ML20209E932
| ML20209E932 | |
| Person / Time | |
|---|---|
| Issue date: | 07/09/1985 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1425, IEIN-85-50, NUDOCS 8507120352 | |
| Download: ML20209E932 (173) | |
Text
- ""
ORIGINAL UNITEMSY'kTES OF AMERICA NUCLEAR REGULATORY COMMISSION In the matter of:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS Subcommittee on Reactor Operations Docket No.
.O U
Location: Washington, D. C.
~
N$5Nhhhl)fYN 2030t lemove from ACTS Office l
ANN RILEY & ASSOCIATES Court Reporters (q
1625 I St., N.W.
j Suite 921 i
- shington, D.C.
20006 71 g 2 850709 (202) 293-3950 f
T-1425 PDR
1
)
/~N.
i g
)
i UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION
.3 4
Advisory Committee on Reactor Safeguards 5
Subcommittee on Reactor Operations 6
7 Room 1046 8
1717 H Street, N.W.
9 Washington, D.C.
10 Tuesday, July 9,
1985 11 The Subcommittee on Reactor Operations c o n v e,n e d,
12 pursuant to notice, at 1:00 o' clock, p.m.,
Mr. Jesse Ebersole, 13 Chairman of the Subcommittee, presiding.
14 PRESENT:
15 Jesse Ebersole, Chairman 16 Glenn Reed, Member 17 Charles Wylie, Member 18 Dade Moeller, Member 19 David Ward, Member 20 21 ALSO PRESENT:
22 Herman Alderman, Designated Federal Employee 23 24 25
lA1; f
'1 PRESENTERS:
2 E.
Hernan 3
E.
Jordan 4
S.
Minor 5
D.
Powell 1
6 D.
Holland 7
J.
Sullivan 8
G.
Rivenbark 9
D.
Jaffe 10 V.
Hodge 11 E.
Weiss i
12 A.W.
DeAgazio 13 J.
Stols 14 A.
Dromeric 15 16 17 e
l 18 19 20 21.
22 23 24 25
2 OV 1
P ROC EED I NO S 2
(1:00 p.m.)
3 MR. EBERSOLE:
This meeting will come to order.
4 This is a meeting of the ACRS Subcommittee on Reactor 5
Operathons.
6 I am J.
Ebersole, Subcommittee Chairman.
7 The other ACRS members in attendance are 8
Messrs. Dade Moeller, Glenn Reed, David Ward, and Chuck Wylie.
9 The purpose of this meeting is to discuss recent 10 operating occurrences at the reactors.
11 Herman Alderman is the ACRS Staff member for this b(
12 meeting.
13 Notice of participation for this meeting has been 14 announced as part of the notice of this meeting that was 15 published in the Federal Register on June 21, 1985.
16 It is requested that each speaker first identify 17 himself or herself and speak with sufficient clarity and 18 volume, so that he or she can be readily heard.
19 We have received no written comments from members of 20 the public and have received no requests for time to make oral 21 statements from any member of the public.
22 I would like to ask if any subcommittee member needs 23 to make a statement prior to opening the session here?
24
[No response.]
25 MR. EBERSOLE:
If not, we will proceed with the
%,a 2
"S t
.pfl' 3
Ug 1
meeting in accordance with the schedule.
2 I want to call your attention to the last topic of 3
the day, which is this fascinating situation at Davis-Besse, 4
an event probably long in coming to many of you, in addition 5
to myself.
So we are trying to economise on our time in the 6
early part of this meeting, so we may have a little bit of an 7
expanded session on that one.
8 I would like to turn the meeting now over to Mr. Ed 9
Jordan, who will introduce the rest of the program today.
Ed, 10 it's yours.
11 MR. JORDAN:
Okay.
As you indicated, Dr. Ebersale, 12 we're here to give you the bimonthly events briefing.
I wculd 4
13 indicate so that you don't expect too much on the last item, 14 that is a status briefing.
There is an ongoing investigatien, 15 and there will be a report issued July 22 with the full 16 findings of the Staff, subsequent to making recommendations 17 for the Commission meeting scheduled on or about July 24th, 18 and 'obviously an ACRS meeting scheduled at about the same 19 time, the next available meeting.
So that will be a special 20 meeting aside from this session.
21 MR. EBERSOLE:
At that time, when we take it up, I 22 would like to ask the committee members to contribute any
[n}
23 questions they might want to include in the investigations now 24 ongoing on Davis-Besse.
25 MR. JORDAN:
So the first event, of several events 2
O 1
that we will discuss, are those listed as (6),
(7), and (8) on and Sid Miner is the presenter Rancho Seco 2
your listing 3
from NRR.
His schedule necessitates catching a flight after 4
this presentation, so we will make him first.
5 The point is made that Sid will not be available for 6
the full committee meeting, so if those items or some of those 7
items are selected, we will have some other presenter make 8
those presentations.
9 MR. EBERSOLE:
One other thing.
As these topics are 10 taken up, I would like to ask the members to make notations as 11 to whether they regard them as suitable for the full committee k' _/
12 meeting.
We can only have a few, as you know, in the limited 13 time we have.
14 MR. MINER:
My name is Sid Miner.
I am the NRR 15 Project Manager of Rancho Seco.
16 I am going to discuss three items that occurred at 17 Rancho Seco while they were coming out of the current 18 refueling outage.
19 The first one happened on June 1st.
20 CS11de.3 21 The plant was shut down, and they were running a 22 surveillance test on the emergency diesel generator.
When f
23 you finish running the diesel generator, when you push the v
24 button to shut it down, it goes into what they call a 25 maintenance shutdown mode.
It means that the output breaker
5
[
1 opens.
The diesel generator idles down from 900 to 600 rpms, 2
stays at 600 rpms for about fifteen minutes, and then closes 3
to rest, stays there fifteen minutes to cool down.
4 So while the diesel generator was in a maintenance 5
mode, at the same time the Licensee was doing some work on a 6
newly installed Class 1(e) bus and had the energizer bus --
7 this bus was parallel to the bus that the diesel generator was so an undervoltage signal, as a result of the 8
connected to 9
generator -- generated an undervoltage signal which would 10 bring the diesel generator up to 900 rpms, closing the output 11 breaker.
b)
(_
12 The UV signal drops off, but the maintenance 13 shutdown signal has a 30-second timer, so it's still active.
14 So when it sees some of the UV signal shutdown, it then starts 15 to slowly open the breaket and slow the diesel generator down 16 again.
17 At that time, the UV signal comes in and closes the 18 breaker and speeds the generator on.
19
[ Slide.]
20 It looks something like this, sort of a circular 21 path, and the diesel generator just oscillates.
It just keeps 22 oscillating, 23 They have solved the problem by putting in a relay, 24 a 100-voltage signal 25
[ Slide.3
6 l
\\'-,/
1 That will bypass the maintenance mode.
2 This was seen for the first time at Rancho Seco.
3 When you test diesel generators, you generally don't test them 4
when the offsite power is off.
So this particular problem 5
existed from when the plant started up.
It could be a 6
potential problem with anybody that has these type of GM --
i 7
General Motors set, and the 9(e) notice on this is in 8
preparation.
9 MR. EBERSOLE:
Can you tell me why they have a 10 peculiar maintenance shutdown mode position, when it could be 11 just' a straight manual maintenance shutdown?
12 MR. MINER:
I really don't know.
You need this 13 timeframe to burn off carbon.
14 MR. EBERSOLE:
That's easily manually supervised.
15 It sounds to me like a degree of automation which is 16 excessive.
17 MR. MINER:
The machine and the startup circuitry 18 was designed that way originally.
I think other ones nay have 19 the same thing.
20 MR. EBERSOLE:
You said "the other ones."
In the 21 ordinary modes, don't they simply just take manual precautions 22 to burn up the carbon and so forth before they shutdown?
You 23 mean they automate this shutdown like this?
\\
24 MR. MINER:
I think every time they shut it dcwn, 25 when you go up and push the shutdown button, it goes down to
1 l
7
( ~\\
\\,,)
1 1
the 600 rpus and stays there for fifteen minutes, whether you 4
2 shut it down from the control room with a normal shutdown or 3
with what they call maintenance shutdown -- same thing.
4 MR. EBERSOLE:
Okay.
l 5
MR. WYLIE:
Just a matter of curiosity.
Whose 6
machines are these?
7 MR. MINER:
General Motors.
J 8
MR. EBERSOLE:
One would observe, I think, that 9
sounds like a most ordinary and routine manual operation you 10 would execute without all the automation.
It sounds like a 11 lot of 12 MR. MINER:
It could very well be.
13 MR. MOELLER:
How crucial is the fifteen minutes?
14 MR. MINER:
You mean to burn off the carbon and cool 15 it down?
16 MR. MOELLER:
Yes.
Say you let it go forty minutes?
17 MR. MINER:
don't think it would make a bit a 18 difference.
19 The Licensee was checking their reactor trip 20 breakers prior to startup.
21 CSlide.]
22 During this refueling outage, all the reactor trip 23 breakers were refurbished at GE in accordance with the B&W f
24 Owners Group program and then were sent on to B&W for 25 certification and then shipped to the plant.
The Licensee
. _ _.. -. _. _. _ _. _ _ _ _ _, ~. -
l 8
1 installed these breakers in their cabinets, and while running 2
a' trip test, one breaker failed to trip.
3 The undervoltage paddle --
4
[ Slide.]
we're talking about this undervoltage paddle 5
6 jammed against the armature, and when the breaker trips, it 7
trips up this way (indicating) and pushes the paddle.
The 8
breaker stayed in the non-trip position, remained operable.
9
[ Slide.]
10 The Licensee then checked the breaker and found that 11 the dimensions, the clearance between that and the armature f
(
12 here (indicating) was excessive.
I think the specs called for 13 that dimension to be something like one to ten mils.
I think i
14 they measured 59 mils.
And that would cause this to raise up 15 (indicating), so the paddle could jam it.
16 All the reactor trip breakers were pulled, then.
I 17 think there were six of them.
They set the failed one aside i
18 and used a spare in its place, checked them all, ran a whole 19 series of checks on them, made a few adjustments, reinstalled 20 them, and they all functioned properly, i
21 This says "the failed breakers to be evaluated by 22 the B&W Owners Group."
The B&W Owners Group had a meeting,
,~
23 and the decision was that the SMUD could easily do it at their 24 shop, as well as the Owners Group show.
So SMUD now is 25 running a series of tests, trying to decide when it was that a
--.,,-,.nn.-.
,n -,,,,, -.,,,
3 e,
1 m
t; t e 9
i came out.
dim ns, ions i
the 2
T, sis breaker was refurbished at GE, again checked at 3
B&W, When it turned up at the plant, it was out of
.s.
4 specification.
So they're trying to establish when that 5.
occurred.
> ~
6
- )
An I&E Information Notice is also in preparation on x
7 this item.3 e
MR. WYI,I E :
Has any similar occurrence occurred in s
o 9
any of the other_ Regions?
10 MR. MINER:
Not that I know of.
I think this is one 11 of the first ones of the refurbished trip breakers that failed
^
12 complete 1y, This failed completely.
It wasn't a matter of s_,
i 13 response time,being off.
This was a problem of actually 14 failing completely.
15 MR.
EBERSOLE:
Can we have a momentary interruption q
16 here?
17
.-[ Discussion off the record.]
18
[ Slide]
F 19 MR. MINER:
In order to measure this little thing i
20 here, you have to go in, get your finger on here, push this i
21 down and get a feeler gauge in there.
1 22 MR. EBERSOLE:
Could you describe the functional s
a 23 movement that thing executes when it does its thing?
l 24 ML. MINER:
Thistfults down here.
This whole thing
\\
25 pul'Is down, hits that paddle wheel I think I have another
10 N"
i one that shows the whole breaker.
2 MR. EBERSOLE:
Just when it's energized.
3 MR. MINER:
Yes.
When it's energized, it folds down 4
pulls away.
When it's deenergized, she pulls up, hits the 5
paddle, rotates up, hits the paddle, and that rotates.
6 MR. EBERSOLE:
That rotating trip shaft is what does 7
the tripping, trips the main device; right?
8 MR. MINER:
Yes.
9 MR. EBERSOLE:
And that was stuck.
10 MR. MINER:
This may not be the exact thing, but I 11 think it's similar.
Py the way, the shunt trip works.
This n/
N_
12 is the undervoltage trip.
13 MR. EBERSOLE:
That's about as close an 14 approximation,to a Rube Goldberg picture as I've seen yet.
15
[ Laughter) 16 MR. MINER:
That's right.
This is a better 17 one. This is the actual cross-section of the undervoltage 18 shaft.
19 MR. MOELLER:
GE routinely refurbishes these?
20 MR. MINER:
This is mainly the problem with 21 lubrication, hangup response times, so they sent us to GE.
I 22 think I have a little note.
They mainly changed the 23 lubricant, worked on the bearings, that type of thing.
We are
)
As part of it, they were supposed to make some 24 supposed to 25 of these key measurements.
A t
i 11
/
1 MR. EBERSOLE:
Had it passed operational tests out l
2 of the GE plant?
l 3
MR. MINER:
Operational tests were mainly conducted 4
at B&W.
GE, I think they must have run a torque.
Here is As they come out of maintenance, here are the things 5
some 6
that they were supposed to do.
7
[ Slide]
8 The operational tests, I think the verification 9
pickup, some of the torque testing that was done at B&W, GE 10 took care of this shaft bearings.
I think operational 11 verification was done at B&W.
The last two were what the i
~%
f (( )
12 Licensee was conducting.
They did the examination and 13 cleaning of the breaker enclosure, and this is when it 14 failed.
Running the functional test, it failed.
15 The other ones here were done at GE and B&W.
16 MR. EBERSOLE:
Since the shunt trip will obscure the 17 fact that this trip didn't work in a normal scram, how long 18 would it be this condition would exist in the plants before it 19 would discover that you weren't getting the trip?
20 MR. MINER:
Probably where you catch it is 21 surveillance, just as we did here.
Running the undervoltage 22 trip.
23 MR. EBERSOLE:
How often do they do that?
I'm just 24 talking about good, average number.
Months?
25 MR. MINER:
No, I don't think so.
12 s
1 Does anybody know how often they run these trips, 2
the surveillances?
3 MR. JORDAN:
It's a monthly surveillance, yes.
4 MR. MINER:
I think maintenance takes it out and 5
works on it every six months.
6 MR. REED:
I don't think the monthly will answer 7
Jesse's question.
You are talking about removing the breaker 8
and putting it on a stand.
9 MR. MINER:
This is not removed, just done in place.
10 MR. EBERSOLE:
Just in the operation of the UV trip.
11 MR. MINER:
I think that is done in place.
I think 12 they have a way when they do the verification they can check 13 both.
14 MR. REED:
They check individually?
15 MR. MINER:
Yes.
I know this one was installed, 16 this one, when they found it, was installed in the cabiret.
17 They were running this' functional test in the cabinet.
It was 18 installed in operating position when they first checked UV.
19 When that failed, they went back and checked the shunt trip on 20 the same breaker to make sure it functioned, and it 21 functioned.
22 MR. HERNAN:
I'm Ron Hernan with NRR Staff.
ACRS l
23 has a newly-formed Subcommittee on Scram System Reliability.
24 Their first meeting is scheduled for next week, the 18th.
It 25 occurred to me yesterday during the drive around that this may
13 1
be an item that that subcommittee may want to discuss as 2
well.
I don't know if Mr. Miner will be available next week 3
for the meeting, but we would propose to have somebody that 4
can speak to this.
I think it is a subject that that 5
subcommittee should delve into a little deeper.
6 MR. EBERSOLE:
Thank you for that suggestion.
I 7
agree with you.
8 Any other comments?
9 MR. WYLIE:
I believe with the B&W trip system that 10 some breakers are going to trip by shunt trip originally.
11 They have done some modification on those, and some also use 12 the undervoltage trip as backup, as I recall.
I'm not sure 13 which one this breaker was.
14 MR. MINER:
This one here would go both ways.
The 15 last one, which happened on June 23rd, was a leak, an 16 unisolatable leak.
t 17
[ Slide) 18 In the high point vent on the B steam generator 19 hotleg, these vents were put in as part of the TMI 20 modifications during the last refueling outage in 1983.
21 MR. MOELLER:
And this was to bleed off air?
22 MR. MINER:
I think this is part of the system to 23 restore natural circulation.
24 MR. MOELLER:
Right.
Okay.
25 MR. MINER:
Here is your 30-inch pipe.
14 (3
1
[ Slide]
2 And coming out of that is a nozzle, and they had I guess it's sort of a spacer for a 3
welded this little 4
lining, the pipe going to the high point vent, and they 5
installed this key on top of that spaoer.
The leak occurred 6
in this weld here, above 120 degrees around the periphery and 7
throughwalls, a 20 gpm leak.
8 MR. EBERSOLE:
What is the size of the pipe?
9 MR. MINER:
One inch.
10 MR. EBERSOLE:
Had that attachment been subjected to 11 the usual stress analyses?
12 MR. MINER:
Yes, it is.
It was analyzed.
The 13 problem that they ran into, the fact that some of the support something 14 sections supposedly there were not there, and so 15 I don't have as part of your package but I got in the mail 16 this morning may help explain it a little bit.
17
[ Slide]
18 We are talking about coming out right here, off on 19 this nozzle here.
What we are talking about here is the one 20 that hooks on here.
What they had and they added on is this 21 system right here, which is the two you saw.
You can push a 22 button to vent off.
What was there originally was this whole n
23 system here, which had a nitrogen connection coming in.
They 24 had a spool piece here, so during shutdown when they wanted to 25 be able to purge the vessel to keep it moisture free, they
15 O
1 pulled that.
2 During the design, somebody designed in a rigid 3
piece. It never was put in.
So that sort of stiffened this 4
side.
They also had a channel with some bands coming around 5
these two pipes to stiffen this arrangement.
That was not put 6
in.
This horizontal support to this -- I guess this is sort 7
of a channel -- also was never put in.
This was never put 8
in. So these were missing.
They were designed into the system 9
but were missing, were not installed.
It looked like a QA 10 problem.
11 MR. EBERSOLE:
It doesn't look like it might be just
(
12 localized to the single point.
It might be all over the 13 place.
14 MR. MINER:
Except you get something nonsupported.
15 You get that type of moti6n 16 MR. EBERSOLE:
I understand.
It just happened to 17 focus at that point.
But having had the crack occur there, 18 why should I believe there isn't potential for cracks all over 19 that. rat's nest of piping.
20 MR. MINER:
They have gone back and done a stress 21 analysis on the thing without the supports.
22 MR. EBERSOLE:
You mean all that pipe network, they 23 have gone back and looked at all of it?
24 MR. MINER:
They have done more than that, i
25 MR. EBERSOLE:
I'm trying to look for the procedural i
J
16 1
error.
2 MR. MINER:
Two things occurred.
Number one, 3
although this was on the stress analyst's drawing, it never 4
turned up on the construction drawings.
5 MR. EBERSOLE:
Was there a method that should have 6
-been used as a check-off that said it should have beent 7
MR. MINER:
Apparently some foul-up.
8 MR. EBERSOLE:
I'm looking for a larger narrative 9
than just the fact that they fixed this.
10 MR. MINER:
They have gone back and have looked at 11 all' supports modified since about 1980, in 1979.
They were D(,)
12 required to go and look at certain supports.
They went back 13 at that time, walked down and' looked at something like 350 14 supports modified since that time.
Now expanding that to 15 include some number of items that they didn't do as part of 16 the ISE walkdown, they have something like another about 175 17 pipes that they are looking at.
They are in that phase right 18 now evaluating.
19 They did some stress analysis on this and also the A 20 side.
21 MR. EBERSOLE:
It looked to me just from the picture 22 that you had the basis for that nasty classical uncontrolled 23 small LOCA.
24 MR. MINER:
No question about it.
As a matter of 25 fact, that's what they had.
17 3
\\~
1 MR. EBERSOLE:
They almost had a real one if a pipe 2
had broken off. It makes a messy leak.
3 MR. MINER:
One-inch pipe you can take care of --
4 MR. EBERSOLE:
Yes, but it makes a mess in the house 5
and creates havoc.
6 MR. MINER:
It was a 20 gpm leak that went until 7
they depressurized the thing.
I don't remember how many 8
gallons they leaked, but it was a substantial number.
9 MR. REED:
Jesse, I can't think of a better place 10 for B&W designed systems to leak.
If I'm going to have one, 11 that's a beautiful place to have it, although it should be O.~
12 bigger, like 4 inch.
Anyway, you didn't say what the l
13 metallurgical examination showed, whether it was fatigue 14 failure or whether it was oxygen stress corrosion cracking 15 from the inside.
16 MR. MINER:
Fatigue failure.
17 MR. REED:
In cher words, was it an outside-in 18 crack?
19 MR. MINER:
That I didn't ask.
I really don't 20 k n o w '.
But they cut that piece out and sent it off to GE for 21 metallurgical examination.
Word came out to me it was 22 fatigue failure.
- [~'\\
23 MR. REED:
I want to make a point here that in a 24 vent like this, a high point vent like this in the normal 25 sequence of events for filling and venting a PWR system can
i 18 fsU 1
get concentration of oxygen in the pipe, depending on how well 2
you vent, how long you vent. whether you vent up to 250 3
degrees or not, so I'm concerned whether it was an oxygen 4
crack initiated from the inside or whether it was fatigue.
5 MR. MINER:
Remember, this thing was installed at 6
that particular place a year ago, 1983, at the fueling 7
outage.
It may be, at the most, two years.
It wasn't there 8
very long. Stainless steel.
9 MR. EBERSOLE:
It's a fatigue crack, isn't it?
10 MR. MINER:
That's what the licensee indicated.
11 MR. EBERSOLE:
It's due to the absence of supports 1
7s l
'\\,
12 which were on drawings but never put in physically.
13 MR. MINER:
This was not on the drawing. The rest 14 were on the drawings and not put in physically.
15 MR. EBERSOLE:
So the root of the problem is where 16 is the guy who said they were put in physically when they 17 weren't?
l l
18 MR. MINER:
Yes.
Who is the guy who said they were 19 there?
20 MR. EBERSOLE:
Have they found him yet?
21 MR. MINER:
Well, you know, that's one of the 22 reasons I won't be here.
They are having a big management f
23 meeting tomorrow afternoon to discuss this whole problem.
N 24 MR. EBERSOLE:
So this is at the early stage itself.
25 MR. MINER:
It's a question of how much more they l
19 r3U 1
do, how much more examination; why wasn't this caught?
They 2
have found a number of others not as significant as 3
these. They don't have anything that they think could cause 4
failure, but they have found some deficiencies.
5 MR. EBERSOLE:
I think I would like to mark this one 6
down for the full subcommittee.
7 MR. REED:
I think so, too.
8 Another thing is, after this was installed and you 9
go into your warmup and operation, it should have been sort of 10 looked at and walked for vibration.
Did anyone observe any 11 vibration on this system?
In order to get fatigue crack like
/~x
(%,)
12 this, you have got to have some vibration, a lot of it, during 13 the year.
14 MR. MINER:
I was told you could go up and push this 15 section of pipe here and it would wave back and forth.
16 MR. REED:
But it was sitting in the operating mode 17 without being pushed.
Was it vibrating?
18 MR. MINER:
They are inside the containment, which 19 is not accessible during operation, so the fact that they 20 didn't see the missing supports would indicate nobody really 21 looked at it after they installed it.
I hadn't asked those 22 questions specifically.
f~%
23 MR. REED:
Isn't the containment accessible during 24 hot standby when you are coming up and shaking it down?
25 MR. MINER:
Oh, sure.
You could even get in there
20 pV) t 1
for limited times during operation.
2 MR. REED:
No one saw it shaking?
3 MR. MINER:
Nobody saw it shaking.
It's sitting up 4
in the top of a steam generator.
5
[ Slide]
6 You are right up in this area here.
I suppose there 7
are platforms to get up to it.
It's not something you would 8
walk by normally and look at.
9 MR. REED:
They must have access walkways to the 10 thing.
11 MR. MINER:
I'm sure, but again, I have to look at s
\\,,
12 it.
Probably through some kind of a platform.
That isn't 13 something you would be walking by looking at.
You know, the 14 fact that they missed those supports would indicate to me that 15 nobody took a good look at it after it was installed.
16 MR. MOELLER:
Is a 20-gallon per minute leak of this 17 nature enough to increase the temperature within containment?
18 This was just spilling ultimately onto the sump?
19 MR. MINER:
I don't think so.
I think they had 20 something like 100 degree weather outside anyway, so I don't l
21 think they saw anything --
22 MR. EBERSOLE:
It appears it was borderline to 23 actually breaking off.
Had it broken off, what would be the 24 difference in the treatment of the accident?
25 MR. MINER:
They would probably have had to rush out
21 1
there and got a second makeup pump.
2 MR. EBERSOLE:
I'm talking about in the treatment of 3
the post-accident analysis.
4 MR. MINER:
I don't think it would be anything.
We 5
are looking at this 6
MR. EBERSOLE:
As if it were broken off?
7 MR. MINER:
We are looking at it like an 8
uncontrolled leak in a primary system boundary.
9 MR. EBERSOLE:
As though it had been a trip 10 break-off?
11 MR. MINER:
Right.
12 MR. WARD:
Maybe I might have missed somathing.
I 13 apologize.
But did you tell us or could you tell us a little 14 bit more about the event?
What gave them the first indication 15 and how did they respond and how long did it take to get the 16 thing put to bed?
17 MR. MINER:
How long did it take them?
Well, they 18 had to pull the reactor down.
They were coming up.
They were 19 in hot standby.
They had been about 15 percent power and they 20 had to shut down because they had a problem with lubrication 21 on theik main turbine.
They were coming up, getting ready to 22 come up.
They were in hot standby.
But if I remember 23 rightly, the first indications were loss of level in the 24 pressurizer as well as some increase in level in the sump.
25 MR. WARD:
They did get both of those?
22
'1 MR. MINER:
My understanding was -- as a matter of 2
fact, they were able to verify the 20 gpm by measurements in 3
the sump, so they immediately notified us and started cooling 4
down, coming to cold shutdown.
That was the only way you can 5
isolate the thing and stop it, is to come to cold shutdown.
6 So I don't remember how many hours they cooled the plant 7
down.
They went into an ordinary cooldown.
8 F~
REED:
In answer to the question about how many 9
gallons came out, this preliminary notice says 16,000 gallons.
10 MR. MINER:
Yes, I know it was substantial.
(
11 MR. EBERSOLE:
Is that it?
12 MR. MOELLER:
Again, 16,000 gallons.
Quickly 13 dividing that means it went on for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />?
Or much longer?
14 MR. REED:
Longer.
15 MR. JORDAN:
The next presentation, I would like to i
l 16 skip to Items 2 and 3,
and the sad story there is, Dave Powell i
17 is the Operations Officer, who, I think, two of you met with 18 briefly this morning in the review of the Operations Center.
l 19 He's been on shift since midnight, and so we'd like to get him 20 h o m e ', so he can go to bed.
21 So Dave Powell will give the presentation on Oyster 22 Creek and on the Westinghouse Part 21, the consideration of l
23 potential seismic interaction.
l s
24
[ Slide.]
25 MR. MOELLER:
To help us on the schedule, we are
1 going by the agenda listing or by the handout listing?
2 MR. JORDAN:
By the handout listing.
3 MR. MOELLER:
Which ones?
4 MR. JORDAN:
These are No.s 2 and 3 on the handout 5
listing.
6 MR. POWELL:
The first event I'll be discussing.this 7
morning i s the uncontrolled leakage primary coolant outside 8
the primary containment at Oyster Creek that occurred on June 9
12th.
This particular event is very similar to an event that to occurred at Hatch in August of
'82.
11 At that time, I believe there was another event that
-s (j
12 occurred at Browns Ferry that precipitated these particular 13 valves right here (indicating), these additional series valves 14 being placed in -- at all the various BWR plants.
15
[ Slide.]
15 This particular drawing here is not of Oyster 17 Creek's specific design.
It's just there to add a little 18 picture there to show you what valves we're looking at and the 19 scram outlet valve.
20 Of course, this particular event is of concern 21 because it represents an uncontrolled leakage of primary 22 coolant outside of containment that went on for some 38
[\\,)T 23 minutes.
It states in the handout it was 36; actually it was 1
24 38 minutes.
25 MR. EBERSOLE:
Would you clarify one thing?
Did T
wv v
,-yy
,y----ww-w-y--
-,w
- - - - ~ ~
24 (d
4 1
this plant have the two valves on it?
2 MR. POWELL:
Yes, they did.
These valves had just
]
3 been put in, in fact, in 1984.
4 MR. EBERSOLE:
What is the probablistic number for 5
the occurrence of some leaking of these valves?
6 MR. POWELL:
I don't have that number.
7 MR. EBERSOLE:
We have a case of two redundant 8
valves failing to close.
I think we need to dredge up that 9
number and look at it against the PRA studies.
10 MR. POWELL:
Okay.
11 MR. EBERSOLE:
By the way, this is similar to the O
12 Hatch event some months ago when they had the same sort of 13 leakage.
14 MR. POWELL:
August of '82 is where they had 15 almost, from the standpoint of the leakage, it's almost 16 identical.
17 MR. EBERSOLE:
I think it's fair to say that we 18 should take this one to the full committee.
But go ahead.
19 MR. POWELL:
A brief synopsis of the event that 20 occurred:
Reaction trip occurred on loss of condenser 21 vacuum.
Loss of condenser vacuum caused an MSIV closure, and 22 evidently at the time they got the MSIV closure and the 23 scram, these particular valves right here (indicating) did not 24 fully shut.
25 I'll explain.
They both have different failure
25 O
1 modes, and I'll explain how they failed.
2 The original precipitating event of the scram was 3
the failure of the electric pressure regulator on the turbine 4
control circuitry.
The operator took manual control of the 5
system, but evidently it was too slow to do that.
The thing 6
tripped out on low pressure.
7 MR. EBERSOLE:
May I ask you about that?
Isn't it a 8
fact of life that any trip would have called for the same 9
event, irrespective of how it got started?
10 MR. POWELL:
That's correct.
11 MR. EBERSOLE:
So you cannot say this triggering 12 event is any other than any normal turbine trip.
13 MR. POWELL:
What I mean is, what gave them the 14 scram originally.
15 MR. EBERSOLE:
I know that.
But that scram could 16 have come by a manual scram.
The frequency of the initiating 17 event would have been very high indeed.
18 MR. POWELL:
F. rom the standpoint of the initiating 19 event, that's correct.
20 In fact, Oyster Creek has had a scram prior to this 21 one, but for some reason we didn't see those particular 22 events.
m 23 MR. MOELLER:
When you lose condenser vacuum, does 24 the pump just quit pumping?
What is the reason for that?
25 MR. POWELL:
What pump are we referring to?
l l
26 the really 1
MR. MOELLER:
You said you lost 2
initiating event was the loss of condenser vacuum.
3 MR. POWELL:
Let me back up.
The problem was the 4
electric pressure regulator problem, which caused the bypass 5
valve to open.
That bypass valve opening caused a' reduced 6
pressure in the reactor.
7' Once they reached the loss-pressure setpoint, the 8
MSIVs go shut, giving them a low condenser vacuum also, and 9
also the scram at that point on MSIV closure.
So that's the 10 series of events that occurred there.
11 MR. MOELLER:
Thank you.
)
12 MR. POWELL:
When that occurred, the operator at the I~
13 plant did have indication that this valve did not go fully 14 closed.
This particular valve, though, he did have l
15 indications of being shut (indicating.)
l 16 Ten minutes into the event, they received high l
I 17 reactor pressure scram signals.
The reason that occurred was 18 the main feedpumps were still functioning at the time, filling 19 up the vessel.
As it filled up, it pressurized the vessel.
20 They eventually did trip off the main feedpumps.
The feed reg i
21 valves did go shut, but they had quite a bit of leakage, and j
i 22 it just allowed the reactor vessel level to increase.
23 When that occurred, they were unable, at that point
[ '}
V 24 in time, to use their isolation condensers due to a problem i
25 with waterhammer at that high level
i 27 O
1 So what the operators did, in accordance with their 2
procedure, was to activate the alpha and delta electromatic 4
3 relief valves.
At the same time, they tried to bring on their 1
4 reactor water cleanup system for letdown purposes to try to 5
reduce level 6
MR. EBERSOLE:
Did they invoke what we now call 4
7 semiautomatic blowdown, then?
Is this right?
8 14R. POWELL:
Basically, yes, sir; that's what they 9
did.
10 MR. EBERSOLE:
That's not considered to be a very 11 nice thing to do.
I s
12 MR. POWELL:
They really had no choice, I don't l
13 think.
14 MR. EBERSOLE:
That's what Hatch didn't do, and they 15 got into worse trouble.
So here's an interesting parallel to 16 draw, or rather a difference.
17 What that did was enable them to shut down the rate 18 of leakage out of these valves by depressurising the primary.
19 Did that work?
20 MR. POWELL:
They never really got it depressurized 21 that much.
22 MR. EBERSOLE:
How did they ever shut the thing off, i
23 then?
24 MR. POWELL:
That all ties up with their ability to i
i 25 reset the scram system, and I'll get into that in a second.
28
-s 1
MR. EBERSOLE:
As I recall, Hatch couldn't do it, 2
because they had a high building pressure.
3 MR. POWELL:
They had a high drywell pressure, yes, 4
4 sir, that precluded them from resetting the scram because of 5
the scram signal was there.
It's a similar event, but a
'6 different scram signal was present.
7 Twenty-six minutes into the event, they received J
8 fire alarms in the reactor building at the 23-foot level, 9
which indicated that they had initiation of their fire deluge 10 system.
At that time, had they not known before that that 11 they had leakage, which I think they must have known, they J
12 would have known then, because basically that was caused by 13 the coolant flowing down into the reactor drain tank, coming 14 back up through the vents, flashed and mixed with the 15 blistering paint off the scram to start volume piping.
16 Those two things combined initiated the deluge 17 system in the reactor building, which is another problem.
At 18 any plant that initiates,that, they have potential problems 19 with equipment malfunctions.
20 MR. EBERSOLE:
Can you tell me why paint is being 21 blistered under what I thought were conditions that were the 22 design basis for the paint?
23 MR. POWELL:
Okay.
I can't answer that.
My 24 understanding of this system is that it's not considered a 25 portion of the pressure boundary per se, especially downstream
29 s
1 of the scram discharge isolation valves, the drain valves 2
there.
This would all be probably low-pressure piping.
I 3
don't know if they would be required to meet the 4
specifications.
5 MR. EBERSOLE:
What about post-LOCA fluid 6
temperatures?
7 MR. POWELL:
I don't know.
That's something we have 8
to look into, to verify whether or not they meet the 9
standards, and I haven't done that.
10 MR. EBERSOLE:
Go ahead.
That's just a peripheral 11 m a t t'e r.
12 MR. POWELL:
They also received, 35 minutes into the 13 event, a low reactor water level scram.
The9 had been 14 oscillating back in here trying to get the w:ter level 15 situation under control The reason that occurred, they had 16 continuous flow down through the reactor water coolant system, 17 and they had brought on isolation condensers, started the 18 cooldown.
All those factors -- the cooldown, the letdown 19 system in operation --
20 MR. EBERSOLE:
This is still being developed in 21 higher detail, analysis of this accident?
22 MR. POWELL:
I'm not sure.
23 MR. EBERSOLE:
It sounds like they're still looking U
24 into some higher details.
25 MR. POWELL:
You mean as far as the way
30 1
MR. EBERSOLE:
Yes, what they did, what happened, 2
and so forth.
3 MR. POWELL:
I have an actual document which they 4
have made up, which is the sequence of events that occurred.
5 MR. EBERSOLE:
Is that part of the handout?
6 MR. POWELL:
No, it's not, but I can get a copy.
7 MR. EBERSOLE:
I wish you would.
There will be lots 8
of questions about this in the full committee.
9 MR. POWELL:
Okay.
I'll have that available.
10 MR. HOLLAND:
Excuse me, Mr. Chairman.
My name is 11 Drew Holland.
I'm the Oyster Creek Licensing Manager.
We are O)
\\,_
12 here and available to answer any questions that you might have 13 to further clarify this.
14 MR. EBERSOLE:
I imagine the greater detail going 15 into here would be consistent with our short-term meeting 16 here.
What I rather expect we would need is a higher level of 17 detail at the full committee meeting.
18 MR. HOLLAND:
I see.
So the scenario that we have, 19 although it corresponds very closely with Mr. Powell's, is 20 somewhat different.
We have a few minor differences we might 21 want to get into.
22 Would you recommend that we come back for the full 23 committee meeting?
34 MR. EBERSOLE:
I think so.
25 One thing, for instance, if you invoke use of the
31 g-U 1
SAR system to blow down, you know, the primary release, what 2
did you do?
Cycle it?
And how low did you get the pressure, 3
and how low did you get the water, and did you cycle it back 4
and forth?
5 MR. HOLLAND:
I have the Director of Plant 1
6 Operations here with me, and I'd like to turn this over to 7
him.
8 MR. EBERSOLE:
We might have a few words on it, but 9
we'll run out of time real quick.
10 MR. SULLIVAN:
My name is John Sullivan from Oyster f
11 Creek.
12 The scenario is basically how Mr. Powell related I
13 it.
We isolated the reactor main steamline isolation valve 14 closure.
With the reactor isolated, the pressure built up.
15 We reached a high-pressure condition.
The A&D electromatics 16 automatically initiated reduced pressure, and it popped twice l
17 automatically.
18 The operator tripped two feedpumps initially on a 19 scram.
He was a little bit late tripping a third feedpump.
20 MR. EBERSOLE:
Are these feedpumpa electric pumps?
21 MR. SULLIVAN:
Yes, sir.
22 MR. EBERSOLE:
Main electric?
[
23 MR. SULLIVAN:
Yes, sir.
24 MR. EBERSOLE:
You would have been worse off if they 25 had been turbine driven?
32
,f-1 MR. SULLIVAN:
Yes.
We wouldn't have had the 2
steam.
3 He tripped the third pump.
His level had gotten a 4
little bit high, about 200 inches on top of the reactor fuel.
5 Above 181 inches, we're concerned about possible waterhammer 6
to isolation condensors, so we put it in the manual mode.
7 Realizing he needed pressure control, he was into symptom --
8 more into procedures.
9 He then manually initiated the electromatic to bring 10 pressure down.
Meanwhile, we had our cleanup system out of 11 service.
We were getting that back in service to afford 12 letdown.
We got it back in, we got letdown, we secured the 13 electromatic valve, went on the isolation condenser, and 14 continued the cooldown.
15 We couldn't reset the scram until it got below 600 16 pounds pressure.
Thereby, we were leaking by those two 17 valves that Mr. Powell referenced.
One valve had indicated 18
" fully shut," but it, in essence, was slightly open, even 19 though the indication was " closed."
The other valve gave us a 20 double indication, and we weren't sure what the position was.
21 And that water from the scram outlet valve went into 22 scram dump volume, into our reactor building equipment drain 23 tank, forcing steam vapor up through the hub drains on 24 Elevation 23 in the reactor building.
The hub drains, steam 25 from the hub drains, coupled with the fumes from the i.
33 g
1 blistering paint, caused activation of our deluge system on 2
the 51-foot elevation, the next elevation up, which set off 3
the fire system.
4 MR. EBERSOLE:
I see.
5 MR. SULLIVAN:
Once we got below 600, we reset the 6
scram and were able to close all the scram outlet valves, the 7
the leak was terminated.
8 MR. EBERSOLE:
The parallelism with the Hatch event 9
is interesting, except you had a few features on your side 10 here.
11 MR. MOELLER:
I didn't follow.
The blistering paint 12 caused the deluge system to cut on?
Give that again?
(
13 MR. SULLIVAN:
The blistering paint -- we're still 14 looking at the paint.
The paint was actually upstream of the 15 valves in an area that woold be the coolant boundary.
So the l
16 specs for the paint, we're still researching to find out why I
l 17 it blistered.
I i
l 18 The fumes got picked up on a product of combustion 19 detector.
l 20 MR. MOELLER:
Okay.
So they thought a fire was --
l 21 MR. SULLIVAN:
We had gotten a report of the l
l l
22 blistering paint, because there were operators in the area.
f 23 They had seen the steam coming out of the hub drains.
They l
24 had seen the paint blistering.
So we dispatched the fire l
25 people, but we were more sure that that's what caused it,
34
- fN 1
because we had that event happen before.
2 MR. MOELLER:
Thank you.
That helped.
3 MR. EBERCOLE:
Any other questions here?
4
[No response.]
5 MR. EBERSOLE:
Thank you.
6 MR. SULLIVAN:
You're welcome.
7 MR. POWELL:
Okay.
So that basically was the event, 8
and the problem, as the gentleman referred to, the reason they
.9 couldn't reset the scram is that they had the scram signal 10 input in there until they reached the 600-pound interlock 11 setpoint.
At that point, the operator promptly reset the 12 scram.
13 I mention the other three scram signals that were 14 input into the system, just to show you, even had they been 15 able to reset the scram, they would have been reinitiating 16 scram signals which basically would have opened this valve 1.7 (indicating) and required these two valves -- actually these 18 four valves (indicating) to go to shut.
19 So at various periods through the event, they could j
20 possibly have had the same situation, but in this case, it was 21 just for the duration of the event basically.
22 The safety significance of the event is threefold.
23.
One, of course, is -- I use LOCA h2re, but in this case it's 24 uncontrolled leakage outside of containment.
25 Second is potential equipment malfunction due to
35 gg
\\
l 1
actuation of the fire deluge system.
2 Also the excessive CED seal temperatures that 3
occurred in the event.
Those occurred primarily because they 4
had this extended flow through the CRD system.
It's my 5
understanding they had intermittent high-temperature alarms on 6
the CRD system.
'7 MR. EBERSOLE:
Let me ask, on the 600-pound,the 8
critical pressure at which you can reset the system, the 9
scrammed up volume level detectors of the focus of 10 vulnerability in all these designs, because they are the only 11 things that tell you the dump volume is clear for a second 12 scram.
13 MR. POWELL:
That's correct.
14 MR. EBERSOLE:
Irrespective of the pressure, if 15 those things say the scrammed volume is empty, are you all 16 right, or do you need 600 pounds or lower?
17 MR. POWELL:
For resetting the scram?
No, sir.
You 18 cannot reset the scram, because the scram signal that's in 19 there is part of the RPS logic.
20 MR. EBERSOLE:
So you need not merely the fact 21 that the. level switches say they're empty, the dump volume is 22 empty, but also that they're lower than 600 pounds?
[}
23 MR. POWELL:
That's correct in this particular v
24 event.
25 MR. EBERSOLE:
So there is a sort of subtle l
4 36 OO 1
additional requirement on scrammed up volume capacity, which 2
is a pressure requirement.
3 MR. POWELL:
You mean in terms of dumping the 4
volume?
i 5
MR. EBERSOLE:
Yes.
How many devices measure that 6
pressure?
One, two, 'hree?
7 MR. POWEl.L :
I can't answer that.
I'd imagine 8
probably two.
9 MR. EBERSOLE:
But you just told me is, it's 10 overriding a requirement of pressure to get a reset.
11 MR. POWELL:
We're talking here system pressure for 12 interlock of MSIV closure.
13 MR. EBERSOLE:
I'm talking about the pressure inside 14 the scram dump volume.
15 MR. POWELL:
That's a different interlock, and it's 16 there for a different purpose.
The two don't have any 17 relation in terms of --
18 MR. EBERSOLE:
At least I'm inclined to look on the 19 level switches as being the only meaningful signal which says 20 you can scram or not scram the dump volume.
We all know if 21 the pressure in there is too high, you can't scram either.
22 MR. POWELL:
The point is, on the scram discharge
[~\\
23 volume, you have override.
If you go down to shutdown mode, 24 you can override the scram signal on the dump volume for 25 the scram discharge volume.
You cannot override the scram
s 37 1
signals for the MSIV closure interlock.
2 So normally what happens is, at BWRs is, as soon as 3
all their scram signals clear, the only way you can clear the 4
scram signal for the dump volume is to go down to shutdown or 5
refueling with the mode switch, then hit the reset button.
6 That temporarily overrides that, allows these valves to open 7
up.
This valve (indicating) will already be shut and allows 8
the volume to drain down, indicates that they do, in fact, 9
need to rescram at a later time.
10 The 600-pound interlock at Oyster Creek, as far as I 11 know, is rather plant-specific.
Most BWRs are able to 12 override that interlock at 1050.
13 The actual cause of the problem -- I put three 14 causes here -- the two direct causes and also one that's an 15 apparent cause.
16 First off, the Valtak valve, which is these three l
17 valves here, I understand were all Valtak valves, had an 18 improperly adjusted stroking distance for the valve.
There 19 was about one-eighth of an inch that the valve lacked from the disc lacked from fully seating onto the 20 fully seating 21 seat.
22 The Velan valve here, which was the downstream 23 valve, had an improperly sized spring in the actuator.
I
(
24 believe these are air to open -- they're either spring assist 25 to close, or air to close spring assist, or just strictly l
38 3
_j i
spring to close.
I'm not sure on that.
The size spring they 2
had in there was for 400 pounds.
And as part of the 3
corrective actions, they changed that 400-pound spring to an 4
1100-pound spring.
5 It's my understanding from resident people at the 6
site that this valve probably went shut, as it was required to 7
do, and due to this, the upstream valve right here 8
(indicating) not seating tightly, they got system pressure 9
developed up underneath the disk of this valve, and it forced 10 the valve open, and that's probably why they got the double 11 indication.
It was at some intermediate position.
O-(,,)
12 So they had a one-eighth inch leak right here 13
'(indicating), and this valve was in some intermediate 14 position.
15 The underlying cause, though, of all this is that some sort of modification 16 these valves were part of a post 17 that was done in 1984, and I presume it was due to the Hatch 18 event and similar events like that.
19 These valves are all stroke-tested in accordance 20 with the IST program.
The stroke test is to make sure these 21 valves function as required on the scram.
22 They do that periodically, either monthly or 23 quarterly, whatever.
It depends on whether or not they have a
)
24 relief set up, that sort of thing.
25 When I first got this, I had a little bit of work in
1 IST.
I thought these valves were also being leak-rate 1
1 2
tested.
Apparently, as far as the Staff is concerned, the
)
t 3
pressure boundary for this whole system ends at the CRD 4
seals.
So instead of categorizing these valves as Category A 5
valves, which were required on a leak-rate test as far as 6
post-installation and periodically on a refueling basis, they 7
are categorized probably as Category B valves, which only 8
require stroke testing.
9 So after post-installation, apparently they did a 10 hydro on the system with the valves full open just to check l
11 the integrity of the piping, never really did a leak-rate test
(')/.
(_
12 on the valves.
Had they done that, this particular problem 13 probably would have shown up during that test.
14 I don't know.
It depends on how long this 15 particular valve was In that position.
I don't know when they 16 17 MR. EBEREOLE:
You mean as far as the Staff is 18 concerned, the endpoint of the primary system boundary is the 19 seal on the individual rod?
20 MR. POWELL:
That's my understanding.
21 MR. EBERSOLE:
There are 185 rods on some plants.
22 MR. POWELL:
That's my understanding.
I understand
/
\\
23 there's a document that I haven't been able to get yet -- I'm h.
24 t rl'i ng to get from AEOD -- that verifies that.
At the very 25 least, I would have thought at least the scram outle' valve --
40 b
1 MR. EBERSOLE:
There's been a huge controversy about 2
potential of breakage of this dump volume in any aspect, not 3
to mention just the valve failure, but in terms of 4
metallurgical failure of the pipes or the main header.
Of 5
course it is denied that such a leak can occur with numbers 6
like 10 to the -15 or something.
Yet they are routinely 7
occurring at Hatch here and here at this plant.
8 MR. POWELL:
It's my understanding, at the meeting 9
we gave last week, personnel within NRR and, I guess, AEOD are 10 getting together to reevaluate their concerns on this.
11 MR. EBERSOLE:
I'm quite certain the full committee g_
12 would like to hear this in its lurid details.
l 13 MR. POWELL:
NRC follow-up.
From our standpoint, we t
l 14 have an I&E notice detailing the event sequence and potential 15 problems involved with it.
And that's it for this l
16 presentation unless there are other questions.
17 MR. EBERSOLE:
Any further questions on this one?
18
[No response.]
19 MR. EBERSOLE:
We are running just about on time.
it's really not an 20 MR. POWELL:
The next event 21 operational-type event per se.
It involves the potential 22 seismic interaction of the in-core flux mapping system of
[ )\\
23 primarily Westinghouse plants with the seal table.
That's the w
24 guids to event seal table.
25 Before we get going on discussion, let me point out
41 1
a few things here so everybody is on the same base.
2
[ SLIDES]
3 What we are talking about, here is the seal table.
4 Everything below this point is seismic Category I.
It's also 5
safety grade.
For some plants, these isolation valves may 6
also be part of the seismic Category I boundary.
Some may 7
not.
It depends on our tech engineer.
But any point beyond 8
this, it's all considered non safety-related equipment and
'9 it's all -- well, as we found out, at least for most plants 10 that have not been properly seismically supported, normally 11 there are six 10-path transfer devices, six bi-path rotary i
12 devices, and also six drive units.
13 For most plants at Westinghouse -- I have a picture 14 here -- only the 10-path' devices are normally located above 15 the seal table, and they would pose the greatest problem.
a 16
[ Elide]
l 17 However, as we were notified by Westinghouse, not l
the same. It 18 all plants have designed '. h e i r seal table syst 19 depends on the architect engineer.
So scme plants may not 20 have any of these above their seal table, and some plants may 21 have this entire drive system above the seal table.
Those, we 22 expect, are few, and most of the plants probably have this 23 sort of structure.
l l
24 As you see, the bi-path and the drive units are off 25 to the side.
These are all seismic boundaries here.
So they l
i I
~ -.
42
~g 1
would not interact with the seal table down here.
Just these 2
10-path devices, which are on rails.
The rails are there so 3
they can remove these 10-path devices, get them out at the top 4
so they can withdraw the guide tubes for refueling purposes, 5
so they have the carts and the rollers there and everything 6
that's on there.
7 That's part of the problem with the seismic 8
qualification of that system.
9
[ Slide]
to To give you an idea why this is of such concern, 11 this system, as you see, although this is just a very general 12 decriptive picture of it, the seal table is somewhere above 13 the outlet nozzle for the vessel, but the actual leak rate 14 path is from the bottom of the vessel, ll-l 15 Now, most Chapter 15 analyses only consider a 16 breakage of one of these guide tubes at the seal table, much 17 like the Sequoyah event that happened last year, and that is 18 supposedly bounded in their Chapter 15 analysis.
But 19 according to recent events and everything, I guess the 20 possibility is that this thing can swing loose during a 21 seismic event, come on down and impact on several of the guide 22 tubes, thereby creating possibly breaking more than one of 23 these guide tubes.
)
v' 24 If that were to occur and they were to complete 2S sever at the compression fitting, then you would have leakage
43
\\
i out of the bottom of the vessel, probably within the 2
capability of the high pressure safety. injection pumps at 3
Westinghouse plants.
If you go any more than four tubes, 4'
though, that becomes hazy.
You don't know how much leakage 5
rate you are going to get there.
6 At Sequoyah they had an average leakage rate of 30 7
gallons per minute with one tube completely severed to totally 8
eject it out, so if these were just to break and remain in 9
place, you-would probably see leak rates around there, more 10 around a 20 gallon per minute leak rate around each guide 11 tube.
But in the event that were to occur, what would happen 12 is the water would be coming out at a rate that may not -- you 13 may not be able to keep up with it by using the high pressure 14 safety injection pumps.
But due to the action of the 15 pressurizer, you would sit 11 maintain the system pressurized.
16 It would be kind of a feed and bleed, but the 17 pressure of the system would be continuously being maintained 18 at whatever the operating pressure is.
Maybe a slight 19 decrease, but the potential is for completely draining this 20 area down right here or the pressurizer possibly forming a 21 bubble up here in the top and still not be able to use your 22 intermediate head pumps yet.
23 That kind of a scenario has to be looked at from a
\\
24 thermal hydraulic standpoint.
That's just something that some 25 people at NRR came up with, possible problem with a leak, or a
44
(
1 LOCA where you have it coming out the bottom of the vessel 2
instead of up here on this piping somewhere, where once you 3
get down to this point, you are basically steaming and not 4
draining the vessel.
5 So that's a big problem.
It hasn't been analysed, 6'
as far as I know.
I'm not sure where that particular subject 7
is at the Staff right now. I understand it is being looked 8
at.
Anyway, we originally found out about the problem from 9
potential Part 21 that we received from Shearon Harris.
I 10 will skip all the prelinaries here.
We got the actual Part 21 it on February 12th.
That's when it was dated by the licensee.
(
12 We actually got it about a month later for review and that 13 sort of thing.
14 Once we got it, personnel in I&E contacted 15 Westinghouse to try to get some feedback as far as the generic 16 implications of the prchlem.
Westinghouse finally notified 17 the licensees at the insistence of the NRR Staff.
First it 18 was just a phone conversation with the various licensees, and 19 a week later they put out a written document indicating that 20 this was indeed an unreviewed safety question for operating 21 Westinghouse reactor plants.
22 At this point, Westinghouse only reviewed the 23 problem with the interaction.
That is -- well, we had it up 24 there before.
But that is what would happen if the thing 25 swings down and hits the guide tubes.
They didn't do any kind
45 O
1 of analysis on what would happen if more than one guide tube 2
were to break as a result of that parti:ular event.
3 The reactor regulatory group or Westinghouse Owners 4
Group came back and gave a brief scenario stating that that 5
particular event was a low probability event and there was not 6
enough energy in the 10-path devices if they were to drop down 7
to do damage to more than three of the guide tubes, and if 8
they were, they would sever the guide tubes at above the the isolation valve.
They would not impact per 9
discharge 10 se on the seismic Category I portion of the seal table.
11 They also said, of course, if more than those
(~
t
(,/
12 occurred, they were outside the bounding analysis of the 13 Chapter 15.
14 MR. EBERSOLE:
Does-it seem to you that the essence 15 of all this is we find that in these guide tubes a breaching 16 of the rationale but there is no point of egress for liquid 17 coolant that amounts to anything below the incoming low 18 pressure lines?
In fact, what did you say, about four of them 19 means you are in trouble?
20 MR. POWELL:
Four of them, if you were to completely 21 sever the guide tube, you would be in excess of your high 22 pressure charging pump capacity.
23 MR. EBERSOLE:
And in order to get to low pressure,
)
24 there is no real blowdown capability other than through the i
25 secondary system and that is not safety grade, so you are just
-,-e+
~
c,.,
.,e m--,
--.,.y7,_
_,w,,.
yw.-
___.r_w-
-.-n-
46 ON-i leaking water out faster than you can put it in from the low 2
point in the vessel.
3 MR. POWELL:
That's correct.
4 MR. EBERSOLE:
Any comments on this, Glenn?
5 MR. REED:
First of all, I have been hearing about 6
this and I'm a little shocked at the spaghetti that makes up 7
the track above the seal table.
It's got enough mass to do a 8
lot of damage were it to come loose, or it could come loose 9
from the seismic event.
10 MR. EBERSOLE:
That raises another question.
Other 11 multiple failures of these that might be generated by other
)
12 phenomena.
13 MR. REED:
Yes, I suppose so.
14 MR. EBERSOLE:
.Like one failing and shearing the 15 res t -by jet action.
16 MR. POWELL:
That was brought up when the event 17 occurred at Sequoyah where they totally injected out the 18 vented tube.
It depends, basically, I would imagine, on the 19 valves here and how stable and how sturdily-that portion of 20 the pipe from this point on down to the seal table --
21 compression fittings like this are great if you have got 22 tensile loads on them, but as soon as you start putting 23 lateral loads on, if they have any defects on how you put them
\\~
24 together, those things are going to break.
We saw it five 25 times last year.
They were all catastrophic-type events
Ic 47 i\\
1 because they all eventually culminated in leakage out of the 2
reactor.
3 MR. EBERSOLE:
Something about other things.
Are 4
they someplace vulnerable where a truck or something, a 5
conveyance, could run into them, or other events sever them?
6 MR. REED:
Generally they are in a tunnel by 7
themselves except for air flow.
Sometimes you have air ducts 8
in there but there is no pressure in the ducts.
Sometimes 9
they actually make up a tunnel where the duct is at or near 10 the seal table top, too.
They are generally well restrained 11 and separated.
There is no truck access, in fact no people Ok,/
12 access.
It is normally a non-access area beth during 13 refueling and during normal operation.
14 MR. EEERSOLE:
This would show at first a 15 pressurizer level decrease.
Then due to the new 16 instrumentation we got on these things, subsequently it would 17 show a loss of level of inventory, I guess, in the primary 18 vessel.
You know, loss of subcooling and then a subsequent 19 loss of level 20 MR. POWELL:
The first thing they would pick it up 21 on would be the radiation monitors.
22 MR. EBERSOLE:
I'm talking about degrading core 23 cooling.
What would the operator do?
24 MR. POWELL:
The operator can fix the problem, I first of all, by turning off the heaters 25 think, by starting
48
.'Q 1
in the pressurizer and opening up the PORV so they could bring 2
in intermediate head safety injection pumps.
3 MR. EBERSOLE:
He will try to depressurize as best 4
he could.
5 MR. POWELL:
Yes, sir.
pre-.sure in all 6
MR. EBERSOLE:
By lowering 7
directions, secondary side as well, I guess.
8 MR. REED:
Jesse, what you are saying is he would 9
like to have PORVs.
It's a nice comment.
10 MR. EBERSOLE:
I'm thinking about Palo Verde and 11 wondering if this got this configuration.
O 12 MR. BETA:
Paul Beta, ACRS Staff.
We have got 13 Bryan scheduled to come in on the 31st on the ECCS 14 Subcommittee meeting. He is going to discuss this from the 15 thermal hydraulics standpoint about how many tubes he can 16 rupture.
17 MR. EBERSOLE:
You will get the whole picture.
18 We will have to move on with this one ualess there 19 are other questions.
I would like to get a reaction or at 20 least for you to note down whether you think the full 21 committee should hear about this one or not in the context of 22 this generic undercooling.
[V
\\
23 MR. REED:
I wonder what the subcommittee -- maybe 24 that would take care of it.
Maybe it's a subcommittee issue.
25 MR. EBERSOLE:
You mean Paul's meeting?
49 O
1 MR. WARD:
Why don't we let the ECCS Subcommittee 2
MR. EBERSOLE:
We will let the ECCS pick it up from 3
there so we can pull away from this.
4 MR. HERNAN:
Jesse, could we go back to the Oyster 5
Creek for a section?
Mr. Powell mentioned he thought NRR had 6
some follow-up commitment to look into this in response to 7
that event.
I talked to our Operating Reactor Assessment 8
Branch folks.
The action was generic in terms of the scram 9
volume discharge, and the implementation was to install 10 redundant valves, which Oyster Creek had already done.
Beyond 11 t h a t',
and beyond the information notice that's going out, we
)
12 presently don't have any other new and different 13 investigations as a result of the recent Oyster Creek event.
14 MR. EBERSOLE:
Thank you.
15 MR. JORDAN:
The next item is back on the normal 16 agenda. The first item of the handout.
George Rivenbark, 17 Hatch 1's stuck open fuel RV.
18 MR. RIVENEARK:
I'm George Rivenbark.
I'm the Hatch i
19 project manager from NRC.
20
[ Slide]
21 On the evening of May 15th, while the Hatch Plant 22 Unit 1 was at full power, a crane was passing overhead in the 23 control room, bumped into a line that connects to the deluge 24 system, the fire deluge system, and the charcoal filters in 25 the air cond'.tioning and the air recirculation system in the
50 p)
\\'/
1 control room.
When it did this, it caused the deluge system 2
to actuate.
The water flooded the charcoal filter.
There 3
were plugged up drains in the base of the filter box, so the 4
water didn't follow its normal path to the drain system.
5 Instead, it leaked through the air conditioning ventilation 6
ducting, and some it dripped into the control room and onto 7
the top of an instrument cabinet.
8
[Slidel 9
The system that started this thing looked about like 10 this.
This is, of course, a rough sketch of the way the 11 charcoal filter plenum looks.
The air -- the charcoal filters 12 were in this plenum in a room directly above the control 13 room.
The water pipes are in here.
I don't have them 14 sketched on here.
But because some drains in the bottom of 15 this thing were plugged up, the water level filled up to the 16 first point that it had any freedom to move out with, and that 17 happened to be into a return air duct from the control 18 room. This is the pipe t h e.t came down and went pretty much 19 directly into the control room.
20 MR. EBERSOLE:
I doubt that that evolution is in any 21 PRA study now existing.
22 MR. RIVENBARK:
This is titled " Systems Interaction 23 Event."
It dripped into the top of an analog transmitter trip 24 system panel.
25 MR. EBERSOLE:
I can't really believe this. The
51 1
worst accident so far I thought about was totally on the 2
floor.
3
[ Laughter]
4 MR. RIVENBARK:
Dripped onto the top of this panel 5
and through some little holes in the top of the panel and went 6
down the cabling that went through the panel vertically.
As a 7
result of this water dripping in there, it caused the SRV to 8
open.
Now, the SRV opened several times and eventually 9
remained open.
I 10 The people at the plant scrammed the plant according 11 to procedures whenever the SRV remained open.
Meanwhile, they 12 had already gone to look for what was going on, how to stop 13 the water.
They put some plastic over 14 MR. EBERSOLE:
-They had known it was water now?
15 MR. RIVENBARK:
The first thing they saw before the l
16 SRV went off, they saw that there was water dripping.
17 MR. EBERSOLE:
Into the control room?
l l
18
'MR.
RIVENBARK:
Into the control room.
They put f
19 plastic up while they were running around looking to try to 20 find the' place to cut off the valves which were in the room 21 above the control room.
The water, meanwhile, dripped into had already dripped into this panel, apparently, and 22 this
'[ /)
23 was beginning to take its action, and the SRV opened, as I r
l
\\-
24 said, opened several times and then stuck open.
i 25 They tried to close the SRV.
They went through
l 52
/%
's 1
procedures to try to get it closed.
They could not get it i
2 closed.
They later learned that the procedure was not l
3 correct.
The fuses were not the correct fuses that they were 4
trying to pull. For about 30-some odd minutes, this thing was 5
open, and it closed by itself eventually without their having 6
taken an action that would cause it to close.
7 MR. REED:
What kind of relief valve was this?
8 Electromagnetic?
Air operated or 9
MR. RIVENBARK:
It's a regular -- it's a two-stage 10 target rock relief valve that is operated electromagnetically.
11 MR. REED:
So it's a heat terminal pilot operated 12 electromagnetic heat valve.
13 MR. EBERSOLE:
The primary pressure blew on down.
14 MR. RIVENBARK:
The primary pressure continued to 15 blow down.
During the process, in order to try to maintain 16 some of the pressure in the liquid, they lost liquid level 17 whenever they did this.
The feedwater pumps came on 18 automatically to regain --
- 9 MR. EBERSOLE:
Which feedwater pumps?
The main 20 feedwater pumps, are they steam driven?
If so, what steam 21 pressure did they have?
Weren't the main isolation valves 22 closed?
23 MR. RIVENBARK:
I'm sorry, I don't know.
(
24 MR. EBERSOLE:
Did the main steam isolation valves 25 close?
53 (x
1 MR. RIVENBARK:
Yes.
1 2
MR. EBERSOLE:
They couldn't have been the main --
l 3
MR. RIVENBARK:
They closed those intentionally.
4 MR. EBERSOLE:
I guess as most plants are, they have 1
5 got turbine-driven main steam pumps.
So you have got HPCI and 6
DRPSI and the low pressure pumps.
How low did the pressure 7
get in the primary system?
8 MR. RIVENBARK:
It went down in the neighborhood of i
9 250, 300 pounds.
That was lost level 10 MR. EBERSOLE:
So they kept inventory where it 11 should be.
l (,f 12 MR. RIVENBARK:
They kept inventory up.
There was a 13 never a signal that caused the HPCI to operate.
14 MR. EBERSOLE:
That's because the leak didn't get 15 over to those contacts.
I
~
16 Any questions about this curious thing?
17 MR. REED:
You said a crane initiated this all?
18 MR. RIVENBARK:
Yes.
19 CSlide3 20 I don't know how the crane system is set up in 4
21 detail, but this is a faded drawing of the plant.
The control 22 room is here. The turbine room for one unit, Unit 2,
is 23 here. The turbine building for the other unit is here.
They 24 have the ability to move a crane, not to carry loads on it but 25 to move a crane from one end of this plant to the other.
-]
I 54
-~
G 1
And apparently they surmised -- and I must point 2
out that they haven't said for sure they know this is exactly 3
what happened, but it is their best analysis of what might 4
have happened that the crane was passing over and was dragging 5
a hook, and that the hook bumped into a line that controls or 6
maintains this valve closed.
It's a line that has water under 7
pressure.
8 When the hook bumped onto that line, it broke a 9
l i t t l e-nipple on the pressure gauge at the end of it.
It 10 didn't break it off, it cracked it, and that caused the 11 pressure to bleed off, caused the deluge valve to open.
12 MR. REED:
So this ventilation system is not housed 13 in a block housing above but is unhoused so there are pipes 14 and things sticking out?
15 M2. RIVENBARK:
That's right.
This is up in the top 16 of an open aree.
17 MR. REEL' :
I'm a little surprised that it's not in a 18 room of some type with a ceiling.
19 MR. RIVENBARK-I don't know if there is a ceiling 20 above the crane or not.
21 MR. REED:
You mean below the crane?
22 MR. RIVENBAPK:
There's not a ceiling below the
(/)
23 crane, obviously.
- m 24 MR. MOELLER:
The Air System Subcommittee looked at 25 this a week or two ago, and there were several items that I i
55 s
1 would like to comment on.
One is that this system, the HVAC 2
system for the control room, the charcoal and so forth, 3
follows Regulatory Guide 1.52.
About a year or so ago, the 4
committee wrote a letter requesting that Reg Guide 1.52 be 5
updated, and the Air System Subcommittee pointed out or 6
concluded at their meeting a couple of weeks ago that Reg 7
Guide 1.52 is being inappropriately applied to the design of 8
the HVAC systems for the control room.
9 Reg Guide 1.52 applies to the containment building 10 and was written for that, but the utilities and the Staff had 11 gradually moved it over and applied it to the control room 12 HVAC system.
13 The point I am making is most~ people conclude that a 14 control room HVAC system ~should not have a deluge system for 15 the charcoal, so had it been properly designed, without the 4
16 water there, this never would have happened.
17 MR. EBERSOLE:
Dade, doesn't it get into the boiler 18 category?
Can the HVAC systems, in fact, become sources of 19 liquid?
20 MR. MOELLER:
Certainly.
I 21 MR. EBERSOLE:
If not through this route, then 22 through chillers or whatever.
In fact, it seems even through
)
23 this clear route here that they can become liquid sources.
24 MR. MOELLER:
Yes, but I want to stress once again l
-25
.that the committee wrote a letter saying Reg Guide 1.52 needs l
l t
f i
56 1
to be updated and corrected and so forth, and nothing 2
happened, or something happened.
3 MR. EBERSOLE:
So there is potential for 4
water-to-air interchange, not merely in this weird mode here 5
but in other modes, and the control room panel boards are not 6
protected from overhead.
7 MR. MOELLER:
The subcommittee also pointed out that 8
the control room charcoal systems have preheaters on them, and 9
they said they don't need preheaters, that the preheaters 10 could be a source of problems.
11 MR. EBERSOLE:
Sure.
(,/-
12 MR. HERNAN:
Dr. Moeller, we have an effort that is 13 nearing culmination on control room habitability.
Isn't this 14 Reg Guide part of our implementation plan?
15 MR. MOELLER:
Yes.
The Staff agrees and has now 16 called for revision of 1.52.
I want to just keep supporting 17 it in any way possible.
The Staff is doing a very good job on 18 this.
The Control Room Review Group and so forth has done a 19 superb piece of work.
l 20 MR. HERNAN:
So you are saying we have responded 21 to your request; it's just very slow.
22 MR. MOELLER:
Yes.
I guess I was disappointed that I would have to look up 23 our letter of a year ago, roughly j
24 the date.
It didn't get any action.
But yes, they are doing 25 a good job.
~
~
57 O
I s/
1 MR. EBERSOLE:
Any further questions on this one, 2
and give your thought as to whether this should be brought in 3
front of the Full Committee.
4 HR. MOELLER:
We are going to be doing it when we do 5
the control room habitability letter, which I think is next 6
month, so we could probably do it through that avenue.
7 MR. EBERSOLE:
Then we will get it to the Full 8
Committee through that route, then.
9 MR. MOELLER:
Right.
10 MR. EBERSOLE:
Let's have a ten-minute break and 11 come back at about 2:30.
(
12
[ Recess.3 13 MR. JORDAN:
Can I make a correction?
On the Hatch 14 stuck-open safety relief valve, it was stated to be solenoid 15 operated or electrically operated.
It is electrically 16 operated but it is an air solenoid at the valve itself. It's a two-stage target rock with an air solenoid at 17 three-stage 18 the valve itself.
It doesn't change anything else that you 19 know about.
20 This is Vern Hodge, who is going to give the 21 presentation on Pilgrim and the water hammer and the HPCI 22 steam turbine exhaust line.
Item 4 on your list.
~]
23 MR. HODGE:
Thank you, Ed.
J 24 I am going to talk to you this afternoon about a 25 coupla of events in recent months at Pilgrim in the HPCI steam
58 1-1
. turbine exhaust line.
2
[ Slide]
3 I have shown here a simple diagram indicating the 4'
physical layout, systematic layout of this part of the 5
system.
Steam is exhausted from the turbine through a check 6
valve and stop valve into the torus pool The turbine is 7
about 12 feet below the level of the torus.
and there 8
The safety significance of these events 4
9 were about seven of them over a three-month period -- is not 10 large, but because it is involved in the safety system and did 11 cause some damage to pipe support systems, we are discussing
~
12 them.
i 13 During this time, the events were all occasioned by i
14 surveillance testing and the system was declared inoperable a 15 number of times, but it could have performed its function if 16 called for.
However, it was not called for.
17 MR. EBERSOLE:
May I ask about that rupture disk 18 down there at the bottom. That normally is in the same room as 19 the.HPCI turbine, isn't it?
20 MR. HODGE:
Same room.
21 MR. EBERSOLE:
Should it rupture because of the 22 plugging of the downstream line, it will rupture into the roco 23 proper?
24 MR. 1(ODGE :
That's correct.
25 MR. EBERSOLE:
That would be fatal to anybody in the
1 room, I believe.
Is that correct?
-2 MR. HODGE:
I don't know about that, but there are i
3 actually two of them.
The in-board one, if it gets ruptured, 4
will emit steam --
5 MR. EBERSOLE:
That's on the grounds you have a 6
gradual degradation of the rupture disk and not on the grounds 7
that you have a true stoppage; is that correct?
8 MR. HODGE:
Yes.
9 MR. EBERSOLE:
If you have a true stoppage, they 1
10 will both rupture because that's what they are supposed to do.
11 MR. HODGE:
That's correct.
s 12 MR. EBERSOLE:
This is an old issue that came up 13 here a few years ago.
I have wondered if there were 14 appropriate personnel warnings if you go in that room, on 15 automatic recall, you are told by appropriate signs that you 16 will be dead if that happens?
Is that the case?
17 MR. HODGE:
I don't know that.
18 MR. EBERSOLE:
I don't know what the probability of 19 rupture disk failure is, but it's not zero.
20 MR. HODGE:
That's right.
In fact, in these events 21 that happen 22 MR. EBERSOLE:
That brings out the point.
Carry on.
23
[ Slide]
24 MR. HODGE:
The sequence started March 31st with an i
25 overspeed trip of the turbine, which the licensee found, a
60
\\-
1 broken connector between the governor valve control and 2
actuator a s s en.b l i e s.
They repaired that and resumed the 3
testing.
On April 2 they did a test and received a high 4
pressure alarm.
After the test, they went to look at the 5
system and foun3 one snubber damaged, in-board rupture disk 6
broken.
7 So they had to repair that.
They modified the 8
procedures also to extend the time from two minutes to three 9
minutes when they ourged this exhaust line with nitrogen to 10 eliminate any water in the line, then decided to inspect all 11 the pipe hangers after each use of the system.
12 In a ddi t i ott, they decided to do a daily purge of 13 nitrogen, but no one nas really decided whether that 14 corrective action is effective.
15 MR. EBERSOLE:
Is the vertical profile of that line 16 such that on nitrogen ptrge it guarantees emergence of all the 17 water in the system?
Yott know, you can purge a line with 18 nitrogen and still have s.ugs of water in it.
19 MR. HCDGE:
Yes, I believe it is.
I don't know 20 exactly where they put the purge in.
21 MR. EBERSOLE:
Is it a self-draining line?
22 MR. HODGE:
There i ts a drain on the line, yes.
1 conf i turat ion I see up there bothers 23 MR. REED:
The t
24.
me a couple of ways.
First of all, I see a check valve in a 25 vertical line.
If it's that way, it can hold up water above
61 3
l 1
it.
Second of all, this is an exhaust discharge line which 2
will go hot and then will cool, and then if you got that line 3
stuck under water in the torus, you will suck water back into i
4 the line.
5 Now, is there a drilled hole at the top of that line 6
inside the torus to break the vacuum?
7 MR. HODGE:
The answer is there is no vacuum breaker 8
at that location at this time.
9 MR. REED:
Not just a drilled hole like a half-inch 10 hole up there?
It looks to me like you could just suck water 11 right back into the system.
l (j
12 MR. EBERSOLE:
By the way, is this a standard design 13 for this exhaust complex for all of these or is it different 14 from plant to plant?
15 MR. HODGE:
I believe it is fairly standard.
16 MR. EBERSOLE-So if you fix one of them, you could 17 fix all of them?
18 MR. HODGE:
Yes.
19 MR. EBERSOLE:
For instance, if they discharge and 20 the rupture disc could be sent somewhere other than the room l
21 environs themselves, they could be done universally.
You 22 know, that kind of so-called nuclear fatality would be marked 23 on as such did it occur.
That rupture disc discharge could go w
24 to some other place besides the room environs.
25 MR. HODGE:
That's correct.
(
. ~
62
.O 1
MR. EBERSOLE:
Go ahead.
I'm wondering at the
'~"
2 moment, though, whether the nitrogen purge does guarantee 3
scavenging of the water, and I don't see any particular reason 4
to believe that it does, certainly not with that profile.
5 MR. HODGE:
To respond to one of the comments, I do 6
have an isometric drawing of the system.
7
[ Slide]
8 The check valve is located in the horizontal run of 9
.the piping.
It is mostly horizontal, in fact, with a couple 10 of vertical changes, right there and right here.
11 MR. REED:
Show me where the torus is, the torus N,/
12 connection is.
Right there?
13 MR. HODGE:
Down here someplace.
14 MR. EBERSOLE:
The torus vent is in the center 15 MR. REED:
The check valve is on the same elevation 16 as the top of the torus piping?
17 MR. EBERSOLE:
Is that the point of torus entry?
18 MR. HODGE:
N o..
The torus entry is right here.
I 19 This is at elevation sero minus three, which is the elevation 20 of the check valve.
The turbine exhaust is minus 12.
21 MR. REED:
So there is still potential for the check 22 valve to hold water up in there.
23 MR. HODGE:
Yes.
N_
24 MR. EBERSOLE:
Is the nitrogen purge a volumetric 25 purge of quite high rates to blow the water out or just to
n 1
show you have got continuity of flow path?
~
2 MR. HODGE:
I'm sorry, I don't know those details.
3 MR. EBERSOLE:
They do different things.
One acts 4
as a carrier and the other just says maybe am I in your flow 5
path.
But go ahead.
6 CSlide]
7 After repairing the damage and continuing their test 8
on May 18th they received a trip of turbine.
Only a few 9
seconds elapsed and then it restarted.
This introduced a 10 severe transient to the system and there was some more 11 damage. Operators heard a loud noise.
On examination, they I
i 12 found two snubbers damaged, some concrate expansion anchors 13 displaced.
14 They repaired this damage and changed the procedures 15 to establish manual control for the turbine to start. This had 16 the effect of decreasing the severity of the transients.
17 Another more important modification was the bypass 18 of the internal pumps of the governor actuator assembly.
This 19 allowed pressure to be increased downstream of that pump and 20 reduces the magnitude of the spike of speed on the start.
21 Before this time, it had been accelerating to 5000 rpm, which 22 is close to the setpoint, and after the modification, only
(
23 3000 rpm.
24 MR. EBERSOLE:
Over what period of time do they come 25 up to this speed?
64 V
1 MR. HODGE:
I don't know that.
2 MR. JORDAN:
That's very quickly.
It accelerates 3
very rapidly.
4' MR. HODGE:
I would expect so because fast start 5
means you just open the stop valve, get the turbine to release 6
all the steam you have, and spike is 3000 rpm.
7 MR. EBERSOLE:
What is the real value of the super 8
fast start?
Do they have to have it?
9 MR. JORDAN:
In terms of --
10 MR. EBERSOLE:
The damper, I'm thinking, even after 11 it's running, doesn't do things very fast, so what is all the 12 gusto by getting going so quick with all this potential 13 damage?
14 MR. JORDAN:.
I really can't answer that in terms of 15 the urgency of seconds.
The acceleration is rapid because of 16 the energy available on fairly low inertia in the system.
2 17 MR. EBERSOLEI I'm talking about time to let valve 18 openings preclude all this stuff.
Does anybody know why we i
19 have got to get going so fast and why a gradual start wouldn't 20 virtually erase this water hammer problem?
21 CNo response.]
22 MR. HODGE:
The normal operation pump in the case of i
23 a LOCA might be addressed the same way, then rapidly restart.
24 So just simulating reality may be the reason for the test.
25 Now, the licensee did test a little bit differently before
65
+
/s 1
)
'\\
1 this year. This year is the first time it has had the quick 2
starts; hence, the water hammer.
3 MR. EBERSOLE:
So quick starts do appear to be the 4-root of the water hammer.
5 MR. HODGE:
Right.
I think so.
6 MR. REED:
I still don't like this layout.
I'm 7
trying to figure out whether the water hammer is coming from 8
water in the turbine or water in the pipe, and I just don't 9
like to see a turbine discharge uphill with its exhaust.
You 10 normally like to think it is going to discharge downhill so 11 you don't get water.
Water will run downhill, somebody says.
(O) 12 You are discharging uphill. That's bad in the fiast 13 place.
The second place is you haven't told me there is any 14 way to break the vacuum that will be created.
Once you make a 15 discharge with this unit, you will pull water back into it.
16 You should have a drilled hole inside the torus at the top of 17 that pipe, that vertical pipe, so you will break the vacuum.
18 MR. EBERSOLE:
I thought, in fact, it was such a 19 hole except it had a check in it.
Are you sure there isnt' 20 such'a hole with a check in it?
21 MR. HODGE:
I believe there is not such a hole.
The 22 vacuum breaker on the line is right here.
The licensee is 23 currently investigating whether to put a vacuum breaker right 24 here.
25 MR. REED:
You know, I have seen pressure relief
66 1
valve, safety valve discharge lines that go under water.
2 Whenever you do that, even though they are running downhill to 3
an underwater pool, you always drill a hole right inside the 4
tank of the torus in order to keep them from vacuum sucking 5
water right back up the pipe.
6 MR. EBERSOLE:
If it looks like that, we will always 7
have a shot of water in it.
8 MR. REED:
Maybe it's going to be in the turbine 9
because you are going uphill and there must be some trap drain 10 on that turbine, I hope.
11 MR. EBERSOLE:
Then if it's a fast start, that's 12 going to suggest a verf fast rise in the last stage turbine 13 pressure, and certainly challenge the rupture disc to some 14 degree.
I'm not sure but what there are not a bunch of men 15 standing around when they do this.
16 Carry on.
I'm interested 17 MR. HODGE:
It has operated successfully up to this 18 year.
19 MR. EBERSOLE:
How many years is that?
20 MR. HODGE:
The plant went commercial in 1972.
21 MR. EBERSOLE:
So it's time is just coming, then.
22 By the way, in all those years did it have a relatively slow 23 start compared to the one we have now?
24 MR. HODGE:
In testing, yes.
25 MR, EBERSOLE:
That's when it's needed most of the
67 1
time, certainly, isn't it?
Yes.
2 MR. HODGE:,
3 On May 23rd an event occurred that didn't have any 4
relation to the seals, and we think air was entrained into one i
5 of the sides of a DP flow switch, giving a biased high flow 6
output system, causing that high flow isolation there.
Not 7
understanding the water hammers, the licensee decided to put 8
on instrumentation package to gather more data, and they did 9
that on May 24th through June 23rd.
10 On J';ne 6th, they had another trip and found a part 11 of the instrumentation package was interfering with mechanical O(j) 12 linkage of the control valve, repaired that, made the system i
13 operable after that, tested successfully several times.
14 On June 30, they found that one of the rupture discs 15 had become inoperable, probably because of one of the previous 16 events.
17 MR. EBERSOLE:
What do you mean by an inoperable 18 rupture disc?
19 MR. HODGE:
As I understand, the pressure indication 20 at the point netween the two discs showed that this was not 21 holding pressure.
22 MR. EBERSOLE:
It indicates the first pressure N
23 membrane had failed.
24 MR. HODGE:
Yes.
25 MR. EBERSOLE:
So they were riding on their reserve.
68 O
1 MR. HODGE:
And if I didn't mention, I meant to, 2
that the rupture disc did flow at one of the first events, the 3
one on April 2nd.
4 The thinking is, the source of water is from the 5
torus pool Water is syphoned up here.
But extensive work 6
has been done with controlling speed of the turbine, operates 7'
more smoothly now, and the Licensee feels that the problem 8
essentially has been solved.
But at the same time, they're 9
still investigating whether to put a vacuum breaker at this 10 point.
11 These are interesting waterhammer events.
There 12 have been others at other plants, and we are including this s
13 one in an Information Notice in which we will also d e's c r i b e 14 those other events.
15 MR. EBERSOLE:
If that one-inch valve there is 16 closed between the relief valve and the pipe, inevitably you 17 will then fill up that hole quite full of water, won't you, 18 after each test run?
If you happen to have that inadvertently l
19 closed, you are now going to 20 MR. HODGE:
This check valve will close, given 21 enough time.
22 MR. EBERSOLE:
That check valve opens to prevent i
(
)
23 steam flow toward the torus.
24 MR. HODGE:
To permit steam flow.
25 ME. EBERSOLE:
Right.
But the water will have
w 4
O -
69 tAf it will leak back 1
filled the entire system, except I see no it will leak past the check valve.
It will be seated, if 2
4 3
it seats, into the open last-stage turbine chambers.
+
4 Is there a drain out of that?
f 5
MR. HODGE:
I don't believe so.
=
I 6
MR. EBERSOLE:
So it will drip on back t?nless it's I
7 well sealed, and fill up the turbine?
It looks like we've got i
8 a stage here for waterhammer.
9 MR. REED:
A comment was made that it worked all 10 right for ten years.
But I don't know the condition of those 4
1 11 valves the potential for leakage, the check valve potential 12 for leakage and so on and so forth.
Maybe it's just the wear I
i 13 of the leakage that puts the water back into the turbine.
I 14 don't know.
It seems to me somebody ought to look at it.
15 MR. HODGE:
The check valves have been replaced this I
16 year.
1 17 MR. REED:
Was it before or after the events?
18 MR. HODGE:
Before.
4 19 MR. REED:
Have you had any events since?
20 MR. EBERSOLE:
It was replaced before the event.
21 Maybe you put in a lousy valve.
22 MR. HODGE:
However, neither the Licensee, I guess, 23 nor the Resident Inspector feels that this check valve had
\\
24 anything to do with these events.
25 MR. REED:
I'm not so sure about that.
If you don't i
70 n
'/
1 have a drilled hole up there and you have the potential to get 2
water running back by vacuum, you might even syphon it right 3
over.
I'm not so sure.
I think somebody has got to look 4
carefully at this layout, who understands water and gravity 5
and vacuums and things.
6 MR. EBERSOLE:
When you look at this event at 7
Pilgrim, do you go back and look at the same class of things 8
at the other boilers that have this configuration to see 9
what's happened with them?
10 MR. HODGE:
Yes, we do.
11 MR. EBERSOLE:
Has anything happened?
.0
( /
12 MR. HODGE:
There have been several events in 13 boilers.
14 MR. EBERSOLE:
Is it at a point now where we need to 15 bore in and fix things?
16 MR. HODGE:
I believe not.
The damage mostly is 17 restricted to pipe hanger assemblies.
Piping failures have a 18 bit of it.
19 MR. EBERSOLE:
If you broke the pipe, there would be 20 a momentary discharge of steam, and then, I guess, you would 21 isolate the HPCI, turn it off; is that correct?
That would be 22 the end of the road.
Then you'd go into semiautomatic release 23 if you needed it?
\\
24 MR. HODGE:
Yes.
l l
25 MR. EBERSOLE:
Just a singular pump here.
So one is
- __ - _ _. _ _ _. _ _ _ _.... _ _ _ _ _, _. ~. _. _ _ _ _ _ _.._ _
l 71 1
depending on the main steam isolation valves to terminate this 2
aooident; is that correct?
3 MR. HODGE:
We have backup systems for the HPCI l
4 systems.
5 MR. EBERSOLE:
I don't mean that.
I'm talking about 6
the stop steam discharge into this area.
You're counting on 7
turnoff steam up forward.
8 MR. HODGE:
That's true, yes.
9 MR. JORDAN:
As far as the generic look at this, of 10 course there was the waterhammer, USI, and now i '. ' s two 11 plant-specifics.
12 You know generally the utilities are able to cope f
13 with those problems, and other than finding that we have 14 several plants that have the same failure mode and then 15 bringing it to their attention, we don't plan a generic 16 action.
We will communicate this to the utilities, but wo 17 believe it's a single plant issue, that they have a situation 18 of a change in valve operation or some feTture of their plant 19 that is different than it was up until this past year.
20 MR. EBERSOLE:
Ed, don't you think you could grant 21 them relief from freshstart on this thing?
22 MR. JORDAN:
If that's the problem, yes.
A n'd I T
23 think we are not able to tell you what clearly is the U[
24 problem.
Certainly I agree that a simple vacuum breaker that 1
25 prevents you from getting water back from the torus seems like l
72 0
1 the easy solution out of that.
So we are at a single plant 2
situation and would continue to attack it on a single plant 3
basis through the Resident, following the Licensee's actions.
4 MR. EBERSOLE:
As I recall, upstream of this, there the main steamline feed to this is kept normally open.
5 is 6
Both high-pressure valves are open, and there is steam 7
condensation in the lines for the turbine at the high-pressure 8
side, and a condensate drain has to be put on them.
9 MR. HODGE:
Yes.
A lot of events have occurred in 10 that part of the system.
11 MR. EBERSOLE:
However, that water condensate flow
/
(,
12 into the turbine proper is stopped by the stop valve, isn't I
13 it?
14 MR. HODGE:
Yes.
15 MR. EBERSOLE:
So there is a forward equivalent to 16 this problem of steam and waterhammer.
l 17 MR. HODGE:
That's right.
18 MR. EBERSOLE:
Thank you.
19 Any further questions, and do you think the full 20 committee would be interested in this?
l 21 MR. REED:
I don't think it needs to go to full 22 committee.
I think there's some detail here and good sense 23 with respect to piping, exhaust, and vacuum.
So Staff ought 24 to follow up on it a little bit hard to make sure the i
25 Licensees are looking at the proper thing.
I don't like the l
l
i e
73 f--
I
\\
1 sound of a vacuum breaker. 'You don't need a vacuum breaker.
2 That's too complicated.
I'm a mechanical engineer at heart.
3 MR. JORDAN:
A vacuum breaker, to me, is, like you, 4
a hole in the pipe surely.
5 MR. EBERSOLE:
I've been long concerned about that i
i 6
rupture disk and the potential for personal hazard in the 7
safety context.
4 i
8 How does that go over with the Staff?
They don't 9
take that into any particular account, do they?
4 10 MR. JORDAN:
No, sir.
I've been in the room with 11 the turbine starting, and I have seen the diaphragm.
I feel D
- k 12 uncomfortable in that position.
s, 13 MR. EBERSOLE:
I don't like to think with a few feet 4
14 of pipe that you couldn't get rid of that hazard potential.
15 MR. JORDAN:
You bring it to another room 16 MR. EBERSOLE:
If nothing else, no more than that 17 across the wall 18 MR. JORDAN:
Duly noted.
19 MR. EBERSOLE:
All right.
Thank you.
20 MR. JORDAN:
Our next event is regarding Calvert 21 Cliffs Units 1 and 2 diesel generators.
Dave Jaffe from NRR 22 will make the presentation.
23 MR. JAFFE:
Good afternoon.
My name is Dave Jaffe, i
24 I'm the NRR Project Manager for Calvert Cliffs Units 1 and 2.
25 During May of this year, an incident occurred at l
i
74 g-1 Calvert Cliffs that resulted in all three diesel generators 2
being declared inoperable.
3
[ Slide.]
4 On May 14th, Calvert Cliffs Unit-1 was in a 5
refueling outage, and Unit-2 was being operated at full 6
power.
On that date, the Unit-1 diesel generator, which is 7
referred to as Diesel Generator No.
11, was being tested for 8
overspeed, and at that time, a bar located in the rotor, which 9
we refer to as an interpolar connector, broke loose and caused 10 substantial damage to the stator.
11 On investigation, the Licensee found that thia O
(_,)
12 interpolar connector had initially become disassociated fcom 13 one of the poles, had broken loose, and had damaged the 14 stator, and following that, the second weld had come undone, 15 causing this piece to fall into the bottom of the generator.
16 This is a diagram that shows the orientation of the 17 bar.
There are a total of sixteen poles, each of which is 18 connected by one of these interpolar connectors.
These are 19 amortisseur windings located on each pole.
These generators 20 were part of a skid-mounted unit that was supplied by
~
21 Fairbanks Morse as original equipment for the Calvert Cliffs 22 units.
(.m) 23 There are three diesel generators.
Normally each 24 Calvert Cliffs unit has a dedicated diesel generator, and the 25 third unit is what is referred to as the swing diesel.
75 t\\#
1 Following this failure, the Licensee determined that 2
the generator would no longer be useful for continued service, 3
and they set about to replace that unit, using a spare that 4
was in stock at that time.
In addition, they had retrieved 5
this interpolar connector and set about to try and determine 6
the failure mode in order to assure themselves that the 7
remaining diesel generators were, in fact, capable of 8
performing their design function.
9 In performing those tasks, the Licensee obtained the 10 services of a consultant from the generator vendor.
On the 11 26th of May, they determined that, in fact, the cause of 12 failure had been high stress induced cracking.
13 CSlide.]
14 And they set about to inspect the Unit-2 generator.
15 That's the generator referred to as the No. 21 generator.
16 Those cracks did, in fact, exist in the same locations as had 17 been observed on the Unit-1 diesel generator.
18 At that time, they declared the Unit-2 generator 19 inoperable.
Shortly thereafter, the swing diesel was also 20 inspected, and similar cracks were found.
At about the same 21 time, work on replacing the Unit-1 generator was in an 22 advanced stage, and, in fact, the Licensee was able to return 23 the Unit-1 generator to service, s
24 Using a collection of internal breakers, they were 25 able to align the Unit-1 generator to Unit-2, such that in
1 the event that onsite AC power was needed, that a reliable f
2 source would exist.
That alignment, in fact, indicated the
)
3 need to immediately shut down Unit-2, since at that time 4
Unit-2 was still operating at full power.
5 Following consultation with the generator vendor, it 6
was determined that, in fact, these interpolar connectors were 7
not necessary and could be removed, and reliable operation of 8
the generator could be assured.
That modification was, in 9
fact, undertaken, and the Unit-2 and the swing diesel were 10 tested and returned to service.
11 MR. EBERSOLE:
What function was lost, then, in 12 declaring that these were nonessential?
They were put there i
13 for a purpose 14 MR. JAFFE:
Absolutely so.
15
[ Slide.]
16 The purpose of the interpolar connectors were 17 determined by the vendor to be twofold.
First, they are 18 important when dissimilar diesel generators are in use.
And 19 secondly, they're important when total three-phase unbalanced 20 load is excessive.
21 The generator vendor determined that that was not 22 the case for Calvert Cliffs, nor was it likely to be the case
[/
23 for any large generator in nuclear service.
l N_
24 MR. EBERSOLE:
This is kind of a general requirement i
l 25 for commercial DGs out on the linet i
i
77 O
1 MR. JAFFE:
That's exactly so.
2 MR. EBERSOLE:
It's a commercial complication which 3
is unnecessary at large.
4 MR. JAFFE:
Yes, sir, that's exactly the case.
And 5
at Calvert Cliffs, those interpolar connectors were 6
successfully removed, and the diesel generators were found to 7
perform fine.
8 MR. EBERSOLE:
Do they generally exist at all DGs?
9 MR. JAFFE:
That's not universally the case, no.
We 10 found out, by way of a Part 21 submittal from Colt submitted 11 on June 5th, that there were other facilities where they did 12 exist.
13 MR. EBERSOLE:
Is it a fact, even if you had an open 14 surrogate, there was no current in these things anyway?
15 MR. JAFFE:
It wouldn't have been any problem at 16 all.
17 MR. EBERSOLE:
Actually, there was no threat to 18 operability at any time?
19 MR. JAFFE:
Except for the fact that there was 20 substantial mechanical damage that occurred, so you could not 21 sustain the damage.
22 MR. EBERSOLE:
That was not due to any 23 electromagnetic load or anything?
24 MR. JAFFE:
Absolutely not.
It was found to be a 25 result of high stress induced cracking as a result of
78
(,
\\
1 centrifugal force.
2 MR. WARD:
I don't understand.
When dissimilar DGs 3
are used, what does that mean?
4 MR. JAFFE:
You could have reciprocal drivers of 5
different sizes in use in parallel, in which case the 6
interpolar connectors would dampen the voltage oscillations 7
which might otherwise occur.
8 MR. WARD:
It's another way to get an unbalance.
9 Okay, I understand.
10 MR. WYLIE:
It's a speed dampening.
In other words, 11 one machine has a little different characteristics than the 12 other, so therefore one tends to hunt for the other one.
13 The dampening har was put in there to dampen out the 14 oscillation.
15 MR. EBERSOLE:
Charlie, would these always be found 16 in these nuclear dos or not?
17 MR. WYLIE:
I think you'll find them in a lot of 18 them.
They all have dampening bars, I mean, in the pole I
19 faces, but a lot of them have the short circuit in the bands 20 to connect those, so that you complete the circuit.
l 21 MR. JAFFE:
That's exactly right.
They almost i
22 always have these amortisseur windings, but they don't always 23 have the interpolar connectors.
24 Now there were other nuclear installations that were 25 identified, and we are very interested to see what those
3 79
- (
1 Licensees were dong about it.
2
[ Slide.3 3
MR. EBERSOLE:
You mean with this particular 4
generator?
5 MR. JAFFE:
Yes, a similar generator.
6 MR. EBERSOLE:
This turns out to be really a 7
mechanical weakness, doesn't it?
A mechanical weakness in the 8
generator design?
9 MR. JAFFE:
Yes, I would say so.
10 Here is a list of similar installations.
Vermont 11 Yankee has two diesel generators of similar design.
After I
12 hearing from Colt, the Part 21 process, they removed them.
13 The same for TMI.
Peach Bottom has four of these units in 14 service.
They inspected the diesel generator that had the i
15 1argest number of service hours, and based on what they saw, 16 they removed the interpolar connectors.
They will inspect the 17 r em a i n i r.g diesel generator units in July.
18 Calvert Cliffs, as I mentioned, has three of these 19 units.
They removed the interpolar connectors on two units.
20 Now the new unit which was installed at Calvert Cliffs Unit-1 21 still has those interpolar connectors; however, those 22 connectors will be removed prior to restart of Calvert Cliffs
()
23 Unit-1, 24 MR. WYLIE:
Mr. Jaffe, all of those generators there 25 are the Louis Allis design?
t
80 A,
1 MR. JAFFE:
That's correct.
2 MR. WYLIE:
The reason you didn't find more that 3
have these connections is, 21 went out just on this machine, 4
right?
5 MR. JAFFE:
That is correct.
6 MR. WYLIE:
So there are probably a lot of them out 7
there that still have those, of other vendors.
8 MR. JAFFE:
That might very well be true.
I'm not 9
aware of exactly what that list would be.
10 MR. WYLIE:
I don't know either.
Of course, this 11 happens to be a deficiency of this particular vendor's design.
12 There are other things on those machines that, if 13 not designed properly, could do the same thing as far as 14 damage on the machines, like fans on the rotors.
The polar 15 pieces all have fans on them, for example.
Those are just 16 bent metal plates that are bolted to those polar pieces.
If 17 they're not designed properly, you could have the same problem 18 as far as fatigue and them falling off and damaging 19 MR. EBERSOLE:
What you're really relying on is 20 inconsistency of the failure.
21 MR. W\\ LIE:
You're relying on somebody doing the 22 proper job on design manufacture.
f 23 MR. EBERSOLE:
But if it's g o i n g '.t 'e fail, you don't 24 want more than one to fail at once.
25 Any questionst
~
81 1
[No response.]
2 I don't think the Full Committee would want to be 3
hearing about this.
4 MR. JORDAN:
The next event is Turkey Point, Units 3 5
and 4,
loss of offsite power due to swamp fire, I'll call 6
it. Henry Bailey from I&E will give a presentation.
7 MR. BAILEY:
Good afternoon.
I would like to 8
discuss loss of offsite power that occurred at Turkey Point 9
last May 17th of this year.
We don't attach any special 10 safety significance to this particular event other than the 11 challenge to the safety system that normally occurs with loss 12 of offsite power.
13 This event is presented at this time due to the 14 previous concerns there have been on loss of offsite power 15 event at Turkey Point.
16 At the time of the loss of offsite power, Unit 3 was 17 in a refueling mode for defueling, and Unit 4 was in 100 18 percent power.
19
[S11 del 20 This is a simplified diagram of the Florida Power 21 and Light grid.
Of course, this is Turke ' Point way down 22 here at the end.
I don't know how well you can see it.
You 23 can't see it at all here.
But there are seven 240 kv lines
,A 24 that leave Turkey Point to go north, and for some length of 25 time, all these 240 volt lines are in one corridor.
82
)
1 It turns out the failure was not here, but I wanted 2
to point that out.
The fire actually occurred up here near 3
these three 500 kv lines.
These dark dotted lines are 500 kV 4
lines.
The fire occurred near these three.
It turned out 5
that it shared out these three lines almost simultaneously.
6 The mechanism of shorting out these lines was either 7
hot ionised gas and/or soot collection on connectors from the 8
fire.
I understand this was actually a grass fire, but the 9
fire suddenly shifted due to a wind shift and burned under the 10 lines.
Combination of the hot ionised gases on the soot 11 caused these lines to short out.
12 Of course, when these lines shorted out, that caused 13 this entire southeast portion of the grid to isolate from the 14 rest of the Florida Power and Light grid.
This situation 15 resulted in an immediate collapse of the voltage in this 16 section, the reason being the load in the section at the time 17 was over 4000 megawatts; generating capacity in the section 18 was a little over 2000 megawatts.
So they were importing more 19 power through this grid than they were generating, almost more 20 than twice as much power being used down here as was 21 generated.
22 So they had an immediate voltage collapse in this
\\
25 section.
This voltage collapse, of course, caused a loss of
[J 24 offsite power at Turkey Point.
25 CSlide3
.. ~
83 1
This is a simplified one-line diagram of the onsite 2
power at Turkey Point.
These are Units 1 and 2,
the fossil 3
units over here. They are connected through these busses to 4
the same yard as the nuclear Units 3 and 4.
5 The loss of offsite power, what it did was you lost 6
the power to this 3-C bus here, and this 3-C bus, it turned 7
out, feeds to this line here over to the 4-C, 4160 bus, and 8
this 4160 bus feeds the 4-B main steam generator feedwater i
9 pumps. The loss of that pump initiated a turbine runback.
In 10 the middle of that runback, or actually early in that runback, 11 the reactor tripped on a steam flow / feed flow mismatch
\\s_/
12 coincident with the low steam generator level and the fossil 13 units over here were also tripped at the same time.
14 The response to the system was normal, as expected.
15 Everything worked as it was supposed to.
The diesel 16 generators, which are not shown here, came up and fed the 17 emergency busses.
Natural circulation was established.
Like there were some minor equipment 18 I said, I don't know of any I
19 failures, but nothing that affected the operation of the 20 plant.
l 21 The reports that we have gotten so far show voltage l
22 was actually established through the grid from the north to l
23 Turkey Point within 44 minutes, and then the fossil units were
=
s i
24 put on the !!ne.
They would supply unit power to the nuclear 25 units within about two hours.
l I
84 1
At about two hours, the reactor coolant pumps were 2
restarted and the plant was put in normal hot shutdown 3
conditions.
4 The !!oensee is studying ways to minimize or prevent 5
this type of event in the future.
One of the things he is 6
looking at is ways to better detect the loss of these high 7
voltage lines, and in the event they are lost, ways to shed i
8 the load so you don't get into a situation where your load is 9
twice what your generating capacity is and have a collapse of 10 the voltage.
11 They are working on a report which is not out yet, kN,)
12 and I anticipate that there will be more recommendations to 13 come out of this.
14 Are there any questions?
15 MR. EBERSOLE:
This was due to extensive brush fires 16 all over the country down there?
Was this what 17 MR. BAILEY:
That was the time I think there were 18 fires several places in the Everglades.
19 MR. EBERSOLE:
Was there any suggestion that these 20 hig fires and all this smoke and so forth could permeate the 21 diesel plants themselves and do anything funny theret I have 22 been looking at the California osse out there, and wondering 23 about whether there is brush around some of those i
24 MR. BATLEY:
In the report that I have heard, the 25 reports I have seen so far, I haven't seen anything that said
.. _. _, _. ~. _,.., _. - -.,, -
_m,~,.......,.. _ _ _ _. _..,. _ _ _ _. _ _,,,,,.., -..
85 1
it was a problem with the smoke or anything around the plant 4
2 itself, although that is certainly something to think about.
3 I don't really know how much grass there is right around the 4
immediate vicinity of Turkey Point.
5 MR. EBERSOLE:
The California fires make it l
6 spectacularly revealing how bad it can be.
7 MR. BAILEY:
That would certainly be something I i
8 would hope they would consider in the study they have going on 9
now.
10 MR. EBERSOLE:
So this really is about the same sort 11 of thing you might expect in a hurricane or something.
i 12 MR. BAILEY:
Pretty much loss of offsite power.
i 13 Everything worked the way it should.
They responded the way i
14 they should and apparently did a real good job on it.
The 15 only thing, if there is anything that maybe didn't go the way 16 it should, there was some indication that maybe there wasn't 17 good enough communication between the people that were 18 fighting the fire and the plant staff to know that those lines 19 were, in fact, in danger to the extent that they were.
If 20 they had known that they were, in fact, going to lose those 21 lines before they did, they could always shed some load so 22 they wouldn't have a complete collapse of the system.
23 MR. EBER80LE:
Right.
Any questions?
24 (No response.1 25 MR. EBER80LE:
Routine.
Thank you.
86
\\-
1 MR. JORDAN:
The next event is Sequoyah Unit 2,
2 reactor trip due to improper use of test instrument.
This 3
will look somewhat familiar to you, I am sure.
Eric Weiss 4
will make this presentation.
5
[ Slide]
6 MR. WEISS:
Good afternoon.
7 We are presenting this event because it demonstrates 6
not only a recurring pattern of a problem but also how, 9
despite safeguards such as communication between control room 10 personnel and test personnel out in the plant, you can still 11 get a reactor trip even when you are following an approved 12 procedure. As Ed said, I'm sure you will recognise this as 13 similar to one that we discussed with ACRS last time.
14 The reactor was at 100 percent power on May 22nd 15 when it got a reactor trip on overpower delta T.
The 16 instrument techniolans involved were performing a calormetrio, 17 an approved procedure that had recently been revised, and were 18 taking voltage measurements off of the resistance temperature 19 voltage elements in the reactor protection set cabinets 1 20 through 4.
21 The cabinets were physically separated by what I am 22 told is about 15 feet, and it was necessary for the instrument 23 techniolan to take all four readings within a relatively short 24 period of time, within about three minutes, in order that he 25 would have a snapshot of the plant condition for the
87
, - ~
k)
i 1
calormetric.
So that gave him about 45 seconds at each 2
cabinet.
3 He opened the cabinet door to the first cabinet and 4
connected the leads and took his reading.
Unfortunately, he 5
had a digital voltmeter that had two sets of connections, i
6 CBlide) 7 One for taking either voltage or resistance reading, 8
and another for taking current readings.
He as we are now probably had the leads in the current 9
able to reconstruct 10 jacks, so the internal impedance of the instrument was much I
11 lower than it should have been for taking these voltage 12 readings.
It, in effect, caused a trip of that channel on low 13 temperature and then proceeded to the next cabinet.
14 The people in the control room noticed the trip 15 condition, noticed the apparent reduction in temperature, and 16 there is a time constant associated with these devices, and 17 they are starting to climb back up, but before anyone in the 18 control room can get ahold of him and before the time constant 19 recovers, he has got the second cabinet door open and again 20 the leads connected, making up the two out of four logio, 21 As you recall, we discussed an event where a 22 voltmeter at Sequoyah shorted out some output transistors in 23 the RPS 1 discussed with you at a previous ACR8, and we issued 24 an information nottee on that.
This relatively 25 straightforward event is something that those of us who follow
88 O
1 reactor events, review the events every 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, see quite 2
frequently, although not this specific set of details. It's 3
very common.
4 In reviewing the events from the previous 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, 5
we see an instrument techniolan has inadvertently caused the 6
reactor to scram even though he is following approved 7
procedure.
We were interested in the corrective measures 8
taken --
9 MR. REED:
You said following approved procedure 10 with his instrument set up wrong.
11 MR. WEISS:
Yes.
12 MR. REED:
I have been having quite a running battle 13 with Professor Kerr with respect to all thumbs maintenance,
.,14 and he has labeled it "thumby maintenance."
Here it looks to 15 me like a personnel issue of people who have perhaps low l
16 mechantoal comprehension who are running around with oranes 17 and voltmeters and all kinds of things.
So I wish that 18 Dr. Kerr would be reminded by the rest of us that here is 19 another "thumby maintenance."
20 MR. WYLfE:
In this cause it is electric i
21 maintenance.
22 MR. REED:
It's the same thing.
You don't want 23 bulls in the china oloset.
And, if they have low mechantoal I
24 comprehension and low electrical comprehension, they ought not 25 to be in there.
89 1
MR. EBERSOLE:
I think it is fascinating.
The 2
45-second time constant is just enough to override the 3
recovery of the systems in the initiated two-channel trip.
4 He was doing that deliberately to get a snapshot, as you said.
5 MR. WEISS:
Yes.
6 MR. EBERSOLE:
And we walked right into communizing 7
the RTDs.
8 MR. WEISS:
Part of the corrective measures taken as 9
a result of this particular event was to include a precaution 10 in the procedure about not only the consequences of what might 11 happen if you make an error but also a precaution to check the 12 readings.
That was one thing that struck me when I first 13 heard about the event.
I wondered how the voltmeter could 14 give them a reading anywhere close to what he would expect if 15 he had the leads connected up incorrectly.
16 MR. EBERSOLE:
Could that be a coincidence?
17 MR. dE188:
kt turns out the readings weren't 18 anywhere near correct, but he wasn't in a position to know 19 what the correct values were.
Part of the corrective measure 20 is to include a precaution as to what the expected value 21 should be, and if it is not within the orpooted range, not to 22 proceed further.
23 MR. REED:
It sounds to me 1the you are overocoking 24 procedure unless you also add a statement for check of each 25 finger of the person on his hands, or thumbs or fingers.
90 1
MR. WARD:
Glenn, you perhaps have some information 2
I don't have, Do you have some information on this particular 3
mechanic which indicates that he would have been weeded out by 4
a natural ability selection test?
5 MR. REED:
I'm looking at events at TVA with respect and I passed on to Professor Kerr a number of them that 6
to 7
indicate to me that there is a people problem with respect to 8
their natural abilities.
9 MR. WARD:
It is the last part of that which I think 10 is kind of conjecture on your part, though.
There is a people 11 problem.
We would agree with that.
But you have somehow l
(_,/
12 concluded that it is tied in with natural ability.
13 MR. REED:
Sure.
I jump to conclusions all the 14 time, and my batting record, I will let it speak for itself.
15 MR. EBERSOLE:
This operator, does he know that he 16 is carrying that kind of potential influence in his hands with 17 a malset digital voltmeter?
I don't know the specifies of this l
19 particular instrument technician except to the extent I asked 20 a similar question of the resident and the resident told me he 21 thought that this particular techniolan performing the test 22 and the test engineer who was present in the immediate
)
23 vicinity of where the readings were taken were not as familiar 24 as he would have 11ked them to have been with the layout of 25 the equipment and what they were doing.
I
m 91
,s 1
That was his personal statement to me.
I understand f
2_
shift manning at the time was quite adequate and that there 3
were lots of people in the control room.
Incidentally, this 4
test engineer, I understood, was also an STA, if that helps.
5 MR. EBERSOLE:
That makes it worse.
6 MR. REED:
Just another part of the lower mechanical 7
comprehension, probably.
4 8
MR. EBERSOLE:
Doesn't anybody tell him: You're 9
walking around with an instrument in your hand and you carry i
10 with you the potential for a million dollar shutdown if you 1
11 don't check in the right probes?
12 MR. WEISS:
You are asking me a question that I 13 caw't answer.
I know 14 MR. EBERSOLE:
It sounds so casual.
l 15 MR. WEISS:
I know that those of us who review 16 events, those of us that have been operations of'acer, 17 definitely see a trend.
As a matter of fact, one of the first the plant 18 things I was taught as an operations officer was 19 cannot tell you what the problem is; we had a trip and we that one of the first things you come back 20 don't know why 21 with is were there any surveillance or maintenance activities 22 under way when that event occurred?
And then they will come 23
- back and typically say something like, oh, yes, we had 24 somebody out fixing such and such or a surveillance was under 25 way, and then you can ask some follow-up question.
92 1
MR. EBERSOLE:
I suggest whatever they do might be 2
in part predone by installed instrumentation that displays the 3
information they are seeking, like, you know, what was the 4
voltage reading on the RTDs by an installed voltmeter.
5 MR. WYLIE:
You can't always get that.
Sometimes 6
the problems eliminate themselves.
They kill each other.
I 7
have known some electricians that did something like this and 8
they got on some hot busses and it killed them.
9 MR. WARD:
You are not proposing a solution --
10 MR. WYLIE:
No.
Sometimes it eliminates intself.
11
[ Laughter.3
(
12 MR. EBERSOLE:
I don't think this would be for the 13 full committee except as a human factors problem.
14 MR. JORDAN:
The next presentation is on 15 Davis-Besse, loss of main *feedwater and auxiliary feedwater on 16 June 9th.
Al DeAgazio will give the presentation.
He's the 17 Licensing Project Manager.
r 18 I'd like to say a few words about how the NRC is 19 going about this particular investigation, which is different 20 than we had previously done.
There is a Commission paper, 21 SECY-85-208, which describes a proposed incident investigation 22 program, and part of that paper indicated that the Staff would i
23 perform investigations during this intervening time until the 24 Commission adopts this or some other program in this fashion.
25 MR. WYLIE:
What's that SECY number?
93 1
MR. JORDAN:
85-208 dated June 10th.
2 The basic attendance, I guess you would say, of this 3
program is that when there is a significant operating event 4
and these are of the level of abnormal occurrence type events, 5
we would expect to set up a multiple-discipline team using 6
technical experts from the various NRC offices.
This was done 7
in this case.
8 Ernie Rossi was selected as the Team Director.
The 9
team is established by the Executive Director's Office, by 10 Bill Dircks, based on recommendations from one of the Office 11 D i r e'c t o r s, in this case Harold Denton.
The team, then, is 12 composed $of' personnel from various offices -- NRR, from ISE, from AEOD','and, in fact, in this case, we used a member from 13 14 the Chattanooga Training Center as one of the participants.
x 15 It'was a team of four.
They went to the site three 4
16
- days, fter this particul.ar event,'and the object was 17
. f ac t f;indiny.
Tife i n'd i v i du a l s liia t are on the team have no l
\\. " i' I
x s
l 18' I dther, duties ~ durini'this time.
Investigation is presently l
N
\\
,%g A('s oge d dl e d tc b e ~c omp l e t e d.'Ju l y 22nd.
This would be by issuing 19
=
,t 20 iaifu'rSal report.'
s a
i f
The_ report'would focus \\ on
't h e description of the 21 3
w.
s.
1 2,' r, f '4 v e n t, the identification of t h 'e root causes, and findings and
+(
~,
i.,
g 3
'p 3
'? copcitzsions.
It will be forwarded directly to the Commission
's s
and ts the EDO and 8. o ;t h e ACRS,for independent review.
24 j
S t.;
i 25 Ths tsam is taking-or has taken statement of the i<1 p,-
1
's so
94 N-1 personnel involved.
That was their first order of business.
2 And these were statements that were recorded by a court
(
3 reporter in the process of these interviews.
4 Insofar as was practical, the equipment was placed 5
in an as-found condition and left there until released by the l
l 6
team.
And so after the initial event was over, then the l
t 7
plant, other than safety responses, was frcson until the team l
8 released those components.
j 9
A Regional team was also there to review other
(
10 matters and to assist the team, and of course the Regional 11 personnel were dispatched to the plant the following day and 12 were there and have been there essentially ever since.
13 So the description today is based on information 14 that the team has been providing to the Staff and, I believe, 15 the ACRS has been getting wopies of these daily or I
16 every-few-days repcrts, 17 The Staff has issued an Information Notice, and it I hope i t.' s attached to your package -- which 18 is attached 19 provides -- page 36 of your package, issued yesterday 20 provides a brief description of the event and a couple of the 21 important chart records.
22 So this was our feedback to the industry, in 23 addition to the INPO feedback that has occurred.
24 So with that, it seems appropriate to go ahead and 25 give you, then, the description of this particular event.
95 0_s 1
MR. REED:
Ed, there are two issues that have been 2
much discussed in the last year that come into play here.
One 3
is the NTSB-like organisation, which we had not crystallized 4
in any new way, and apparently you put out a SECY on June 5
10th, the day after the incident, to say, "This is the way 6
we're going to structure all this."
Is that what you just 7
said?
8 MR. JORDAN:
Those two things were independent.
9 MR. REED:
They were independent?
10 MR. JORDAN:
Obviously we can't put out a SECY that 11 quickly.
12 MR. REED:
Anyway, we've been chewing on this t
13 NTSB-type thing, and.it looks to me, from what I read and what 1
14 you just said, you're trying to put your best foot forward 15 with respect to sort of doing the kind of things that came out 16 of the Brookhaven report, except one, and that is, you didn't l
17 involve parties; is that correct?
18 MR. JORDAN:
That is correct.
I 19 MR. REED:
So that is the big NTSB-like controversy, l
20 which I think some of the full committee will want to address 21
-- how does that work out?
l 22 The other thing, of course, is the B&W system, the l
(
23 low-set stesa generator system and its potential 24 vulnerabilities, which, of course, has been highlighted by 25 TMI, by Davis-Besse before Three Mile Island.
Now Three Mile 1
96 j (p i
s-1 Island has kicked it back to Davis-Besse.
These things have 2
been ohewed over by ACRS members quite a bit, and, of course, 3
I'm very interested in these potential vulnerabilities, and 4
I'll be very interested in your -- the NRC Staff answer to 5
NTSB-like organization and their findings.
I will be very 6
interested in those findings.
7 MR. HERNAN:
I'd like to say a few words about one I
8 of the investigating team members, Larry Bell, the person Ed 9
mentioned who was from the Training Center in Chattanooga.
10 For the information of the committee, Larr/ is a former i
11 operator in Arkansas, one which is a B&W plant.
He's been fh
'\\,)
12 with the Training Center for about four or five years, and he 13 has set up the NRC's course on B&W reactors, and he's also the 14 instructor at the B&W simulator, which is at Soddy-Daisy, i
15 Tennessee, that being a Bellefonte simulator.
16 I can attest personally that he is very 17 knowledgeable from the standpoint of B&W plants and I
18 operations.
19 MR. EBERSOLE:
May I ask a couple of questions?
20 This point has been under the gun for a good many 21 years, like seven or eight, recognized early.
22 Were those two turbine driven aux feed pumps buried 23 down in the basement with electrio coolers. and I think 24 original electric valving.
That it had any number of ways 25 that it could fail, tetality of failure of feedwater.
97 h
J 1
So I guess it's been under fire a good many years.
2 MR. JORDAN:
Yes, sir, it has.
3 MR. EBERSOLE:
Nothing, of course, ever got done, 4
except, I think, somebody generated some number-crunching and 5
decided via the PRA that it didn't look too bad.
6 Will there be a reconsideration of that aspect of 7
that prior justification?
8 MR. JORDAN:
I'm sure there will.
9 The Staff is waiting for the findings of this 10 committee -- of the investigative team before launching into
~
11 any prograq.
)
12 NR. EBERSOLE:
It occurs to me, maybe this is a 13 focal point in time, as well as the events here, to reexamine 14 the whole PRA approach and see whether it gives us the right 15 answers or not, versus a judgmental consideration of 'he 16 rather obvious shortfalls, which have long existed at this 17 plant.
18 I just now noticed here, it's boobytrapped by the 19 operator having a couple of hand switches in front him that he 20 can just guarantee turnoff of the feedwater.
21 MR. JORDAN:
Yes, there are those boobytraps, and 22 perhaps we'll let Al walk through them and eliminate some of
)
23 them.
d 24 MR. REED:
Just to follow up on a comment, I will be 25 very, very particularly interested in Mr. Bell's findings.
98 1
You can warn him in advance.
2 MR. DeAGAZIO:
Before I start the sequence, just to 3
put.a simplified diagram of what the Davis-Besse feedwater 4
system looks like.
5
[ Slide.3 6
There are two once-through steam generators, two 7
main feedwater pumps that are turbine driven, two steam driven 8
main feedwater pumps feeding through feedwater, heating 9
trains.
There's a main control valve for feedwater flow 10 control.
Bypassing that, there is a smaller valve for startup 11 feedwater flow control.
The main stop-valve for each
( j/
12 feedwater -- for the feedwater to each steam generator.
The 13 feedwater in the once-through steam generators comes in at 14 about the lower third of the steam generator, flows down 15 through the downcomer region, mixes with some recirculation 16 flow, and then up through the steam generator shell.
17 Auxiliary feedwater is injected at a higher level at 18 the top of the steam generator.
It is sprayed in over the the auxiliary feedwater at 19 tops of the tubes.
This is 20 Davis-Besse consists of two steam driven feedwater pumps, one 21 pump for each steam generator.
There are provisions for being one turbine 22 able to crossfeed the flow from one train 23 driven pump by closing a valve to the one steam generator, 24 opening up another valve, and feeding, for example, Steam 25 Generator No.
1 with Feedpump No.
2.
1 There is an electrically driven startup feedwater 2
pump which feeds through either the feedwater trains or, not 3
shown here, it can bypass either train and follow the normal 4
feedwater flowpath.
This is the purpose of the startup 5
feedwater control valve.
6
[S11de.3 7
At the time of the event, the plant had been 8
operating at 90 percent power.
There was no special 4
9 surveillance test or unusual operations going on at the time.
10 One main feedwater pump was in manual control, and one main 11 feedwater pump was in automatic control.
12 It's my understanding, the reason why one was in 13 automatic and one was in manual was, Toledo Edison had been i
14 having difficulties with the control systems for the main l
the speed governors had 15 feedwater pumps, and these had been 16 been replaced on both pumps at the last refueling outage.
In 17 fact, the week before, there had been a trip also as a result 18 of governor problems at.the main feedpumps.
So to minimize l
19 the. possibility of tripping both main feedwater pumps, Toledo i
20 Edison had one in operation.
21 The event started by the tripping of one main 22 feedwater pump, the one that 9,s s in automatic control on 23 overspeed.
The plant automatibally started in a power runback i
i 24 at 50 percent per minute; however, the runback was not fast 25 enough, and approximately thirty seconds later when the
100 v
1 reactor was at 78 or 80 percent power, the reactor tripped on 2
high pressure.
and I'm 3
Approximately a second later, there was either what's called a 4
not certain of this information 3
half-trip or a full trip of the steam and feedwater rupture 6
control system.
The purpose of the steam and feedwater 7
rupture control system is to detect either steam main 8
failures, feedwater line failures, failure of feedwater flow, 9
or failure of all four reactor coolant pumps, and take the 10 appropriate action as far as the secondary system in 11 conc'erned.
k 12 MR. WARD:
You said that includes the failure of the the reactor coolant pumps are
.13 reactor coolant pumps 14 included?
15 MR. DeAGAZIO:
Yes.
Let me describe what the steam l
l 16 and feedwater rupture control system does.
l 17 In the event of a low steam main pressure, 600 psi 18 or less, the steam and feedwater rupture control system will 19 i s o.l a t e both steam generators by closing the main steam I'
20 isolation valves, which are not shown on this diagram.
It j
21 will close both isolation valves on the feedwater lines.
(
22 Yes, it closes the auxiliary feedwater and the main
(
23 stop valve for the feedwater flow, for the steam generator 24 that gives the signal, so if the Steam Generator No.
1 went l
l 25 down to 600 psi, it would be isolated both on the main steam
=. - -
4 101 Ow i
flow, main feedwater flow, and auxiliary feedwater flow.
2 The flow from the aux feedpump for No. 1 would be 3
diverted by opening this valve (indicating), and would be 4
diverted to Steam Generator No.
2.
So the steam generator 5
that presumably has been ruptured would be totally isolated.
6 MR. EBERSOLE:
Is the purpose behind all this to 7
preclude overpressuring the containment with continued 8
discharge into the secondary side, plus extracting here out of 9
the secondary as well as the primary?
4 l
10 MR. DeAGAZIO:
I believe that's part of the reason.
11 The other reason for isolating the steam generator in the w
12 event of a steamline rupture is to prevent a rapid cooldown of 13 the reactor coolant system.
14 MR. EBERSOLE:
I see.
But the bottom line, though, 15 what they have introduced with all this, is the potential for 16 cutting off all water to the secondary, haven't they?
it's possible that's 17 MR. DeAGAZIO:
That's what 18 what happened.
19 MR. REED:
This is an automated sensing system that 20 decides when a steam generator is in trouble, Then it closes 21 main stop on the auxiliary system?
22 MR. DeAGAZIO:
The system consists of pressure
)
23 switches in the main steamline, pressure switches in the 24 feedwater system, level sensors in the steam generator, and 25 contacts on the reactor coolant pumps.
t' 102 0
1 MR. EBERSOLE:
By the way, it's also predicated,
'2 isn't it, that the isolation valves that isolate the main 3
steam water lines to the turbine, in fact, perform the 4
isolation function, so there's no cross-connection on the 5
discharge?
'The system is sectionalised by isolation valves on 6
the main steamlines.
If they work, then you can have 7
different pressures; if they don't work, you can't.
Is that 8
correct?
On the output to the main turbines?
9 Obviously, if you are headed together 10 MR. DeAGAZIO:
If you had a large rupture, steam 11 pressure in the main nearest the rupture should drop faster.
j 12 MR. EBERSOLE:
It would drop faster, so okay, you're 13 racing with time, then, depending on whatever the pressure 14 drop is.
But certainly,-if you don't get sectionalization on 15 the main steam isolation system, the secondaries will come to 16 uniform pressure, won't they?
17 MR. DeAGAZIO:
That may be possible.
18 MR. EBERSOLE:
Because they're tied together.
And 19 then both of them will go down, and that guarantees you're 20 going to dry up; is that right, because you're now precluded 21 from pumping water from any source.
1 22 MR. DeAGAZIO:
There are also turbine stop valves 23 in this plant which will close.
24 MR. EBERSOLE:
Do they prevent crossflow?
25 MR. DeAGAZIO:
The mains come together at the
2 103
. f-)s
('_
1 turbine, and the turbine stop valves would, I believe --
2 MR. EBERSOLE:
There are two designs.
The stop 3
valves sometimes prevent crossflow between some areas; 4
sometimes they don't.
Sometimes the crossflow is permitted 5
ahead of the stop valves.
I think it would be interesting to 6
show that complex in an enhanced diagram when we show this to go back and look at the main steam 7
the full committee 8
network and see whether the turbine stop valves precludes 9
crossflow.
10 MR. REED:
I don't think the valving arrangement 11 here is complete, and it would seem to me, when it comes up to O)
(j 12 the full committee, we ought to have some check valves on l
13 those main feedpumps somewhere.
We ought to indicate motor 14 operators and non-motor operators.
15 MR. DeAGAZIO:
This is a very simplified diagram.
16 We'll see if we can improve the detail, so that it's not too 17 complicated and maybe set some of these questions to rest.
18 MR. EBERSOLE:
The effort was obviously to try to i
19 get, pure steam-driven energy eJurces to get water in the l
20 boilers.
It brings up, I think, an earlier question that the t
21 ambient temperatures on these aux feedpumps was controlled by 22 alternating current boilers, which made them dependent upon 23 the AC system, which I don't know whether it's been changed or l
24 not.
25 It also brings up the question of what drove the l
i 104 s
1 valves on the steam turbine-driven aux feed system, what 2
operated the valves, what electrical system operated the 3
valves?
4 MR. DeAGAZIO:
The valves in the auxiliary f e e d*'a t e r 5
system, one train is operated by AC, one train is operated by 6
DC.
7 MR. EBERSOLE:
On AC power failure totality, the AC 8
system is dropped, isn't it?
9 MR. DeAGAZIO:
Yes.
10 MR. EBERSOLE:
I just wanted to say what Glenn 11 said.
An enhanced diagram for the full committee, which would
)
12 show the main steam secondaries, as well as what works the 13 valves, would be helpful 14 MR. DeAGAZIO:
On low steam generator level and high 15 feedwater pressure differential, which would indicate a i
16 failure of the feedwater system.
The main steam isolation 1
6 i
17 valves are closed.
The maan stop valves for feedwater are 18 closed.
The respective isolation valves for the auxiliary 19 feedwater flow are not closed on a loss of reactor coolant 20 pumps, auxiliary feedwater flow, only is initiated to go with 21 natural circulation.
22 As I said, the reactor coolant system created a 23 high-pressure trip at 78 to 80 percent power, and 24 approximately a second after the reactor trip, the steam and 25 feedwater rupture control system either gave a spurious
105
\\
half-trip or a full-channel trip on low steam generator 1
2 level.
I'm not sure, based on the information that I have 3
right nc7, which is correct.
4 The team had indicated that that was a low steam 5
generator level trip, and it was a full trip.
- Now, 6
considering the initial conditions at which the reactor was 7
operating at this time, this should not have been a low steam 8
generator level 9
The team also indicated that approximately a second 10 or so later, that trip had cleared.
But in the meantime, the 11 main' steam isolation valve had closed.
With the closure of 12 the main steam isolation valve, the one main feedwater pump i
l 13 which was running at manual control eventually ran out of 14 steam.
It took approximately four minutes for that pump to 15 coast down and trip.
16 So approximately four minutes into the event, the 17 steam generators now have no feedwater and the steam generator I
18 level is beginning to drop.
At approximately six minutes into l
19 the. event, one steam generator reached the low level trip and 20 the steam and feedwater rupture control system generated a l
21 signal to start the auxiliary feedwater pumps and isolate the I
22 steam generators.
(
23 At this time, the operator, recognizing that a low l
24 level was being approached in the steam generators, took I
25 manual action, and unfortunately, he hit the wrong buttons on l
.. _ ~ -
106 1
ths steam and feedwater rupture control system.
He hit the 2
low pressure isolation switches instead of the low level 3
isolation switches.
4 This had the effect of telling the system that steam 5
generator number 1 was ruptured and the aux feed system should 6
be aligned to steam generator number 2.
It also said that 7
steam generator number 2 was ruptured and the aux feed system 8
should be aligned to steam generator number 1.
9 So this closed all auxiliary feedwater valves.
He 10 recognized approximately a minute later that he had made this 11 error.
The design of the steam and feedwater rupture control 12 system is that the low steam pressure trip has priority over 13 the other trips, over the low level trip.
14 MR. REED:
Could you go back to why was that low 15 pressure trip system there anyway?
Apparently it affected 16 both steam generators.
17 MR. DeAGAZIO:
The isolation of the steam generators 18 is there to prevent a rapid cooldown of the reactor coolant 19 system.
20 MR. REED:
Part of the problem of the B&W system is 21 you have no inventory, you have no cushion.
Things are going 22 to go like crazy on the fly.
)
23 MR. DeAGAZIO:
Also to preserve steam, to run the 24 steam-driven aux feed pumps.
No inventory system
[
25 MR. REED:
It introduces that i
t
107 O
1 really introduces a lot of things, doesn't it?
this 2
MR. DeAGAZIO:
Twelve minutes into the event 3
is approximately eight minutes after the main feedwater pump, 4
the second main feadwater pump had tripped.
The steam 5
generator had reached approximately eight inches, which is 6
considered by procedures in B&W plants to be a dry steam 7
generator.
Preliminary information I had indicated the steam 8
generator pressure in at least one steam generator had dropped 9
to something on the order of 750, 800 psig.
10 The operators at that point recognized that they had 11 to restore feedwater.
They dispatched personnel'to the 12 auxiliary feedwater pump rooms to reset the pumps.
They 13 dispatched an individual to the startup feedwater pump, the 14 electrically-driven startup feedwater pump, to attempt to get 15 that into operation.
That is a system that was valved out for 16 other considerations, mainly a high energy pipeline situation 17 in the auxiliary feedwater pump room that was an unanlyzed 18 situation.
19 So to remove that unanlyzed situation, the system 20 was valved out by license condition.
It is manual.
They were 21 locked out, the valves were locked out, so the operators had 22 to go down and manually operate the suction and discharge 23 valves to the startup feedwater system and provide some 24 cooling to the pumps.
25 They also had to replace, or I should say insert
108
\\-
1 control fuses to get the startup feedwater pumps started.
The 2
operators at the startup feedwater pumps and at the auxiliary 3
feedwater pump appeared to have been able to restore some 4
feedwater flow at approximately the same time.
5 It's not clear the moment which pump actually 6
provided flow first.
let me back up for 7
When they attempted to restart 8
a moment.
I think in the questioning of the steam and 9
feedwater rupture contrcl system sequence. I missed one point 10 that is important.
When the operator had recognized that he t
11 had made an error and isolated on low pressure instead of low
(
12 steam generator level and corrected it, the isolation valves 13 should have reopened by themselves, should have reopened 14 automatically.
They did not.
3 15 These valves were subsequently opened by the let 16' operators who went down to the auxiliary feedwater area 17 me back up.
I'm not sure it's the same area.
They were 18 manually reopened.
They were started manually, and then once 19 the, valves had been moved off the seat, the valves continued 20 to open by themselves.
21 MR. REED:
Is this a case of a Carl Michelson 22 operability of valves against maximum pressure differential?
23 MR. DeAGAZIO:
Exactly what the reasons for the the team is 24 valves not opening, I don't believe we have any 25 pursuing that.
~..
109
,s
\\-
1 MR. REED:
The pumps were running, right?
The aux 2
feed pumps 3
MR. DeAGAZIO:
The pumps at that time were not 4
running.
There was back pressure from the steam generator.
5 The pumps had not been restarted.
When the operator cleared 6
the signal, the valves should have reopened at that time.
7 MR. REED:
What did you have, about 900 pounds in 8
the steam generators?
9 MR. DeAGAZIO:
700 to 900.
10 MR. REED:
It sounds like a case of differentials.
11 MR. EBERSOLE:
On page 38 here, they think the e
12 torque switch was malset, which would do the same thing.
13 MR. REED:
Not for two valves, right?
14 MR. EBERSOLE:
Does it seem that there should be 15 some sort of a countermanding switch somewhere that says I 16 will never refuse to drive water into at least one of the 17 secondaries?
Because if you do that, you are in trouble, 18 aren't you?
MR. DeAGAZIO:
You certainly need feedwater.
19 20 MR. EBERSOLE:
There is no other way to cool a PWR, 4
i it looks like you would need a 21 and B&W is particularly 22 roadblock someplace that says I'm not going to cut this water 23 off no matter what.
24 MR. DeAGAZIO:
The operators said after they 25 restarted the auxiliary feed pumps, they had difficulty in l
t.
110 0
1 controlling one aux feed pump from the control room.
It was 2
manually controlled, and both pumps were reset locally. It is 3
my understanding that one pump operated properly; the other 4
pump had to be on manual control.
5 MR. EBERSOLE:
Can the valving network for the aux 6
feed system be fully manually operated, do you know, since 7
they are electrically dependent at this time?
8 MR. DeAGAZIO:
Which valves are you referring to?
9 MR. MOELLER:
These are both the steam issue valves 10 as well as the discharge valves.
I'm getting back to the j
11 point of their being dependent on the electrical network.
12 MR. DeAGAZIO:
They can be opened manually.
13 MR. EBERSOLE:
You can presumably work these pumps, 14 I guess, just on DC to support the turbine auxiliaries; is 15 that correct?
16 MR. DeAGAZIO:
No, I don't believe that's correct.
17 I believe one train is dependent upon AC; one train is 18 dependent upon DC.
19 MR. EBERSOLE:
What did they ever do about room 20 cooling for these pumps to keep the room from getting so hot 21 that the engines would be excessive?
They were originally 22 cooled by HVAC systems that were driven by diesel power 23 electric system, which made them interdependent on the AC 24 system.
25 I think the valves were originaly AC operated. I
-m
,..~.. -
,-.cy-----------r
--r
111 (N
(v) 1 thought they changed one of them to a DC system.
2 MR. DeAGAZIO:
They were changed back in, I believe, 3
1979.
They may, in fact, have been changed earlier than 4
that.
There was a license condition --
5 MR. EBERSOLE:
Usually the applicant will put these 6
turbine-driven aux feed pumps out in open draft room, which 7
accounts for their survivability if you have a steam line 8
failure or if you had other ambient problems, but it happens 9
in Davis-Besse these things are buried down in the bowels of 10 the plant, down someplace in an enclosed concrete cavity.
11 MR. REED:
Jesse, this gets into the conflicts 12 between decay heat removal and security.
All I'm saying is 13 you are damn lucky there was access to these valves for manual 14 operation.
I don't know what the roadblocks of security were 15 that were in existence.
16 I would like to go to this issue of one pump ran 17 okay, the other one didn't.
These are steam-driven pumps, 18 which I don't like in the first place.
I don't understand 19 hanging your hat on two steam generators, one of them making 20 steam and one of them ruptured or unruptured, one getting i
21 feedwater and whatnot.
Is it possible that you didn't have 22 steam enough because of dryout to run both of these 23 turbine-driven pumps reliably?
steam 24 MR. DeAGAZIO:
When the steam feed rupture i
25 and feedwater rupture control system signal was generated,
~.
112
(
1 both pumps started successfully.
Both of them tripped out 2
shortly thereafter on overspeed.
3 MR. REED:
Do you have an explanation for that?
4 MR. DeAGAZIO:
The team is looking into that.
The 5
plant is investigating that.
6 MR. WARD:
They say something about it in the 7
information notice, but it's kind of vague.
8 MR. REED:
You mention 700 to 900 pounds pressure.
9 was it pressure because of dryout? Was there not enough 10 inventory and heat transfer surface in the low levels to be 11 able to maintain a steam of reasonable pressure condition?
12 MR. DeAGAZIO:
I don't really know the answer to i
13 that.
All I can say is that there was sufficient steem 14 pressure to start the pumps and get them up to the overspeed 15 trip.
16 MR. HERNAN:
Mr. Reed, I think this question falls 17 in the category of answers we don't have yet.
Anything we say 18 would be conjecture.
19 MR. WARD:
The information notice says something 20 about condensation in the crossover line, which I don't 21 pretend to understand.
1 22 MR. JORDAN:
This was a statement the licensee made 23 rather than the team.
They conjecture, I believe, that there 24 may have been condensation of the crossover lines which 25 contributed to the overspeed by slugging through the turbine,
. ~. - _... - -
...~-- --- -- - -.
i 113 O
i and that's not confirmed by the team yet.
So we were just s
2 stating it was a licensee statement.
3 MR. WARD:
If the aux feed pumps had not tripped out 4
on overspeed, they would have been shut down at the next step 5
anyway when the operator erroneously hit the wrong button; is 6
that correct?
7 MR. DeAGAZIO:
No, I don't believe so.
I think they 8
would have continued to run.
9 MR. WARD:
They would have continued to run then?
l 10 MR. DeAGAZIO:
Oh, yes.
11 MR. WARD:
I guess I'm confused, then.
What did 12 that action do?
l 13 MR. DeAGAZIO:
Basically, his error had the effect 14 of closing both the discharge valves to the aux feed pumps.
15 MR. WARD:
So the pumps would have run but they 16 wouldn't have had any water in them.
i 17 MR. DeAGAZIO:
The pumps should have been running i
18 against a closed valve.
There is a recirculation valve for recirculation line for testing purposes.
19 testing purposes 20 MR. EBERSOLE:
He has an easy capacity within arms 21 reach for drying the plant out.
}
22 MR. DeAGAZIO:
He has demonstrated that.
l 23 MR. WARD:
That is what you call control room g
24 MR. EBERSOLE:
At some point you have to stop the 25 control room function.
114 I) 1 MR. DeAGAZIO:
Toledo Edison has completed a review 2
of the control room design review and the steam and feedwater 3
rupture control panel was one of the items they had identified 4
as having human and engineering defects.
5 MR. WARD:
HE deficiency.
6 MR. DeAGAZIO:
I have a lot number of what this 7
particular portion of the system looks like.
8 MR. EBERSOLE:
When I look up there. I see one 9
electric-driven feedwater supply source.
It's called the 10 startup pump.
It must run its water through the main 11 feeedwater lines, the heaters, the level control valves and 12 the stop valves; right?
13 MR. DeAGAZIO:
Yes.
14 MR. EBERSOLE:
So it's subject to all those 15 intercedent devices, and even when it's running, what kind of 16 pressure and flow can it develop?
17 MR. DeAGAZIO:
This has full discharge pressure 18 capability.
The capacity of flow is approximately half of the 19 aux.iliary feedwater system.
20 MR. EBERSOLE:
About half?
21 MR. DeAGAZIO:
Approximately.
22 MR. EBERSOLE:
You mean half of full power?
23 MR. WARD:
Half of the aux feedwater.
24 MR. EBERSOLE:
That's what I thought it was, half of 25 the aux feed.
115 g~sv)
I 1
MR. WARD:
Now, I think you said that that probably 2
didn't get on much before the aux feed pumps got restarted; is 3
that right?
4 MR. DeAGAZIO:
The information from the team, the 5
timing is such that the operators, they were in two different 6
areas of the plant, were taking action to restore both 7
auxiliary feedwater flow and to start up the startup feedwater 8
pump, and it's not clear from the information I have available 9
which one actually established flow first.
10 MR. WARD:
I see.
But I understood the aux feed 11 pumps had to be restarted from the field, or at least from 4
12 outside the control room.
13 MR. DeAGAZIO:
That's correct.
14 MR. WARD:
Where the startup pump could be restarted 15 from the control room, couldn't it?
16 MR. DeAGAZIO:
That is not correct.
Because of the 17 license condition that currently exists to eliminate the high 18 energy pipeline hazard, the existing startup feedwater pump is 19 in one of the auxiliary feedwater pump rooms, and when the 20 startup feedwater system is valved in, there is a suction i
21 pipeline that has high energy fluid in it because of an 22 unanalyzed situation on that.
Toledo Edison had that system 23 valved out and these are manually operated valves.
I i
24 It is my understanding because the suction valve to 25 the pump was closed, Toledo Edison procedures required the
)
1 l
116 0-s 1
fuses to be pulled on the control circuits of the pump, so the 2
valves had to be opened manually and locally and the fuses had 3
to be replaced.
4 MR. EBERSOLE:
If the circumstances of secondary 5
feedwater insufficiency had extended how many minutes later, 6
what would have been done?
You would have tried to evoke 7
feed and bleed?
8 MR. DeAGAZIO:
The procedures that would be 9
followed, I believe, are that the PORV would be open to reduce 10 the system pressure, the makeup pumps would be started to 11 provide full capacity makeup.
These are high pressure I
w 12 discharge pumps with 2500, 2600 psi discharge pressure capable 13 of lifting the safety valves, and the high pressure injection 14 system, which in this plant has a discharge pressure of 15 approximately 60/50, I believe it is.
Shut-off would also be i
16 started in an attempt to feed and bleed.
17 MR. EBERSOLE:
But you would have been then 18 dependent on the standard PORVs on the primary loop and the l
19 block valves being open to get above that to reduce the 20 pressure to the point where you had sufficient flow.
l 21 MR. DeAGAZIO:
That's correct.
22 MR. EBERSOLE:
Would you have had sufficient flow 1
l 23 with open PORVs?
l 24 MR. DeAGAZIO:
I couldn't tell you that.
I would be l
25 speculating at this point.
117 f-1 MR. EBERSOLE:
In short, could you feed and bleed?
2 MR. DeAGAZIO:
There have been calculations which 3
have showed with the startup feedwater pump and the existing 4
PORV capacity after 30 minutes there is the capability in 5
conjunction with the makeup pumps to cool the plant.
I'm not 6
certain what the situation is without the startup feedwater 7
pump.
8 MR. EBERSOLE:
You are telling me there is really 9
not'a straight feed and bleed capacity without assistance from 10 the startup pump?
11 MR. DeAGAZIO:
I don't know whether it'has that or 12
- not, i
13 MR. EBERSOLE:
I see.
14 Are there any other questions?
15 (No response) 16 Do you have any further questions, Glenn?
17 MR. REED:
I have about a million of them, and I 18 guess we wouldn't finish in time.
I guess there is no great big 19 But I think here 20 stir, there is no core melt.
Yet, I don't understand how you 21 can go without feedwater for 11 minutes on a system with an 22 inventory capacity in the secondary of only about three of 23 four minutes.
And I have to assume that somehow the primary 24 kept this thing from overheating the fuel.
And I have to 25 assume that, and I have to say isn't it lucky that these are
118
,,s f'd I
1 raised steam generators instead of low set stem generators.
2 Because, if you had low set steam generators, I think it 3
would be a much more complicated scenario here.
4 I,
quite frankly, have to say this again.
I have 5
said it before, that I don't think the B&W system should be 6
without a dedicated primary blowdown system.
7 MR. EBERSOLE:
Any further observations or 8
questions?
9 Dade, do you have any?
Or, Dave?
10 Let me ask the committee opinion about taking this at least the front end of this accident to the full 11 to
(
12 committee.
I presume we want to do that.
13 And, get an enhanced diagram when we want to do 14 that.
And talk about the interdependencies which we didn't 15 stress much here.
And whatever the additional information you 16 will have picked up.
17 Well, goe whis, we don't have any time.
18 MR. WARD:
I get the impression Mr. Jordan wanted to 19 w a i,t. But, I think the committee ought to hear this.
20 MR. EBERSOLE:
Don't you think end we can get a 21 front end exposure on this, because it has attracted a lot of I
22 attention, I guess.
23 MR. JORDAN:
Certainly, we have no objection to
\\
l 24 providing the same sort of presentation.
We are just simply 25 unable to give you the full story, and I apologise for that.
L
I 119 x
\\
J 1
MR. EBERSOLE:
Well, you can't do everything at t
2 once.
I understand.
3 How much time have we got for the full committee?
4 MR. ALDERMAN:
An hour and a half.
let me put down 5
MR. EBERSOLE:
I had down here 6
the ones I had put down.
7 Rancho Seco, unisolatable event; the Oyster Creek 8
scram discharge volume failure; certainly this event here; and 9
I had a questionmark about the interaction of the incore flux 10 mapping system as topics.
11 I have four down.
What is the other members of the 12 subcommittee 13 MR. WYLIE:
What did you have down for the diesel 14 generator problem?
15 MR. EBERSOLE:
Calvert Cliffs.
16 MR. WYLIE:
That is broken straps 17 MR. EBERSOLE:
I thought that degenerated into no 18 more than a random failure mechanical components.
19 MR. WYLIE:
Only one other comment I would make 20 about that.
Part 21 went out just on that one vendor.
But it 21 seemed like to me it has the potential of affecting other 22 vendors' machines, and perhaps the Staff should investigate.
23 MR. DROMERIC:
We are evaluating this problem, and 24 it looks like we will issue an information notice.
25 MR. EBERSOLE:
Is that reasonable then, in view of
120 1
the subcommittee here that we take the items that I noted, 2
which were items 2,
3, 8 and 11 and sort of distribute the 3
time across the hour and a half that we have, with any 4
amplification that you care to put on any of these things?
5 MR. JORDAN:
The hour and a half will be, probably, 6
more than sufficient.
7 MR. EBERSOLE:
It will be all right with you if we 8
fall short?
9 (Laughter)
.s 10 MR. WARD:
I guess I would vote for number one, the 11 Hatch --
12 MR. EBERSOLE:
Tne Hatch stuck open PORV, SRV,-that 13 was the one that I think, Dade, you said you were going to 14 take up in the course of additional --
15 MR. MOELLER:
We-won't do it -- maybe we will touch 16 upon it in our discussion of the control room habitability 17 work of the staff.
I will take it back.
If you want to hear 18 about it --
19 MR. EBERSOLE:
I think it might be a practical 20 example of an intricate system, interaction.
21 MR. MOELLER:
Let's stick with it then.
22 MR. EBERSOLE:
I think we will do that.
23 MR. WARD:
Wasn't 3 the one you said we were going 24 to hear more about through the ECCS?
25 MR. EBERSOLE:
We said that as well.
That is why I
3.
'\\
(.
s s
I s
w q
)
s-
-f\\*
.. y '
-gg,.
y
(
s m
s, 121 qq y
'N 4%
\\ g%t.
s'2-v
.s e-
't ' },}h a v s 4 4 ge.se s t i onma r k ptst th So I take the subcommittee's s.
1
- (Iw -
4
's
,(
\\
vi on;it?
2 t'
3 i.
MR. WARD:
$I'think we will get more from the g
t t
analysis anpi e od that.'
%;hermohyt'raulic
(-
x,st Q'
t t-5 3.MR.
EBEFSOLE:
Yes, I am sure we would.
I put it
'\\
s
(
' '6 Y own here wN N,a, quest'ionmark on it.
., g~
7 --=.. I
.\\
We' c an cut out, if you want, the seismic interaction 3 ' A' i
3 3
- s. ; s A.
s i
'3 on i t;[o z e flux mapping system to a later date.
! 'l 2
-[
-9
- h
\\MR. MOELLER:
So, wh..tch ones are you doing?
'"~
,g
' 'MR.
EBT.R50LE :
Now I,have got the Hatch stuck open
,10 l'1 P,ORV., That is i tem 1.
I've got 2,
and I've got 8,
and I've y
y
'12 g o tL I q ' '
s mt 13 And ge will have at least4 an hour and a half and if 14 we can use less ofsthat, fine.
1 one of these diagrams to be ask(d 15 I think we
,i 8
5, 3
16
,}xpandedla little bit, and the discussions extended, 1
- \\ l A
3 17 3especiaD1y oE tho' case o.f the scram discharge volume.That
'w g
\\
(.
f al c i na t i ng l y '1'i ke, t he old Hatch event, yet with 18 looks so 5,
4 1C d i f.f o r erse e's where they did use blowdown to stop the leak and g-t 20 get into a reset mode.
You know, some operational aspects of s,
' s.x.
s 21 that.
Herd it Idoks like at least they knew what they wanted A'
s 22 toNdo to ge t 'the pressure down.
23' 2
,i ME' JORDAN:
Yes.
24 MR. EBERSOLE:
In the other case they just let it
(
25
. sit andN1eak, and.that was an interesting part of that m
.-s s
s y
m
l 122 t'"
i evolution.
2 Any further comments about the subcommittee?
3 MR. REED:
I know there is a screening operation by 4
the staff before we get this list, But, I would just like to 5
point out I was a little surprised to see the two stuck rods 6
at Wisconsin Electric Power Point Beach was not on the list.
7 I'm a little surprised to see that Browns Ferry 3,
high 8
pressure coolant injection, HPCI torus suction valve shunt 9
windings roversed was not on the list.
I was a little 10 surprised to see that the Commonwealth Edison LaSalle failed 11 expansion joints flood cribhouse is not on the list.
Or, was l
%_/
12 that taken before?
13 MR. JORDAN:
No, that one has not been taken up.
14 That was somewhat similar, I guess fou would say, 15 to a Quad Cities event of ten years or so ago.
There were no 16 complications in that particular case in terms of surprises of 17 equipment that was located that malfunctioned.
18 MR. REED:
All service water pumps and circulation 19 pumps were lost due to flooding.
20 It sounds like it has some significance. And it 21 relates to where these expansion joints are located.
22 Well, I was a little surprised that those three were 23 not on your list.
Apparently they were screened out.
24 MR. JOREAN:
And we do interact with the ACRS staff 25 in terms of events as well So, we would be pleased to bring
123 1
those back or look at them again and bring them to the 2
committee subsequently.
3 MR. HERNAN:
Glenn, one of the things we look at in 4
the screening process, totally objectively, is when an event 5'
has generic implications at other plants.
Of course, you get 6
two people, they can have two different opinions of the 7
generic implications or the overall safety.
8 MR. REED:
I think the Point Beach stuck rod, which 9
is a design manufacturing modifications issue is going to 10 happen to a lot of Westinghouse reactors, so it certainly has 11 generic implications.
12 I think flooding the cribhouses is generic.
l 13 What was the other one?
14 MR. JORDAN:
That was the Browns Ferry Torus the shunt line was reversed.
15 suction valve reverse f
16 MR. REED:
Well, here on the Browns Ferry thing, it 17 is another TVA, and there is this running argument, and I 18 focus on it, whether it is "thumby-maintenance" or not.
l MR. EBERSOLE:
I guess that is going to be a new 19 20 word, "thumby-maintenance."
21 Ed, in this matter of Davis Besse, has that had a 22 PRA done on it?
23 MR. JORDAN:
I do not believe that it does, no.
24 MR. EBER, SOLE:
Sooner or later I am sure we are 25 going to run into a cross examination of the PRAs versus real 1
124
(
1 life.
I can't see a PRA ever producing a stuck-open SRV that 2
you just described through the ventilation system, without a 3
good deal.of straining of the examiners on the real physical 4
interactions.
and I just don't remember --
5 But I had thought 6
that Davis Besse had passed judgment and given a blessing on 7
reliability analysis that they have adequate feedwater 8
systems.
9 MR. JORDAN:
There was, in fact, a PRA done on 10 the system itself.
But there was a subsequent commitment by 11 the licensee to install a motor driven feed pump --
12 MR. EBERSOLE:
Which has not yet been 13 MR. JORDAN:
Which has not yet been iststalled.
14 So there finally, I understand, was a commitment.
(
15 But physically has not yet been installed.
16 MR. EBERSOLE:
I don't know whether this is a point 17 of comment about it or not, but sooner or later the PRA 18 conclusions versus the r.eal-life conclusions are going to have I
19 to be put side by side and we will come to grips with the 20 reality of these PRA assessments that we are doing, i-21 I seem to see a number of incredible event 22 combinations that are coming up, and this looks like a I
l 23 fascinating combination of coincident factors that wouldn't i
24 ever get on the books.
I 25 MR. REED:
Jesse, what you are talking about is PRAs l
t-
125 t
1 under Murphy's Laws.
2 (Laughter) 3 MR. WARD:
If the electric driven aux feed pump had 4
been installed, its discharge valve probably would have been 5
closed by this incorrect action of the operator anyway.
6 MR. JORDAN:
That has been hypothesized by the 7
staff, that perhaps it wouldn't have been independent of the 8
other two trains.
It could share valving 9
MR. WARD:
If there is a commitment to install them, 10 there is probably a design for it.
Or, isn't there?
11 MR. JORDAN:
I'm sure there is.
12 MR. STOLZ:
I am John Stols.
There are a couple of 13 items I wanted to clarify.
14 First, there was no PRA prepared for the Davis Besse 15 plant.
There were, however, reliability studies prepared 16 that dwelt on the aux feed system.
And the latest, of course, 17 was prepared in December of 1981 that was submitted to the 18 staff for its review.
That was subsequently reviewed by the 19 s t a_f f using the Brookhaven people.
20 What was the second point that you raised, sir?
21 MR. JORDAN:
The motor-driven feed pump.
22 MR. STOLZ:
The upgraded pump was really a decision 23 to replace the startup pump into the turbine building, because 24 of the relocation problem with regard to the high energy line 25 break.
Things like Al DeAgazio mentioned.
126 Oi
~ have a rather preliminary proposal to i
We just 2
relocate them, and there are no details regarding it, except 3
that it is going to be a full capacity pump, it is going to be 4
a manual start, you can hook it up to vital power.
There is 5
no automatic start proposed.
It is not tech spec'd. So, if we 6
look into that as a third pumps with all the safety grade 7
ingredients, that has to be reviewed by the staff.
8 We need a full fledged submittal on that.
9 MR. WARD:
But would it be tied in to this automatic 10 isolation system?
l 11 MR. STOLZ:
No.
As a matter of fact, the obvious i
i (m
12 conclusion is that one has to straighten out the auxiliary 13 feed system and all the controls quite apart from the third 14 pump issue.
And I think that is certainly clear.
15 MR. EBERSOLE:
It certainly seems odd to me that you 16 can so easily in essence dry the plant out of any cooling 17 function other than bleed feed by just straight out 18 manipulation of switches.
And somewhere you don't do that 19 mechanical cross bars like on transfer switch says you do one 20 thing you counteract it with another so you don't dry out.
21 By the way, I didn't hear any discussion of what had 22 been the merits or demerits of trying to enhance the electric 23 pump function by depressurising the secondary, which is 24 invoked by Palo Verde, you know, as the road home.
Since they 25 have no PRVs, they have invoked depressurization of the
T 127
-w v
i secondary as a means to enhance getting secondary water in.
2 Unfortunately, when you depressurize a secondary you 3
also lose secondary inventory.
You might cool it down, but 4
you lose some water int he process.
So, it is a contradictory 5
process.
Shall I throw away some water while I am trying to 6
cool down, or keep it while I am still hot?
7 And I don't know what was thought about here in this 8
case.
I didn't detect any attempt to depressurize the 9
secondary to make it easier to get water in.
10 MR. DeAGAZIO:
Jesse, the team did come up with some 11 information that the atmosphere dump valves had been
(
l
\\
12 operated.
I don't know exactly what the purpose of the L
13 operations were, or for how long the valves had been opened.
14 I just know that the operators had operated them at some point 15 during the sequence.
16 MR. EBERSOLE:
Before you operate it, you must 17 argue, do I want to depressurize to get more water in, or am I 18 willing to lose what water I still have to do that. And I 19 don /t know how you determine that.
20 Anyway, I guess we will get a more complete story as 21 the thing unfolds.
We will do the best we can at the full 22 committee meeting and we will tell them that it is an interim 23 story.
24 MR. JORDAN:
We appreciate the ACRS patience on this 25 matter.
_. _. - ~ ~ _.. _ _.- _ __
128 1
MR. EBERSOLE:
Any other comments.?
2 MR. HERNAN:
I have one more question on Oyster i
3 Creek.
4 You mentioned some of you wanted for the full 5
committee, some additional information on the probability of 6
both valves failing at once?
7 MR EBERSOLE:
Oh, yes.
That is the nominal 8
conclusion when you put in redundant valves, one or the other 9
will surely work with some low numbers associated with 10 coincident failure.
11 I wondered what they were here.
12 MR. HERNAN:
This is information that would come I
13 from a PRA if one exists.
14 MR. EBERSOLE:
It might come from GE or somebody 15 that has now seen a failure with a probability of X,
or 16 whatever it was.
17 Any other comments?
18 MR. JORDAN:
With regard to the three items Mr. Reed 1
19 passed to me, perhaps the most expeditious thing would be for 20 us to consider those for the next meeting with you.
21 MR. REED:
Sure.
22 MR. JORDAN:
I would say any events that the ACRS i
23 subcommittee feels should be brought to this meeting 24 MR. EBERSOLE:
We will have to point them your way.
i i
25 MR. JORDAN:
-- we will be glad to consider it.
. - - _. - _ - - -. -, - - - - _. - ~ - - _ _ _...
t 129 1
MR. EBERSOLE:
Thank you.
2 With that, unless there are any other things to be
(~
3 undertaken, we will just close this' meeting.
I appreciate 4
everybody coming by and giving us reports on all these t
j 5
matters.
6 The meeting is adjourned.
7 (Whereupon, at 4:30 p.m.,
the meeting was 8
adjourned.)
9 i
10 11 12 13 14 i
15 4
16 17 18 19 20 21 4
22 2
23 4
24 5
25
- r--m
+e-r*
g.----yg%+*ww--g----.w.-epr -m-e. e. e w-e-y-yw w w g e%e, tem erw-.,w--w..=.m
>>.+g-r----+.+e, 3m. e ewem erem e eweg-
--e-e-N-
1 A
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
matter of ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8
9 Name of proceeding:
Subcommittee on Reactor Operations 10 11 Docket No.
12 place: Washington, D. C.
13 Date: Tuesday, July 9, 1985 14 15 were held as herein appears and that this is the original 16 transeript thereof for the file of the United States Nuclear 17 Regulatory Commission.
13
/
(Signature)
/j yf f-g g,
~
(Typed.Name of Reporter)
Mimie Meldzer 20 21 22
(
23 Ann Riley & Associates, Ltd.
\\
24 25
AGENDA FOR ACRS SUBCOMMITTEE MEETING
,q 1
ON REACTOR OPERATIONS JULY 9, 1985 1:00 P.M.
ROOM 1046, H STREET RECEllT SIGNIFICANT EVENTS Facility / Title Event Date Presenter / Phone kC 1.
Hatch 1 Stuck May 15, 1985 G. Rivenbark, NRR d-Open SRV.
(492-7136) 2.
Oyster Creek Scram June 12, 1985 D. Powell, IE 7
Discharge Volume (492-8373)
Isolation Valves Failure 3.
Westinghouse - Potential February 12, 1985 D. Powell, IE 9
seismic interaction of the I
incore flux mapping system 4.
Pilgrim - water hammer April 2, 1985 V. Hodge, IE
/]
in HPCI steam turbine May 18, 1985 (492-7275) exhaust line.
O V
5.
Calvert Cliff Units 1 and May 26, 1985 D. Jaffe, NRR
/7 2 emergency D/G inoperable.
(492-8140) 6.
Rancho Seco - EDG control June 1, 1985 S. Miner, NRR 33 circuit design error.
(492-8352) 7.
Rancho Seco - Reactor June 5, 1985 S. Miner, NRR
.2.4 Trip breaker test failure.
8.
Rancho Seco - RCS High June 23, 1985 S. Miner, NRR 1h Point Vent Leak
-Q 9.
Turkey Point Units 3/4 May 17, 1985 H. Bailey, IE f
Loss of offisite power (492-9006) due to fire offsite.
10.
Sequoyah Unit 2 May 22, 1985 E. Weiss, IE 31 Reactor Trip due to (492-9005)
Improper use of test instrument.
11.
Davis-Besse loss of June 9, 1985 E. Jordan, IE 3]
(492-4848)
A. DeAgazio, NRR (492-8945)
- The I.I.T. will discuss this event in detail at a later meeting.
HATCH UNIT 1 - STUCK OPEN SAFETY RELIEF VALVE OF MAY 15, 1985 (G, RIVENBARK)
O
~
A SYSTEMS INTERACTION EVENT UNIT 1 OPERATING AT FULL POWER CONTROL ROOM EMERGENCY VENTILATION SYSTEM CHARC0AL FILTER DELUGE VALVE ACTUATED WATER LEAKED THROUGH VENTILATION DUCTS INTO A HATCH UNIT 1 ANALOG TRANSMITTER TRIP SYSTEM (ATTS) INSTRUMENT PANEL CAUSING SRV TO OPEN REACTOR MANUALLY SCRAMMED FEEDWATER PUMP RECOVERS REACTOR WATER LEVEL SRV CLOSED - WITHOUT OPERATOR ACTION O-CAUSe LOSS OF INSTRUMENT WATER SUPPLY CAUSING DELUGE VALVE TO OPEN TOGETHER WITH PLUGGED DRAINS NOT SURE HOW WATER CAUSED THE SRV TO OPEN ACTION REPLACED ATTS POWER SUPPLY, CLEANED PLUGGED DRAINS AND INSPECTED DRAINS IN REDUNDANT FILTER UNIT l
LIbENSEEPROPOSESTOADDCLEAN0UTCHECKPROCEDURES 1
FOR PLENUMS AND THEIR DRAINS ORAB WILL DEVELOP TIA TO COORDINATE:
IE NOTICE
'n FURTHER INVESTIGATIVE EFFORTS GENERIC REVIEW l
4 0
o 4
g
'* t o
a V'a a t
.h s he 3
ha td t
e4 iW t l%
k a
s =5 g
%o q
/
>u t
3 4
~
{
i g
m h,
ek' O
mN/
Rh y
=
na t
-_e
~
e k
6 Nhg i
m, X
O
'3! ?
3 l
[ll.. af.:f t.44,*.*f-f ti=%;. i;Uti <c t.
m
- e i ma d A tl 'N
- i-
.n.
a i
a
'-Y.'.g
'+'s'**<^*n-*-'**
. b..
_3 4
i
- .
- i
.*t
,T, E.-
ig t
... -' -S F ,'~
m
- :; T* *.
s n.
1 z zg h ":4 l
Z
^
,+
E a.
a<N r_N
- i- - >i>'i>
5 - - e I
==g
+
w r'.s.2.= ~_--sBt:= h e. C l I
gy g
<s q
i i4 V
j.
I P
j.lt y @e _ x 3 m ].. -,.i...i. - -
l 1
wa s e s rcm=m.-m.m--m
,, i g
4 i '#
T]i I
j4
.tu 3 ;< a,,. r.,
a, r
_}
1 f
e u
a g._.-p E...,s e 4
-. ~w,_
1 o
i I
j g:7l p
m
,)
V cb.
l l
i gy e
s @ )I. 4 ep w
1 o
4 i!
'7
.J, q
e o
r:.pl.g /
,i g
c i
i t.
pk ~;j,[e,,
i i
I p!
--k l 1*.,,
N
,I3 i II
<7 N' ;* -fh Ni[.Sl l d j;-[.
,8 j
. Q.. H -
t;, ' l- -
P I
4.,,, -
y t-g gg,1..'
l.
7 d.
! l P
E 4 o
) '!T,i,,
Ir,
,,G,s i
. 1,.. _. -
3 s
i
- i-i-
=
.t il
.1.
i rt.1
., 2.,..= x
.s
.I i.
l i
i 3
I T
1 l
.j
) k.;
i
..p
.c
~
- h. '
I I
t /
4 l;4 >! I k.
d.J O
ilL ::
i i
i xa
.- + --- - ~
k)!,-
-4
- d; --
I
+
i rw
.. -$. "j@.,
- j
- :
ll p.h :__I t
J m
in i-.";,-i-d, e@k' y'
< m -.- i
.tr..
,@ ~~ A=R ~~ M=r _:, I r e.==-E's.r=%,,.. q) l3,
- ,. y i
J
~-
m,.
..g re-.
- 3 7,..
-,,, - m e.m m... y%pt
. p r-
.; q
=
i q
_- := _.. m _.l
.m muae..
A
- l
/
,. g^ ^ '.^
l
^ ^ ^ ^ ' ^.
- t flid i!!iro ill !):(
.! 6
'l,
bl l,$i_3 ! ',
k vr Fr "
i:
I. 6 O'
, ;. q..{j: Tri.-
i---
j' $:
.L3 ww.3 I
l b.
lifililli::i j
liia :
s v i j
'H.,',,
i.
j'j
_.5--f,.
I
/
I.8 lj i:f tjtlil?.:
i 1
.i r.I=,1.
,d.
.. i e u :,
r, l'Q si i
- .r-
. c.
i_
.a
.. ra
'l
...Nvirrre-g g-t-
-' ~
d ' 4 :!M H f.Il fi l, :
e n:..
- 1:
,,3 J
3
g g
b
,0
55 3
E4 b
5R e
p t:
a hlII
$$ '~I 1
-I i:' q. 3I l !;I ;!
'b li!
2 I
i f
14,
.{'
t :
g w
.h!
.i i
I!' a f.
Il 8; ; 'l Nl,N.
ill ill l
"I M
aw
,pd....
h g
a ei g g q
g.g,
y 4
'i V 8> ! V -
d.4,Li!.L min,6
!Djinv,..!*
l 8
a<
_,1C
.o d
- 6. I 1.
9 iP l'l
- -<EI *
',}!! ! !! {
r a..=..gFgy.. 4..e;-
?
.2:.
isu o.,w.K o,y:A 1
,,p,,i,;
n
-e 4 iI F,,, d 1
a ij{ !;II.i si 10
- g......<u.4, i
d;' Dg,t l ii,,nig H
p
...: =
'e M
i i
-i
..=; y i.rr nilllllilij4 h
n y;; ;i er-c+:<
ii s
i l,i.s g.e y. 4 k. 0. 0.. a n.. e< <
i c.
f i
i IL
< p.... g e, _
i 9
! 00:
i e
7-Ir m 3 "pr<....--, f.
\\
I f t fi * $ T i..'.dWI+'d, e d,: 4< -
ITW--n_e d M *6 ea
.3 !*' M -A h h.-
I'i I+
0 liR I
4 e
o r
+r,-
N
<v t r
pN; :/;y d v$ w "4..ni s
N,v bY@h, fs il Lit.....1.,..i:
p.. g
.:.1
....i.....b---t :"ig g.,,r......:..........g@
N' i,
..:.g,i i
1 I,
^
i 4,3 Q.
r~$p,>i *, i ';
4.
f A>
L:
35 ]
4!
c qca i.1
,, iW j ':i
..3 111.. ~....:.
g.. +...
.3 g
frig *./q..
..... i,
,. i 3
y 9
3 it' s.;..q)iB 1
d.),
I.Ohk h.l hi u
3
!'l i Tl
[
11 E G-
'd
. l ! 11. '4..
- -il p;'
c.
d i
iV -iY ry[i g
E,E- ! il T lH #
4 i 'iyl.Di I'
l.'
g
.i i
.: iD i I
I
./
t i1
,4 ;
ti
. 4, 1, 1, l.
j.
il li Y,i_
L. II In 11 I
y i41 r.
1 e
- I,l.,. 3
,t i
n ;) ;i il r
i.
.,f.ti ^i l n *n...'.;,, ig! -
>i i
i.i II.
n-t,1. i g
+a _ +.-
.....Il i a. i nfig. di
,~ ':.idi t
.A v
m-J 'll j
1 i...T..!!iM.>.....I.j...ji.il*Cdd Fi..$
i
.n..o ;
s
,. 6c
.a i.
r, g^
,l w.:.lg pj,,
il.,
. ;.i f,.:L '. I;; _ w,, in.:.@1..I. !a r w.3 g
t....../i$:. i 1
- u., ml I,.-C,lg'--
4 I
n.
i1 io t
a,. 1. -a e
I, m.
h'bb
. 311*>t...;'.fk 4-k lg 5
i
^
^I NS) a
.u-
'1 jS;.;
1-
.y
'y
[_,, -
- a no -dTQ-a.
h*"'I s-s Y S
r"
~
r i
p 3
M --
3 T*
! :. a. a!,
i 8
- f
- e i
C m
laI i
~
u.~
f I
/H i o
b 8
.. 0l lT M..gg _
2et: : 2*st i 2H11 2w:?
2>75 x75 H11 HI1 ;n !-
g y
C!
- p 3
P ;25 P327 PS21 P923 P200A P200A PS23 PS21 P92'.P1 2' g
t
{
s.
y-4...
t.
y -s-i 2 -s-I 2 -s-12 -s-t 2 -s-12 -s-I'.,2 -E 12 -s _12 -6;.7
- 7, 4
~
V,-.:t g
. uc. y i"
In'. s -o. r -o.2 -o.2 -o.1 -s-s -o-4 -s 4 -s ;,,,i
- g
')
- t
. g 4'
o i
g g
_ z_,
g.
- o I'
r
.z,, '... $
4
- I
- d 5y b
m i
sit I
3
.f l8 a!
. :r75?2x15 2x75 2r75 p33 2175 P605A PSOSA I.,
pf r
h,.g s..
- -i F
1 401 P502 P603 PGO4 'P603 l
PI O4 o
~t go ef.
~~
m".
e v
i
- )p. o
,=
.f.tf i.%,,,
aar w a Lsra.
l:
- s, r
l
- 81
- 2-ata cae 1
1
- *i ng
-p
=
/
/
-A '
_xS._
1-W,..
i i;.-
~-
-=
o
}
. h
< =,. -
/
t ir i
g l
e P33 x15 i
- o
- x. s i
J 2,.
i r75 ITS,x75i 175 Pine,,9 PIC4 2H1'
-il t r.'._.
-8 k-z i:.
..t l.,
o l
'9 a8 J
8 b31 P602aP6031P604 PECSB CGQ5h I
_,s.
i i
i.
g
., 1 I
n
~
4pA g.
a f
_ al. " [, --
~
I 42
- _* h.!
//
i E -o-4-s-
.c4 2 -oJ l2*-o-2 *"
2
.g r-t.t m,. _
e P
e r
. }f
-g Mi t-
~. ( *-:
g I,
g
- g Ag w-se,-
e 4:a ;;p -,
a L
- g a y"'
4 a
I.t_
E,'
,3, >Jgy 9
I
- f.. g l
4
- b, o*,,
l 2H11 2H11 2H11 2,41 :
2tf5
.175 Hts. H11 H11 N
~
l
_Z,1 d
3
'I f926 PS28 PS22 P9,24 P2000 P2006 0924i P922 P923lg 4
r-
- n
- f a-o.
M-s,r-s-l2 -c_ I 2 -s '. 2 -e 12 -s-[2 -s _1I -s i 2 -s-l 2 -ei -d 4
sj go.-
., g@
2 2
o
[
-1[.
z
.x
?
_ = > _
r s.f i.4 1G
)3"' "" q.
w/
--~'-~~~~~~~~-~~~~
7
. :?.%
,r
-s c t- -
)
/
Q
<//// 7/////L 7.'//yp,,,,.,,,,,,,,,/,
t
~
J 6
.a o
ra o
f 4
L./
A h
7}' Q I
l I'
il L
a a 3 -s-
.2 -5'!n1 3 -S - 1e,2 -4 : 4 -5 % -
.. -w- -)
7
-WI
~' '
'8 sta cuci hs.N! ROCl4
-[.tI l.Aottsnoou f y,
r4
/
//-
l i
- .is.1 el- -
3 k
8 Ta
+.s I
y 6
h~
'i i
t*3,-
.; p,.* a==== p,
-. T.*' *W.-
t-
~
itfWh %
?fh 1.*-
~
?N.
~
I
}4. A A/ V/f W pesc-r wogg /A TT.$
PAW S
-,, a.., < n - o, s. ~ A
(,
H
=-
_ ~ -
a OYSTER CREEK - UNCONTROLLED LEAKAGE OF REACTOR COOLANT O
OuTSiDe CONTAINMENT r
JUNE 12, 1985 (D. POWELL, IE) f e
i WITH REACTOR AT 997. POWER, FAILURE OF THE ELECTRIC PRESSURE REGULATOR CAUSED A TURBINE BYPASS VALVE TO OPEN RE3ULTING IN j
A REACTOR PRESSURE DECREASE, FOLLOWED BY MSIV CLOSURE AND REACTOR SCRAM.
l SCRAM DISCHARGE VOLUME DRAIN VALVES FAILED TO FULLY SHUT /
SEAT CAUSING REACTOR COOLANT TO BE DISCHARGED.TO THE 4
REACTOR BUILDING DRAIN TANK.
RELEASE OF' STEAM FROM FLOOR DRAINS AND' PAINT BLISTERING ON l
HOT PIPE Cl,USES, PORTION OF REACTOR BUILDING DELUGE SYSTEM l
TO ACTIVATE I
SCPAM S'IGNAL NOT RESET F'OR 36 MIN ALLOWING CONTINUOUS REACTOR COOLANT FLOW TO THE DRAIN TANK.
CAUSE WAS 600 PSI lHTERLOCK ON MSIV CLOSURE / LOSS OF CONDENSER VACUUM.
l SAFETY SIGNIFICANCE - (1) LOCK OUTSIDE CONTAINMENT, (2) POTENTIAL EQUIPMENT MALFUNCTION DUE TO FIRE DELUGE SYSTEM, (3) EXCESSIVE CRD SEAL, TEMPERATURES.
CAUSE-VA,LVE SPRING ON VALVE V15-134 (VELAN) VALVE UNDERS1 ZED l
-VALVE V15-121 (VALTAK) STROKE DISTANCE INSUFFICIENT TO TIGHTLY SEAT THE VALVE. (1/8" OPENING)
-!MPROPER POST-INSTALLATION TESTING OF VALVES "CdRRECTIVEACTIONS-REPLACED 400LBSPRINGWITH1100LB i
l SPRING, ADJUSTED VALVE STROKE DISTANCE CHECKED CRD SEALS FOR DAMAGE, CHECKED G
EQUIPMENT, NO DAMAGE FOUND.
(
NRC FOLLOWUP ACTION - IE NOTICE IN PREPARATION.
7 c
0684 oV i'
air I;
y
- h:
1 I-
.x no e
i 2
en
+f
\\
\\
l.
v h
\\'n i gir
(
M
/
Ti:
i scuc 3
=
,,,,c 1Qv::
7 lii
!r t y/
/
ENs
\\ :gg !f ll:
!.s
~
! !; !s t
1 Ei X
3 e
2 Ols
@X
.i e-H
- s 3-N 2.1 0
- c 3
x:
r
,3 i
i vg om o
x
.s m
w m
c g
4.'. s
!ll l
3 I
3 if I
j(g:!
is e
xx w
is
!s w
- i. !!-ii
!f
.i l.!
g-1i a
i;e x
i i
i::
Wir
)
r
=
d, i
f i
i il!
- --@X!!
i.
~
!.-_..!____..i'_.I_..i...I
+
n
~
t c
.y
.t
!:2 gi s.
is.11 @
I*
si!
g, il g!;; e
.a
(
N u,
O a
I 2
/
s
- [
- !fk
/'
~
5
'E Y
ei so 3
h 2.3-35
WESTINGHOUSE - POTENTIAL SEISMIC INTERACTION OF THE INCDREFLUXMAPPINGSYSTEM FEBRUARY 12, 1985 (D. POWELL, IE)
PROBLEM AT W FACILITIES - NON-SEISMIC PORTION OF THE FLUX MAPPING SYSTEM MAY INTERACT WITH INCORE GUIDE TUBES AT THE SEAL TABLE DURING SEISMIC EVENT.
SAFETY SIGNIFICANCE - POTENTIAL SB/IB-LOCA DUE T MULTIPLE FAILURES OF THE GUIDE TUBES AT THE SEAL TABLE.
PROBLEM DISCOVERED AT SHEARON HARRIS, NRC NOTIFIED OF POTENTIAL PR0BLEM ON JUNE 22, 1984..
FEBRUARY 12, 1985',' LICENSEE NOTIFIES NRC OF WESTINGHOUSE FINDINGS.VIA 10 CFR 21' LETTER.
MAY 30, 1985,- WESTINGHOUSE NOTIFIED LICENSEES OF " POTENTIAL UNREVIEWED SAFETY QUESTION."
CORRECTIVE ACTIONS - PERFORM STRUCTURAL INTEGRITY ANALYSIS 0FTHEPORTIONOFSYSTEMABOVESEALTAilLEJMAKESTRUCTURAL MODIFICATIONS AS REQUIRED.
GENERIC IMPLICATIONS - APPEARS TO AFFECT ONLY WESTINGHOUSE PLANTS ON PLANT-SPECIFIC BASIS.
DIABLO CANYON /SHEAP.ON HARRIS KNOWN TO HAVE POSSIBLE INTERACTION.
i NRC FOLLOWUP ACTION:
NRC; WESTINGHOUSE AND REGULATORY RESPONSE GROUP CONFERENCES) t IEINh0RMATIONNOTICE85-45ISSUEDONJUNE5,1985; STAFF CONSIDERING GENERIC LETTER.
f
STORAGE REEL A[
NELICAL-WRAP ORIVE CA8LE I
I ll DRIVE WNEEL DRIVE NOTOR
- 1) /
[/
- 4 5 FATM ROTART TRANSFER iaTERCONNECTfNG TUBING 7
10 PATH ROTARY TRANSFER e
'~
W l
WYE UNIT i
s
/;
i l
nicn PRESSURE stAL 180LATl0N VALVE
' 1
\\
MinlATURE stuTRON OETECTOR i
1
$EAL TABLE.J,.
Figure IV-2.4 In Core Drive System
\\
/0 N'2* D WNTO-7606 m
THE RMOCOUPLE CONDUlf SEAL AND N DISCONNECT PLUG PORT EXTENSION]
pmivE MECHANISM VERSEL MEAD PORT
/
\\
\\>
h l
PORT COLUMN g STANDOPPS 5 %
SOCKET PLANGE.
y
~*
PLATE 3EAL TABLE '
vEgSEL 3EAL LINE-s ll Mq s
-T, e
L
\\
THERMOCOUPLE CONOUlf N
otviCE b
- THOMSLE GUlbE AT COLUMN 8 ANO. MIMING m EwEm gm p
UNION y
Q l
Toe or Active PutL --
l g
PUEL ASSEMBLI-,
s Ill l
UPPEn T
S cone f
s
((]
O 4
s PLATE s/
s Ill l s 1
s i
s l ggg
+
s s
CORE SUPeOnT -
s l 111 l s o-COLUMN s
1 N'
ggy N
/'
w u
i
' l s
INSTRUMENT TM8MSLE =Nk LOWERl l
n u
I i mi.e Gulot mumwo
% CONE-
~
)
.Y PLATE:
I 3
/
, Al l
THIMsLE OulDE s,
/
VERSEL PENETRATION * -
j TuSE MOUNT 8
..\\
% I 7
Tust
/
\\ SUPPORT PLATE THIMELE GUIDE TUGE -
h POAGING TO VESSEL WELD JOINT n
I THOM5LE GUIDE
~~. Tvet mount e
l l
5 j'
THIMSLE GUIDE '
Tutt WELD UNION TYPICAL ARRANGEMENT FOR INCORE INSTRUMENTATION' FIGURE 17-2.3
].
IV-2.18
}INTO-7606
l
.pj m
7
-r.---
I7
/
.,l.
.p
,, V
'i
~
=
.t I
i
[-=
l
./
~
==
=
M L a, li'
=t
- .' i il NIIIIII l
Ir
~e 3
C C E
a t u g
l t n.c i t a%
4 i
i a' s s-s e
i u l-
'1
/
l H
i j
N,,
,\\_ s r
E
/
l i
si e C
.C. t m
i e
d Q,I f
E I.
'w 7
C l (
Gl. x i)
C 1
i c.:
.41
. '-'ai i
w-a c
.= y JI
/
O>
j 4
[*
gg e
- ni 9
i
~
i l
l l
.e.
s 5
.. _-=r i
,i gi s.
U.
, t
\\' N
)==
I 'Ju
.~.:ji n
(.!
..Il-t e >}
~...
l
'8 W&.a l
1 t
,gg g
==
sf2 e
1
.r 4,
E="
9.:
3
' :* I
'it!
-j lFf.
,3,,
IIe i
gt, (
g r,
l L*s. j 8
I l
t e
I N
.fi. M W.
-,~
/1
8 t
i R
$p i e o.
e j
3 3
[
I
[
IN l
HPCI TURBINE EXHAUST i
I I
AT l
I i
i PILGRIM l
1 i
l.
4 b
i l
BY l
1 VERN H0DGE, IE j
i I
1 f
k h
k i
I t
I h
f
?
m i
I I
L
/3 t
t
O HPCI EVENTS AT PILGRIM OVERSPEED TRIP (MARCH 31)
HIGH PRESSURE ALARM (APRIL 2)
DAMAGE TRIP, RAPID RESTART (MAY 18)
DAMAGE LQ l
i l
I i
O
/f
O HPCI EVENTS AT PILGRIM HIGH FLOW ISOLATION (MAY 23)
INSTRUMENTATION INSTALLED (MAY 214)
TRIP ON INTERFERENCE'(JUNE 6)
INDICATED IN0P RUPTURE DISK (JUNE 30)
~.
4
/[
7-.-.--__
m J
O O
.O HPCI TURBINE EXHAUST LINE DAMAGE AREA 1"
TORUS
~
- sgy
- ~ x
( ORA. oT )
=" age s
.n 1"
VACUUM w
RELIEF VALVE I
I RUPTURE. DISC HPd TURBINE l x l
GN i
l l
s O
O O
David H..Id fa, W R i
- (
CALVERT C_IFFS tRilTS 1 AND 2 s.
~
Ali DIFRFI GR4ERATORS (DGs) IN0PERABLE' ON FAY 26, 1935,,
1r
.~
JillIIIL COMDITIONS
'- ?-
x c.
~.' '
s
~
UNIT 1 Ill REFUELING OUTA3E o
o UNIT 2 AT FULL P0 tier o
UNIT 1 DG (#11) FAILS DURING BVERSPEED TEST ON i%V 10 1985
_ 1_
~
4 A
- THREE lgs FOR TWO UNITS M
~
O O
O DIAGNOSIS o
FAILURE OF DG #11 DUE TO FRACTURE OF All INTERPOLAR C0iMCTOR (CONNECTS AMORTISSEUR WINDINGS BETWEEN GENERATOR POLES) o FAILURE MODE IDENTIFICATION COPPLETED ON MAY 26, 1985 (CAUSE:
HIGH-STRESS INDUCED CRACKING)
\\
l
O O
O ANAL'fSIS OF EVENT GENERATOR TYP5i LBUIS ALLI TYPE TGZDJ,14083KVA -
o INTdRP0i.ARCONNECTORSREQUIREb:~
o 4
.o NiiEN DISSIMILAR DGs ARE IN USE -
~
+
.x.
-ro WHEN 10TAL'THREE PHASE UNBALANCED' LOAD. !S a-
, '1
/
EXCESSIVE-o LOUIS ALLIS'RECOMENDS REMOVAL OF INTERPOLAR CONNECTORS i
j (JUNE 5, 1985~, 10 CFR PART-21 SUBMITTAL FROM"COLTF i
v'
.r '-
- .s i
+.
1 4
4 N
4 i
l
O O
O i
R:5 IIIAL ACTION o
TWO REMAINING DGs (#12 AND #2D REMOVED FROM SERVICE TO INSPECT INTERPOLAR CONNECTORS ON MAY 26, 1985.
CRACK FOUND ON #21 DG.
I o
REFURBISHED #11 DG REALIGNED FOR UNIT 2 SERVICE.
o INTERPOLAR CONNECTORS REMOVED FROM #12 AND #21 DGs.
i i
o ALL DGs DECLARED OPERABLE BY MAY 27, 1985.
1 i
l l
t) b
O O
O STATUS OF SIMILAR INSTALL ATION FACILITY NO. OF 3Gs STATUS VERMONT YANKEE 2
CONNECTORS REMOVED.
TMI-l 2
CONNECTORS REMOVED.
PEACH BOTTOM I4 ONE DG INSPECTED AND C0t;NECTORS REMOVED.
REMAINING DGs TO BE INSPECTED IN JULY.
CALVERT CLIFFS 3
CONNECTORS REMOVED ON 2 DGs.
THIRD DG TO BE MODIFIED PRIOR TO UNIT 1 RESTART t
f s
RANCHO SEC0 - EDG CONTROL CIRCUIT DESIGN ERROR
()
JUNE 1, 1985 (S. MINER, NRR) 4 PLANT SHUTDOWN FOR REFUELING OUTAGE.
EDG CONTROL CIRCUIT CYCLES CONTINUOUSLY UPON RECEIPT OF LOSS OF OFFSITE POWER (UV) SIGNAL WHILE IN MAINTENANCE SHUTDOWN MODE.
MAINTENANCE SHUTDOWN MODE EDG OUTPUT BREAKER OPENS EDG IDLES DOWN FROM 900 TO 600 RPM EDG IDLES AT 600 RPM FOR 15 MINUTES EDG SHUTDOWN AND COASTS TO REST WHILE EDG IN THE MAINTENANCE SHUTDOWN MODE, EMERGENCY BUS WAS DE-ENERGIZED DUE TO PRE-PLANNED WORK ON NEW PARALLEL BUS.
UV SIGNAL BRINGS EDG UP TO 900 RPM l
EDG OUTPUT BREAKER CLOSES - UV SIGNAL DROPS OUT MAINTENANCE SHUTDOWN MODE DOES NOT DROP OUT FOR 30 SECONDS AND AUTOMATICALLY OPENS EDG OUTPUT BREAKER; UV SIGNAL RECURS EDG CONTROL CIRCUIT REPEATS CYCLE DESIGN DEFICIENCY HAS EXISTED FOR LIFE OF PLANT.
LICENSEE INSTALLING RELAY TO DE-ENERGIZE MAINTENANCE SHUTDOWN MODE ON UV SIGNAL.
( ))
POTENTIAL GENERIC PROBLEM IN ALL PLANTS WITH SIMILAR GM SETS.
IE INFORMATION NOTICE IN PREPARATION.
)L
O UV signal MINTENANCE SHUTDOWN MODE:
EDG RETJRNS TO EDG (lUTPUT BREAKER DOES NOT DROP OUT FOR 30 SEC OPERAT1 1G SPEED OPENS UV SI NAL RECURS EDG ID E DOWN BEGINS l
EDG OUTPUT BREAKER CLOSES 9
O d3
m(d RANCHO SEC0 - REACTOR TRIP BREAKER TEST FAILURE JUNE 5, 1985 (S. MINER, NRR)
RANCHO SECO HAS GE AK-2-25 REACTOR TRIP BREAKERS.
j DURING REFUELING OUTAGE, ALL RTBS WERE REFURBISHED IN ACCORDANCE WITH B&W OWNERS GROUP PROGRAM, PREPARED IN RESPONSE TO GL 83-28 0F JULY 8, 1983.
PRIOR TO PLANT STARTUP, ONE REACTOR TRIP BREAKER (RTB)
FAILED DURING POST-MAINTENANCE OPERABILITY TESTING.
ALL RTBS HAD BEEN REFURBISHED BY GE-ATLANTA AND CERTIFIED BY B&W-LYNCHBURG.
l UNDERVOLTAGE TRIP PADDLE JAMMED AGAINST ARMATURE WITH ARMATURE IN ENERGIZED POSITION (SHUNT TRIP REMAINED OPERABLE).
LICENSEE EVALUATION INDICATES UV TRIP ASSEMBLY ARMATURE / ROLLER-RIVET MEASUREMENT GROSSLY OUT OF l
SPECIFICATION.
1 ALL RTBS REINSTALLED AFTER PASSING REVESED POST-MAINTENANCE l
TEST PROCEDURES.
l FAILED BREAKER TO BE EVALUATED BY B&W 0WNERS GROUP.
IE INFORMATION NOTICE IN PREPARATION.
~
k
O
,, FIGURE 1 1
I
- s i
TRIP SHAFT
.v 4
F
\\-
unDERvCLTAGE TRIP PADDLE A
e b
W
\\\\\\\\\\\\\\\\\\
g(\\
- RotLEA-AIVET er, v%,.
f F
l 3
e-FRAME v
l h=ap
[
4 a
A
@)
ictu n p
/
o UNDERVOLTAGE TRIP OEVICE O
cat' oe-easaar2to 2
db j
RANCHO SEC0 - RCS HIGH POINT VENT LEAK
()
JUNE 23, 1985 (S. MINER, NRR)
PLANT IN HOT STANDBY RESTARTING FROM A REFUELING OUTAGE 20 GPM NON-ISOLATABLE PRIMARY COOLANT LEAK ON HIGH POINT VENT ON B STEAM GENERATOR HOT LEG TMI MODIFICATION INSTALLED 1983 REFUELING OUTAGE 120* THRU WALL LEAK AT WELD CAUSE APPEARS TO BE MISSING SUPPORTS AND FATIGUE FAILURE LICENSEE ACTIONS:
STRESS ANALYSIS TO IDENTIFY OVERSTRESSED AREAS (BOTH HOT LEG VENTS)
REPAIR SYSTEMS INSTALL SUPPORTS WALKDOWN TO INSPECT AND EVALUATE OTHER SYSTEMS REGION V, IE TEAM PARTICIPATING IN WALKDOWN.
26
O FROM Nr. SU PPLY)
TO MANUAL VENT w-D 4u h 'TO HIGH-POINT
%/ ELD DETAI L VENT VALVE 6 a
h i
c T_.. i m,,.l) f
^
- M1.
f^ _
^
__;)
/
~
SCALE : FULL g
N l
/ GTEAM GENERATOR E-205 5 SCALE: I"=l' a'7
(~')
TURKEY POINT-FOREST FIRES LEAD TO LOSS OF 0FFSITE POWER v
MAY 17, 1985 (HENRY BAILEY, IE)
PLANT STATUS PRIOR TO EVENT:
UNIT 3 FUEL UNLOADED, UNIT 4 AT 100% POWER CAUSE:
FIRE SHORTED OUT THREE 500KV TRANSMISSION LINES, THE LOSS OF THE 500KV CAUSED SOUTHEAST FLORIDA TO ELECTRICALLY ISOLATE FROM THE GRID, V0LTAGE IMMEDIATELY DROPPED IN THE ISOLATED AREA AND CAUSED THE LOSS OF 0FFSITE POWER, SAFETY SIGNIFICANCE:
THERE IS N0 SPECIAL SAFETY SIGNIFICANCE TO THIS PARTICULAR EVENT, OTHER THAN THE CHALLENGE TO THE SAFETY SYSTEMS ASSOCIATED WITH A LOSS OF 0FFSITE POWER AND REACTOR TRIP.
THIS EVENT IS PRESENTED DUE TO PREVIOUS CONCERNS ON THE LOSS OF 0FFSITE POWER EVENT AT
(}
TURKEY POINT.
EVFNTS AT TilRL'EY POINT:
0 LOSS OF 0FFSITE POWER DE-ENERGIZED THE UNIT 3C TRANS-FORMER WHICH FEEDS UNIT 4C 4160 BUS, O
LOSS OF POWER TO 4C BUS TRIPPED THE 4B MAIN FEEDWATER PUMP, WHICH INITIATED A TURBINE RUNBACK t 0
RE. ACTOR TRIPPED AT 11:47 AM ON STEAM FLOW / FEED FLOW MISMATCH WITH COINCIDENT LOW SG LEVEL.
PLANT RESPONSE TO THE TRIP WAS NORMAL, NATURAL CIRCULATION WAS ESTAB-LISHED.
THE EMERGENCY DIESEL GENERATORS AUTOMATICALLY STARTED AND PROVIDED POWER.
O
- 12
0 FOSSIL FUEL UNITS 1 AND 2 ALSO TRIPPED UPON LOSS OF 0
Ores 11e e0Wea.
ONe e0SSIL UNIT WAS ReSTAareD AND SUPPLIED POWER TO UNITS 3 AND 4 AT 13:52.
0 A RCP WAS RESTARTED AT 14:01 AND NORMAL HOT SHUTDOWN CONDITIONS ESTABLISHED, 4
k a
h 4
9
)
i: O 4
7h
\\ ~.~ GE0RGl A
.f
., Q c
.._.._,--_.~*-7.%,,
g
\\
- m.,
(
j
- '"a=
./'
.),,,,,, 9, q
3 j
j.,,,,,
rn N.,.,.,s,'*==*~==s'
,i
.n.
.u p
w =.
"K.
g s*
- - - ~. - _, _ _
p.--
--)
a
'i g
.g, w,
=_
, g f-a==<<
l,,,,,,,, \\.,
- ' f d
d""
,===
e g
.- C L-- ~.~. N; g s'5.,T m ::=.
=
s m
o J
T'
- w..o s
cb,/
j-'
)
j
.m u-y v.
i i
i gU
'-- g'b[/ *~ N"k.$
m.'
~
~
4g o
':.. ' =w.m. p.-
c
~w"
==
x i * **" h O
Q 7
, P(
=
.,,N i
\\
=
I N
- /
r-
/
- _.)
f "E"
.i s,
r-
,~
i'f*... ~,
igr r.:=
i=
\\
' t
_g" M
,'Q
\\
i e**
.i j
(
. h ~ aL**
-~
- \\
l
.T
\\
i=
s A
p-y-.
.,L.-. 4.,
]'
('
"*"' i ~ e.'
Y.4%
"==
".".T' d"J "".*",.
\\
1",*i'
"",'""E i
,m
" 'T
- E**-,,.
N S'"*
~'
,,;,g 2
W,... _...
g'a "" ~
'"=3 a
." q ~s
=
.p
.. u u a,"'"'" M.3-T.
1
'E.
agg m e<.
.our T.
(".,(
. '. " -'PJ/,.3 7
\\. g=, g.j.
- " p*-... -. __ rni m.
a Jrgy p
g:.;
i O
i O
E, 'u a
%g%'s,'
- C:.
- '*w
.9 er h
i J:. LL
%s l
Ltstmo
(
I p
,=
m'"
~
8 {=
- ,, -m aa.
ihmsu ssio.e U.ts 6
- w:
.u. -
tg-a.-
m
=-
_a m,,,.cas M
/ ay.
g t.g u.
==.8 e.:.s q
~q],p;.y or.ititaineo. tuts a suesvaroes o.
-. ~..,,
o%#:
3 o.
-w uw y
.~
/
D.
ens k
ist s s.e,
r
- e. s cs,,
240 kV TRANSMISSION NETWORK v
p
.....m p..,
FIG,UR,E 8 2-1
.s
. e d
V
O O
O
~
..... s
,.g....
g.,.. -..,
un
.. y.
g..
. y.
.y.
.il I
/-oil
/--o il
/-o li
/-o
/-o ]?
/-o!! l
/-o 1
I I
.t 8
I J
l 1103
.il
?Ab Nab silh dll3
,il.LJ
- of b
O ;'
'Mg APy R
g "il g
g m
O d
!3 t]
u i
...... s.,
k i
6'*
- i sos tie e A 51 6'..
=...
ElVT.
=....
ulu-.....,
. u ul
.a r l u r t; 1
aut m;n '
m:.tr.
tJ::'or4
'Kan "T" ~
0;' '.'T*
tII:M
'f;mt. '
lc.D..".
c"a'r", ra "r, r"a","
r a 'r,i-e ra ry l
'., ' ^
~
r r,r, i
i I
j j
(!
l sn <,
uh 4 i, A_ e, m ',
c, s e, <, <,.
u i, i,
e MAIN ONE LINE DIAGRAM-TURKEY POINT PRESENT 1983-1984 'C' BUS ADDITION s
W*
5 SEQUOYAH 2 - IMPROPER USE OF TEST EQUIPMENT AND REACTOR TRIP MAY 22, 1985 (E. WEISS, IE)
REACTOR TRIPPED FROM 100% POWER ON OVERPOWER DELTA T AND TCOLD DIGITAL VOLTMETER (DVM) USED TO MEASURE TH0T LEADS FROM.DVM INCORRECTLY CONNECTED TO ONE RPS CABINET BEFORE TIME CONSTANT RECOVERS, LEADS FROM DVM INCORRECTLY CONNECTED TO 2ND RPS CABINET em O
O e
O n-,
,.-,,--,n
..--~,.,.
DAVIS-BESSE - LOSS OF ALL MAIN FEEDWATER AND AUXILIARY FEEDWATER-JUNE 9, 1985 (A. DEAGAZIO, NRR)
Loss of one main feedwater pump at 90% power.
Reactor trip at 78% power on high pressure.
Both MSIVs close spuriously tripping remaining main feedwater pump.
Steam generator low level starts both auxiliary feed'iater pumps but both trip on overspeed.
Operator erroneously manually trips SFRCS on low pressure No feedwater available for about eight minutes and steam O
generator levels fell to about eight inches.
pnRV cyclot three times - did not reseat on third cycle, operators close block valve.
Startup feedwater pump used to feed one steam gener*& tor.
Operators restart auxiliary feedwater pumps and restore normal post-trip conditions.
J t
No indication that subcooling margin was lost or that reactor coolant activity was abnormal.
Plant now in cold shutdown.
O 33
KNOWN EQUIPMENT FAILURES OR MALFUNCTIONS Main feedwater trip Spurious half trip of SFRCS MSIV closures (2)
AFW pump trips on overspeed (2)
AFW isolation valves fail to open (2)
PORY failure to reseat SUFP valve failure to open AFW speed governor after r.eset Switchover to service water backup supply Damaged turbine bypass valve O
O 29
D, O
b**
De..
C..Jem.'t.
~
ce,;ts, 5sh 3
s,.,%r %%\\
Fe alw.ie e-2 2
W Ve.ters ~
OT5Gr
- i ar..i
->d M
pg w:.
Maia y
- c. w,.1 S'*P Ve t.e.
% '* L-i g
c r., Wa. b~ia Feta g 54c /Fet/. Ps.7s h
.1-E O
sI,' v.1,s al v.. t V*
ov s t,
=
j Y
q-6
~
tv. 1.
r ti w tar e
N
(
N'* 5
@-X St..v.,, C..et,.)
Fe o.a D e. -
V, t,4.
--7
- Ated h y
(oJa.414.
5y14 i
i l
t..".It ekit, X
g, 3 EMI Z
57,,, (
T Z
X-B 7vas,94 Daiwd l
i
%y, htsr Deds.
A T V ?"*
E 5 t.,T.,p b y
u,,
pycm cowsn,shk t
~
s toues (d
t
SSINS No.:
6855 IN 85-50 UNITED STATES NUCLEAR REGULATORY COMMISSION O
OFFICE OF INSPECTION ALD ENFORCEMENT j
WASHINGTON, D.C.
20555 July 8, 1985 IE INFOP.MATION NOTICE NO. 85-50:
COMPLETE LOSS OF MAIN AND AUXILIARY FEEDWATER l
AT A PWR DESIGNED BY BABC0CK & WILC0X ADDRESSEES:
.All nuclear power facilities holding an operating license (OL):Cr construction permit (CP).
Purpose:
This information notice is being provided to inform licensees of a significant reactor operating event involving the loss of main and auxiliary feedwater at a pressurized water reactor.
Information in this n6tice is preliminary and was obtained from the special NRC fact finding team which is investigating the event.
A comphte report of findings will form the basis for further communi-cations or actions related to this event.
The NRC expects that recipients will review this notice for applicability to their facilities.
Suggestions O
contained in this notice.do not con.stitute NRC requirements; therefore, no specific action or written response is required.
Description of Circumstances:
On June 9, 1985, the Davis-Besse plant was operating at 90% power with Main Feedwater Pump 2 in manual control becaust. problems in automatic had been experienced.
A control problem with Main Feedwater Pump 1 occurred, and it tripped on overspeed.
Reactor runback at 50% per minute toward 55% power was
~
automatically initiated.
Nevertheless, 30 seconds later, the reactor tripped
^
at 80% power on high pressure in the reactor coolant system.
One second after reactor / turbine trip, one i:hannel of the Steam and Feedwater Rupture Control System (SFRCS) was automatically initiated due to 'a' spurious signal indicating low water level in Steam Generator 2.
Both Main Steam q
Isolation Valves (MSIVs) closed. Three ' seconds after the actuation, the SFRCS' automatically reset.
Closing of the MSIVs isolated the turbine of the operating main feedwater pump from its source of steam.
The pump continued to supply
'feedwater to the ster.m generators for a few minutes as it coasted down.
Four and.a half minutas after reactor trip, water level in the steam generators began 'to fall from the normal post-trip level which is 35 inches.
After MSIV closure, steam release to atmosphere continued to remove decay heat. One minute later, Channel 1 of SFRCS actuated when the water level in Steam Generator 1 actually reached the SFRCS setpoint at 27 inches (See Figure 1).
SFRCS started Auxiliary Feedwater Pump 1 and initiated alignment of it to Steam Generator 1.
.G 36
IN 85-50 July 8, 1985 j
Page 2 of 4 Within seconds after automatic initiation of Channel 1 of SFRCS, the operator actuated both channels of SFRCS; however, he inadvertently actuated both SFRCS.
I channels on low steam pressure instead of low water level. When an SFRCS channel is actuated on low steam pressure, a rupture of the steam line asso'ciated with that channel is presumed td have occurred.
The SFRCS closes the steam generator isolation valves, including a valve in the auxiliary feedwater line, 4
and aligns the auxiliary feedwater pump to the other steam generator.
Because both channels had been manually actuated on low steam pressure, both steam generators were isolated from both auxiliary feedwater pumps.
Five seconds after the operator's inadvertent actuation of both channels on low steam pressure, SFRCS Channel 2 received an actual low water level actuation signal.
Because low pressure initiation takes precedence, alignment of the auxiliary
. feedwater pumps remained unchanged.. At six minutes into the event as both auxiliary feedwater pumps were accelerating, they tripped on overspeed.
i In summary, all main feedwater had been lost, both steam generators were isolated from feedwater and were boiling dry, all auxiliary feedwater pumps were tripped, pressure of the reactor coolant system was rising, and reactor coolant,. system 1
temperature was increasing.
Within one minute after the operator's inadvertent actuation of the SFRCS on 4
low steam presinare, the mistake had been recognized and the SFRCS had been
. reset.
If equipment had performed in accordance with system design requirements, I
the operator's error might not have had a significant impact on the event.
O' The auxiliary feedwater isolation valves should have reopened automatically, but the valves did not reopen.
The operator then tried to reopen the valves 4
l from the main control panel, but the valves would not reopen.
Operators were dispatched to locally start the auxiliary feedwater pumps, open the auxiliary l
feedwater isolation valves, start the nonsafety-related motor-driven startup feedwater pump, and valve it to the system.
Pres.sure and temperature in the reactor coclant system continued to rise because there was not sufficient water in the steam generators to provide an adequate heat sink.
At 13 minutes after reactor trip, reactor coolant system
-l l
pressure reached 2425 psig, and the Pilot Operated Relief Valve (PORV) opened three times to limit the pressure rise. Ort the third lift, the valve remained open.
The operator closed the PORV block. valve and reopened it two minutes later after the.PORV had closed.
Approximately16to18minutesafterreactortrip,theoperatorshadthestariup and auxiliary feedwater pumps running and the valves aligned.
Water levels were
.beginning to rise in the steam generators.
Reactor coolant temperature reached t
a maximum of 594* F and then started to decrease to normal.
Refilling of the.
steam generators caused the reactor coolant system to fall to 1716 psig and about 540*F before, returning to normal (See Figure 2).
j At 30 minutes after reactor trip, plant conditions were essentially stable.
O 3
I
'l 37
IN 85-50 1
July 8, 1985 Page 3 of 4 1
' d Discussion:
I For several minutes after reactor trip, the steam generators were unable to cool the reactor coolant system adequately.
4 l
The first problem contributing to this event was the loss of all main feedwater due to closure of the MSIVs. The licensee's hypothesis, based on information from Babccck & Wilcox, is that turbine trip caused a pressure transient upstream from the turbine stop valves which caused the outputs of the redundant steam generator level instrumentation channels to oscillate widely for several seconds.
The licensee believes that this caused a spurious low level actuation i
of SFRCS which closed the MSIVs.
Three additional problems contributed to this event by affecting the availability of both trains of the auxiliary feedwater system.
The first occurred when the i
reactor operator pressed the wrong SFRCS buttons.
The second occurred when both auxiliary feedwater pumps tripped on overspeed. The thirrd occurred when both auxiliary feedwater isolation valves did not reopen when SFRCS was reset.
Control buttons for the SFRCS are arranged in two vertical columns.
Each column of buttons controls one SFRCS channel.
The operator should have pressed the fourth but, ton from the top in each column.
Instead, the operator pressed the top buttons causing isolation of both steam generators.
1 Both auxiliary feedwater pumps are driven by Terry turbines which tripped on overspeed early in the event.
When this occurred, steam was being supplied to the turbines via crossover lines, which are longer than the normal supply lines and include long horizo'ntal runs.
The licensee believes that significant condensation may have occurred in the crossover lines.
Further, the licensee believes that the quality of steam arriving at the turbines may have been affected significantly by the configuration of the crossover lines and may have caused the overspeed trips.
T:a auxiliary feedwater system isolation valves have Limitorque motor operators.
1r.c motor operators have torque switches which prevent overtorquing of the valves by disconnecting power to the motors,. When the valves are being opened, additional torque is required to overcome friction while the gates are being unseated and while a significant pressure differential may exist ac.ross the gates. During the initial part of the opening stroke, the torque switch in the '
motor operator is bypassed by a bypass switch so that full motor torque is' developed if necessary.
The licensee believes that these bypass switches went off bypass too early.
The valves did not reopen until an operator unseated them by hand.
~
!O s..
l i
_______.___i,____.__
IN 85-50 July 8, 1985 Page 4 of 4 No specific action or written response is required by this information notice.
If you have any questions about this matter, please contact the Regional Administrator of the appropriate NRC regional office or this office.
ward L Jordan, Director Divisior of Emergency Preparedness and Engineering Response Office of Inspection and Enforcement Technical
Contact:
R. W. Woodruff, IE (301) 492-4507 Attachments:
1.
Figure 1 - Steam Generator 1 Level and Pre'ssure 2.
Figure 2 - RCS Temperature and Pressure 3.
List of Recently Issued IE Information Notices-.
O i
en l
l
~
O O
l s
27 IN 85-50 LB83
$3 1 SU RANGE LYL. S$3 m o July 8, 1985 O
2.5" SO 7f tod 7:ng 150
/75 2co 1.1r y,
(s)
.i s.
i P932 SG 1 OUT Siri PRESS.PT1282 PSIR 600 6.10 70 0 7So' goo SSb 900 96D
/000
/aS*o
//co 8
~~-i M. s
%i~i i.
I i
e i
e i
e l
l l
5 i
e...(
i i
i i
l C !
.I
[
.'W f,
i i
i 1._..
l 1.__!
I i
1 I
i 1
I I
I I
I I
I I
I I
I l
1 I
i l
t l
- I i
I i
i I
h.
I
!.. -2 g
p.
1 J.
1 _.p p___.
l l
l l
l l
I m
t i
e d
i
.i t
g 4
L l
l l
N
~
._a_.._
I
_.a g
l l
W
.! - i
.i l.i
.I I
8
.t-._,1 !4_.!.
g i
i e
i i
e i
i i
i g
i.
i, I
.i.
l i
1 i.
i i.
i i
l
- u,
i l
I f
j i
i i
b4 i
.i i
i i
I 8
I
.i
.t..-
.I t
^
2
==
g i
j
!hl t
i I
4 i
l g
"'at l
5 f
i I
I
..i
,i
.e.
. _. i
.i is t
i i
a
{
i i
i j
j:
i
'j l
1_
I
?
. _p i
p
.... ___.7 g.
. _.. 7 _.i.
.i___ 7._ _...._ 7
.i i
7... _)...
I i
i t
i I
l l
l i
l t
i 1,
I l
t t
i i
i i
I I
l
'i i
i i
.._1 i _.i _.
y,
~
i 3
.' i i
i 4 T-i, I
i.
l 1
i i,
i t
n.
I I
i i
i m
T (G
FIGURE I: STEAM GENERtJoR I t EVEL AND PRESSURE
. f IN 85-50 July 8,1985 i
~
i s
(
)
P725 RC LOOP 1 HLG VR PRESS.SFRS CH 3 PSIR
~,
/YOO
/500 ffed f) tie
/ foe fyon M
.3/00
.3 M ATOC 40 i
i i
T7ES RC BYG NR TEnP ',
F S70 580.
syo 600 610 610 SM C30 SYo SSo
&so l
f 1
i
?
I t
e l
l l
l
!l l-I o-
--4 l
l l
I I
i i
j l^
).
w_
l I
I i
- -i l
l l
1 i'
1 I
I l
[I.
.-_.___t--
l m-l
)
(,k !
l l
i i
i
~
I I
I' g
4
~l I N _!
I
!! l e
i I
I I
t
~
g 53
.!____!____.t I
i 1
i 5
P-l l
I l
l 1
I I
l l
l 1
J' 1
Q l
i i
. __1._4 g
Q-i l
i j
i[
i i
i i
i N
~
l
- i l
I i
i i
e-i imi
.4 i
.o i
i.i i
l
.-t !,,
F
~
r, sr.
z 9
i i
i I
I i
d i
i i
j i
l.
I, i,
i t
i i
i i
i i
i i
i l,
i i
I i
i j
t l
l i
l Q-1-
l
,q 6
i 1
i i
_!i_ _ : -
-.i i
i I
e i
h
.I
.i
. _.i_ _.i
- - ' ~ ~ ~
._7-. T l
I n
9-b)
! !N!
(
I7 ;
.C
/
\\
[
a n
t il f~1GURE 2: RC.S TEMPERAT URE AND PREESVRE.
m_ _ _ _ _ _
i Attachment'3 IN 85-50 July 8, 1985 i
I LIST OF RECENTLY ISSUED IE INFORMATION NOTICES Information Date of Notice No.
Subject Issue Issued to 85-49 Relay Calibration Problem 7/1/85 All power reactor facilities holding an OL or CP 85-48 Respirator Users Notice:
6/19/85 All power reactor' Defective Self-Contained facilities holding Breathing Apparatus Air an,0L or CP, research, Cylinders and test reactor, fuel cycle and Priority 1 material licensees 85-47 Potential Effect Of Line-6/18/85 All power.. reactor Induced Vibration On Certain facilities holding Target Rock Solenoid-Operated
_ Valves 85-46 Clarification Of Several 6/10/85 All power reactor Aspects Of Removable Radio-facilities holding O
. active Surface Contamination an OL Limits F,or Transport Packages 85-45 Potential Seismic Interaction 6/6/85 All power reactor inuniving The Movable In-Core facilities holding Flux Mapping System Used In -
an OL or CP Westinghouse Designed Plants 85-44 Emergency Communication 5/30/85 All power reactor facilities holding System Monthly Test an OL 85-43 Radiography Events At Power 5/30/85 All power reactor Reactors facilities holding,
85-42 Loose Phosphor In Panasonic 5/29/85 All power reactor 800 Series Badge Thermo-facilities holding luminescent Dosimeter (TLD) an OL or CP Elements e
OL = Operating License CP = Constructi. n Permit 4,1
.