will find 11 in a lot of plants where control system malfunctions result 12 in a valve going closed. They say, well, that is just a r~N (v ) 13 premature safety action, so it is a no never mind. 14 It is obviously intended and hoped to do that way 15 because the great white father in Washington has blessed 16 it. 17 MR. EBERSOLE: Yeah. , 18 M R'. COLBURN: But I think this is one of those 19 situations where -- like we ran into this problem four or 20 five years ago on the plants out in California, where they 21 had exactly this, a spurious safety injection signal could 22 come on and bring the ECCS system on, and also bring it on 23 immediately into the recire phase, and shut these valves. On the CE plants they have these things called ( r~)/ 24 s recirc actuation signal, which does this very thing, and
6-6-JoeWal 121 1 the corrective action way back then was disabled data is p h 2 an automatic feature. If you do get to a big LOCA and have 3 to close them, or get to the recirc mode of ECCS, you got 4 time enough to do that manually, but you really, on those 6 valves, there is no fail safe position. 6 Because in one case you are talking containment 7 isolation, and in another case you are talking about burning a up ycur ECCS pumps, so there really is no fail safe e condition. 10 And that is the problem that I see in this, is 11 that we get locked in on this thought that there is some 12 fail safe, and that all failures in that direction are fine. 13 MR. EBERSOLE: Yeah. 14 MR. REED: I think my history is coming back is to me. And I now think I remember why this was done. It is was done because the requirement was that you should shoot 17 concentrated boric acid at the accident of a large break. 18 Now,-the deal was the relative elevations of to these tanks were such that those check valves just downstream 30 of the restrictors, would leak backwards and what would at happen was the pipelines, because of slight packing leakage, at seal leakage on the pumps would get diluted -- and you see ss this is carrying a technicality a long way -- they get diluted se down in concentration, so you might have less than full as concentration boric acid tank level in the lines to shoot at k
122 6-7-JoeW21 1 the accident scene, and as I reflect on the setting here it l l . (_) 2 looks like it was over cooked to try to prevent a rather 3 questionable dilution of a line. ! 4 And the other thing was to get back into the 5 tanks because there are periodic test requirements to cycle 6 those closed valves you see up there in parallel, so that 7 is why this was done, and now modifications are sometimes 8 over done. l 9 MR. EBERSOLE : Yeah. 10 MR. COLBURN: Right, as I said, the short term 11 fixes are to keep these valves open, and if you had a 12 spurious safety injection signal or a small break LOCA
/ N
( ) 13 you would always get flow back through here until such time x_/ 14 as you needed -- until such time as you brought the refueling 15 water storage tank down into about six percent needed to get is into the recirculation phase, at which time you would then 17 manually shut these valves and open these from the control 18 room. 19 MR. REED: Either you or I as a part of the 20 regulatory should call the present people up here and tell 21 them they have now undone, or or intending to undo something 22 that was done to prevent dilution. Do you understand what 2 I am talking about. 24 ('~'3 MR. COLBURN:I understand. Right now the licensee v' 25 is looking at long term fixes, and their schedule is to
(-8-JoeWal 123 l 1 have something chosen and submitted to us for approval by
)
/ 2 the end of September.
3 Westinghouse was informed of the deficiency by 4 letter dated July 24th. 5 MR. EBERSOLE : Can't they just lock the valves 6 open? 7 MR. COLBURN : Pardon? 8 MR. EBERSOLE: Just lock them open. g MR. COLBURN: Yes. 10 MR. EDERSOLE: Any further questions. 11 MR. COLBURN: They are manually overridden. You 12 have to go to the local station to shut them. (.
;v; 13 MR. MICHELSON: Can you get to them during the 14 loss of coolant accident?
15 MR. COLBURN: Yes. 16 MR. MICHELSON: We are beginning to add more things 17 for operators to do, more things for them to remember and 18 so forth. 19 MR. COLBURN:This is a short term fix, and they 20 are looking into a long term fix right now. 21 MR. REED : Well, you know I have the answer to Z2 that. 23 MR. MICHELSO!!: Give them an examination. (^N 24 MR. REED: Think a new principle. v 25 MR. EBERSOLE : Any other questions?
i I 124 6-9-JoeW 1 MR. ROSSI: Okay. The next item will be given d 2 by J. T. Beard of the Office of Nuclear Reactor Regulation, 3 and he is going to talk about a couple of problems that are 4 related at Maine Yankee on common mode problems with 5 steam generator indication. XX INDEX 6 MR. BEARD: Now -- 7 MR. MICIIELSON: Events based on two months of a experience, is that about right? g MR. ROSSI: Yeah, I think that is -- 10 MR. MICI!ELSON : The last meeting was about two 11 months ago? 12 MR. ROSSI: It is basically July and August. - 13 MR. MICIIELSON: This is two months of principal 14 events. la MR. ROSSI: Right. 16 MR. BEARD: Okay. 17 (Slide.) la MR. BEARD: There are two events at Maine Yankee 13 that I would like to bring to your attention. Primarily 30 because they are common mode problems, not because by 21 themselves they were so terribly significant. 22 The events involved the instrumentation on the 23 steam generator measuring its pressure. The first event turned out that it was discovered in August just as the f] 34 36 P l ant was coasting down at end of cylce.
.P 6';-Jo;Wcl 125 1 The plant was down to 78 percent, and if I
( 2 remember the steam generator pressure was down to 650 or 3 something like that, and they noticed one channel was I 4 sagging down about a hundred pounds lower than that. 5 So, they called their I&C people in and they 6 'found out that root valves weren't open. 7 MR. EBERSOLE: Let me ask a question about 8 that.
\.
9 $ MR. BEARD: Already? ! 10 MR. EBERSOLE: Yeah. Why is it that -- these l ] 11 are root valves, you know, little faucets. Aren't they, or 12 should they not be always rigidly safety wired, fully open? t () 13 MR. BEARD: May I come back, Jesse? j 14 MR. EBERSOLE: Yeah, go ahead. 1 15~ MR. BEARD: After they found this one transmitter
- 16 - had its root valve, quote, not fully open, they checked the i 17 other transmitters, and nine of the twelve pressure trans-18 mitters had their root valves not fully open.
19 I am going to come back to that in a minute, but 30 let me go on. 21 MR. MICHELSON: Did they tell you what not fully l 22 opened meant?
'M MR. BEARD: Yes, sir.
i j S4 MR. MICHELSON: Okay. Are you going to tell us? ! (
- \_
- 26 MR. PEARD
- Yes, sir. I am-trying to do this k
s 6-ll-JoeW21 126 1 newspaper style. I am going to hit you with the big stuff ( ) 1../' 2 and then come back and fill in the gory details. So bear 3 with me a minute. 4 The second problem they found is that after they 5 shut down in August, when they found this problem, or 6 shortly thereafter, they were in the refueling outage and 7 they did their 18 months surveillance on the pressure 8 channels, and they found out that lo and behold, they would 9 not trip. 10 And the reason they wouldn't trip is they put 11 in a post-THI improvement and they didn't design it right 12 and they got themselves caught, n () 13 It had to do -- I use the word, ' grounding.' 14 Actually, it is instrument commons is probably a technically ' 15 better word, but you know what I am talking about there. 16 Of course, the safety significance of this is 17 simply that your pressure instrumentation would not have 18 functioned properly for a steam line break accident, and 19 coupled with the fact the plant had run the whole cycle, 20 15 months or thereabouts, with this unknown to the operating 21 staff, and we do need to say that in terms of safety 22 significance, there are diverse backup instruments such as 23 steam generator level. M 7'T r ; There are trips in the primary side that are
's.J 25 going to cause a re-entry trip. Now, they arc going to
6-12-JoeWal 127 1 respond slightly later, but they will occur before you get
/"
k ,%) m 2 into any big troubles. 3 3 Now, let me go back to the details. 4 Okay. I made up a little matrix over there to I 6 try to show you exactly which valves on the root valve 6 problem were affected.
; 7 (Slide.) , 8 Okay. It was the D Channel on steam generator 1 9 that they first noticed the problem on, and you will see 10 from the matrix over -here on the other cartoon, the D Channel 11 for steam generator 1 was one that had its root valve i
12 essentially fully closed. Fully closed. () 13 It is interesting to note that although the root 14 valve was fully closed, the instruments in the control room 16 were all reading good, solid, normal values, okay? 16 And when you first hear that, it may come as a 17 surprise, but when you stop and think, valves are designed 18 to leak. Or they all turn out to leak. 19 So, that what happens was is where the valve was , se really torqued down tight and closed, or just not the 21 prescribed number of turns closed, there was enough leakage 22 getting by that on a steady state basis, it would basically 23 charge the line up and read the proper pressure, so that l 24
- r% the crux of the problem really comes back to
- What will it b 25 do when you have a transient?
I l _.y -n. , . 4 w,-- m - w q - ~ ,- -u--
6-13-JoeWal 128 1 of course, obviously with a big pressure drop 2 across the valve it is going to respond slower. 3 The reason they got into this box was they did 4 a hydro test refueling a year ago, and they closed all these 5 various valves and if you will notice over here - 'where 6 did that little stick go -- here we go. 7 The nine affected valves are all in Channels B, 8 C, and D. They have four instrument channels for each of the 9 three steam generators. The A ones were not affected. 10 An interesting tidbit was after they did the 11 hydro, they had another modification, or another piece of 12 work, rather, that involved I think some Appendix R work, () 13 but it specifically said in the procedures go back and open 14 these root valves, okay? 2 - So, because that procedure said it, the A Channel 16 was okay. 17 The other thing that is worthy of note here was 18 the root valves, and this is the answer to your question a 19 minute ago, Jesse, the root valves were not on any 30 procedures. They were not on any check list, and the reason 21 was, and utility management just says this flat out, their 22 assumption was that if the instruments are reading -- and 23 that is prima facie evidence that the root valves are open, 24 right? 25 Wrong. i i t
I i 129 l
..6-14-Jo Wal !
1 MR. EBERSOLE: Wrong. I 7._
'\ s, ) 2 MR. BEARD: And they recognized that they had a 3 fallacious assumption. They have egg on their face, and 4 they are ' tiki6g rather significant corrective action about 5 it.
6 MR. EBERSOLE : What about a common requirement, 7 though, that you have a standard ritual for opening and a shutting these things, and when you do either one, especially 9 opening them for the running mode, that you make a visual 10 record of this by safety wiring them, by drilling the handle 11 and winding wire around it, causing it to be open and locked 12 open with safety wire? (Oj 13 Then you see it every time you pass it. 14 MR. BEARD: That is one way to solve the problem. 15 MR. EBERSOLE: Right. 16 MR. MICHELSON: You can't use safety wire, though, 17 Jesse, because I am not sure that these are all traveling 18 stem and so forth, and it depends on the design of the 19 valve -- 20 MR. EBERSOLE: I am assuming you turn the handle, 21 or faucet. 22 MR. BEARD: I think it is like your common garden 4 23 . faucet.
<3 24 MR. MICHELSON: Yeah, but it takes more than part U 25 of a turn, so how do you tell whether it is in its first
6-15-JosW21 130 1 turn, or second turn, or third turn.
'ss 2 MR. EBERSOLE: Oh, you don't do that. You just 3 say that it was visibly locked in place.
4 MR. MICHELSON: After somebody made real darn sure 5 it was open. 6 MR. EBERSOLE : Yeah, right. 7 MR. MICHELSON: Oh, okay. 8 MR. BEARD: What you could do, and you know as 8 an engineer like everybody else, I tend to be prescriptive 10 because I have my own ideas on how to solve it. I am trying 11 to avoid that. 12 But on these valves, turning it stop to stop O ( j/ 13 is five and a half turns, and one way you could solve the 14
. problem is turn it to -the right position.
15 Usually, if you are careful, you can look and 16 you can see the unoxidized part that sticks out. It is 17 small, but -- and then you just safety wire it, normally not 18 with wire but something plastic like a seal wire on your 19 electric meter, so it can be broken. i 20 l That is the way to do it. We had the same problem 21 you remember on the hydrolic control units with scram valves.
- 22 MR. EBERSOLE
- The best thing to do is to do it 23
) with wire with a lead seal, with the guy's initials in it, 24 (~'s and he prints it. V 25 MR. BEARD: Well, we could ask him to leave a o
6-16-JoLWal 131 1 pint of blood there, too.
\_ 2 MR. EBERSOLE: Yeah.
3 MR. BEARD: Let's see. They did run some tests 4 on these instrument root valves, and interesting enough 6 depending on how far open or closed they are, they 6 either sec no perceptible change, and these were not rigorous 7 time response tests, but just eyeballing it, and some of I 8 them were up fairly slow. ! 9 The worst number we heard was it could be as 10 much as 30 or 60 seconds slow. i i 11 So, that is the story on that one. 12 MR. EBERSOLE : I think the full committee would () 13 be interested in the significance of root valves not being 14 regarded ao being on. 15 MR. BEARD: Let me add one more sentence, and 16 maybe I can help you answer it. 17 The second half of this story I am going to give 18 to you, I am going to tell you why all the A Channels didn't 19 work, and the third -- the next.part is this particular l 20 licensee believes this is a very generic problem, at least 21 to the family of Yankee reactors, and they think it is very 22 widespread. 23 We have taken a little bit of a survey, very c s 24 informal, just to give.ourselves some feel of it, and we l- s 25 called two utilities this afternoon, and they said you have L 1 l'
~ . __
l l 6-17-JoLWal 132 1 to be kidding. If you don't have your root valves on a O - %- 2 check list of procedures, you are in the stone ages. 3 MR. EBERSOLE: Uh-ugh. 4 MR. BEARD: But at the same time, this utility 5 said they have checked with other plants that are serviced 6 by Yankee Atomic System, and they found that what happens 7 is all those other plants now have them on a ' check list,
~
8 but it is because in the past they had rome little occurrence 1 9 where they learned this lesson',"and as a result, have put 10 on there. 11 They are safety related valves. They should be 12 on procedures, and it is inconceivable to me that they are 13 not. 14 MR. EBERSOLE: Are there any orders or instructions 15 or warnings, or a ything going out to standardize this 16 approach? 17 MR. BEARD: This thing does have generic t 18 implications, as I mentioned. Especially if th'e utility is -; 19 correct in their assessment. . 20 We don't know exactly how we are going to handle , 21 the generic implications, but we are working on it. i M MR. EBERSOLE: All right. 23 MR. MICIIELSON : It is a broader problem, of course, i
/s 24 of any manual valves in any location being left in the close I N., '
[ 25 when they should be open, or vice versa. t s
6-18-JoeWal j 133 l 3 1 You can't just walk by and necessarily tell they 2 are there. You have to turn them to find out if they are l 3 there. 4 MR. BEARD: That is right. Or wire them, or 5 comething. 6 MR. MICHELSON: You have to turn them real well 7 in some cases, because they might have froze where they 8 were. 9 MR. EBERSOLE: I think you should turn them with 10 a companion to say you turned them. 11 MR. MICHELSON: You have to have some kind of 9 12 a formalized administrative procedure to first of all verify
- r f~h y g) 13 they are correct, and then some method of making sure they 14 aren't moved.
15 MR. REED: Let me give you a little operating
!, 16 history and the differences between departments.
17 These kind of valves, which are instrument valves, l
/ 18 would normally not be attended or opened and closed by 19 operators who are quite familiar with the check-off list
- 20- routine.
21 They would be done by instrument and control i ,
- 22 - personnel who are doing things to calibrate, and this that r 23 d$dsthe other, and so forth, so there is a bit of a difference.
24 And sometimes -- I notice the words, ' partially (~')
,g'.- '
25 closed.' Sometimes, I&C people even do intentionally l I
!6-19-JoeWal- 134 , 1 partially close valves on some instrumentation to reduce
'- flutter and wear on tubes and stuff like that, and gearing 2 3 and so on. l
.Now, I have to assume that this was a clear '
4-5 cut case of the instrument. department group not having a ' 1 6 . check-off list for their safety equipment and returning it 7 to service after calibration appropriately, is that 8 correct?. End 6. 9 M. Simons fois 10 11 12 1 .' O >> y q 14 r 15
; 16 17 18
, 19 h,
m i t
; - 21 '
i 23 ri 24 26 N-
.b ' ' ' -
135 l
' MR. BEARD: That is not correct. In fact, this !
1 2 particular utility, that is exactly the opposite of what 3 happened. s
'< 4 At this utility, the way that they work,it, you 5 know, around the transmitters you have got a'whole manifold 6 of valves equalizing fill, drain and all that good stuff.
7 Now all-those valves are normally operated by your 8 garden variety INC text. Nowtherhotvalvesatthis 9 utility are-cperated only by Licensed operators. Now that 10 may be unique with this utility, but that is the way they 11 told us that i,s the way they do business. 12 But the point I think here is they weren't on 13 anybody's list, and the cunningness, the subtlety in this t' 14 one is the fact that they are closed and the leakage i is through the valv'e can cause them to indicate they were j 161, confined says they dadgum well got to be on the list. 17 < MR. EBERSOLE: I am reminded of the Salem ignoring 18 of the fact that the breakers were safety grade. 19 MR. MICHELSON: When the utility says that the 20 root valves are controlled by the operators, did they
, 21 define for you what they meant by root valve versus 22 instrument valves?
s 23 MR. BEARD: Yes, sir, very clearly. 24 MR. MICHELSON: I know of one utility which puts 25 in both a root valve and an instrument valve, the root !\ p i ,
136 l 1 valve operated by the operators and the instrument valve by 2 the instrument mechanics. 3 MR. BEARD: The only thing I can really say is 4 that, you know, all these plants are different. They are 5 operated like, who was it, Asselstine used to say 70 6 different Companies, and they are all different. 7 MR. EBERSOLE: Indian tribes is a good way to ' a describe them. . 9 (Laug hter. ) 10 MR. BEARD: Nothing is constant.
' Let me go on to the second part of the problem, if 12 I may.
13 The second problem that they go into is an l' instrument design problem wherein they wanted to put in a is subcooling monitor. 16 (Slide.) , 17 I am going to put you a cartoon up here that you is have got in your book and then try to hide the bottom of it i, from you because it comes later in the plot. MR. EBERSOLE: That is this new equipment? 20 MR. BEARD: Well, it is somewhat new. 21 22 Let me talk about your basic instrument loops 23 where you have got, you know, pressure transmitter, and 24 inside the dotted box up he-<s I just tried to hide the i 25 power supply and test switches and all that staf f that is 1
- 0
i 137 () i not related to this scenario. But basically the transmitter puts out a current 2 3 throughout the whole loop, 4 to 20 milliamps, depending on i 4 what the pressure coming into the transmitter is. And the 5 way you get signals out of it is you put a resister out 6 there any place in the loop and run wires to whatever your 7 circuitry is. It is pretty standard instrupentation. ! e In this particular plant's design they have three i 9 transmitters all in one instrument Channel, one for each to gene rato r . . They go through, and here is your commonality, it what they call an auctioneer low circuit. And what this l 12 does, it says I think that my safety purpose over here is l j i3 to trip, if any of these pressures go low. So I will look ! 14 at all three signals and at .this point I don't care which is generator. If any of the three go low, that signal goes 16 onto the bistable and trips and you are off to the races. ] 17 Now what the utility wanted to do was to tie these i is same pressure transmitters in with some temperature signals l , i, and use them for subcooling monitors on the secondary i i 20 side. That is where it ties into a TMI type fix.
- 21 So they took some resistors in here, and I. am not 22 exactly sure how this scenario unfolded. One story was
- 23 these were existing resistors that were used to provide the l 24 signal to the plant computer just for monitoring the 25 value. Another story says no, they just went in and i
O i a _.m.__-.,,...,--.,...n.-__ -
._.w... - . - -cm.~,- m, ,,n. ., , , , . . .. ,. . , - _ _ . . . . . y .
,,.,m._.% , , . _ , . _ . . - - _ , _ . , , . _ _ _ . - _ _ . - . -
1 138
) i disconnected two jumpers, so to speak, as it was wired 2 around the plant and brought down an added resistor and 3 then took it off.
4 But at any rate, they put in a resistor, took the 5 signals and put them down here to inputs to their 6 subcooling meters and with switches and this, that and the
; 7 other. And thinking that it doesn't make any difference, a they said well, we have got a switch between these three, 9 and it Would be a lot easier to buy Cheaper hardware if we to ground, or common it, if I can use that term, and then we 11 switch to the other side.
12 So apparently they stuck a common in there and i3 ignored what might be over to the right. And low and 14 behold Murphy's law caught them and there was a common over , is here, but they never looked in that box. 16 So they made a mistake, they got caught, this 17 common and this common were in the wrong place such that
! is now these resistors that they were using were in effect in 19 parallel. So instead of getting an auctioneer low 20 function, they were getting an averaging function.
21 And it turns out that when you test this circuit, 22 you take one steam generator and you simulate its pressure 23 going down and the bistable doesn't see it and you don't 24 get this separate. 25 Now there is another interesting little tidbit on 1 0 i
l 139 () i this story that I think is sort of interesting. this mod last spring, and I believe it was like May. They did The 2 3 month prior to that they did the 18-month surveillance on l l 4 all this instrument going all the way through. That 18-5 month surveillance would have picked up this problem. 6 After they put this modification in, their post-7 mod testing said well, we wanted new subCooling meters, and l a lets look at those, and yes, they work. They did no 9 functional test of the reactor protection system systematic io gear, all those modules to see if they had been impacted, it and they did no calibration test to see which would have I 12 been the test that would have picked this up. I i3 The reason I say it that way is the so-called , f i4 monthly functional test would not have detected this problem. The 18-month surveillance test would have, but is 16 they had just done that the month before and saw no need to 17 repeat it. 18 So it is an interface question where you are i9 making mods on protection system stuff and you don't do 2a enough homework on the interface. 21 Now did I cover everything down here? Yes. ; 22 MR. REED: This is in the same ball park of return 23 to service af ter modification, as the Duke Power changeout 24 of level instrumentation on their accumulators in a way. i ~ 25 What is the seriousness of this by comparison? What the i O
140 4
) i thing is here is they created a design change and didn't 2 see that there was a slight flaw in design change, and then j 3 they returned to service without a complete test. Now is 4 this a serious issue with respect to what it might have )
s done, the consequences? 4 6 MR. BEARD: I think it is serious, and I will tell 7 you why. If I go back to this little matrix that I made up a of the affected channels and throw in up here that now I 9 have got nine of the 12 instruments affected by root 10 valves, they may work, they may not work or they may be l 11 slow, you know, and I don't know how slow. Then I say that i ! 12 all the "A" channels have been negated by this design l , i3 problem we just discussed, and I have got 12 channels . 14 affected. is So in my humble opinion, the safety function based 16 on steam generator pressure going low was wiped out. So if 17 you want to talk about an abnormal occurrence criteria that i is says a major degradation of a system important to safety, I l 19 think you are getting pretty close. l 20 Now on the other side of the coin, there are these
- 21 diverse other systems, like steam generator level. It is 22 very likely that if you have a main steamline break, which 23 is the accident that is relevant here, you probably are 24 going to get swell on the generator and go on high steam 25 generator level. Thac is probably going to be the first
~w l ss) t
) s
141 i thing that gets you. 2 If that doesn't get you, then containment high 3 pressure will come in after that. The licensee has done 4 some analysis of this and they think that the containment 5 high pressure signal will come in something on the order of 6 five seconds. 7 So the diverse systems are there, but the primary a system is basically severely impaired. So, in my opinion, 9 it is a significant event. to MR. ROSSI: And it is another case where you have is a breakdown of controls over safety related equipment. I 12 mean they didn't have adequate controls for the root valves i3 and they apparently didn't have adequate engineering review i, and so forth to do the post-design change testing that is needed to be done to make sure everything was working, and 16 that is at least as big a significance in my mind as the i7 actual function that was affected. 18 MR. BEARD: I have got to tell you another side of 1, this, Glen, and that is that they decided to modify the 20 steam gererator side and wanted just one channel. If the 21 NaC had issued the requirements, my esperience suggests 22 that maybe we would have said for redundancy we want 23 multiple channels. So this mistake would have been 24 repeated on multiple channels, and the problem would have 25 been compounded. f O I
142 9 Q 1 MR. REED: Every since gross modifications and 2 backfitting came into play in about 1973 or so, and then it 3 exponentially increased, I have been very concerned about l 4 backfitting because from ground zero development doesn't 5 take place. I don't know. Maybe plants should never be l 6 back fitted . ) 7 MR. BEARD: Well, this utility is very, very upset a by this occurrence. I mean it has really gotten their ; 9 attention, and rightfully so. And one of.the reasons it to has gotten their attention is in the last few years since 11 TMI we have hung a lot of mods on these plants. We have 12 got people scouring the countryside doing control room i3 design reviews saying more indicators on it. l 14 so there has just been a huge number of these . L is things, and the utilities indicated some 50 different mods 16 in this area. So they are concerned that they don't know 17 the status of that protection system right now well enough is to start the plant up, and they are undertaking two major I 19 corrective actions. 20 One is an independent design review of all the 21 mods since day one on the plant and, secondly, is a 22 comprehensive set of tests to show they don't have these 23 diseases. And the utility is doing that on their own 24 because they are very concerned. 25 MR. EBER80LE: That is the very same thing that he O P b k
143 is talking about. The mods have carried some flaws with ' [Vl i 2 them. 3 MR. BEARD: Well, you know, philosophically as a 4 regulator I have~to say that if we learn lessons from 5 accidents like TMI and we decide that safety improvements l 6 are necessary, it shouldn't be necessary to tell them don't , 7 screw up the mod. That is really what we are talking 8 about. 9 MR. EBERSOLE: Well, every alteration carries a io price as well as a benefit.
- MR. BEARD: That is all I have.
12 MR. EBERSOLE: Okay. Thank you. i3 MR. ROSSI: The next event is loss of offsite 34 power event that occurred at Dresden 2 on August 16th, and is that will be given by Bob Gilbert of the Office of Nuclear 16 Reactor Regulation. 17 MR. GILBERT: I am going to put up a couple of, as
- is J.T. calls them, cartoons.
l ig I am Bob Gilbert, by the way, from the Office of j 20 Nuclear Reactor Regulation. I am the Project Manager for 21 Dresden. 22 (Slide.) 23 Early in the morning on August 16th, and actually 24 it was 12:21 Central Daylight Time -- well, first of all, 25 let me preface this a little bit. ('] (m, i
144 i This is the tie hook-in of the reserve aux 2 transformer to the 138 KV lines leading in and out of 3 Dresden, and this is the one that was closed. 4 This one is indicated as closed in the FSAR and it 5 was not, and if it had been, this wouldn't have occurred at 6 all. 7 The thing that happened was this transformer, a which feeds Unit 1 as some minor loading still for Unit 1, 9 probably of old age gave up the ghost and flashed and all to of these, this ring of breakers, here opened and obviously r 11 lost power to the reserve aux transformer, went dead. So 12 we lost power to one of the two electrical divisions. i3 Now here is where the problem really occurred. l 14 Normally in a plant this circuit would have fast f is transferred over and brought in the station aux 16 transformer. There was a design deficiency in that the 17 lock-out relay was not wired in properly, and it would not i is allow that fast transfer to occur. So it did not fast t 19 transfer over here. It lost feedwater and low reactor ; 20 water level developed and we had a plant trip and then the , I 2 loss of power to that aux transformer and finally a 22 complete loss of offsite power to Unit 2. 23 .W ell, everything worked properly. Again, I hate l 24 to worry about shared things, but Dresden has a swing 25 diesel. This swing diesel and the Unit 2 diesel fired up C : l
1 145 i and activated and picked up the load. 2 The operators closed the MSIV's to conserve 3 inventory, and they activated the isolation condenser, 4 which is a nice little device because it has no moving 5 parts. The only valves in it are DC operated. This one is 6 maintained normally in the open position so that clear up 7 through the tube side and all in here is generally at a reactor pressure. So the one they opened here is the 9 return line, and they get a circulating flow. 10 There are three sources of water for the isolation a condenser. You have your clean domin water fill here, you 12 have, well, secondly, and the one that was used is the i3 water from the condensate storage tanks, and the reason i4 this is used is that the pumps for it are loaded to the is emergency buses, a This is essentially an infinite supply of water i7 for the fire service system, which has got another diesel is generator that will fire up and serve water. It was not i, used in this situation. (Slide.) 20 21 The makeup was supplied by a single CRD pump, and 22 reactor , level and pressure was stabilized three hours into 23 the event. The offsite power was restored af ter four hours 24 and the incident response center, because they did declare 25 a site alert, was brought into action. I got a call at
146 1 three in the morning an went down and several people were 2 already there. 3 One thing that helped in getting the offsite power 4 back on as quickly as they did was the fact that that night 5 a Region III securing inspector was out wandering the site 6 with one of Dresden's security people and they noticed 7 where the flash occurred and were able to direct them a quickly to the area where No. 12 here had gone. 9 So finally at about 4:30 in the morning their time to offsite power was restored and they started to bring things 11 back on line and they downgraded it to an unusual event. 12 They did not go all the way to take everything off because i3 they felt until their domin's and everything back on line, 14 which was at 9:30 in the morning, they would not do that, is and then they finally did at 9:30 in the morning. 16 (Slide.) . 17 I would like to discuss the staff concerns. is obviously we didn't let them go back up again. They stayed 19 somewhere slightly above, oh, between cold and hot shutdown 20 until they had that design deficiency replaced or repaired, 21 let me put it that way. It was repaired quickly and they 22 are now back on. 23 Well, another problem you have is that you don't 24 have a quick access over here because with this down and in 25 order to get power into this transformer, you have to O
147 i disconnect these things. It turned out that that, 2 according to the licensee, takes at least 24 hours and you 3 have to build scaffolding and it is a rather complicated 4 procedure. 5 So we were somewhat worried about conformance to 6 GDC-17. Unfortunately, you probably can't see it, but down 7 at the bottom there is an indication of a cross-tie to Unit a 3. So we had a tenuous second source of offsite power. It
, was a calculation made by a member of the Power Systems to Branch to indicate that if that had been used, then there it would be a little more loading available than was available 12 from the use of the diesel generators, but things were i3 working well with the diesel generators, and Unit 3 was at i4 90 percent power and they decided not to hook that in to i$ avoid perturbations.
36 One reason that this thing was hooked up here ir instead of here where it is indicated, as I say, in the is FSAR was that they had asked the load dispatcher which he i, preferred from a systems point of view, and it was his 20 choice that this was the particular spot. These lines go 2i to such places as Joliette and Lazan, Kankakee and there is 22 one that goes to industrial customers, and two lines go to 23 Joliette. 24 But this line here and this one over here go to 25 transformers which connect to the 345 line, the other set O L
148 l 1 of lines which are used for Unit 3. And so if, say, they 2 had been hooked up here and all these breakers hadn't gone, 3 they could have used that other. 4 The licensee has been asked to review their 5 conformance to GDC-17. They also asked, and since this I 6 situation also exists for Dresden 3, to be allowed to 7 continue to operate Dresden 3, and the staff is allowed to l e , that because Dresden 3 is nearing the end of its cycle, j I
- 9. That are in coastdown now and they don't have situations to like this that can cause that sort of a problem to exist. (
11 They also agreed that if they would go down for 24 I 12 hours that they will repair this design deficiency. So j i3 both the units will be fixed. O u They also checked all of their other Commonwealth f t is stations, and Dresden is the only one, unfortunately, that [ 16 has this particular design deficiency. All the others l 17 would have fast transferred under this situation. is MR. EBER80LE: You didn't mention makeup. There l 19 is not much makeup needed, is there, with that design? ( i 20 MR. GILBERT: ' Pardon? j 21 MR. BBERSOLE: Makeup. 1 22 MR. GILBERT: Yes, from the condensate storage tanks. ; 23 24 MR. EBERSOLE: No, no, no,-I don't mean that : 25 makeup, the primary makeup. Is there any primary makeup I O - i l z.-_..-...._-------,----,.-..--.---...-.---..-----
149 i needed? 2 MR. GILBERT: Well, they closed the MSIV's and it 3 was made up by just the CRD pump, and as the pressure 4 d ropped --- 5 MR. EBERSOLE: There was virtually no makeup,
! 6 right?
7 MR. GILBERT: There was hardly any --- MR. EBERSOLE: There was just seal leaking or i 8 9 something. in MR. GILBERT: They were able to bring it up. In is fact, they overshot. 12 MR. EBERSOLE: You can't beat that design with
)
i3 that return from that isolation condenser. That was before
\ i4 the days of the MBA's, Glen.
1 i$ MR. GILBERT: Well, this was replaced by RCIC i6 systems. This is a BWR-3. 17 MR. EBERSOLE: Right. te Any questions? i, (No response.) 20 MR. EBERSOLE: No questions. 21 MR. ROSSI: Next is the problem with Browns Ferry 22 with the seismic design of cable trays, and Eric Weiss will 23 cover that. 24 MR. REED: I have trouble following electrical 25 stuff. So I would just like to s'k the question, is this a I U l l
150 m) ( I total loss of offsite AC or a partial loss? Well, it is a matter of semantics to 2 MR. GILBERT: 3 some extent insofar as they could have used that cross-tie 4 to have a tenuous connection through Unit 3 to offsite i l 5 powe r . 6 MR. ROSSI: Well, they were totally dependent on 7 the diesel for some period of time. So I think it is a a total loss of --- 9 MR. GILBERT: Yes, we called it a total loss of to offsite power because they never did use that cross-tie. 11 MR. REED: Well, on the Point Beach one you 12 labeled the same way, loss of offsite power. Was that a i3 total loss or a partial loss? () 14 MR. GI LBERT: I can't answer that is MR. ROSSI: Well, that one they lost the offsite 16 power to the safety related buses, but the plant wasn't 17 tripped. So had the plan tripped, it would have been a is total loss of offsite power for some period of time until i, they cross-connected to the other unit. I would have I 20 guess been about the same kind of an ever.'t as Dresden had 21 the plant tripped because of it, but the difference is that 22 at Dresden apparently the bus that they lost caused the 23 trip and then they were totally without offsite power. 24 So Point Beach was without the offsite power, but 25 they still had the generator running and, you know, the 7-~
151 l i plant was still running. Had they tripped, then they would 2 have been totally dependent on diesels except I understand 3 they could connect hoses to the other unit. 4 MR. REED: I hope that some day we will 5 differentiate and get our nomenclature in line with 6 respect, shall we call it, total loss of outside AC and 7 partial loss of outside AC and total loss of all AC. I a hope we get these things straightened out because there are o reports being generated and numbers being put together and to incidents being tabulated which, in my opinion, don't it reflect the real situation. They sometimes tend to grab a 12 general term like loss of offsite power, and that is not i3 really what happened. It is something less than in many i, cases . is so I think we ought to be very careful with our 16 categorization because Oak Ridge will come up with a report 17 some day and everything will look wrong. is MR. GILBERT: We11, I agree with your point yery 19 well. One thing that commonwealth is very proud of 20 relative to Dresden is that they have got 11 lines leading 21 into the station and, therefore, wiping out all 11 lines 22 even with a tornado is highly unlikely. So they are very 23 proud of their loss of offsite power and maybe we gave them 24 a little dig here because of this particular incident. 25 ' MR. REED: Well, you understand what I am talking O
. _ _ ._ ___ _. _ _ _ - _ _ _ _ . _ _ ~ . _ _ _ . _ _ _ _ _ _ ~ . . _ . - _ . -
l j 152 1 1 about. I think we ought to really differentiate'these 2 things and make sure that we are not putting out l 3 overzealous reports or something when they.are not really l 1 } 4 the case. 4 I 5 MR. ROSSI: Well, sometimes it'is hard to tell 4 f 6 exactly, you know, what the situation is. Dresden really 7 was totally dependent on diesels for some period of time. ! 8 Now it is our understanding that because the diesels were i l 9 working properly and they weren't having any problems, they j l 10 chose not to try to cross-connect from the other unit i 11 because then they might have problems with both units, and , 12 that is very appropriate to do.
- 13 But how you count this is not clear to me because, i f 14 you know, they seem to feel they could have connected to 15 the other unit. We have had cases at Turkey Point I guess i i,
1 j 16 where they have tried that where what they have ended up ] j 17 doing was bringing down the other unit. So it is a little ( t
- is hard for us to know how to count that one.
i 19 MR. WYLIE: Just a question. They don't have an : l 20 emergency plan that says how they are going to do it? l i 21 MR. ROSSI: I think they have that, yes. I 22 MR. WYLIE: Well, what was their plan? 23 MR. GILBERT: I am sorry, in terms of doing what? ! , 24 MR. WYLIE: Well, I mean meeting GDC-17. i l 25 Seventeen says that you have to have a second source of j .. l l 4 ~ . - . - - _ - - - - . - - - . - . - . _ - . _ - _ - . _ . - - - -
-. .- -__ .-_ =
153 () i power available within the time limit consistent with the safety analysis of your plant. 2 3 MR. GILBERT: Well, of cou rse, if they had had 4 that fast transfer, they would have had the sources. 5 MR. WYLIE: Well, not really. 6 MR. REED: That would have only connected them 7 over to their own generators. That would have made it like a the Point Beach case I think. 9 MR. GILBERT: We are still looking into their to review of their conformance I would say to GDC-17. ii MR. WYLIE: I understand that, but in essence they 12 haven't done that analysis. i3 MR. GILBERT: No, we haven't seen that yet. 14 MR. WYLIE: Okay. I mean that is their problem. 15 MR. BEARD: Charlie, one of the shortcomings is I 16 remember being over in the instant response center when i 17 Indian Point went down and we thought we were going to have is another one of these New York blackouts, and they were i, keeping the plant up on diesels. 20 I don't know for certain, but it is my general 21 feeling that because the NRC requirements have these 22 dictates on how to get onto the diesels and how to transfer 23 buses and this, that and the other, that is as far as the i 24 utilities have carried it. 25 By and large I don't know of utilities that have
154 i procedures on how to recover from that. We reached the 2 conclusion during the Indian Point episode that if the grid 3 is back and available, the best posture for you to get into 4 is to put one division back on the grid as soon as you 5 think it is stable and keep the other one on the diesel and 6 let it run in that configuration for 4 to 12 hours or 24 7 hours, some period of time, get the plant really good and a stable and then switch the second division back on the 9 offsite power. 10 But by and large most plants do not have those 11 procedures because I believe they think the events are 12 never going to happen and the only reason they got the i3 first end of it is because the NRC required it. 14 MR. WYLIE: Well, that may be, but you know it as is well as I do that a lot of these electrical arrangements, 16 they meet the GDC-17 configuration, but invariably the 17 thing that happened here could happen in a lot of those la stations and you have got to demonstrate that you can get 19 that second source back in there to meet the 17, or 20 otherwise you are down flat on your diesels and that is 21 where you are. 22 MR. BEARD: The real crux of this problem was, and 23 in particular the utilities with the second source with 24 all those various lines that came through the switchyard, 25 and they didn't realize because all those lines came O
)
155 ( ) i through one transformer that that would only count as one 2 circuit, and they didn't have procedures to cover all the 3 various configurations and no plans to use them. 4 MR. WYLIE: I understand. i 5 (Slide.) 6 MR. WEISS: The next event is the problem that i 7 occurred at Browns Ferry, all three sites, and it began a with a call to the Operations Center on July lat. The call 9 said that some bolts were found missing in hold-down clamps i io for cable trays. , ti On August 15th we received a follow-up 12 notification in the NRC Operations Center, the substance of l 33 which was that there were bolts missing and that there was i s- i4 also some design problems. 15 Region II dispatched an inspector to the site who l g was a specialist in this area, and he found a number of 17 problems . He found that not all seismic loads were j is considered in the design of the supports, and that indeed
! i, many of these clips not only had bolts missing, but well 20 the cable trays themselves had their design capacity 21 exceeded in some cases by the combination of the weight of 22 the cables and in some cases the trays were overfilled, l
23 Pl us the cable trays were covered with a flame retardant i 24 material, Flamemastic, plus the conduit, plus the junction 25 boxes exceeded the design weight of the trays. 1
156 i The cable ampacity was brought into question, 2 particularly on the power cables, by the application of 3 this Flamemastic. There was no slack allowed for relative 4 motion of the cables in the trays. So that where a cable 5 went around a corner in a tray, if the cable and tray moved 6 separately, the tray might cut the cable insulation. 7 The anchors specified in the drawings that he a found were different than those used in the installation in 9 the non-conservative direction. He found errors in the to seismic calculations where he was able to find the 11 calculations. In one instance he found the seismic 12 response spectra of the reactor building used as a spectra 13 in the diesel building. 14 And beyond that he even found some trays where is there was no seismic criteria applied. Dead load weight a only was used in the ca'iculations. 17 TVA's analysis, preliminary analysis was that the is plant was not seismically qualified for interim or a long-19 term operation and they brought in a contractor, eight UE&C 20 personnel to help with an interim evaluation. 21 This is still in its preliminary stages and we are 22 telling you about this even before all of the dust has 23 settled, but Region II decided to dispatch an inspector to 24 Sequoyah, the same inspector, by ths way, and see how 25 Pervasive the problem was, and he general he found that 2
l l 157 () i although Sequoyah had some problems, they were not of the same magnitude nor as pervasive. 2 3 MR. EBERSOLE: Is this part of the reason they 4 shut down the plant recently? 5 MR. WEISS: No. This was not why they shut down l 6 the plant, but this may very well delay the restart of the 7 units. l 8 MR. EBERSOLE: If this Flamemastic a material if 9 they have sprayed water on it from a fire it would 1 io practically quadruple its weight? 11 MR. WEISS: I don't know. 12 MR. EBERSOLE: Wouldn't that be important to find 33 out? j u MR. WEISS: I am sure it would be. I asked my i Region II contacts for a rough feel of just how far off { 15 , a they were. Were they off by 10 percent or 20 percent, and 17 he said well, they were off in some cases by a factor of 18 three in what should have been the design capacity of the i i, trays versus what they were actually loaded. 20 MR. EBERSOLE: That material looks like it might 21 be like a sponge, and if you put water on it, you can 22 imagine the weight increase if that is the case. The end 23 result of spraying it with water might be to have all the 24 trays fall down. 25 MR. WEISS: I am surprised from what I heard over O
~ - -- -
_ _ . _ . ~ . _ __._ _.._______ ___ i. L
- 158 1
() i the phone that some of them still are there, but this is in i 2 the early stages of evaluation. , 3 MR. EBERSOLE: The problem is the superficiality l 4 of this . They find this to be the case, but do they I 5 backtrack and find the points where they aberration took i 6 place? t 7 MR. WEISS: Well, some of this came about as a i a result of the Browns Ferry fire when cables were added to a 9 the trays. The same inspector found a paper trail, a stack j to of correspondence that indicated that at least prior to
- 11 1981 TVA knew that it had overfilled trays. It didn't i
j 12 specifically address the seismic issue, but they knew they l 13 had overfilled trays and there were requests to Knoxville l 14 for engineering evaluations and possible design changes. 4 i l 15 But it was a long series of correspondence of requests and ; j 16 requests for extension and delays, and nothing has been i 17 done. l 18 MR. EBERSOLE: Yes. It is an enormous l 19 bureaucracy, and each leg of it is semiautonomous and they i ! 20 don't link up well. l 21 MR. WEISS: I understand prior to this or earlier ; i . 22 to the discovery of this problem TVA has made some changes, ! 23 including moving some of the design organization down to l 24 the site. 1 i 25 MR. EBERSOLE: Yes. Any questions? i i k i I t
..- --.,- ,-., - ._ ,- -.,-.- ..-,-.- --- - ,.-.---.- - -.- - ----..-- =
i 159 () i (No response.) MR. EBERSOLE: I think we are seeing what amounts 2 3 to the bureaucratic approach. i 4 MR. ROSSI: Next we have an electric power 5 transient that occurred at Callaway. This will be given by 6 Tom Alexion from NRR who is the Project Manager. i 7 (Slide.) 8 MR. ALEXION: Well, you can't see it all, but I 9 will move it up later. io On August 20th,1985 with the plant operating a it full power, the operator was bumping down the generator 12 output voltage and this was a request from the load i3 dispatcher. i () i, Then there was a sudden loss of voltage regulator, i 15 loss of the main generator field and the switchyard 16 voltage fell from 350 KV to 280 KV instantaneously. 4 l i7 The generator tripped on loss of field and that is was followed by a reactor trip, there was loss of power
- i, from the auxiliary transformer to the non-safety related 4
20 buses and the reactor coolant pumps tripped. ! 21 Since the power loss was fast, a sync-check relay i 22 dropped out and that prevented the automatic fast transfer 23 from the aux transformer to the startup transformer. This 24 is a proper response. 25 However, there was an automatic dead bus transfer 1 t I
160 i V
\ 1 after about two to three seconds and that re-energized the 2 non-safety related buses.
3 There was, however, no loss of power to the ESF 4 buses. S (Slide.) i 6 As you can see from this figure over here, one bus 7 was already on the startup transformer, and the other bus a comes directly from the switchyard. 9 MR. EBERSOLE: Is that a standard configuration to that they use, one on one and one on the other instead of I 11 both on the switchyard? 12 MR. ALEXION: You mean one here and one here? t
,, i3 I am not sure I understand your question.
. 14 MR. EBERSOLE: I am looking at the ESF buses, and is one of them is on the switchyard and the other is on non-i 16 safety buses. 17 MR. BEARD: Jesse, that is not typical. is MR. EBERSOLE: Okay. Well, the ESF buses here j 19 look to me like they are looking into -- no, wait a 20 minute. One ESF bus derives its power from the unit l 21 output, doesn't it, when they are on line; am I right? 22 MR. ALEXION: No --- 23 MR. EBERSOLE: Do they both? 24 MR. ALEXION: Both EFS buses get their power from 25 offsite. 0
- . ' . _- 4 a.- .o -
161 1 , () ~ 1 r. MR. EBERSOLE: conservative way. From offsite. That is the modern
.So that you retain power if you trip the 2
3 unit. 4 MR. ALEXION: Yes. In fact, the point of this l , 5 figure is to show that the EFS buses were never lost.
, ' , 6 MR. EBERSOLE: Yes, right, sure, and that is what
/
7 you want to show. - Yes, that is wanted to show, right. 0 / MR. ALEXION: 9 MR. EBERSOLE: And you would have lost them if you to tripped the' unit, if you had it on one of the unit boards. p You would have loat them if you had them on the unit j 12 boards. , The electric power system would have been lost. i3 MR. ALEXION: I assume so. ja This .shows the fast transfer that was prevented 15 from taking place, and this is also the dead bus transfer 16 that took place two,to three seconds afterwards. And this i7 is the generator that was lost. 18 (Slide.) j i9 MR. EBERSOLE: Well, aren't you telling us this is 20 the way it should have been? 21 MR. ALEXION: Yes. Everything happened, this 22 'whole event, everything happened that should have happened, a i 23 except the regulbtor failed. 24 MR. EBERSOLE: Yes. I am just hearing that T
] 25 everything went right off on schedule and you had a good O
i '
- -- - _. . - . . = . , -- .-,- . -
k i l 162 1 desig n. 2 MR. ALEXION: That is right. I guess we will come 3 to something else a little later. , 4 In going on with everything happened as normal, 5 there was a feedwater isolation in aux feedwater auto-start r 6 on the. reactor trip and low TAVG, and that is normal. 7 MR. ROSSI: Let me point out that that is the one 3 8 we were talking about earlier, that here is a modern 9 Westinghouse plant in reactor trip. They did both actuate io auxiliary feedwater normally and they isolated main feed, 11 which is the way most of those plants are, and here is a 12 good example. i3 MR. EBERSOLE: Do we call this now a legitimate 3-- 14 challenge when we summarize challenges? i 15 MR. ROSSI: Well, I don't know that you do 16 because, you know, you aren't actuating the auxiliary u feedwater system because you have had a failure in the main
, i
- 18 feedwater system. _
t ( i9 MR. EBERSOLE: It is just a matter of the better ', 20 'way.
~
21 MR. ROSSI: Yes, you can get it-back-in some l 22 reasonable short time if you needed the main feedwater, but 23 you do actuate the auxiliary feedwater, and.primarily you ! 24 actuate it because of the shrink in the steam generator. 7 25 level that gets you down to the low level setpoint-just . I i l
* \
+.
.) i 'i i
- - 't i h i 163 Lh it
) t, I
9 " because you have tripped. i 2 I wouldn't really consider this as a challenge 3 myself becaus'e it is not due to a failure in the main 4 .feedwater system. 4-MR. BEARD: I think the crux to understanding this 5 6 one, Je'sse, at least in my mind is that the auxiliary 7 feedwater system as an engineered safety feature is e intended.to take care of accidents and what-not, but it 9 also serves as the truly auxiliary feedwater system in that io it. comes ,on when you don't want the full main feed on. 11 So you end up with a safety related system that 12 serves a safety related purpose, on one hand, and an 13 auxiliary function, on another hand, and you are really i4 challenging it or calling it to work in that auxiliary 15 capacity and not because you have had an accident. g MR. EBERSOLE: But you do throw the system into a 17 nunber of transfers. Westinghouse also has a feature where
, is they-don't like transfers, which is where they refuse to g take the AC power supply off the turbine generator even e 20 after it has tripped because they don't want to-face an
- 21 auxiliary transformer to the station output and miss, and I C think the main reason is they missed the main coolant pumps 22
' ~
; 23 and they may have had an overpower transient.
< 24 You know that 30-second time delay ---
25 MR. BEARD: That is intended just to give them a O c y
- a. r I -
164
) i little extra so that if a transfer is missed they have 2 bought a little extra.
3 MR. EBERSOLE: That guy that wants that is not the 4 one that designs the feedwater system, because he throws it l 5 into these transients, and I really don't know which is the 6 more reliable or better way, but I can see points on both 7 sides. s MR. ROSSI: Well, again it is my understanding 9 that they isolate the main feedwater following the reactor to trip when you get down to a temperature that is about seven 11 or so degrees above the normal no-load temperature, and the 4 12 reason they do that is if you keep pumping feedwater in to l i3 try to rapidly recover level after a trip, you are going to u overcool the whole system, and if you overcool it enough is you empty the pressurizer. u MR. EBERSOLE: I've got you, sure. 17 MR. BEARD: I think you have to also realize, is Jesse, that a good operator can keep you above that low TS 19 setpoint if all of his equipment is working right. So that i ! 20 doesn't get challenged every time you trip the plant. f 21 MR. EBERSOLE: You ought not to list-this in the 22 challeng es . - 23 MR. BEARD: Well, you know, I can let the PRA guys l 24 worry about that one. I don't know how to handle that one, l l 25 and I am not sure I want to. 0
165 () i MR. WARD: Well, the PRA people would have to show it as the sort of challenge that it is. 2 3 MR. EBERSOLE: It is a benign challenge. It is 4 kind of a test. 5 MR. WARD: It ain't all that benign if something 6 doesn't work. 7 (Laughter.) MR. EBERSOLE: If it don't come back, it is not 8 9 benign. 10 Okay, carry on. 11 MR. ALEXION: I will just pick up the tail end of 12 this. 33 The licensee sent a couple of voltage regulator 14 components to GE and replaced them. He sent the ones he is suspected were faulty to GE for analysis. The a potentiometer was high suspect and the circuit card was i7 thrown in for good luck. is The thing that is interesting about this event is i, that a similar event happened in January, except instead of 20 bumping down the generator voltage, the operator was 21 bumping up the generator voltage, and we had a loss of 22 voltage regulator and loss of generator field and trip and 23 everything else. Everything was identical. 24 What they did at that time was they cleaned and l 1 25 checked the voltage regulator circuitry. They still don't l lO l' l
166 n ( ,) i know what the route causes were, and they sent it to GE for 2 analysis of those components. 3 MR. EBERSOLE: But by and large this was a 4 transient that went off very well. 5 MR. ALEXION: Right. 6 MR. ROSSI: Well, let's see, I guess this one ; 7 really is a true challenge to the protection system, a because if the reactor hadn't tripped because of an ATWS, 9 they really had a loss of flow and they would have gotten a io loss of the reactor coolant pumps, would they not have? 11 MR. ALEXION: I am sorry? 12 MR. ROSSI: I mean they loss the reactor coolant 33 pumps. So they lost all flow. So it was a true challenge u in that they really needed a reactor trip for this one. 15 MR. ALEXION: Yes. 16 MR. ROSSI: I mean I just throw that out for u general philosophical discussions of this event. is MR. ALEXION: I think it is a first of the kind 19 that we have seen. I guess that is why it was brought up. 20 MR. ROSSI: And they were on natural circulation I 21 guess for some period of tine until they got the pumps 22 back. 23 MR. ALEXION: Five to ten minutes. 24 MR. ROSSI: About ten minutes? 25 MR. ALEXION: Yes. 0 (
167
~
\
' 1 MR. WARD: Would you explain, it is a first of the (7-<) i 2 kind that you have seen?
3 MR. ALEXION: Well, I heard a comment made that we i 4 haven't seen this type of event before. 5 MR. EBERSOLE: The loss of exitation? 6 MR. ALEXION: No, the voltage regulator, but maybe 7 that is not true. I just happened to hear that, and maybe a it is not true. 9 MR. ROSSI: I think trips caused by the losses of io the reactor coolant pumps are relatively rare, too, aren't ti they, J.T.? I don't think that we-see that in many of 12 them. i3 MR. BEARD: In my memory there has only been one 34 other really big transient on loss of the main generator < is regulator, and that was down at Brunswick where they swung 16 the grid all over the place. But that is the main thing, i7 it was sort of a challenge to the whole plant system due to is a component that is usually very reliable, and we don't see i9 many failures on it. 20 I think another thing that brought it up and got l 21 it on the agenda here today was that we had had a large 22 rash of loss of offsite power problems, and here is one 23 that went right because of its design capabilities. I
! 24 think that was the main thing we wanted to tell you about 25 it.
!O
168 MR. ROSSI: The next presentation is Bill Jones is f) 1 2 going to give you a summary of startup experience 3 comparison on five plants. 4 (Slide. ) 5 MR. JONES: This is a startup comparison for these 6 five plants, and all the information is from the issuance 7 of full power license for the five plants. 8 The information is based on the 50.72 9 notifications that we have received on non-significant 10 events. Events which were not considered significant were 11 removed from consideration. And the selection of plants is 12 based on selecting some new plants that have recently gone __, i3 critical and received their full power license, and it is _ 14 based on an agreement, a consensus between NRR and IEE. 15 It just turns out that all but one is a 16 Westinghouse plant and Waterford is a Combustion plant. 17 (Slide.) is The first thing you looked at was a just gross 19 reported event comparison. These are 50.72 notifications 20 in the first, second and fourth month af ter issuance of 21 license, and one might expect that there is a high level, 22 and then it sort of goes down, and nominally I would say 23 that is the case except for two things. 24 Catawba was shut down a month. Significant 25 shutdowns have been eliminated. So there is a period of l b^3 l
. _ _ _ _-,m _ . ,
169 () i shutdown here that is not considered. These six events may indicate restart af ter significant shutdown where people 2 3 may have forgotten things and things may have happened that 4 that were due to shutdown. There were seven Waterford 5 events. And so for what the average is worth on five 6 Pl ants looks like this. 7 (Slide.) a What is more significant probably is to look at 9 reactor trips and to look at events that involve reactor 10 trips and losses of feedwater. And for the number of ii events that we had in that four-month period, most of them 12 were trips. In fact, I think the majority of the events 13 were reactor trips. i These reactor trips were greater than five percent i4 is Power. And if you take a four-month period and you look at 16 an AEOD study for '83 and an NRR study for -- let's see, it 17 is an NRR study for '83 and it is an AEOD study for one is quarter of '84. The average for this same comparable four-19 month period would have been 2.1 trips for all the 20 reactors, for all PWR's that were operating during this l 21 Period of '83 '84. And that 2.4 is directly comparable to 22 this n'imber here. So this is a significantly higher number 23 than the average for those that were operating. l 24 MR. WARD: That is 2.1 per month? l 25 MR. JONES: No. This 2.1 is directly comparable O
l ', 170 p i to the number here. In other words, it.is 2.1 over a four-2 month period, and I made it four months because that is the 3 number that is here. 4 MR. EBERSOLE: You mean after they have shaken > 5 down and so forth they get --- 6 MR. JONES: In '83 '84 all the plants, all PWR's 7 that were running, it turns out that was the area. 8 MR. EBERSOLE: This is the bath tub curve? 9 MR. JONES: Right. And, again, I think that is I to mostly due to the fact of their initial criticality to full-l 11 power license. i i 12 I don't have any comparisons for loss of 13 feedwater, but there is a significant number of events, at 14 least for these three plants that involve loss of 15 feedwater. 16 As far as significant events, reactor trips and k 17 loss of feedwater for these five plants, none of them seem 18 to stand out to me, and that is maybe one conclusion. 19 (Slide.) 20 MR. EBERSOLE: You didn't figure out which was 21 Operator error? 22 MR. JONES: I did but, unfortunately, I didn't l 23 divide out the reactor trips. I have tried to look at, you I i
- 24 know, of all of the significant events, and not just the i
! 25 trips, I have tried to define for you, based on what we i O l l l,. ___--.._._.._,__-__.._____.,._.._.__.._____.___--._._____--J
171 ( ) i could figure out from the 50.72, and some, I guess I would 2 say some of these columns may shift when you get the LER, 3 the licensee event report. 4 But, at any rate, for what it looks like right 5 now, the causes of the events are about 50/50 here and 6 50/50 here. < 7 For Callaway these two numbers are a little out of a sync with the others, and maybe Catawba was a little the 9 other way and Byron a little the other way. But I guess io for the data that is here, nothing jumps out at me. in MR. EBERSOLE: None of these stick out as being 12 indicative of anything fundamentally different from the 33 others? i~ s a MR. JONES: Right. That is the way I look at it. is You do have this one, but that is a positive sense, and we 16 have this one and this one and this one and this one. 17 MR. EBERSOLE: I guess I am a little bit trying to is relate this to the experience at Palo Verde where things i, fell apart all over. 20 MR. JONES: They fell apart all at once. 21 MR. EBERSOLE: Yes. 22 MR. JONES: If you-look at a list.of new plants, Palo Verde is on that list and there are about five others, 23 24 but these happen to be the ones that were selected. Also, 25 Palo Verde has been shut down and there just is not that O l
b 172 7-1 much information on it. 4 2 MR. EBERSOLE: Sometime ago I was just wondering
; 3 what sort of criteria do you use to say one of these plants t
] 4 has got too many things wrong with it or something and we 1 a 5 have got to go and have another hard look at it? l I 6 MR. JONES: I think in this case for these new ^ f 7 ' plants 'the events are reviewed on a daily basis. And the )' s basis of selecting these plants was somebody felt we ought i 9 to look at that plant. t i 10 As I said, that was a consensus between NRR and ! 11 IEE. All I can say is it is being looked at and --- l 12 MR. EBERSOLE: What I am saying is where would you
- 23 have to go on one of these plants before you would open up f
l "- 14 a new, you know, another level of investigation? \ ' j 15 MR. ROSSI: One of the things we are trying to do 16 in our branch is to try to develop a more quantitative way 17 of looking at events on plants to identify what we are is going to call problem plants which are plants that ought to 19 be concentrated on in the inspection program. , i i j 20 So within IEE we are looking at better ways to
- 21 identify plants that are having enough problems that they 22 ought to receive additional attention through the i 23 inspection program.
i 24 MR. EBERSOLE: Palo Verde, how does it stand at 25 the moment? O
173 () i MR. ROSSI: I am not able to comment on Palo 2 Verde. 3 MR. JONES: If you just look at the events, and I i 4 guess it hasn't stood out in anybody's mind. I did take a 5 look at all the events that I could find on Palo Verde for 6 a reasonable amount of time, and I guess we h&Ve to go back y and take a look at it. 8 MR. EBERSOLE: I sort of figured as much. l 9 (Slide.) in MR. JONES: For those plants that have operated in out through seven months we did add some events, and these 12 are the number of events and these are the reactor trips. At least within a reasonable kind of behavior, I think this l i3 l u is the kind of behavior that you would expect. is Again, this is the month after receipt of a full-f 16 Power license, i 17 MR. EBERSOLE: Okay. Got you. is MR. JONES: I guess, as I say, the bottom line, to i 19 me at any rate, of the plants selected, there didn't seem i 20 to be anything that jumped out. 21 MR. EBERSOLE: Good. Any questions? 22 (No response.) 23 MR. ROSSI: The last item that we have is a i i 24 discussion of pipe crack indications at Peach Bottom, and 25 Gerry Gears from NRR is going to talk about this one. O f
174
) 1 (Slide.)
2 MR. GEARS: We finally wanted to present this 3 afternoon a brief overview of pipe crack indications as a 4 representative of certain problems that we have been seeing 5 going on in the field of detection, UT detection, and also 6 highlight a recent episode at Peach Bottom Unit 3 that sort 7 of culminated in and focused on our Concern. 8 As you may recall, back in'83 and actually '82, 9 there was a general concern of IGSCC found in boilers, and 10 I&E came out with Inspection and Enforcement Bulletins 82-i 11 03 and 83-02, and subsequent to that last April the staff 12 came out with a generic letter, 84-11, describing a i3 reinspection program to be undertaken by all boiling water 14 reactors. is Currently at Unit 3 they are 'u ndergoing and still u are inspecting pipe welds for IGSCC, and we got indications 17 starting in mid-August that'they have found new cracks, is which is not surprising I guess. But what was surprising 19 is the number end in some cases the size of the cracks on 20 welds which previously back in 1983 had no indications or 21 at least were called to be clean or free of any 22 indications, and I will talk about that a little later. 23 Generally we have looked at some data that was 24 Presented by the licensee last Friday, and in almost all 25 cases there wan some information provide on UT. It was \ < 0
l 175 i called generally ID geometry, root geometry and indications 2 like that. ! 3 As I said before, one of the welds here that we 4 are talking about so far apparently has a fairly 5 substantial crack in length that is called a 360 degree 6 intermittent circumferential, and it is sized upwards to 55 7 percent through-wall on a weld that back in 1983 again was a considered to be without any indications. 9 I guess this information that we received at Peach 10 Bottom sort of added the last nail in the coffin, so to it speak, because there has been some ongoing information 12 Presented to the staff and knowledge to the staff that UT 13 inspections have been performed and that things have been u in fact so-called missed, not identified. 15 I can give you one episode that I am familiar with 16 at Peach Bottom where back in 1983 a variety of welds were 17 selected for induction heat stress improvement, a technique is that has been claimed to prevent IGSCC. i, These welds were considered to be without in some 20 cases, at least two of the welds have been inspected 21 already that IHSI performed, and then a post-UT inspection
- 22 done, in which case they found cracks.
23 There was indication of course back then, a 24 concern that for some reason those cracks were probably 25 missed. l l
176 l 1 We now have more samples based on utilities going 2 through and replacing non-conforming pipe with conforming 3 pipe. We nave some cut-out welds that actually have gone i 4 through a much more detailed analysis in laboratories and 5 found again another example, for example, at Peach Bottom 6 where a particular weld had back in 1983 three independent 7 UT teams look at them. Each team showed on this weld a fairly corresponding crack locations. In other words, 9 there wasn't full agreement, but the majority of those to teams indicated and had some pretty good correlation that 11 there were cracks. ! 12 Upon further examination, laboratory examination i3 it was agreed that they did a good job in calling that for ; u the cracks they called, but then there was a bunch of ! 15 cracks that were totally missed that none of the teams r 16 Called. , 17 So based on that evidence and, as I say, finally j is on the evidence that we have gotten at Peach Bottom, we are
- 19 somewhat concerned with the whole reinspection program l 20 under 84-11, and to that end the NDE Center in Charlotte, i
2i North Carolina, together with EPRI and IsrE and our NRR 22 staff have been trying to put together what has been called ! 23 requalification program or retraining program to put the 24 inspectors through a more rigorous testing procedure. 25 Basically we are able to do that at this point O l l
i 177 J () i because there are a lot more samples out there and a lot more actual field cracking that is available, cracks that 2 l 3 are available that perhaps were not available a year and a 4 half ago. So at least the industry has taken the stance at 5 this time, it is ry understanding, that there is a need for
! 6 sending the inspectors through a more rigorous defined i
j 7 testing program > 8 In NRR we have asked the BWR Owners Group to come 9 in next week and further discuss what they plan to do on the plants that are coming up. There is something like f io n seven plants that are coming up in the near future for 12 their 84-11 inspections and we will get some more detail on 4 33 what they play to do to address this issue of j 34 requalification.of the inspectors. Generally, also, we are concerned about the l 15 i 16 direction that the pipe crack issue is heading,.at least a l i7 year or a year and a half ago, let's say in that time l ts frame. There was a generation indication that many BWR i
! j9 owners were going to also opt for replacement.
I 20 It is not clear today is that is the case . In l 21 fact, there are some utilities on record who have indicated 22 that they would like to operate the life of the plant as l 23 with overlay weld, repairs, et cetera. So that is another 1 24 issue that will be discussed in this meeting. 25 MR. WARD: Are you pretty convinced that these new O
-,gw+m- es?-e *+--e-, we- -.-mw r*e, .
ase cM.9 ----+,,--y.--r,---yvy, gr
178 [) i cracks are things that weren't found, or is there some 2 indication that the chemistry is not under control as it is 3 thought to be? I don't which is more alarming. 4 MR. GEARS: Well, it is a difficult issue. As I 5 said, most of the cracks so far have been found in welds 6 that had been IHSI, and if you make the presumption based 7 on data presented for induction heat stress improvement, a that it is supposed to prevent or at least retard cracks, 9 then the evidence or at least the feeling is that they were 10 probably there all the time and they were probably missed. 11 We are still out. 12 I mean the Philadelphia Electric addressed that 13 issue on Friday and made a strong presentation, and strong O" s- 14 in the sense of saying that they have done so much IHSI and is they have found a correlation between IHSI in causing 16 cracks and that all the correlations so far would indicate 17 that in fact it does what it is supposed to do. So if you is take that tact, then you would have to believe that they 19 were probably there all the time. 20 MR. REED: I think this is one that should come to 21 the full committee --- 22 MR. EBERSOLE: Oh, I agree. i 23 MR. REED: --- because it has all the hydrogen 24 chemistry aspec,ts and all the others. 3 One thing, I am interested in a comment you made,
\
_ _ . _ - - ~ _ -___
179
/3 and it is a thread that runs through our industry in the V i 2 regulatory, and that is you immediately say well, more 3 training for the inspectors will whip up the inspectors to 4 find the cracks. But all the same, the cracks will be 5 there one way or the other.
6 We don't seem to focus very much on perhaps there 7 is a design issue here that should be flogged as much a training of inspectors. 9 Now we do the same thing with operations of io plants. We don't look at the frailties or vulnerabilities 11 of the plants and their equipment. We say flog the 12 operators to a higher plateau of performance. You know, it l 33 is a threat. I just wanted to point that out. ja (Slide.) MR. GEARS: I wanted to show one other piece of 15 16 information. 17 MR. WARD: Let me ask you a question on that, is What is the status of Peach Bottom, Unit 3? i9 MR. GEARS: Unit 3 is currently in a refueling 20 outage, and they have come in to us and asked in an 21 informal meeting setting that they would like to repair 22 based on criteria given in 84-11 cert welds. They would 23 like to leave in an unrepaired state based on 84-11 2'4 guidance other welds to last for approximately 18 months or 25 the next refueling outage. Then they have at least
180 l N) i informally committed to us that a certain number of these 2 welds unrepaired will be replaced in the next refueling 3 outage. 4 Now the welds that we are talking about that are 5 of significance in this issue, not in terms of cracks, but 6 of the unrepaired state are safe ends which when they say 7 replace safe ends, they are really talking about,a major e outage where they have to drain the reactor vessel. 9 Essentially a commitment for that would mean at least to marginally a partial replacement of their recirc and RHR 11 lines. 12 So Unit 2, they have totally replaced all their ___ 33 non-conforming welds. That outage was supposed to last eight months and it has ended up lasting 14 months. So I s 14 15 their position is they would like not to go into another 16 full outage. So they have asked us essentially to let them 17 operate with either repairs or as is for another 18 months, is and that is still under review. 19 MR. WARD: Is this weld that you have described 20 here repairable under 84-117 21 MR. GEARS: Yes, that one we would allow. We 22 would assume based on the criteria that that is a -- well, 23 it has been called 360 degrees, and we would assume it is 24 360 degrees, and we would assume it is through wall,100 25 Percent through wall, and then do a full-strength overlay
~N ia
_c
i 181 weld. We have allowed from cycle to cycle that type of f() i
- 2 repair.
3 The safe ends issue is a trickier issue I think 4 and the staff has to give a lot of consideration to that. 5 MR. EBERSOLE: May I ask you this. The staff, as i 6 you know,' is moving in the direction of leak before break
- 7 1091 c, and they have gone down that road pretty well on a PWR's. As it usual, they are trying to treat all reactors
- 9 as though they are the same. They have been pushing along l io or dragging along the BWR behind that, and I haven't ti thought that that is really up to that level yet. How does 12 this impact on that?
i3 MR. GEARS: I don't know. If Warren would like to
\~s u handle that, Warren Hazelton.
15 MR. EBERSOLE: -This tendency to try to force fit 16 all these reactors into the same box runs me up the wall. l l 17 MR. RAZELTON: Well, we are not fitting them all is into the same box, No. 1. 39 MR. EBERSOLE: Good. l 20 MR. RAEELTON: We take the attitude that for BWR 21 recirculation piping and other associated piping that as 22 long as they are having problems, special problems that we . 23 know about, IGSCC, we just don't want to take that extra l 24 step and talk about leak before break. 25 MR. EBERSOLE: -All right. They are different O
I 182 O,g i cats. 2 MR. HAZELTON: That is right. 3 MR. EBERSOLE: And they will probably always be 4 that way. 5 MR. MICHELSON: This GE Smart UT system, is that 6 being used by most of the utilities now? 7 MR. GEARS: Yes. I didn't really mean to give a a plug on the GE Smart system here. I think that when this 9 was made up a week or so ago we had not a full, good 10 understanding of it.
~~~~
11 It has been used partially I think in other 84-11 12 inspections, but this is the first time I think it has been i i3 full time at a plant. Seventy-five percent of the welds, I 14 understand, are being at least initially looked at by this is system. 16 MR. MICHELSON: Well, other people are looking for 17 cracks. Is it correct, are they using this new system? 18 MR. GEARS: No, we aren't mandating their using 19 it. 20 MR. MICHELSON: Well, how are you rationalizing 21 that now that you have a system which seems to be better, I 22 guess, in finding more cracks, why aren't you using it to 23 look at other utilities? 24 MR. GEARS: Warren, do you want to handle that? 25 MR. HAZELTON: Let me answer that. First, there l O 1
l I 183 , l f i are a lot of competing smart systems out there. 2 MR. MICHELSON: The difference here is that this 3 is smarter than the rest in finding more though, isn't it? 4 MR. HAZETjrON: Let me answe r . The important thing 5 here on the findings at Peach Bottom, I don't think we can 6 attribute it just to the use of the General Electric 7 Proprietary Smart system. 8 The fact is that other people even doing the 9 manual UT are far better trained than they were before, and in they are using more sophisticated manual methods than were ii used before. 12 The important thing that we are saying here is , 33 here is a case where the utility specifically is using the a nest up-to-date inspection methods and they have been is finding things that they rissed the last time when they a weren't using the up-to-dete methods. 17 Now not all utilities are .using these up-to-date is methods or plan to use these up-to-date methods in their 84-l i, 11 reinspections. That is the point we are trying to make, 20 and that is why we are saying we have to do a better job on 21 the 84-11 inspections. 22 MR. MICHELSON: Well, you kind of ran around the Problem. You almost answered it at the end there, but not ! 23 24 quite. What are you doing about those who elect not to use 25 as good of systems and so forth? r
104
) 1 MR. HAZELTON: What we are doing is saying that 2 these people have to be requalified and they are going to 3 have to pass a test that is much tougher than they had to 4 pass before. In other words, they are going to have to use 1
5 the up-to-date methods to pass the test and get qualified 6 to go out and inspect.
- 7 MR. MICHELSON
- Okay. What you are I guess a telling me is that they will have to use a smarter system 9 than they have been using, whatever it is ---
10 MR. HAZELTON: That is correct. 11 MR. MICHELSON: --- and be better at its use 12 before they can pass. Are you having an actual i . i3 requalification examination? 14 MR. HAZELTON: Oh, yes, and we have fought this 15 through with the ' industry. We said that people have been not doing a good enough job 'and therefore we need to have a
~
16 17 qualification program at least to make sure that these ! la operators know.how to find cracks. That is just part of 19 the problem. f 20 The next part of the problem is it is the 21 responsibility of the owner to see that when he gets out 22 there inspecting pipes that he does it the way he knows 3 23 how, and we have got problems in that area that it-is going l 24 to take a long time to discuss. 25 MR. MICHELSON: Well, clearly, the NRC has a basic 0 .
4 185 () i safety obligation to be sure that if there are cracks out there of serious safety significance that we find them. 2 3 MR. HAZELTON: Well, yes, but I look at it that 1 4 the owner has a responsibility to find them. 5 MR. MICHELSON: Well, I mean you are responsible i 6 to see to it that they are found, and presumably the owner i 7 is the one to do the looking. g MR. HAZELTON: Well, I think the owner has the { .l 9 primary responsibility, and we are just looking over and 30 trying to determine whether or not they are finding them. l 11 MR. MICHELSON: What do you think your 1 12 responsibility is in a case like this? Here there are 33 potentially serious cracks in these systems and so forth. u You as the staff or as the NRC, how do you feel or what do 15 you think your responsibility is in a case like this, you 16 know, to the public I assume? 17 MR. HAZELTON: That is a long question really, but l is clearly one of my responsibilities I think is where I see a i, situation where cracks may not be discovered, I whip 20 whoever I have to whip to try to make sure that they are i 21 doing as good a job as they can. Now that is one thing. 7 l ( 22 Obviously if somebody asks me what would I like l 23 them to do, I would like them to replace the pipe, but that ! 24 takes rulemaking. It doesn't take rulemaking to whip 25 People. O 1
' 186 1 r, 3 1 (Laughter.) , 1 ! 2 MR. EBERSOLE: Suppose that there were one 3 proprietary system that could see cracks, and this may be 4 one, and there was only one that could really do a good job 5 of it, you can't tell them to use it, can you? - 6 MR. HAZELTON: The only thing I can do is make a 1 7 test that is reasonable, but a test so difficult that this a guy is the only guy that can pass the test. 9 MR. ROSSI: Or anybady that is better. 10 MR. HAZELTON: Yes, that is right. Now there are 11 many systems that will do a good job and that will pass the 12 test, but it just so happened that this GE Smart system is l i3 one of the better ones and it happened to be used at Peach
} u Bottom. So it was mentioned.
l i 15 MR. MICHELSON: Most of what you told me here, did 16 it occur before or after they discovered that they were i 17 missing things at Peach Bottom? The requalification l 18 program and the requirements and so forth, was that before 19 Peach Bottom? 20 MR. RAZELTON: We covered this in a meeting that l I 21 we had last June at the NDE Center. We have been finding
- 22 problems at Duane Arnold, at Fitzpatrick, at Quad Cities 23 and Brunswick where it appeared from the best available 24 knowledge that we had and the region people had that some j
25 of the inspectors were not doing a good job. They were i, b " "%
' 187 () i missing cracks that subsequently somebody else found and f 2 things of that nature. 3 So in June we said we think we need a
- 4 requalification program to make sure that everybody out i
l 5 there inspecting pipes really knows what to do. 6 The early bulletin performance demonstration tests 7 were we felt perhaps too lenient and, furthermor e, because 1 ! a we now had many more field crack pipes available, we could 9 set up a better training program and a better inspection. 10 So we felt we had to do it, and I mentioned this, the it information to EPRI people, et cetera, and I understand the i 12 owners group was told about it that we were going to ask $ 3 33 for that. l 34 Then we did bring the subject up on the Hatch case 15 where Hatch 2 said they were going to come in and request 16 to go a third cycle with-just overlay and then do IHSI and 17 no forth. And we said, gee, if you are going to go that 1 is far, we think you really need to have your people 1 ig requalified to find cracks. If you want to continue to l 20 Operate and try ensure that you meet GDC-14, extremely low f 21 Probability of abnormal leakage, then if you have to depend . 22 on inspection, you are going to have to do a better job 23 than we see a lot of people doing. 24 So we brought that up and I guess that was maybe a ' l ! 25 month and a half ago or so, and we were still fighting with O
188 O Then Peach Bottom happened and they gave up. g_) i people. 2 MR. MICHELSON: When did the GE Smart system 1 3 become available to the industry? 4 MR. HAZELTON: It was out on some field trials and 5 so forth probably about six months ago. 6 MR. MICHELSON: And when you talked about 7 requalification last June you had in mind that you knew at a least one system that was finding --- 9 MR. HAZELTON: Oh,'we knew a lot of things that to were available. The technology had really improved. 11 MR. HAZELTON: Okay. So you knew there was enough 12 improvement that a requalification made sense? i3 MR. HAZELTON: Oh, yes. 14 MR. MICHELSON: Okay. 15 MR. HAZELTON: And the problem is that there are 16 literally hundreds of inspectors out there, most of whom 17 hadn't, or we felt hadn't been exposed to the new is technology and the new equipment and the new methods. So [ 19 we felt we had to do that. i 20 MR. MICHELSON: Maybe some utilities weren't 21 trying too hard to find cracks. Is that possible? 22 MR. HAZELTON: I wouldn't say that. 23 MR. REED: You know, I certainly agree with one 24 thing, that it is a GE Smart system. Here is a company 25 that creates the boiling water reactor full of oxygen which O
l 189 A i causes the cracking, which then they device a system to
, 2 find the cracking. It is a smart system.
L3 (Laughte r. ) 4' MR. HAZELTON: And then they sell the pipe 5 replacement. 6 'MR. EBERSOLE: I think it is a tribute to the 7 company that has the cracks to have the best system to find i a them. 9 MR. HAZELTON: I think if we could put equal to attention on water conditioning or hydrogen use or ii something to try to stop the cracks we would be better off. 12 MR. MICHELSON: We are hoping that that isn't just t 13 another piece of pie in the sky, too, and that it will 34 really work. I don't if there is any new information on 15 the hydrogen, you know, its promise or not, and it is i 16 Probably too early'~yet. ( l 17 (Slide.) i is -MR. GEARS: I just wanted to show this. We were i ,~ : talking about this whole issue just recently or just now,
;o /and we essentially asked the same question. The question 21 really care to the floor last Friday by Philadelphia 22 Electric an~d not by any nudging to us. They presented 23 these data here thati generally address the issue of why l 2.5 they were missed and what was' their rationale.
25 By the way,.most of'the General Electric people 4 '
> I, i
s,_ - -- -
~
190
) I who did the 1983 inspections were the same ones that did 2 the-1985. So in some ways they were on the line, or. they 3 could have made some rationale of why they missed it. But 4 this is a chart that they presented us. ij 5 I think to summarize it generally, tile feeling is 6 that in 1985 there are several areas that they feel they 7 are better trained in and the level two and level three a people have gone through the EPRI training and 9 Certification program. These same people have had two 10 years, or many have had two years of actual field 11 experience of trying to identify IGSCC.
12 And then the equipment, as we talked 1about before, , _ _ , i3 is much better it appears. Not only the Smart system, as s- 14 you can see on the bottom, but just the use of a variety of 15 new transducers that apparently have a-much better 16 capability of not only finding IGSCC but also sizing it. 17 Sizing was a big issue and it came to the floor on Peach is Bottom a year and a half ago. It was the forerunner of the 19 sizing issue when they called a particular crack 30 percent 20 through wall and it went into repair and it leaked. So l 21 there was a question about whether it was 30 or 100 percent 22 through wall. So I think they are also forefront here of 23 finding some interesting information. That is all that is
?
24 Presented here. It is just to give you an indication. l 25 Southwest was the other group that worked back in l-l i 8 j
i 191 8 [ , i 1983 and currently is doing a part of the 1985 inspections. 2 What Peach Bottom is doing is that they go in 3 basically with a Smart system and determine the locations 4 where -- the system determines locations where there are 5 cracks or indications of cracks, and then they follow it up 6 by manual calls, both by General Electric and Southwest. 7 And so far, from what we understand, the Smart system is not overcalling. They are right on the mark. Everything a
' that'has been called so far has been confirmed by manual , 9 'i '
t- io UT.',So it seems like it is a fairly good system.
- The advantage of that system is it takes your 12 People out of the field, the radiation field. You have 13 more time actually to do a more detailed assessment of what i4 is coming across, and just maybe in that there are better 15 working conditions and much more time to look at what is 16 actually coming across, and I think they probably do a 37 better job.
.is MR. MICHELSON: Ie there any way of predicting the 39 extent to which they are undercalling the cracks?
2c MR. GEARS: We tried to ask a question, in other 21 words, are they doing any sort of independent manual only observations and corresponding to that. No, they aren't 22 23 reallyldoing that.. All they could tell is that this Smart 24 System with the operators have gone through the NDE Center 25 in the last year and have all successfully passed. O - 8 . , - - ,--y
192 1
,~5 l But how do I know that this system
( ) 1 MR. MICHELSON: 2 isn't missing cracks? What other independent check is 3 being done to look for that possibility? 4 MR. GEARS: I guess the only way I could answer 5 that is that apparently there is no hundred percent i 6 guarantee with a UT. 7 MR. MICHELSON: This is just finding more cracks a than they used to find. 9 MR. GEARS: That is right. 10 MR. MICHELSON: But whether they are finding all 11 of them or not, you don't know. 4 12 MR. GEARS: That is still the question. 13 Apparently we are resolving and fine-tuning it enough that u at least we are identifying the fairly large ones or even is ones that are on the category of the two and three-inch in 16 length ones. Are we still missing the one incher or the 17 two inchers, that is debatable. 18 MR. WARD: There is some indication of its ability 19 to detect any crack from the piping that has been removed, 20 isn't there? 21 MR. GEARS: Right. 22 MR. WARD: It is not entirely just blind. 23 MR. MICHELSON: But are they actually doing a 24 verification examination then to see if there were cracks 25 at-the --- l l 77 l N~/ l I
_ _ -. . . . =_-_ -_ 193 i MR. WARD: Yes. They cut up the pipe and looked. 2 MR. MICHELSON: How much of a sample are they l 3 doing? 4 MR. WARD: I don't know. 5 MR. GEARS: Warren maybe can handle that. As I 6 said before, at least this particular system has gone
- 7 through the Center which has cut out examples and ,
a identified cracks, and it has passed successfully. Whether 9 it got 100 percent of all the cracks in there or passed to somehow, I can't answer. t 11 MR. HAZELTON: What was the question. I was out l 4 12 in the hall. j 33 MR. GEARS: I guess the question is what sort of 14 certitude do we have at this point that the GE Smart system 15 is still not undercalling or still is missing cracks? 16 MR. HAZELTON: Well, we never have complete i 17 assurance. What has happened is that -- and I will tell f 18 you one of the first instances where it was used. It was 19 used out at Brunswick where a Region II inspector went down i 20 there and was very unhappy with the way they were 21 inspecting the pipes. He didn't think they were doing a ! 22 good job, and this is one of the few that we have that l 23 takes his own equipment and goes and scrubs the pipe and i 24 looks for cracks. And he found cracks in welds that the 25 utility's inspection agency said were free of cracks. ! l lO l
~. . - _ . . . . . . . -.-.- - .- .- - -.- - - .. - . . . .
I t 194
) i So they took appropriate action and got them I' .2 moving.
1-3 MR. WARD: You have got a regional inspector that j 4 has his own UT equipment? ,
)
5 MR. RAEEUTON: Yes. There are several, but in + 6 Region II there is one regional inspector that does that,
! 7 and he is very good at inspecting.for IGSCC. You know, he a happens to be very good at it.
9 And then they brought in the GE Smart system. So l io he had an opportunity, and he was watching it very closely ; 11 obviously and working right with it, and he saw how good it i l i 12 wo rk ed . He gave me a call and we talked for about an hour i
- i3 on this subject and he extolling the virtues of it. !
l is He said that it called some cracks that he had b So he went back to look for them and he said he t is missed.
! 16 had an awful time trying to find those cracks, but he ,1 17 . finally did. But he came to the conclusion that the Smart i is system was doing a better job than he could do. '
i ig Now that system, as well as any other, this is an
~
l ] 20 automated scanner, you understand, as well as hooked up to 1 21 a computer that analyzes the signal. So they set the thing 22 up on the pipe and the operator then goes back out of the 23 radiation zone and works the thing. And of course that is - 1 24 really fine because he has all the time in the world and he i, i , t 25 isn't going to get burned out. ; I I
195 i Now anything that they want to use has to be 2 qualified through the NDE Center. The Smart system did 3 qualify through the old test at the NDE Center, and it is 4 now going through to be qualified under our requalification 5 Program that is a tougher test. They all have to do that. 6 Now as to whether it is overcalling or not, you 7 don't know until you cut the pipe out and look and see, a MR. EBERSOLE: Any other questions? 9 (No response.) 10 Well, I want to thank all of you for coming and I it giving us this information. 12 I have identified some topics here that I would 13 like to get your agreement or disagreement. I numbered i4 these topics 1 through 14, and I picked the following. 15 MR. REED: Without knowing your numbering system, 16 we are going to have a problem here. 17 MR. EBERSOLE: Well, I will just read them off. 18 MR. MICHELSON: There are 14 listed I guess. j9 MR. EBERSOLE: Yes, there are 14. I just lined 20 them up and numbered them as they were on the sheet. , 21 I thought we should take to the full committee the 22 Post-trip loss of auxiliary feedwater, and coincidentally 23 carry that on with the interrelated matter of the Turkey 24 Point potential MSIV failure. That is Items 3 and 5. 25 MR. REED: I agree. O
_ _ . . - .- .. _ - = _ _ - - . . - _ . _ _ . 196 MR. EBERSOLE: And we would tie that together.
) 1 2 Then the inadvertent actuation of the fire suppression 3 system, I am sure that would be interesting and important 4 to the full committee.
5 I then picked the common mode problems with the 6 steam generator pressure indication and request the Maine 7 Yankee representative, who I think it was said to be doing
; 8 extensive investigations of the implications of this, that 9 he be prepared to say a few things about what they are io doing.
11 MR. ROSSI: Did you talk with him, because I think ! 12 he has left now. 13 MR. EBERSOLE: Oh, I understand he is coming back. ! N- 14 MR. ROSSI: Oh, okay. But somebody has told him is that he is expected to talk on Thursday? 16 MR. EBERSOLE: Right. He has been alerted for 17 that. 18 MR. ROSSI: Okay, fine. 19 MR. EBERSOLE: And certainly this last topic, new 20 pipe crack indications. Bill Kerr wants to hear about that 21 criticality accident, and only for that reason I would add 22 it. I don't know why, but if he needs to hear it -- I 23 don't really see that it is important. 24 Why don't you just do this in a very casual way, l 25 Ernie. Without having somebody come up here for that, why i O
. ~ - - - - - .- - . - . - -- ..
197 i don't you just do it. 2 MR. ROSSI: Okay. 3 MR. EBERSOLE: You don't need to spend a man's I 4 time for that for heavens sake. 5 MR. REED: I agree with your choices. 6 MR. WARD: How about the Dresden loss of offsite i 7 power? 8 MR. ROSSI: Well, you now have six, I will point 9 out -- well, five and a half if you keep Fermi 2 as a half. 10 MR. EBERSOLE: Right. i 11 MR. WLIE: I find offsite power interesting, but } 12 not alarming. MR. EBERSOLE: Right. It went through its course i3 O 14 or rather standard pattern. 15 MR. WLIE: If we are going to inject another one, 16 the inverter failures point up I think a real problem to i7 me, and that is the proper training of maintenance . 18 Personnel to handle this complex electronic equipment in i, these plants. I don't think it is done. 20 MR. EBERSOLE: You think that should be brought 21 into that. 22 MR. W LIE: Well, I don't know if that is that 23 important. 24 MR. REED: I don't think so. j 25 MR. W LIE: Well, the staff will probably take i !O l
198 O(j 1 care of it. 2 MR. EBERSOLE: Okay. Any other comments? 3 MR. ROSSI: It is five and a half. We would like 4 to point out that we spent 45 minutes today on the Turkey 5 Point one, we spent 21 minutes apparently on the MSIV --- 6 MR. WARD: Well, you will have to do a lot better 7 than that. 1 a (Laughter.) 9 MR. MICHELSON: You won't get all the questions to from the subcommittee. 11 MR. ROSSI: Well, obviously, that is the reason 12 they took the longest is because I guess they were the most i i3 interesting. We will try to get, especially the loss of
, (>
L- 14 feedwater one on Turkey, down to be a little more of 15 substanc e . We will try to get it a little more concise. ; 16 Do we have two hours? 17 MR. HERNAN: Yes, two hours. 18 Jesse, I have talked to the Project Manager on i 19 Turkey Point, and I am trying to arrange for him to be here 20 Thursday afternoon. He has got one problem. He has a late 21 afternoon plane to go down to the plant to have this exit 22 interview on Friday, but he sounded like he would be able 23 to be here, and I think it would help the full committee if 24 we could display some concise schematic diagrams of the 25 systems we are talking about, particularly aux feed. i ? 9 U 1 l-- - - - .- , __. .._____ - - _ - _ __
199 () i MR. EBERSOLE: Right. 2 MR HERNAN: And we should be able to answer some 3 of the questions in a little more depth. 4 MR. EBERSOLE: I would be interested in asking him 5 how he got into this box of diminished diversity. Was it 6 economic, or what was it? 7 MR. HERNAN: We feel that, we, NRR at least feel a that the Turkey Point aux feedwater system is better than 9 Davis-Besse from the standpoint of the concerns expressed. 10 The Project Manger has been down there for a week within it the last month or so and done a detailed study of the 12 system. So I think he can answer some of your questions on i3 a personal basis. O (/ i4 MR. EBERSOLE: Well, let me postulate that we have
/
is a transient and we trip it off, you know, one of these 16 weather transients -- well, the hurricane, and I am trying 17 to get that safety valve stuck open -- or, sorry --- is MR. ROSSI: The MSIV. ig MR. EBERSOLE: The MSIV stuck open. I am trying 20 to lose your steam pressure. 21 MR. HERNAN: I think he will be able to answer 22 some of your questions. 23 The other thing I wanted to say to you is we knew 24 you would be interested in the Peach Bottom crack findings, 25 and particularly for the full committee with Dr. Shewmon O
200 ( i here, I am sure he is going to be very interested --- 2 MR. EBERSOLE: Oh, I am sure he will be interested 3 in that. 4 MR. HERNAN: We would like to be able to fit all s these things in the two hours we have allotted, and 6 wondered if you had an idea of what scope you wanted to go 7 into on the cracks. I know we could spend a whole day on a this subject, and we are setting up a subcommittee sometime 9 with the Pipe Crack Committee within the next two or three to months. ii MR. EBERSOLE: I should think that we will go into 12 two or three times as much time on the cracks that we have i3 spent here with Shewmon there. 14 MR. HERNAN: Well, with two hours and these other is things, I am not sure how you can. We would like to be 16 responsive to whatever you want to spend your time on, and i7 maybe we should have Bill Johnston come down and talk to is the regulatory stance of this whole thing, or at least lay 19 out where he thinks the thing is going. 20 MR. MICHELSON: Are you suggesting that we have
! 21 too many subjects for the two hours?
22 MR. JONES: I think they could all be covered if i 23 we have the right people here and if the questions are 24 somewhat focused. 25 MR. EBERSOLE: Well, the consensus seems to be I
201 i that we wouldn't spend a great deal more time, and we could 2 Probably expand it into another meeting on that topic 3 alone. 4 MR. HERNAN: I think we will have to.
~
5 MR. EBERSOLE: I guess the whole matter of finding ; 6 refined inspection technig;ues, and what you do with them is ) i 7 you improve them, and what do you do with the old records, a MR. HERNAN: Well, I perceive this as an evolving 9 sClence of technique. It is one reason, for example, that jo our Pipe Crack Committee report is being held up somewhat, j 11 It is kind of a moot target. i l 12 MR. EBERSOLE: Rig ht . Well, with that we will just adjourn the meeting.
~g i3
\s 34 Thank you very much, all of you.
The meeting is adjourned. l 15 36 (Whereupon, at 5:35 p.m., the meeting adjourned.) 37 1
} IO ll 4
20 21 22 23 1 24 25 l
--__.--m .
/~N) s l \y- 3 CERTIFICATE OF OFFICIAL REPORTER l
2 S 4 5 This is to certify that the attached proceedings 6 before the United States Nuclear Regulatory Commission in the 7 matter of ~ ADVISORY CO:L'iITTEE ON REACTOR SAFEGUARDS 8 9 Name of proceeding: Subcommittee on Reactor Operations 10 11 Docket No. 12 plac*' Washington, D. C. t la Date: Tuesday, September 10, 1985 14 15 were held as herein appears and that this is the original 16 transcript thereof for the file of the United States Nuclear 17 Regulatory Commission. 13 (Signature) [g % (Typed Name-of Reporter) Mary C. Simons 20 21 ( 22 i 23 Ann Riley & Associates. Ltd. f} V 24 25
1 1 CERTIFICATE OF OFFIC1AL REPORTER l 3 4 5 This is to certify that the attached proceedings 6 before the United States Nuclear Regu l a tory Commi ss i on in the 7 matter of- ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 1 9 Name of Proceeding: Subcommittee on Reactor Operations 10 11 Docket No.r 12 PIace- Washington, D. C. 13 Cate: Tuesday, September 10, 1985 14 15 were held as herein appears and that this is the original 16 transcript thereof for the file of the United States Nuclear 17 Regulatory Commission. 13 / 99 (Signatur, j , l (Typed Name of Repor'tg) GarrettJ.Wd , Jr. 20 i . 21 i 22 O 23 Ann RIley & Associates. Ltd. U i 24 25 l - - . - . . - .. - . - . - . .- - .- . - . . -, .- - - - - . . -
1 l 1 CERTIFICATE OF OFFICIAL REPORTER
- 2 ,
3 4 5 This is to certify that the attached proceedings 6 before the United States Nuclear Regulatory Commission in the 7 ma t ter of : ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8 9 Name of proceeding: Subcommittee on Reactor Operations 10 11 Docket No. I 12 PIace: Washington, D. C. 13 oste: Tuesday, September 10, 1985 14 15 were held as herein appears and that this is the original 16 transcript thereof for the file of the United States Nuclear 17 Regulatory Commission. 13 (Signature) % g . (Typed Name of Reporter) Myrtle H. Walsh j 20 21 22 1 23 Ann Riley & Associates. Ltd. 24 25 _ , . , - . . . , , _ . _ , , , . , _ . _ , . , , _ , . ...,,c . _ . _ . . , _ . , , _ _ . . . - _ _ . _ . . . . . -. -.
Ag:nda fer ACRS Subcommittoa Meeting on September 10, 1985 1:00 p.m. Room 1046, H Street V RECENT SIGNIFICANT EVENTS Presenter / Office ' l Date Plant Event telephone
- Pg .'
12/84 Seabrook Main Steam Safety Valve Flow G. Hammer, NRR 2 Deficiency 492-8963 7/2/85 Fermi 2 Premature Criticality D. Lynch, NRR f 492-7050 - 8/8/85 Turkey Pt. 3 Post Trip Loss of Auxiliary J. Henderson, IE 5 Feedwater . 492-9654 . 9/4/85 Hope Creek Inadvertent Actuation of Fire J. Henderson, IE 6 1 Suppression System 492-9654 7/23/85 Turkey Pt. Potential MSIV Failure At V. Hodge, IE 7 Low Steam Flow 492-7275 4-8/85 Turkey Pt. Recent Inverter Failures H. Bailey, IE 12 492-9006 7/25/85 Point Beach 1 Loss of Offsite Power H. Bailey, IE 15 492-9006 7/23/85 Point Beach Safety Injection Recirculation T. Colburn, NRR 17
- Line Deficiency 492-4709 8/7/85 Maine Yankee Common Mode Problems with Steam J. T. Beard, NRR 19 9/3/85 Generator Pressure Indication 492-7465 8/16/85 Dresden 2 Loss of Offsite Power R. Gilbert, NRR 25 492-7128 8/10/85 Browns Ferry Seismic Design of Cable Trays E. Weiss, IE 29 492-9005 8/20/85 Callaway 1 Electric Power Transient T. Alexion, NRR 30 492-8929 6-8/85 Waterford Startup Experience Comparison Wm. Jones, IE 32 Wolf Creek 492-7613 Callaway Catawba Byron 2 (O,/ 8/28/85 Peach Bottom New Pipe Crack Indications G. Gears, NRR 37 492-8362 l
l
SEABROOK - CROSBY MAIN STEAM SAFETY VALVE O et0W DeriCieNCv - December 198a
~
(G. HAMMER, NRR) ' PROBLEM - FULL FLOW TEST RESULTS INDICATE SPRING-ACTUATED - -' MAIN STEAM SAFETY VALVES MAY NOT ACHIEVE RATED FLOW i CAPACITY, , SAFETYSIGNIFICANCE-PbSSIBLEINADEQUATEOVERPRESSURE PROTECTION OF SECONDARY SYSTEM IN PWRs OSING THESE VALVES WYLE LAB TEST RESULTS: INADEQUATE LIFT OF VALVE DISK (ABOUT 50%) WITH THE VENDOR (CROSBY) ADJUSTED RING SETTING ADJUSTMENTS. TESTS WERE CONDUCTED TO DETERMINE ADEQUACY OF DISCHARGE PIPING, i i CORRECTIVE ACTION - RINGS READJUSTED. OBTAINED FULL LIFT ON SEABROOK VALVES I l - GENERIC IMPLICATION - SEABROOK VALVES AND DISCHARGE PIPING SIMILAR TO OTHER PWRs. FULL FLOW TESTS NOT NORMALLY RUN TO ADJUST RINGS, - NRC FOLLOWUP ACTION: - (1) DEVELOPING IE INFORMATION NOTICE (2) STAFF MAY PURSUE AS A GENERIC ISSUE Q (3) DISCUSSIONS WITH CROSBY BY REGION 1 AND NRR REGARDING ADE00ACY OF VEND 0R GUIDANCE AND SRV i RING SETTINGS, g ; l
. _ , . _ _ , _ . _ _ _ _ _ _ . _ _ . . ___ _ _ _ _ _ _ _ . _ . - _ _ _ _ _ _ _ _ _ _ , _ _ _ _ _ _ . . _ . ~ _ _ _ . _ _ . -
( O s' c fs
- oc " o i
Jo) F, 1 v r % i
.i <
P I O
) ,
x ~ >
'TT] IT~T' -
~
d
" L~ 3 ,
lI f- g GUIDE RING "n i
#:s ~ pr j
1 T j % NCZZLE RING
=~ c 9 ,
\ % r ' J w i - [
\
L-.s 1 ' FIG.I TYPICAL HA1N STEAP SATETY VALVE 3 i
FERMI-2 PREMATURE CRITICALITY EVENT , JULY 2, 1985 (DAVE LYNCH, DL) o PROBLEM - REACTOR BROUGHT CRITICAL PREMATURELY FROM STARTUP o SAFETY SIGNIFICANCE - SIGNIFICANT PROCEDURE VIOLATION BUT SAFETY LIMITS NOT VIOLATED . o REACTOR BEING RESTARTED IN SOURCE RANGE AFTER SHUTDOWN. LOW POWER LICENSE ISSUED ON 03/20/85; INITIAL CRITICALITY ON 06/21/85. o OPERATOR PULLED ELEVEN RODS IN GROUP 3, ONE AT A TIME, FROM NOTCH 00 (FULL IN) TO NOTCH 48 (FULL OUT). PULL SHEET O CALLED FOR PULLING GROUP 3 RODS FROM 00 TO 04 AS FIRST INCREMENT. o ROOT CAUSE WAS R0 FAILURE TO FOLLOW PROCEDURE. AMONG CONTRIBUTING CAUSES WERE FAILURE T0: (1) ADEQUATELY MANAGE , CONTROL ROOM; (2) OPERATE AS TRAINED; (3) TRAIN AS OPERATED; (4) USE HARDWARE (RWM) TO FULLEST POTENTIAL. o SUBSEQUENT ACTIONS - COMMISSION BRIEFING ON 07/10/85; FULL POWER LICENSE ISSUED ON 07/15/85; RIII INFORMED OF EVENT BY DEC0 ON 07/15/85; CAL ISSUED BY RIII ON 07/16/85. o STATUS OF RESOLUTION - AWAITING DECO RESPONSE TO R III CAL Y
. TURKEY POINT 3 POST-TRIP LOSS OF AFW JULY-21, 1985 (J. B. HENDERSON, IE)
PROBLEM - ALL THREE TURBINE-DRIVEN AFW PUMPS BECAME INOPERABLE DURING POST-TRIP RECOVERY PERIOD. SAFETY SIGNIFICANCE - IMPAIRMENT OF DECAY HEAT REMOVAL CAPABILITY. . BRIEFING SIGNIFICANCE - THE ONLY SITE OTHER THAN DAVIS-BESSE WITH NO ELECTRIC DRIVE AFW PUMPS.
.- TWO TURBINES TRIPPED TO LOCK OUT ON MECHANICAL OVERSPEED TRIP.
(#'t
%)
FLOW CONTROL VALVE ON THIRD UNIT FAILED.
-- LICENSEE CORRECTIVE ACTION - 1. RETRAINING ON TURBINE GOVER-NOR RESET. 2. PERFORM OVERDUE
- MAINTENANCE ON AIR SUPPLY SYSTEM TO CONTROL VALVE.
OVERSPEED TRIP PROBLEM MAY BE GENERIC TO A SIZEABLE CLASS OF TURBINES
- NRC FOLLOWUP - NOT DEFINED YET O
Nh
J HOPE CREEK CO2 RELEASE
~
4 4 O 9/ae85 J. B. HENDERSON, IE-EGCB i RELEASE OCCURRED ABOUT 8:45 AM [ A SITE EMERGENCY RESPONSE DRILL WAS IN PROGRESS
/
! C02 RELEASED FROM STORAGE TANK TO PROTECTED VOLUME THROUGH NORMAL PIPING SYSTEM PROTECTED VOLUME, FUEL TANK ROOM UNDER DIESEL-GENERATOR l CO2 ESCAPED TO OTHER PARTS OF BUILDING COMPLEX l PERSONNEL EVACUATED FROM BUILDING COMPLEX iO* i f COMPLICATION - MOST THOUGHT THIS PART OF SCENARIO, NOT REAL 23 PERSONNEL TRANSPORTED TO HOSPITAL ! INADVERTENT ACTUATION SIGNAL i PROGRAMMED DISCHARGE 2 TONS , ACTUAL - 10 TONS 1 t LICENSEE BELIEVES PROBLEM MAY BE IN LONTROL PANEL PANEL SENT TO MFR (CARD 0X) FOR EVALUATION I i I a .t 1 ) b
- .---..<-._--,,..~.,,.n.,,.._. - - - _ - - . - . , , . . - . . - - . , _ . _ . , . . . - - - . _ . - , . , , , , . - - . . - - -
i 4 1 !O ~
\
l l \. t -
' l
- P0TENTIAL MSIV FAILURE .-
l AT LOW STEAM FLOW l ~ i TURKEY PT. 3, 4 ; JULY 23, 1985 ,
- BY VERN H0DGE, IE t
j . l i l ! s TWO ISSUES: t
- 1. UNANALYZED CONDITION FOR MSLB ACCIDENT f
l I I
- 2. INADEQUATE TESTING PRACTICE i
1 ( I l L 4 O i i 7
4 L UNANALYZED CONDITION FOR MSLBA !O 4 1 l 1. MSIVs WON'T CLOSE WITHOUT STEAM ASSISTANCE BECAUSE-AIR-ACCUMULATORS ARE T00 SMALL. j .
- 2. PROBLEM APPLIES TO TURKEY POINT AND ROBINSON ,
1
/
i j 3. MSIVs MUST CLOSE IN 5 SEC, EVEN AT LOW STEAM FLOW - i ! 4. THRESH 0LD STEAM FLOW NOT KNOWN i ASSUMED TO EXCEED AFW CAPABILITY l i
- CONTINUED BLOWDOWN POSSIBLE lO
I i i i l i i . , !O '
- JUSTIFICATION FOR CONTINUED OPERATION Q
j . ASSURE SUFFICIENT INSTRUMENT AIR:
- 1. DIESEL AIR COMPRESSOR BACKS UP PLANT SYSTEM - '
i l 2. FOSSIL PLANT SYSTEMS TIED IN FOR BACKUP
- 3. PLANT SHUTDOWN ON FAILED INSTRUMENT AIR SUPPLY CORRECTIVE ACTION l EXPEDITED DESIGN MODIFICATION T0:
l - t !O ! 1. ASSURE MSly CLOSURE IN 5 SEC W/0 STEAM ASSISTANCE i I ! 2. ALLOW TESTING MSIV CLOSURE W/ SECURED ELECTRIC AND
- INSTRUMENT AIR ACTUATOR POWER f f i
l0
INADEQUATE TESTING PRACTICE O MSIVs STROKE TESTED WITH INSTRUMENT AIR CONNECTED TURKEY PT, ROBINSON CITED IN FEBRUARY 1985 TO VIOLATE
- 10 CFR 50.55A(G):
IST OF SAFETY RELATED PUMPS AND VALVES MUST COMPLY WITH ASME BPV, SECTION XI / Fall SAFE VALVES MUST BE TESTED BY LOSS OF ACTUATOR POWER (PARA, IWV-3415) ACTUATOR POWER INCLUDES BOTH ELECTRIC POWER AND INSTRUMENT AIR POWER O NRC RESPONSE O AN INFORMATION NOTICE IS BEING PREPARED i l lO . t .
1 l , ( s
' 1 CLOSE ,;
c g , ,;' ," AIR i ,f' + RESERVE ) , TANKS
\
i
/
+ t j ATM W jA f ,,
N MANUAL RUPTURE TEST VALVES - s e DISC ,, BYPASS LINE s i ,,
/,e , e
+ \ ,t' ,
ll-I T-
\
n i OPEN , ? if e 5 :
- =
, s ,- s
, ,- - jn l ,
' II l
! ==-
y
~
~
l
- MSIV , 5 i
^ LIMIT SWITCH MSIV CONTROL DIAGRAM INSTRUMENT AIR rJ i
! -s \ ~ . i .
TURKEY POINT - RECENT MULTIPLE INVERTER FAILURES ,g V (H. BAILEY IE) TURKEY POINT-HAS EXPERIENCED 16 INVERTER TRIPS IN 4 MONTHS (APRIL 4, 1985 TO AUGUST 1, 1985) . FIVE INVERTER LOSSES RESULTED IN REACTOR TRIPS (SINGLE . UNIT) IN TWO MONTHS , SAFETY SIGNIFICANCE - CHALLENGES TO THE SAFETY SYSTEMS AND INTERACTION BETWEEN PLANTS AND POTENTIAL DUAL-PLANT TRIPS TYPICAL SCENARIO FOR REACTOR TRIP FROM POWER: TURBINE RUNBACK ON R0D DROP SIGNAL Q - HIGH TAVE DUE TO LOSS OF AUTO ROD CONTROL HI OR L0 SG LEVEL DUE TO LOSS OF AUTO FEEDWATER CONTROL REACTOR TRIP DUE TO REAL AND/0R FALSE SIGNALS SI ACTUATES DUE TO FALSE SIGNALS PRESSURIZER SPRAY VALVE FAILS AS IS PRESSURIZER HEATERS LOCK OUT
- DIESELS START; BUT DO NOT LOAD SPARE INVERTER IS PUT ON LINE AND SYSTEMS RETURNED TO NORMAL.
OLD "EXIDE" INVERTERS BEING REPLACED BY NEW " SOLID I STATE CONTROLS" INVERTERS O i V l /2
t ( $ l g , 2-U 8 NEW INVERTERS PLACED IN SERVICE SINCE LAST REACTOR TRIP ON 8/1/85, LTHE REMAINING 4 INVERTERS (TOTAL OF 12) WILL BE REPLACED SHORTLY, , NEW INVERTERS ARE DESIGNED TO MINIMIZE INTERACTION BETWEEN i UNITS 3 AND 4 WHEN LOADING AND UNLOADING INVERTERS, , PROCEDURES FOR LOSS OF 120V INSTRUMENT PANEL HAVE BEEN UPGRADED AND OPERATORS, RETRAINED, ,.
- GENERIC APPLICABILITY UNDER INVESTIGATION, i
\'
\O r i l .1 { l l lo .
. /3
l TUR K E Y fo t A/ T 0 I26 V V I TA L POWER BATTERf CtlARGERS - 3A 3S 4B 4A MS 38 : '
\/ \/ \/ \/ N/ N/
3A ,_. 4B 44 3B
) '
l)
' )- - - - ) ) ) ) :) ') - - I) ,)
! I I I l I t i I i) t IJA P-c 8=r 14B l 144 'ys
') ') I) I) I) h I) ') ') I) l) I)
~
O 3A AS 4h 30 DS 40 38 BS 48 3c. c3 4c INVE TERS hsw-s 1 r-
/
L usur > unir 4 I2av A-c 12.6v A-c I cP 4 I eF 4 - suses s u.s e s . I i i O lY . !
-- _ .. - - - - -- 1
i, ( e . 1 POINT BEACH 1 LOSS OF 0FFSITE POWER JILT 25,1985'0ENRf PAllEU' ~J D - PLANT OPERATING AT 100% POWER 07:04 EDT - LOSS OF OFFSITE POWER RESULTS FROM LOCKOUT OF IX04 LOW VOLTAGE STATION AUX TRANSFORMER - SEVERAL BUSES INCLUDING SAFEGUARDS BUSES WERE AFFECTED. EDGS START'IMMEIDATELY AND LOAD ONTO SAFEGUARDS BUSES. / 3 LICENSEE COMMENCED SHUTDOWN PER TECH SPECS. , 07:49 EDT - POWER RESTORED TO REMAINING BUSES FROM MAIN GENERATOR. 09:57 EDT - 1X04 TRANSFORMER DECLARED OPERABLE OFFSITE POWER RESTORED. UNIT BEGINS RETURN TO FULL POWER. 10:17 EDT NORMAL OFFSITE POWER SUPPLY RESTORED TO SAFEGUARDS BUSES. EDGS SECURED.
.ALL SYSTEMS FUNCTIONED AS DESIGNED.
t e oO m t
/5
Offsite Power
~
G ; o o o o L y _ 345KY g g g O " 1XO3 1X01 2XO1 0 2XO3 ww ww ww vGw mm mm mm nnnn
/
13.8KV O ' O n 19 KV 2XO4 2 02 1XO4 3,ggy 2X02 Mein Gen- ww ww ww ww Main Gen. 4.16 o 1 o . . x E3 O O 1E4 O' O 2E 2T3O
^
LJ I O l l
, l l
? _? emergency l'A01 1A02 Diesel Gen.
l ! G02 1A05 1A06 l l..................... l
- Engineered l Safeguards Emergency -
Bus Diesel Gen. G01 I S.lMPLIFIED DI AGRAM OF POINT BEACH O ELECTRICAL DISTRIBUTION SYSTEM m
POINT BEACH 1 AND 2 - SAFETY INJECTION RECIRCULATION LINE DESIGN DEFICIENCY JULY 23, 1985 (T. COLBURN, NRR) (V7 BOTH UNITS AT 100% POWER, PROBLEM: SINGLE FAILURE OF CONTROL CIRCUITRY FOR EITHER SAFETY INJECTION RECIRCULATION LINE ISOLATION VALVE . (SI 897 A OR B) COULD RESULT IN OVERHEATING AND DAMAGE TO BOTH SAFETY INJECTION PUMPS, FAILURE OF CONTROL CIRCUITRY ALSO FAILS CONTROL ROOM / ANNUNCIATOR FOR "SI 897 A OR B SHUT", (VALVE SHUT, OPERATOR UNAWARE) PROBLEMONLYAPPLICABLEFORSPURIOUSSISIGNALSANDSM5LL BREAK LOCAS WHERE PRIMARY PRESSURE IS MAINTAINED ABOVE SI PUMP SHUT OFF HEAD (I.E. THOSE REQUIRING MINI-FLOW RECIRC) PROBLEM DISCOVERED DURING E0P REVIEW AND VERIFICATION ADMINISTRATIVE CONTROLS ENACTED TO PROVIDE SHORT-TERM FIX A) MINI-FLOW.,RECIRC VALVES MANUALLY LOCKED OPEN (RESULTS IN DEFEAT OF INTERLOCK WITH CONTAINMENT SUMP ISOLATION VALVES SI 851 A & B) B) LOCKED CONTROL COVERS FOR SUMP ISOLATION REMOTE AND MANUALOPERATbRS C) PROCEDURES ENACTED AND OPERATORS TRAINED ON MANUAL INITIATION OF SUMP RECIRC FOLLOWING LOCA , D) WESTINGHOUSE INFORMED OF DESIGN DEFICIENCY (LICENSEE LETTER OF 7/24/85) SEVERAL LONG TERM FIXES BEING CONSIDERED, SCHEDULED TO BE SUBMITTED BY 9/25/85. O - eART 21 NOTICE ISSUED BY LICENSEE; PROBLEM MAY BE GENERIC TO OTHER WESTINGHOUSE TWO-LOOP PLANTS
/7
DIAGRAM OF POINT BEACH ECCS O . REFUELING AJf"11K A K
/
e X A F.C. 8,7A M: 4: F.C.
=
.. .p i
O ,, C.
= C')
{. Ag x, - [ k J lf 1 (!) l 4 x x x _ s a _ 1.
=
W (":9 A .s, A _ RCS rm - C a CONT N M E NT i (":") ,,
.s,.
=~
O 4
3 MAINE' YANKEE: TWO COMMON MODE PROBLEMS WITH STEAM l, - GENERATOR INSTRUMENTATION (J. T. BEARD)
SUMMARY
AUG. 7, 1985: PLANT AT 78% POWER (E0C), FOUND ROOT VALVES "NOT FULLY OPEN" FOR 9 0F 12 PRESSURE TRANSMITTERS. -
/
SEPT, 3, 1985: REFUELING OUTAGE TESTING REVEALED THE "A" PRESSURE CHANNEL FOR EACH STEAM GENERATOR WOULD NOT TRIP---
; DESIGN MODIFICATION ERROR---GROUNDING.
SIGNIFICANCE: PRESSURE INSTRUMENTATION WOULD NOT HAVE RESPONDED PROPERLY TO STEAM LINE SREAK ACCIDENT; BOTH CONDITIONS EXISTED FOR WHOLE OPERATING CYCLE. DIVERSE INSTRUMENTATION (E.G., S/G LEVEL) WOULD HAVE ACTUATED, LATER. ROOT VALVE PROBLEM f
"D" PPESSURE CHANNEL ON S/G #1 NOTICED TO BE " SAGGING";
FOUND 3 VALVES FULLY CLOSED (BUT LEAKING), 6' VALVES -
" PARTIALLY OPEN."
VALVES WERE CLOSED FOR HYDRO-TEST; SPRING, 1984. I
i t Roof VALVG PKoB L GM . ,
/Afsyt . CHAndct-
. ./
A B C b ! pt4LG CLosEb CLaLKb I l
! cuarb SMM y,7ase , psgratty A "
WWUW CL M CMEb l Aerm maiwetr nz: raw / l CLesab CLoSO Closel l e O k E 4 l l l l . l l 1 l l l en. oe 2D . . e ,e, .gm- ea e m me 4 t
f 2-CJ INSTR. ROOT VALVES NOT ON ANY CHECKLIST. LICENSEE ASSUMPTION WAS: "lF INSTRUMENT INDICATES, VALVES MUST BE OKAY " LEAKAGE CAUSED SATISFACTION STEADY-STATE INDICATIONS. SLOW TRANSIENT RESPONSES (UP TO 60 SECONDS).
~
d
~
DESIGN PROBLEM 4
- EACH S/G PRESSURE CHANNEL " AUCTIONEERS-LOW" SIGNALS FROM ,
TRANSMITTER LOOPS FROM ALL THREE S/G'S, MODIFICATION MADE TO TAP OFF ANOTHER SIGH'AL FROM EACH LOOP TO PROVIDE INPUTS TO SUB-COOLING METERS. DEFICIENCY IN DESIGN OF MODIFICATION---ANOTHER SIGNAL COMMON ESTABLISHED WITHOUT CHECKING COMMON IN EXISTING CIRCUIT FOR RPS. DESIGN REVIEWS FAILED TO FIND ERROR; MOD. INSTALLED AFTER 18-MONTH SURVEILLANCE TEST; NO POST-MODIFICATION TESTING , OF RPS. UNDETECTED FOR ENTIRE OPERATING CYCLE. t i 9 O
.;l/
s -.
.r- - , , - . e ,n - - ,- , ..w,e-,- , - - - ,v.-
. - s -,w ap. ,..
.g
^
M64 M.+-4*4. N. 6_ee.e e.> ;M N'8' **EM
.. M =W6 .M* . w. e.
e .em o @ MM
.9
.-..... . _ ........M
.. y
. ....g w&
w . . ..
.N D. . ... .._ _ .._._ .
y
~
4
._-.. t l J
T 4 4 M > 0 4 g / w L <
. . . . < g
_ . _ . ..F J at
-U
..g.- O. .
e a. e.. e e C .. w
%b h . -
a a x .
. ._. ._ .% g
,$ ,_.1 i .l e g ,
I . O . . . a e E i j i
. i g
e l ,> g (
. .J e. - -- ., k T 1
\
O . l ' N a p
) , g }3 I L d 4 a o9 l Q.
T %g so U l
=
4 r
%J -
w i , i oc . ' o -
. .y . .
u m ,
.w wu e2 u m =u d < 88 y w M K et, w M
.) A b ..
A 6 4 0 - W y
._y4
% b< J1
__ 4'. . - .- gy - . , . l W S l I
a --u_. 4 a- a eh._- ...___,,s- w _ a. gg _g . .6.b._ .9 &_ ..Sep&* .M g. It a6 . mm ee w -<. ell *w .a. == _ = . D N D8
..g . _ .
--_- _. . . . _ . . . ~ . . - - . -. , g . ._
_ . . _ _ .. . . . . . _ . . - . . . . . . _ _ . . . . . .g . w
.._ g -- . . .
- b. . . . . ...,_...._. .
p
.r - . ,
T. >- O . - . .
.. 4 .s
} $
o g
.. . ..t J SC U -
n U - i 4
! \
W C M de - 4-D .. oc . or et b g
. ., ,4 8
.1 ;
49 _ . .-_. e g. . U l I I. . . . 4 *
' . t e
.d ..
4 . _ . . ,
, I l g %
g 1 . . .t *. --- ., > \ k I
* . _ .. .> \$'
o m
~
N t 4
.d R et I
y to
--Q q P
=,%
I *J g ; ---- - _
- i b
_. . . . _ . . g. .w. . . . _
- 4. b. ...
wh h M T T w 88 < y og - gy = - y A b n. t- S. w - - - ,- m . . . . .. . . ._. a3-
-g
,g .. . .. . . _ .(
1 l
Roof VALVG PKo 5 L CA - i
/Af572. . CHANdfL
. _ _ _ . ./
.A E c h ActredtcY l CLostb CL*L*b csaccb D~"# marmatt notrgtty
$ WOW $ ""
esan cwo i AmrM MutTwetar Ma:reetty 3 CLasch l ek10 CLa$r) I 1 l l . o a e D e 49 *'b
7s1 . DRESDEN UNIT 2 - LOSS OF 0FFSITE POWER
AUGUST 16, 1985 (R, GILBERT)
EVENT SEQUENCE
- UNIT 2 AT 70% POWER, UNIT 3 AT 90%, UNIT 1 SHUTDOWN '
AT 0030 FAULT OCCURRED IN UNIT 1 TRANSFORMER - TR12 UNIT 2 RESERVE AUXILIARY TRANSFORMER (TR22) WENT DEAD DUE TO POINT OF CONNECTION TO 138 KV BUS ,/ LOST POWER TO ONE OF'TWO ELECTRICAL POWER DIVISIONS AUTOMATIC TPANSFER TO UNIT 2 AUXILIA'RY TPANSFOPMER - (TR21) FAILED DUE TO DESIGN DEFICIENCY
- LOSS OF POWER TO FEEDWATER PUMP; LOW REACTOR WATER LEVEL; PLANT TRIP
- LOSS OF POWER TO UNIT AUXILIARY TRANSFORMER; COMPLETE LOSS OF OFFSITE POWER (LOOP) TO UNIT 2
( }) - RECOVERY SEQUENCE UNIT 2 AND SWING DIESEL ACTIVATED AND PICKED UP LOAD OPERATOPS CLOSED MSIVS TO CONSERVE RCS INVENTCRY
~
ISOLATION CONDENSER ACTIVATED USING DC OPERATED VALVES
- MAKE-UP TC 150. COND, VIA DEMIN NOT AVAILABLE PUE TO LOOP; USED CONDENSATE STORAGE TANK-(CST) AS BACKUP (PUMPS Ot' ESF BUS); WATER SLIGHTLY CONTAMINATED RCS MAKE-UP SUPPLIED BY A CRD PUMP REACTOP LEVEL AND PPFSSUPE STABILIZED THREE HOURS IMT0 EVEUT .
0FFSITE POWER RESTORED AFTER FOUR HrDRS
- IUCIDENT RESPONSE CENTER ON STANDBY; REGIONAL ADMINISTRATOP MENT TO SITE.
d
STAFF CONCERNS AUTOMATIC TRANSFER DESIGN DEFICIFNCY - CECO HAS FIXED
- TO RESTORE TR 21, MAIN GENERATOR DISCONNECT LINKS f MUST BE REMOVED. TAKES AT LEAST 24 HOURS; SCAFFOLDING AND PROCEDURES NEEDED ,
' - 4160V CROSS TIE TO UNIT 3; SECOND SOURCE OF 0FFSITE POWER; GDC 37 CONFORMANCE / HOOK UP POINT OF RESERVE AUXILIARY TRANSFORMER TO 138KV BUS i - LICENSEE ASKED TO REVIEW CONFORMANCE TO GDC . lo I, l u 1 4 i { < t !O
- a , _ - 2 a aL..-, a
~
- O l
9 . O 3+ > rA* i 4-C 8 oi av r _
- ~
7 EE
~
+
g 4 s
*g.,
r-}j
~
+
o g 7- p- ,
,B1
,J ,^: -
+> t.* _ .
2W
=
+> . .
O #. i a_ - 4 l :
; -.a
. n W^!
" W s.I .e.
&h h 5 t< a ~
p2 m -- w 2
-s 7
*> 0 < i h we C i p
=#
2 y
#> , 2 r x v :: ,
=
h 3L g*5 O s. 2* 2
O i
~ ~ ~ ,,. .
N E T ? -8 C ' T. .Y .
- u_
^ ;...-. n 6
J [.. Til'1
@=--
c.t::. f .D
" ~
LL .'.". I
- ' G mgm ptepeW . - - -
g , 2 - 9 q ,, T
- f ')"
r- : (= i:: t ' t' @ L'" ,": , ,
. $g =
!, e
_? _ m ..
,c, -
e 1
- v :
l __ _
= .: : , , = > T ::ra
.= .
n .e l-- _ ,. h, m u ,v.
~
g
. x# . t g ,_
h ., f.f V-g6 ' ,X n .. _
~
l %- l l l FIGURE 4.6.2:1 ISOLATION CONDENSER - PIPING DIAGRAM O 28
O BROWNS FERRY CABLE TRAYS SEISMIC QUALIFICl AUGUST 19, 1985 (ERIC WEISS, IE) PROBLEMS - CABLE TRAYS "0VERFILLED" - KNOWN j
- ALL SEISMIC LOADS NOT CONSIDERED# IN DES
- MANY HOLD DOWN CLIPS HAVE MISSING BOLTS
-- CABLE +FLAMEMASTIC+ JUNCTION BOXES EXCEED WEIGHT OF TRAY
- CABLE AMPACITY QUESTIONABLE DUE TO FLAMEMA
- NO SLACK.MAKES RELATIVE MOTION CAPABILIT TO TRAYS QUESTIONABLE Q - ANCHORS SPECIFIED ON DRAWINGS DIFFERENT T IN INSTALLATION
- ERRORS FOUND IN ORIGINAL SEISMIC CALCULATIO
- SOME TRAYS DESIGNED WITHOUT SEISMIC CRIT LOAD ONLY)
SAFETY SIGNIFICANCE - NOT SEISMICALLY QUA
- OR LONG-TERM OPERATION
- STATUS OF RESOLUTION - TVA HAS HIRED 8 UEaC , PERSON INTERIM EVALUATION REGION 2 HAS INSPECTED SEQUOYAH - DOES N THAT PARTICULAR PROBLEM, O
)
. . -- ._ -_ . ._ . __ . _ = _ .
CALLAWAY 1 ELECTRICAL POWER TRANSIENT AUGUST 20, 1985 (T. ALEX10N) PLANT OPERATING AT FULL POWER OPERATOR " BUMPING DOWN" GENERATOR OUTPUT VOLTAGE (REQUEST
/
FROM LOAD DISPATCHER) , LOSS OF VOLTAGE REGULATOR, LOSS OF MAIN GENERATOR FIELD, SWITCHGARD VOLTAGE FELL FROM 350 KV TO 280 KV INSTANTANEOUSLY GENERATOR TRIPPED ON LOSS OF FIELD FOLLOWED BY REACTOR TRIP LOST POWER FROM AUX. TRANSFORMER TO NON-SAFETY RELATED BUSES, RCPS TRIPPED DUE TO FAST POWER LOSS, SYNC.- CHECK RELAY DROPPED OUT PREVENTING AUTOMATIC FAST TRANSFER FROM AUX. TRANSFORMER TO 4 S/U TRANSFORMER-AUTO DEAD BUS TRANSFER AFTER 2-3' SECONDS, RE-ENERGIZED NON-SAFETY RELATED BUSES ' NO LOSS OF POWER TO ESF BUSES (1 BUS LREADY ON S/U TRANSFORMER, 1 BUS DIRECTLY ON SWITCHGARD) FEEDWATER ISOLATION AND AFW AUTO-START ON REACTOR TRIP AND LOW TAVG (NORMAL), MANUALLY CLOSED MSIVS TO LIM'IT COOLDOWN RCPS RESTORED IN 5-10 MINUTES VOLTAGE REGULATOR COMPONENTS (POTENTIOMETER AND CIRCUIT CARD) REPLACED AND SENT TO GE FOR ANALYSIS , SIMILAR EVENT IN JAN. '85, CAUSE INDETERMINATE, VOLTAGE REGULATOR CIRCUITRY WAS CLEANED AND' CHECKED 3o
[ l i l I
!'i!
j , ,
! l l
- f. ..
J4(;xv Swircumab
;i ; I l: '
l.
- ; '!,,., ; j
.~ '
! J4fkV ;
l l ;i ; .]'.ll . i n.,,wl. .. i lI ___J !;! i. I ' t i i
' 1 .
l l L - i i ; ,
! l - ! Il ! l t -
lil l s
' j l mar # j; l- l, l w g. l lxmz t i -
; i
, i i , 4
( l
- l e
l
. i l - -
' l / ,
- l. l l i * -
i gja
, Aux.,
l
;i WM XMA. M ggg j !
34rav - l
'. l -
Is.r kt i l I i- , , l I I I-
, ). +) ; .) ,.+) l . . , . ! ,
' i
; I I i !
i - F ;Mut - SANTY s Butit . l l i l
! ! j ~~
l i l I (!,IJ.Pav) ' I ~ fST i i ! ~~ xmps misz.r a v; <
"*4 I
M.t n Gam) f
- ; l 4cd. .1, .
i l \(' . I l ! l, i i' j t i l}'
. I i
I i i i I. rsosur< M<e n o v. %. li!l lI 1
. l .
i i , i t ! . ; i; i l ' - 1 I - + i l!
; I i : 1 : '
I I
; ,i! l !
- ! . : j l !
i
' ' ; . i
. l ,i ! j i
i 6
. s .
. ! l-
, . i t '
[d&Id I.IY(N . i i l . j- .. M dMMMI' i ; [dI(* - I :
, f ,
l '% '. , i : i ' I !
! I ! ,
! , l f l f .
i i i i l
- i , : ! i , . . i. I ,
i c , l , l
_s -4 A-- -4 .A - -- - amm--- m - o a- 2 a - --- -m 1--, 4 4 j h STARTUP EXPERIENCE COMPARISON (W. JONES, IE) FULL POWER FACILITY INITIAL CRITICALITY LICENSE DATE , l CALLAWAY 10-02-84 (10-18-84) CATAWBA (01-17-85) b BYRON 02-03-85 (02-14-85) . WATERFORD - 3 03-04-85 (03-16-85) WOLF CREEK 05-22-85 (06-04-85) BASED ONLY ON 50.72 NOTIFICATIONS-- SIGNIFICANT EVENTS ONLY (I.E. , NOT CONTROL ROOM VENTILATION ISOLATIONS, ETC) () l I f i . I a g 4 3 2.
4 f 1 i . REPORTED EVENT COMPARISON
SUMMARY
I i MONTH l
> l 1st 2ND 3RD 4TH l .
4 4 WATERFORD 4 2 3 7 l WOLF CREEK 5 6 2 l CALLAWAY 11 2 3 1 ./
~
7 3 2 3 BYRON CATAWEA 5 0 1 6* i. I i, MONTHLY 6.4 2.6 2. 2 4.25 AVG. i ) l *1 MONTH SHUTDOWN NOT INCLUDED IN PERIOD l I ! . l l ! l . i t 33
J . i REACTOR TRIP AND LOSS OF FEE 0 WATER COMPARISON DETAIL SIGNIFICANT REACTOR + INVOLVING MONTH EVENT REPORTS TRIPS LOFW ~ 1 4 4 3 WATERFORED 3 ' ' 2 2 2 0
- l 3 3 '2 1 4 7 7 3 .
SUBTOTAL 16 15 7 WOLF CREEK 1 5 2 5 .I. 2 6 5 5 l 3 2 2 0 13 9 10 CALLAWAY 1 .11 6 8 2 2 2 0 3 3 3 2 4 1 1 0 17 12 10
.~
CATAWBA- 1 5 3 1 O 2 3
'O 1 -1 O O 1
2 4 6* 3 12 7 4 BYRON 1 7 6 2 2 3 3 0 3 2 1 1 , 4 3 3 0 15 13- 3 : ! *MAY REFLECT RETURN TO POWER AFTER LENGTHY SHUT 00WN , I i
+In 1983-84, AVERAGE NUMBER 0F TRIPS FOR 4 MONTH PERIOD WAS 2.1 FOR ALL PWRS l (BASED ON AE00 & NRR STUDIES)
O 3'l 1 .
. . -- - - .. _. .- - . .--. - _ - . . . . . . . - - - . - . - . . - - . . . . . . . =_ ~
PERSONNEL ERROR & EQUIPMENT PROBLEMS COMPARISON DETAIL l O SIGNIFICANT PERS EQP MONTH EVENTS ERROR PROB OTHER l - 1 I, WATERFORD 3 l 4' 3 1 0 , 2 2 1 1 0 , l 3 3 2 1 0 i 4 7 2 4 1 16 8 7 1 WOLF CREEK 1 5 3 2 0 ./ i I 2 6 2 4 0 4 3 2' 1 1 0 13 6 7 0-1 11 2 7 2
- CALLAWAY 2 2 0 1 1 ,
l ! 3 3 0 3 0 i 0 1 0 4 J . 17 2 12 3 - l CATAWBA 1 5 3 1 1 l 0 0 0 0 2 3 1 1 0 0 4 6 3 2 1 12 . 7 3 2 1 7 2 4 1 BYRON 2 3 2 0 -1 3 2 0 1. 1 4 3 0 3 0 15 4 8 3 l I i [ 1 O 35 ; F ___--___m----_..-<_r
,w,-- ,,w- ,-,w-,-mm..y%-,,, ,_,e.m.3 , . , _ , , - - . . ,,- v- e c-# +~_q- -.--4 ---rw.<gm~m --
c - a w- .- w H e-mm
. _ . _ .__ _ _ _ . . . _ _ _ . . - _ _ _ . _ , . . ~ . . _ _ _ _ . _ . _ _ . _ _ _ . . . . _ _ _ _ _ _ _ _ _ . . . _ . . .
?
1 i i
- t 4
kC a 4 I-h LONGER TERM EVENT REPORT / TRIP DATA t e d
- i. . EVENT REPORTS 1,
- ; l ,
j
- MONTH 1 2 3 4 5 6 7 4
. CALLAWAY 11 2 3 1 4 2 1 ,
i i. ! CATAWABA 5 0 1 6* 2 0 -
/
i 3 0 l' BYRON 7 3 .2 3 5 4
- i j
REACTOR TRIPS . l
- MONTH 1 2 3 4 5 6 7 I
4 . CALLAWAY 6 2 3 1 3 1- 1 CATAWABA 3 0 1 3 0 0 - BYRON 6 3 1 3 5 2 0 l. h, f l AVG 5 1.6 1.6 2.3 f.6 1 i i f i ! 1 l ! I f l i . 6 ! l I e t i t 1 I l i i
- - t i.
36 ( w . - r ere -. v wrw y, - m:--- _eww- p.,-wwwe-,e--s-,-,.y.%w.--ww+ -e.w'e+.---w-, e,: e+ mw er e-W-
PEACH BOTTOM UNIT 3 --NEW PIPE CRACK INDICATIONS O' AUGUST 19, 1985 (G. GEARS, NRR) PROBLEM - REINSPECTION OF UNIT 3 RECIRCULATION AND RHR LINES FOR IGSCC (GENERIC LETTER 84-11) REVEALED NEW INDICATIONS. J NUMEROUS CRACKS INDICATED BY NEW GE " SMART" UT SYSTEM. MANY CRACKS DISCOVERED ON WELDS WHICH HAD NO INDICATIONS .
~
UNDER PREVIOUS INSPECTIONS (1983, IEB 83-02). 4 ONE CRACK WAS CALLED AT 360*, 35-55% THRU-WALL. O SIGNIFICANCE - HIGH PROBABILITY CRACKS WERE " MISSED" UNDER EARLIER INSPECTION PROGRAMS.
- THESE FINDINGS CONFIRM RECENT TESTS RESULTS ON CUT-0UT WELDS (1.E., CRACKS HAVE BEEN
! " MISSED").
- QUESTIONS NOW RAISED ON QUALITY OF REINSPECTIONS.
j BWROG BEING ASKED TO ADDRESS THESE ISSUES IN A SEPTEMBER 20, 1985 MEETING WITH NRR. O
- s7
PEACH BOTT g NIT NO. 3 (V~') ULTRASONIC TECHWDUE COMPARIS0N b,) , GENERAL ELECTRIC SOUTHWhSTRESEARCH No specific IGSCC detection or PERSONNEL: No specific IGSCC detection or P_ERSONNEL: sizing training. Only procedure demon- sizing training. Only procedure demon-1 stration by representative personnel re- stration by representative personnel re-g quired by I.E. Bulletin 82-03. quired by I.E. Bulletin 82-03, 8 All examinations performed EQUIPMENT: All examinations performed EQUIPMENT: 3 manually utilizing 1.5 MHZ transducers, manually utilizing 1.5 MHZ transducers. No sophisticated sizing transducers No sophisticated sizing transducers available for indication discrimination ovallable for indication discrimination or sizing, or sizing. . All Level IIand III personnel PERSONNEL: All Level II and III personnel PERSONNEL: 1 are EPRI trained and certified in the are EPRI trained and certified in the All Level I per- detection of IGSCC. All Level I per-g detection of IGSCC. sonnel are~ trained to the level of their sonnel are trained to the level of their
' All personnel involvement in exams. All personnel involvement in exams.
utilized for sizing of indications are utilized for sizing of Indications are 5 EPRI trained and certified in planar EPRI trained and certified in planar flaw sizing. flaw sizing. All exams are performed with EQUIPMENT: All exams performed manually EQUIPMENT: Where configu- utilizing 1.5 MHZ transducers. All sus-2.25 MHZ transducers. rations permit, the exams were performed pected indications are scanned utilizing with the " SMART" automatic system. All the "SLIC-40" transducer for discrimina. tion questionable indications are scanned and subsequent sizing. utilizing sizing transducers such as - U "SLIC-40", "WSY-70", OR "RTD" to aid in @ determining indication origin. _}}