ML19308C071
| ML19308C071 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 05/09/1979 |
| From: | Judith Weaver HOUSE OF REP., INTERIOR & INSULAR AFFAIRS |
| To: | |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001200081 | |
| Download: ML19308C071 (91) | |
Text
___ _
.s.
~d'UM5P Uf $$PyrPE?niati1125 i.
.O G
Report of Proceedings 4
Hearing held before 1
COMMITTEE ON INTERICR AND INSULAR AFFAIRS TASK FCRCE CN t
THREE MILE ISLAND ACCIDENT O
/
t i
i e
Washington, D.
C.
i WFDNESDAY, MAY 9, 1979 i
4 i
d ACE FEDERAL REPORTERS.INC.
Offi:ial Reponers
- 12ooort
~ ~ c8 c== -,,,,,
Wcsnn gren, O.C. :ccc!
NA*!CNWICE COVERAGE Dally Teleoncav:
(200):47 n:o
. ~.--
1 1
1 (1732 1 !
TA'SK FORCE ON 1
2 THREE MILE ISLAND ACCIDENT l
3I WEDNESDAY, MAY 9, 1979 i
4 i
U.
S. House of Representatives, S
Committee on Interior and Insular Affairs, 6!
I Washington, D. C.
7 The committee met at 3:40 p.m.
in room 1336, Longworth si l House Office Building, the Honorable James Weaver, presiding.
Present:
Representatives Weaver, Carr, Markey, Kostmayer, 10 Vento, Marriott, and Cheney.
Il {
I Staff present:
Messrs. Myers, Burnham, Terrell, Reis, (v')
12,
4 Scoville, Rogers, and Sadleir.
13 Also present:
Messrs. Mattson, Stello, Eisenhut, and Combs.
14 15 i
16 17 18 i
i 19 l
I 20
?
21 I
i i
22 l
()
l j
s_
08ecers Aeoorters, Inc.
25 l
s i
1 9
h
1CR 473'2 2
Craig/ cs l
t ksN )
1l Mr. Weaver.
Gentlemen, thank you very much for coming.
j A
i 1
i 2i Would vou identify vourselves, c. lease.
I am Jim Weaver.
i 3,
(Introductions.).
6 4l Mr. Weaver.
Thank you very much.
I I
t 5:
What we want to do here is have an informal, albeit we i,
6! will be recording the remarks that we make, review of what 7
happened.
What this task ' force is concerned with --
3 (Discussion off the record.)
9' Mr. Weaver.
But I would like to keep this informal.
l I
10 Let's talk about the details of what happened at Three Mile I
1 I
II Island so that as we comeile this information c.erhac.s a i
i.
i (m
12 picture will begin to form in the minds of the members of this
\\_)
i i
committee 'o we will be better able to analy e and deduce frca l
13 s
14 that.
I i
15 I would like to ask Dr. Myers to commence his line of
{
i 16 questioninc.
If any of you have anything to of fer or volunteer,'
17 that would be fine.
i 18 ;
Mr. Myers.
Do v.ou think it wculd be useful if the.v went 1
19 lthre agh the chronology as they best understand it now, as 1
20.; occ.osed to the way Denten did it the other dav. where he was i
21 ! sort of relating it as he had experienced it?
22 Mr. Weaver.
That is excellent.
I think that is a very I
23 '
fine starting.
That is what we are after, the facts en Three g3 l
\\-)
\\
sa' i
Mile Island.
So I think we can start.
p N:erv meoonen. :re.
oc Mr. Myers.
Darrell Eisenhut has been briefing the Cc=micei-i o
3 11 on this.
Darrell, I don' t know who would want to do it or j
(])
2' wno is most familiar with it.
3 Mr. Stello.
Let me make scme opening ccmments first.
41 In order to truly have an accurate list of facts exacEly l
5 as they occurred during the accident, you are going to have 6t to go back through all the records, all of the strip charts, 7
pull them together, interview all the people to get their 8!
best recollections, and pull together the best sequence of i
I 9:
events you can based on that information.
That is being done i
i.
I 10 '
right now.
My understanding is that that actual identification 11 llof the factual chronology will be available about next week.
12 That is going to represent the best information we have to s
i 13 identify that sequence.
14 Otner than that, I think it's going to have to require 15 very crude, rough ideas based on individual recollections, and i
t i
16 like the piece of paper that you have now, which is-GPU's l
l 17 recollection and identification of the sequence of events l
i 18 based on their record search.
But anything we do say we are i
19 subject to refinement and identification of the facts based I
I 20 en the investigation that ccmes out next week.
21 l With that, Darrell, maybe you can highlight.
i t
22 l Mr. Eisenhut.
Yes, I was going to make the sar.e kind of I
i f
ccament.
Mr. Stello, of course, just returned fe;m the site 23 g~
kJ 24 yesterday.
I guess ycu and : were contacted early on the 28th l
n Fwe,.i m emem. inc.
25 !
about 3:30, 3:45.
We have been preceeding in parallel since f,
O
4 1
t I
()
Il then.
We first came back tcget~ner yesterday chysically to us 1
2' where we can be working trying to put the piaces together.
l 3i Mr. Weaver.
I would like to ask you:
At what hour of 1
I4j what day did you first get on site?
Mr. Eisenhut.
5 Mr. Eisenhut.
I didn't go to the site.
i I
6!
Mr. Weaver.
You did not.
Mr. Stello did.
If Mr. Mattson, you were not at the site either?
I BI Mr. Mattson.
Yes, I was on the site.
I 9l Mr. Weaver.
Mr. Stello, what hour of the day did you 10 'l arrive at the site?
i II !
Mr. Stello.
We arrived en Friday, I guess about 3:00 o' clock.
I3 f Mr. Weaver.
You stayed there frcm then on; is that Id correct?
15 '
Mr. Stello.
I stayed there frcm then on.
I6 l Mr. Weaver.
Mr. Mattson?
I l i Mr. Mattson.
11:00 a.m.,
Sunday, April 1st, through I
l 13 Easter' Sunday.
19 f Mr. Weaver.
You were, however, all three of you, working I
20 !
at NRC headquarters in a coordinaticn role on this?
i 4
i Mr. Mattson.
When I wasn't at Three Mile Island I was j
,1 l
22 woe.-king on Three Mile Island frca Sethesda, yes, sir.
i
{~s
{
Mr. Weaver.
How about reciting, beginning with Mr.
23 s
Mattson, what nappenec as far as your knowledge goes and what
,s p s c m, u m m,s.sec.
e your role was in it -- I mean what you saw, what you actually
'3 I
4 i
5 i
r l'
saw personally.
Why don't you just begin reciting to us.
i i
1 2i Mr. Mattson.
Maybe it will help kick off the chronology 3
sort of thing you want to hear about.
4l Mr. Weaver.
When did you first hear a. bout it?
You can i
5 s tar t o f f with tha t.
Where were you and what was happening?
6!
Mr. Mattson.
I was in my offices on Wednesday, the 28th.
I 7'
I tried to reach Mr. Denton that morning; got ahold of him.
3, Heard Mr. Case, Mr. S tello, Mr. Eisenhut had gone to the 1
l 9:
Emergency Center.
I believe you had gone that morning, i
10 Darrell; is that right?
II Mr. Eisenhut.
No, I didn't go early.
I was running the A
12 other end.
U l
13 !
Mr. Mattson.
That there.had been some kind of event at 14 Three Mile Island.
i 15 !
Mr. Weaver.
What time was thir?
i i
16 Mr. Mattson.
Mid-morning, Wednesday, the 28th.
i 17 Mr. Weaver.
9:00, 9:30?
13 Mr. Mattson.
Probably closer to 10:00 o' clock.
It would 19 not be normal f or me to be called on an oceratine reactors l
~
l 1
20 l kind of problem. Mr. Stello is the Director of the Division of l
l 21 '
Cperating Reactors.
I am Director of the Division of Svstem 22 :
Safety.
I believe I was talking to a secretary so I knew I 1
23,
wasn't getting firsthand inf ormation.
I asked her to have Mr. l t
t 24 !
Denton talk to me later whenever he knew waa was go:.ng on.
OE9CIf 31 ROCCfttrl, Inc '
agi I
4-He did tnat later in the day and said that the people were
.h
t 6
l I,
()
I working hard in ene Incident Center.
He planned to go en i
i 2!
cver a little later in the afterncen and that I should keep I
3' working the five-plant shutdown.
You may recall we'd been down talking to you a week or'so i
4 5
before.
I i
6 The first priority for my day, that day, was working en the l
Y i seismic analysis of the five plants that were dcwn.
I, in fact, 1
2:00o'clockThursdayafterncoE, 8'
stayed on that verk until about i
9 at which time I went to the Incident Center to get Mr. Stello 1
i i
i 10 [
to sign a piece of paper having to do with the five-plant.
f II l shutdown.
From that time on, I have been working Three Mile I
)
I i
12 Island.
j
(])
i r
13 1 Mr. W'eaver.
Mr. Stello, when would you say you first --
i i
14 '
where were you and what was happening?
l i
15 Mr. Stello.
I got to work about 7:00, 7:30 that morning.
I 16 The first phone call I received, I think, was identifying that i
I7 there had been an incident at Three Mile Island, probably near f,
i I
IS j the hcur of 8:00 o'c1cck in the morning.
1, 19 Mr. Weaver.
Who was it from?
l 20 {
Mr. Stello.
Mr. Mo s e ley.
e 21' Mr. Weaver.
Wno is Mr. Moseley?
j i
i 22 I Mr. Stello.
Mr. Mese:ey is the director of all of the cperating reactors in the Inspection and Enforcement Division, 23
("N w/
t 3d wno called and activated cur Operations Center.
As par of n.; w ero ceoer m i. ire.
the standard precedure for activation of that center, he is to
-e i
i, l
I t
i
{])
call the various members, and I am one of them.
So-he called I
i 2.
to notifv me that ths lenter had been activated.
i l
3l Mr. Eisenhut was in the office with me thac morning.
I I
I I
4' immediately called Mr. Grimes, since he knew there had been a l
5 release of fission pr ducts t--
the environment.
I told Mr.
6 Grimes to immediately go to the center.
i 7l I asked Mr. Eisenhut to set up the various pecple that t
l 3!
wculd be needed to understand Three Mile Island in his office, ;
l t
9 and as soon as I gave him that word I immediately lef t to go i
i l
10 !
to the center, j
I I
11 !
Mr. Weaver.
The center is Bethesda?
3 i
12 '
Mr. Stello.
It is in Bethesda in our headquarters of: 1ce f
)
/
l i
13 l of Inspection and Enforcement on East-West Highway, some 14 distance frcm my office which is on Norfolk Avenue.
15 l Mr. Weaver.
How many times has that center been activated?
16 l t
Mr. Stello.
I can recall, I guess, three occasions.
17 Mr. Weaver.
Three occasions in the past?
4 la Mr. Stello.
This being the third.
There may have been 3
i 19 !
more.
20 Mr. Weaver.
In the past how many years?
i 21,
Mr. Stello.
Since its existence.
22,
Mr. Weaver.
Wnich is
--?
I
)
23 '
Mr. Stello.
Abou: three years.
Brcwns Ferry fire being,
~
7sO 24 '
I think, the first one that I can recall.
banere secomn. tne. '
25,
Upon arriving at the center, tried to obtain some I
l 8
I
(")i 1l information, understanding what was happening both to the I'
u 2l reactor and to the environment.
We had a communications 1
I 3
system set up where. cec.cle who were in the control rocm at I
4 Three Mile Island were communicating f actual information as to i I
5 wnat was happening to the people in Bethesda.
We were doing 6',
various kinds of studies, trying to understand the process i
I 7l that was going on in the reactor as well as what was going on j
i i
8; in the environment.
I 9!
Our immediate effort was probably heaviest in terms of I
t 10 understanding what was happening to the reactor, since the I
I, 11 '
normal processes that you expect if you have a very had i
12 accident in the reactor is one where the crimar.v envelope has O
i t
i 1
13 ;
been breached and the water and. fission prcducts are emptied I
i 14 !
into the containment and envirorment.
I 4
15 i Here we had a clearly different situation.
We had an i
i i
16
_ntact primary system with signi:1 cant release of fission i
l 17 l picducts.
Early throughout the day it became apparent that j
l 18 l conditions had developed where the core was being badiv l l I
.s i
l i
l, 19 '
damaged.
i 20 '
Mr. Weaver.
This is still on Wednesday?
21 !
Mr. Stello.
I am now giving a very general picture of l
22 Wednesday, rather than specific hours.
Withcut going back 23 thrccch the actual taces --
I
(~s)
~
(_/
i 24
,gr. weaver.
- 3urely, saheerv necornes. sec.
l 23 Mr. Stello.
and refreshing my own recollection with I
3 1
l
I 9
i i
I respect to hours or times, I couldn't.
But throughout the day,,
(
I the clear impression was created that there was significant I
damage being caused to the cere.
The principal reason, I 3
guess -- I am giving you a personal f eeling rather than -- I
-i can't remember whether the group truly felt this way or not.
c' 6 !
Mr. Weaver.
I understand.
7 Mr. Stello.
I will try to come back --
t' i
3i Mr. Weaver.
I have to leave.
Dr. Myers, proceed.
I will j i
9l be back.
i 10 I Mr. S tello.
The reason for believing you had significant f
11 l damage was the fact that we had a clear indication that there -
was superheated steam ccming cut of the reacter vessel.
The la '
)
q 13 i only way you can, in fa~t, get superheated steam out of the l
i I!
14
vessel is to have the core uncovered.
And the core, just I
15 frca recollection, was uncovered for substantial periods of
}
time throughout the day, which left me with the clear impres-16 :
J I,' I sien that significant damage had occurred in the core.
l Mr. Myers.
Did you" at the time have a sense of, you know, I8 what v.ou were thinking of bv significant damage, whether it 19 i
20 was as it turned out to be -- were you thinking at the time i
21 i it mighu be as extensively damaged as it later develcped it t
-e i was?
1 Mr. Stello.
I guess my initial reaction was that
,3
(~3 i
%)
24 i
l censiderable f ailure of the fuel had cccurred, with large bewee = eoone,
- 4. e.
e l
nc numbers of failed fuel r:ds.
And that the potential for I
I
l 1
oxidation, metal-water reaction, was cleariv the re.
)
I 2
The periods of time involved were, even if the heat-up were i
lower temperatures, with the top of the l
a 3
not significant, at t
I 4j core uncovered for extended periods of time, metal-water g
I 5 I reaction was clearly possible.
The rates -- how high the I wouldn't even indicate-- I could have guessed 6i temperature go t
t But odixation of the cladding was scrething I clearly 7!,
at.
1 8'
e xpected; failures of the cladding were clearly evident.
9 Mr. Terrell.
When did you suspect this?
This is on I
f 10 Wednesday you're talking about that you expected this fuel i
a I,
.I 11 damage?
i I
I
(])
12 ;
Mr. S tallo.
Yes.
I think just f cm kncwing what was 13 happening to the core, you had to expect that there was quite 14 a bit of damage' to the fuel.-
15 i Mr. Terrell.
Was this communicated to anycne?
16 Mr. Stello.
I think yes.
I think we talked about it.
17 Mr. Terrell.
To whom?
Mr. S tallo.
People tnat were in the Incident Center.
18 i i
i 19 l Mr. Terrell.
How far did that go up?
l t
I I
20.'
Mr. Stello.
Oh, dear.
I.
?
i I
21 Mr. Terrell.
No, I mean did it reach the Cc= mission?
I i
22 Mr. S tello.
Well, there were Cc missioners presen from
/~N 23 l time to time in the Incident Centar.
Probably at one time er l
\\-]
t 24 ano ther I =ade my views known, that I expected that the core f
a Federsi A toor'ers, Inc.
25 had been uncovered and the fuel damaged.
I suspect so.
l t
11
/ \\
.I.
Mr. Reis.
Was that contested?
No: that vcu sat arounc
,~,
\\_)
)
2i the table for debate, but was there a discussion as to whether ;
i 3l there was or wasn' t damage during that time?
'i 4
Mr. Stello.
I don' t know of anyone who suggested there i
i 4
5' was not damage to the core.
I guess I am confused -- maybe i
there's a history of something that had gone on which suggested i
6' 1
7
-- I don' t know why you' re asking the question.
3 Mr. Mattson.
I don' t know how anybody could contest that question because there are fission products all over the place. }
9 !
i 10 I is generally accepted and known and they caly ccme frem That Il <
failed fuel.
So bv definition there is some core damage.
The I
12 l extent of the core damage was subject to further infcrmation
{}
I3 and furth'er conj ecture, I suspect, in a number of people's 1
14 minds.
IS i Mr. Stello.
If you are asking me the question which i
16 there is scmething behind I could get to -- I just don't know of.
l' any reason anybcdy would have contested it.
I think tne l
t 13 I people who were there generally knew there was damage to the l
l l
I9b core.
I don' t know of anyone who would have argued to the e
i 20 '
The early concern was what could we possibly de to
{
contrary.
t advise the licensee as to hcw to try to correct the situation 21 l
since you had, as I indicated before, an arrangement which is no i
23 not what you would acrmally have expected if you had an acciden:
,e w s-)
\\
witn a ic of fissica prcduct release, which is a broken I
- 4 e
- , r,ce.s. n eween. we.
De primary boundary.
k I
j i
I 1
39 1
1 N
1!
You clearly had an intact primary boundary with the system
)
I j
l 2,
pressure having varied from 2000 psi dcwn to 450 or 500 psi en i
l 3,
back up to in excess of 2000 psi.
So clearly you had an intact i i
4 primary boundary.
i 5<
Now, how to deal with the situation was one of cencern.
I I
6:
Mr. Reis.
Intact except in the sense of the valve?
J 7l Mr. Stello. Well, they were able to correct that by i.
8!
putting a bicek valve.
You knew you had a primary boundary I
9 that was in gecd shape in terms of holding the coolant.
10 [
Mr. Scoville.
We are jumping around here a gccd bit, which' I
11 l is part of the nature of the informality of it.
Perhaps we 12 could focus more and stick to the point.
I think if we re-gw0 13 turned to' the chronology of the, event as each of you understand.
14 l it --
15 Mr. Stello.
I was trying to do that.
16 :
Mr. Scoville.
I understand.
I am not being critica' l
1 l
17 ;
But I think we shculd proceed in that order and we can probably, i
a t
i 13 keep our eyes on the ball.
i 19 l So.to go back to that, you were about to defer to Mr.
i i
20 Eisenhut to discuss --
l i
21 !
Mr. Stello.
No.
Well --
I 22 '
Mr. Scoville.
-- the chronclegy, or were you discussing i
22 vcur eersonal invcivement?
O('N i
24 Mr. S te lle.
I was discussing my perscnal involvement
.a pnam newnen. u c.
25.
which is abcut all I can de to identify the chronolcgy.
I i
i 13 t
I i
l' Mr. Scoville.
Ckay.
s I
s/
l 2[
Mr. Stello.
Other than tha t, why, without going back and I,
2, going through the records, I would be hard-pressed to put 4
times against these things.
5l Mr. Myers.
The briefings have been presented to the i
6 6!
Commission.
So I think Darrell has done that.
Is that
}
7; possible to sort of run through the major events as you under-i s'
stand them, starting with the initial malfunctions?
9' Mr. Eisenhut.
Sure.
I 10,
Mr. Myers.
I think there are certain details of this that '
11 I maybe are not well-understccd now.
But I think maybe that is,
I 12 not so important for this discussion, but to say that these O
s_/
i 13 j are the major events that happened along the way, and this i
i 14 :
is what people did, and maybe interj ecting here and there l
15 ;
where you think they might have done something different i
l 16 wculd be useful.
t i
17 Mr. Scoville.
Before you start that, because t's going i
18 to take quite some time, there are a couple of points involving 19 !
what vou were doing in the Response Center as to whether one 20,
particular topic was mentioned.
I think maybe the snswer t
i 21 would be fairiv simple.
We could get it on the record and go 1
t 22,
back to it again.
That is, as I understand it, you were at the i
t 22 {
Response Center on Wednesday.
[)
l e
24 '
To vcur kncwledvei did anyone mention :
v.ou the r.ossi-on,aro necceen. inc.
bilityofahydrogenburnordetonationwithinthecontainmebt 25 i
1 i
6
j 14 I
I l
'I
!l' at Three Mile Island that day, that it may have occurred?
Do l
i fs
(_)
3l you remember that?
I 3l Mr. Stello.
I don' t recall any specific discussion of a i
4 hydrogen burn inside the containment.
But there was con-i l
Sl siderable discussion that hydrogen evolution would be taking 3
i 1
I 6l place with a metal-water reaction.
But I don' t recall ever t
7 hearing that there was a hydrogen burn.
So the answer is I 3
have no recollection of hearing that on Wednesday.
I think 9
Fridav was the first time I heard it.
I 10 Mr. Scoviile.
Mr. mal' son, you were at the Incident i
i I
i Response Center, as I understand it, Wednesday afternoon?
I 1' !
Mr. Mattson.
Thursday afternoon.
/~'T t
\\/
Id Mr. Scoville.
I missed it., Okay, I needn't ask you the i
l Id !
question except to ask you when was the first time you heard t
I 15 l about the possibility of a hydrogen detonation or hydrogen 16 burn?
I7 Mr. Mattson.
App cxima tely 10 : 0 0 o ' clock a.m., Friday the la 30th.
I I' l Mr. Scoville.
Then the other question is:
What about i
A i
I I
20 !
the reporting of a pressure spike in the control room indi-l t
I i
l 2I l cating increased pressure in the contaimacnt?
i 22 Mr. Mattsen.
Oh, I should correct my answer.
I heard of the spike, pressure spike, at abcu: 10:00 a.m. on Friday l
23 !
('~1 2[
morning, the 20th, from which we conjectured that there had or,eewas 9eocrtees, Inc.
t Sc g
teen a hydrogen explosion.
i l
I l.
t
+
i 15 t
i
\\
_)
I' Mr. Reis.
Is that conjec'ture something that was 2j immediate?
I mean if you heard of a spike in containment, 3'
v.our trained mind wculd immediatelv. go to hydroc.en?
I
\\
4{
Mr. Mattson.
Mine did, yes.
5 Mr. Stello.
If it's a spike -- the thing that I saw on 6
the chart finally would be scmething that identified with a 7;
sudden release of energy such as a hydrogen burn.
I couldn't Si think of anything else, so I would have that association.
i 9'
Mr. Scoville.
When did you learn of the spike?
1 10 l Mr. Stello.
To me the spike has new become synonymous i
l II '
with the hydrogen burn.
No longer is it separable in my mind --
i i
12 they are one and the same and the answer is the same -- Friday '
i I3 '
morning.
Although there could be -- and again I would have to Id go back and check we did know that sprays came on.
I'm i
t 15 l trying to remember, when did we knew the sprTys came on?
I i
16 l thought we knew they came on Wednesday cr Thursday.
I 17 '
Mr. Eisenhut.
Yes.
I8 Mr. Stello.
Which would mean we knew that the pressure I9 had to go over 5 psi.
l l
20,
Mr. Eisenhut.
Four or five.
I t
21 f
' Mr. Stello.
Four or five.
I don't remember folicwing thatl t
99 i u,
in any great detail.
i I
23 Mr. Terrell.
Wculd that imply in the slightest bit that
(}
.d there was hydrogen involved here?
o, 3MJMetaf R eportert. l*C.,
i
,e i
Mr. Stello.
It would imply that the conta nment i
i
}
6 l
l
16 1
l (N
l' pressure had to be raised above the set point for the system.
I i
1 2
Mr. Terrell.
What could cause tha*'
I t
i
'"a I
3 Mr. Stello.
Without going back and checking through I
i 4
records, I could recall that containment pressure had been f
5 rising.
I don' t remember whether it had been getting close
{
l 6;
to tne set point by itself.
l 1
7j Mr. Eisenhut.
No.
I i
3 Mr. Stello.
Or whether you wculd have had to have seme-j i
i9i thing like the hydrogen burn.
i 10 Mr. Terrell.
I just wonder if there could have been any i
i I
II other phenomena that could have created the situation other I
l 12 !
than hydrogen.
i i
13 '
Mr. Stello.
Yes.
~
14 Mr. Eisenhut.
Let me try to bridge the gap for a second.
15 As Vic said, we took the phone call Wednesday morning.
I got 16 l to9 ether a 9roup of technical people at our office.
In fact, i
i 17 !
we were running a couple different groue.s, one on radiological,
IS aspects, one on sys tems performance.
I went to the Incident 19 Center on Wednesday afterncen.
l 20 I share Mr. Stello's view that there certain:.y was not any 6
21 real discussion, that I can recall either, of spikes in the 1
22 containment on Wednesday.
It seems like there may have been i
i i
23 probably -- time runs together because we we: s there e3 b
i 24 '
essentially en a 24-hcur basis. Either late Thursday or a enee necon,n. we.
- c Friday morning, there was considerable discussion abCut new t
f
3-t, si t
i f
l ',
- f. st the spike was.
If, in fact,it was a very rapid spike, l
i i
2{
it would be associated logically.ti th sc=ething like a hydrogen 6
3 burn.
The pressure in the containrent 2 know we'd been 4 l!,
=cnitoring and were following.
It was varying anywhere from 5,
2 to 4 psi and, in f ac t, appeared to have gone up and con.e 6!
down.
There was no question in any of our minds as to waether 7
there was fuel damage, becau;e I can recall as early as i
3l Wednesday morning hearing the values that the monitors were i
.I i
-l reading :or radiation in the containment, which can only be ti I
10 '
associated with one thing.
That is core damage -- in the I
t 11 i numbers that were being recorded.
Those values were increasing' 1
12 throughou Wednesday.
j
()
l 13 I seem to recall I joined Vi.ctor at the Incident Center i
l i
14 i ir.:e in the afternoon, around 4:00 o' clock, probably.
It i
15 !
became apparent that we were going to move several of our i
t I
16 staf f that had been working in our offices to the Incident i
17'
' Center.
i 13 '
Mr. Scoville.
Do you mean 4: 00 o' clock Wednesday?
l 19 Mr. Eisenhut.
Wednesday evening.
We both became involved, 1
I 20 ;
a little after 3:00 o' clock, 3:30, something like that, on E
21 '.
Wednesday morning, when we happened to have been working on i
22 another problem and got the same phone call.
We basically I
l
(~}
23 worked in two directions.
I
\\vi I
- d d
Mr. Ste.lo went to the Inciden: Center where he is a v. ember i
M f tt MofT1. f M..
?
a5.
on ne response team.
I s ta' red a the office with a team of t
A
_38 4
i l.
- c. eo.cle, c. rchably anywhere from twentv. to thirty technical I
i
~
I people.
When Mr. Stello went on the TMI site, I went to the 2'l 3
Incident Center to the respense team.
And I stayed there, I for what? -- three weeks, essentially.
4j
- guess, I
5j Mr. Stello.
Thirty-nine days.
I, 6!
Mr. Terrell.
I am not sure I understand.
We had some 1
7 sprays ccming on.
I was under the impression that those l
8i sc.rav.s were actuated uren certain thinvs hac. c. ening.
If it's t
6, I
9I sonething other than gas bubble --
l l
10 i Mr.
Stello.
Anything that can cause high pressure in the I
II l containnent would cause the sprays to co off.
The pressure j
i i
.t I'
was increasing because you were discharging the heated fluid rN 12
(_)
l l
into the quench tank which was raising the pressure inside of 13 14 the building.
Nov, that had been ccming up, but I don't recall 15 that ever getting to a high enough level to actuate the sprays.
16 But then, again, I will have to say parenthentically, please reserve me the opportunity to go back and review the 17 I
13 record because I'm not sure, but I don't recall that the i
19 f pressure due to the addition of the heated water ever.got high.
(
t 20 l encugh.
i i
21 1 Mr. Eisenhut.
I think that's right.
l i
1 22 f' Mr. Stello.
So that would leave ycu with something 1 ke f
2 i.
this burn to raise the cressure high encugh to actuate the J;
/'
V i
24 '
sprays, which is clearly pcssible.
t D EMtess 4 oorte's. loc.
.r.-Scoville.
What you are saying, then, is that there 23 I
1
19 I
1 i
1 could have been other explanaticas for the increase in l
2f pressure resulting in the turning on of the sprays.
3l, Mr. Stello.
By all means.
l 4
Mr. Scoville.
So that did not send to you, Mr. Stello', a 5,
signal that there had been a hydrogen burn that day?
But when i
i 6l you learned about the increase in pressure, the so-called 7
pressure spike --
8j Mr. Stello.
Very narrow, high -- right.
Very narrow band.'
9 It was about 20, 29 psi, the exact number --
i 10 Mr. Eisenhut.
About 28 pounds.
t 11 i Mr. Stello.
Fairly high and very narrow width.
That is j
i 12 indicative of a burn.
(2) i 13 Mr. Scoville.
That sent a very clear signal to you at 14 that point?
15 Mr. Stello.
Yes.
16 Mr. Eisenhut.
That's the point I was naking earlier that I
i 17 we were, in fact, talking to the site, in f ac t, specifically
- i t
13 seeing how fast the so.ike was.
19 Mr. Terrell.
But there was no correlation between the i
i I
20 1 spike and sprays --
i i
21 Mr. Myers.
There was.
i r
22 Mr. Terrell.
I'm saying in their minds.
I am trying to I
i 23 !
find cut if there was a correlation.between the spike. and (Z) i 24 sprays coming on because they came on al= cst simultaneously.
D ecerst Aeoorters, tec. ;
t 25 Mr. Stallo.
- de didn't know-about the spike.
There could l
I
i 20 I
I l
/ )
1l have been other reasons the sprays came on.
It could have m
I 2;
been caused by others.
I remember being precccupied with how i
3l high was the ccatainment pressure getting.
My best recollection getting high enough to actuate the 4
is I never remember it not, S'I sprays.
6' Mr. Reis.
Just to clarify, when did you knew the sprays t
i 7
came on?
8i Mr. Mattson.
Wait a minute.
Reactor building high-I 9,
pressure isolation signal actuated and isolated the reactor I
10 building at four minutes until S:00 o' clock at 4 psig.
That is.
f.
I 11 i the only pressure that shows on their chronology.
That is when'
(~)
12 ;
they finally got reactor building isolation, was high pressure, s_-
r 13 Vic.
l 14 Mr. Stello.
Reactor building isolation is listed on this t
i 15 :
one, and at the same time containment spray went up.
i 16 I Mr. Mattson.
At what time?
'I s
17 Mr. Stello.
9:49, which is 1350.
1 13 Mr. Mattson.
This is 7:56 in the morning, four hours into 19,
the event.
i 20 l Mr. Myers.
While we are on this pressure spike, I have i
l 21 ;
one question.
Does the f act 1
22 I Mr. Stello.
Go ahead.
23 !
Mr. Myers.
The fact that it measured 29 pounds on tha 3
s) 24 chart, was that number determined by the response of the 4.s ceras necone,,, ene.
e 25 instrument, or does anybcdy knew what the actual pressure 1
i i
e
2.'
l
,r I'
l 1j might have gotten to?
\\
2>i Mr. Stello.
That is, as I recall, a strip-chart recording.
t t
3 which would be an instrument' recording.
How much higher it might have gone and the instrument would pick up depends iin 4
i 5
the instrume.t response time.
But I would think in the narrow '
i l
6i width that that would rot likely be significantly different.
I i
7!
Mr. Mattsen.
It t/as six minutes wide.
So the ceak i
3, shculdn't have been nissed by tcc much.
But I den ' t know the t
i 9!
specifics of the instrument. Missed it sc=e, but six minutes, I
l 10 '
not six microseconds, as you often think about instrument II l resconse times.
~
O 12 '-
Mr. Sceville.
Coulc everv.bcdv. withhold for a minute i
s-I 13 i here?
14 (Pause.)
15 Mr. Weaver.
Mr. Stello, I want to go back to the spike i
l i
i and ask you -- I may be covering sc=e cid territory here, but 16 e
t I
17 I would like to hear it.
The people in the plant have main-l 18 tained that their personnel saw the spike at the time it 5
19 occurred, 1:53 p.m., on Wednesday, and that the: 2 were NRC I&E i
i' i
20 I,
' ecc. le c. r esent.
Can you "ust speak c.enerally on that?
What J
e i
c t
21 l do.vcu think haccened?
22 l Mr. Stello.
Not having been in the control rcom at the i
.O 23i time the s=ike cccurred or at the time the conversat:.cn toc.<
U l
1 24 place puts me at scme disadvantage, so I will have to --
s>.cm, a ece,mi. ice. '
25 Mr. Weaver.
Let ne ask you this, then:
Did you discuss 6
s a
i
22 l
I f
('
11 this with the I&E people on the spot, either at the time or
(_)
l j
2l afterwards?
j l
3' Mr. StellC-Aa*
-'d "eard about it, afte your visit, I
I 4
I went over and asked the person who was in charge of the i
i.
5 investigation if they have really concentrated on when did the I
61 inspectors truly know that there was a, quote, ' hydrogen l
l 7
burn," a problem in the containment.
I told him that I 3:
thought this was going to be a point of significant interest I
1 I
9 for people, and I thought he ought to make sure that he
[
i I
10 'i covered it.
That was.cretty much the first time that I had I
i 11 l spoken to an.yone about tne need --
i 12 Mr. Weaver.
Who was it vou scoke to?
i
()
13 Mr. Stello.
Bob Martin.
{
14 :
Mr. We aver.
F.e was one of the people actually present' 15 Mr. Stello.
No, he is in charge of the investigation tha
?
16 'i is taking olace uo there to make sure that he goes back and I
8 17l co;ers this point in detail.
I 18 l Mr. Weaver.
Right.
Did you ever, on Thursday or what-i i
19 '
ever, have any discussions with the I&E reccio?
Did vou ii 20 e cenmunicate with them in any way?
21 :
Mr. Stello.
I communicated with them throughout, from the i
22 time I'd been there.
I never discussed this hydrogen burn i
22 question with them.
("
,s 24,
Mr. Weaver.
The.y never volunteered?
.e 4ceni aeecreett inc.;
25 Mr. Stello.
And I never got it from them that I am aware i
6
23 i
[
t l
e-(_T)
I of until Friday.
Now, there are.erobablv several possible I
2 i explanations for that.
i 3'
Mr. Weaver.
I would like to hv.r.othesize.
i 4!
Mr. Stello.
One is that it wasn' t known.
l.
5-Mr. Mattson.
Could I interject for a minute?
I had read i
e 6'
to me about tnree hours ago over the telephone a memorandum to
.7i tne Ccmmission from the office of Inspection and Enforcement 3
which speaks to this question, what the two. rec. ole from I&E I
9 in the control recm were doing on Wednesday, Thursday, and 10 hcw thev first came to know of the spike.
I can't recite the i
i
- Ilf, letter to v.ou,
~
but it was 9enerally that the.v were in the i
()
12 shift superviscr's office and in the control room with the i
i 13 !
primary responsibility being to gather information for i
t Id !
transmittal to the Bethesda Incident Center.
And it goes en to I
jO ;
speak abcut their first awareness and transmittal of the r
i i
16 information concerning a spike to Bethesda.
i I
I7 '
I think rather than any of us conjecturing on what those I
i e"eople knew when, it is probablv. better to rely u on what those!
I3 e
i Uj people said,
- 0' Mr. Weaver.
I would like to hypothesire.
This will i
i i
21 l enable us to ask better questions, you see.
I would like to i
i
'n y-hypothesire, Mr. Stello.
i
/~
22 !
Mr. Stello.
Okay.
The I&E people, as Dr. Mattson has d'
24 indicated, were in the superviscr's office.
They were no:
n.sme,es n ecceen. sn.
^5 chv.sically catrollin~v and independen:lv readin~ matters e
t
24 i
I
(~
=atter by matter by matter to the best of my knowledge.
s i
9i
- i Mr. Weaver.
That is not their job.
I
, i, Mr. Stello.
That is not their jcb to be reading the 4
j individual matters.
In fact, in a situation such as the cne i
S'l we had, I wculd suspect that you would want to be a little bit t
6.
removed frem there to observe the actions just a bit further 7ll back rather than getting physically in the operator's way.
0[
Mr. Weaver.
Would it therefore --
9 'i Mr. Stello.
When we asked them for information, a specific:
I 10 !
i question, they went and obtained the information we were l
+
11 l looking for, and they were focusing on the information we were.
l l
asking for.
[}
3 13 Mr. Weaver.
So you would assume, therefore, that it's the 14,
operating personnel's duty to report to the 7.&E person seme i.
I 't unusual chart reading that has occurred, or not?
16 l I
j Mr. Stello.
If the operating crew knew something signifi-
$/
cant would have happened and appreciated, understccd that la.
I significance, yes.
i i
19 Let me finish my hypothetical.
When did this become known i
1 r
20 t
as a hydrogen burn an the containment is really the issue.
't l
Could it have been a spurious indication?
A wire that got i
l shorted cut and caused it?
Certainly.
There wasn't a dis-l 2e l-
\\
charge of liquid in there, but there were other reasons you
~
g3 l
()
,4 l e e_ecital 4 e:cr'ef t, i.*C.
could have had the problem.
With the acciden: that vc
- had,
- .c t
I dcn' t think anycne would have been able to, while Observing i
i e
a
25 i
t 1:
the instruments, draw conclusiens immediately without scme i
i 2;,
quiet time to study what they had.
3 My suspicien is that some trae ar ter the charts were re-i i
4l moved frem the control room where the people could 1cck at 1
5 tnem, and tried to get the correlation that scmething had l
l 6
happened, that it was some time af ter the hydrogen burn occurred I
7 when it was truly underst0cd to be a hydrogen burn.
3, I am not persuaded or convinced yet that the people in i
I 9l in the control recm knew that 'they had a hydrogen burn much i
j 10 l before Friday.
I heard nothing while I was up there that i
t i
I II.
suggested to me they knew they had a hydrogen burn much before' l.
i 12 wa knew it on Friday.
13 Mr. Weaver.
They certainly would have seen the spike, i
I4 l Let me ask you this:
If something goes wrong in a plant, l
IS just wrong, and you are right there and you rush into the f
control rocm, what are the major gauges you are going to lock i
+
16 i
I i
17 I at immediately?
What are the mest important ones to see?
I i
Mr. Stello.
I don' t think vou can decide that without 13 l
t.
~
l9 having some additional hypothetical in front of me.
Let's use i
20,
the one we had at Three Mile Island.
I I
21 i Mr. Weaver.
What are the ones you are going to have to t
22 1cok at?
/
23 !
Mr.Stello.
The program meters I was most interested in Three Mile Island was the primary system pressure, tne at 09 *E9C t* 44 EfOQf *tf t, I f'C.
'S tem erature of the het and cold legs and temperature of the i
i
2r r
rN.
t, )
1 incore thermoccuples with a confirmation that the reactor i
2l scrammed and all that stuff, too.
l 3i Mr. Weaver.
What is the gauge the spike occurred on i
i.
4!
called ?
l 5j Mr. Stello.
The gauge the spike occurred on is a gauge I
6 that measures containment pressure.
i 7l Mr. Weaver.
That was the first one you mentioned.
j i
BI Mr. Stello.
No.
The crimary system pressure.
l.
9i Mr. Weaver.
Within the pipes, then.
I 10 Mr. Stello.
Hot and cold legs of the primary system, I
11 -
incore temperature dis tribution.
And I say a lot of other
()
12 things you just ought to confirm.
You want to know that 13 you had the scram and vou want position indicating all the 14 rods went in.
You want a reactor building pressure.
Was i
j 15 pressure changing with time?
Did you see any increase in con-I, 16 i tainment pressure which would be an indicator of a LCCA, as an '
l,i,
example.
t IS Mr. Weaver.
But that i s --
t 19 I Mr. Stello.
We knew we didn' t have a loss-of-coolant l
20.
accident in any major sense because throughout the whole day,
}
i 21 '
cace the block valve was closed on the electromatic relief i
i i
22 !
valve, the system was fairly tight.
There was no -a=' '^ss i
23 h of system -- of inventcry.
()
24 So we had, as I described earlier, a situation where you y
co Fmes. a.cenm. inc.
25,
had a lot of fissica prcducts released in a closed and t
l 27 l!
isolated system.
l c
2j Mr. Weaver.
Would you call the containment pressure or I
3 chart, among the hundred things that you wculd look at, a
i 4l minor or majcr ccmponent?
5, Mr. Stello.
Well, you know, it's difficult now because i,
6, you are viewing everything with the wonderful --
I 71 Mr. Weaver.
Yes, I understand.
1 8 !
Mr. Stello.
-- hindsight.
Knowing that -- I guess I i
I 9,
believed you had a highly damaged ccre.
I believed you had i
I i
10 '
metal-water reaction.
It is not surprising to me now, nor was !
l II ;
it surprising to me then, that you would have had significant.:
l 12,
quantities of hydrogen to cause a burn.
Dcesn't surprise me.
s I
f 10 i That was af ter I learned of,it, very useful information to l
14 l help me decide what I ought.to do next, because I already 15 j had decided I had a highly damaged core.
16 f Mr. Weaver.
I have no further questiens.
17 Mr. Scoville.
You say you have already decided that you i
13 had a highly draaged core.
Was this because you learned about !
I i
19 the burn on Friday and other information had added to the i
20 information on Wednesday to make you believe the core was i
l 21 !
highly damaged, or --
1 2 2 ',
Mr. Stello.
The thing that persuaded me that the core I
23 !
had to be badiv. damaged was the fact that the exit thermo-
,\\
\\-)
~
-44 couple was above saturation temperatures.
The caly way you
- *w mi meoonen.i c.
can get that condition to exist is for the core to have been 4
i i,
~
28 1
4 I
I i
I
(_w)
I' uncovered, to get superheated steam coming out or the core.
Given I had superheated steam ccming out of the core, I knew 3;
the rods had to be running at signi::cantly higher tempe-=~"-as.
l 4!
to do that.
I knew that all of the incore thermoccuples were i
t 3l reading question mark, which means they were reading off the l
6 range of high scale, 7l Mr. Weaver.
What was that?
8' Mr. Stello.
I think that was 700 degrees.
9 Mr. Eisenhut.
Presumably it was 700, i
i 10 !
Mr. Stello.
On the computer it was 700.
They could have i
l i
11 '
actually been down and did make measurements, and the measure
(])
12 ;
ments, as I recall, indicated temperatures were significantly I
I3 '
higher.
Mr. Weaver.
They had a board up when thev were there dav I#
i I
before yesterday, flashing little red numbers on each thermo-15 i
\\
couple.
And they said that board had just been put up after 16 1
i I,'
the accident.
Where was the readings for the thermocouples?
i I
i I3 Mr. Stello.
On the ccmputer I
Mr. Weaver.
Nothing but the computer?
Thev. didn't have
'9 I
?
l
'O any of those flashing --
i All of those real-time reccrding '
21 Mr. Stello.
Yes, right.
I instruments that you saw were placed in the centrol roca as an se i
23 '
addition to moniter tranferring the plan: fr:m its forced
(-)
~-
circulating mode to natural circulation.
It was cre= ara:Orv to.
- e Acero 4ecortsrs Inc.
SC tha t step.
I i
h.
a
t 29 I
l l
i I
Mr. Weaver.
The ccmputer.was turning cut dollar signs s'
so all you knew it was over 700.
i 3i Mr. Stello.
There are three ways in which it could have l
4l been printing out the dollar signs.
It was off-scale low.'
i 5
You immediately say that is not possible.
It was off-scale l
6; high, or it was possible that they had somehow become disabled,'
7 !
breken, which to me would mean they were ranged high and they I
3I were brcken, because if they were broken you would expect the 1
9' temperatures to have gone well in excess of the 700.
The best l
i 10 guess I had is they were still functional but theywererangedj i
i ll i l
4 off their scale, and the ecmputer, because of the electronics, i
12 '
wouldn't take those above 7 00, s
s ')
13 Mr. We aver.
When did you find out the readings, the dollar 14 i
signs?
15 Mr. Stello.
Oh, boy.
16 i
Mr. Weaver.
Wednesday?
j l'
l
Mr. Stello.
Clearly Wednesday.
The hour en Wednesday f
13 i
I'm not sure.
I think it was Wednesday acrning.
~
/
~
19 i
Mr. Weaver.
Wednesday morning?
j i
i 20 Mr. Stello.
- Yes, i
i i
Mr. Weaver.
Knowing what I know now, it is evident to 2I l 22 l me that that was a real tip-off, as it seemed to be to you, i
i i
$} }
i Way do you think that wasn't a tip-off to other people?
i l
l km) a#
Mr. Stello.
I have no reason to believe -- I thcught I e
n n ere ne m m es.see.
nad persuaded people that we had a damaged ccre en Wednesday.
l e
30 i
i l
l l
1 1 I didn't knew of anyone wno b'elieved to the contrary --
)
i i
2!
Mr. Myers.
On Wednesday, I guess, we had people that told :
t 31 us the damage was like a half percent.
Then on Thursday they 4
tnought it might be up to 1 percent.
5 Mr. Mattson.
Half a percent or 1 percent of what?
I 6
Mr. Myers.
I guess half percent --
7{
Mr. Mattson.
Of the fuel was failed?
8l Mr. Myers.
The number was given without ever saying i
9{
exactly what was meant, half percent.
j 10 i Mr. Stello.
No.
I don't knew why anybody would be 1
guessing at percent of f ailed fuel.
l Il 12 '
Mr. Mattson.
Is part of the problem here the Mattson i
(~T s/
13 transcript whien says you couldn't convince B&W until late 14 last night we had a badly damaged core?
Is that the problem?
15 Mr. Myers.
No.
What we are trying to do is reconcile 1
16 in part our own recollection of trying to follow it at the l
I 1
17 !
time with, you know, what your recollection is.
Also just t
i 13 trying to understand what it is that people knew when all this I
i 19 1
was going on.
~,
I
)
Mr. Mattson.
The only point I would make is there is 20 1
21 damage; there is badly damaged and there is badly damaged.
22 i There was scme nuance of -- it got worse as we understocd more.
22 Encwing about the hydregen explosion', knowing about the
( j]
u i
d' primary coolant sample activity, knowing that the.emperatures o*nere sex,ws. me.
9C.
came down slowly and monocenically but tcok_a long time to i
31 I
4 i
1 ccme down, all those things put together ultimately led to
(~S s/
a 2'!
the picture we have today of what the core locks like.
That I
I 3i is generally worse, I think, than what was earlier thought, i
I 4 l But significant damage was realized early on.
I i
5, Mr. Myers.
Did anyone, say, on Wednesday actually seek i
i i
6!
to measure directly the voltage of the thermocouples?
You l
i t
7'l sa*v there was a ccmputer that was orogrammed to cut off at 700,1 1
8';
and if they were measured directly, that would have gone up I
9 higher.
t i
10 Mr. Mattson.
My recollection, we asked them for that l
1 11 on Thursday afternoon.
And because they were running around l
12,
doing a lot of other tnings, we didn' t get it until Friday b
j I
t 13 some time.
They were able to go behind and recalibrate all 14 the thermocouples eventually when theY had the time to do it.
s l
- Vic, do you remember whether they did any on Wecnescay?
l 15 l
16 We asked on Thursday and it wasn' t until Friday we got an 17 answer that I recall.
18 Mr. Stello.
The first recollection that I have, I think, i
19 was the folicwing week, that I had heard that one of the I
i 20 i technicians went dcwn during the transient and put an amplifier 21,
directly on the instrument, and I think it indic= ad emperature i
22 !
ranges on the order of 2000 degrees.
That is the oniv i
23,
(^)s recollection --
i
'\\_
24,
Mr. Weaver.
What was this, sir?
i o ne.<. meconen.im e 25 '
Mr. Stello.
It was during the transien:.
It would i
a I
32 t
I.
i
)
I in (J
1l probably have to have been Wednesdav.-
1 2'
Mr. Matts e n.
Had to have been Wednesday.
a 3l Mr. Weaver.
I see.
l I
4 Mr. Eisenhut.
Let me try to glue these two pieces to-i 1
5, ge th e r.
There was also, while these discussions were going 6
on some time Thursday, understanding what was being printed out t
7 and what was meant.
We were mapping in the core the actual I
i i
8i thermocouple readings as they are coming back on the scale.
i t
1' 9
As Dr. Mattson said, later on Thursday and Fridav. we were i
I i
l 10 '
asking the utility that when they had somebodv available to co
(
t i
- l 11 l in and actually take a reading, to calibrate those nu-bers, to i
l 12 do that, too.
x 13 So we were asking them.
Both of these things were 14 happening.
We were trying te not only get a handle on what 15 the nurbers were pretty precisely; we wanted them to go in and i
e i
16 find out exactly what it was.
l, 17 Mr. Weaver.
You mentioned a reading some place o f 2000.
18 Mr. S tello.
My memory is that seme time later af ter the i.
19 transient, I heard that one of the instrument technicians went f
l 20 l in and placed an amplifier during the transient en 9.e i
f! thermocouple wires themselves and had cotten a readine that i
21 i
i l
22 l indicated a reading on the order of 2000, I
n 23 Mr. Weaver.
Those thermccouples can read any temperature.
(v) i 24 ! Ecw hich do thev co, the actual ther=occuple itself, as opposed ps.eersia:ecorteri.inc.
25, to what it reads cuu?
I l
t
I 33 1
I
[
1 I!
Mr. Stello.
The instrument technicians tell me that you i
i I
2 I will continue to get reading so long as the wires are intact.
I 3,
Mr. Mattson.
Ibout 2300 degrees.
I think they said 2300.
i I
4' Mr. Eisenhut.
Certainly about 2000.
5l Mr. Myers.
But was the 2000 ever reported to Bethesda?
I 6'
When did you hear that in Bethesda?
i 7;
Mr. Mattson.
I had never heard it before in my life.
i I
3!
Mr. Stello.
I'd never seen any discussion or analysis, j
l 9l But again, I don't attach tco much to it.
If you had any
[
t i
10 metal-water reaction started of any proportion, which we i
i 11 clearly believed we had, temperatures of the crder of 2000 t
i
{])
12,
degrees are not the least surprising.
13 Mr. Mattson.
The postaccident analysis says there were I
i temperatures in excess of 2000 degrees over a significant 14 15 portion of the core.
The number doesn't surprise me.
This is i
i I
I have heard it was actuallv measured durinc. the 16 the first i
I 17 accident.
Mr. Weaver.
What about the figure I heard of 3700
{'
t 18 i
i 19 degrees in certain portions of the core?
I l
20 Mr. Stello.
Yes, the characteristic of fuel 's such that a
21 ;
there is a very sharp knee in the curve, that unless you take 22 !
the exide matrix to a temperature probably in excess of about l
22 3300 cc 3500 degrees Fahrenheit, the fission product releases 3
J 24 :
are very, very low; that the release rates frca the fuel a.enew nwenen. tee.
25 matrix itself stays low until you cross the knee of the curve
f 34 i
i 1
in that range.
After you cross the knee, then the release
(]}
I 2l rate frcm the oxide material goes very, very high, up to 100 i
I t
3; percent.
j l
i Based on the amount of fission product material that h'as 4
5I been released, it is now believed that there was a f
6!
substantial fraction.
That might mean 10, 20, 30 percent of I
Ii 7
ene fuel had to get to high temperatures to release to I
3; volatiles, the gases that were in there at the quantities which i
I 9:
were measured.
So that the matrix of the fuel clearly had to t
i 10,
get to elevated temperatures.
i 11 i Mr. Weaver.
Why didn't that temperature then destroy the,
I 12 i wires?
Why are the thermecouples still working?
i fs(-)
i 13 Mr. Mattson.
Let me interject here that there are several schools of thought as to how high the temperature 14
.I
-i I
i 15 l reached.
What Vic has just described is one that has been i
16 I crevalent now for several weeks.
There is another school of f
I 17 thought that says the volatiles can be released when there is i
18 a formation of a uranium circonium UTIC which occurs at a 4
19 !
slightly lcwer temperature than the 3700 degrees 'Jic has i
2 0 ',
described where One knee in the curve is for the cooking out i
21 f of the volatiles frcm the cranium dioxide.
That temperature 22 '
is in excess of 3000 degrees.
I think Mr. Levine told the 23 ;
ACRS tcday it was on the order of 3100 or 3300.
- escapes d,s There are several schools of technical though:
24 my memory.
- . 7mou nexnm. inc.
25 as to exactly what was the peak temperature so you shouldn' t i
h
35 i
l
('
Ij be surprised if you hear that number bounce around a little--
I 2
bit between 3000 and 4000.
3' Mr. Weaver.
I understand.
Very good, and I appreciate I
i 4!
that.
I am trying.to clear up that the thermocouple melts, I
5l the wire melts at 2300 degrees; yet you are talking abcut 6!
over 3000.
7 Mr. Mattson.
The thermocouple is in a housing which has a l
i 3i irconium sheaf --
1 I
9l Mr. Weaver.
So it is protected?
10 Mr. Mattsen.
The circonium reacts with water and the II,
cladding was badly damaged, inside the circonium sheaf or 12 i additional sheafs.
Finally, the thersccouple is a stainless O,
i 13 l steel tube in whien there is magnesium oxide, a refractory 14 material which is good to have high temperatures.
Inside of i
15 !
tnat magnesium oxide are two little wires of chromal-alumel I
16 !
whica join at the top.
So there was quite a lot of protection 17 l for the theraccouples, and they are not right in close contact i
t IS !
with the uranium, l
19 (
Mr. Weaver.
How hot do you think it would have had to get I
i 20 '
to put the thermocouples --
21,
Mr. Mattsen.
2300 to 2500 degrees.
l 22 Mr. Weaver.
We are ta'. king about over 3000 degrees.
t t
23 :.
i, i
Mr. Mattson.
The theraccou=les didn' t see that.
(~)N l
(_
i 24 -
Mr. Weaver.
How het wculd it have had to have gotten to m seere neocaws. nee.1 l
.g put the theraccouples cut?
l 4-l l
1
-Q I
i (h8 1,
Mr. Reis.
How hot would it have had to get in the core 1
d 2!,
in or.er to get at the 2300 degrees in the thermoccuple?
I I
I e
3,;
Mr. Mattsen.
I don't think you can follow the gecmetry of I
i 4 i what tne uranium did --
5, Mr. Weaver.
It could have gone very high?
l l
l 6'
Mr. Eisenhut.
The thermocouple is raised above the fuel, t
I I
7, too, so it's chvsically removed.
8.
Mr. Mattson.
I don't think anyone has attempted that i
i 9'
calculaticn.
You could probably attempt it.
It would be l
i 4
10,
very difficult, very inaccurate, t
11 Mr. Weaver.
I just wondered what we knew and didn't know.'
i i,
( ')
12 Why was the ccmputer set at 700 degrees, when the thermo-l
~-
I I
13 ;
couple will obvicusly read much higher?
i i
i e
14 i Mr. Stello.
These thermocouples are not put in there as t
15 a device to monitor accident conditions.
They are put there i
i l
16 for the normal operating conditions of the core.
And normal i
i I
17 operating temperatures are considerably below 700 degrees.
i i
I 13 Mr. Weaver.
Surel.v.
0 l
19 I Mr. Stello.
They ane arranged and cptini' zed for that i
20 i cur:cse.
Tnat is what they are needed for; that is what thev i
I i
l 21' are put there for.
22 i.
Mr. Weaver.
Thev are not in all plants, Aus: a few s
e I
e I
(~')
23 ;!
experinents --
(/
]
l
,d Mr. Stello.
That is correct.
Ormt an newws. nee.
at Mr. Mattson.
They are n t required in any plant by NEC.
3 I~'
1 Mr. Stello.
The fact thev were there and able to read,
\\
2 { made use of that technical informaticn because we hac. it.
3 Mr. Weaver.
It wouldn't have been hard on the cc=puter to l
I 4j put it up higher, or nchody thought of it?
5 Mr. Stello.
I think you would have to have electronics I
6!
such that you may have been able to switch from one to the i
I 7{ other.
I would think the electronics covering the range would I
ai have to be different.
I 1
1 9
Mr. Mattson.
As I understand it, they later changed the I
t.
10 -
700 degree cut-off to 900.
That was the limit of the electronics I
11 l in the ccmputer.
12 Mr. Weaver.
I
.se.
New I am going to ask, unless there are
[}
13 further questions by anyone, Dr. Eisenhut to review the accident, 14 j beginning from the opening mcment, with everything you knew --
1 I.t ' not necese trily your personal kncwledge but evervthine. vou know, i
16 i sir, now -- hindsight and present sight and everything.
If you 17 wouldn' t mind j ust sta: Ling --
16 Mr. Marrictt.
Just one question.
Is your definition of i
t 6
~
19 I fue.1 damage the meltdcwn of the fuel?
r I
i t
20 Mr. Stello.
No.
Damage, to me, I would characterize i
i l
21 something as being failed if the fuel were just breachid and the !
l 2 i fission gases allowed to -- but the integritv. of the claddin:
s I
22'was such that I had no question.
I think I would use the words i
O ss 74'that fuel would fail during the accident.
When I' characterize m.i.cen a.=rwn, tre.
- 23. fuel as being damaged, te me that is fuel that has had its t
1 4
e
38 I.
i
_)
1 integrity c o m e r. c = i s e d b v., for' exam.cle, in this situation, l
2l oxidation of the cladding material to significantly change t
3 l its properties and its integrity.
So damage starts, I thi nk,
I I
a where the integrity is, in my op,nion.
1 5,
Mr. Weaver.
Dr. Eisenhut, would you start with everything I
6 you know.
i I
i 7!
Mr. Eisenhut.
Sure, I'll go through the chronology I
g>. basically.
There are a few things we are still debating.
l 9
Mr. Weaver.
Make note of those as you come upon them.
10 t Mr. Eisenhut:
Feel free to interrupt.
11,
Mr. Mattson.
Yes.
l t
l 12 Mr. Eisenhut.
Now, at 4 :00 o' clock from our reference time '
U
13
-- we have made it 4:00 o' clock for ease of convenience --
i I
14 i there was some operators working or engineers working doing l
l 15, maintenance on the feedwater system.
Apparentiv transient i
1 l
16 ) situation occurred where they lost all the feedwater.
The i
17 ' reactor was operating ar around 9 8 percent power.
The transient.
It 13 was apparently initiated by a loss of condensate pumo which l.
19 caused loss of feed pump which caused the turbine trip, as one 20 ' wculd ex=ect.
The auxiliary feedwater pumps started very I
l 21 shortly, as they should, ve promptly; all three pumps startad.'
I u i, Mr. Weaver.
Auxiliarv.
t.
4, 1
i
{
33 '
Mr.-Eisenhut.
Auxiliar.*. f ee dwa te r c umo. s'.
The main c.um:s
-}
24 I are running; they are desj :ced to go to the back-up pumps he,.:, cent nexn.n. ice.
25; immediately, i
i t
i
r 39 l
I yl 1l
.Mr. Weaver.
Thev. c. ump into the primarv. s v. s tem?
i I
2:
Mr. Eisenhut.
What they do is they pump the secondary i
I I
3j system and keep the ficw going in the secondary system so that 1 I
l you keep flow to the steam generator so you keep coolir.g.
l A
4 5!
few seconds into the transient, the electromagnetic relief l
6 ! valve -- this is' a pilot-operated relief valve on the pressurizer i
7 !
on the primary system -- opened at about 2255 psi; it opened I
a!
as designed.
Yet, except for this transient, we would calculate 1
9 today that it would open because of the pressure spike.
10 The cressure increased for about another -- a little over l
I 11 :
100 psi.
The pressure keeps coming up after the valve opens.
l l
6 12 The reactor tripped on high pressure.
That was like about i
1 13 !
9 to 10, 12 seconds into the transient.
At that point the t
i la l reactor scrammed.
The system had shut dcwn.
The main feed-l 15,!
water system had stopped flowing.
The auxiliary feedwater I
16 l sys tem, the pumps had started.
They should be providing l
l 17 coolant to the steam generator.
18 Mr. Weaver.
Up to this point has everything progressed I
19 quite normally with the exception, of course, of the caralyst, I
20 of the condensate pumps going off?
Af ter the condensate pu=ps l
l l
I.
going of f and to eds point in your chronology, has everyrhing l
i 21 i
t I
22 '
progressed the way it should?
23 Mr. Eisenhut.
I think so.
A f ew seconds later the valve i
(-)
s/
24 should have opened the auxiliary feedwater.
'e.a.ec. a ecen.n. rne.
Mr. Stelle. Feedwater pumps had ccte on but the valves f
25 l i.
i.
40 t
i
~
I 1
were isolated.
The valves should have started to open af ter
(]}
i 2
the pumps were on.
I 3!
Mr. Eisenhut.
It's a matter of seconds as to whether it is i
4l or isn't.
At this point, as Vic pointed out, one of the.iext
]
t l
1 5'
items we have on our chronology is, even though the auxiliary 6 1 feedwater pumps started as one turbine-driven pump to electrical-i j
7 !
criven pumps, they had all started at the early time, immediately i
g upon s tarting, and coming up to speed the discharge valves 9:
should start to open.
At some seconds later it was noted the I
10 l valves were not opening or had not opened.
That was, in tact, I
i 11 ;
the failure. As to exactly what second they were supposed to have i
l' 12 '
started, it's a matter et how things were set.
But it's about U,'
13 this time where the first majo; problem with the system began.
14 Mr. Weaver.
All right.
Now, the valve, going into just a 15 bit more detail, what exactly happened?
l 16 t Mr. Eisenhut.
The valves turned out; there was a blocked 17 valve on each of the systems.
18 Mr. Carr.
Mr. Chairman, I'm sorry I'm late.
One thing 19 occurs to me.
I don' t believe our own committee room is being t
4 I used.
20 4
t l
21 !.
..tr. We ave r.
Right.
I was going to wait for the next vote i.
i but --
33 I
i (Pause.)
i
,3 I~')
24 Mr. Weaver.
So far everything is operating according to m.pnere neocaws. x,c.
normal af ter the f ailure of the condensate?
l
'S 1
i i
i f
)
1 Mr. Eisenhut.
That's right.
On the loss of the main 2'
feedwater system shown here, there is a signal to start the i
3 emergency feedwater pumps.
There are three pumps shown en the 4
lef t-hand side down to -- enis is one train of the simolifiec 5l system.
I i6i There are really three feedwater pumps.
One is a turbine-I i
I i
7i.
driven.cumo, and two are electricallv. driven.
Those pumps are i
8i civen a signal to s tart.
Those pumps did receive the signal 9
to start and the pumps did start, i
10 '
Now, a few seconds af;er the pumps have come up to speed -
11 th ev. are designed with an inte5 rated control system that should I
I i
(
12 ;
open discharge valves.
j I
13 l Mr. Weaver.
Where is this valve?
i I
ii 14 Mr. Eisenhut.
There are really two valves.
One shown by i
15 an X on this figure witn an arrcw next to it.
i 16 l Mr. Weaver.
Does everybcdy see that?
i 4
t I
17 l Mr. Eisenhut.
One valve --
I I
i 13 Mr. Weaver.
Right here (indicating).
j i
19 Mr. Eisenhut.
Yes.
i 9
20 l Mr. Weaver.
That valve was closed?
21 f Mr. Eisenhut.
Well, there really are rwo valves there 22 ' The firs t valve -- you could picture it where the X is -- is 1
t
(~)5 23 designed to be closed and autc=atically receive a signal to i
24 l cpen.
There is ancther bicck valve dcwnstream of that.
On ca.s,eere neconers. nee.
25 this chart it would be ibeve it.
i
}'
42 e
i i
i
()
1 Mr. Weaver.
Oh, j us t the
- s traigh t line?
l 2
Mr. Eisenhut.
On the straight line there would be ancther t
3 valve.
That second valve was closed.
There are three --
4 !
Mr. Weaver.
Shoulc it have been closed?
5 Mr. Eisenhut.
It should not have been closed.
I i
6{
Mr. Weaver.
That is the valve we think may have been a t
i 7l violatio n.
8 Mr. Eisenhut.
That'n right.
There should have been a I
I i
9i valve closed that should automatically open, tue in this case 1
10 l, there were discharge valves cicsed which should have been 11 cpened, sort of a second valve.
l 12 !
Mr. Weaver.
Now, did indeed the valve marked by an X i
13 open as it should have?
14 Mr. Eisenhut.
I think it did.
We didn' t eneck.
I cer-15 l tainly don' t know personally firs thand.
I would assume it did.
l 16 l Mr. Weaver.
So it's the second valve that was closed that 17 didn't fail, but the sys tem f ailed because it was closed?
18 Mr. Eisenhut.
That is right.
1 i
i 19 l Mr. Weaver.
All righ t, i
l i
20 ;
Mr. Eisenhut.
Now, the reactor tripped, as the pressure l
i 21 in the reactor went hivher.
It tripped on high pressure.
The I
l I
99 = set :oint was arcund 2355 esi.
--I i
i I
23 Mr. Weaver.
Had this valve been open and the emergency
()
l 24,!
water got into-the secondary system, feedwater, you would not o.snms n exeers. s ~e.
25 have expected the 2350 psi to have been reachec?
i i
I
43 1,
i l
1
()
1 Mr. Eieennut.
You would not have expected such a high I
i l
2l pressure.
The S&W reactor has an automatic run-back system on i
3i it so that the plant, on a turbine-trip situation, can auto-t f
4' matically go to emergency or auxiliary feedwater ficw.
In fact, l
5' in theory it is designed to, you don' t scram the =lant at all.
I 6
You can, in fact, in an economic consideration, return the plan 7
to service fashion.
If you scran the reactor and it trips out, '
i 8,
it gets longer to get back into a power operation.
9i Mr. Weaver.
But the likelincod is if the closed valve had 1
I I
10 been open, you would not have had a scram?
I 11 f Mr. Eisenhut.
Not necessarily.
It should --
f i
I
(~)
12 Mr. Weaver.
The scram occurred how long after the conden-s-
13 sate pump tripped?
14 Mr. Eisenhut.
We estimate between 9 and 12 seconds af ter i
15 '
the time v.ero which, for the sake of our discussion, we called
?
I I
16 !
4:00 a.m. precisely.
The difference in the chronclogy is on 17 the order of seconds.
I la Mr. Weaver.
So we are talking about 12 seconds?
19,
Mr. Eisenhut.
Nine to 12, in that range.
s I
I 20 t v-Weaver.
For the scram to occur, the whcle thing to i,
21 occur.
l i
22 1 Mr. Eisenhut.
Yes, for the reactor to have cotten uo hien j
('N 23, enough in pressure to get a high-pressure scram,
/
s/
t i
w 24 Mr. Cheney.
When we were up at Three Mile Island on
	C1* 41 A fDor*sf t, if*C. l 25! Monday, this loss of a condensate pump was attrib u:ed, as I i
I
+
i
44 6
i I
I recall, to a power outage.
pJ
.i 2
Mr. Weaver.
They answered my ques tien they didn' t know.
1 o
3 Mr. Cheney.
When we were there on Monday, the statement was i
4 made to us by the utility pecple that the loss of the condensate 5
pump was somehow related to an interruption of the flew o' I;
l 6
electrical pcwer to the c.u==..
i t
7 Mr. Eisenhut.
I certainly don' t know what the cause was, i
i 3
The next couple of weeks I think our Inspection and Enforcement 9li investigation will have proceeded to the point where we should 10 have somewhat of an answer to that.
I certainly don' t know now.
i i.
II Mr. Stello.
I heard the same response.
- l i
12 i Mr. Cheney.
You heard the same thing, did vou?
C s
13 Mr. Stello.
Yes.
Their chronology, however, said that the '
i
.t' 14 activity that was going on at the time, they were trying to l
15 clear out a particular line.
In the process of trying to clear '
e 16,
out that line, it is conceivab.le that for some reascn the I
i i
17' pcwer sup.oly to the.cump was interrupted.
But until thev have l
18 gene back and looked some more, I don' t know that we will really 9
1 19 have an answer.
i i
+
20 :
Mr. Carr.
Question.
I 21 i Mr. Weaver.
Surely.
i 22 j Mr. Carr.
Condensate pumps, as I recall during our visit, t
e i
23 :
there were three numes; ri c.h t?
i h'-
24 '
Mr. Stello.
Yes.
- c am si me
- mrms. ide.
25 Mr. Carr.
Did all of them go out or only one of them?
i i
t e
I O
1' Mr. Stello.
As I understocd the answer dat was given en l
v l
u 2'
Mondav, one of tne condensate cume. s trinced which was sue.:.1v. ing 3
pressure to the heacer for the feedwater pumps, and =ecause I
i i
4 I the header pressure dropped, it caused all three feedwater pumps' S
to trip of f the line due to a 1 css of suction pressure, i
l 6i Mr. Carr.
So you had three sequential pump trips, right, i
i
+
1 7l bing, bing, bing kind of tning?
l 8li Mr. Stello.
In theory, four.
On the --
l 1
I 9i Mr. Mattsen.
The enronology at the to.
o f the cage,. car.e f
I 10 l 2, if you nave a copy, says dey were running tao feedwater 11 pumps, three condensate boos ter pumps, and two condensate pumps.
p 12 -
There are, in fact, tnree sets of pumps like that.
The O
13 chronology says only two condensate pumps were in operation i
and one stripped out of service, which led to the further trip-l 14 t
15 pine. of two feedwater n. umes which were all that were in service.!
i l.
16 Mr. Chenev.
Then I assume we will folicw un at some coint i
17 on this s tatement.
They have here condensate pump A stopped.
9 i
la Mr. Stallo.
That's the one that s tarted the sequence.
That i
19,..c ume. s ue.. lies. cress ure to the feedwater pumps.
It was tri.=ced, l i
I 20 '
althcugh there were other nume. s running.
There wasn' t enough I
21 !
pressure to the inle side of the feedwater pumps, and they in i
i.
I 22 I turn tripped.
That was the answer you were given en Monday.
It t
i 23 { certainly is a reascnable answer.
There is no reason to debate v
)
24 '
it.
es,en aeoerms. t.ac.
i 25 '
Mr. Cheney.
Co you expect we will knew why the p u=p I
i
46 5
i
('s 1
(s)
I' tripped at that time?
It's a question --
l 2
Mr. Stello.
Perhaps later.
As part of our investigation i
I 3I that is going on right new, looking into the sequence of events,.
4 ho c. e fullv. it will shed some light on that.
Until we have seen i
5 that report, I won' t know.
l I
6 Mr. Weaver.
Mr. Jack Hermein -- I asked him the same 7l question -- told me that they hoped to find out but they didn't:
t i
8, know then.
(
I i
9 );
Mr. Marriott.
Mr. Chairman, a question.
On this diagram I
I 1
where are the three emer;ency feedwater pumps?
Are they in the 10 1I 11 same location as che feedwater pump that stopped?
I 12.
Mr. Eisenhut.
No, at the bottom of the page you will see I
13
" Emergency FW Pump. "
That is feedwater pump.
This chart only 14 graphically shows parts of the system.
In parentheses. it says 15,
three pumps.
They feed all three.
Similarly --
i 16 l Mr. Marriott.
Does the feed otmo tnat stcos simultaneousiv,
i i
1 4
17 with the condensate pump always go out under all conditions?
Iff 13 you lose, the condensate pt.mp, would you lose the feed pumos as i
I well?
19 I
20 Mr. Eisenhut.
I'm not sure what -- what it is is a hcoster; 21 - one t.umo. runs to the ne:.t cump.
O ne c. umo --
\\
I 22 !
Mr. Marriott.
The f act bo th pumps sto.c.eed simultaneously, i
I
,/~h 23 - v.cu woult have expected that had vou 'icst the firs t pump?
(_.
24 :
Mr. Stello.
Yes, at f ull pcwer, which is where they were.
opecn neoo,ws. wa.
25 es s entially.
If they had been in part pcwer operation, perhaps
i 1
e 1
s 1I not.
But at the full ecwer condition when all of the pu==. s are.
s i
2, operating near their normal design values, the answer is yes.
l
}
4 3!
Mr. Marriott.
Didn' t I lis ten care fully at the plant, dat '
i 4
that is one area that ought to be corrected in this scenario of '
j 5i nuclear ecwer clants. Should that be allowed to hapcen, or I
6I should there be a dif ferent syste= in there to assure tna:
i 4
7!
would not hatten?
i 8l Mr. Stello.
I think that tnat is a question tnat deserves I
I 9!
scne attention.
However, it is not clear to =e that you w:ll I-10,i ever be able to eliminate f eedwater transients, ter sc, even if I
t 11 '
you somehow nave the split.
Because they normally are pumps 12, that' require very large quantities of electricity to drive them..
13 i Thev no r= ally operate on o ff-site power sys tems.
If you lose 14 of f-site power, for instance, if there is a pcwer interruption, i
15 '
v.ou will c.enerally lose all pumps.
It would be interesting to i
16 I see if the number of transients that could be avoided a year I
i 17i could be signi'icantly reduced if it were modified some way or l
I 13 '
ano the r.
19 Mr. Mattsen.
We have been locking at the' generic i= plica-20 t tiens of Three Mile Island for de otner 3&D plants.
I, 21 :
Mr. We aver.
Yes, I wanted to develop that question.
I i
l -
22 l Mr. Ve nto.
Mr. Thairman, I have sc=e ques tions here.
i 1
I i
Q 23 li This is basically a syste=, an' enclosed systa=, is it net, wid V
l 24 i tne condensate pu=ps picking it up from de turbines; :.s tha:
ED 8ece'ai tecor*ers, tae.
8 25 ric.ht?
4 4
h l
l 1
I i,
1 Mr. Eisenhut.
Yes, sir.
{
()
l 2
Mr. Vento.
Would the change or substantial. chan ce in 3
pressure coming out cf the turbine which could have been ca. sed j 4
4i by another secuence of events have any ef fect on the operation i
?
i5' o f thes e.c um.os, a dramatic reduction in c.ressure or an increase.i 6
in pressure or for whatever reason, have any impact on these i
7!
pumps, this condensate pump?
I 3l Mr. Eisenhut.
Certainly at some point in this entire i
9l closed loop, basically it's more generic.
If the feedwater i
1 1
10 l was inceded or storced at some coint in the loop, the 1 Cop will !
i I
11 I shut down.
It's the same type phenc=enon; it a valve closes f
1 12 '
that loop will shut down.
13 Mr. Vento.
What I am specifically talking about is 14 oscHlation or substantial change in pressure.
In other words,
l 15 ;
this is a very simplified diagram.
Do we have, in this i
16 ; pressurized system we are icoking at, any type o f a s a f ety i
17 mechanism which would prevent a blewout in one line of tne l
i 13 sys tem from af fecting another, in other words, in terms of back 19 pressure?
I 20 Mr. Eisenhut.
I'm not sure.
The system we are talking i
I 21 i about easically is a nonsafety grade system.
It is not really e
I.
1 22 i a safe ty system of the clant cer se.
The main feedwater is no t.;
I t
p)i 23l Mr. Vento.
Would changes in pressure af fect the operation u
I 24 o f these pumps, substantial enanges of pressure caused by over-M,c.,
a,oe n.n,tre.
25,
heating?
i
,a i
1 i
"=-
S'*
2
"' =^* "*""*""*= = =""
Y
- =""** ^' "
C' i
2; vacuum, several inches, 20, 30 inches, very low pressure.
So i
3 ene pu:ap taking suction from t is desicned for tha: condi tio n.
i t
4' There are a number o f c. ume. s.
There is the condensate, conden-t i.
b' 5
sate booster pumps, and the feedwater pumps.
If for some I
6, reason the pumps that follcw this pump drvinc off the condenser '
I l
i 7{
sees a condition of icw pressure, low suction pressure, they t
i 8 i, are designed to tric o f f to c. revent them from being damagee..
l 9;
So the pressure transient, if it's one caused by one o f i
10,
tn es e ea rlie r c. ume. s in the train tric. oinc., it will cause the 11 i otner cumes to tr:.c.
That is precisely whst hac. cened in Three i,
i I
12,
Mile Island.
I
)
i 13 Mr. Vento.
But the condenser cume. do es no t --
t, l
14 l Mr. Stello.
Tne condensate pump tripped.
It caused tne i
l 15
i 16,
rapidly.
That caused the feedwater pumps by their cwn i
17 protective circuits to be dropped of f f rom une line.
18 Mr. Ven to.
You don't knew right new.
I guess, in other
'l words, you are saying the condensate pump' doesn' t operate on 19 I,
i 20 that basis, tho uc.h, except that it's operatinc in a vacuum, so l
21 if it lost its vacuum it would stop, would it no t?
i i
1, i
I
'. 2 l Mr. Stello.
That would be more f avoratie.
If the vacuum I
23 is lost, it would have a creater cressure to cump the water, O.
,6
(
s not less.
24 c
3 9 eceras Mccetsrs, f ee.
25 Mr. Weaver.
Mr. Stello or Mr. Ma::sca er Mr. Eisenhut, t
50 1
1 I
I the closed valve
-- now, you say' this is a closed system,
(])
~'
I 2
the secondary system.
Suppose the valve that was closed had i
3l been opened as you say it should have been.
Would the system 1
4{
have worked?
Would it have prevented a scram, even with this I
I I
5t condensate pump tripped?
It strikes me if it's a closed system '
I i
6l the whole thing has failed.
I 7:
Mr. :Mttson.
I don't think you can say for sure.
The i
I I
8l way it is designed to function is if those valves are open and I
i i
9:
the auxiliary feedsater comes into the steam generator the way i
I I
10 it 's suoposed to, then the pilot-cperated relie f valve will 1
11 l open momentarily, relieve the rising pressure, the reactor will r's 12 ramo down in cower, hold a low power level with heat rejection Q')
13 l througn the auxiliary feedwater sys tem.
Pilot-operated relie f I
i i
14 l valve will close in that sequence, and it's a normal feedwater I
I 15 l transient.
l l
16 Mr. Weaver.
Where is the relie f valve? The very one we are '
17 talking about at the top of the pressuriner?
i i
18 '
Mr. bu ttson.
Yes.
i 19 Mr. Weaver.
In other words, what you' ve got is the sys tem.
I i
1
+
20 feeds frem the emergency feed pump on through the steam i
21 '
generator.
22 !
Mr. Mattson.
Out the side of tne s team generator.
fi i
}
I 22' Mr. We ave r.
What?
gs
%,)
f i
24 Mr. Mye rs.
It comes out here.
j
)
- o seceu 9 oce m s.ine.
25 Mr. Eisenhut.
Really what it is -- let me try to make it l
)
51 e
(
1; very simple.
I 2l Mr. Wec."er.
I see.
l I
3 !
Mr. Eisenhut.
It's basically, if you don' t remcVe the l
4 heat in tnis --
1 I
5i Mr. Weaver.
Would you mind coming up here and showing us 6
on this graph exactly what happens to the water?
7!
Mr. Weaver.
That has failed.
You are now going to send j
i 8!
water through he re.
This valve is open instead of closed.
Whad 1
i.
9 happens?
10 Mr. Eisenhut.
What you really need, on the primary system
[
t l
11 side -- remember, tne c. rimary system is a closed loop.
I i
O 12 '
Mr.
Weaver.
- Richt, j
U 13 Mr. Eisenhut.
Water comes out of the too of the reactor I
i l
14 ;
s team -- two s team generators on the plant.
The water comes t
15 l;tnrcugh the steam generator which consis ts o f thousands of small'.
l 16 I tutes.
Water from the primary-system side goes through these I
i e
i 17 '
thous ands o f tube s.
What you need to do is you need to have i
l 4
13 cold water en the outside of these tdes so that the heat is 19 l removed.
That water comes in as cold water in the bottom, goes 5
20 '
en tne outside of the tub e s,
cicks uo. the heat, and goes cut I
i 21 i anf. turns the turbine.
If you lose that secondary water, what I
22 happens is this sys tem, the centrol rods co in and it scrams.
i i
p 22 Mr. Weater What you are saying is the heat is no:
d 24 '
trr.s fe rred.
Q3920'51 NCCCr*t?1. '?C.
25,
Mr. Eisenhut.
Ri gn e.
Even though the reac:Or scrams even
i 52 i
i
()
l ahead of time, on sota plants you can have the reactor scram --
i 2
af ter tne reactor scrams there is anough residual neat, generally t
t 3
we call it roughly decay heat -- even if tne reactor scrams you ;
l 4
must remove that heat or the plant is going to keep going up.
r 5
Mr. Weaver.
I understand..
That is a diff eren t issue.
Thel l
6 issue I am trying to establish is where is this water going?
I.
7 Mr. Eisenhut.
Comes in through the same pipe -- this one i
8' goes tnrough.
It's actually a separate pipe but it goes I
t 9
through the s team generator on the outside of the tubes to cool ;
10 '
those tucer.
11 Mr. Weaver.
You mean there is another sys tem tnat circu-l l
t i
O 12 !
lates here?
l sj 13 Mr. Eisenhut.
Yes.
14 Mr. Weaver.
It doesn' t circulate through here, is my 15 que s tien.
i 16 i Mr. Eis ennut.
It dce s no t.
This can actually pump water.
i e
17 If, for example, a sys tem is broken, this pume can cet water i
i 13 from a large tank, condensate storage tank, from --
l t
19 :
Mr. Weaver.
F.eally this diagram is incorrect.
It shows the 1
20 ; feedwater going into --
i i
r 21 !
Mr. Eisenhut.
-- the same line.
i i
22 Mr. Weaver.
It is not; it's a totally diff erent sys tem.
I f
23 !
Mr. Eisenhut.
Actually strays on the top o f the tube.
('"s) s 24 Mr. Weaver.
F.cw is tnis water eccled?
s
- w ecers neoar en. tec.,
25 Mr. Mye rs.
Oces it go out?
4 h
I e
i
53 i
(])
g I
Mr. Eisenhut.
No.
Tne relie f valve 'is only a relief valve 2
in the primary sys tem. :n the secondary system,. for example, a
3-the press ure would build uo.
l 4
Mr. Mye rs.
This water is being pumped in through the 5
secondary --
1 6:
Mr. Eisenhut.
NO, it goes to a spray header.
Auxiliary I
7:
feedwater goes into the spray header, top o f the s team j
i i
9 generator.
9 Mr. Mat tson.
This isn' t drawn correct.
i 10 Mr. Eis e nh'u t.
This is a very simplified diagram, t
11 i Mr. Myers.
Where does the water go?
As it ccmes in by wayI l
I 12 ~
o f the emergency f eedwater cume, it goes through the secondarv,.
s i
g 13 out of the secondary?
14 Mr. Eisenhut.
Yes, it goes out of the seconcary system 15 and can go around.
However, it doesn ' t have to rely on water i
16 continuing the loop.
i 17 Mr. Weaver.
It ~ j us t draws continuous water, can go out the ;
18 pressure valve.
All right.
19 Mr. Myers.
Does it circulate in the system?
I 20 Mr. Eis enhut.
What if there is a creak in the feedwater i
21 system?
f 1
8 22 Mr. S tello.
The aux liary feedwater?
Mr. Eisenhut.
No, a breax in the main feedwater system.
g 23 l m
~J 24,
Where would the auxiliary feedwater f eeding the. steam generator orwersi a.oert n. ine.,
25,
always go to?
i l
i 24 l
i i
i 1
I m) 1!
Mr. S te llo.
The normal valve line-up for the auxiliary l
I 2l feedwater is to feed in its own -- it has its own secarate i,
.i i
I 3'
no::le.
i t
1 Mr. Eisenhut.
No, no, after that.
I l
5; Mr. Stello.
At the highest paru of the s team generator.
4 I
6 !
Then it comes out.
It's normally designed -- when you are usingl l'
7 auxiliary feed, you would expect tnat your main condensor is 81 unavailable, and you wouid be c.oinc. out v. our atmospheric dumps.
i 9i Those are shown --
l 10 !
Mr. Eis enhut.
Suppose there is a break in the line on' the 11 '1 way out?
There is another back-up way t
12 Mr. Stello.
The valve line-ups are two incec. endent caths, Oj i
13 ;
one o n e ach s team ge ne rato r.
And each s team generator -- it i
i 1.1 i can either be suc..clied throueh the main feedwater ring or 15 l auxiliary feedwater ring.
Each s team generator has two flow I
i 16 ;
caths.
Eacn flow c. ath is indecendent o f the o ther.
The valve I
17 l line-ups are such that you can get any combination of those you l
13 ' desire.
19,
Mr. Eise nhut.
Even if the coint is -- vou reallv. don't i
I 20 need this other ma:.n system.
Even "a-e wo uld be, for example, I
i i
21 i a b re ak in the line, then in fact these sys tems c: pump:.ng water l 1
I i
22 l and circulating it through the steam generator are -- essen-I t
i p,
- n.. i tially vou can 'us t keec cum =:.nc water.
s
(/
I 24 Mr. Weaver.
And the water is actually spewing out the 4 9v:n no:ccen. we 23 Ourbine telding ficar or scre hing like that -- well outs:.de I
k 6
1
l 55 l
1 the containment.
]
t 2
Mr. Eisenhut.
If the lines uc. s tream o f the.
l 3!
Mr. Steilo.
You could isolate the main steam line so that
+
4 if there were a Dreak downs tream of it, the break would be t
5!,
isolated.
Then you could adjust the pressure in the steam i
6!
generator by throttling tne relief valve which is shown 7l; senematically on the steam line anc drop tne steam generator i
i 1
3; pressure to any level you desire.
I a
1 9l Mr. Carr.
I am j us t wondering.
I am going to really try to, i
i i
10 ;
understand this.
I think the purpose of this task force is to j
Il I cerhaos r. eel of f a layer of simplicity --
12 Mr. Eisenhut.
Yes.
(^
i 13 !
Mr. Carr. -- and get to a more fundamental and basic and I
la !,
in-death deta: led understanding of how this P ant operates.
l l
15 '
I am wondering, without c.oing ric.ht back to the blueorints,
i I
16 which I ' m s ure no ne o f us wo uld unders tand -- at leas t I i
i 17 wouldn't -- can we pael off a layer of simplicity and get our-18 selves a diagram?
l i
19 Mr. Eisenhut.
Yes.
We have a set which are sort of one i
i
\\
20 l step away frcm his ultrasimple sketch which gces to the next 21, level.
I i
22 '
Mr. Carr.
I would even cc dcwn a courle of levels.
I would.
i 23 '; like tc see one where all t".e valves are, and we can ask I
g
( -
i
- 24 l ques tions about what this valve d es and --
- .4.c.m a.:en rt inc.
25 Mr. Eisenhut.
I think you can get tnere in two s tages.
If I
0 4
56 i
j
\\
i i
(~'y I
i l' you try to go to tne big one with 500 valves on it, you lose
-(
s 2:
ce rs cec tive.
i 1
3l Mr. Carr.
You're right.
I, f or one, wo ulc liX e us to i
4l peel off a layer of simplicity here.
l 5!
Mr. Stello.
I think the one we nave here is about where i
I 6-you would want to wind up.
You might need to go a little bit 7l behind it to unders tand hcw it works, but this is good enough.
t.
Mr. Carr.
Except that, you know, looking at it, I would 3l 9L have assumed, I guess, that the emergency feedwater pumps 10 feed right into that same line.
l 11 li Mr. Weaver.
Dr. Eisenhut, this is the chart you actually 1
I 12 used to brief the NRC, the Commission itself, at the early
(_
I I
i 13 ; stages?
i 14 Mr. Eisenhut.
Yes.
I didn' t rememcer exactly where it was 15 ;
but I certainly recalled the sketch.
It was a sketch meant to l
I 16 ! he extremely simple so we could, in fact, basically snew what 17 really was going on.
It was also used, I believe --
18 Mr. Carr.
I think to some extent you can simplify for 19 people who are familiar with these systems and will read the l
20 1 data absolutely 'dif ferently than we will.
But I con' t knew r
I 4
f 21 '
anything about these plants, and I' m learning.
So I can ' t nave,
I any shortcuts here.
Otherwise my unders tanding will be all t
('T 23 ;
screwed up.
1 i
i x/
24 lir. We ave r.
It's going to take several times over and over a swerei a.oenen. ice.
25 - be f o re we e ve n --
i i
57 l
i
{
}
i i
I Cw 1!
Mr. Carr.
That's what my motaer said about me.
I 2i Mr. Weaver.
Would you continue, unless there are further I
i i
3 questions.
We have now reached the point 12 seconds in.
4' Mr. Myers.
Let me ask one quick one.
Do I unde rs tand, 5
to folicw up, if you were to do what Mr. Carr was indicating, l
l 6
there would ce two lines all the way around this thing?
Wnere 7
there is only one shown, there is actually two, o r two j us t in 3
this one area?
I 9!
Mr. Eisenhut.
Tao.
One turbine and one condenser, but I
i 10 i they go through multiple systems so they feed both s' am genera :
I 11 I tors.
t 12 Mr. Weaver.
Well, no, I think what Dan was talking
]
l 13 l about is your emergency feedwater.
That is a separate system t
14 frc the sys tem that feeds the turbine.
15,
Mr. Stello.
In other steam generators, see, the lines are I
! not snown on the --
16 17 Mr. Eisenhut.
The right-hand side o f the page.
I I
la Mr. Weaver.
Rign t.
But he was not talking about that.
i i
i e
We knew that on the other steam generator there is also a i
19 I
20 turbine and -- unere isn't?
l I.
21 !
Mr. S te llo.
No.
The same condenser.
But the emergency i
22 :
feecwater sys tem, there are two completely independen ficw t
I I
pN 23 j paths.
l t
24 '
Mr. Weaver.
Righ t.
s>s,e.rs m.oorms, n c.
25,
Mr. Stello.
One for eacn s team gene rator t".at are capable l
I t
I 6
4 53 i
i O
V I:
c f cperating as totally indepencent --
i i
8 2'
Mr. Weaver.
Ri nn.
They do not go enrough the turmine.
l 1
3l Mr. Eisenhut.
That's true.
Yes.
They do not go through 4
the turbine.
I Sl Mr. S te llo.
They are isolated.
l l
i 6
Mr. Weaver.
They just get that heat out.
That is their i
1 7;
toint.
t t
8!
Mr. Stello.
Yes.
l i
I 9-Mr. Myers.
May I continue for a second?
On tnis chart, if l
i 1
10 !
the second line comes un like this, where does it end uo?
I i
l 6
t II Mr. Eisenhut.
It goes um into a scrav. header at the too. o f,!
i 1
12 the s team cenerator.
I 13 Mr. S tello.
Higher than the steam line pnysically.
It's I4 acove it -- above it physica'lly, i
15 ;
Mr. Myers.
Yes.
l l
16 '
Mr. Stello.
It's located pnysically in the same generator 17 at an elevation higher than the steam line.
i I
I i
i 18 '
Mr. Weaver.
What haccens to the water?
I 19 Mr. S tello.
The water is sprayed inside the tube bundle 20 ;
when it's used as an auxiliary feedwater, changed into steam 4
21 '
by the energy of the primary coolant inside of the tube.
1 22,
Mr. Weaver.
Aga:.n, it's got to go comeplace.
I I,
r p
23,
Mr. Stello.
Then it ccmes out the s team line.
Ncw, if
.s a.t the condenser is availanle, like it is rignt new, it could be c.J.c.,.i a.oe,mi, inc.
i
.c discharged into the condenser.
If - t:'.e condenser is no t 4-t 4
l l
59 I
1 I
i I
(])
1l available and they must design it assuming it is not available, i
2l then the water can ce discharged out througn what are snown 3'
here as relief valves.
They are called atmospheric dumps.
I i
i j
4 Mr. Weaver.
Ri gh t.
Wnere is that water going?
Oh, it's S,
outside as steam.
I 6'
Mr. Stello.
As s team, i
7l Mr.'Eisennut.
You go througn the by-pass valve label there '
I sj and bypass tne turbine if you have the condenser.
If you don' t j There are reveral contingencies built in, l I
9' you go out as steam.
I l'
10 that it can go here, can go there,can go there.
Tnat 's wny 11 !
it's a little conf using.
I i
12,
(P a us e. )
s 13 Mr. Mattson.
It might pay at an early opportunity to have 14 one of our systems people come down here with the diagrams and i
15 { give you a half-hour or an hour of basic, now the sys tem worxs, 16 !
and what is where.
i, i
17l Mr. Weaver.
Ri gh t.
We are finding out what we don ' t know, f i
i la '
Mr. Mattson.
It takes all of us a bit of time --
l 19 Mr. Weaver.
Well, would you proceed?
We are 12 seconds 20 into the transient.
We have covered tne first 12 s eco nds.
i' I
i i
21 ' New wnat nappens, tne 13th second?
i i
22 '
Mr. Eisennut.
Se: ween 12 and 15 seconcs, as we centioned
.1 I
23 a minute ago, the plant should have pressure in the primary
~
i m.
24 syste=, the rignt side, should have s tarted to drop, shoul: have a verv neoo,ws. see.
25 dropped past the-peint where the relief valve, tne valve that I
L e
1 1
(~}
1 opened on the tcp of the pressurizer, should have been closed.
~
s-I l
2' Mr. Weaver.
This is the valve on top of the pressurizer, 3
the top of the right-hand page.
4i Mr. Eisenhut.
Yes.
New, there are various indications --
i 5
it's a little confusing _ as to what the operator really saw.
6 Sone records say they are indications the valve had closed.
He,
may have thought it had closed.
That is an area that is a 7l 8!
little confusing.
t I
Mr. Mattson.
That is probably what they heard about at the +
9li t
10 site tnat might be worth explaining.
l t
i 11 ;
Mr. Weaver.
Right.
I 12 l Mr. Mattson.
There are several kinds of indications.
One ha 13 is a simple light, that the solenoid that is driving the valve i
s 14 has moved.
That light says the valve should have closed.
15 It's not a positive indication, however.
It only says the 16 l solenoid has done its job.
Evidently the solenoid did its job i
17 and tne valve still didn' t close, i
18 Mr. Weaver.
Excuse me.
What is a solenoid?
i 19 Mr. Stello.
The device that is used to make the valve open '
i i
20 -
and close.
There is an air supply that is actuated when the l
21 solenoid opens up.
When that opens up, it allows the mechanism 22 !
for the valve to cause the larger valve to come ocen.
i 23i Mr. Weaver.
So we knew that that worked.
We don' t knew the
/~N i
%-)
(
24 valve actually closed?
I orecew Foxmn, sec.,
25 !
Mr. Eisenhut.
We are not sure exactly hcw f ar it went.
We I
i I
I f
1l do.<new there is conflicting information.
There is some indi-I 1
2ll catien that says it may nave told the operator the valve nad I
l 31, closed.
In any event, we new know the reactor coolant sys tem i
t 4'
apcears to be pressuri::ed for either continuousiv. for the next l
S two hours or certainly part of the next couple hours.
That is, '
I 6
it stuck open because a couc. le of hours later they closed the i
7l; l
back-up valve and even tne exact time when they closed, tha t 8
is a little bit --
j l
i 9l Mc. Weaver.
You are saying that this valve remained open
+
i 10 {
for several hours?
l t
11 Mr. Eisenhut.
This valve cces into a ci.te and that c. i.te I
i l
12 '
simply has another valve in it.
Two ho urs, rouchiv, into the 1'
13 transient'he closed that second valve.
s 14 Mr.
Burnam.
Could there have been an instrument that said !
15 the valve was closed?
Would that have been technically 1
16 possirle?
I 17 j Kr. Eisenhut.
Yes, u rtain valves --
l e
18 Mr. Burnam.
This one did not?
l 19 Mr. Stello.
I.t did no t.
The indication it had was not 20 f positive.
i, i
21 '
Mr. Mattson.
There was another indication, however.
The-22 > other indication you will see in here is the temperature in the,
I t
23 discharge pipe of the pilot-operated relief valve.
There is a O
N.
24 i th e racco uple, tenperature-neasuring device in that discharge o s.ceri=,oen n. =
t 25 ; pipe that does --
I t
I l
e 62 I
{)
1l Mr. Carr.
Is it possible. to get, say, the manufacturer
)
I i
2 of this valve er one similar to it -- would they nave a cut-i 3
away of the valve assenbly and solenoid?
I l
4' Mr. Mattsen.
I believe we nave it.
5 Mr. Stello.
Yes.
1 i
6, Mr. Carr.
At least a fairly goed cutaway drawing of it, I 1
7 think, would be helpful to me.
3, Mr. Cheney.
On this point, when the pressurizer valve
!9! cpe ns, there is cn indicatio:. on the board that it has been j
i i
10 closed af terwards.
It discharges into a pipe?
11 l Mr. Eisenhut.
Yes f s, 12 l Mr. Cheney.
Tuare are two separate places, th e n, two ways
(/
.i 4
13 in which the operator could have been misled.
One, the 14 solenoid triggered the lign: that showed it was closed when it 8
15 was n' t and, secondly, temperature reading of the discharge I
i 16 4 pipe?
i 17 Mr. Mattson.
Yes.
There are actually more.
If you want t
la to go into more detail, we will tell you about more indications..
l I
19 Mr. Eisenhut.
We'll get there.
That is a few seconds
,l 1
20 i aWay.
{
l i
21
.Mr. Cheney.
Ycu are going to cover that, though.
l This line -- the pilot-operated relief valve !
22 Mr. Eisenhut.
23 '
gces into a tank on the pressuri:er.
That is called sc etimes I
()
~
24 j a c uench tank.
It's got various differen names.
e,.s,aere n ewte,s. u~e.
23 :
Mr. Weaver.
Is that inside the container?
4
+
i t
6
1 6
1 li Mr. Eisenhut.
Yes.
I i
i 2:
Mr. Weaver.
All right.
I J
3i.
Mr. Eisenhut.
That tank has on it level indication; it has 4 i on it temperature indication.
So there are a couple of means l
l l
I 5t to say first -- an operator if ne had cone to enose could have i
i 6'
said the water is hot, indicating that the valve is open.
So
}
7' there are ways he would have known.
Pilot-operated relief valve 8l was stuck open, running water out the pipe into a tank.
9 Mr. Cheney.
Are you confident that those numoers were i
10 :
accurate?
Or do we know yet?
11 ;
Mr. Eisenhut.
I j us t don' t know.
I 12 '
Mr. Weaver.
Henry has a ques tion.
13 Mr. Myers.
The ins truments for the quench tank they said s
14 f the other day were behind the main console over in the corner.
I 15 !
Mr. Eisenhut.
I'm not sure where they ar.e physically.
i 16 Mr. Myers.
I got the impression the other day daat one of 17 the problems was that the instruments for that were all way over; 13 in the corner of the room behind the console and maybe they 19 we ren' t watching.
l
?
20 Mr. Weaver.
I just want to say that is an extremely impor i
4 21 tant point, because one of the things we wanted to find out was !
I 22 l how well they were in control, the possib4 ' 4 *i as o d control
-)
23 he re.
So you are raising the issue, Henry, I think, tha t s_/
1 24 because these gauges or whatever were of f at one side they Osnere n eoorws. va.
25 : didn ' t see them.
Is that a possibliity?
i
l 0%
V 1
Mr. Eis enh ut.
It's ce rtain a pcssibil:.ty.
In fact, dat l.
2l is why I said that is an area -- the reason I nighlighted it, I
3i it is a ques tion area as to wnat nappened to the pilot-operated n
i 4
relief valve, what kind of indications they had that the ' '.l ve 5
was still open.
i 6l Mr. We aver.
But they could have seen them?
i 7i Mr. Eis enhut.
There were at least a couple of oder in-I l,
a!
struments they could have used if those instruments were working.
i l
9i I don' t know whether tney were.
l l
l i
i 10 i Mr. Weaver.
The indications are that they didn't see them i i
11 for several hours; is that ccrrect?
Q 12 Mr. Eis enh ut.
At least a couple o f hours.
13 Mr. Ma tts o n.
Wait a second.
We 've got some of dat infor-l 14 mation in front of you.
You've got the chronology.
If you I
i 15 '
icok on page 8 -- there are several in here.
This is one I 16 just happened to find. In Me tropolitan Edison's own words,
17 at 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> 17 minutes 5 3 seconds de operator requested the 18 ;
outlet temperature reading in the tail pipe, in the discharge i
1 19 pipe; it's 228.7 degrees F.
l i
f 20 ;
Mr. Weaver.
What does that tell us?
l I
i 21 ;
Mr. Ma ttso n.
Well, it should have said that there is s team '
i tt i
22 in the tail pipe, which we now knew, o f course, there was.
I 6
p 23 :
Mr. Weaver.
Which would tell us the valve was open?
v 24 '
Mr. Matesen.
It should have told him the valve was open.
s>secem =ecormi, inc.
e l
25 Mr. We aver.
That was t.vo hours?
i i
t i
I i
i
J 61 i
l I!
Mr. Mattson.
If you go back to page 7, vou will see he l'
I 2
asked at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> 20 minutes also, and he received word that the i
3!
reading was 293 degrees F.
l 4
Mr. Weaver.
So at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> 20 minutes he read it.
i I
5l Mr. Mattson.
He also asked at 24 minutes 58 seconds --
6{
page 6 -- and got the information it was 285.4 degrees 7 1 Fahrenhei t.
Again, recall I am referring to the Metropolitan I
i 81 Edison chronology.
Most of this information comes frcm com-I 9
puter records.
i 10 '
Our I&E people are doing their own investigation, which in-11 cludes these computer records, plus interviews, plus strip 4
12 -
charts, and all the other information they can put their hands I
13 :
on.
We can't swear and declare that these are accurate until 14 we see that frcm our I&E people, but this is what Metrcpolitan -
15 Edison says is true.
16 Mr. Weaver.
Eo we have the nearest time from point zero to l 17 when somebody read a gauge that said there was steam in the i
f la guench tank?
19 Mr. Mattson.
Well, you ought to get on to the other i
I 20 j indicators.
If you look at 3 minutes 26 seconds at page 4, i
l i
I i
21 :
the reactor coolant drain tank, which is the discharge tank for.
1 i
i I
72,
the pressuri:er, high temperature alarm went off at 127 degrees.:
I i
i 23 '
Mr. Eis enhut.
I think the real answer to vcur cuestion, A)
(.
i 24 -
thou-h, we won ' t really have as to how soon there was an i
ces,wes neconm. v.c. '
1 25 '
indicarion.
}
a e
66 i
i I
p d
1 Mr. Ven to.
Is it common that this relief valve would vent i
i 2j pressure?
Is that a common occurrence, something that happens l
t 3l daily at the plant?
l i
t 4l Mr. Mattson.
It's happened 150 time s, ro ughly, in the 38 4
5 reactor years o f B&W plants, in terms of the 38 reactor years ;
1 6!
for all B&W olants.
I don' t knew the number for Metropolitan l
7',
Edison.
3:
Mr. Weaver.
It averages four a year.
I 9,
Mr. Vento.
So it's a f airly unus ual occurrence, is it no t, i
i 10 '
this particular phenomenon?
11 :
Mr. Mattson.
It would han..cen for events like loss of i
C 12 feedwater wnich happen about twice per year.
I 13 :
Mr. Ven to.
So it's not a regular thing that would be I,
i 8
14 happening.
Cne might reasonably expect that they ought to i
l 15 '
look for some additional information af ter that relief valve I
i 16 ope ns, a fairly unusual event, pe rhaps the first time it happened i
i 17 { at this plant, that they would look at --
l I
t Mr. Ma ttso n.
I t 's no t the firs t time at this plant.
la I I
19 i Mr. Vento.
I t 's no t the firs t time at this plant?
l 1
20,
.Mr. Mattson.
No.
4 l
21 ;
M.r. Weaver.
Cn page 4, 3 minutes 26 seconds from point 22 ;
- ero, an alarm was received.
What is this alarm?
Is it an p
23 audio ?
J 24 '
Mr. Ma ttso n.
I do n ' t knew.
It may be that at 3 : 26 -hat co sec2rv Geoo, ers. tne.
25 is an expected alarm for a normal opening.
That is very early l
a i
I k""%_)
1 into tne transient.
It's later on at 24 minutes 53 seconds
+
i i
2l the PORV normally would have been open1d by them.
He knew I
i 3.
he had auxiliarv f eedwater, schew he would interpret that alarm i
i i
I 4 i I don't knew.
I i
5 Mr. Weaver.
He should have done something here, according i
i 6 l to hindsight, at 24 minutes 58 seconds ; is that correct?
An 1
l 7l action was omitted, was it not?
3f, Mr. Mattso n,
Maybe; maybe not.
I recall reading somewhere I
9i from some conversations with the operator that they were used i
10 l to seeing the temperature indication in the tail pipe lag the i
11 I closing of the PORV. ' So he may have read the indication at 24 *
()
12 minutes 58 seconds to be a lag in the decrease in temperature of i
13 i the tail pice which would go up any time the PORV opens.
But j
14 li then he asks again at 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> and 20 minutes, and the temperature' 15 is s till high, and he still didn' t do any thing about it.
i 1
1 1
16 !
Mr. Weaver.
Going back to the 24 minutes, does he have any,
l 4
17 I other ability to knew s team is going into the quench tank?
l I
13 '
Mr. Stello.
Le t me try to c.et scmething clarified.
~
j i
i 19 ;
think we will want to focus very heavily on this.
The reason 4
20 ;
the operator continued to go back and query the tail pipe i
i I
i l
21 ;
reading is, was it readin g too low?
If the relief valve were i
I 4
22 '
open, should he he getting a higher reading?
The location of I
m
("T 23 '
the thermecouple and behavior c f that particular dr.ermoccuple
%J 24 -
in tte ail pipe is a particular area tha: needs some real a.sm:,a = ex,tm. t-c. '
25 '
hard lecking.
I don' t believe we want to conclude that when
68
,r t
i
(
1l he lecked at de tail pipe he should have kncwn it was open.
2 If they were discharging steam into the pipe, tne s team ccm:.ng i
3l out of the pressuri::er is in the neighbcrhood of 600 degrees l
i 4lf F.
The thermocouple is reading 28 3.
It would be a reasonable I
5' conclusica to say the valve is closed rather than open.
f 6l Why was the thermocouple reading what it was reading?
I 7
Should it have been reading something different?
What was the '
1
?,
pnysical location of it?
l i
9 This is an area we are starting to lcok at very, very i
=
10 carefully.
I don't believe we are prepared now to try to say I
11 I the operator at 24 minutes should have kncwn by that the rmo-i 12 couple reading that the valve was open.
That thermccouple
&Q 13 '
reading perhaps may have been telling him it was closed.
i 14 I think af ter the interviews that our investigators are g
i 15,
going through right new -- and now I mus t tell you I am specu-1, 16 l lating; I don' t know this for a fact, conjecture -- perhaps i
i 17 !
what they may get from the operators f rom that interview is la '
that when they saw that thermocouple reading, what they were t
i, 19 !
celieving is that the valve was closed because the thermo-i 20,
couple was reading too icw a value for a condition which would i
21 have represented the valve be open.
i 22 l So again, conjecture; I don' t knew.
We are prepared to sit l I
I m
23 : here and go over this to the de s t o f o ur abil:.ty, b ut we are (J
1 24 really going to need cur investigators ' report which is really
- e. m ess necce w s. u.a.
25 dealing with this very, very questien in building up this i
i e
i
1 69 i
l l
l
(})
1i f actual data base.
i-
!2j What I am concerned about in the way we are trying to I
t 3i respond today is will we leave you with a confused impression
\\
i 4!
of the record because of the inability for our own recollections-i S
and s tudy to have ha d the be ne fi t o f the re po rt - tha t the in-I 6;
ves:1gators hope fully will be coming oun with next week.
I wany l
i t
7:
to make sure the record clearly sav.s that what we are tellinc I
9l you may, in fact, turn out to be somewhat incorrect, and later 9{
facts will just have to modify it, i
t 10 Mr. Weaver.
A caution much. appreciated.
11 I Mr. Carr.
To j ump to a conclusory kind o f s tatement, we i
I 12 ;
were told the o ther dav at the plant ds.at there was no direct C-)s 1
13 level indicator in the reactor at the level of the coolant, tha t 14 that had to be inferred from 1 coking at other data, s ay the 15 ;
p ress urize r, pres s ure.
Now we are saying that there is no i
j i
16 ;
direct indication o f this valve at the toc o f the press urizer, I
a 17 that we only_ got an indication, a direct indication on its i
I la,
mechanism rather than on its functioning, and that there were I
l.
19 ! some back-up inferences vou could draw from other instruments i,
20 l which would lead someone to assemble sore data.
I t
i And I s upcos e j u= ping way, w ay ahe ad o f o urs elve s h ere, I ami 21 !
22 wo nd.e ring, is ttis a special feature of Babcock & Wilcox plants l
1 1
e i
23 that en their gauging and instrumention they don' t use direct i
fw/
n s.
24 '
data, and then use the other stuff for cackue?
Or is this eree... = e:c,m i. :-c.
i 25 ' scmething throughout the nuclear industry, where there -is a lot ;
I
[
+
7 l
)
1 of data points and it leaves it to the cperator to sort of I
2 assemble all of this, what is haoneninc inside that building, 3
in his head.
I 4
Mr. Stello.
Valve positions normally are indicated by a j
i Si mechanism which looks at the travel of the steam, which is to i
6 some degree an indirect reference to what the valve is actually !
7, doing, but a more direct method than you would get frcm looking
- S at the solenoid which starts the e ven t.
i 9
Whether this is typical of the valve arrangement for valves ;
I i
10 -
as a general matter, it is not.
It is atypical.
Valves 1
i 11 i normally are measured by position indication, as to what appears f'
12 '
open.
This is a very, very special kind of valve, the electro I
t 13 matics that are on the pressurizer.
Fo r tha t cla s s o f valve s,
l i
14 I don' t knew if the indirect way is used by all.
Butitclearlyl i
t 15 <
was used in this situation.
i I
i 16 If you go back and look at all of the information you have, since we are now charged with the ability to do this with hind- {
17 -
18 sight, I think that there was sufficient information to knew l
19 that he had a relief valve open.
Wi th the b ene fi t o f tha t hind '
I i
20.
s igh t, by looking at all of the things that there are to 1cok i
21 l at, I think there was a pretty clear indication.
I think there t
22 '
were too many things going on for the operator, and he was l
t T
23 !
getting misleading information.
And, c f co urse, we are locking
(/
\\_
I.
i 24 ! very hard at that, to o.
But I don ' t want to try to suggest i
es=.ca m emrmi, ve.
25 that he should have had the same knowledge at the time of the i
I
l 71 :
I f
1, accident that we have today.
Clearly he didn' t.
I j
2 Mr. Weaver.
Mr. Stello, how many alarrs were going of f?
l i
i 3'
How many buttons were flashing and audio alarms coinc off at this I
e 1
4 i time?
l l
5 Mr. Stello.
I don' t knew the answer to that.
l 6
Mr. We ave r.
Can you giw us an idea?
I was told by the i
7 plant manager what his guess was, but I would like it.
l r
a Mr. S tello.
I think it would be large in nu-ler.
f i
I 9
Mr. Weaver.
Ecw many?
Did you know hcw many audio alans 10 there are?
l 11 '
Mr. S tello.
I would perhaps guess on the order of i
i i
l 12 hundreds to a thousand.
i 13 Mr. Eisenhut.
I would guess somewhat slightly less than l
l 14 th a t, only : rem go:.ng to the simulated accidents.
I f you go l
15 through the simulated program, it's clearly on the order of 4
i.
16 maybe a hundred.
17 Mr. Weaver.
Yes, this is the --
?
18 Mr. Carr.
Where is that simulator?
19 Mr. Eisenhut.
B&W's is at Lynchburg.
I went through a r
20 l SWR course on one.
The types o f ala=s you get and nurbers you 21 get, I imagine, don' t va:y greatly, but each of the major i
22 vendors has a computeri::ed --
23,
Mr. Che ney.
Is this the kind cf facility where you go
)
j 24,
get a feel of it?
\\
los.ceni aeoonen. Inc. ;
25 l Mr. Eise nhut.
Yes.
I 6
72 i
i i
l j
'1 Mr. Carr.
I s uggest, Mr. ' Chai rman, some time we 70 up 2'
the re.
3 Mr. Stello.
That would be a good idea.
i 4
Mr. Eiser. hut.
I think it's an excellent idea.
They can,
,i 5
from the mas ter. control, simulate a LOCA.
You can really see l.
6!
what happens.
Or they can simulate any kind of a transient 1
i 7!
situation and you can see the types o f alarms.
l I
3!
Mr. Mattson.
The TMI-2 event had been prograr:aed on a l
9i simulato r at Lynchburc.
All the B&W ocerators in the United
+
i a
10 l States are being trained today in an accelerated way on the s
I t
11 '
TMI e ven t.
1 Q
12 Mr. Wea ve r.
Lynchburg, Virginia?
13 '
Mr. Matts o n.
Yes.
Your best idea is to go down and watch i
14 one of those cperations with some of the operators from another '
i 15 i plant being put through their paces.
I 16 Mr. We ave r.
The plant manager told me he thought a hundred i
17' audio alarrs were going o ff at one time.
i la Mr. S te llo.
Is that what he told you?
l 19 Mr. Weaver.
- Yes, i
l 20 t Mr. Stello.
I wouldn' t be s urpris ed.
1 1
21 'l Mr. Mattson.
"'here would be alarrs over a period o' ** -a i
l 22 : associated with the primary and secondary sys tem, over tne --
I t
d p
23 ;
Mr. Weaver.
All occurring in the control room?
I d 24 Mr. Mattson.
Yes.
e.7.wu necorms. inc. '
25 Mr. Carr.
And there is no instrumentatio-
' =' ' elps the L
t f -
~
73 t
1 i
1 l
~
i
]
1:
operator sort out the o rde r o-f i=o rtance o f the s e alarms, o r i
2j the order of importance of gauges or lights or switches; right?
I' 3
Mr. Mattson.
Only the training.
i f
i 4
Mr. Carr.
Only the training to assemble all o f these i
I 5'
instruments and react in split seconds,
i I
6' You would agree, in hindsight and orderly procedures 1
I' I
7*
planning, you would order your sets of gauges; and other I
l i
Si, gauc.es, while important, may be clutter, may te noise in the s
9 sys ten.
I i
10 '
Mr. S tello.
Let me give you my experience, while I l
I i
i i
11 :
folicwed the course of the accident.
And I won' t cretend that !
i i
l p
12 ;
I knew B&W designs well enough.
But as trie d to understand
\\
\\
t 13 !
the accident, I reduced the nummer of program meters tha t I i
14 '
really wanted to know about to very few to try to make the l
15 j udcment of what was going on.
I think the ope ratcrs can do i
l i
16 ;
th a t,
i 6'
i 17 For a LOCA, for example, you know you are going to get I;
18 decreasing pressure and probably some level swell, so you might.
19 !
ge t i nc re as ing le vel i n -he p res s uri ::e r.
But af te r the l
20 transient is going on, af ter you know you nave an intact 21 sv. stem, which he knew he had, at some c.oint, you start to know 22,
that v.ou are interes ted in the c. ressure o f the c. rimarv. sys tem, 1
t 23 the he t and cold leg temperatures, and any incore temperature O
^
v 24 respense that you have.
25 You can reduce the nu-2:er -hat yo u rea'.ly nea'
" cus on
+
l 1
i i
74 t
l l
I i
to trv to understand the event that you haw had to
-al = " vely 1
i I
2l few.
For example, if you had a major LOCA, he would knew 3
very quickly what program meters.
That would then tell him 4I what he ought to focus his attention on.
He wo uld haw ve r/
i 5
quickly a response in the reactor building that says I have a 6'
di s cha rge o f a lo t o f high-e ne rgy flui d.
I either have a break i l
7i in my secondary system in the reactor building or the primary 3:'
system.
I imme diate lv. 1cok a t mv c. rimarv. sys tem mete rs.
Then.
.6 l
9i I can tell if that was a primar.i system rupture.
If it's a i
10 !
prir. arf syste:a rupture, I then know what program meters to II l track.
i f
12 l So there is a hierarchy of logic built into the very many i
i 13 i alarrs that you see that point him at the right direction.
14 He had some thing which, again, in the end result, with the I
15 ! bene fit o f hindsight, wasn' t some thin 9 you would ha te excected.
t I
I' 16 That is, an intake primarf system but a very, verv larce 17!
fission c. roduct inventerv..
You norma 11v.. exc. ect larc.e fission
.I 13 product releases to be associated with a very large less-of-I 19 coolant accident where you would have had a substantial i
20.
fraction o f fuel rods f ail, not damaged, f ail initially and -
21 :
qui cklv.
So he had something c.oi=. on tha t his trainine. didn't i
i 22 '
really direct him in that e ffort.
l I
I 23 !
But this whole family of alarms and gauges da; you see, d'
24 '
cere is a hierarchy of logic buil into it that, as you are I
- w s.o.,c aeoo,m s. ire.
25 '
trained to be an operator, you know what cri tical p rogram me te rs i
e I
I 1
to lock for very c'uickly and eliminats or focus your attent'on in the right direction.
3, 3
Mr. Carr.
In the location of the alar.r.s and gauges and I
a buzzes and switches and all that s tuf f, doer the location on 5'
the panel follow that logic?
i 4
Mr. S tello.
Yes.
i I
7 Mr. Mattson.
I would like to add something to what he said.:
I i
S, Mr. We ave r.
Surelv.
i i
j 3
6 9'
Mr. Mattsen.
The disabling f eature of this accident, in my i
10 l mind, was the fact that it was an unmitigated small-break less-I 11 l or coolant accident.
Now, that plant was reviewed, designed, I
(
12 i and analyzed for a small-break LOCA.
Un fo rtuna tely, the j
i
(
13 operator took action, evidently, on the basis of the informationl i
ja,
that is here, to either defeat or degrade the emergency core j
I ic cooling svstem that was desicned for just this kind of accident.
l i
p-l And he was trained to knew that no ocerator actions are recuired
=
i i
t I
17l with the ECCS for a small-break LOCA.
So what evidently j
i i
nappened was that his training was insufficient, or the i
.a }
19 l transient that caused the small-break LOCA, tha t is, the
+
1 feedwater ::a:isient with the s tuck auxiliary feedwater valves, 20 21 i so confused tne situation that he was unable to recognize he 22 had a small-break LCCA, and hence was able to intervene in the -
i operation of the autcmatic system.
n J
t
-.i Those systems are made au cmatic.
Thev are made operator-i om,ei.=eoe,ms, i, c.
23 free for jus t the re ason you are conc =--ad
=" cut.
That is,-all y -
1-M P
i
()
1l cf these indicators, all o f these bells going on.
So in the I
2' licensing process we normally assume no operator action to i.
3:
mi tigate the accident for pe riods o f minutes, 10 minutes to a 1
I 4
hal f-ho ur, typically.
5 What we haven' t looked at is the likelihood e.at the i
6!
operator wo uld intervene, thinking he had something dif ferent 7l than what he actually had.
It appears, in =y j udgment, at s
+
8';
least, to be what happened in this ins tance.
He had a system i.
9l fully capable of handling the break that occurred, but for a i
t 10 l1 varie ty o f reasons, all of which need to be explored, including '
i i
11 ;
all o f thes e ligh ts going on, that sys tem was not permitted to
({}
12 do i ts j ob,
t' 13 Mr..s@ers.
A question along that line, if I may.
All o f l
14 a sudden, everything was normal.
Then you sav 100 or 200 l
15,
lights and bu::ers and whatever else came on, indicating that 1'
i 16 there was an emergency of some kind, j
17 I want to ask just two questions.
Number one :
What was
[
18 the atmosphere and condition of the control room at that point?
i 19 Was it chaos?
Did r. ecole ac. cear to have their heads on?
When I
20 you say one guy made all these decisions, it would seem to me 21 if an emergency came on you might have two or three supervisors, i
22 i everybody else involved.
We wouldn' t be relying on one in-i,
(~)
23 l experience'
'C '-w to make the wrong j udgme-~
"rw do yo u
\\_/
1 24 respond to that?
a ;,w a necem n. ire.,
2' V-Stello.
First, with respect to what was the at=csphere i
i l
1 l
_O in the control rcom, how do you characteri:e the demeanor of V
1' 2,
the people the re, I think as calm, compe ten t, and sure that i
i 3l they were moving in a course which was going to recover the l
l I
f situation.
All the people I ha a had who have followed what's gone on in the. early part of the accident have again reiterated 5i I
6 to me the confidence.
i 7I I don' t me an by that to sugges t tha t, in hindsight again, 3
we won' t find that some o f the acticns that were taken were i
i 9
inco rre ct, that better actions could have and should have been ;
10 taken.
We had a situation which we clearly had analyzed i
i Il be fo re.
There were a whole variety. of things that have happened.
f-kQ 12,
A feedwater transient by itself is a classical transient.
I
\\
13 '
It is analyzed.
A small-break accident is an accident that is a
14 l analyzed.
By itself, it's okay.
But when you start to icok at l
15 these critical events that have happened, you had a feedwater j
i 16 l trans ient; then you had a power-operated relie f valve which I
l 17j failed to operate.
Somewhere later in the transient we are i
la i going to go where he started to adjust the other engineered features put here for that purpose, hich-oressure '
{
19 :
s a fe ty j
20 coolant injection system.
21 As Dr. Mattson said earlier, you have intervened several
-l
?
22 times in the course of this transient and caused this transien:'
I I
v, 's/~ 1 23 to take on a prettv new and dif ferent character.
We have to 24 avoid this f rom happening in de future.
I think some of de e,.s.en a eco,ms, ine.
25 s te s that can be taken to avoid this are clear.
i i
f 4
78
.I I
f f
O 1l Mr. Myers.
Who was ultimatelv responsible fo r the
~
V 2:
de cis ions that were made when this c=erc.encv. was firs t shcwn up?
1 l
3 Mr. Stello.
' he highest-ranking method o f ficial in the l
4' control room --
i l
5' Mr. Mv.ers.
Was he the one that v.ou were referring to that i
t 6'
was inexperienced?
l I,
7' Mr. Ma tts o n.
Did I say inexperienced?
I 8i Mr. Myers.
Yes, I tho ught yo u s aid ine xpe rie nced.
i 1
9, Mr. Mattsen.
I didn' t mean --
10 i Mr. Marriott.
Or he added tha: One train of events were j
i i
f i
11 :
so complicated it was impossible --
(N 12,
Mr. Mattson.
I didn' t re an to i=civ. he was inexterienced.
i s
i 13 '
I don' t believe his training was ins uf ficient for the condition I
l 14 j he got in.
l l
8 15 Mr. Mye rs.
Who was that?
The superviso r, or someone under ;
i l
16 the su=ervisor?
l.
t 17 Mr. Mattson.
I don' t think we can sit here today lacking t.
i the IE investigation and tell you who was in the control rcom..
13 t
19 '
I read the newspaper af ter you folks were up there a couple f
20 :
of days ago that there were two in the control reem at the 21 cnset of the event.
The best knowledge I had before that was I
i f
I 22 I that there were four.
tJntil I see -ne record o f who was there i
I, i'
23:
when, I don' t think we knew the answer.
i l O i
\\J l
24 Mr. Mye rs.
Are you inferring that this event was really a lcCS E9d!*38 $ 9 porter 1. IPC.
t 25 hu-.an e rror?
i 5
1 4
g-1 Mr. Mattson.
What we have said, as an agen cy, i t was a i
2e combination o f design sho rtcomings, equipment malfunctions, and i
3.
h uman e rro r.
I think I can show you specific instances of all 4
o f tho se.
t i
5 Mr. Eisenhut.
I think you recall the six items we iden-i 6
tifie d earlier.
We are not really yet prepared to say hhat i
i i
7 here is how the system would have performed given you remove at, any one of those or any combination of the six items.
It's a
!i9' combination o f all three.
i 1
i 10 ;
Mr. Scoville.
Whe n yo u s ay " h uman e rro r, " do yo u me an i
11 :
error on the part of the operator, dhat he did sociething he 12 wasn' t told to do in his training?
Have you identified things s
where he made mistakes from what he was told in his training?
q 13 l.
la :
Mr. Mattson.
I used the word "humcn" because I am not i
I l
15 '
ready to conclude that he did things he was not trained to do.
t I
r I
16 I t may be tha t he did.
I am not personally ready to do that.
17 So I will sweep operator error, lock closed, auxiliary feed-18 water all together under " human erro r. "
19 Mr. Sco ville.
You used " human error," saying he did things 20 l he wasn' t supposed to do in terms of training.
You are saying 21 1 he did what he was told to do but it's a mistake?
i i
22 Mr. Eis enhut.
That is certainly possible.
It is also t
h
- l i
e i
22; pessible to go the o ther way.
That is, he took actions when his' (a~)
24 '
ins tructions were silent.
n pnes. necorws. nee.
25 Mr. Ve n to.
One of the points. here that is in conrention, f
I l
e l
80 I
i i
l, 1l obviously, is that there is an automatic sys tem that is supposed x
2l-to take over when certain events take place.
Mr. S tello com-l 3'
men te d regarding that.
And that the operator at these ins tances P
G 4
changed the enaracter of the transient by actually turning'it o f f i
1 i
5 automatic and putti.T it, in essence, on manual, if I can l
i r
6' simplify it by saying that.
I 7:
I guess the real question is:
Did he wait the proper l
}
I 8!
amount of time, wh:.cn :.s a guagment that would have to be based t
l' e
9' on it.ct?
Did he wait the proper amount of time before makinc 1
I 10 l changes to bring en what he was instructed to observe, or should i
i 11 :
have been Irving; it was not roving in the direction or didn' t 12 '
rescend on the basis of the automated systems.
p l
d i
I 13 i That is, indee d, a point in contention.
I think that is a -
3 14,
point we are going to have to e:<plore.
The impression that is 15,
le f t here is that he interrup ted these sys tems which would l
16 naturally take ova-
" deed, th at is h uman e rro r.
17 '
We have already pointed out certain malfunctions with 18 l regard to the equipment.
You sugges there are also the same 19 :
questions regarding design, s
i 20 =
We are heavy on the human error here, it seems to me.
The.
s 21 question is:
What he did, was it rerAr3hle based on the type i
22 of training he had, based on the characteristics of this i
i i
~3 par ti c"' = -
-a acto r?
Do we have a reactor design that really
' f)
V 24 caused him to mow in this d'-a on :.n an unreasonaole way?
l ? E M t' 41 44Dorttft. IMC.,
25 1 Yo u k new, tha t is really what it all hinges upon, sho -kncws.
i L
i
' 'T 1
They have suggested to me dr.at a large reactor li'ke this re-
-)
s.
2 sponds very sicwly to changes, whether automated changes or I
3i whether they are done manually.
Is that accurate, especially 4
if some thing is cperating at 9 8 percenu capacity?
I s
5 Mr. Mattson.
One of the difficulties we found in looking 6!
closer at the B&W desien since this accident is that it does I
l 7!
respond rather rapidly to upset conditions in the secondary l
l l.
i 8l sys tem.
If you heard or read the newspaper where we said the l
9:
B &W design was more sensitive, that is what wt me an t.
It has l
i 10 !
less water in the secondary side, which means for delay or i
11 degradation o f the auxiliarv. feedwater deliver./, the.crimary I
l 12 system can see rather large pressure and level swings cop. pared I
(
13 ;
to other PWRs in a relatively short period of time.
14 The B&W machines, tne operators for those machines are 15 l specifically trained for that sensitivity.
But the question I
i l
16 l be comes :
Is that an unwarranted burden to o. lace on the ccerator I
+
i 17 to expect quick demanding things to be acccmplished in an j
i 13 unreasonable time?
i l
19 Mr. We a ve r.
I think Mr. Vento 's point, howe ve r, is eis t 20 l some of the gauges and other readings have a lag time.
l l
21 l P.r. Ven to.
I think what they are saying is that there is i
22 not as much inertia in this system as in some of the others.
I 23 Mr. Stelle.
I think that's right.
24 li Mr. Ma tts o n.
Your point about procedures is well.taken, a-secere Aeaorms. ine.
25 There are several kinds of procedures we knew he had and we knew i
i.
i.
l I,
/]
I he was trained in.
Fi rs t, he was trained in a
.CCA.
i 2'
Second, he was trained and had procedures for a s tuck-open i.
3 pilo t-operated relief valve, exactiv. what he had.
i I
4; Third, he had procedures for feedwater transient, which is 5
what he had.
I 6i The dif ficulty -- I get back to the point I made earlier --.
7l is he had all three simultaneously.
Now, was his training j
t 3
suf ficient for him to knew what to do in that peculiar circum-9l' l
stance of having two or three things, an accident, transient i
10 nature, going on all at once?
That is an open question, I think..
l 11 Mr. We ave r.
Excuse me.
We ha ve a vo te.
I had planned on I
12 adjourning at 6 :00 nnyway.
I am going to le t the s taf f, while s\\
i
+
13 we cc vo te, continue on the sequence.
If anybody wants to come !
14 back for the few minutes remaining -- I intend to -- that's j
15 fine.
But basically the task force would be adjourned for 16 '
to day.
I want to take this up.
Oces everycody have our i
i t
I i
17 l s che dule ?
We are going to be interviewing actual operators on 2 13 l Friday.
And on Thursday we will be --
t i
s 19
- All ri gh t, you' ve go t this.
That's fine.
I don' t need to t
20,
go over that.
I I'
I 21 The quest:.cn :.s, next Wednesday, a week from today, seems to l
1 22 l be an occortune time to to interview the o f ficials o f the 1
l i
23i govern =ent o f Pennsylvania, de Lieutenant Governor and de m
i O
24 !. Cabine t s taf f =enbe rs on the conmi tte e, and c.ossibly the 3
c ;,ee,.. = eco,:en, re. '
25 Go ve rno r, as to the informatien they were receiving and hcw i
l l
1
t 83 I
h
/^
1!
their decision =aking process was arrived at.
Te have a full
(
l unders tand, '
i 2'
conmittee :.ee. ting that day, b ut with no busine ss, I 3'
that will take up the cc =ittee.
Is th at s atis f ac to ry to 4
eve r2'bodv. ?
5 Mr. Carr.
I have a question.
l 6,-
1 i
Mr. We aver.
Sure.
l 7!
Mr. Carr.
On Friday the sensitivity is to weekend j
i 4
8 s chedules, et cetera.
Co we have assurance we will be done f
l 9
by 3 :00 o ' clo ck ?
i 10 Mr. Weaver.
We will be done whenever you want.
When do f
11 i we start Fridav?
At 10 :00?
I can assure you we will be done 1
I I
l 12 '
k v. *es.
(
P 13 Mr. Carr.
I hope we will have a dif ferent room.
i 14 Mr. Weaver.
We will trf. to get the best possible room we 15 can.
t i
16 Sv. the wav.c let me go into this other point firs t.
That i
l fo r us.l 17 '
is, as to press, as of now these meetings are, you knew, I
18 New, if somebody makes an issue of it on the press, we will u
i 19 l probably have to vote whether to close them or not, but I don't i
20 i intend to raise that issue new.
Richt new thev are informally 21 I, o rivate.
I 22 !
New, as to cress, I can j us t tell vo u f rom my ;cint o f l'
s c
23i view I would like to see as little cress as ecssible until we w) i 24 :
-- this is f act-finding.
We are not taking valtMs 4, u.dc.=en ts.
.= J,es v s.:c,ms, is.
25 i I don't intend to hcid any.c res s co nfe ren ces o f any kind un-il
'W l
i i
1 t
s, I
i I
f 9
1:
the task force decides at the end of its journey that we have i
i 2'
sore thing to communicate if we do.
So I don ' t think e.e press 3:
need be involved at this ceint.
Naturally, eve ry mencer is 3
I 4!
free to do what he wants.
I just want to tell you what the i
i i
5, philosophy o f the enairman is, that I don't intend to indulge or,
l, 6l engage in press at all during our time.
Each member can 7'
respect his own iudgment on this.
\\
l l
l 8
Mr. Ven to.
Mr. Chenev and I want to be inc.luded in the i
i l
9l initial press release in terms o f names.
i I
i 10 Mr. Che nev. '.
I am no t s ure I do.
t, l'
11 (Laughter.)
I 12 Mr. Weaver.
I am going to go ahead and schedule the trip I
i l
13 4 to Pennsylvania en Wednesday =crning.
If there are any s trong i
14 and general objections, le t us knew.
15 Mr. Myers.
I want to ask just one thing.
On the two 1
l 16 '
val'as f rom the eme rgency flew cume, the one that did not ocen, 17! is it v.our understandin9 that someone closed that pump during a i i
l i
i la routine maintenance operation and did not open it?
i.
I 19 Mr. Stello.
That is correct.
y I
20 '
Mr. Eisenhut.
That is ri ch t.
I 21 l Mr. Ma tts e n.
I wo uld s ub s titute the wo rd " s urve il'anc e" 22 f 0 r " maintenance. "
t 22,
Mr. Eisenhut.
" Maintenance is righ t.
O" 24 0 Mr. Myers.
We have no informati0n on who that was or hcw adecerai sexr ers. tre. n 25 ' that ha=. :. ene d o r how we ne ver ge t to the point o f de te ctine. th a t,-
i 9
85 l
4 1f is that correct?
t 2l Mr. S tello.
Tha t is co rrec t.
i s
t 3j Mr. Mye rs.
Is there no 'way to manually open that?
I l
t t
i 4;
rean that is not within the chanter here, is it?
Is there no I
t 5!
otner way?
If an emerc. enc.y came up, if you knew there is an i
t 6!
emergency through your bu :er, ligh ts a nd so fo r th, no way l
i I
7i that could be corrected?
8 Mr. Stello.
He could open it f rom the control rcom, 9-pusninc a cutton.
t 10 :
Mr. Eisenhut.
It's a manual coeration.
i l
11j Mr. Mye rs.
He tried that, didn' t he?
12 -
Mr. Eisenhut.
No, that was, in fact, the ope ration, a i
I 13 couple more steps in the sequence where, in f act, he did do 14 that cperation.
15 Mr. S te llo.
After about eight minutes he opened it from i
\\
i 16 i the control room.
l 17 I Mr. Ve nto.
There are three pumps there but only one line; i
i n
13 I is that accurate?
l I?
Mr. Stello.
No, there are two lines.
20 l Mr. Ve nto.
One 9ces dcwn -- the e=erc.ency ficw c u==
was I
4 I
I i
21 ;
down, ycu s aid?
i I
22 !
Mr. Eisenhun.
No, tnere are three pumps -- two electric i
22 ' Ouses and one turbine-driven p ump.
The two. electrics go --
O V
i i
24 Mr. Vento.
The'y feed into this one, but there is only cne e..s.e.,.. seco,ees.<ne.,
25 line in essence --
i I
i
1 86 i.
t I
I Mr. Stello.
There are two steam generators.
Each s team
({}
2 generator can be arranged with a valve arrangement so that 3;
each steam generator has 100 percent auxiliary feedwater flow t
i4i to it.
Its ficw path is pretty independent and redundant to the i
S' flow path in the other s team generator.
This is a schematic.
t t
I 6i Mr. Ven to.
- Yes, i
7l Mr. S tello.
There is a sirple arrangement of pipes and t
3i valves.
I 9'
Mr. Ven to.
If one valve is closed, that valve that we are 10 talking about that was closed that Mr. Marriott mentioned a 11 i moment ago, would that preclude the cperation of water going 12 frcm o ne o r the o the r pumps?
{-)s 13 l!
~
Mr. Stello.
No.
1 14 Mr. Ven to.
That by itself, there could have been, sho uld j
i o
15 i have been ano ther flow of water that would have picked up in i
i 16 !
the absence of that one cperator?
17 Mr. S te llo.
Yes, he had to isolate two block valves, one i,
13 in each independent flow path.
I t
i 19 Mr. Vento.
You also commented about the other one should 1
20 ! have been open.
In this schematic, an extra valve, one should i
t 21 have been open.
Do you think it was cpen right now?
i, 1
22 '
Mr. S tello.
Normally, both of those valves would ha a been r
I 23 ;
crenine, and he sho uld have had two independent ficw paths,
()
I 24, one in each steam generator.
i o.s,eer a oenen. ime. '
l i
25 '
Mr. Ve nto.
Th at 's ri gh t.
i 1
37 O
1l x=
s eetto -
s1==e eese vetves we=e ciceee, dota or *ose 2
flow paths we re blocked.
3l Mr. Ve n to.
One valve was closed manually.
The other one 4!
was closed --
5!
Mr. Stello.
Tao valms were closed, one in each flow path.
6, Ecth of those are capable of being opened from the co ntrol l
7-room.
3; Mr. Ve n to.
Okay.
But he tried; dey were both closed.
I 9
Cne was closed manually, but you are saying bcth of dose were I
I I
10 :
closed manually?
11 !
Mr. S tello.
Manually means remote manually.
8 U^
12 Mr. Ven to.
It was n' t autcmatic.
13 Mr. Stello.
No.
i 14 i Mr. Vento.
In o ther words, he didn' t go down -- in other
,l 15 words, he nad a button up there.
i 16 !
Mr. Stello.
He could have opened dem from the control a
17 -
room, ves,
i i
13 Mr. Ve n to.
Fo r whatever reason, both of dose were closed i
19 manually.
i 20 !
Mr. S tello.
Yes.
I I
I 21 l Mr. Vento.
Sod were closed manually.
I.e t me j us t pursue i
I 22 '
a point.
Cn this relie f valve, you are tal.<:.ng acoitt this pipe n
23 temperature which became quite an issue here.
What is' de V
i 24 nature of that thermocouple dat you are talking about?
Cc es 1
I:e-Fec:ess t eoo,ters !ac.,
25 it j ust measure pipe temperature, or does it measure de conten-I l
i a
88 t
I in the pipe?
Mr. S tello.
Pice tem erature.
2li
)
3 Mr. Ve n to.
In other worn., if you have s team coming out at i
I 4
500-and-some-degrees, v.ou wouldn' t ex=. ect the cine to be de i
l I
5!
sare tampe rature.
It would be somewhat eccler, I don't know I
l 6
how much, depending upon what its thermal qualities we-a 7s l
l 7'
tha t aluminum?
What is i t?
Stainless s teel?
3, Mr. Stello.
Carben s teel.
"h a t is de controversy, what i
9 should they have read.
I I
i 10 '
Mr. Ve n to.
There is only ona
-a'ia' valve on this, one 11 -
.crima:v. relie f valve, althouc.h, as yo u said, an electromagne tic 12,
with a solenoid valve, there is only one valve on it.
What if h
13 I that relief valve didn' t open?
i l
14 '
Mr. Mattson.
The re are a nurbe r o f o th e r s a f ety valves.
I i
15 i
Mr. Ve n to.
A nurter of other safety valves that operate.
i 16 l:
Mr. Mattson.
At highe r pressure.
1 l
17 Mr. Vento.
There is no backup sys tem if a valve s tays open..
la i If a valw stays open, what is the backup system for da ?
Then i
19 ;
you have dat going to that quench tan'< with alarms and so i
.i 20 i forth, but there is no other mechanism in terms of valves s tav-I 21 !
ing co.en really o de r dan tan'<s.
f i
22 i Mr. Mattson.
For de safety valves, no.
l i
23 :
Pr. Stello.
Sa fe tv. valves cannot be isclated bv. ccde.
If
\\
t 24 de safety valve fails open, it will renain opqn, because you M.E tc ee gt 2fMf tt?1, I *C, i
5 cannot put an isolation valve :or a sa:ety valve.
i f
v l
39 l
i l
I i
s 1.
Mr. Vento.
You couldn' t do that because that obviousiv i
[
would double v. our chances o f -- that's ric.ht.
Yes.
t 2
j I
I 3i I have to go in a minute, but you implied his particular 4,
type c f plant had had a certain number -- Mr. Mattson, I think, j 1
1 5;
mentioned 150 incidents where the relief valve had opened on i
I 6
the s e, and in 38 reactor years ; I guess that was it.
I won' t I
7i hold you exactly to the numbers.
I S
Mr. Mattson.
That 's abo ut ri gh t.
l 9
Mr. Vento.
Do you know if that is high, is that low?
Do b
i 10 l you find that unusual or is that a usua-1 event?
I d
is a:y nusual number of PORV openings /
11 l Mr. Ma t tson.
That O
12 for a ?WR.
\\
l 13 I Mr. Vento.
Yes, that would nake about four a year on these i
i i
14 l particular tv.. es o f =. lants.
t i
15
- You also implied this particular plant had that occur.
i 16 I How many previous to this had that occurred?
l 17 l Mr. Mattson.
At Three Mile Island?
Several.
I don' t know la i.
the exact number.
Understand, the FORV is put there in-t l
l 10
-antionally as a design feature intended to open --
t 20.
Mr. Vento.
It's a backup svstem?
I i
t 21 Mr. Mattson.
No, it's no t a b ack up ; intended to increase i
i 22 ;
de availability of de reactor for upseu conditions in de 4
t i
_p 22,
seconda:,/ system.
I t' s a unique f e a ture o f de B &W desi gn, de 24 way :. 's used wid the B &W :.ntec. rated control sys tem.
O the r I
- wercew newws. wa.
i
~
25, plants have pilo t-cpe-=-= A -='i=' valves, but the extent to j
I l
)
.. _... ~ _.. =. _ _ _.. _. -. _ _ -. _ _... _. _ _ _ _ _. _ _
i w
a.
i 90 l
l i
3.
4 l
1 which 3&W uses them is unique.
Therefore, 3&W has had.cre i
l 2'
PCRV openinc:s than have other PWR desiens.
i l'
3 Mr. Vento.
I have to gc.
Thank you very much.
I
[
4 (Whereupo n, at 6 :0 2 p.m., the hearing was adjourned.)
j i
i i
5 l
l i
6'i i
7 1
i-I 8'
l i
9,!
i I
i 10 l I
11 l s
12 '
O
\\'
13 i
i 14 l
f i
15 1
.- l 16 i l
17 18 19 1
2o l l'
21 l 22 i i
i 23 i O
21!
~
l-i m.;ecess< nesonen. va..
25 !
- i
.l
,t i.
..:-6
,