ML20211C731
| ML20211C731 | |
| Person / Time | |
|---|---|
| Issue date: | 06/06/1986 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1520, NUDOCS 8606120377 | |
| Download: ML20211C731 (106) | |
Text
.
OlGINAL 0
UNITED STATES NUCLEAR REGULATORY COMMISSION IN THE MATTER OF:
DOCKET NO:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 314T!! GENERAL MEETING
- P ;
.i:: r:1 Is.
.1 ! !. -
u.,,-
')
- s p, -
- ;. -i!
';1. d. :
s i :. u._ i
'1.+.
LOCATION:
WAS!!INGTON, D. C.
PAGES:
1 - 92 DATE:
FRIDAY, JUNE 6, 1986
?
KCR80FRCECOPY
?oKlemove ' rom ACRS Tice gk ACE-FEDERAL REPORTERS, INC.
Official Reporters 0
444 North Ca itol Street Washington,.C.20001 (202) 347 3700 61 7 060606 NADONWIDE COVERACE T-1520 PDR
CR27106.0 1
REE/0jg 1
UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION 3
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
5 314T!! GENERAL MEETING 0
Nuclear Regulatory Commission Room 1046 7
1717 !! Stroot, N.W.
Washington, D. C.
l 9
Friday, Juno 6, 1986 10 l
The 314th General Moeting reconvend'at 1:55 p.m.,
11 Mr. David A. Ward, chairman, presiding.
12 Q
13 ACRS MEMBERS PRESENT:
C' 14 N!R. DAVID A. WARD l
MR. JEESE C.
EBERSOLE 15 l DR. MAX W.
CARBON DR. WILLIAM KERR 16 DR. IIAROLD W.
LEWIS DR. J. CARSON MARK 17 l MR. CARLYLE MICl!ELSON l
DR. DADE W. MOELLER 18 j DR. DAVID OKRENT MR. GLENN A.
REED 19 DR. PAUL G.
SHEWMON MR. Cl!ARLES J. WYLIE 1
20 '
21 22 '
23 24 nU 2s ACE. FEDERAL REPORTERS, INC.
202 347 3700 Nationwnie Covernse 8%33N
l e
PUBLIC NOTICE BY THE i
l UNITED STATES NUCLEAR REGULATORY COMMISSIONERS' ADVISORY COMMITTEE ON REACTOR SAFEGUARDS FRIDAY, JUNE 6, 1986 The contents of this stenographic transcript of the proceedings of the United States Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards (ACRS), as reported herein, is an uncorrected record of the discussions recorded at the meeting held on the above date.
No member of the ACRS Staff and no participant at O
- at eias cc 9==
av r 9aa tattier car rr=====
inaccuracies of statement or data contained in this transcript.
O
27072.0 2
l REE 1
P, R O,C E E D I,N G S 2
MR. WARD:
Come to order again.
The first topic 3
for this af ternoon is discussion of recent operating 4
experiences at nuclear plants.
5 (Discussion off the record.)
l l
6 MR. WARD:
Mr. Ebersole.
l 7
MR. EBERSOLE:
We held a meeting June 3rd 8
involving the reporting on the operating reactors and 9
selected the most significant ones, and out of those, we 10 picked four which we are going to discuss -- three we are 11 going to discuss today.
Before we get started on those, 12 however, we are going to mention an event which we did not 4
13 discuss today which is an update.
Just to expedite this, I
' ()
I l
will just get it started as fast as possible.
Mr. Dennis 14 15 Allison is going to handle that.
16 I remind the Committee that there may or may not 17 be a letter which we should write.
This has been selected 18 as being of the greatest interest and significance to the 19 full Committee, and some of us are quite -- well, I think l
20 the one about the minimum flow logic problems that disable l
21 flow.
I will give it to you, Dennis.
And you can take 22 over.
l 23 MR. ALLISON:
The first thing is to give you a l
l 24 very short briefing on an event that happened at LaSalle l
l 25 last Sunday.
The plant was operating about 93 percent 1
ACE FEDERAL REPORTERS, INC, 202 347 3700 Nationwide Coverage 800 336-66W
~
/
27072.0 3
REE
()
1 power and had a feedwater transient that brought the 2
reactor water level down very close to the trip set point 3
of 12.5 inches, instrument level.
And at that point, the 4
operator turned it around and the level went back up.
5 MR. MICHELSON:
How far above the core is 12.5 6
inche s?
7 MR. ALLISON:
155 inches.
s 8
The zero is about 155 inches.
I am sorry.
It s
9 was a long way from any problem of that type.
10 The operator turned it_ around and in the process, 11 one of the four RPS channels indicated a low level and 12 tripped and gave a half scram.
The other three channels 13 did not.
Apparently at the time, the operator thought he O
14 had made it through the transient.
15 Later in the day, as other people at 16 Commonwealth Edison were reviewing the records of the 17 transient, they Cegan to think that it looked like the 18 level had gone down to about 6 inches.
So all four 1,
19 switches should have tripped.
Finally it was concluded 20 that the reactor protection system may have malfunctioned 21*
and declared an alert, shut the plant down slowly.
22 We have an investigation team at the site now 23 and they and the lico'nsee are looking at the event.
24 Testing so-far indicates that the pressure switches that 25 give the low level trip signal vary a lot.
Their se t
~ -
ACE-FEDERAL REPORTERS, INC.
202-347.3700 Nationwide Coverage 800-336-6646
)
27072.0 4
REE
(])
1 points vary by a few inches.
So that it looks like that is 2
what caused the lack of a trip.
3 The signals are going through the logic all 4
r ig ht, so it doesn't look like there is a problem there.
5 Those switches are made by a company called Static 0 Ring j
6 and we have had problems with the set points varying at 7
Oyster Creek on the same kind of switches.
We are now 8
looking at the factory experience.
There is similar l
9 expe rience there.
10 So those switches are drifting around a bit.
11 Not nearly enough, from anything we have seen so far, to 12 cause a real safety problem in this particular application.
13 MR. MICHELSON:
They are drifting?
It must have
('-#
14 been downward because you missed the transient.
15 MR. ALLISON:
They seem to be going in the wrong 16 direction.
I think it is partly, it is probably a time i
17 response thing, if you do it fast.
The level was only 18 under 1.5 inches for about two seconds in this transient at 19 LaSalle.
So the time delay tends to enter in there, too.
20 MR. MICHELSON:
They don't have hydraulic 21 snubbers in their instrument line, do they?
22 MR. ALLISON:
I couldn't really tell you for 23 sure.
24 MR. MICHELSON:
Some people had problems with 25 the pressure variations, and the snubber is to slow down O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 NationwiO Coverage 830 336-664i
27072.0 5
REE
()
I the response?
2 MR. EBERSOLE:
If the trip sensors flunked out, 3
what would be the next pickup of a trip signal?
4 MR. ALLISON:
The tests indicate they would trip 5
low and not flunk.
The next would be the double low level 6
and then the triple low.
Similar switches in those 1
7 applications at this plant.
8 MR. KERR:
What is the significance of the 9
inspection team?
10 MR. ALLISON:
Well, we think it is something 11 that needs to be looked at more than usual.
12 MR. KERR:
You think it really wasn' t because of 13 the failed switches or you just want to make sure it was 14 because of the failed switches or what?
15 l
MR. ALLISON:
The augmented inspection team 16 really just got there to find out what happened.
17 MR. JORDAN:
Dr. Kerr, the augmented inspection 18 tedm is to assist the Region.
It is a little bit of a 19 formalization in terms of having headquarters' expertise 20 j
assisting the Region.
They have a confirmation of action 21 letter to execute with the licensee which would cause them 22 to freeze equipment so that they are able to obtain the 23 neceseary information and a formality in the reporting, 24 feeding back of information.
25 Their course in this case is a question as to
(r v
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33 6 6646
27072.0 6
REE
()
I whe ther we ought to be taking some generic action on Static 2
O Ring switches.
We don' t have the information yet that 3
tells us that we should, to cause other licensees to do e plab to issue an 4
anything other than to be aware of it.
W 5
information notice promptly communicating what we know at 6
this point, and then further to consider other actions.
7 MR. SHEWMON:
Are the plants closed down until 8
we all make up our minds?
9 MR. JORDAN:
Until the plant and we make up our 10 mind.
They have the same problem in terms of understanding
(
11 whether this is a big or little problem.
They have 130 of i
12 the se types of switches used in that particular plant.
l 13 MR. SHEWMON:
Po s t-te s t you did say that they O
14 functioned properly when the level was lowered somewhat 15 further af ter the reactor was down?
16 MR. ALLISON:
It functioned sometimes at the set 17 point and sometimes off.
18 MR. SHEWMON:
Off by inches?
19 MR. ALLISON:
By a number of inches, yes.
)
20 MR. SHEWMON:
What is the spec supposed to be?
21 One inch pressure may or may not be a lot of head.
l 22 MR. ALLISON:
The trip set point is 12.5.
The 23 technical specifications require it~to come in by 11 inches.
24 This might be down to about 6 inches without a trip in this 25 case.
i O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 7
REE
()
1 MR. KERR:
I got the impression that these were 2
fairly newly installed switches?
3 MR. ALLISON:
That is something I forgot to
~
4 cover.
The se have been newly installed at LaSalle and 5
Oyster Creek in this application.
There is an interesting 6
thing about them.
They were put in to get an upgraded i
7 environnental qualification modification.
And they are 8
unlike the previous installation in that you can't tell if i
9 they are drif ting except by doing a calibration check.
It 10 is what I call a blind switch.
It is just a switch sitting 11 there.
12 The previous installation was a Yarway level 13 indicator which you could read every shift or every day and 14 see whether that indicator was tracking with the others.
15 You can't do that with the switches.
All you can do is put 16 in a test pressure and do a calibration check.
17 MR. KERR:
Is this because they are 18 environmentally qualified: in order to environmentally 19 qualify them, you can't see them?
20 l
MR. ALLISON:
It can be done either way.
This 21 just happened to be the selection that was made at this 22 plant and at Oyster Creek.
23 MR. KERR:
All these other switches are also 24 switches that were installed in order that they become 25 environmentally qualified?
U-s ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 8
REE
()
1 MR. ALLISON:
I am not sure why.
It is probably 2
so, though.
3 MR. MICHELSON:
Are they still using the Yarway 4
to indicate the control room reactor level?
5 MR. ALLISON:
I don't know.
6 MR. MICHELSON:
I guess what they have done is 7
replaced the control function of the Yarway with this 8
separa te switch.
9 MR. ALLISON:
That is true.
I don't know 10 whether they are still using Yarway for indication.
11 MR. MICHELSON:
Well, for what else would they 12 be used?
13 MR. ALLISON:
A new instrument, maybe.
14 MR. MICHELSON:
That is a possibility.
15 MR. KERR:
Your investigation is going to do 16 what?
To determine why the switches failed?
17 MR. ALLISON:
That is right.
And why they had 18 the problems of feedwater pumps that gave the transient and 19 how the operators reacted, whether they reported properly.
20 And --
1 21 MR. KERR:
Whe ther they reported properly?
I 22-MR. ALLISON:
Right.
See, it wasn' t reported --
23 it was a long time lag in the company realizing that the y 24 had a problem.
25 MR. SHEWMON:
A matter of hours?
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Naticawide Coverage 800-336 6646
27072.0 9
REE
()
1 MR. ALLISON:
13 hours1.50463e-4 days <br />0.00361 hours <br />2.149471e-5 weeks <br />4.9465e-6 months <br />.
2 MR. KERR:. According to television.
3 MR. EBERSOLE:
Will you tell me about the 4
Yarways that are still in other plants?
5 MR. JORDAN:
The equipment qualification program 6
has ascertained the quality of all of the ins trumen ta tion,
7 so there should not be a doubt at this point.
8 MR. EBERSOLE:
Was it the Yarways were uniquely 9
disqualified at this station and they are qualified 10 elsewhere?
11 MR. JORDAN:
There were a number of transducers 12 in plants that were not qualified and were either replaced 13 or qualified in the overall equipment qualification process.
14 MR. EBERSOLE:
This is no condemnation of Yarway 15 per se?
16 MR. JORDAN:
Not at all.
17 MR. MARK:
Was this 13 nours between recognizing 18 there was something to look at or telling you guys?
19 MR. ALLISON:
It was in recognizing it.
The 20 company reported it to us when they caught it, what the y 21 did.
22 MR. EBERSOLE:
Any further questions?
23 MR. REED:
I would like to make a point.
You 24 know, it is a lot dif ferent being on the doing end than it i
25 is on the critic, media critic or regulatory critic.
And
(~)
\\_-
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6666 I
\\
27072.0 10 REE
()
1 this doesn ' t sound to me like a very bad performance on the 2
part. of the licensee.
In fact,- it sounds like good 3
performance on the part of the licensee.
I wonder -- of
~
4 course, they are always under the gun for instantaneous 5
reporting before they have all the information.
6 Let me ask this question, now that you have the 7
information that you have, do you feel embarrassed about 8
calling in an AIT?
9 MR. JORDAN:
I will answer that.
Not at all.
10 We don ' t have all the information yet, nor does the utility.
11 And there is not, at this point, a condemnation of the 12 utility's performance by the NRC.
We are simply describing 13 to you a process that is going on and a potential problem O
14 with a particular type of instrument.
1" MR. KERR:
What led you decide that an AIT was 16 needed?
17 MR. JORDAN:
A failure to trip.
That a trip 18 point had been attained and the reactor system didn' t 19 respond with the trip and the operator didn' t trip out the 20 plant.
That was the utility's view as well as the NRC's 21 view.
So the utility then notified the NRC that their l
22 plant had failed to trip when it should have and they l
23 /
voluntarily took the plant down to try to find out what had 24 happened.
So there is now this review to fully understand 1
25 the problem.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
l 27072.0 11 REE
()
1 MR. KERR:
How long is the plant likely to be 2
down?
3 MR. JORDAN:
At this point, I don't know.
4 MR. EBERSOLE:
Any further questions?
If not, 5
we can pick up the next one.
The next one has a very long 6
history, like about 15 to 20 years.
7 MR. ALLISON:
I guess if I can interrupt, I 8
think there is one thing I forgot to tell you about LaSalle 9
that I probably ought to mention.
There was a previous 10 trip of the plant on May 9 where the level, in this case 11 the level was going down and kept going down.
And at that 12 time, the operator noticed the level being below 12.5 13 inches and decided to scram -- this is kind of preliminary O
14 and subject to verification -- and the automatic scram beat 15 him.
However, the point is, I think, that, you know, in l
16 that event, it is another data point that technically says, 17 you had a drift of about this much occurring in the 18 instruments.
And I wanted to mention that, and then ask l
19 you if it would be all right if we interchange the order i
20 and take Palisades first?
21 MR. EBERSOLE:
I don't mind.
22 MR. MICHELSON:
Could I ask on the last bit of 23 data you gave us, I thought you said the trip point was 24 about 12 inches?
25 MR. ALLISON:
12.5.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Naticnwide Coverage 800-336-6646
27072.0 12 REE
()
1 MR. MICHELSON:
So you didn't have a chance yet 2
to trip.
The operator intervened before the automatic 3
equipment responded.
Is that what you said?
4 MR. ALLISON:
No, in the previous event, the 5
operator saw it and said, that level is below the scram set 6
point.
7 MR. MICHELSON:
He was not well below 12 inches?
8 MR. ALLISON:
I don't know how far, but below.
9 MR. MICHELSON:
Below where the automatic should 10 have come on.
11 MR. ALLISON:
Right.
He tried to scram it but 12 the automatic beat him.
13 MR. MICHELSON:
Thank you.
14 MR. MOELLER:
I wanted to comment on the 1
15 preliminary notification of event or unusual occurrence, 16 the one da ted June 2nd.
On the fourth reading of the 17 second paragraph, I think I understand what is said, but it 18 says, "and declared an alert under its emergency 19 classification system on the basis of the potential ATWS at 20 about 6:30 p.m."
21 The ATWS did not occur. at 6:30 p.m.
I gather on 22 the fourth reading, I figured out they declared an alert at 23 6:30 on the basis of the potential ATWS that occurred at 4:21 24 a.m.
25 MR. ALLISON:
Yes, sir.
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
l 27072.0 13 REE
()
1 MR. MOELLER:
I think this this needs edited.
2 MR. ALLISON:
Well, it is too la te.
Once the y 3
go out, they are out.
The Region told me they issued 4
another one yesterday af ternoon, but I haven't been able to 5
find it ye t.
It will probably clarify that.
6 MR. MICHELSON:
Why did they declare an alert 14 7
hours after what was going to happen had already happened?
8 MR. ALLISON:
The conservative assumption was to 9
say that the reactor protection system was not working.
10 MR. MICHELSON:
But the reactor was scrammed.
11 MR. ALLISON:
No, it was still operating at that 12 time.
13 MR. MICHELSON:
That is right.
Okay.
It is a O
14 whole 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br /> be' fore they discovered it at that point?
15 MR. ALLISON:
Right.
Once they decided that it 16 might not have worked, you make the conservative assumption, 17 put in a trip.
18 MR. EBERSOLE:
Other things being equal, if you I
19 drop water that f ar in a boiler, you take the static head 20 of f the boiling process and that is going to open up the i
21 void and there should be a power shif t downward.
Had they 22 compensated for that?
What you are doing is increasing 23 void compression because you are reducing static head.
24 MR. ALLISON:
I haven' t really delved into what 25 is true level and pressure compensation.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 14 REE
(
l MR. EBERSOLE:
If you keep on running the water 2
down, it is going to shut down.
3 MR. KERR:
The pressure sensor in the reactor 4
will sense that and will compensate for it with a recirc.
5 MR. EBERSOLE:
It doesn' t accommodate changes in 6
the liquid head vessel, does it?
7 MR. ALLISON:
I think it is sensing the 8
differential pressure.
9 MR. KERR:
You have a power level sensor.
It j
10 adjusts the recirc flow to keep the pressure causing vessel 11 at whatever power level you want.
So I think that would be 12 taken care of automatically.
1:
MR. EBERSOLE:
I was unaware that that did that.
14 That is not automatic, though?
15 MR. KERR:
Ye s.
It is automatic.
16 MR. EBERSOLE:
Is the recirc automatically 17 adjusted?
18 MR. ALLISON:
I am going to ask Wayne Hodges.
19 MR. HODGES:
For the relatively small drop in 20 level that you are talking about, the power change would 21 not have been very great; even for very wide swings in 22 water level, the void fraction changes very little.
As
)
23 long as -- if you can look at it from the standpoint, you 24 stay at zero reactivity, then the only thing that is going 25 to make much difference is the temperature e f fect.
OV l
i ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 15 REE p) 1 MR. EBERSOLE:
It is just in the noise.
q 2
MR. HODGES:
It is basically in the noise.
As I 3
understand it, the operator had started to manually valve 4
the flow down during the event and was cutting back on the 5
recirc flow and the power and that manual action would have 6
overwhelmed anything.
7 MR. EBERSOLE:
Am I correct in saying there is 8
no automatic recirc flow control?
9 MR. HODGES:
Yes, there is an automatic recirc 10 flow control, but it would not have -- it is based upon I
11 what you are doing with the turbine and such.
And it can 12 run manually.
But what we are talking about here is a 13 matter of the operator manually valving the flow downwards.
O 14 And then when you get to level 3, the low level trip, you 15 get a runback -- these are control room pump speed and 16 valve control rather than -- there are two pump speeds, a 17 high speed and low speed, and then valve control.
And so i
18 when it ge ts down to the level 3, you go from a high speed 19 to a low speed.
20 MR. EBERSOLE:
If the reactor held power and got 21 on down below this 12 inches, it would have hurt the top 22 end.
23 MR. HODGES:
If it had gone that far, that is 24 correct.
25 MR. EBERSOLE:
So we were about 16 or 24 inche s
(
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
~ _ _.
27072.0 16 REE
/
1 from trouble.
2 MR. HODGES:
You were -- the instrument zero is 3
about 150, 155 inches above the top of the core.
You were 4
maybe six or so inches above that.
So you were a long way 5
away f rom the top of the fuel.
1 6
MR. EBERSOLE:
I misread this.
7 Okay.
The automatic tran set point of 12.5 i
8 inches?
9 MR. HODGES:
Yes.
Zero on that is the bottom.
j 10 MR. EBERSOLE:
So you are a long way?
11 MR..HODGES:
Yes.
I 12 MR. ALLISON:
I guess we would like to -- we l
13 will probably talk about this again at the next mee ting.
O 14 We would like to present Palisades now.
In 15 introducing Bill Hehl from Region 3, he will give the 16 presentation.
And in introducing Bill, I would like to 17 just mention that at the Subcommittee mee ting, there was 18 some question about whether this event was important enough 19 to warrant an AIT.
We are prepared to address those J
20 questions, but I would like to first let Bill describe the 21 event.
22 MR. HEHL:
What I would like to do, in the 23 Subcommittee meeting earlier this week, Tom Weibach went 24 into quite a bit of detail about the actual transient that 25 occurred, and what I would like to do is present to the OV ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 17 REE
()
1 f ull Committee some of the background information that 2
occurred, go briefly over the transient that resulted and i
3 then discuss an overview of the findings that came out of
~
4 the investigative effort that we did at Palisades.
5 Just to attempt to try to give you some of the 6
data base from which Region 3 headquarters were operating, 7
I guess a brief description of the event on May 19, 8
Palisades experienced a reactor trip of 99 percent power in 9
response to a high pressurizer pressure condition.
The 10 high pressurizer pressure condition was the result of the 11 loss of control power to the turbine EAC system which 12 allowed the turbine valves to close.
Upon reactor trip and 13 1
during the (E'l
, plant recovery, numerous pieces of equipment 14 failed to perform.
And as a result of this reactor trip 15 and the associated equipment failures, the poten tial I
16 I
serious challenges to safety systems that the se type of 17 failures pose, the burden that these failures place on the i
18 operating Staff, the NRC dispatched a f act finding team to 19 Palisades to review the event prior to the unit returning 20 to power.
21 The augmented inspection team was tasked with 22 performing an independent review of the May 19 trip to 23 assure that the scope of the equipment failures were i
24 accurately known, to evaluate the equipmer' failures and to 25 gain perspective regarding the impact of these failures and O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 18 REE
()
1 any existing out-of-service equipment on the operating 2
staf f and their ability to respond to the plant transients.
3 Getting into the history.
In the most recent 4
SALP report covering the period of November of 1984 to 5
October 31, 1985, the areas of maintenance, surveillance, 6
quality programs and administrative controls declined at 7
Palisades from a previous rating of 2 to a category 3.
8 The low ratings in these areas were in part due 9
to a lack of aggressive corrective action by the licensee 10 and poor management controls.
Prior events at the facility, 11 beginning in late 1984, due in part to inadequate 12 maintenance, involved other problems of safety-related 13 equipment.
This included five separate events related to 14 leaking safety injection tank check valves; and despite 15 maintenance on these valves during the cycle 5 refueling 16 outage, during cycle 6, two of eight of these valves had to 17 be refilled.
18 Additionally, during the cycle 6 refueling, the 19 licensee elected not to perform maintenance on the primary 20 coolant pumps, despite indicated seal oscillations.
21 During the week of March 3rd, Palisades facility 22 was returning to operation from a cycle 6 refueling outage 23 with the ' first stage seals not staging on two of the four 24 reactor coolant pumps, a problem which may have been 25 averted had they done some work on those during the outage.
()
ACE FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 19
,REE
(}
l In addition, valve leakage problems were 2
identified in the primary coolant system loop check valves 3
on HPCI injection line, two safety injection tank pressure 4
control valves and a manual isolation valve associated with 5
the safety injection tanks and the three-way valve in the 6
CVCS system.
7 Following discussions between Consumers Power 8
Company and the NRC, the licensee elected to shut down and 9
repair these problems on March 8.
Final completion of 10 those repairs, the licensee returned the unit to power on 11 March 25.
March 26 the unit tripped as a result of main 12 generator voltage regulator problems and was returned to I
13 power on March 27.
O 1
14 On April lo, Palisades again shut down af ter i
15 exceeding the tech spec limit for unidentified primary j
16 coolant system leakage.
The licensee determined the cause 17 to be a failed relief valve.
i 18 During the return of power on April 11, when it 19 shut down, the licensee experienced a packing failure on 20 the condensate pump A.
The pump was repacked twice prior 21 to a decision to finally replace the pump with an on-site 22 spare.
And following this pump replacement, the unit was 23 re turned to power.
24 On April 23rd and 29th, existing valve leakage 25 problems in the primary coolant makeup system resulted in O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 20 REE f~)
1 excursions in the TL which, although later determined to be J
2 enveloped by the accident analysis, at the time of these 3
excursions and at the time of the report from the licensee, 4
there was concern w'ith regard to exceeding the main steam 5
line break analysis assembly.
6 Automatic associated manual valves, pure water 7
to boric acid feeds to the primary coolant makeup water 8
system were leaking to the extent that when an operator 9
attempted to make up to the primary coolant sys tem, he 10 didn't know whether he was diluting or borating.
These 11 valves were subsequently repaired while they were on line,
i 12 a few days prior to the March 19 trip.
13 MR. OKRENT:
Would you say this is far from the 14 average experience you would expect, similar to the average?
15 Can you categorize it?
16 MR. HEHL:
I would characterize, at least from 17 my experience in Region 2 and in Region 3, that this is far 18 from wnat typically would be expected, and -- as far as 19 repetitiveness of equipment failures.
l 20 MR. OKRENT:
Except for the repetitiveness of 21 certain equipment, if they had done more maintenance during 22 the prior shutdown, is there any reason to have confidence 23 that many of these would not have occurred and nothing new 24 would have occurred?
25 MR. HEHL:
Let me answer that by saying that as Ov ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 21 REE
{}
l part of this investigative effort at the site, we 2
concentrated on looking at both equipment histories and 3
interviews with the operating staf f and with the 4
engineering sta f f.
During those in te rv ie ws, T think, as 5
brought out during our review of their equipment his torie s,
6 there was significant concern on the part of operators with 7
regard to the maintenance activities that take place at 8
Palisades and with the reliability of the equipment there.
9 MR. EBERSOLE:
You are looking at a case like 10 Eastern Airlines.
11 MR. OKRENT:
I am looking at the leakage in 12 particular.
The se leakages presumably were either not l
13 there or not so obvious before the --
14 MR. HEHL:
They have experience with regard to 15 the safety injection tank check valves.
They have 16 experienced leakage from those beginning in about 1983 time 17 frame.
During the cycle 5 refueling outage, they went in 18 and rebuilt all of those check valves.
Excuse me, cycle 4.
19 Coming into cycle 5, that lasted for about one month, until 20 they started developing leakage back into the safety 21 injection tank system from the primary coolant system.
22 The leakage at that point, the continuing 3
23 leakage over most of that cycle varied, at times being 24 excessive to the point of requiring plant shutdown; at 25 other times, being less than requiring plant shutdown, but O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
27072.0 22 REE
['J) 1 in all cases requiring additional operator action to go out
~
2 and manually take care of the service on these systems on a 3
very periodic basis.
~
4 MR. OKRENT:
They were building -- the 5
rebuilding, is that something they did in-house or did they 6
bring in someone who was supposedly knowledgeable about 7
these valves?
8 MR. HEHL:
My understanding is that there was 9
vendor assistance to rebuild these.
10 MR. OKRENT:
Why should the problem have 11 occurred?
12 MR. HEHL:
At this point they are looking at the 13 adequacy of these valves for the application.
Whe ther -
.T.
14 guess you can always question whether that had, should have 15 been done three years ago or 10 years ago or what.
I 16 MR. EBERSOLE:
What does the Staf f do about, 17 differentiating between inadequate design -- how does the 18 Staf f dif ferentiate between poor maintenance?
19 MR. HEHL:
I guess I can't speak for --
l 20 MR. ALLISON:
I would just say that from the 21 I headquarters perspective, what we see is new reactor i
22 coolant leak every two or three days at Palisades.
It 1
23 looks like there is some thing wrong.
We don't really i
24 differentiate.
I guess I don' t know whether it is a design 1
25 or a maintenance problem.
But when I see the problem O
ace-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
l 27072.0 23 REE
(' ';
1 occurring over and over again and it doe sn ' t get corrected,
\\_/
2 you know tnat there is some thing that you would like to see 3
corrected there.
~
4 MR. EBERSOLE:
It is the operator's option to 5
put in better equipment?
6 MR. F HL:
Agreed.
7 MR. ALLISON:
Yes.
8 MR. OKRENT:
Is there another valve which is 9
known to be better than the one that is there for this one 10 that was rebuilt and still creates trouble?
I am trying to 11 understand --
12 MR. ALLISON:
I don't know.
Bill, do you know 13 that?
(-J 14 MR. HEHL:
All I can tell you is --
15 MR. OKRENT:
There aren't that many plants like 16 Palisades.
17 MR. HEHL:
There aren't many plants like 18 Palisades and Palisades does have some unique problems with 19 the location of their safety injection tanks, so high in 20 containment.
But there are numerous other facilities that 21 do not experience this type of repe titive and continuous 22 problems with their check valve leakage.
23 MR. EBERSOLE:
Do you happen to know that these 24 models of check valves are used all over the place and 25 there is less trouble with them elsewhere?
,7
)
ACE-FEDERAL REPCRTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 24 REE
()
1 MR. HEHL:
I can' t tell you that.
2 MR. REED:
Well, Palisades is a plant that has 3
been in operation since about 1970 or so, and if the valves 4
were basically good in 1970 and were not perfect in 1970, 5
back-leakage or erosion would have put them in a bad shape 6
by 1986.
So I don't know how accessible the valves are or 7
anything like that, but if they were not pe r f e c t -- and I 8
am sure they weren't -- when they were first put in service, 9
they would have needed some very caref ul maintenance in the 10 intervening years, very careful attention.
11 It is my impression that Palisades' operating 12 organization, the maintenance organization, are not on t
i 13 their toes.
y b
14 MR. HEHL:
You are stealing my thunder.
15 MR. OKRENT:
In this particular case, I 16 understood that they brought a vendor in, and one might 17 argue that they were looking for a quality job.
18 MR. REED:
Well --
19 MR. HEHL:
I have no knowledge of whether the 20 vendor worked the valves or whether the vendor was there to 21 provide assistance.
22 MR. REED:
I have great difficulty in believing l
that a vendor would understand the importance of what I 23 24 will call zero leakage on check valves.
Having gone 25 through this before, I have always asked for, once the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
i 27072.0 25 REE
(;
1 valve is brand new, delivered and supposedly perfect, 2
testing of the valve's back-leakage to try to establish 3
zero before you put that valve into that application to i
4 holo against accumulators.
But vendors wouldn't have an 5
appreciation, if they came in to just get bodies to do a 6
job, they wouldn't have the apprecia tion of the kind of job 7
they should do.
8 MR. EBERSOLE:
You didn't make clear that it was 9
a vendor maintenance program.
10 MR. HEHL:
I don't belie ve the vendor did the 11 I
maintenance.
I 12 MR. REED:
In-house people are not doing the 13 maintenance they need to do.
And it may be very difficult N-)
14 maintenance to do.
It may may be highly radioactive.
It 15 may be a large valve seat.
Hours of time in that radiation 16 field.
17 MR. EBERSOLE:
The stage is set to go into this, 18 isn't it?
There was some premonition that some thing like 19 this would happen.
20 MR. HEHL:
I think we had -- I think we had 21 premonitions that this was -- I think we were happy to find 22 the results of our investigations.
23 Ra the c Lhan going step by step, there is a 24 listing of the equipment failures that occurred and there 25 was some existing equipment out of service.
,~L)
ACE-FEDERAL REPORT ERS, INC.
202-347-3700 Nationwide Coverage 800-336- % 46
27072.0 26 REE
()
f 1
Let me go through the synopsis of the reactor 2
trip event, which is a compilation from the view of the 3
post-trip data and interviews with the operators.
4 That is that shortly before 2:16 p.m. on May 19, 5
both the primary, a 15 volt DC, and secondary,.215 DC 6
electric power supplies tripped, causing the loss of the AC 7
power, allowing the main turbine valves to drift close.
8 The loss of the turbine load initially resulted in the 9
reactor coolant tempera ture increasing.
It is noted that 10 at the time of the event, the plant was operating at an i
11 elevated TL to maximize electrical output.
This elevated 12 TL was necessitated by the number of plugged tubes in the 13 steam generators.
s 14 The increase of the TL was promptly identified 6
15 I by the plant primary control room operator, who was 16 designated as CO 2, who immediately initiated manual rod 17 insertions.
Simultaneously the control room turbine 18 operator observed the loss of governor valve indication, 19 and based on his evaluation of the plant conditions, 20 started reaching to trip the turbine.
Prior to him being 21 able to trip the turbine, though, the reactor tripped on 22 high pressurizer pressure.
23 MR. EBERSOLE:
What span of time was this?
24 MR. HEHL:
This was in the af ternoon, about 2:16 25 p.m. on eastern daylight time.
O ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
27072.0 27 REE
/'T 1
Upon the reactor trip, the operators initia ted U
2 the post-tr'ip re sponse in accordance to emergency 3
procedures, and although not required by the procedures, 4
the operator followed the automatic trip with a manual 5
6 During a scan of the control boards, the 7
operator noted that the turbine bypass valve and one of the 8
four atmospheric dump valves failed to open.
These 9
failures caused lifting of all three banks of main steam 10 safe ty valves.
To limit primary system cooldown, the 11 operators tripped both the manual feed pumps of the 12 motor-driven auxiliary feed pump A which had initia ted 13 following the shrinkage af ter the trip.
But in an attempt 14 to more rapidly restore steam generator levels to normal, 15 and also to extract a little bit quicker return of TL to 16 its elevated condition, the operator initiated the 17 turbine-driven aux feedwater pump manually.
18 During these observations of the PCS parameters, 19 the operator observed that the pressurizer spray valve had 20 indicated open.
21 Noting that pressurizer pressure was fairly 22 rapidly decreasing, the operator tried to initiate this 23 third charging pump.
In spite of the repeated unsuccessful 24 attempts to start the charging pump, the pressurizer 25 pressure stabilized at about 690 pounds.
(3 v
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336 6646
27072.0 28 REE
~
1 During the subsequent plant stabilization and e
v 2
recovery, the operators noted the control rod 34 rod bottom 3
light had not illuminated.
They had verified that all rods 4
had in fact inserted.
5 Following the return on pressurizer level above 6
the letdown isolation set points, the operators attempted I
7 to restore letdown flow.
During this restoration, the 8
letdown orifice regulator closed, necessitating the 9
operator to shif t to a redundant regulator, which 10 functioned properly, although it had been previously 11 identified as unreliable and subject to another work 12 request.
13 The plant was stabilized in hot shutdown pending
/
)
(/
14 post-trip review and, evaluation of equipment failures.
t 15 l
MR. EBERSOLE:
The pressurizer spray valve I
l 16 l
failed to fully close?
l 17 L
MR. HEHL:
That is correct.
I 18 l
MR. EBERSOLE:
Had it failed to close at all, 19 what would have happened then?
Would you have sprayed out i
20 the steam bubble and eventually blinded the pressurizer 21 spray system?
22 MR. HEHL:
The Palisades plant is designed with 23 no charging pumps operating on a trip from 100 percent 24 power such tha t, the pressurizer is such that you won't 25 lose level indication.
But that is assuming that you have pv ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 29 REE
~
l the pressurizer heater, all the banks in service.
And also 2
assumes that the le tdown is correct.
So I cannot give you 3
a definitive answer.
4 If the pressurizer spray valve had, in fact, 5
stuck full, it surely would have significantly increased 6
the cooldown rate and probably the significance of this 7
But I can tell you from interviews with the 8
ope ra tors, that looking over and seeing the pressurizer i
9 spray valve open indication, coupled with his rapid 10 cooldown, I guess it generated quite a bit of consternation.
)
11 MR. EBERSOLE:
How does the operator feel when 12 he realizes his pressurizer is rising to the point where he 13 is blinded?
He doesn't like that, does he?
()
14 MR. HEHL:
No.
15 MR. REED:
I might ask the question, on the 16 pressurizer spray valve and your review, you should look at 17 whether that is a bellow-sealed pressurizer spray valve or 18 whether it is just a packed valve, because that may be a 19 factor in its failure to close.
20 I might also point out that the I&E did not 21 bring be fore us the designed pressurizer spray valve as it 22 would have occurred.
There may be some relationship.
23 Quite frankly, I can't understand why the packing blew out 24 on this spray valve because they at one time had 25 bellow-sealed backup packing spray valves.
So this is
(
l ACE-FEDERAL REPORTERS, INC.
2 @ 347-3700 Nationwide Coverage 800-336-6646 n
27072.0 30 REE
/ ')
1 something I don't quite understand.
Maybe you want to LJ 2
follow up on that and whether the bellows have been removed 3
or what.
4 MR. HEHL:
The licensee notified the NRC of the 5
plant trip and during this notification, which was at 2:54 6
p.m.,
the licensee identified the trip initiated signals, 7
loss of turbine load, turbine trip.
Further evaluation of 8
by the licensee determined that the initiating signal was 9
in fact pressurizer high pressure.
This clarification was 10 subsequently made to the NRC through the E&S.
Subsequent 11 to this determination and phone call at 5:45 p.m.,
the l
12 licensee notified the operations center that in fact, in 13 accordance with their emergency plans, this was an unusual
/";
I
(_j' 14 j
event, based on the type of trip that occurred.
I 15 It is noted also that during these three 16 telephone notifications by the licensee, that in all cases, 17 the headquarters duty of ficer queried the licensee with 18 regard to any other equipment failures, anything abnormal 19 about the plant trip.
The answer in each and every case 20 was no.
I 21 MR. OKRENT:
Excuse me.
I am trying to 22 understand something.
It tripped on pressurizer high 23 pressure, whicn, I assume, indicated tne primary system was 24 heating up?
25 MR. HEHL:
That is correct.
p),
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
27072.0 31 REE
(
1 MR. OKRENT:
But you referred to a cooldown.
A]_/1 2
MR. HEHL:
Yes.
3 MR. OKRENT:
So this was a cooldown occurring 4
after --
5 MR. HEHL:
Yes.
6 MR. OKRENT:
It was not -- was it a very rapid 7
cooldown, something unusual about the cooldown?
8 MR. HEHL:
It was more -- I guess it was a 9
little bit more rapid than what would typically have been 10 experienced following a trip from 100 percent power, based 11 probably on the operator's actions to try to bring TL down 12 fairly rapidly, based on the elevated start point and the 13 type of TL increase that occurred.
The operators probably D.
()
14 left the feed pumps on a little bit longer than they would 4
i 15 have in a normal trip, if anything occurred.
They t
16 potentially would not have started the turbine-driven aux 17 feedwater pump.
So I think probably --
18 MR. OKRENT:
They didn' t violate any cooldown 19 rate?
20 l
MR. HEHL:
No.
I will -- the bottom line fren 21 this event itself is that the failures that occurred did 22 not result in significant worsening of a plant transient, 23 that the performance of the operators and other major plant 24 systems, that they performed as expected and they were well 25 within any design criteria, p
Y_.
i ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646 i
27072.0 32 REE o()
1 As I said, tae augmented inspection or 2
augmenting inspection team was dispatched to the facility 3
with the task of:
Evaluate equipment f ailures and the 4
impact of these failures on the operator's ability to 5
perform or respond to the operational transients.
6 With regard to the equipment failures, the AIT 7
found, based on their review of the apparent failure modes, 8
the maintenance history and discussions with the licensee's 9
maintenance organization, that significant weaknesses exist 10 in the areas of diagnostics, troubleshooting repair and 11 pos t-main te nance te s t ing.
These were contributors to most 12 of the failures that occurred.
13 Interviews with the operations shif t supervision, 14 the opera ting management de termined that historically, the 15 equipment failures have been of significant concern 16 rela tive to the additional burden that these type of things 17 add to the operators' -- in other words, the things that 18 they just have to work around during routine operations and 19 the number of things that they have to remember to operate j
20 manually or are not going to function at all after an 21 operational transient.
4 22 MR. MOELLER:
Your last point on the operators i
23 having the serious concerns, in other words, the plant 24 operators really don' t have faith in the ir support --
25 MR. HEHL:
The plant operators have no O
t_J ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I I
27072.0 33 REE
[')
I confidence in the maintenance those people perform.
They v
2 have no confidence that when they go to start another piece 3
of equipment at the plant that the equipment will start.
4 MR. EBERSOLE:
Are you putting the operators and 5
the maintenance people in a separate category?
It looks 6
like they are charging their own maintenance people with i
7 problems.
8 MR. HEHL:
That is correct.
9 MR. EBERSOLE:
I thought we are supposed to have 10 a unified plant organization which included maintenance.
11 Who is the boss that is in charge of all this?
Can you 12 find him?
At TVA you never could.
13 MR. SHEWMON:
American Electric Power down in O
14 Columbus.
What do you mean?
15 MR. KERR:
This is Consumers Power.
16 MR. OKRENT:
Can I ask a question.
17 MR. EBERSOLE:
Is this a managerial problem?
18 MR. HEHL:
Let me say that at Palisades, there 19 has been, within the last two years, significant management 20 changes at Palisades.
With the demise of the Midland 21 facility, there was a truncation of the pyramid at 22 Palisades.
That was replaced by Midland management.
Part 23 of the problems that we observed and part of the management 24 control problems that exist are probably rooted in the 25 experierce level of both the Palisades plant maintenance n
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3364 4 6
27072.0 34 REE
)
1 operations, eng ineering, and the inexperience of the 2
management staff.
3 MR. EBERSOLE:
I think we have to look for what 4
you are going to do about it and what are we going to do 5
about it here -- I am talking about the f ull Committee --
6 if anything?
7 MR. REED:
Does Consumers Power at Palisades use 8
validated aptitude te s ting for employment and transfer of 9
their personnel at that plant?
10 l
MR. HEHL:
I cannot answer that.
I 11 l
MR. REED:
I think you should find out, because I
12 I hear in the air that there have been years of problems 13 with respect to performance of plant people and operations
,n k
14 and maintenance.
15 i
On the other hand, I don't hear a lot about Big l
16 l Rock being a particular [roblem, although it is a very 17 small plant.
18 MR. HEHL:
That is not an uncommon problem to 19 have differing levels of performance within the same 20 utility.
21 MR. REED:
That gets down to a lack of 22 standardization, of evaluation and processing of new 23 employee s.
You may have an individual that is responsible 24 for promoting, handling, disciplining of employees at Big 25 Rock that has been successful.
On the other hand, on the 73
(_)
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
27072.0 35 REE
()
1 larger scene at Palisades, you need more sophisticated 2
regimentation.
3 MR. HEHL:
That may be part of the problem.
Did you say tbat there has been a 4
MR. WARD:
5 step change in the performance or it is your perception 6
that there has been a step change in performance at 7
Palisades since this Midland organization shift was made?
8 You said that.
9 MR. SHEWMON:
I thought he said a step change in 10 personnel.
11 MR. HEHL:
There was a step change in personnel.
12 MR. WARD:
Why were you telling us about that?
i 13 MR. HEHL:
The jury is still out on the final
(
(~
l 14 outcome of the personnel changes.
15 I think that Region 3 has confidence or a 16 limited amount of confidence or some measure of confidence 17 in the current management's ability.
But I think part of 18 the root problem here is that you have a facility that has 19 been neglected and has been the focus of least-cost fixes.
20 MR. WARD:
Is it your opinion that the step 21 change in personnel has improved the situation or not?
22 MR. HEHL:
I think at this point it is too early 23 l
to tell.
I think that we do see maybe the upturn, but I 24 think back in about 1981-82, we thought we were seeing the 25 same thing.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 36 REE f~N 1
MR. OKRENT:
Are there any objective indica tors v
2 that support your, what I will call opinions concerning the 3
quality of the maintenance?
In other words, are there 4
larger failure rates of valves, pumps, wha tever you think 5
might be related to poor maintenance, is there a larger 6
failure rate here than at other plants, or do you have any 7
indicators that you consider to be good objective measures 8
of what you are saying?
9 MR. HEHL:
During the last ref ueling outage, the 10 outage was held up at the end because of containment sump 11 isolation valves required to pass the integra ted LOCA leak 12 ra te test.
That set of valves was reworked, estimates in 13 the double digits, until someone finally came across with
( >)
14 i
the piece of information that the valve was supposed to be s
i 15 at a slight angle.
It wasn ' t a zero degree flat.
16 Once that was performed, the valve passed.
The 17 biggest concern from the operators and indications that 18 they have of poor maintenance is the amount of rework that 19 is being, having to be done consistently.
20 MR. REED:
The force the outage rate probably is 21 terrible for this plant.
I would expect that is one of the 22 best indicators.
23 MR. HEHL:
Ye s.
24 MR. REED:
In the state of Michigan, it is my 25 understanding tha t there has been very harsh treatment by O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 800-336-6646
27072.0 37 REE f))
I the public service commissions of the state with regard to 2
the financial stability of this company.
Is this a factor 3
in the plant situation?
4 MR. KERR:
You are not addressing the question 5
to me, are you?
6 MR. REED:
No.
Is it a factor in your opinion?
7 MR. HEHL:
I believe that we have, at Region 3, 8
have in the past felt that -- I don't think it is any 9
secret that the company has been in significant straits 10 with the facility.
11 MR. REED:
Financial straits?
12 MR. HEHL:
Ye s.
Their employees have had to 13 take significant pay cuts.
Their attrition rate up until 14 this year in their ope' rating department has been on the 15 order of 25 percent.
75 percent of their maintenance 16 engineering, their engineers and technical staf f engineers 17 are right out of school.
18 MR. EBERSOLE:
You are saying this might go back 19 to rates?
20 MR. HEHL:
The licensee contends on = regular 21 basis that there is no ef fect on the Palisades f acility 22 based on their financial condition.
23 MR. EBERSOLE:
Well, they are pretty much forced 24 into saying that.
25 I guess the bottom line, here again, that I O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 38 REE 1
would like to call out, this is the only chance to get a 2
review in the trenches.
I would like to hear, have you 3
listen, in the context of what are we going to do, if 4
anything.
I haven' t heard what your action might be.
5 MR. HEHL:
With regard to Palisades facilities, 6
in late 1984, early January or February of 1985, we 7
conducted some team inspections at Palisades, looking at 8
the maintenance activities there.
As a result of that 9
inspection effort, there was a significant number of 10 violations identified, corrective action implemented.
The 11 return visit was made in September of 1985, at which time 12 there had not been significant progress made in the 13 reduction of the maintenance backlog that existed at that
,e -8 J
14 point in time, both in sa fe ty-rela ted, nonsafety and 15 I control room-type of deficiencies.
l 16 As a result of that second view, there was a 17 confirmatory action letter issued to the facility.
It 18 addressed some commitments on their part to reducing this j
19 backlog and improving maintenance activities at the 20 facility.
21 We monitored the performance in that area 22 through portions of the ref ueling outage.
At this point in 23 time, based on the history and the May 19 trip, we have 24 another confirmatory action letter in place which requires 25 them, in addition to determining the sequence of the events
{')
v ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6616
27072.0 39 REE
("T 1
and the problems encountered on the May 19 trip, to also V
2 look at and inve s tiga te the status of safety and import. to 3
safety equipment and they are going to -- in fact, they are 4
in Region 3 today discussing the scope of that ef fort, that 5
review ef fort, to try to establish, then to get a handle on 6
what the condition of the plant is.
7 My understanding is that that mee ting today will 8
be followed by a meeting on June 20th, at which time they 9
will present us with the results of their review, 10 corrective action they have taken to date, projections 11 towards a start-up date.
12 MR. REED:
Do you know the solution to what you 13 are talking about?
You almost seem to be telling us of a 14 parallel case to TVA, only it isn' t quite as ripe at this 15 point in time.
The solution to this kind of situation is 16 not easy.
I don't know that anyone yet has solutions to 17 the Tennessee Valley Authority situation.
18 Do you feel that the NRC Staff and the Reg ion 19 people are going to be able to come up with genuine 20 turnaround conclusions?
Or is it beyond the Region and I&E 21 to come up with turnaround conclusions?
22 MR. HEHL:
I think that we can -- we have gained 23 sufficient knowledge at this point to be able to evaluate 24 the licensee 's program for implementing turnaround.
But 25 there again, it is all in the implementation of that sort O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6M6
27072.0 40 REE
^(^
l of a program.
w; 2
I believe that we are committed and I believe 3
that the S ta f f is committed to a turnaround.
4 MR. EBERSOLE:
Are the -- is the paper in order 5
on their design and construction and operation?
6 MR. HEHL:
During the inspection activities that 7
took place back in the -- well, in 1985, several team 8
inspections took place, there were significant problems 9
identified in the paperwork portion of it.
There were 10 problems identified in that the facility did not have any 11 kind of a maintenance history, maintenance tracking type of 12 system that was usable for performing any kind of analysis 13 of failure mechanisms to determine whether or not
-/~'s 1
is 14 preventive maintenance which they weren' t doing would have 15 helped them along those lines.
Whether or not -- there I
16 l
were problems bound in the QA/QC involvement, through the i
17 whole spectrum.
18 MR. EBERSOLE:
Well, we have to move on.
I 19 presume the Committee would want us to follow on and see 20 how this evolves since it looks so potent.
It may be on 21 the --
22 MR. OKRENT:
The fact that they had to reduce 23 salaries, if I heatd correctly, and they are hiring junior 24 people, at least could be suggestive of the fact that they 25 either don't have appropriate rates or even though they
()
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
27072.0 41 REE
']
1 have appropriate rates, they are not running enough at the 2
time or something.
But --
3 MR. EBERSOLE:
Does any of your investigations 4
examine how much the fo'.ks are paid, whe ther they couldn' t 5
hire quality people?
6 MR. HEHL:
No.
7 MR. EBERSOLE:
You don't go that far?
8 MR. HEHL:
I know we have done some studies in 9
the past to look at the averages of pay for operators and --
10 MR. EBERSOLE:
Well, the PUCs are responsible I
11 i
tor this.
12 MR. JORDAN:
Maybe I could help answer that.
13 The utility is obliged to make the ir case with the PUC and
(;'
(_
14 l
they do hire consultants to help them explain the ir case s,
t 15 l
and it is, I think, the NRC's responsibility to make sure i
16 tha t they are making the case and providing qualified 17 people.
18 The Davis-Besse facility, tha t was one of the i
19 issues there, in terms of not providing fully qualified 20 statt.
21 In answer to another question to Glen Reed 's 22 question, the NRC management, including Vic Stello, have 23 revie wed the actlur.a tne t the Region has taken with respect 24 to Palisades, and at this time, we feel that it is an 25 appropriate action.
ACE-FEDERAL REPORTERS, INC.
L 347 3700 Nationwide Coverage 800-336-6646
27072.0 42 REE
(}
l There is a periodic, like about quarterly look 2
at the plants that we feel are most troubled, and so we 3
would like to come back to you with a description of what I
4 is happening here.
5 MR. EBERSOLE:
We would ask you to.
6 MR. JORDAN:
The thing I think is very 7
interesting is that we are finding the problems sith 8
Palisades principally based on balance of plant - equipment, 9
rather than safety-related equipment.
10 MR. EBERSOLE:
There is no NRC requirement on 11 what to do with that?
i 12 MR. JORDAN:
That is right.
But certainly we 13 are getting precursors of serious problems out of large s_J
,14 numbers of balance of plant failures.
15 MR. EBERSOLE:
All right.
I 16 MR. KERR:
One bit of information that has not 17 been mentioned here which may be relevant is the fact that 18 Consumers Power Company has a new president and a new vice 19 president, both of whom have extensive experience in the 20 nuclear business and the new vice president has only been 21 there -- he was with Consumers earlier -- he has been on 22 board and has had his current responsibility for about 23 three months.
24 MR. EBERSOLE:
So maybe things will improve?
I 25 MR. KERR:
I would guess that that would be the ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 43 REE
~
l case.
I simply mention it because it is a fact.
2 MR. EBERSOLE:
It could be.
Well, there was 3
some argument about should we take this up at the full 4
Committee at all.
5 MR. KERR:
I think it is significant.
6 MR. CARBON:
They have 25 percent turnover per 7
year, poor quality people, that seems a problem.
8 MR. EBERSOLE:
Well, I think we have done enough 9
to ge t the color of it.
We will see you again on this case.
10 Did you say quarterly or sooner?
11 MR. JORDAN:
The management review is about 12 quarterly.
The next meeting of the regional directors and 13 Mr. Stello is October at this time.
,O
(_/
14 MR. EBERSOLE:
But you would pick up the 15 resolution of this case earlier than that.
16 MR. JORDAN:
Yes.
17 MP. EBERSOLE:
So it will be earlier than that.
18 MR. JORDAN:
Yes.
We can give you a briefing on 19 the status at the next Subcommittee meeting and decide 20 whether it is good for the full Committee.
21 MR. WARD:
I was just looking here at the SALP 22 ratings for Palisades the last six years.
They seem to 23 have been largely, back in '79 and '80, up to
'82, largely 24 2s and 3s.
Then there was '82-83, is and 2s, but now they 25 have gone back down to 2 and 3 again.
)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336- % 46 c --n-
,--- _ n_- ~ ~ - -. -..
27072.0 44 REE
~%
1 Has the I&E or the Region activity gone up and (O
2 down in response to those ratings?
Are there -- is there 3
more pressure being put on Palisades now or --
4 MR. JORDAN:
The inspection programs are now 5
adj usted based on the poor performers.
The better 6
performers get less inspection; the poor per formers, more.
7 So Palisades is receiving a great deal of inspection 8
a t te n t ion.
9 MR. MOELLER:
One quick comment.
Obviously, you 10 have noted this, as you hear all of this, you realize then 11 that Palisades is down and they must, you know, justify 12 themselves, obviously, be fore they come back to power.
You 13 i
have LaSalle, Pilgrim and TVA units -- those are juut the (ss 14 ones that come to mind.
How many units are down at the 15 moment and cannot come back to power without careful 16 reviews and so forth by the NRC Staf f?
17 MR. JORDAN:
Besides the ones you mentioned, 18 Davis-Besse, Rancho Seco, San Onofre.
You mentioned the 19 TVA.
20 MR. MOELLER:
It is close to 10 percent of the 21 total.
That is not bad or good.
I am just asking.
22 MR. EBERSOLE:
Okay.
We have a new one.
23 MR. ALLISON:
The next pre sen ta tion is the 24 snubber failures.
Terry Chan will present that.
25 MR. CHAN:
I am the project manager for Trojan.
O ACE FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
27072.0 45 REE NJ/N, 1
Wha t I am going to discuss today is failure of certain 2
snubbers at Trojan, specifically the steam generator 3
hydraulic snubbers locking up when not desired.
The 4
concern here is that overstressing of the RCS piping could 5
occur, and we feel, did occur.
I wi'. cover the background 6
of how that determination was arrived and some follow-up 7
activities which have been performed in order to assure 8
that the RCS piping is still adequate to maintain its 9
function.
10 Excuse me for making an impromptu change of a 11 slide, but I think the slide that I hava up here now 12 provides the overview of the issue more succinctly than the 13 copies that you have.
A(_)
14 In February of 1985, Trojan was issued the 15 snubber surveillance inspection tech specs for large bore 16 snubbers.
In April, Trojan went down for its ref ueling.
17 All 16 steam generator snubbers were inspected at that time.
18 Two of the snubbers were te sted ; both failed.
Conse quen tly, 19 based on the management decision by PGE, all 16 steam i
20 generator snubbers were declared inoperable.
21 The se failures were attributed eventually to 22 restrictive acceptance criteria for the control valves.
23 All 16 snubbers were disassembled, inspected, and 24 reassembled.
They were then retested, found to be 25 acceptable and then they were placed back in service.
O ACE. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 46 REE
['/)
1 The plant then restarted for the '85-86 cycle.
s_
2 The basis for coming back up to power and the 3
acceptability of the snubbers was that during the 4
disassembly, marks were evident within the snubbers which 5
indicated that the snubbers had exhibited motion.
And 6
based on that decision, it was not concluded at that time 7
that any of the snubbers had actually locked up during the 8
cycle.
9 MR. MOELLER:
You are saying you could tell that 10 they had moved.
11 MR. CHAN:
The licensee made that assumption 12 based on markings within the cylinder walls of the snubbers.
13 MR. EBERSOLE:
No external markers?
O
\\l 14 MR. CHAN:
Not that I am awa re.
15 MR. EBERSOLE:
Okay.
I wasn' t going to ask any 16 questions.
17 MR. CHAN:
During the April '85 outage, a hot 18 leg to steam generator B pipe whip restraint to lateral 19 support member was found pulled from the wall at about 20 5/8th of an inch.
A connection at that time was not made 21 between that deficiency and the possibility of locked 22 snubbers.
23 Now, during this time period, and since 1982, 24 the Trojan had observed erratic pressurizer surge line 25 movement and over the past three years had been monitoring O
v ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
27072.0 47 REE
N 1
this motion in order to determine its cause.
In the fall 2
of 1985, the licensee hired a consultant to evaluate the 3
pressurizer surge line movement and was -- the consultant l
concluded that in order for such a movement to occur that a 4
5 rotation of a certain amount, one-third of a degree at the 6
hot leg pressurizer surge line connection to the hot leg, 7
would have had to occur.
8 When the consultant was presented with the fact 9
that two steam generators did not pass acceptance te s ts 10 during the snubber testing program, in addition to the 11 j
horizontal member of the pipe having been pulled from the 12 wall, it was then concluded that the surge line movement i
13 was in fact attributable to the locked snubbers, locked 1
x l
(
14 steam generator snubbers, and that the pull out of the 15 l horizontal member of the pipe whip restraint was also a 16 i
direct result or a direct effect.
17 Following up on that analysis, it was determined 18 that overstressing, under the worst-case condition, i
19 assuming the steam generator snubbers were locked f rom the 20 l
cold position at the beginning of '85-86 cycle, that 21 overstressing of the hot leg elbow to the B steam generator 22 could have occurred.
23 In April of 1986, these same 16 steam generator 24 snubbers were reinspected.
The licensee had committed to 25 this reinspection in their LER.
Inspection of these g)
(
ACE-FEDERAL REPORTERS, INC.
l 202-347-3700 Nationwide Coverage 800 33M546
27072.0 48 REE kJ) 1 snubbers revealed -- or functional testing, excuse me -- of 2
these test snubbers revealed that 11 of the 16 snubbers did 3
not pass the functional test acceptance criteria.
4 That determination of failure of the snubbers 5
was attributed to inadequacies in the design of the control 6
l valve.
7 Because the Region and the Staf f were concerned 8
about the possible overstressing of the RCS piping, the 9
Staff requested that several follow-up actions be performed 10 on the piping in order to assure its soundness.
11 The PT test was performed on the steam generator 12 elbow -- excuse me, elbow to pipe weld.
No indications 13 were found.
The licensee, NRR and the Region walked down 14 the portions of the RCS piping and observed some evidence 15 of restrained thermal growth.
16 The UT testing was also performed on all four 17 hot leg elbows.
No indications were found.
18 The snubber control valves were replaced with 19 one of a new design.
This decision had been made by the 20 licensee in 1985, based upon recommendations from the 21 company which tested its snubbers.
22 And as a follow-up, prior to restarting for the 23
'86-87 cycle, the licensee is going to monitor the thermal 24 growth of the RCS during heat-up and during operations to 25 assure that the predictad thermal growth and clearances are O
ACE. FEDERAL REPORTERS, INC.
M-347 3700 Nationwide Coverage 800-336-6646
27072.0-49 REE
(~}
l all acceptable and, in fact, that the assumption of the
\\.s 2
locked snubbers causing the erratic pressurizer surge line 3
movement and the damage which was observed on the pipe whip
^
4 restraint has been resolved.
5 In addition to that, NRR is reviewing the 6
licensee's stress and fatigue reports to assure that the 7
integrity of the RCS piping is intact.
8 MR. ETHERINGTON:
What is the maximum general 9
membrane stress calculated to be?
10 MR. CHAN:
I believe our engineering branch can 11 respond to that.
12 MR. TORREAU:
What was the question again, 13 please?
)
14 MR. ETHERINGTON:
What is the maximum general 15 membrane stress as a result of this lock-up?
16 MR. TORREAU:
Okay.
As a result of this lock-up, 17 the type of stresses that we see in the pipe are secondary 18 stresses.
These are basically characterized as a 19 self-limiting type of stress.
20 MR. ETHERINGTON:
Secondary stress?
21 MR. TORREAU:
Due to restrained thermal growth, 22 yes.
So the stress, the ASME Section 3 design allowable. --
23 I emphasize the word " design" -- has an equation, equation 24 12 of NB 3645.5 -- NB 3645 -- NB 3653.6 equation 12, limit 25 of 3 SM.
('~T
_/
i i
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800 336-6646 s
27072.0 50 REE f'/')
1 That limit for secondary stress is for design s_
2 and it is intended to preclude thermal ra tche ting in the 3
pipe ; in other words, collapse due to incremental loadings.
4 The question here is the applicability of that 5
equation, design equation to an unanticipated event.
6 Clearly, from the analysis, they did not meet this equation.
7 But what the Staf f and what the licensee proceeded to do 8
was to evaluate the strain ef fects in the pipe due to 9
exceeding this equation.
And so we did not look at it from 10 a stress standpoint, but more from a strain standpoint.
11 From a strain standpoint, we determined that it was a 12 maximum of about.8 percent stra in in the pipe elbow.
13 MR. SHEWMON:
So at that level, most of this 14 strain is plaster?
15 MR. TORREAU:
At.8 percent strain, it is in the 16 nonelastic region for this material, 17 MR. OKRENT:
I would like to understand a little 18 bit about the snubbers themselves.
As I look at the
)
l 19 Vugraph, it suggests that the reason tha t these were tested 20 was because there were snubbers technical specs for the 21 first time?
22 MR. CHAN:
That is correct.
23 MR. OKRENT:
When did Trojan start running?
24 MR. CHAN:
I believe it was 1975.
25 MR. OKRENT:
Why is it that the operator would O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 336-6646
27072.0 51 REE 7 'x 1
not have examined the snubber opera tion on his own 2
initiative?
I am missing something.
3 MR. CHAN:
The snubbers were inspected, not 4
f unctionally tested.
They were inspected for seal 5
conditions and fluid levels.
I understand the licensee to 6
say that the main emphasis during that particular time 7
period for them was that snubbers had problems with seals 8
and, thus, leaking fluid.
9 MR. OKRENT:
So this is the first time that this 10 type of failure of hydraulic snubbers was observed in any 11 plant?
l 12 MR. KIESSEL:
I am Dick Kiessel, with the Office 13 of Inspection and Enforcement.
To elaborate a little bit pq
\\_-
14 more on why weren't these snubbers tested earlier, during 15 the '70s, when the initial tech specs were issued on 16 hydraulic snubbers, they called for a visual examination of 17 all snubbers and a functional testing of a sample of 18 snubbers.
They also included an exemption for any snubbers 19 that were in dif ficult or unusual places, and they also had 20 a size limita tion.
Anything over 50 kips was not required 21 to be tested.
22 Part of the rationale at that time was that i
23 there were just not sufficient facilities to perform i
24 adequate testing on the large snubbers.
25 l
In November of 1980, NRR issued a generic letter
/
V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
27072.0 52 REE I
which included revised tech specs.
These included two 2
provisions:
first of all, they took out that 50 kip limit; 3
and, secondly, they modified the campling plans to allow 4
several different forms of plans.
5 It is my understanding that Trojan had not 6
resubmitted their tech specs until late in 1984-85.
So, 7
therefore, this was the first time that they had been 8
te s ted.
9 There have been a number of other instances with 10 the large steam generator snubbers.
There have been 11 instances of the f a ilure to lock up and also of their being 12 continually locked up.
So that there have been a number of 13 other incidences, and Trojan is not the first to have
(")
' _J 14 encountered the problem with steam generator snubbers.
l 15 MR. OKRENT:
I just have one other question.
If l
instead of 1980 it were 1986 when you were to issue your 16 17 generic letter on the testing of these larger snubbers, on 18 what basis would you have demonstrated a cost / benefit ratio l
that justified the added expense?
19 20 !
MR. KIESSEL:
I am afraid I can't answer that i
21 one.
I am not familiar with NRR's cost / benefit procedures.
22 MR. EBERSOLE:
I want to --
23 MR. OKRENT:
Is there anyone in the Staff who 24 can answer that?
I would appreciate learning at some 25 future meeting just what the Staf f would have done with
<m
\\ s)
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336-6646
27072.0 53 REE
y 1
that generic le tte r in 1986, assuming they had not been
\\mf 2
able to test them be f ore so.they didn' t know their behavior.
3 Okay?
4 MR. ALLISON:
We will try to get an answer.
5 MR. EBERSOLE:
Let me ask you, isn't there a 6
generic requirement that snubbers shall not respond to 7
release of static friction lock-up which is what occurred 8
I here?
I 9
MR. KIESSEL:
To the best of my knowledge, there 10 is no such requirement unless it is written into an 11 l
individual design specification.
12 MR. EBERSOLE:
Isn't this what happened here?
13 MR. KIESSEL:
I can't say what was the
(~'N
(_)
14 l
initiating mechanism.
The licensee has attributed the l
15 l
locking up of the snubbers to the very, very sensitive i
16 lock-up velocity that they have which was measured in mills i
17 per minute as opposed to the normal convention, being 18 measured in inches per minute.
19 MR. EBERSOLE:
Pipes, if they are hung, if they 20 release, sometimes it jumps.
21 MR. KIESSEL:
I believe the licensee is of a 22 feeling that they were -- that the control blocks were so 23
- sensitive that almost any motion, whe ther it be just even a 24 gradual thermal growth, without even having to go into the 25 jump, could have caused them to lock up.
n ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646 m.-
~
~
27072.0 54 REE
()
1 Now, I think by way of clarification, " lock-up" 2
is used in two different conditions, when talking about a 3
snubber:
Lock-up in one cose refers that the main poppe t 1
4 has closed and, there fore, you are in the bleed range; and 5
also lock-up can mean that there is no flow at all.
6 MR. EBERSOLE:
Yes.
7 MR. KIESSEL:
For that reason I like to use 8
" activation," meaning that the poppet has closed; " lock-up" 9
meaning that the snubber has gone to a rigid strut.
10 MR. EBERSOLE:
With this potential to lock up, 11 isn't the snubber a detriment to sa fe ty?
12 MR. KIESSEL:
I really can' t answer the question 13 because I don't know both sides of the equation.
14 MR. EBERSOLE:
How many cycles do they go 15 through?
16 MR. KIESSEL:
I believe they indicated 30.
17 MR. TORREAU:
The exact number is 28.
l 18 MR. EBERSOLE:
You are not unhappy with that?
19 MR. TORREAU:
I am not unhappy with the 28 20 number, no.
21 MR. EBERSOLE:
Okay.
22 MR. CHAN:
May I say that the arrangement of the 23 snubbers and their hydraulic lines at Trojan is such that 24 the design intent is that if one snubber locked up, that i
25 would not preclude or -- if the flow would not go through l
O ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverase 800-336-6646
27072.0 55 REE
(}
l the check valves, which would activate that particular 2
snubber, that if that was not operating properly, that the 3
four etcam generator snubbers are arranged in a parallel 4
fashion such that the necessary fluid flow to permit motion 5
could be transferred to the remaining three snubbers; thus, 6
a failure of one snubber would not prevent the motion 7
necessary.
8 MR. EBERSOLE:
Any questions?
We are running l
9 out of our limited time.
10 MR. ETHERINGTON:
How many heat-up and cooldown 11 cycles has the project experienced in this lock-up position?
12 MR. CHAN:
Since it is unclear when the snubbers 13 might have locked up, their analysis -- their analysis n
A/
14 assumes 28 cycles.
15 MR. ETHERINGTON:
And is that used up in any 16 cumulative T factor?
17 MR. TORREAU:
At a 1 percent strain in the pipe,
18 if you use the ASME code for the T design curves, you would 19 end up with an allowable design cycle of 366.
20 MR. ETHERINGTON:
How many?
21 MR. TORREAU:
366.
l 22 MR. EBERSOLE:
Are there any generic letters to i
i 23 warn people of this?
24 MR. CHAN:
NRR and I&E are working together on l
25 evaluating the generic implications of this occurrence.
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4 336-6646
27072.0 56 REE
(}
l MR. EBERSOLE:
And these are used in how many 2
places?
3 MR. CHAN:
As I understand it, Trojan was or is 4
one of the very, very few plants which utilize this 4
5 previous control valve.
As I'have said, most of the --
6 Trojan has changed out the control valve to a new design, 7
as have many other utilities.
8 MR. EBERSOLE:
Any further questions?
l 9
Mr. Chairman, do you want to hold this last i tem,
.I 10 which is very significant, until the next meeting or what?
11 MR. WARD:
I think we could do that.
Could you 12 give us personally a five-minute summary of it.
13 MR. EBERSOLE:
Yes.
Ye s.
Wha t we found is one 14 plant at least, if not several, on boiler design still 4
15 retains a system which we call a loop selection system.
It 16 is a system whereby they found that when you have a large 17 size break on one side of the reactor -- it is just a 18 I two-flow system -- that you have problems getting enough 19 water in for low-pressure injection.
So the y, the key 20 people decided they would put a cross die in with the valve 21 and they would elect to send the water from the 22 hypothetical three pumps towat~d the one but not toward the 23 empty loop.
24 This necessitates an extremely f ast transient DP 25 gradient.
One of the early findings was that the impulse O
ACE-FEDERAL REPORTERS, INC.
202 347-3700
>!ationwide Coverage 800 336-6646
27072.0 57 REE
()
1 line to detect this quick gradient and pressure was wrapped 2
around the rear pipes that were supposed to do the 3
measuring, so you couldn't see anything anyway.
The upshot 4
was --
5 MR. WARD:
This assumes a drastic double-ended 6
guillotine large break LOCA.
7 MR. EBERSOLE:
That is correct.
Which was the 8
basis of the design.
So the signal generation system would 9
never have worked anyway.
And in the course of that 10 finding and subsequent findings that the low-pressure 11 course break system in fact did more than spray, it also I
12 constituted an inventory makeup function as well, it was 13 decided to lock out the selection logic, the loop selection O
14 logic at that plant.
And I thought the staff had locked 15 them out everywhere.
It comes as a shock to me to find out 16 they are still in existence.
17 They found that they are still in existence and 18 in their presence there is a potential for certain valves 19 to close.
That puts all four of the RHR pumps against a 20 closed discharge which fills them all, leaving you without i
21 core cooling pumps, on a single failure malfunction.
22 The RHR pumps in that configuration can be 23 locked up witn zero flow and grind themselves to a state of
)
24 inoperation permanently to enable the plant to move the 25 heat from the core or not.
O ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336-66 4
27072.0 58 REE (J'T 1
MR. ALLISON:
I guess we actually were briefed 2
in the single failure problem rather than the presence of 3
the loop selection logic.
4 MR. EBERSOLE:
If it is retained.
On the other 5
hand, you better look at that selection logic design and 6
see if you are ever going to get to see one.
7 How many plants are there like this?
Do you 8
know?
9 MR. HODGES:
Today all of the BWR 3 plants, 10 including Dresden 2 and 3, Millstone 1, Pilgrim, Monticello 11 and Quad Cities 1 and 2, plus two BWR 4 plants, which is 12 Duane Arnold and Fermi 2, have them.
But not Browns Ferry.
13 MR. MICHELSON:
Don't you ever wonder why?
14 MR. EBERSOLE:
Why is it?
Why didn't that t
15 I
action propagate to the se o the r --
16 MR. HODGES:
I presume -- I don't know all of i
17 the thoughts that went into the decision process, but they 18 were able to demonstrate that with the logic as it exists, 19 they satisfied the regulations.
They satisfied Appendix K.
20 They showed that for breaks down to a.1-square-foot size 21 break, the loop selection logic worked and selected a 22 proper loop.
If it didn't select a proper loop for smaller j
l 23 breaks, they st.ill didn' t go above 2200.
So it satisfied 24 the regulations and they continued on.
25 MR. MICHELSON:
You realize the system works O
ACE FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverasa 800.336-6646
27072.0 59 REE
(~')
I fine on paper.
It is in the hydraulics of the actual
'x_-
2 operation that the difficulty arose.
The system has such a 3
high hydraulic noise level that at any point in time it 4
cannot detect which side is broken.
Just tell your 5
inspectors to go look, go to the panel, look at the readout 6
of the dif ferential pressure.
There is a local gauge there, 7
and ask them if that thing could ever work.
8 What it is, we first got wind of it when it kept 9
breaking the gauges.
So people got smart.
They just 10 snubbered the gauges.
Well, that de feated it comple tely, 11 because the decision, the logic decision has to be made in 12 about a half a second.
And it wouldn't even respond if you 13 properly hydraulically filtered it, it would no longer even
(
14 respond.
15 MR. WARD:
But in a large break, you should get 16 a signal.
The fact that you have noise in normal operation --
17 MR. MICHELSON:
The problem is, the logic makes 18 its decision in about a half a second.
19 MR. WARD:
Okay.
1 20 MR. MICHELSON:
What the dif ferential pressure 21 is at that point in time is totally irrelevant to what is 22 happening in the system.
If the gauges are so far off l
23 range, they don' t even get back to range to record the 24 dif ferential pressure condition in time.
The timing has to 25 be so f a st.
(')
V Act.. FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage '
800-336-6646
27072.0 60 REE
()
1 MR. EBERSOLE:
It was pretty much like the 2
reactivity control system.
3 MR. HODGES :
I won't defend it as being an 4
uncomplicated design.
5 MR. EBERSOLE:
I want the Staf f to tell me they 6
will go fix it at all these other plants.
7 MR. MICHELSON:
It might make an interesting 8
case history if somebody wonders why the industry as a 9
whole got into some of the situations it got into and why 10 the vendors didn't seem, even though a problem arose and 11 was fixed on a given plant, they seemed uninclined to go 12 ahead and make sure that the industry as a whole fixed the 13 problem.
b' s/
14 MR. KERR:
Aside from history, what should we be 15 doing now?
16 MR. MICHELSON:
I wouldn't worry about it.
17 MR. KERR:
We don' t need to do anything?
18 MR. EBERSOLE:
Would this justify a letter from 19 us?
I 20 MR. WARD:
Do you want to ask the Staf f to 21 justify its position and not require that this system be 22 changed in these other -- this list of plants?
23 MR. EBERSOLE:
We have found out that this thing 24 has been laying like a snake in waiting.
25 MR. HODGES:
The Staff has known about it for
()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 61 REE
()
1 some time.
This is not a surprise.
2 MR. WARD:
You don't have any intention to do 3
anything about it or you don't see it as a problem?
4 MR. HODGES:
We are not proposing to go back and 5
tell these plants they have to change.
6 MR. EBERSOLE:
You are going to leave the single
]
7 failure vulnerability in place?
8 MR. HODGES:
No.
9 MR. EBERSOLE:
Or are you going to go at another l
10 gizmo to protect against this single failure mode?
11 MR. HODGES:
I am not going to tell you what the 12 fix would be.
I don't know.
13 MR. EBERSOLE:
Of course, you can patch it.
O 14 MR. JORDAN:
W'e indicated to all the plants that 15 are susceptible with a bulletin asking them to assess their 16 single f ailure vulnerability and to tell us what they plan 17 to do.
They have activated an owners group to try to come 18 up with a solution.
19 MR. EBERSOLE:
Can you estimate how they 20 anticipate detecting this?
21 MR. JORDAN:
The loop selection portion of it?
22 MR. EBERSOLE:
The dynamics of the process?
23 MR. MICHELSON:
Make sure they really go back 24 and look at the dynamics, the -- look at the steady-state 25 dynamics of the situation and then ask if a break should O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 62 REE
()
1 suddenly occur, will that equipment be able to record the 2
break effects.
3 MR. JORDAN:
We issued the wrong bulletin.
4 MR. EBERSOLE:
Remember, the alternative 5
performance of this is that all the water gets poured out 6
on the floor.
7 MR. MICHELSON:
It may be that a lot of things 8
have happened in the last few years in terms of the amount 9
of water required for flooding.
Maybe they can now take a 10 five-second delay, in which case you could snubber the 11 instruments and wait for the situation to settle down to a i
12 true break situation.
13 MR. WARD:
Do you understand the question?
14 MR. JORDAN:
I do.
15 MR. WARD:
Are you willing to respond to it?
16 MR. JORDAN:
We will respond to the question.
17 MR. EBERSOLE:
Then we will see you at the next 18 meeting.
19 MR. WARD:
Let's come back at a quarter to.
l 20 (Recess.)
21 MR. WARD:
Our next topic is a discussion of the 22 source term for power plant accidents.
Dr. Kerr.
23 MR. KERR:
You will recall that in December, the 24 Committee wrote a letter commenting on a draf t report 25 entitled " Reassessment of the Technical Bases for O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
I 27072.0 63 REE
()
1 Estimating Source Terms, NUREG-0956."
This report had been 2
preceded by a number of Subcommittee meetings on which we 3
had reviewed the material in that draf t report.
A letter 4
made a number of comments on the report.
Since that time, 5
the Str f f has received public comments as well as comments 6
from the ACRS and has made a number of significant changes 7
in the format of the report.
Most, if not all of you, have 8
received a copy of what is called " Review Copy of 9
NUREG-0956" dated May 23rd.
A Subcommittee met with the 10 responsible group on Tuesday of this week, and it was the 11 consensus of the Subcommittee that significant improvements 12 had been made in the report.
13 I asked the responsible group in making its 14 presentation to the Committee to go over the comments that 15 were made in our letter and attempt to say how they had 16 been dealt with in this revised version.
That was the 17 approach that I suggested.
It may be the approach that 18 they will use.
We will see.
19 Mr. Silberberg is the cognizant Federal employee,
I will turn things over to him.
21 Incidentally, we have been asked to write a 22 letter at this meeting if possible, commenting on this 23 draft which purports to be a final draft of the report.
24 MR. SILBERBERG:
Thank you, Dr. Kerr.
We do 25 appreciate the opportunity to come before the full O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 64 REE
()
1 Committee today to assist you in any way that we can with 2
this important task you have before you; namely, the 3
preparation of a letter on the final report, NUREG-0956, 4
Since we only have a short time this af ternoon, 5
we would like to accomplish two things.
First, I want to 6
briefly place the state of progress on source term 7
technology in perspective, regarding the implementation of i
8 the Severe Accident Policy Statement.
Following me, 9
hopefully my remarks will be brief, just to maybe focus the 10 Committee, Jocelyn Mitchell will walk through the ACRS 11 letter as you had recommended, December 12, 1985, on the 12 draft report, and describe how we have responded to your 13 comments by revisions in the report and in continuing to 14 redouble our ef forts in our research to address other i
15 concerns that you note in the letter on a somewhat 16 longer-term basis.
17 We have taken your comments seriously and we 18 hope that you will find that we have been responsive.
I 19 thought it would be useful to briefly note for you the 20 major changes that have been made in the final report that 21 Dr. Kerr referred to.
I won' t go into any detail, just to 22 make you briefly aware of them.
23 The most significant change that we made was to 24 add a considerable amount of additional technical 25 information directly in the report to deal with the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverase 800-336-6646
27072.0 65 REE
()
I technical basis behind the source term information, the 2
source term science and methodology so that we could make 3
it very clear that the reassessment in 0956 was more than 4
just a code suite but indeed had a considerable supporting 5
basis that already existed and is continuing to emerge from 6
the research program.
7 We spent considerable time in the document 8
describing the upgraded code suite, whereas we only had 9
just briefly, the draft report had just briefly done,
10 mentioned some of the features of the upgraded code suite.
11 We now have the STCP.
12 We also presented analyses of additional 13 sequences that had been performed with the source term code 14 package for the NUREG-ll50 analysis.
We had removed the 15 chapter on risk and containment based on the advice that we 16 had received from our comments; have a complete appendix on 17 how we responded to the public comments in some detail; 18 provided a chapter that reflects the revision to the severe 19 accident research plan of NUREG-0900, now called Revision 1, 20 which had been going on sort of in parallel as we went 21 through the public comment period and to the public comment 22 response.
Finally, we have an improved statement of 23 conclusions.
24 Now, five years ago, the Staff issued NUREG-0772 25 recommending the need for a systematic coupled analysis of O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
-27072.0 66 REE m
1 fission product release and transport phenomena in severe 2
accidents, along with 'the development of a supporting data 3
base on severe accidents and source term phenomena.
4 We have now come to this point where we believe 5
that the major advances that have been made in this 6
technology, since WASH-1400, with most of it coming in fact 7
in the last five years, has been made, and an important 8
part of this progress is the development of an analytical 9
approach to source term estimation, which we refer to as 10 the source term code package.
11 Let me briefly display a list of some of the 12 advances, more important advances that we have made in 13 source term technology, which is basically table 1 in the O
14 executive summary.
We are showing here the areas of l
15 improvement, for all of the processes from initial heat up 16 of the reactor to --
17 MR. KERR:
Mr. Savio, could you cut the --
18 MR. SILBERBERG:
-- to progressively higher 19 temperatures from release of fission products from the fuel, 20 transport and deposition within the primary system within 21 the containment, within the engineered safety features of 22 suppression pools and ice compartments, all the way to 23 interactions with core concrete, with concrete following 24 postulated melt-through; and, in the case of loads, 25 pressure and temperature loads which impact the containment.
m.
(V ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 67 REE
()
1 In all of these areas, both mechanistic 2
treatment modeling and a considerably enlarged data base 3
has provided us with the basis for saying that we feel the 4
progress has been very significant and the advances have 5
been major.
6 MR. MOELLER:
I notice that a number of your 7
major advances are, in a sense, duplicated when you list 8
the areas of research needed.
9 MR. SILBERBERG:
Yes.
And this has been 10 commented on a number of times.
And we don't feel it is, 11 Dr. Moeller, that it is strange, because in going into --
12 getting a deeper understanding of severe accident phe nomena 13 and working in the areas that heretofore had not been
/~'s
\\-
14 addressed, either from a modeling standpoint or from an 15 j
analysis standpoint, that going deeper now and uncovering, 16 in a sense, what the true uncertainties were in this area 17 gave us the information to stand back and say, yes, we have 18 made progress and we understand a lot about this area.
But 19 there is still some gaps in tough places.
A lot of the 20 places I have listed are pretty tough, because they are the 21 subject of the research that we are now focusing on.
So we 22 realize this is a dichotomy, but yet we are comfortable 23 with that, with the fact tnat we have both of the se.
24 MR. MOELLER:
Thank you.
25 MR. SILBERBERG:
Now, in looking at this O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
27072.0 68 REE
{)
1 progress, the important thing to note is that the -- we 2
have gained considerably, in our understanding of severe 3
accident phenomena.
4 Now, as I noted to Dr. Moeller, tha c we do 5
recognize that more important work remains to be done.
6 Further improvements can be expected from the research tha t 7
is now in place.
But it is important to note that the 8
analytical approach that has been done provides a framework 9
for accommodating improvements that are dictated by the 10 additional information and insights that we will get from 11 this research that is going on now.
So, in ef fect, we have 12 a way of continuing to communicate with the information 13 base that is developing and to allow us to make the
(
14 appropriate changes and modifications that allows us to i
15 benefit from an even deeper understanding of it.
l 16 Now, we believe that we now have a good 17 technical basis to move forward by contrast with the 18 current bases; namely, TID 1984 and 1984-4, 25 years old 19 and WASH-1400, 1974, or basically 12 years old.
Now, we 20 believe that the current methods, from a phenomenological 21 standpoint, are outdated, and we believe the time has come i
22 to use the new source term information to reevaluate 23 regulatory practice as mandated by the Severe Accident 24 Policy Statement.
25 Now, I must bring to your attention the fact i
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
27072.0 69 REE
()
I that the process of using the new source term information 2
and, in particular, the source term code package, has 3
already started.
The first application has been in the 4
analyses for the NUREG-1150 risk rebaselining study now 5
coming to conclusion.
6 The second application deals with the 7
implementation plan for the Severe Accident Policy 8
Statement and the regulatory use of new source term 9
information.
This plan on source term-related changes in 10 regulation was the subject of an ACRS letter on March 18, 11 1986.
The implementation plan, as noted in SECY '86-76, 12 clearly spells out the use of the information and 13 technology that is described in NUREG-0956 and its 1,4 accompanying references.
I 15 l Thus, NUREG-0956 is an important element of the 16 process of moving forward in examining current regulatory 17 practice with respect to source te rms.
18 Let me just close by saying that we hope that 19 you will also share our belief that we can move forward, 20 while at the same time, providing us with any valuable 21 advice on areas needing additional emphasis and attention 22 as we proceed in our use of the technology.
23 Thank you.
24 MR. KERR:
Thank you.
25 MR. MOELLER:
This is probably a detail,-but I O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4336-6646
27072.0 70 REE
()
1 am curious, on page -- if I may cite a specific page, 1-5, 2
the second full paragraph, I just needed a clarification.
3 It lists -- it says, " A considerable amount of Staf f ef fort 4
is being devoted to two technical issues that have 5
particular significance to source term questions and to the 6
resolution of regulatory policy matters.
The se are, one,
7 phenomena that can lead to early gross failure of the 8
containment."
I understand that.
But the second one I 9
don't understand.
It says, "The potential for containment 10 leakage that would prevent major structural failure -- the 11 leak-before-break hypothesis."
12 Maybe I have been too narrow, but I had always 13 looked at the leak-before-break hypothesis as piping breaks, C) 14 not contaimment breaks.
15 MR. SILBERBERG:
Dr. Moeller, perhaps that could 16 be clarified further by improving the text.
You are 17 correct.
We have nominally discussed leak-before-break in 18 the past always in the context of piping.
However, the 19 conta inmen t loads and containment performance working 20 groups that performed in parallel with the source term work 21 back in 1983-1984 focused considerable attention on just 22 that concept or that premise, which said, for example, 23 might there be failures of penetrations or seals or 24 something other than the basic structure that would in f act 25 provide a leakage, suf ficient leakage to compensate for the O
\\
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 71 REE
()
1 rate of rise due to introduction of loads.
2 But I might say that that issue is still open 3
and is the subject of continuing research to look, where, 4
in fact, specific studies are now being done by our 5
division of engineering and technology, to look at 6
penetrations and seals to determine if, indeed, that case 7
can be made and under what conditions.
8 MR. SHEWMON:
Dade, it is my impression that 9
that was a very strong argument back when the gas reactor --
10 any time you get a steel-reinforced concrete structure that 11 would sort of come apart and start leaking out while the 12 seal still held much of it together, the metal may well 13 burst.
O 14 MR. MOELLER:
Thank you.
15 MR. KERR:
Anything else?
16 (No response.)
17 MS. MITCHELL:
I have provided a second copy of 18 the December 12, 1985 letter which Dean has passed out, 19 which has numbers down the side, and if we are going to 20 walk through a letter, I thought it was appropriate.
This 21 way we can all speak about exactly the same paragraph by 22 the numbers.
I will go through them pretty much starting 23 at the beginning and ending up at the end.
I won' t spend 24 very much time on some of them and I will spend a little 25 more on others.
The first paragraph --
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
l 27072.0 72 REE
()
1 MR. CARBON:
I compliment you.
2 MS. MITCHELL:
The first paragraph is a history 3
and we agree with it.
4 The second paragraph sta te s that fission product 5
release and transport is only a piece of the in forma tion 6
which is going to be needed in the future when you are 7
looking at restructuring existing regulations.
We agree.
8 Page 4 of the ACRS letter discusses what the 9
ACRS believes are the other pieces of the puzzle and we 10 will come to them when we get to page 4.
We believe that 11 the other pieces that you need at any time are dependent on 12 the use.
Therefore, our list would be the use and then one 13 or more of the following:
the event f re quencie s, the p/
14 containment matrix frequencies, consequence calculations; x-15 and, very importantly, the uncertainties.
For example, to 16 review the use of spray additives, you don' t need a very 17 good estimation of the event f re quencie s.
18 Paragraph 3 asks whether we find a significant 19 dif ference compared with the reactor safety study.
We are 20 less ambiguous than before in the draf t of the report.
We 21 now state that we believe there are not large systema tic 22 reductions.
We do not see systematic ef fects that are 23 applicable to both plants, that is, Surry and Peachbottom, 24 all sequences and all chemical element groups.
25 Paragraph 4 notes that no guidance on estimating OV ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
27072.0 73 REE
.i 7_.s()
1 uncertainties was given in the draft.
We agree.
We tried 2
to do be tter in the draf t of the final report that you have 3
and we learned on Tuesday that that was probably not enough, i
4 So we have people going back now and we will make further 5
changes in both chapter 4 and in chapter 3 and probably add 6
a very small paragraph in the executive summary to provide 7
as much guidance as we can to the three major sources of 8
uncertainty information.
9 Wha t do we know today on uncertainties?
In the' 10 QUEST study, we developed a method for scanning a 11 calculation for the sensitivities for a particular 12 application.
In some cases, you might need the source 13 terms and you would start at the end of the calculation and 14 work backwards.
In other cases, you might need 15 intermediate results instead of the bottom line, and you 16 would start at that part of the' calculation and work 17 backwards, finding the sensitivities for your application.
18 In QUEST we also start on a method for 19 developing the parameter ranges for NUREG-ll50, which is 20 the first application of the new source term technology.
21 We are developing a method of numerical uncertainty 22 estimates that is based on the judgment of the code 23 developers and the experimentalists, those people who today 24 know the most about the subject.
And very importantly, in 25 the NUREG-ll50 arena, they are documenting the bases used O
i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 74 REE
(}
l by each of the experts.
They take a tape of the discussion.
2 They have a very detailed discussion with each of the 3
experts expressing their opinion and the reasons for it, 4
and all of that will be documented and will be a very good 5
source of insights.
6 In the QUASAR study, we are defining a 7
structured numerical uncertainty method based on the use of 8
the source term code package.
We look forward to a new 9
code coming along called BELCOR which will make that 10 particular effort a lot easier.
Therefore, in the existing 11 studies we have a broad but not deep evaluation and also 12 the deep but not broad view of evaluations.
13 We envision that each new application will 14 l
probably have to be adj usted on an ad hoc basis.
That is, 15 the fourth uncertainty study will depend on insights from 16 the first three.
And the fifth study of uncertainties will 17 depend on the insights that it got f rom the first four.
18 Over a period of time we will develop a method that is 19 developed by pulling ourselves up by our bootstraps.
20 Paragraph 5 mentions that the containment 21 methods that were used in NUREG-0956 were preliminary.
We 22 certainly agree.
We have made changes to clarify the 23 status of containment behavior. as it is used in the fission 24 product release and transport methodology.
We evaluate 25 loads and we input response.
That is, containment failure O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
~
~lr~
27072.0 75 REE
(}
l criteria are determined external to the fission product 2
release and transport and in their input to it.
3 NUREG-1150 is developing quantitative 4
containment behavior for their risk analysis.
A change in 5
the methods between the draf t and the final report of 6
NUREG-0956 is in the development of the source term code 7
package which removed inconsis tencie s in the way previous 8
calculations were done in that a code called CORCON was 9
used to evaluate fission product ef fects and enter this 10 routine to calculate loads.
And t he fact that that 11 inconsistency was present is the reason that there was a 12 caveat in the draf t concerning our calculation of the loads.
13 We have a large experimental program in this area.
m i
14 Paragraph 6 recommended that we discuss the 15 comparison between the conclusions.
We have added a great 16 deal of information about the IDCOR program and technical 17 differences with the NRC.
We added a little.in chapter 2 18 with the history.
Chapter 5 goes through the technical 19 issues.
It is backed up by a lot of information in 20 Appendix B.
We note that sequence definition is a very, 21 very important area, and that a lot of the sequences where 22 they were nominally the same are actually significantly 23 different in their definition.
24 The first comparison that we made. trying to 25 mimic our understanding of the MAT code, the industry code O
%J ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 76 REE
().
I that IDCOR uses was unfortunately not very illuminating and 2
we now state that our discussion and evaluation of the 3
industry efforts is limited.
Paragraph 7 requires no 4
comment.
5 Paragraphs 8 and 9 are a history and we agree 6
with them.
7 Paragraph 10 notes that the Staf f said that 8
NUREG-0956 contained a ncientific bases for source term 9
estimation.
We have made extensive changes in the report 10 to make that first sentence come through.
We believe now 11 that there is some information, a significant amount of it, 12 on the scientific bases.
13 Chapter 3, we have added comparisons of the
\\/
14 codes or models with experiments, and chapter 6, which we 15 now call "Research" instead of " Future Research," contains 16 a description of the status of the present research and 17 also what is upcoming in the future.
18 Some of this information was previously 19 incorporated by reference, but now made it explicit.
20 Paragraph 11 is a f act and we agree.
Paragraph 21 12 contains two thoughts.
The first is that the draft is 22 tentative about the comparison with the reactor safety 23 study on risk.
We removed the chapter on risk and the 24 appendices that supported that chapter.
It was based on 25 very unsatisfactory information and we felt that it was, 7sd ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 77 REE p)
(,
1 being superseded very rapidly by NUREG-1150 and there was 2
nothing to be gained by putting it in.
What we do believe 3
about source terms we have discussed in paragraph 3.
4 The second part of 12 questioned what we meant 5
by the suf ficiency of the five referenced plants analyses.
6 We revised the text to be a little bit more clear.
We 7
exercised the codes for a wide range of conditions:
for 8
high-pressure and low-pressure se quence s, for rapidly 9
developing and slowly developing accidents, for different 10 types of sequences, for LOCA and transients and sequences 11 that have induced failures subsequent to start, and for 12 different plants where the importance of different 13 phenomena changes.
We showed that the codes execute and Us 14 that they give physically reasonable results; that is, all 15 of the heat is accounted for and we don' t account for 200 16 percent of the inventory of some material.
17 It is not validation.
We now have a program for 18 validation of the codes and the models.
That is, we will 19 compare them with existing experiments and experiments that 20 are in our future research program, and they are in place 21 and are described in the chapter on research for all of our 22 codes.
I stress the word "all" because in some cases we 23 intend to look at the source term code package and its 24 comparison with the more mechanistic code.
But if that 25 mechanistic code has not been validated, then the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Natic. wide Coverage 80 4 336-6646
-l
27072.0 78 REE
)
1 comparison is not'particularly useful.
2 I might also mention that we have a verification 3
program that is the Fortran statements do indeed allow one 4
to execute the equations that one thought one was executing 5
and we also have quality control procedures in ef fect.
6 MR. MOELLER:
Is the South Texas project one of 7
the five referenced plants?
8 MS. MITCHELL:
No.
The five referenced plants 9
are Surry, Peachbottom, Sequoyah, Zion and Grand Gulf.
10 Those are the ones that are discussed in NUREG-0956.
11 The NUREG-ll50, after its first publication as a 12 draft, they will be adding material on LaSalle, which is a 13 BWR MARK II.
O 14 MR. MOELLER:
At this mee ting the Committee is 15 reviewing the operating license application for the South 16 Texas project, Units 1 and 2.
And in NUREG-1171, which is 17 the environmental statement for the South Texas project, i
18 there is a complete beautiful description in which the 19 source term code package is compared to the traditional 20 analyses with curves showing, you know, dotted lines for 21 one and solid lines for the other.
22 MS. MITCHELL:
I haven' t seen the DES but 23 probably --
24 MR. SOFER:
I am Len Sofer of NRR.
To answer 25 your question, one of the earliest applications of the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
27072.0 79 REE
()
1 source term code package and the revised methodology was 2
for relooking at accident risks in this South Texas project 3
where there was a discussion using the source term from 4
WASH-1400 together with the revised methodology so that 5
they could be compared side by side.
6 MR. MOELLER:
Thank you.
I have found that very 7
helpful in terms of this review.
8 MR. KERR:
Since that is not a safety document, 9
you have to be careful of interpreting the results in a 10 safety context.
11 MR. MOELLER:
I have been pushing for the 12 Committee to read this because I think it is very valuable.
13 MS. MITCHELL:
Paragraph 13 describes the e f fort O
14 as far from conclusion.
The task is far from conclusion.
15 But it also is well begun.
And now it is even better than 16 before because of the changes in the code suite which we 17 described in chapter 3 which had not been in existence at 18 the time that we wrote the draf t and has only been 19 developed, which includes the removal of the 20 inconsistencies, one of which I mentioned before, and also 21 includes some of the important additional feedback ef fects.
22 But we believe that it is more than a status report, 23 because it describes the technology with marked advances.
24 Those advances that Mel talked about will provide better 25 understanding of severe accident phenomena and give us more CE) i ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33(H5646
27072.0 80 REE 1
confidence in the results.
2 Furthermore, we have provided a framework of 3
coupled and integrated codes that can be easily upgraded as 4
we learn.
This is the reason why we have concluded that 5
the source term code package can be used, provided the 6
uncertainties are considered, and why we believe that the 7
source term code package is good, as well as being better 8
than the reactor safety study.
9 Paragraph 14 noted that the ACRS asked how the j
10 ma terial in the report would be used repeatedly.
We 11 appreciate the ACRS' frustration and the report has been 12 changed somewhat and will be changed a little bit more to 13 clarify that the use in the regulatory arena is ite ra t ive.
d 14 That is, we will start out, see what we need to know, find 15 out how well we know it and, if necessary, go back and 16 refocus our experimental or calculational program to 17 provide the information.
18 The people in the source term area work on a 19 daily basis with the people in research, who are working on 20 1150 as the first application, the second, perhaps, 21 depending, and NRR, to develop the details of the usage.
22 We certainly will find more questions that we 23 have to be answering and we will refocus our efforts to do 24 so.
25 Paragraphs 15, 16, 17, and 19 list the other O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 81 REE g)
(
l pieces of the puzzle that the - ACRS believes are important.
2 We definitely agree that some or all of those things listed 3
will be required for any application.
Those are the 4
fission product release and transport, which is the subject 5
of 0956, the risk, that is -- including the event 6
frequencies, which we will get from 1150, a better handle I
7 on containment performance, which is being handled also in 8
1150; the methods for individual plant analysis, which NRR 9
is working with IDCOR on, and the uncertainties.
And again, i
10 I want to stress that we don't necessarily believe that all 11 pieces are required for every application.
12 Paragraph 18 mentions that the codes in their i
13 present form should not be given much weight in s
14 decision-making.
We agree that codes by themselves are not 15 answers to the important questions.
We just discuss the 16 other pieces that are needed and those needs limit what use 17 can be made of just the fission product technology.
18 We have reflected more of this in the report and 19 will, in the guidance on uncertainties, reflect more.
But 20 the tool that we believe should be used when fission 21 product release and transport information is needed is the 22 source term code package and that it will provide a better 23 understanding than you could get from the 10- to 24 25-year-old technology.
25 Paragraph 20 is on containment behavior.
O a
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
27072.0 82 REE
()
1 NUREG-1150 will provide the frequencies, and we note that 2
the source term code package provides the load and an 3
evaluation of when the criteria are met and, therefore, the 4
source term.
All the material that was previously there on 5
the containment behavior has been removed and the report 6
has been clarified.
7 Paragraph 21 has two thougnts.
The first was 8
that the ACRS would like to have known what effect does 9
each of the major improvements have on the source term.
10 This sounds like a very reasonable request and a request 11 that would be easy to satisfy.
But this is not the case.
12 The reactor safety study methodology was used as 13 a set of small stand-alone codes that were stitched O
14 together to provide results by hand.
And we feel that it 15 is really not practical to try to go back and look at the 16 advances one by one, compared with the old methodology.
17 What we are trying to be responsive to is what you want to 1
18 see.
And I want to note four things.
19 The first is that for those sequences that are 20 comparable, we have provided a comparison of the bottom 21 line.
That includes the effects of all of the advances 22 including the putting together of the framework, taken 23 together.
24 The second is that we are evaluating the ef fect 25 of the iodine chemical form assumption with a O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 83 REE ps,
(,)
1 forward-looking chemistry package rather than going back 2
and looking at what the reactor safety study has said.
And 3
we have in the present draft the first part of that 4
parametric study evaluating that.
5 We are getting some more results.
We have 6
gotten some in the last week which are not reflected in 7
what is in there, and we will be putting two more sequences 8
in the final report.
9 There is, in the present NUREG-0956, the May 10 draft, an interesting insight on what happens when you use 11 an integrated code in that the S2C calculation that was 12 done originally for Surry in the reactor safety study, when 13 the calculation was over, they concluded that it might be a 14 very important sequence and it is one of the important 15 sequences that they discuss.
However, putting the thermal 16 hydraulic package together and doing an integrated 17 calculation shows that it is really not going to be a core 18 melt sequence that is going to be not important at all in 19 the risk for that reason.
20 Finally, the QUEST study, which is the set of 21 four volume documents that we incorporate by reference,
22 does provide some insight on some one-by-one sensitivities.
23 However, they used the then-existing code suite and so it 24 Goesn't have the advantage of having put toge ther the codes 25 in a more coherent fashion.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
i 27072.0 84 REE 3
\\_)
1 The second thought in paragraph 21 is a 2
reiteration of IDCOR comparisons and we have already 3
discussed it.
Paragraph 22 talks about natural circulation, 4
steam generator tube ruptures and high-pressure melt 5
injection as it might be important for containment failure.
6 These are all important areas.
W'e totally agree.
7 Natural circulation is listed in NUREG-0956 as a source of 8
major uncertainty.
We have major code develop efforts 9
under way, and their validation will have to come from 10 experimental programs such as the one sponsored by EPRI at 11 Westinghouse.
When we have information, our source te rm 12 codes, like the source term code package, can be modified 3
13 to include insights about natural circulation.
14 Steam generator tube rupture accidents can be 15 considered in the source term code package framework.
One 16 such sequence was considered for the Sequoyah plant and is 17 talked about in the NUREG-0956 May draf t.
18 As usual, it requires determination of failure 19 criteria outside of the source term code package, and in 20 this case it is modeled by a failure, an input failure as a 21 function of time.
22 Direct containment heating is handled in a very 23 crude fashion.
Part of it is today modeled by the source 24 term code package in the thermal hydraulic conditions, but 25 then it is corrected by hand to account for what we believe l
(3
\\
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
27072.0 85 REE 1
crudely today to be chemical ef fects.
And some of that is 2
now talked about in the May 23rd draf t.
I also note that 3
there is a large experimental program in this area.
4 Paragraph 23 notes that fission product release 5
is temperature-dependent and WASH is very crude in this 6
area.
It may be very crude in general.
We agree.
Core 7
melt progression is a major source of uncertainty also.
8 And it is being considered in the uncertainty study for 9
NUREG-ll50 and is a major emphasis in our code development 10 and our experimental program.
The code development has 11 been in progress for some time, but the models that are 12 used in those two codes are still mostly user-controlled.
13 We have experimental programs appearing now and 14 more will be coming in the future from various of the 15 in-pile experience.
The experiments in melt progression 16 are difficult because of the questions of scale, and we are 17 necessarily investigating them with experiments that 18 significantly reduce scale.
When we get new information, 19 again, we can either modify the crude models or we can 20 adjust the way we use the present models, whichever seems 21 more appropriate.
22 Paragraph 24 discusses core-concrete 23 interactions.
W'e agree that it deserves further 24 investigation and we are actively pursuing the 25 inve stiga tion.
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
L 27072.0 86 i
REE
)
1 Calculations with CORCON and VANESA, those are i
2 the models that are used in the source term code package in 1
this area, both for the Sandia and the test at KFK, show 4
reasonable results but indicate that the models could be i
5 improved.
The problem with them is that there seems to be l
6 a discrepancy between the split between the axial and the 7
radial penetration.
And that kind of a dif ference may have 8
only a small ef fect on the source term.
9 Text has been added to chapter 6 to describe 10 this.
I do want to note that some of the perceived 11 differences in aerosol processes for the Sandia KFK test 12 was based on some early work and the results are very 13 dependent on the type of concrete used and CORCON and gs 14 VANESA calculates this dif ference very well.
15 Paragraph 25 says that some documents were not 16 readily available and that complete documentation of the 17 bases of NUREG-0956 is very important.
18 At the time of the draf t, some of the 19 documentation was not in final form.
Some of it was in 20 letters that were in the public document room.
Some of it 21 was in drafts that were only out for review and were not 22 really complete.
We are now going to issue NUREG CR 23 reports.
That should eliminate or alleviate at least some 24 of the concern about availability to the general public.
25 The draf t of May 23rd for the final report is O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
27072.0 87 REE
)
1 incomple te in its references. "There are some of them that 2
are inscrutable, to say the lesst.
We are correcting those.
3 We recognize that we didn' t have the details of those 4
references and they are just cryptic notes, so that when we 5
go back we can get the right reference in place.
We are 6
going to clean them up, but we chape also instituted a 7e review where we are going _to look;for instances where we 8
cari provide references for-other statements that were made, 1
9 mainly in chapter 3.
10 Paragraph 26 is about external events.
It 11 speaks directly to NUREG-1150 and not to NUREG-0956, and we i
12 have passed on the comment as such to the developers of 13 I NUREG-1150.
14 The answer to paragraph 27 is, you are welcome.
15 MR. KERR:
ThanA -you for a well-organized 16-pre sen ta tion.
Are there questions?
17 MR. MOELLER:
Who is leading the, ll50 18 development?
19 MS. MITCHELL:
Mel Ernst.'
20 MR. SHEWMON:
Can you tell me again what you 21 told me in response to paragraph number 3?
That is, the 22 comparison of WASH-1400 will M laund in an appendix in the 23 new final report or where7 24 MS._HITCHELL:
The difference -- the differences 25 in source terms for those sequences which we find to be
(')'
\\
s.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
-27072.0 88 REE em k) 1 comparable is at the end of chapter 4.
But in the m
2 statement, as to what we believe about whether there are 3
significant dif ferences compared with the reactor safety 4
study, we are less ambiguous than we were.
And we now 5
state that we believe there aren't large systematic 6
reductions, and we don't see systematic effects that apply 7
to both plants, all sequences and all chemical element 8
groups.
9 MR. KERR:
Are there other questions?
10 Thank you very much.
I turn things back over to 11 you.
12 MR. MARK:
I wonder if I could hear or is there 13 any point to raising here the question I had about the i
m 14 hydrogen --
15 MR. KERR:
This is your last chance forever to l
raise any questions on 0956.
16 17 MR. MARK:
Chapter 4, at a number of places, in 18 particular on page 4-50, you refer to "during the multiple 1
19 hydrogen burns."
L 20 MS. MITCHELL:
Yes.
21 MR. MARK:
I could not find any description of 22 the assumptions made decreeing that there was to be a 23 hydrogen burn.
What kind of a burn was it?
Under what 24 conditions did it occur?
{
I 25 MS. MITCHELL:
The assumptions in most of the
{
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336 % 46
27072.0 89 REE
(~(_)
/
1 cases were that the hydrogen concentration and the volume 2
should be 8 percent.
3 MR. MARK:
That sounds reasonable.
That comes 4
straight from TMI.
5 MS. MITCHELL:
And the steam, moisture content 6
should be below, I think it is 55 percent?
Is that -- I 7
would have to go back and check.
8 MR. MARK:
If the steam is in the --
9 MS. MITCHELL:
In some cases, you might be steam 10 inerted.
11 MR. MARK:
If if the steam is anywhere close to 12 50 --
13 MS. MITCHELL:
Hydrogen more than something and 14 steam less than something else.
15 MR. MARK:
Is this hydrogen coming from metal 16 water reaction or coming from chlorine-concrete?
17 MS. MITCHELL:
It depends on the time of the i
18 accident.
In vessel it is coming from zirconium oxidation.
l 19 And --
20 MR. MARK:
There you get solid, pure hydrogen 21 with steam.
In concrete you get at least as much water and 22 carbon dioxide as you do hydrogen pl.us carbon monoxide.
23 MS. MITCHELL:
Rig ht.
24 MR. -MARK:
There is only room for about -- well, 25 at the most two hydrogen burns from the reaction because ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336,6646
27072.0 90 REE
(.-
1 you are inerting carbon dioxide and water.
2 MS. MITCHELL:
I don't think in all cases you 3
are doing that.
You start in at the bottom of the melt --
4 molten pool with CO-2 and H20, and as it goes through, 5
i those two oxidize material, and you will end up with carbon 6
monoxide-hydrogen.
In some cases, you reduce it all the 7
way to carbon.
8 MR. MARK:
I am reading the amps per second of 9
gas that you have in your draf t here some where.
Slightly 10 more inert stuff than there is hydrogen?
11 MS. MITCHELL:
That graph on the DF for the pool l
12 is a function of particle size.
It is meant to be -- to l
13 show you that as a function of particle size, it spans 14 decades in DF.
And it certainly is applicable, that i
15 particular set of conditi ns is applicable to one point in 16 time in one accadent, but it is not meant to be indicative 17 that that is the set of conditions that you would see going 18 through the pool at all times.
19 MR. MARK:
Okay.
20 Now, the water never condenses once you have got 21 it, because you are heating the gas while you go.
So there 22 is no tendency for the steam to condense.
There is no 23 tendency for new air to come in because you have warmer 24 stuff inside than you have outside.
25 MS. MITCHELL:
In -- certainly in Sequoyah, O
4 ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage 202 347 3700 800 336-6646
27072.0 91 REE
_)
I there is a graph in there for Sequoyah, the release, the 2
airborne fission products and the released fission products 3
as a function of time.
The steam will condense because it 4
is going through the ice condenser with ice in it.
So if 5
you start out with water and hydrogen at the bottom, you 6
are not going to have water vapor --
7 MR. MARK:
Okay.
I didn't want to claim that I 8
knew what one should assume here, but it seems to me that 9
you did not explain what you assumed.
10 MS. MITCHELL:
Okay.
We will note that.
11 MR. KERR:
Mr. Ward, do you have a question?
12 MR. WARD:
No.
13 MR. KERR:
Any further questions?
Speak now or --
14 MR. WARD:
I guess I do have a question.
Do you 15 plan to have a letter --
16 MR. KERR:
Yes, sir.
17 MR. WARD:
Do you know what you are going to say?
18 MR. KERR:
Yes, sir.
Well, I know what I am 19 going to say in the draft.
I have a draft.
20 MR. WARD:
All right.
21 MR. KERR:
But I would assume that the Committee,
22 which has not seen the draft, would have additional wisdom 23 to add or subtract.
24 MR. WARD:
Do you want to look at the draft for 25 a first run-through now or --
CC r
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6M6
27072.0 92 REE
)
1 MR. KERR:
It is almost in final form, but it is 2
partly handwritten.
3 MR. WARD:
You would rather wait?
4 MR. KERR:
I think --
5 MR. WARD:
You started half an hour late and you 6
finished half an hour early.
7 MR. LEWIS:
Well, that is wonderful.
8 MR. KERR:
That is the old bicycle shed axiom.
9 MR. LEWIS:
If you want some trivia to talk 10 about for a half an hour, any of us can help.
11 (Laughter.)
12 MR. WARD:
Okay.
No, we are in fine shape.
13 Thank you very much.
Let's go on to the next topic.
14 MR. SILBERBERG:
Thank you very much, 15 Mr. Chairman.
16 (Whereupon, at 4:45 p.m.,
the meeting was 17 adjourned.)
18 19 1
20 21 22 l
23 24 25 O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4336 6646
CERTIFICATE OF OFFICIAL REPORTER O
This is to certify that the attached proceedings before the UNITED STATES NUCLEAR REGULATORY COMMISSION in the matter of:
NAME OF PROCEEDING:
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 314TH GENERAL MEETING DOCKET NO.:
PLACE:
WASHINGTON, D.
C.
DATE:
FRIDAY, JUNE 6, 1986 were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission.
(sigt)
M C
(TYPED)
REBECCA E. EYSTER Official Reporter ACE-FEDERAL REPORTERS, INC.
Reporter's Affiliation O
I c-
.c-,
.., - -, - - - -. - -,,.--. n _,,----
i
=
0 Agenda for ACRS Meeting on June 6, 1986 1:00 p.m.
Room 1046, H Street i
RECENT SIGNIFICANT EVENTS Presenter /0ffice Date Plant Event telephone M
i 5/19/86 Pilgrim Single Failure Could Disable E. Weiss, IE 2
All Redundant RHR Pumps 492-9005 l
6/85 Trojan Repeated Snubber Failure T. Chan, NRR 5
492-7136
^*4'> e
) O I
e l
i d
O A
v
--y y
--..,r.,1,
,,+-,--,e-
.,,~-.,g.
..,-y.,g p.-,w-,,,mr.f-e,y,-.,-
y,
,w-,,
,y.ry.w+,-wwew,,-,,_.-.,
,,.,m.,,_..--rw..
,--e-s.--,-,,,---,mm
PILGRIM -
SINGLE FAILURE COULD DISABLE ALL REDUNDANT RHR PUMPS
(_)
MAY 19, 1986 (ERIC WEISS, IE)
PROBLEM:_
SINGLE FAILURE OF MINIFLOW LOGIC COULD DISABLE ALL REDUNDANT RHR PUMPS DURING SMALL OR INTERMEDIATE SIZE BREAK LOCA SIGNIFICANCE:
POTENTIAL SINGLE FAILURE CAUSES LOSS OF MULTIPLE SAFETY FUNCTIONS POTENTIAL FOR NO LONG TERM COOLING FROM SAFETY SYSTEMS CIRCUMSTANCES:
LICENSEE REVIEW (PROMPTED BY INF0 NOTICE 85-94) DISCOVERED THAT SINGLE FAILURE OF EITHER MINIFLOW SWITCH COULD PREVENT ALL AUTOMATIC LOW FLOW PROTECTION FOR ALL RHR PUMPS; PUMPS COULD BURN UP IF MANUAL ACTION NOT TAKEN IMMEDIATELY DURING SOME ACCIDENTS OR SPURIOUS ACTUATIONS, RHR PUMPS WOULD BECOME DEAD HEADED FOR EXTENDED PERIOD CURRENT MINIFLOW LOGIC DESIGNED TO BE CONSISTENT WITH LOOP SELECT LOGIC FOR LPCI
<3 kJ*
WHEN FLOW DETECTORS IN EITHER LOOP SENSE ADEQUATE FLOW, BOTH RHR MINIFLOW LINE VALVES CLOSE CONSEQUENCE OF RHR PUMP LOSS IS LOSS OF LONG TERM COOLING WITH RHR HEAT EXCHANGERS, AND OTHER FUNCTIONS INCLUDING:
-SHUTDOWN COOLING MODE
-LOW PRESSURE COOLANT INJECTION
-HEAD SPRAY (REMOVED FROM PILGRIM)
-CONTAINMENT SPRAY
-TORUS SPRAY
-SUPPRESSION POOL COOLING WHICH EVENTUALLY WOULD CAUSE LOSS OF:
-LOW PRESSURE CORE SPRAY
-HIGH PRESSURE COOLANT INJECTION
-REACTOR CORE ISOLATION COOLING GE FIX IS TO ELIMINATE "CLOSE" SIGNAL TO MINIFLOW VALVES; COULD INCREASE PEAK CLAD TEMP 50*F IN SOME BREAK SIZES; NRC CONSIDERS THIS TO BE INTERIM ACTION
(]) FOLLOW-UP IE BULLETIN 86-01 ISSUED 5/23/86 IE AND.GE ARE DETERMINING GENERIC SIGNIFICANCE NRR WILL REVIEW RESOLUTION FOR PLANTS WITH PROBLEM, INCLUDING TECHNICAL SPECIFICATION ISSUES
- }_
SIMPLIFIED DIAGRAM OF PILGRIM MINIMUM FLOW FOR RHR PUMPS w2 VM l '
E L2 La V3 V3 RX LPCI V PRESS.
V LPCI VESSEL AINJECT INJECTd 3
A RECIRC.
REClRC.
PUMP P U M,P
)f____________________>_,
____}(
OR GATE FLOW n
[
Y
- i FLOW I
INDICATOR l
TORUS l
l TORUS j
INDICATOR l
MIN. FLOW
____1____
MIN. FLOW BYPASS VALVE BYPASS VALVE CROSS TIE es eus sus es 1 r 1 r RHR
($
(
'1 f P U..M.P. S PU..M.P. S lfQ g
=
s r 7,i RHR RHR HX "B" H X **A" "EITHER SENSOR DETECTING FLOW WILL CAUSE MINIMUM FLOW VALVE TO GO CLOSED O
3
TROJAN - REPEATED SNUBBER FAILURES JUNE,1985 ( E GNAN, NRR)
PROBLEM:
STEAM GENERATOR HYDRAULIC SNUBBERS LOCKING UP DUE TO DESIGN INADEQUACY.
SIGNIFICANCE:
DAMAGE TO HOT LEG PIPE WHIP RESTRAINT (1985)
OVERSTRESSING 0F HOT LEG ELB0WS PREVIOUSLY UNACCOUNTED FOR MOVEMENT IN THE PRESSURIZER SURGE LINE (1982-1985)
CIRCUMSTANCES:
NRC RECENTLY LEARNED THAT RCS HOT LEG PIPE RESTRAINT HAD PULLED FROM WALL IN 1985 LICENSEE, NRR, AND REGION V WALKED DOWN RCS PIPING DYE PENETRANT TEST PERFORMED ON "B" SG ELBOW.
NO O
. INDICATIONS FOUND,
}f PERFORMED UT ON ALL 4 HOT LEG ELB0WS AND FOUND NO INDICATIONS CRUSHED GRAPHITE SHIMS FOUND ON 3 0F 4 HOT LEG PIPE WHIP RESTRAINTS INDICATING HOT LEG TO RESTRAINT BINDING 11 0F 16 SG SNUBBERS FOUND TO HAVE FAILED AGAIN IN SAME WAY BACKGROUND:
1982 - LICENSEE REMOVED THE THERMAL SLEEVE ON THE PRESSURIZER SURGE LINE; HOWEVER, SURGE LINE DID NOT SETTLE OVER NEXT FEW CYCLES, AS HAD BEEN EXPECTED IN W ANALYSES; MOVEMENT CONTINUED 1985 - LICENSEE H! RED IMPELL TO REVIEW THE SURGE LINE MOVEMENT; UNABLE TO ACCOUNT FOR CONTINUED MOVEMENT 1985 - A HOT LEG (TO SG "B") PIPE WHIP RESTRAINT HORIZONTAL SUPPORT WAS FOUND PULLED FROM THE WALL Y
TROJAN - REPEATED SNUBBER FAILURES JUNE, 1985 ( f CHAN, NRR), (CON'T
([)
' )'
BACKGROUND, (CON'T.)
1985 - SNUBBERS TESTED PER NEW TS REQUIREMENTS
- 2 0F 16 SG HYDRAULIC SNUBBERS WOULD NOT RESPOND T0 100 KIP LOAD; SHOULD HAVE nESPONDED AT <;10 KIP; ALL 16 WERE DECLARED IN0PERABLE AND REBUILT
- SNUBBER FAILURE ATTRIBUTED TO CLOGGED HYDRAULIC LINES; CLEANED j
WHEN ASSUMED THAT ALL SG SNUBBERS WERE INOPERABLE,IMPELL ANALYSES WAS ABLE TO ACCOUNT FOR THE SURGE LINE MOVEMENT AND THE DAMAGE TO THE PIPE WHIP RESTRAINT THE LICENSEE CLAIMED (1985) THAT ALTHOUGH HOT LEG STRESSES EXCEEDED ASME SECTION III ALLOWABLES, STRAIN IS WITHIN 1%
LIMIT, WHICH WAS NRC-APPROVED LIMIT FOR SONGS-1 ON SEISMIC CRITERIA AND METHODOLOGY
(- ) FOLLOW-UP:
SNUBBER CONTROL VALVES TO BE REPLACED WITH NEW DESIGN LICENSEE TO PERFORM PRE-STARTUP WALKDOWN OF RCS IN A HOT CONDITION NRR TO REVIEW RCS PIPING STRESSES AND APPLICABILITY AND ACCEPTABILITY OF LICENSEE'S ANALYSIS
($)
S'
/
/
l
/
/
o l
aC V %,
s,.
8 n,,.
- a.
l.:.c.
l q.% 7 Q~ \\
h'(_
l z.
's
> ~.
8...
'\\,.......'
j
\\
\\.\\
\\
h ',b
\\
l N
I
\\\\
ifl i
i t
.i i
\\' )
Eni
(,\\
\\,\\
d
(
\\\\
\\\\
4
\\\\
1
\\\\
\\
/
4
\\
\\
(
l
- /
(
I!!
I 5
~
'E.
D 4
li !
u,! c
/
[
/~s, N-T T1
)
)
fs x
l\\\\h-
--( q!
g.
.y O
\\
MOTE:
CONTA0L VALVE 15 Fu2NISHED WIT H (4) CUTLETS (2 EA. SiCE)
O
,/, _
MORE SuPAAE55045 IN P*AALLEL.
TO allow MOUNTING (2) CA RESERVOIR THE OUJLETS ARE DRovlCE O WITH CHECKS, SPDING BIASEO NO A M AL;Y OP EN.
"nQWP_"-.,-
R E F.}'T k A a 4,A2,Bl & B2.
E-Z_ ~_ =~=~.::,
SEE
. PAGE I2 a FC4
-]_
TYPCAC CONTROL C 6 A Cult.
-E--
CAPPED CUTLETS.
_q l
) (A2 V2 VI l
I x; x
10
'S
~
0, 75 i
_r j
l l-- CONTROL l
VACVE.
.~
\\
vl O
Iv
.O
^ i
^2 E
S2 I^
l i
L___
__a i %
\\
N
\\
_ _ =,
__=_ _ _:
\\
Etz -E Ei =-r.-r z-N r
r N
i 3_~.=_r =-'_: :1--
= - ~ -
T
_ z:E t.- i ;
s\\.
._. =. _
y\\
N NN N
r
.=_ =__-
== = 2 _=_=.
N.
N
_ = _ _ _
_=_=_=__._._:
N.
= _ _
_=
N
_=,. _._ ____- _
\\
._ _. _ _ =. =
N
\\
1 u
TYPICAL SUPPRESSOR t
FIG. 4. SCHEMATIC DIAGRAM OF CONTROL VALVE 1
O soox sueeResson
~
7
g' ntstavo R RESPECTIVE CONTROL
,b VALVE COMPONENTS SAME AS SHOWN i
B Etow.
\\
kW
[
l x
gl l}
-t l_
_J
'I r
Ju!!R M !L
//
o I
to
/
M1 EMT
%/
/
u N'
lk-ordc>-4l
/
\\,_ \\
l g' l
~
m
~~
sys L_
'J seN R$ff
$$NL
((
=
2 O
l
\\
(o j
\\l\\
' EMi@
EEK N/'~
axs O
s=s i-
"-- M->,
'E
/ w'
\\[\\
lhMdY li o
l s
s;s L
- gI pse
. \\., N I 'ip.
x v
_J
/5 e
M4 Gssi?
((
y s r
o I
(o s"gN
' F~$!R ET
. N/
N
/
~
n2 uvt v
R2 A
i CONTROL fcoNNtcron I
Rt
=
c opro.
' VALVE
\\
Js323 ssfL g/
O I
I (o
j
"@bE$
E!?#T
%f N
j Soon suppRtsson
~
O FIG. 4a. SCHEMATIC DIAGRAM OF' CONTROL CIRCUIT 4 c. id.
8
O PALISADES PLANT BACKGROUND:
SALP CATEGORY 3 - FAINTENANCE, SURVEILLANCE, QUALITY PROGRAM LACK OF AGRESSIVE CORRECTIVE ACTION POOR PANAGEPENT CONTROLS i
CYCLE 5 RECURRENT EQUIPVENT PROBLEMS - 1985 SAFETY INJECTION TANK SYSTEFS (SIT)
VARCH 1986 STARTUP FROM REFUELING / MAINTENANCE OUTAGE Q
TWO 0F FOUR PRIl%RY COOLANT PUMPS WITH SEAL PROBLEM 3 PCS LOOP CHECK VALVE LEAKAGE ~
SIT SYSTEM VALVE LEAKAGE CVCS DIVERT VALVE LEAKAGE APRIL 10, 1986, SHUTDOWN - PCS LEAKAGE APRIL 11, 1986, DERATING - CONDENSATE PUVP PACKING FAILURE APRIL 23-29, 1986, - VALVE LEAKAGE PROBLEMS IN PCS PAKEUP SYSTEM 3
i O
1
O A
PALISADES PLANT - REACTOR TRIP OF MAY 19,1985 PROBLEMS:
MJLTIPLE FAILUES TURBINE BY-PASS VALVE FAILED TO OPEN 1 STEAM DUMP VALVE FAILED TO OPEN BACKPESSUE EGULATOR IN LET-DOWN LINE FAILED CLOSED PESSURIZER SPRAY VALVE FAILED TO FULLY CLOSE VARIABLE SPEED CHARGING PUMP TRIPPED 5 TIES ROD BOTTOM LIGHT FAILED TO INDICATE ONE RCD FULL IN TURBINE LIFT PUMPS FAILED TO START AllT0FATICALLY O
(
EXISTINGOUTS_F_SERVICEEQUIPENT CONDENSATE ECIRC VALVE AUTO OPERATOR INOPERABLE l
BANK OF PESSURIZER HEATERS INOPERABLE i
l SIGNIFICANE UNNECESSARY CHALLENGES TO SAFETY EQUIPENT INCREASED BURDEN ON OPERATORS TO C0FFENSATE FOR FAILED OR DEFICIENT EQUIPENT IELICATIONS CONERNING TE QUALITY OF MAINTENANCE AND POST-FAINTENANCE TESTING O
O SEQUENCE OF EVENTS PM ON TURBINE VALVE CONTROL CABIET FANS 14:15:14 TURBINE VALVES CLOSED REACTOR TRIP ON HIGH PESSURIZER PRESSUE TURBINE TRIP FIRST ATMDSPHERIC DUMP VALVE OPENED, AFW PUFP P-8A STARTED 2ND ATMDSPHERIC DUMP VALVE OPENED 3RD ATMDSPHERIC DUFP VALVE OPENED CHARGING PLFP P-55A STARTED (55B & C ALREADY RUNNING) i PRESSURIZER LEVEL LOW LAST LETDOWN ISOLATED CHARGING PUFP 55A TRIPPED; THIS PLFP WAS ESTARTED 4 MORE TIES TRIPPING 30 SECONDS LATER AFTER EACH START t
14:22:15 PRESSURIZER LEVEL NOFFAL Q
PLANT PARAETERS:
PESSURIZER PRESSURE PAX 2245 PSIA, MIN 1689 PSIA T/ HOT MAX 594*F, MIN 535'F T/ COLD PAX 557'F, MIN 535'F S/G PESSURE PAX 1025 PSIA S/G LEVEL DROPPED FRCM 70 TO 12 PERCENT NRC RESPONSE:
EGION III ISSUED A CONFIPPATORY ACTION LETTER REQUIRING LICENSEE CONDUCT THOROUGH INVESTIGATION INTO CAUSE AND IMPLICATIONS OF THE PAY 19 TRIP EGION III APPROVAL PRIOR TO RESTART O
O CONCLUSIONS:
PERFORMANCE OF PLANT OPERATORS AND THE OPERATION OF OTHER MAJOR OR SAFETY-RELATED SYSTEMS WEE AS EXPECTED AND DESIGED CONSIDERING THE EQUIPENT FAILURES THAT OCCURRED.
SIGNIFICANT WEAKNESSES EXIST IN P%INTENANCE FUNCTIONS OF DIAGNOSTICS, REPAIR, POST-FAINTENANCE TESTING. THESE WEAKNESSES WEE CONTRIBUTORY TO FUST OF THE ECUIPENT FAILUES.
EGUIPMENT FAILUES AND DEGRADED EQUIPENT HAS PLACED VARYING LEVELS OF ADDITIONAL BURDEN ON PLANT OPERATORS.
FOR MAY 19,1986, TRIP, EQUIPMENT FAILUES DISTRACTED OPERATORS BUT DID NOT SIGNIFICANTLY Q
JEOPARDIZE PLANT SAFETY.
PLANT OPERATORS HAVE SERIOUS CONCERNS EGARDING THE ADEQUACY OF P%INTENANCE ACTIVITIES AND EQUIPENT ELIABILITY, O
_