ML20210E231
| ML20210E231 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/18/1986 |
| From: | Atomic Safety and Licensing Board Panel |
| To: | |
| References | |
| CON-#386-836 LRP, NUDOCS 8609220166 | |
| Download: ML20210E231 (150) | |
Text
-
Uh1TED STATES O
NUCLEAR REGULATORY COMMISSION OR/GWAJ_
IN THE MATTER OF:
DOCKET NO: LRP INQUIRY INTO THREE MILE ISLAND UNIT 2 - LEAK RATE DATA FALSIFICATION l
l O
f LOCATION:
BETHESDA, MARYLAND PAGES:
1758 - 1905 DATE:
THURSDAY, SEPTEMBER 18, 1986
. /0 t i
, i O'L ACE-FEDERAL REPORTERS, INC.
OfftaalReporters 444 North Capitol Street Washington, D.C. 20001 hDR D
0 20 T
PDR NATION %TDE COVERAGE s
CR28124.0 1758 BRT/cjg 1
UNITED STATES OF AMERICA 8
_J 2
NUCLEAR REGULATORY COMMISSION 3
BEFORE THE ATOMIC SAFETY AND LICENSING BOARD 4
X In the Matter of:
5 Docket No. LRP INQUIRY INTO THREE MILE ISLAND 6
UNIT 2 - LEAK RATE DATA FALSIFICATION 7
- - - - - - - - - - - - - - - - -x 8
9 Nuclear Regulatory Commission Fifth Floor Hearing Room 10 East West Towers 4350 East-West Highway 11 Bethesda, Maryland 12 Thursday,-September 18, 1986 rx 13 i
k/
The hearing in the above-entitled matter convened'Et 8:30 a.m.
15 16 BEFORE:
17 JUDGE JAMES L.
KELLEY, Chairman j
Atomic Safety and Licensing Board U.S.
Nuclear Regulatory Commission 18 Washington, D.
C.
19 1
JUDGE JAMES H. CARPENTER, Member Atomic Safety and Licensing Board 20 U.S.
Nuclear Regulatory Commission Washington, D.
C.
JUDGE'GLENN O. BRIdHT, Member 22 Atomic Safety and Licensing Board U.S.
Nuclear Regulatory Commission 23 Washington, D.
C.
24 I) 25 v
ACE-FEDERAL REPORTERS, INC.
3)2 347 3700 Nationwide Coverage 800-3364646
1759 1
APPEARANCES:
2 On behalf of GPU Nuclear Corporation:
3 ERNEST L.
- BLAKE, JR.,
ESQ.
JOHN NASSIKAS, ESQ.
4 Shaw, Pittman, Potts & Trowbridge 1800 M Street, N.W.
5 Washington, D.
C.
20036 On behalf of the Employees:
6 HARRY H. VOIGT, ESQ.
7 MICHAEL McBRIDE, ESQ.
LeBoeuf, Lamb,.Leiby & MacRae 8
1333 New Hampshire Avenue, N.W.
Suite 1100 9
Washington, D.
C.
20036 10 On behalf of Jack Herbein:
JAMES B.
BURNS, ESQ.
11 Isham, Lincoln & Beale Three First National Plaza 12 Chicago, Illinois 60602 13
, (~]
CHRISTOPHER W.
FLYNN, ESQ.
\\/
Isham, Lincoln & Beale
~'
14 1150 Connecticut Avenue, N.W.
Washington, D.
C.
20036 On behalf of Gary P. Miller:
16 MICHAEL W.
MAUPIN, ESQ.
17 M.
CHRISTINA HENSLEY, ESQ.
Hunton & Williams 707 East Main Street 18 Richmond, Virginia 23221 19 On behalf of Former Metropolitan Edison Employees:
SMTIH B.
GEPHART, ESQ.
21 Killian & Gephart 217-218 Pine Street 22 Box 886 Harrisburg, Pennsylvania 17108 23 On behalf of the NRC Staff:
24 JACK R.
GOLDBERG, ESQ.
MARY E. WAGNER, ESQ.
f
25 w/
U.S.
Nuclear Regulatory Commission Washington, D.
C.
20555 ACE-FEDERN, 9EPORTERS, INC.
E-347-3700 Nationwide Coverage 800-336 6646
82124.0
.BRT 1760 0
1 PROCEEDINGS 2
JUDGE KELLEY:
Good morning.
Two quick 3
announcements.
After lunch today we'll announce our ruling 4
on the legal question discussed the other day about the role 5
of the Staff and follow-up questions.
Also, it was suggested 6
yesterday afternoon by Mr. McBride that we would like to 7
right after lunch, assess where we are and see what we think 8
we should do in the way of witnesses for next week so we can 9
notify those people.
10 MR. MC BRIDE:
Judge Kelley, in order to assist in 11 that discussion I wanted to tell you and the parties that we 12 hope but are not certain that we will be able to file 13 Mr. Seelinger's testimony in advance of the date that the 14 Board had previously allowed us.
We hope to do so, to assist 15 the Board, in its questioning of Mr. Havercamp.
16 JUDGE KELLEY:
Thank you.
That might be helpful.
17 We can resume where we were.
18 JUDGE CARPENTER:
Resuming the questions on the i
19 individual tests, I believe the next one is NRR test 69.
20 MR. STIER:
That's our test 89, Judge.
21 JUDGE CARPENTER:
Thank you.
The first question, 22 number 392:
On what basis do you conclude that the alleged 23 hydrogen addition had a 100-gallon effect on the makeup 24 level?
Mr. Stier saw no such effect.
On what basis do you 25 conclude that there was a hydrogen addition at all?
{}
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80 4336-6646
82124.0 BRT 1761 0
1 MR. CAPRA:
The 100-gallon effect on the makeup 2
level indication'is based upon the change in indicated makeup 3
level which could be caused by either a hydrogen addition or 4
level transmitter switching.
If a hydrogen addition caused 5
the effect on makeup level indications seen on the strip 6
chart, it is unknown whether it had any effect on the makeup 7
level transmitter providing input to the computer.
8 Based on the slope of the makeup level indication, 9
assuming constant leakage and taking into account changes in 10 T av and pressurizer level, the makeup level indication 11 printed out on the computer is about what would be expected.
12
.Thus, while there was an obvious effect on the strip chart 13 the actual effect on the other level transmitter is not 14 clear.
15 As stated earlier in our testimony, since hydrogen 16 did not always have the same effect, if any, on makeup level 17 indication, it is not possible to confirm a hydrogen addition i
18 was made unless there are log entries to confirm the 19 additions.
In this case, there are no log entries to confirm 20 NRR's conclusion that this was, indeed, a hydrogen addition.
21 Although somewhat more erratic, it does have the 22 general characteristics of a hydrogen addition seen during i
23 some logged additions.
These are NRR test 152, 134, 123, 24 120, 63 and test 23.
25 Except for possible level transmitter switching,
( },
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6 4 6
82124.0 BRT 1762 I
which I'll discuss next, there are no logged evolutions which 2
would explain the observed trace.
3 Leak rate test 69 took place during a period where 4
LT-1 was known to be exhibiting unusual behavior, one 5
possible scenario for level transmitter switching which could 6
explain the erratic trace I'll go through now.
If you start 7
before the leak rate test began, level transmitter 2 is on 8
the strip chart.
At time 0805 was the start of the leak rate 9
test.
At that time, LT-2 was reading 66 inches.
TL-1, read 10 by the computer, was 64.4 inches.
11 At time 0844, switch to LT-1.
At that time, LT-1 12 is reading 67.5 inches.
13 0855, switch to LT-2.
At that time, LT-2 is 14 reading 65.5 inches, which is consistent with the initial 15 slope had no switching taken place.
16 At 0857, switch to LT-1.
LT-1 is now reading 69 17 inches.
18 0905 -- that's the end of the leak rate test --
19 LT-1 is reading 67 inches.
LT-2, now read by the computer as l
20 64.9, which, again, is consistent with the initial LT-2 slope l
21 had no switching taken place.
22 At time 1000, switch back to level transmitter 2, j
23 LT-2 remains on the strip chart for the rest of the excerpt.
I 24 If this test was run according to the scenario
[}
25 which I just described, during the test period the change in l
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646 l
82124.0 BRT 1763 0
1 makeup level seen by the computer was plus
.5 inches, while 2
the stable LT-2 change was minus 1.1 and the unstable LT-1 3
level change was plus 2.6 inches.
4 This test was conducted, as I mentioned, on 5
December 20, 1978, between 0805 and 0905.
This was during a 6
period where the LT-1 was tagged "out of service"; the 7
out-of-service sticker was not removed until 1335 that 8
, afternoon.
9 JUDGE CARFENTER:
Any comments, Mr. Stier?
10 MR. STIEN:
Yes, Judge.
We are not certain what 11 caused the deflection in this trace, but I don't believe it 12 was a hydrogen addition.
The reason'I don't believe it was a
(
13 hydrogen addition is because the trace is not similar, in our 14 judgment, to what we know to be the trace caused by the 15 effect of a hydrogen addition on makeup level indicator that 16 is malfunctioning because of the loop seal condition.
17 As a point of reference, we have a trace which we 18 looked at yesterday, which we believe is characteristic of a 19 hydrogen addition under the circumstances that I described.
20 And, in addition, as another point of reference, we had a 21 similar trace obtained in another investigation in which the 22 same kind of makeup level instrumentation was used.
When you 23 compare that trace with the known trace that we have at TMI-2 24 with this trace, I don't see -- MPR doesn't see the
()
25 characteristics that we would look for.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336 6646
I
'82124.0 T
1764 l
1 There is a more important reason why I don't 2
believe there's a hydrogen addition.
As you know froni the 3
testimony, a hydrogen addition should have no effect 4
whatsoever on makeup level indication, unless there is a loop 5
seal or some other malfunction that we -- that might cause 6
the effect that we have been describing.
If you look at i
7 MPR's cover sheet for test 89, you'll see under " water 8
addition," the last two entries are " ratio of indicated to 1jlogged additions for the test" and " ratio of indicated to 9
10 logged additions for the day."
11 Both of those numbers indicate that in terms of 12 this loop seal effect, the nakeup level indicator was showing 7-s d
13 no s'ign that it was affected by that condition.
Therefore I
14 you shouldn't see any effect at all on the strip chart, from 15 j the addition of hydrogen to the makeup.
16 The deflection could have been caused, as 17 Mr. Capra suggests, by switching the makeup level 18 transmitters during the course of the test.
That's quite 19 possible.
We know during this time that LT-1 was 20 malfunctioning, was giving that very erratic trace, and in 21 ltryingtodeterminewhichleveltransmitterwasusedforthis 4
22 particular test, MPR made a judgment that it may very well 23 have been run using LT-1, the bad transmitter.
And during 24 the course of the test the operator may have switched
()
25 transmitters, perhaps to see how LT-1 was reading.
We are t
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationaide Coverage SO4336-6M6
~,.
82124.0 RT 1765 l
1 j not certain.
That's'our best judgment on this test.
2 We did not draw any inferences from the evidence 3
related to this particular test.
4 MR. CAPRA:
May I make two comments, sir, on what 5
Mr. Stier said?
6 JUDGE CARPENTER:
Yes.
7 MR. CAPRA:
Two things.
As I mentioned when I 8
I talked about the possible hydrogen addition, I kn'ow I've 9
stated before that I do not believe that all hydrogen 10 additions had the same characteristic trace as the experiment 11 run, which Mr. Stier referred to.
That is based on observing 12 logged hydrogen additions and their effect on the makeup O
13 level indicator on many occasions where there was not a leak 14 rate test run.
I don't think we can expect all hydrogen 15 additions are going to have that effect.
I'm not saying in 16 this particular case that I'm certain this was a hydrogen 17 addition.
My testimony was that that was one possible 18 explanation.
I believe the level transmitter instrumentation 19 switching is a more credible explanation, more reasonable 20 Iexplanation.
21 The other point, if it were a hydrogen addition, 22 while I agree with Mr. Stier that, based on the additions 23 j during the day it did not appear that a loop seal condition l was present, that may be true for LT-2; that does not mean 24 O
2s that it was present for LT-1.
ACE-FEDERAL REPORTERS, INC.
I 202-347-3700 Nationwide Coverage 800-336 6646
i 82124.0
/T;T 1766 k_)
1 As I said, if a hydrogen addition was made, I 1
i l
2 believe it.only affected LT-1 in this case, and did not 3
affect LT-2.
4 MR. MC DRIDE:
In light of those answers we would
. ithdraw the next question.
j 5
w 6
JUDGE CARPENTER:
Since you are free to 7
hypothesize several possib.le explanations for the observed 8
trace, what inferences did NRR draw?
The question said "on d
9 what basis do you conclude?"
Do ycu really come to a 10 conclusion here or do you merely present speculation?
11 MR. RUSSELL:
We did not present inferences with g]
respect to willful manipulations with respect to any 12
\\_/
13 operators based on this test.
14 JUDGE CARPENTER:
To make sure I understahd, 15 examination of this woul~d show that it violates the " steady 16 state requirement for 60 minutes," since the slope went down 17 and then up, et cetera.
So, by inspection, there should have 18 been an exception or --
19 HR. RUSSELL:
We didn't conclude On the steady 20 state.
We believe the plant was shut down at this time.
It 21 appears the trace does not appear to be steady state.
That 22 could be explained by the switching of the transmitters with 23 the out-of-service tag on one of the transmitters.
That 24 would not be in accordance with administrative procedure 1036
()
25 and would not meet the intent of the surveillance procedure.
ACE-FEDERAL REPORTERS, INC.
[
M-347-I'00 Nation + ice Coserage 80 4 336-6646
e-.
I l
82124.0 1767 gBT -
l W
l I
1 i
When I talked about.
" Inferences drawn," I was
-l 2
I using it in the same context that Mr. Stier has; that is, an Y
3 inference of a willful manipulation.
We did not reach that 4
conclusion.
5 It is one of the tests which we characterize as 6
questionable, with respect to one which we believe was not 7
I conducted in accordance with the procedures.
8 JUDGE CARPENTER:
That's the point of my 9
question.
Is it really necessary to reach a conclusion about i
10 what the most probable speculation is in order to reach the 11 conclusion that this clearly was not -- this clearly is a 12 questionable test?
It should have been flagged.
()
13 MR. RUSSELL:
It was a questionable test.
It was 14 flagged in that context, and we identified it as either a 15 hydrogen addition or possible level switching.
We considered 16 that as circumstantial evidence but we did not use this test, 17 in particular, to reach a firm conclusion on an operator.
18 JUDGE CARPENTER:
My question is, is it really 19 necessary to go that far?
Is it necessary to speculate on-20 what the cause was in order to reach the conclusion that the 21 operator did not, since he accepted a test which didn't meet 22 the procedural requirements?
23 MR. CAPRA:
This gets to where we were talking 24 about yes cerday, and did we consider a test or not consider a
()
25 test in drawing a conclusion.
There are many of these tests ACE-FEDERAL REPORTERS, INC.
202 3C 3700 Nationwide Coverage 800 3E6646
i 1
1 l
82124.0 RT i
1768 l
l 1
jwhen, lumped together under individual operators or under l2 individual shiftc, that led us to conclude that many of these 3
tests were not run in accordance with the precedure.
But i
4 it's different when you are trying to make a call, whether t
5 j this tert was intentionally manipulated, to try to get a 6
better leak rate test result.
l'
~
7 MR. RUSSELL:
If I could characterize it, we 8
censidered all the tests, and those which were questionable 9
or were obviously not done in accordance with the procedure, 10 we used th'ose in reaching conclusions generall*j about 11 negligence, with the way the test was conducted.
We did not 12 draw the inference in all.of the questionable tests that they 7-)
\\J 13 were willfully manipulated.
That's a much higher threshold I
14 that we used for those tests, and we felt that we needed to 15 have a repetitive pattern of activity from the analysis, some 16 statements by at least an operator on the shift that that 17 method was used, to draw a conclusion of deliberateness with 18 respect to a willful manipulation.
19 And it goes to an issue that we felt was fairly 20 significant, and that is:
Operators perceived or 21 rationalized that the hydrogen additions, because they 22 weren't precluded by the procedures, may not have been as 23 wrong as adding water which was -- it was clear that a water 24 addition during the test that was not recorded would be
()
25 incorrect.
So we used a scale in judging each test and how l
ACE-FEDERAL REPORTERS, INC.
_ g,.n
~gg,
k_
?
l i
I fb 82124.0 i
fS T
l 1769 r
t b
1 I we used that in drawing conclusionv about willfu1Aeaa, He l
[ -
2
! used all of thee in reaching concluiert about conduc.t of
(
i n
l operations, Jtspons-lbility of the opeu tiou departasnt 3
i e 4
nanagement for the ecnditior.s that existM at the tir,e.
5
{
JUDGF CARPENYM2:
Mr. Stier?
h 6
a NE. BTMP.:
Judge, you said hwOin;r that I"a not
{h 3
1
- l 7
sure I'm in egreneM. Vith.
You 341d t.'M
- locking at the j (
6
] strip chart, one CM10 concittle tMNt t.% eAot we run dating f y-a
\\ i 9
nonsteady state condition.t.
I"n r;ot stile we ucyld conclude j
\\
=
1G that.
g
) L 11
?.i tNr le Ys.i E n t n d 't u f c% #w'.10 Lad Art'170 th j(
g 1 g o
12
, test and tM toglaning rGH We wy.. t (Ao M s.s l g O
I
^
13 transnitt*J, Oc Mdirq Ncdibim W3f A N M on Ancther,
{
14 l clearly it'7 t Violation 97 yra cedt.f..%
J. Mn't Aave p y I
}
15
, doubt aborst that.
I'A ndr nVe fM t'f uMt h0ppefed, j 1 l
(,
16 though.
17 If tha apertror.w!rd.%id LLerwmt6 duriag.St L
iB I coJrse et.the teet %:(, teos The 54<patiM rCAd2ngs us' onk
' i l
19
.! transmitter.att tMe Ja e -~ EJ & creding rpA-]tne
<>r. the t
1 C
20
'same transmitttT. even thou@ (hs tynq.4 it c Teese in
! 0 21 betwaen,.I wouldn ' t wsMar *. raat 4 M M gertu wr# ander
}
I ;
22
[nonetsady c ate corn:E Uchs, I
23 g
When seu !eas a t 3 w gr w;..At that wrs reccrded i
24 in the test, in the JenNt ed tsC pM ared, rw see that it o
l eoes not indicate e <> tinae :.: c:2.ra e s e in preauretzer 2s
)
I Aa-FEDERAL RiFORTERS, INC.
1
~,c *
- r G t wr th).WMA M-347-3?M s
- E
[,.,
,7
g me it f
k 4
87.124..3 j
gRT 1770 I'
W h'
.1
) Swel, makeup tur.A 3 ervel, and aW& age temperature that would c.
1 F
2
) tuggest that t.he plabt is not at steady state.
b i
g 3
I just wanted to make that point because we did h
4
)not -- our approach, perhaps, was a little bit different.
We f
i F
tried tc understand as Jr.uch as we could about each test, i
k-6 crtract evidence froze tr.e test document.ation that was l
4 h'
7
{re, levant to any cf the iswes that we were looking at, t
S
)whetherthwasmanipulationornegligence,runningatest l
,1 0
9 during boastesdy state conditions, whntever it happened to be 20
~- extra:t a s mbch e:vidence ar we could from the test data, k
{ M we fdit t.r.at that evidence was reliable, we were certain 11 ll '
12 a>Wt it, we would use it.
If not, we just didn't draw any 1.3 interendy whatsoever froat the test.
We did not try to g
H p tegcries t:sts, group tests together, perhaps, in the same l$
way thet the Staff did.
L-16 Looking at this particular test we did not infer L
t 17 I that this test was run during nonsteady state conditions.
We L;
16
- ald it was possible that the test had been performed using l P..>.c bad makeup level transmitter.
L-j But, even there, I did l
}
E 4
20
?not, when it came to actually drawing an inference that an j
m>
M cperater used the bad level transmitter for this particular
(
22 test, I didn't feel that even that war strong enough to draw
)
lI i
j g
23
- an inferent e from so I didn't use the test, even for that
(
24
- purpcse.
i F
I l
?S h
JUDGE KELLEY:
The Staff indicated that they n
l l
l ACE-FEDERAL REPORTERS, INC.
l'
_ _ -_ _-3700 Nationwuie Coverate M336-6646 i
202 347
I l
- 82124'.0 RT.
1771 I
thought that one. interpretation of this test, the strip 2
chart, would be the hydrogen addition occurred and cited half 3
a dozen other tests where they thought the hydrogen 4
indications.were similar where, I gather, they thought 5
hydrogen had been added.
I 6
You, Mr. Stier, as I understand it, did not agree 7
with that and referred to a particular test which, I gather, 8
was a test where an experiment was run and someone wrote on 9
the strip chart that they actually had used hydrogen.
10 Apart from that particular test where you had that 11 l kind of pretty clear extrinsic evidence, did you conclude l'2 that any other test involved hydrogen additions, in all (1~)
13 likelihood?
3 14 MR. STIER:
Yes, we did.
There were several, I
15 iseveral tests.
X c4n get you the numbers of those tests.
16 I'll give you the numbers of the tests and 17 indicate whether MPR determined that there was an effect; le that is, where we believe that a hydrogen addition took place 19 during the course of the test, it may not have had an 20 effect.
The reason we concluded it occurred during the test 21
/was that the time it was logged corresponds to the test
[
22 period.
When they examined the strip chart in some of these 23
! cases they could see no effect whatsoever.
24 JUDGE KELLEY:
I guess what I'm really after is,
{)
25 are there any cases where your basic conclusion -- I realize y
ACE-FEDERAL REPORTERS, INC, aw en.ames~
pum-_
l
(
I' 82124.0 1772
("_)T 1
you don't conclude that a person did something.
You conclude
~
)
2 that the evidence is strong that the person may have done 3
something, something like that.
i 4
MR. STIER:
Right.
5 JUDGE KELLEY:
But bearing that in mind, is there 4
6 any hydrogen addition test, test that looked suspect on that 7
ground, where you concluded that, basically or primarily or 1
strongly on the basis of the strip chart, or did you always 8
1 9
have to have some sort of supporting -- like a log, for 10 example?
If I were trying to sneak in hydrogen I wouldn't 9
11 log it, I suppose.
I'd just put it in.
r 12 Are all your cases logged cases?
Legged l
13 hydrogen?
14 i
MR. STIER:
No.
15 JUDGE KELLEY:
Okay.
16 MR. STIER:
First of all, let me give you a couple 17 where there are log entries and you can look at those strip 18 l charts and you can see the strip charts that we think show i
19 l the effect of a hydrogen addition during this loop seal l
20 condition.
21 Test 36, tes t 35, are two instances where MPR l
22
[ found a logged addition of hydrogen and observed what they l
l thought was an effect on the strip chart trace.
There we are 23 i
24 really certain that those are hydrogen additions and that had
(/
25 an effect.
i 1
ACE-FEDERAL REPORTERS, INC.
I s uw~
.a-~~r w-
m I
i i
8?.124.0
(
it 4
17?3 r^
U 8
1 There Are two other inntancec where MPR was not l
2 certain that there was A hydrogen addition, but I did draw an 3
inference that there were hydrogen additions.
And I drew the i
.I 4
inference from evidence that was extrinsic to the 5
documentation, the data that MPR had.
6 I had some additional testimony and a shift 7
turnover note that indicated to me that a hydrogen addition 8
had taken place during the course of a particular test.
9 Those tests are, I believe, 33 and 34.
10 If you look at the strip chart traces for those e
11 tests I think you'll see what we believe are the
,f-12 characteristics of the effect on the strip chart trace of a 13 hydrogen addition.
14 JUDGE KELLEY:
Are those characteristics sodehow 15 distinct and different from other F.inds of traces?
16 MR. STIER:
I,think they are.
Perhaps I ought to 17 let Dwight Harrison describe his sense of what a hydrogen 18 addition looks like under those conditions, because he is the 19 one who very carefully examined each of these strip charts to 20 try to make those judgments.
h JUDGE KELLEY:
Sure.
Mr. Harrison?
21 l
22 j
MR. HARRISON:
Let me make sure I understand what l
23 Mr. Stier wants me to answer.
24 (Discussion off the record.)
O 2s MR. axR,2sG :
what we did, because we were ACE-FEDERAL REPORTERS, INC.
'02-347 3700 Nat;ocwide Coverage 800-JM6646
m I
-92124.0
/"I 1774 i
r i
\\J i
i concerned about the hydrogen addition -- and it is summarized 2
in the report in table VII-1, which is in volume IV-A of the 3
Stier repcrt.
Because of the uncertainty in what a hydrogen 4
, addition looked like, we went and looked at every logged 5
hydrogen addition in the CRO log and the AO -- the auxiliary 6
ioperator's log, that we could find.
We looked at each and 7
every one of those, a total of 102 different instances, to 8
see if we cauld, from that, conclude that there was some very characteristic conditions.
9 10 Now, as well as the characteristic one that is 11 called out on the strip chart reenr4 and in the log that rm 12 say3:
I added the -- the one that Mr. Stier drew.
And we U
13 did find, in some cases, not during tests where hydiogen 14 additions were logged
- we certainly have two here, one on 15 the 21st of February and another one on the 2ist of February, 16 and some other possible ones, where there is a distinctive 17
- trace, 18 l
As far as the cther ones, we simply could find i
19 nothing characteristic anout them or we simply found nothing 20 at all.
This was the basis on which we came to conclude that 21
! a hydrogen addition probably -- if it is effective in 22
- affecting the test -- shows -- moved the slope n offset in 23 some Cashion.
And that is what we looked for when we looked 24 at four hydrogen additions.
l
()
JUDGE KELLEY:
When you say "nothing at all," you 25 ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nanonwide Coverste E63W686
82124.0 GT 4
1775
, V.
1 mean there's a 1ogged hydrogen addition and you look at the strhpchartandyouseenothing?
2
3 MR. HARRISON:
That's right.
Nothing.
No blips, 4
no dips, no movement of any kind.
In'some cases -- and I 5
have indicated a number as "possible" because there is some 6
kind of movement, but it is nothing that has any t
7 l characteristic to it as you see from the makeup charts that 6
you.have now been subjected to, they move around.
9 MR. COLE:
Since this was one of the earlier main 10 l charges of manipulation, I think the mere fact that we went 11 through some 100 or more of these tests, meticulously trying
,q 12 l to see if we could get a unique signature, this is the thing V
13 that has led us -- it is only in very isolated cases that we 14 can come up and pinpoint:
We feel that's a hydrogen 15 addition.
It wasn't that we very quickly passed over this 16 issue.
We dug in<until we thought we wrang it dry.
I'7 JUDGE KELLEY:
The coup 1e of cases where you 18 concluded there had been an addition, in those cases was j
19 there something distinctive atcut the strip chart?
2D MR. HARRISOh's The strip che.rt, in the cases where
.21 it was distinctive --
22 JUDGE KfLL2Y:
Again, those 7.re numbers what?
23
]
MR. HARRISON:
Well, ther aren't in the test.
24 There are two testa that have these.
But there are three O
2s instances which we have in ou= tan 1e which are veer c1eer.
I
?
ACE-FEDERAL REPORTERS, INC.
1 x
.-.#m
.w s
. _ yu
l 82124.0 l
T i
1776 l
1 which are not during any test of any kind.
2 JUDGE KELLEY:
Right.
I 3
~
MR. HARRISON:
One of them on the 4th of February 4
and two of them on the 21st of February.
5 JUDGE KELLEY:
Do you include those traces?
6 MR. HARRISON:
Those traces were not included in 7
our report.
8 JUDGE KELLEY:
They are not in your report?
9 MR. HARRISON:
We could make those available.
10 MR. STIER:
If you would like us to make copies of 11 those and supply them to you, we would certainly do that.
12 JUDGE KELLEY:
We'll come back to you on that.
13 But there are a couple of tests -,
14 MR. HARRISON:
Two tests.
15 JUDGE KELLEY:
-- that do exhibit something 16 distinctive about the strip chart?
17 MR. HARRISON:
Basically what Mr. Stier has drawn 18 previously.
Basically his offset picture that he drew.
19 (Indicating.)
20 JUDGE KELLEY:
We would find that in the record, 21 those two tests?
22 MR. STIER:
Those two tests; yes.
And there are l
23 also cases, two cases vhere MPR was not certain but I found 24 hydrogen additions, tests 33 and 34; traces we identified in
()
25 our testimony just now.
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646
82124.0 RT 1777 1
JUDGE KELLEY:
Any comment, Mr. Russell, 2
5 Mr. Capra, on'the Stier/MPR comments?
3 MR. RUSSELL:
I think we are going to be getting 4
to some subsequent tests which involve hydrogen additions S
i which we don't believe the trace continues with an offset 6
over a long period of time.
We believe that the tail of the 7
offset does come back down; that the hydrogen test is 8
transitory.
Whether that's a 30-minute period or longer, I 9
think for the purpose of the discussion, during a leak rate 10 test for the ones we have seen, the record is adequate at 11 this point.
12 I would comment that one of the things we looked 7_
t i
13 at was what happened to makeup level over the period of time 14 of the test.
It is obvious that there is some switching 15 toward the end of this test.
Whether you ended up on a 16 different transmitter than the one you started on, if that 17 were not the case, then we can't explain the situation where 18 the levels would essentially be constant -- that is, the 19 printed result on the computer of 64.4 and 64.9, with the other transmitter only increasing one inch from 66 to 67, 20 l
l 21 does not appear to be consistent with the gross leakage 22 pattern that's going on during that period of time when the l
23 plant is shut down.
That's why we feel that it is possible 24 that a switching did occur, as we previously indicated.
([-)
25 MR. CAPRA:
I would like to make one comment also t
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6 9 6
82124.0 gRT l
1778 W
l 1
on the hydrogen addition.
We did not look at all hydrogen l additions in the auxiliary operator's log.
We did, in the 2
i 3
course of plotting leak rate tests, look at all hydrogen 4
additions, whether or not they were during a leak rate test 5
period; the ones that were logged in the control room 6
operator's log.
7 The only problem that you have with that is that 8
you then only have information for one level transmitter 9
because you don't -- if you don't have a leak rate test 10 ongoing at the time, the only thing you have to observe is 11 the makeup strip chart recorder.
The advantage of being able 12 to look at a hydrogen addition during a leak rate test is you e
13 can see that, in some cases, it may affect one level 14 transmitter and not affect the other.
15 JUDGE CARPENTER:
Mr. Stier, numerous employees 16 have a brief follow-up that would fit right here.
It is a 17 follow-up to the question that I asked a few minutes ago.
18 The question is:
Are steady state condition 19 parameters applicable to the reactor or to the makeup strip 20 chart?
21 MR. STIER:
Well, I think it's applicable, in a 22 general sense, to the reactor, the reactor coolant system and 23 the related systems that are incorporated in the leak rate 24 test that we have been discussing.
I mean, it's what is O
25 going on in those systems that the term " steady state" ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
82124.0 RT 1779 l
1 relates to.
2 MR. RUSSELL:
Staff would agree and point out that 3
the precautions and limitations in the procedure which we 4
have used as a basis for the definition of steady state do 5
not address effects on the makeup level transmitter per se.
6 JUDGE CARPENTER:
And you completely reject my 7
notion that if I were sitting there running the test I would 8
look at the time rate of change of the makeup level d' ring u
9 the hour to see whether there had been perturbations?
I 10 might not conclude what had caused them or some plant 11 parameter that violated steady state, but at least if there 12 were significant changes in slope during the one hour of the
(_
13 test, it would raise a question as to whether steady state 14 conditions prevailed.
15 MR. RUSSELL:
Absolutely.
16 JUDGE CARPENTER:
And it's incumbent on the 17 operator to question it.
Not necessarily to conclude --
18 MR. RUSSELL:
The observation that the makeup 19 level indicator is a good indicator of what is going on in 20 the reactor system, assuming that there are no leaks or other 21 changes, is a valid one and it would cause you to look to see 22 whether unstable plant conditions had caused those changes or 23 not.
But the fact that temperature remains constant, 24 pressurizer level remains constant, and the other precautions
()
25 and limitations are met, that would be an indication that you ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33(H5646
82124.0 T.
1780 I
are at steady state.
If you then had a change in the makeup 2
level it would be as a result of additional makeup to the 3
primary and potentially because of a leak.
4 It is an indicator that would cause one to 5
investigate and question whether you had steady state 6
conditions or not.
7 JUDGE CARPENTER:
Let's say a leak was abruptly 8
repaired in the middle of a test -- an extreme hypothetical.
9 Wouldn't that still invalidate the concept of steady state?
10 MR. RUSSELL:
If a leak were repaired halfway 11 through the test the leakage that you would calculate would 12 be one-half of what the leakage was, because you'd have a O
13 leakage rate of one value for the first 30 minutes and you'd 14 be dividing by 60 minutes.
15 JUDGE CARPENTER:
What I'm trying to get at here 16
-- I'm surprised that it wasn't a simple criteria, if I 17 understand all this, in the training of the operator to look 18 and see whether the trace was a smoothly changing, level over 19 the 60-minute period.
If it wasn't smoothly changing to 20 question the test to try to figure out why and then decide 21 whether it was a valid test.
22 MR. RUSSELL:
I agree that that should have been 23 done and that that is a way of judging whether you are at i
24 steady state conditions.
It is also possible that something
()
25 other than not being at steady state can cause oscillations ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33 & 6646
82124.0 BRT 1781 0
1 in a makeup level.
You may be having leakage past other 2
systems that causes water to go into the makeup that could 3
cause oscillations in level.
You may have left a water 4
addition pump on for a few minutes.
It's not always the case 5
that oscillations in the makeup are caused by oscillations in 6
7 JUDGE CARPENTER:
Thank you for helping me learn.
8 Mr. Stier?
9 MR. MC BRIDE:
Judge Carpenter, I was writing out 10 a further follow-up because I think this is an important 11 point.
I didn't get finished with my writing here; if I 12 could orally finish the question here or write it --
O 13 JUDGE CARPENTER:
If you wish to state it, that's 14 fine.
15 MR. STIER:
Before he does that, Judge, I would 16 like to make a comment on the last point.
I must have 17 misunderstood the point that was made before.
I had assumed 18 that the question was whether perturbations in the makeup 19 indicated by the strip chart trace would, in itself, ind'.cate 20 nonsteady state conditions.
I agree completely with what you 21 just said.
In fact, the operators testified that they relied 22 very heavily on observing the makeup strip chart trace.
That 23 was a very important source of information for them and I 24 agree completely that, if you see wide swings or a change in
(}
25 slope during the course of the test, it would -- it should ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80433(H5686
i 82124.0 RT 1782 1
prompt one to question whether that test was run during 2
nonsteady state conditions.
3 JUDGE CARPENTER:
That was the point.
That should 4
trigger a question, not lead you to jump to a conclusion.
5 MR. STIER:
Right.
6 JUDGE CARPENTER:
Trigger a question.
7 MR. STIER:
I agree.
8 MR. RUSSELL:
Staff agrees.
9 MR. MC BRIDE:
Judge Carpenter, the question would 10 be this:
Under a different set of hypothetical circumstances 11 that the parameters applicable to the reactor, power level, 12 et cetera, were steady, but that during the course of a test g_
13 one level transmitter was beha,ving erratically, would it 14 necessarily follow, to an operator, that he was not at steady 15 state merely by observing fluctuations on the strip chart?
16 MR. STIER:
Was that directed to --
17 MR. MC BRIDE:
To both.
18 MR. STIER:
If you had one level transmitter that 19 was not operating properly, as we've testified, the operator
[
l l
20 should have used the good level transmitter to run the leak 21 rate test, which means that it would have been feeding the 22 computer, which means that the strip chart trace would have 23 been drawn by the malfunctioning level transmitter, which 24 means that the trace would be meaningless to the operator as
()
25 an indication of whether they were at steady state ACE-FEDERAL REPORTERS, INC.
l 202-347 3700 Nationwide Coverage 804336 4 46
82124.0 BRT 1783 1
conditions.
2 JUDGE CARPENTER:
That seems clear.
3 MR. MC BRIDE:
Yes.
4 JUDGE CARPENTER:
Moving on to the next test, NRR l
5
' test 72 --
6 MR. STIER:
Judge, that is our test 86.
7 JUDGE CARPENTER:
Thank you.
Question 394 has 8
been withdrawn.
9 395:
Which deflections on the strip chart 10 indicate that an unstable level transmitter was used during 11 this test?
How do you distinguish deflections on the chart 12 that indicate a possible feed and bleed operation during this
)
13 test?
How do you distinguish deflections on the chart that 14 indicate any water additions at all during this test?
15 MR. CAPRA:
There are no deflections on the strip 16 chart that indicate unstable level transmitter was used 17 during the test, since the level transmitters were not 18 switched during this period.
Except for short switching 19 periods, the good level transmitter was left lined up to the 20 strip chart and the bad level transmitter lined up to the 21 computer between December 17, 1978, and December 26, 1978.
22 However, the difference in readings can be seen by comparing 23 the makeup level trace with the values printed on the 24 surveillance test sheet.
In this case the difference is
()
25 approximately 7 to 8 inches.
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
82124.0 1784 l
1 If you look at the strip chart trace in the NRR 2
report for this test you'll see I have the values printed on 1
3 the surveillance sheet indicated on the strip chart extract.
l 4
That's the line that is -- that runs above the normal makeup 5
level trace, beginning at 86.5 inches and ending at 87.5.
6 That's what the computer was seeing during the leak rate 7
test.
8 With respect to water additions, for approximately 9
one hour before the test and most of the test period, the 10 general slope of the makeup was approximately minus 3.5 11 inches per hour.
Taking into account the changes in T av and 12 pressurizer level at the end of the test, the makeup tank gs t
)
13 level should have decreased approximately 2.4 inches.
14 However, the actual change seen on the strip chart is 0 15 inches.
16 The surveillance test sheet values show a plus 17 1-inch change.
18 As a result of this, it is believed that part of 19 the 800-gallon addition that is logged in the control room 20 log at 0045 may have been added during the test.
In this 21 case, the large additions, which are logged at 0045, at 0130, 22 0150 and 0200, are actually feed and bleed operations.
23 I'm not sure if you can read the printing on the 24 strip chart but what I'm referring to are the arrows with the
()
25 numbers underneath them right after the test period.
Those ACE-FEDERAL REPORTERS, INC.
20:4 47-3700 Nationwide Coverage 800 336-6646
82124.0 BRT 1785 1
are all logged additions.
2 JUDGE CARPENTER:
The reference is clear.
i 3
MR. CAPRA:
Examples of water additions that are 4
not part of feed and bleed operations can be seen on the 5
strip chart at 1746 and 1825.
Both are 300-gallon additions 6
to the makeup.
7 There is another batched addition which is shown 8
on the strip chart as 2030, approximately 420 gal'lons.
9 However, that is not a logged addition.
10 There is one other complication on this particular l
11 test which leads to some question.
If you'll notice, there 12 is an arrow at 2110.
The notation underneath it says, O
13
" power," with an up arrow, "50 percent."
At this time 14 there's a log entry that power level was being increased up 15 to 50 percent.
Power level increases, and depending on the 16 rate of the power level increase, does affect the rate on the 17 strip chart.
While I don't have it on the strip chart.
18 There's another entry on the control room log at 0230, which 19 is right near the end of the strip chart extract.
20 The control room log entry states that power is 21 now steady at 70 percent, so somewhere between 2110 and 0230, 1
22 power level was changed initially from approximately 35 23 percent to 50 percent, and eventur.ily to 70 percent.
24 JUDGE CARPENTER:
I think you have given us a lot
(}
25 of information about this particular record of this test.
I ACE-FEDERAL REPORTERS, INC.
l m.7 7 y.e_m.c_...
82124.0 BRT 1786 j
O 1
want to be sure that three questions were specifically 2
answered.
What deflections in the strip chart indicate that 3
an unstable level transmitter was used during this test?
4 MR. CAPRA:
I think I answered that in that there 5
are no deflections to indicate that because there was not the 6
phenomenon of level transmitter switching.
7 JUDGE CARPENTER:
Yes.
8 MR. CAPRA:
However, if you look above the makeup 9
level trace itself you will see the printed values that I 10 have handwritten in, of the beginning and ending times seen 11 by the computer.
That was the 86.5 inches and the 87.5 12_
inches.
That's what the computer was seeing while the trace
(
)
13 was doing something else.
14 JUDGE CARPENTER:
And that's diagnostic that there 15 was an unstable transmitter being used?
16 MR. RUSSELL:
That, plus looking at the periods 17 where it was switched before, which are off this small 18 extract.
We believe that the unstable transmitter was on the 19 computer for this entire time period, as shown on this small 20 extract.
21 MR. CAPRA:
This was not a period where the level 22 transmitter was tagged "out of service."
However, in 23 reviewing the entire strip chart you can see periods where 24 the level transmitter is switched back and forth momentarily
()
25 and you can still see that it is exhibiting an unusual ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336 6686
t I
82124.0 1
ERT 1787
]
1 characteristic and that can also be seen by the plotting of i
2 the values here during the leak rate test.
3 JUDGE CARPENTER:
Thank you.
4 Moving on to the next individual test, NRR test 5
77.
396 was withdrawn.
397:
Which deflections --
4 6
MR. MC BRIDE:
Excuse me, Judge Carpenter.
We 7
didn't ask Mr. Stier whether he had any observations on the 4
8 prior test and I wondered if you wanted to do that.
9 JUDGE CARPENTER:
Well, I looked at him and --
l 1
10 MR. STIER:
I looked relaxed?
l 11 JUDGE CARPENTER:
Yes.
Do you have any i
l 12 observations, Mr. Stier?
O 13 MR. STIER:
Well, I might direct your attention to 14 figure V-5 of volume IV-A of our report.
That is a figure 15 which graphically depicts MPR's best judgment about the 16 periods during which the bad level transmitter was feeding 17 the computer and indicated on that figure are all of the
'i 18 tests that were run during that period, so that you can tell 19 the period of time during which we think the bad level l
20 transmitter was on the computer and which tests were 21 performed using that bad level transmitter.
l 22 I agree with Mr. Russell's description about how 23 you make that determination.
Clearly, the level transmitters 24 were not operating in the same way because the beginning and 25 ending points of the test don't match, and observing this
(}
l l
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 80 4336-6646
82124.0 T
1788 4
1 strip chart trace and comparing it with other strip chart 2
traces from around that period, MPR believes that the bad 3
level transmitter was continuously on the computer during 4
that entire period.
4 5
JUDGE CARPENTER:
Thank you.
Moving on to i
6 question 397 on NRR test 77.
What's your number, Mr. Stier?
I 7
MR. STIER:
81.
8 JUDGE CARPENTER:
Thank you.
397:
Which 9
deflections on the strip chart indicate that the reactor was 10 not at steady state operation during the test?
Can you 11 distinguish any deflections that indicate that a small water 12 addition occurred in.
)
13 MR. CAPRA:
The pressurizer level decrease of 13.9 14 inches and the makeup level increase of 4.4 inches during the 15 test was the basis for NRR's conclusion of the unstable plant 16 conditions, since normal steady state conditions would result 17 in a decrease in makeup level during the test.
I 18 With respect to the water addition, there are no l
19 apparent deflections to indicate a water addition.
- However, i
i 20 prior to this leak rate test the average slope of the makeup 21 was minus 1.5 inches per hour.
The expected change in makeup t
j.
22 level, as we've described before, is equal to the gross l
23 leakage plus any decreases caused by T av minus decreases 24 caused by pressurizer level changes minus any additione.
In
(])
25 this particular case, the gross leakage during the test is i
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 804336 6M6
82124.0 RT 1789 1
383 pounds mass or minus 1.5 inches per hour.
2 The minus.359 degree drop in T av is equivalent 3
to 345 pounds mass; the minus 13.9 inch drop in pressurizer 4
level is equal to 1679 pounds mass.
Assuming that no 5
additions were made, this equals an expected mass change in 6
the makeup of 951 pounds mass, or plus 3.7 inches.
7 The change in makeup level read by the computer 8
was 4.4 inches and the change in makeup level indicated on 9
the strip chart was approximately 5 inches.
These 10 differences could be caused by a 20- to 40-gallon water 11 addition.
However, the more significant reason for 12 concluding the test was invalid or the use of the unstable
(
13 level transmitter, LT-1 as input to the computer and the 14 unstable level -- I'm sorry -- the unstable plant 15 conditions.
The calculated small water addition does have 16 substantial uncertainty because of the large changes that j
17 transpired in plant parameters during this test.
18 I guess the bottom line is we do feel confident 19 that the unstable level transmitter was used.
We do feel 20 very confident that, in classifying this as unstable plant 21 conditions there is substantial uncertainty whether a water 22 addition took place or not.
That was based on the calculated 23
-- the expected -- the difference in expected makeup level e
i 24 from the calculated plant parameters during the test.
()
25 JUDGE CARPENTER:
Comment, Mr. Stier?
i ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coserage 800-336 6646
i l
82124.0 T
1790 i
1 MR. STIER:
Yes.
We ' agree that the unstable level 2
transmitter was used.
We agree that the -- that the test was 3
run during nonsteady state plant conditions; very significant 4
change in pressurizer level, for example.
We could not find 5
evidence that a water addition had been made.
I would note, 6
if you look at the trace deflections during the period of the 7
test, they are approximately 2.5 inches, peak to peak, which 8
would account for -- or I should say mask any indications of 9
water additions during that period.
10 MR. MC BRIDE:
Judge Carpenter, in light of those 11 answers, we would withdraw question 398.
12 JUDGE CARPENTER:
The next test is NRR test 87.
O 13 MR. STIER:
That's our test 71.
14 JUDGE CARPENTER:
Thank you.
399 was wrong.
l 15 Question 400:
What is the relevance of "a marked 16 change in the slope of the makeup level approximately 15 17 minutes before the end of the test"?
l 18 MR. CAPRA:
For approximately 2-1/2 hours prior to 19 and during this test, the slope of the makeup level was 20 approximately minus 5 inches per hour.
Assuming constant 21 leakage, and taking into account the changes in T av and i
22 changes in pressurizer level, the makeup level should have 23 decreased approximately 5.2 inches during the test.
- Instead, 24 because of the use of the unstable LT-1, the computer saw an
()
25 increase of plus
.2 inches.
In addition, the makeup strip i
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-336A646
I 82124.0 l
1791 RT 1
chart, LT-2 decreased only 4 inches during the test.
2 Approximately 15 minutes before the end of the 3
test the slope of the makeup level shows an upward shift from 4
minus 5 inches per hour to about ninus 2 inches per hour.
5 This effect appears to last only about 20 minutes after the 6
test.
Since there are no logged evolutions to account for 7
this change and based upon the expected change in makeup 8
level, NRR believes this change could have b'een caused by a 9
water addition or hydrogen addition about 15 minutes before 10 the end of the test.
11 It is not clear from looking at the makeup strip 12 chart, if you are looking at the NRR version you can see that
-s
(
)
13 I've drawn a line which extends at 5 inches per hour through 14 the test.
About 15 minutes before the end of the test there 15 is a marked change in the slope.
After the test, there is 16 also a chango the slope.
It is not clear whether --
17 exactly whe re t'lat change in slope took place on a permanent 18 l basis.
We believe that the slope change which started 15 I
19 minutes before tht. end of the test -- the trace appears to 20 l come back to almost meet the 5-inch per hour line, about 10 21 minutes after the test; and tha c the actual slope change may I
l have appeared after that test, tae permanent slope change.
22 23 j
JUDGE CARPENTER:
C o.a m e n t, Mr. S ter?
24 MR. STIER:
Yes.
I think we are dealing with two f~')
25 separate issues here.
One is, what does the strip chart xs j
l ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserag 23364M6
82124.0 T
1792 1
trace indicate?
And the second is, what might have affected 2
the test result?
3 on the strip chart trace, we are not certain what 4
caused a change in the slope.
We do know that this was a 5
time of high actual unidentified leakage.
Our report
'6 discusses the leakage during this period quite extensively.
7 Whether any efforts that were made during that time to 8
identify leakage had any effect on this slope, we just don't 9
know.
But I don't think that that's relevant to what 10 happened during the course of the leak rate tests.
The 11 reason I say that is because we believe that this test was 12 run using the bad makeup level transmitter and, if'that's the c
(~'s
)
13 case, the result -- which I agree shows a difference between 14 recalculated total leakage and estimated total leakage of 15 approximately 2 gpm, which should trigger some question --
16 that's meaningless because the result that you got was simply 17 the random effect of running a test on a bad level 18 transmitter.
We discussed this at great length.
It improved 19 the odds of estting a good test result, having nothing to do 20 with the reality of the conditions under which the test was 21 being run.
22 MR. MC BRIDE:
Judge Carpenter, in light of that 23 testimony, we would withdraw question 401.
24 JUDGE CARPENTER:
Thank you.
()
25 JUDGE KELLEY:
We'll take a 10-minute break.
ACE-FEDERAL REPORTERS, INC.
m.m..
s.s _ m. c _,.
yum j
82124.0 RT 1793 1
(Reces s. )
2 JUDGE KELLEY:
We can resume.
Judge Bright will 3
take up where we left off.
4 JUDGE BRIGHT:
Next in line is NRR test 92.
5 MR. STIER:
Our test 66, Judge.
6 JUDGE BRIGHT:
Question 402 has been scrubbed --
7 withdrawn.
8 403:
The annotation on your strip chart 9
confidently states, " water addition of 150 gallons at 0218,"
i 10
'yet you agree with Mr. Stier that the trace at 0218 is not 11 characteristic of a batch addition.
On what basis do you 12 conclude that there was a water addition?
What is the
-s 13 relevance of a 5-inch offset in the strip chart level slope 14 between the beginning and the end of the test?
Which way was 15 it offset?
Did the strip chart always correspond accurately l
16 to the makeup level?
Has the correspondence ever varied 17 inconsistently?
How can you be certain that the strip chart 18 supports your allegation of a water addition during this 19 test?
20 MR. CAPRA:
For several hours during and after 21 leak rate test 92 the slope of the makeup strip chart was 22 minus 8.5 inches per hour.
Assuming constant leakage, and 13 taking into account the changes in T av and pressurizer 24 level, makeup level should have decreased a minus 9.6 inches
()
25 during the test.
However, as a result of using unstable LT-1 ACE-FEDERAL REPORTERS, INC.
202 347-)?00 Nationwide Coverage 8 @ 33 4 % 86
82124.0 T
1794 1
the computer saw a decrease of approximately minus 3.1 2
inches.
3 The strip chart also shows a decrease of only 4
minus 3.5 inches during the test.
Beginning at chart ting 5
0225 there is an increase of approximately 5 inches in the I
6 strip chart.
Following the increase, the slope of the strip 7
chart generally follows the original slope; however, it is 8
offset by approximately 5 inches.
This offset normally J,'
3 9
indicates an addition of water equal to the amount of the 10 offset.
11 NRR agreed with Mr. Stier that the trace is not as l
12 steep as a typical addition; however, assuming a flatter n'
13 slope associated with,no significant change in gross leakage 14
-- that's the average slops of the makeup level trace over a 15 four-hour period rather than the one-hour period used by NRR 16
-- would result in NRR's classification of this event as a 17 possible feed and bleed rather than a batched water 18 addition.
19 There were two other unrecorded changes in reactor 20 coolant system inventory during the watch.
At 0120, it was j
21 180-gallon batched addition, and at 0400 it was a feed and l
r 22 bleed operation.
23 This is another one which I might point out,
[
24 depending upon which strip chart you look at, there is a l ;
()
25 different visual presentation which one can gain, or a ace-FEDERAL REPORTERS, INC.
20M47500 Nttion vide Ceerage
%A)M 6W6 u
I l
l 82124.0 I
RT 1795 1
different perspective one can gain from looking at our l
2 depiction of this leak rate test in the NRR report and the 3
, trace drawn by Mr. Stier.
I think I've answered most of the l
4 questions.
This is a several-part question.
5 It says:
Which way was the offset?
I think if 6
you look at our strip chart we are talking about the offset 7
being in a positive direction, starting at 0118 -- I'm sorry 8
-- 0218, there's a rise of 5 inches.
We believe the slope 9
was offset above that.
10 Did the strip chart always correspond accurately 11
- to makeup level indication?
No.
We have stated that 12 pteviously.
There does not appear to be anything unusual O
13 about this trace which would lead me to believe this was some 14 type of a sporadic offset, or some level tank fluctuation.
15 Has the correspondence ever varied t
16 inconsistently?
I think we have seen that during periods 17 where we nave had the unstable level transmitter on the strip 10 l chart.
You can see that that correspondence is i
19 l inconsistent.
l 20 Also, on this strip chart you'll see plotted above i
21 the makeup level indicatica, above the strip chart 22 indication, there are the points plotted for what the values 23 were that the computer saw during this leak rate test.
24 JUDGE CARPENTER:
Did you draw any inference ot
(])
25 willful intent on the basis of this test?
ACE-FEDERAL REPORTERS, INC.
2024*?1700 Nationwide Coverage 800 33H646
82124.0 1796
}
l I
1 MR. CAPRA:
We categorized this. test as a.
2 questionable test.
We do believe it represents an unlogged 3.
water addition of approximately 150 gallons.
The test was 4
run by Mr. Mell, and the individual who had the panel at the 5
time was also Mr. Mell.
6 We did not draw any conclusions regarding 7
Mr. Nell, since he was not the subject of our investigation.
8 MR. MC BRIDE:
Judge Carpenter, at the time we put 9
these questions we did not have the Stier follow-u,p report of 10 August 29th, the MPR report of August 28th.
We, therefore, 11 did not have the strip chart as clearly available to us as we 12 now do, and I would suggest that the question about the O
13 rectangular portion adjacent to the test area, which is 14 question 404, therefore be deleted.
t 15 JUDGE BRIGHT:
Mr. Stier?
16 MR. STIER:
This is another example where you have 17 to separate the data that you see on the test from the strip 18 chart.
There really is no connection between the two, 19 because this test was run using the bad makeup level 20 transmitter and the result has no connection with the actual 21 conditions that were going on at the time.
At least we think 22 that the odds on its having any relation to plant conditions 23 are very, very remote, because of the way the bad transmitter 24 was behaving.
()
25 This occurred, actually, the day before the level ace-FEDERAL REPORTERS, INC.
3C 347-1700 Nationeide Coserage atu3.WM46
- l1
)l 1
I
=82124.0 I
BRT l
1797 1
transmitter was repayed, aand W neew w lut'er.eWii trea the 2
information avbilable %#: this 53t-vas indeed, run on th?
j I
s 3
bad. level transmitter;.el Chat to T.iocwpi N our tepMt l,
e 4
l where we discuss the heL9.'10t of Fl.a"4 J trscN;at who u'vad l
5 that transmitter at ti'nes phrd, thq keev 4.hp. j t Was behaving l
6 l erratically nevertheless need it t.J.1;un.ledd Agte tette.
l J
7 As far as the 1.raco ld! thy # tI t p Chart h 8
concerned, there is a def!L S W Qe Oct't /;nov precitsly 9
what caused that deflection.
It dow not appea ? to be 10 characteristic of a water addition, vej the slopeu 'ar>d the 11 effect of whatever happened on the sicp.es depends 4ft how you 12 j draw lines on the strip chart to try to ceso*tre the rate of
( -
13 change in makeup level during this,perisp $1 t i.t<3,
l 14 If you look at our strip eka;rt VDu*ll anc that MPR l
15 has drawn lines that suggest that there way a rither charp 5
16 drop in makeup level of perhaps 2.5 inctes, and M en a rather i
17 steep rise but not as steep and not as straight or you 5.ould 18 find normally with a water addition, which peako at nbatt 75 19 or 76 inches; and then another sharp drop at ahcut the sine 20 i rate as before, e
7 21 When that is completod, when thet drop ie g
\\
22 completed the trace begins to level off and r69unes rougnly l.*
i 23
! the came rate of decline that it had shown before that up xvi 24 down trace began.
]
25 l
It is possible that that trace could have tenst e
ACE-FEDERAL REPORTERS, INC.
~
.T 347-)h D Nanen*ide Coverage 80MWes6
_ _- ~
i 82124.0 f"7T-l 1798
'k_)
I caused by some sort of a feed and bleed operation.
Small 1
2 bleed, larger feed, and then a large biced.
And if you look l
l 3
at the overall rate of decline in makeup level throughout 4
that entire period you'll see that, according to our 5
estimate, it. remains roughly the same.
6 Whether that hypothesis is correct or some other 7
j hypothesis, we simply don't know.
The point is that we oniy i
8 used this test for the purpose of our discussion of the use 9
of a faulty makeup level transmitter.
10 JUDGE BRIGHT:
You say this was the day before the 11 transmitter was fixed?
12 MR. STIER:
I ought to check the precise date in s
l 13 l my report.
As I remember, it was repaired on January lith.
14 We discussed that in a section of our report.
Let 15 me refer you to it.
That is discussed beginning on page 79 16 of volume I of our report; that is, the whole period of time 17 during which the bad makeup level transmitter was used and we 18 discussed the work requests and when we believe it was 19 L actually repaired.
As I indicated, January lith is my I
20 recollection of that date.
21 l
JUDGE BRIGHT:
Can we assume with some confidence 22 that the bad transmitter was tagged out?
l-23 MR. CAPRA:
I was just going to mention that.
I 24 There is some question about that.
While the work request
()
25 itself was signed off as being completed and the inst.rument ACE-FEDERAL REPORTERS, INC.
202 347 3700 Ntionwide Coverage 80 N % 6646
I 82124.0 RT 1799 i
'l returned to normal service on the lith of January, the 2
out-of-service tag itself was removed on January 10th, which 3
is the date of this leak rate test.
4 I
There is a place for the shift foreman or shift 5
supervisor to indicate the time as well as the date that the 6
tag was removed in the index.
In this particular case it is 7
either not indicated or we couldn't find it.
There is an 8
entry that shows it was removed on January 10th.
9 However, this particular leak rate test was run on 10 the mid-watch between 2300 and 0700.
Normally, instrument 11 technicians do not work on the mid-shift.
They normally work 12 during the day chift, so, if I would have to makg a -- hold 7-13 on just a second.
i 14 M3. STIER:
While he's looking for that, Judge, I 15 think one very strong indication that it was still 16 malfunctioning at this time is where the beginning and ending 17 readings on the comritter printout show up when you put them 18 on the strip chart.
Clearly there's a vast difference 19 between the levels indicated on the test and the level 20 indicated on the strip cliert, which is one very strong
'21 indication that the level transmitters were not reading the 22 same.
23 JUDGE BRIGHT:
Well, it is true that if they want 24 to do a leak rate test using the computer and get an accurate
()
25 one, that they have to use the good transmitter on the ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationside Coverage 2336e46
I 82124.0 g yT I
1800 v
1 computer.
2_
MR. STIER:
The good one.
Yes.
And we believe 3
the good one was on --
4 JUDGE BRIGHT:
So you automatically say the strip 5
chart is on the bad one or there's nothing on the strip chart 6
at all; is that right?
-7 MR. STIER:
No.
We believe this is the. trace made 8
by the good level transmitter.
When you compare this trace 9
with the other traces you can see a significant difference.
10 This is the trace of the good one.
11 If the good one is en the strip chart the bad one 12 has to be on the computer.
If the bad one is on the 13 computer, you ca,n see the difference between the two sets of 14 readings.
15 The test should have been run using the level 16 transmitter that is shown on the strip chart in this 17 particular case.
They should have switched that to the 18 computer to run the leak rate test; and, indeed, there are a 19 number of instances during this time period that, for a 20 one-hour period, the bad level transmitter is switched to the 21 strip chart and you can see the erratic trace.
The good one 22 is on the computer for the purpose of running the test and 23 then after the test is done they switch it back.
24 This is an instance in which the good level
()
25 transmitter was left on the strip chart; therefore the bad ACE-FEDERAL REPORTERS, INC.
202 3 4 3700 Nationede Coversy m))MM6
1 e
i 82124.0
}
7'ET 1801
'w]
k l
1 i one was the one used to run the test.
2 MR. CAPRA:
I may have some additional evitience 3
here on when the tag was cleared.
4 l
As I indicated from the out-of-service log, I
5 January 10th is the date indicated that the tag was removed.
6 The initial, under the " shift supervisor / shift foreman"
-I 7
column is "CA," which I believe stands for Chuck Adams.
8 Chuck Adams was on watch on the 1500 to 2300 watch that day, 9
.which is the swing shift, which is after the time pericd of r
10 this leak rate test.
11 JUDGE BRIGHT:
Let me see if I have this l
l straight.
There's something here that just doesn't -- there I
rN 12 L.;
13 are two channels from the makeup which, supposedly either one 14 l of them, if everything is right in the world, they both will I
15 give you the same indication.
And they are checking one 16 another, which is why you have two of them.
l 17 l
MR. CAPRA:
Yes, sir.
I l
JUDGE BRIGHT:
One of them goes bad and is " tagged 18 I
19 out."
l 20 MR. CAPRA:
Yes, sir.
I 21 JUDGE BRIGHT:
Now, even if it is, what is the 22 significance, then, of that tag out?
Does that mean that it 23 l can't be used for the computer thing?
Or that it can't be 24 used for the strip chart?
(q
_j 25 MR. CAPRA:
What it is is --
ACE-FEDERAL REPORTERS, INC.
M 34*.3'00 Nationwhje Cosmge 800-13/A646 5
I 82124.0 BRT 1602 a
1 JUDGE BRIGHT:
Which takes precedence, if you are
+
2 doing a leak rate test?
i 3
MR. CAPRA:
It's essentially a caution tag or 4
warning to the operator not t rely on tha,t particular l
5 instrument, the output.from that particular instrument.
So 6
common sense would dictate that, during periods you were not 7
running a leak rate test, since makeup level was used by r
8 operators to monitor-t?.e plant, you would normally leave the i
9 sele'ctor switch in the control room lined up to LT-2, meaning 10 that the makeup strip chart seen by the operators would --
11 the trace that was presented would be from the good level 12 transmitter.
()
13 However,.if you are going to run a surveillance 14 test which requires the use of that instrument to demonstrate 8
15 compliance with the technical specifications, it is necessary 16 to provide reliable and accurate information as an input for i
j 17 the survelliance test, in which case it would be necessary i
18 for the operator to switch level t'ransmitters to provide 19 input to the computer from the good level transmitter, at l
20 least during the data collection period.
I 21 JUDGE BRIGHT:
Is this a part of the procedure?
i 22 MR. RUSSELL:
Yes.
If I could I would like to t
23 supplement this.
It is not just common sense, it's a l
24 requirement of the procedures.
i 1
25 Administrative procedure 1036, which we've
()
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationaide Coverage 800 336-6646
3
~ - -
~.
?
- 02124.0
- BRT.
i 1803
~t i
f 1
M 1
Nstified to last week, Three Mile Is1'and liuclear Station 2
administrative procedure 1036, instrument out-of-service 7
3 Icontrol.
This is Revision 0, which was issued March 1, 4
1978.
5 The purpose of the. procedure is to describe tho 6
method c.f' control of readout devices which become inoperable.
7 or are strongly suspected of being inoperable, such that they 8
are marked, documented and controlled until repair is 9
effected.
That's under section 1.1; purpose.
10 Under section 2.,1, responsibilities:
" Shift 11 supervisors, shift foremen and control room operators.
The 12-shift supervisor and/or shift foreman and/or control room, 13 operators are rerponsible to assure out-of-service 14 instruments are identified and logged "out of service."
They 15 are also responsible to assure that work requests are 16 rubmitted to effect repair and remove the out-of-service t
17 stickers from the instruments and close out the V
i 18
'out-of-servico instrument log entries."
19 This'is the procedure which promulgates the log
,_i 20 that we have been discussing, and I think the procedure is 21 quite clear and it indicates that out-of-service instruments 22 and suspected instruments are not to be used and they are to 23 be tagged and indicated and those tags are to remain on them 24 until such time as repairs are effected and they are again 25 reliable for the purpose of operating the facility.
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage WW
82124.0 BRT 1804 1
MR. CAPRA:
I believe Mr. Russell may have given 2
you the date for Revision 0 when he said March 1, 1978.
3 Revision 1 was issued August 23, 1978.
They are essentially 4
the same procedure.
The only difference is the modification 5
to the actual index for the out-of-service log itself.
6 MR. RUSSELL:
The cover sheet shows pages 1 7
through 4 of Revision 0 were still in effect.
It is only 8
page 5 which was affected by revision 1 on August 23,.1978.
9 So it's a page-change process to the procedure and the page I 10 read from was unchanged from Revision O.
11 JUDGE BRIGHT:
Mr. Stier?
12 MR. STIER:
I would like to add a little something
()
13 more to that.
Aside from the tagging procedure I think 14 there's a more fundamental requirement that leads to the 15 conclusion that you cannot accept leak rate tests run using 16 that faulty makeup level transmitter.
If you ge back to the s
17 technical specifications that establish the purpose for this 18 surveillance and the procedure that describes its purpose, it 19 is to document actual leakage.
You can't measure actual l
20 leakage because you don't have an instrument capable of doing 21 that; you can't document it, therefore you can't effectively 22 run the surveillance test and an exception has got to be l
23 filed in order to assure that the instrument necessary to run l
l 24 the surveillance test is repaired in order to comply with the l
25 requirements of the technical specifications, ACE-FEDERAL REPORTERS, INC.
i
e
~e*ema m-..
82124.0 BRT 1805
[')
\\/
1 I mean, within the procedural scheme at TMI-2, 2
this kind of situation was contemplated and there are very 3
specific requirements for how this kind of an issue is 4
supposed to be dealt with.
You can't run a surveillance 5
test.
There are procedures that tell you what it is that you 6
are supposed to do and who is supposed to make decisions 7
about that and how youiare supposed to then repair the 8
equipment and document new test results.
And that was not
~
9 done here.
10 JUDGE BRIGHT:
Well, then it seems to me that what 11 we have here is a system where you have to have -- if you are 12 going to do leak rate tests you must have two instrument
()
13 channels.
14 MR. STIER:
No.
I don't believe so.
15 JUDGE BRIGHT:
I'm not talking about hand 16 calculations.
I'm talking about the way it seems to be done 17 here.
The computer needs one input and you also need the 18 strip chart input; is that true?
19 MR. RUSSELL:
No.
The technical specifications j
20 would allow continued operation of the facility with only one 21 makeup level transmitter operable.
The use of the computer 22 with it automatically selecting the instrument which is not 23 lined up to the strip chart was a part of the design of how 24 they operated.
It is possible to switch to the computer for 25 the data collection period, switch back to the makeup strip ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336 6646
82124.0 BRT 1806 O
1 chart in between the data collection periods, and then switch 2
back to the computer for.the final set of data.
3 It is also possible to taka and run a hand 4
calculation by taking data from the patch panel or by using 5
the data logging system of the computer, if you wish to leave 6
the level transmitter switch with the good one lined up to 7
the computer.
8 So that there were other options.
It was not'a 9
requirement that you have two instruments.
It was necessary 10 to have two instruments to run it the way they did on 11 automatic, with just punching in "RCSL, carriage return," and 12 starting the test.
()
13 MR. MC BRIDE:
Judge Bright?
One follow-up on the 14 line of questioning you just proceeded with.
(Handing.)
15 JUDGE BRIGHT:
The question is:
During the period 16 in which the good level transmitter was being read by the 17 computer and the bad level transmitter is being read on the 18 strip chart, would it be reasonable for an operator to switch 19 the level transmitter during the test to see the accurate 20 makeup level on the strip chart?
21 MR. RUSSELL:
Yes.
22 MR. STIER:
I agree.
With the understanding that 23 it is switched back for the final readings.
24 JUDGE BRIGHT:
Is that laid out explicitly in the 25 actual procedures?
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646
82124.0 BRT 1807 O
1 MR. RUSSELL:
The actual procedure does specify 2
that the instrumentation that's used for the initial set of 3
data readings is the instrument used for the final data 4
readings.
It doesn't specify using one or the other; it just 5
says make them the same.
6 JUDGE BRIGHT:
It doesn't tell them what to do if 7
one goes bad?
8 MR. RUSSELL:
Administrative procedure 1036 tells 9
them to tag it out and gets it repaired before it's used 10 again and the tag is removed, so there are two procedures 11 that cover that.
12 JUDGE BRIGHT:
NRR test 125.
()
13 MR. STIER:
That is our test number 33.
14 JUDGE BRIGHT:
Question 405:
You agree with 15 Mr. Stier that a hydrogen addition was possible during this 16 test.
Why did you not note that independently?
17 MR. CAPRA:
This was noted.
NRR classified this i
18 as a possible water or hydrogen addition.
Since it was not 19 logged in the control room log and the trace was somewhat i
20 different than the response to~ leak rate test 124 -- that is, i
21 the response was transitory -- NRR classified it as involving 22 either possibility.
23 However, based upon Mr. Chwastyk's turnover notes 24 of February 19, 1978, which NRR did not have at the time, 25 Mr. Stier's classification is more probable.
ACE-FEDERAL REPORTERS, INC.
=w ae-v? c-?m m
82124.0 BRT 1806 i
/~T kl 1
chart in between the data collection periods, and then switch 2
back to the computer for the final set of data.
3 It is also possible to take and run a hand 4
calculation by taking data from the patch panel or by using 5
the data logging system of the computer, if you wish to leave 6
the level transmitter switch with the good one lined up to 7
the computer.
8 So that there were other options.
It was not a 9
requirement that you have two instruments.
It was necessary 10 to have two instruments to run it the way they did on 11 automatic, with just punching in "RCSL, carriage return," and 12 starting the test.
()
13 MR. MC BRIDE:
Judge Bright?
One follow-up on the 14 line of questioning you just proceeded with.
(Handing.).
15 JUDGE BRIGHT:
The question is:
During the period 16 in which the good level transmitter was being read by the 17 computer and the bad level transmitter is being read on the
~
18 strip chart, would it be reasonable for an operator to switch 19 the level transmitter during the test to see the accurate 20 makeup level on the strip chart?
21 MR. RUSSELL:
Yes.
22 MR. STIER:
I agree.
With the understanding that 23 it is switched back for the final readings.
24 JUDGE BRIGHT:
Is that laid out explicitly in the 25 actual procedures?
l ACE-FEDERAL REPORTERS, INC.
w2 3 2 3:oo s.1... c...,...
.co.33 e
D 82124.0 BRT 1807 i
O 1
MR. RUSSELL:
The actual procedure does specify 2
that the instrumentation that's used for the initial set of 3
data readings is the instrument used for the final data 4
readings.
It doesn't specify using one or the other; it just 5
says make them the same.
6 JUDGE BRIGHT:
It doesn't tell them what to do if 7
one goes bad?
8 MR. RUSSELL:
Administrative procedure 1036 tells 9
them to tag it out and gets it repaired before it's used 10 again and the tag is removed, so there are two procedures 11 that cover that.
12 JUDGE BRIGHT:
NRR test 125.
()
13 MR. STIER:
That is our test number 33, 14 JUDGE BRIGHT:
Question 405:
You agree with 15 Mr. Stier that a hydrogen addition was possible during this 16 test.
Why did you not note that independently?
17 MR. CAPRA:
This was noted.
NRR classified this 18 as a possible water or hydrogen addition.
Since it was not 19 logged in the control room log and the trace was somewhat 20 different than the response to leak rate test 124 -- that is, e
21 the response was transitory -- NRR classified it as involving 22 either possibility.
23 However, based upon Mr. Chwastyk's turnover notes 24 of February 19, 1978, which NRR did not have at the time, 25 Mr. Stier's classification is more probable.
ACE-FEDERAL REPORTERS, INC.
r:
msm -
mm-. -
82124.0 BRT 1808
/
1 I believe this test will probably be discussed in 2
more detail by Mr. Stier, so that's the Staff response for 3
this particular question.
4 JUDGE BRIGHT:
Mr. Stier?
5 MR. STIER:
Yes, Judge.
This trace was identified 6
as a result of examining a number of separate pieces of 7
evidence.
First of all, when MPR first examined this strip 8
chart, the trace did not appear to be so characteristic of a 9
hydrogen addition that they could say that:
Yes, they were 10 certain that it was.
11 During the course of our investigation we obtained 12 a very early interview with Joseph Chwastyk, who.was a shift
()
13 supervisor.
In questioning Chwastyk about his knowledge of 14 the effect of hydrogen on makeup level indication, he 15 indicated that, yes, he was aware that it could have an 16 effect and that he had learned of it from discussions with a 17 CRO on his shift.
He was not certain of the identity of the 18 CRO, but believed that it was an individual by the name of 19 Illjes.
20 Chwastyk said that when it was brought to his 21 attention he ran an experiment to determine whether it had an 22 effect and, indeed, observed that it had an effect on makeup 23 level indication.
He couldn't remember when it happened but 24 that was a rough outline of what he described to us.
25 We then went back and checked all the evidence O
l ACE-FEDERAL REPORTERS, INC.
202 Nationwide Coverage 800-336-6M6
.. 347-3700
=_ _
i 82124.0 BRT 1809 s
O 1
that we could think of to try to pinpoint when that 2
occurred.
In checking his shift turnover notes we found a 3
rather cryptic notation.
Let me read it to you so that we 4
are absolutely certain about what we are talking about.
5 The note says -- and the note is written for the 6
February 19th shift turnover:
"At 60-inch level in MUT and 7
5-6 psig H2 over pressure we get good LR."
That means to me 8
that at about a 60-inch level in the makeup an increase in 9
makeup over pressure by approximately 5 to 6 pounds pressure 10 resulted in obtaining a good leak rate.
11 If you look at the strip chart for February 19th, 12 the only time at which the makeup level is in a position
()
13 where it matches up with that description, roughly 60 inches, 14 is during the time that tests 33 was run.
15 In examining the strip chart trace we see a slight 16 flattening and a curve to that trace that suggests the 17 possibility of a hydrogen addition.
That test was run on 18 Mr. Chwastyk's shift, of course, and Mr. Illjes participated 19 in that.
I can tell you specifically what his role was on 20 that day.
Mr. Illjes was on the panel at that time, and of 21 course as we have indicated, that is the individual who would 22 be in a position to change hydrogen over pressure.
23 That test was filed very shortly after another 24 successful test and we believe that in that earlier 25 successful test on the same shift the hydrogen addition was ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 3364M6
82124.0 BRT 1810
. -m 1
made.
That's another test in which Mr. Illjes participated.
2 Based on all of that evidence, we inferred that on 3
that day, Illjes performed a test, added hydrogen to the 4
makeup during the first test, got a successful test result.
5 There was some communication with Chwastyk.
The second test
.6 was run as an experiment; another successful test was 7
obtained during which a hydrogen addition was made, and 8
Chwastyk indicated that fact in his shift turnover note.
9 Now, we have been unable to get back to 10 Mr. Chwastyk to go over all of this in detail with him to try 11 to confirm it.
He has refused, despite very intense efforts 4
12 on our part to convince him to be reinterviewed on this, he
(])
13 has refused to submit to further interview.
So we have not 14 gone back and confronted him with these details.
15 It is based on all of that that I reached the 16 conclusion that the evidence was very, very strong that 17 hydrogen addition was made during the course of test 33.
18 MR. CAPRA:
I might want to point out one 19 additional piece.
.We had also concluded that there was 20 circumstantial evidence that shift B, which this test 21 occurred on, may have been involved in hydrogen additions.
22 That was based on our review of four successive tests 23 performed by this test.
They correspond with test 119 on 24 February 14th, in which Mr. Kidwell was the individual who 25 signed the test and had the panel at the time; there was a ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
82124.0 BRT 1811
%A-I hydrogen addition during this test that was not logged in the 2
control room log but it was logged in the auxiliary 3
operator's log.
4 The next test that was retained that they 5
performed was three days later.
It was NRR test 123.
6 Mr. Illjes was the individual who performe'd the test and also 7
had the panel.
That was another time where hydrogen was 8
added during the test, it was not logged in the control room 9
log but was logged in the auxiliary operator's log.
10
.Then we come to NRR tests 124 and 125, which are 11 the tests that Mr. Stier just described.
12 MR. STIER:
I would also like to add that we used
()
13 the earlier tests.
If you look at the tests from our numbers 14 39, 35, 34 and 33, you will see a series of hydrogen 15 additions which were confirmed and also the hydrogen 16 additions that I have identified in the manner that I'just 17 described.
18 JUDGE KELLEY:
The strip chart trace here during 19 the test time, which you conclude indicates a hydrogen i
20 addition, does this contain some distinctive characteristic 21-that sets it apart from other strip chart traces where the 22 Staff earlier thought it at least might have been hydrogen 23 and you thought you really couldn't make that conclusion?
Is 24 there something different about the strip chart as opposed to 25 notes and statements and other extrinsic evidence, if you ACE-FEDERAL REPORTERS, INC.
- - - m ~
swec, -
. ~ ~
e
i 82124.0 BRT 1812 1
will?
2 MR. STIER:
It is not so different that MPR could 3
look at it and say:
Yes, we are sure that's a hydrogen 4
addition.
It is not that distinctive.
AJ you can see from 5
looking at strip charts when you are talking about a 6
deflection of 2, 2-1/2 inches, we think it is very 7
difficult.
However, when you have the extrinsic evidence and 8
you focus on this particular portion of the strip chart, the 9
trace fits the general characteristics that we've observed in 10 other traces.
So that MPR, in looking at that, would not 11 say:
No, that isn't a hydrogen addition.
12 The general characteristics are there but it is
()
13 not so distinctive that it would jump out at you if you are -
14 examining this particular strip chart.
15 JUDGE BRIGHT:
NRR test 126.
16 MR. STIER:
That's our test 32, Judge.
17 JUDGE BRIGHT:
Question 406:
Is it your position 18 that a water addition is less probable than a hydrogen 19 addition for this test as well as for test NRR test 125?
20 Upon what basis do you reach the conclusion that hydrogen was l
21 added during NRR test 126?
22 MR. CAPRA:
The answer to the first question is 23 yes.
The makeup level trace observed during test 16 is 24 almost the same as seen on test 125.
25 This test was conducted by shift C.
Both nv ACE-FEDERAL REPORTERS, INC.
202 Nationwide Coverage 800-336-6646
...... -347-3700
82124.0 BRT 1813 O
1 Mr. Cooper and Mr. Congdon have admitted using hydrogen 2
during leak rate tests, but deny adding water for the same 3
purpose.
Thus, NRR believes that a hydrogen addition is more 4
likely.
5 JUDGE BRIGHT:
Mr. Stier?
6 MR. STIER:
We concluded that the trace appears to 7
be very much like a hydrogen addition.
We think it --
8 there's a probability of that.
However, I did not use this 9
particular test to draw an inference from because I did not 10 have some independent confirmation.
I did not want to rely 11 totally on a technical opinion that this trace appears to be 12 characteristic of a hydrogen addition.
I felt that there was
()
13 enough subjectivity in that kind of-judgment that I didn't 14 want to base a finding solely on that.
And so in this case, 15 because it was not logged -- there wasn't a hydrogen addition 16 logged at that time -- and because I didn't have some other 17 extrinsic evidence and because it wasn't critical to my 18 overall findings, I did not draw any inferences from this 19 particular test.
20 l
MR. CAPRA:
I might add that in reaching our 21 conclusion, it was not just the observation of the trace l
22 similarities to the previous test.
We also did our 23 calculation of expected makeup level and the calculation, as 24 shown on the strip chart, indicates that there was an 25 approximately 60-gallon difference, or 2-inch difference in O
ACE-FEDERAL REPORTERS, INC.
M-347 3700 Nationwide Coverage _
800-336-6646
82124.0 BRT 1814
,~
I the strip chart from the expected value.
2 MR. STIER:
Maybe I ought to indicate, based on 3
our threshold of 1 gpm difference, because of normal 4
fluctuations in test results, this did not meet our threshold 5
-- the difference between recalculated and estimated total 6
leakage was
.7 gpm, and so it did not reach our threshold.
7 That's not to say that this -- that that indicates that this 8
wasn't a hydrogen addition.
All it indicates is that, based 9
on test results as compared to expected test results, it 10 could be explained by normal variations in the test results.
11 JUDGE KELLEY:
I suppose, Mr. Stier, this isn't 12 anything new in terms of investigating improper conduct, but
( )'
13 this is a case where there's an indication of improper 14 conduct in the trace but there's really nothing else because 15 the person who ran the test didn't log it and, I gather, 16 didn't admit improper performance of the test?
17 MR. STIER:
Well, we -- yes.
18 JUDGE KELLEY:
I don't mean to reflect on this 19 particular case, t'It in a sense it's sort of an example how 20 crime pays.
If you don't log the thing you might get away 21 with it; isn't that right?
Or is it not right?
22 MR. STIER:
Well, there are ways in which 23 manipulation of leak rate tests could have been --
24 JUDGE KELLEY:
I don't make any reference to this 25 test or any person.
Just, in general, if you are ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80433MM6
82124.0 BRT.
1815 1
investigating, you say I'm not going to draw any adverse 2
conclusions unless I can corroborate my conclusions by a 3
strip chart, then if there isn't any such evidence, there's 4
really no place to go with it.
5 MR. STIER:
I think that that's true.
- And, 6
clearly, a creative person, intending to manipulate test 7
results, could have done so and left very little, if any 8
trace behind.
Very little if any evidence.
9 The whole process here was to take fragments of 10 information and try to pull them together and reach some 11 conclusions.
12 There are many unanswered questions.
There are
()
13 many tests where you have suspicions but your suspicions 14 can't take you any further because the evidence simply won't 15 support the kinds of inferences that we were looking for, 16 And I think you run into problems where you try to make your 17 subjective judgments fit patterns that you expect to find.
I 18 think the technical analysis has got to have its own weight 19 in order to satisfy me, anyway, that it is legitimate to draw 20 inferences from it.
I 21 JUDGE KELLEY:
Thank you.
That's the point I I
22 wanted to make.
23 MR. CAPRA:
I think I mentioned this when we were 24 describing our conclusions on this particular one.
Not only 25 did we believe that the technical analysis showed that this ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coserage 800-336-6686
82124.0
~BRT 1816 28
(%-)
1 situation was quite possible, but the test was performed by 2
the shift who -- the two operators indicated had, in fact, 3
added hydrogen during leak rate tests.
4 JUDGE BRIGHT:
Follow-up question:
Is it a fact 5
that the shift that performed this test, NRR test 126, logged 6
hydrogen more frequently than any other shift?
7 MR. CAPRA:
That's correct.
8 MR. STIER:
I would agree.
9 JUDGE BRIGHT:
NRR test 132.
10 MR. STIER:
That's our test 26.
11 JUDGE BRIGHT:
Question 407:
The strip chart time 12 error used by Mr. Stier for NRR test 132 differs by 15
()
13 minutes from that used by you.
In light of this discrepancy, 14 can you determine conclusively that NRR test 132 was begun 15 during a feed and bleed operation?
16 MR. CAPRA:
This particular test is the first test 17 on a new strip chart roll.
The strip chart timing is 18 somewhat in question, since the error is large and increasing 19 with time.
However, the trace which NRR believes is a feed 20 and bleed operation appears on the test period captured by 21 Mr. Stier as well as NRR.
Thus, the timing error is not a 22 factor here.
23 In response to the next question, I will address
! more specifically why we believe the feed and bleed may have 24 i
i 25 occurred during the test period.
But with respect to the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6 4 6
r 82124.0 BRT 1817 1
timing error, I don't think it is significant.
2 JUDGE BRIGHT:
Mr. Stier?
3 l
MR. STIER:
We would agree that it is possible i
4 that there was an unlogged feed and bleed that took place 5
during the course of this test.
6 I would note that the evidence that we used frem i
7 this test had to do with a logged and accounted-for water 8
addition during the course of the test.
We did not deal with 9
the question of whether there was an unlogged feed and 10 bleed.
11 JUDGE BRIGHT:
408:
Mr. Stier does not discuss 12 the presence of a feed and bleed operation in connection with 3
13 NRR test 132.
What evidence do you have that such an 14 operation occurred?
15 MR. CAPRA:
If one assumes that the gross leakage 16 during the test period is constant and no additions were 17 made, makeup level should decrease approximately 4 inches 18 during the test.
19 However, taking into account the 162-gallon 20 l addition which was made during the test and included in the 21
! calculation, and taking into account changes in T av and 22 pressurizer level, makeup level should have increased 23 approximately 1.6 inches during the test.
However, makeup level actually decreased approximately
.5 inches.
24 25 Based upon a review of the strip chart during the l
I 1
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 33M686
82124.0 BRT 1818 Ih
\\_/
1 test there is a sharp drop in level of 5 inches during the.
2 first five minutes of the test, followed by a slower rise in 3
level of about 3 inches over the next 10 minutes.
There are 4
no log entries to explain this trace.
5 Between 1700 and 1905, power level was reduced 6
from 98 percent power to 90 percent, and then raised back to 7
94 percent, initially, and then to 98 percent again.
During 8
and following these power level changes, boron concentration 9
was being changed.
10 Based upon the above, NRR believed that the test 11 was started during a feed and bleed operation, the net effect 12 being a net reduction in inventory that was not accounted for 13 in the, calculation, thus making the calculated value of.
(}
14 leakage higher than otherwise would have been the case.
15 In other words, we believed during this test that 16 by running the test during the feed and bleed operation the 17 calculated value for unidentified leakage was higher than 18 would have been expected.
So it was not, obviously not done 19 to manipulate the test.
l 20 l
JUDGE BRIGHT:
Mr. Stier?
21 MR. STIER:
We believe that this was a manipulated 22 test and it was manipulated by addition of water late in the 23 test and accounted for in taking advantage of the offset 24 which,-for the day, was about almost 30 percent.
That is the 25 effect of the loop seal conditions.
o
(_-
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
3 s
82124.0 BRT 1819
,-m l
We think that the water addition is quite clearly
(_)
1 I
2 depicted on the strip chart just before the end of the test 3
and those were the significant facts that we were looking 4
at.
5 our calculation of the difference between expected 6
and recalculated leakage do'es not show that much difference.
7 In other words, that water addition really made up for the 8
high unidentified leakage that would have resultqd had it not 9
been for that water addition during the course of that test.
10 We think that accounts for whatever difference there may have 11 been.
12 MR. RUSSELL:
I would like to ad,d that this is a 13 test that was performed by shift D.
There were, I believe, a
{}
14 series of 13 where they had this characteristic of a water 15 addition that was recorded towards the end of the test.
16 Because of our concern about the feed and bleed 17 which may have affected the beginning of the test, we did not 18 use this as one of the 13.
We used 12 out of 13 to establish 19 that pattern for this particular shift.
And because of the 20 questions about the feed and bleed and the effect of it we' 21 did not consider this one in reaching conclusions and include 22 it in the pattern of testing for shift D.
We felt that was 23 conservative.
There was more than sufficient evidence with 24 the other 12.
25 JUDGE KELLEY:
Is this an example of a test where O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwife Coverage 800 336.Ma6
1 82124.0 BRT f
1620
,o 1
there's'some divergence of opinion between Stier and l
2 Russell /Capra, but it goes.in the opposite direction?
That 3
is to say, stier and MPR find it as evidence of manipulation 4
and you do not?
5 MR. CAPRA:
As Mr. Russell explained, this does 6
fit the pattern that we have observed in the other 12 tests 7
regarding test manipulation.
So we agree with Mr. Stier that 8
this test may well have t6en intended toebe manipulated, It 9
was early in the pattern.
It was one of the first.
- However, 10 we believe that if this test was begun during the feed and 11 bleed operation, the effect of the water addition later in 12 the test was not as significant because of the bleed -- the
' ( ')
13 water that was bled out of the system at the beginning of the 14 test.
15 JODGE KELLEY:
I guess what I'm harking back to is 16 a couple of days ago, in attempting to summarize differences 17 between the two studies, I recall that there was some number 18 of tests where the Staff found improprieties or at least 19 questionable tests and the stier/MPR analysis did not; and 20
, conversely, there was some smaller number where the opposite 1
21 was true.
I'm sinply asking for the record whether that's 22
. what we are talking about h(re.
I don't know that it's of I
23 critical importance, but it seems to me all discussions ought 24 somehow to fit what was said before, if that's the case.
I i
25 MR. RUSSELL:
I don't think there's a technical CE) i ACE-FEDERAL REPORTERS, INC.
(
2f2447-U00 N.uwmwide C6verage 80 4 336-6646
I l
1 82124.0 SRT 1821 (h
\\l 1
disagreement on the-test.
The pattern is clearly there, the 2
water addition.toward the end of the test.
That's a pattern 3
that we saw and we feel that that is indicative of a 4
manipulation.
Our reason for not including it in that 5
pattern was because of the effect of the feed and bleed at 6
the beginning of the test.
We felt it was more conservative 7
to leave that test out because we had a very strong pattern, 8
not involving feed and bleeds.
So we have felt the evidence 9
we were basing our conclusion on, on manipulation, was 10
- stronger by leaving it out,than by including it.
11 JUDGE KELLEY:
Okay.
But I'm still not clear 12 where this fits in what I would call the matrix that appears j
()
13 on page 3 of the attachment to your memorandum of September 14 20, where you say that, in 21 cases, NRR finds tests invalid 4
15 and Stier finds them valid; and in six cases Stier finds them 16 invalid and you find them valid.
Maybe that's not what we 17 are talking about here.
That's what came to mind so that's 18 why I asked it.
19 MR. CAPRA:
I think the way we classified this was 20 we both agreed in our terminology that the test was invalid 21 but for slightly different reasons.
That's how it is 22 characterized in the attachments to the September 20th memo.
23 JUDGE KELLEY:
All right.
It is not, in my mind, 24 a complete numbers game; meaningless, that is to say.
25 If one set of investigators consistently reaches ACE-FEDERAL REPORTERS, INC.
202 347 3?00 Naionwide Covenge 8J0 336-6646
82124.0 CRT 1822 l
l
,r \\.
\\~)
1 conclusions on debatable tests one way and the other 2
consistently the other way, that might indicate some kind of 3
leaning or different approach.
If, on the other hand, they 4
break out, as you suggest they do, that sends a somewhat 5
different or implies a somewhat different message, it seems 6
to me.
That's the only reason I was raising the question.
7 Okay.
I 8
MR. STIER:
I would like to add one other thought t
9 to that.
As I have indicated, classification of valid or 10 invalid is not really applicable to our tests.
The four i
11 tests -- we discussed four teste yesterday at great length 12 where the Staff was convinced that jogged water additions had
(~N 13 occurred and I took a somewhat different position on those NJ tests.
That is-not intended to suggest that I don't find 14 15 those tests suspicious or feel that the Staff's approach to 16 analyzing those tests is unreasonable.
It is just that I saw 17 it in a slightly different way.
I don't want to suggest 18 there's a sharp difference between us.
19 JUDGE KELLEY:
I think that came out in the 20 testimony.
I understand that; right.
21 MR. RUSSELL:
We have a similar situation in the 22 four tests we talked about this morning involving Mr. Illjes, 23 which were the hydrogen additions we gave for those tests.
24 Those were not tests that I included in a pattern indicating 25 willful manipulation.
And, in fact, including Mr. Illjes in c^1 m
t ACE-FEDERAL REPORTERS, INC.
202 347 3'00 Nationwide Cover:ge 8043.VN
I<
li x
'l 62124.0 g
BRT.
1823 l
3 es A)
\\
- the group that we felt had not -- may have, but didn't.
And j:
'+'
1 l'
latsometimeyoumaywanttohearwhat my reasoning was for 2
3 l why were -- even though that pattern existed, the Staff did L
4 f not and I did not include him in the category of willful 5
manipulation.
I think it's a fine line, but we have reached l
6
.different judgments, basically on the same evidence.
7 JUDGE KELLEY:
In the light of this discussion, l
8 might one not conclude, having heard the testimony of the i
9 past several days, that what I referred to in the '85 memo on 10 i page 3 and what is captioned the "Overall Summary" and which l
I l
11 gives you a sort of quick breakdown, is, for our purposes, 12 maybe a littleover simplified and that this was intended to
()
13 lgiva 6 very capsule summary to the Commission and not~to J
14 spell oct in gfeat detail approaches, nuances, and the like?
15 MR. RUSSELL:
Yes.
The messags I was intending to 16 fsendwiththat memo was there were not basic differences i
17 between the twc, and based on everything Mr. Stier had done I l
t 18 did not see a reason to go back and revise any of the 19 technical analysis or revise any of the conclusions the Staff 20
! had reached.
21 JUDGE KELLEY:
Thank you.
I 22 l
JUDGE BRIGHT; NRR test 141.
I 23 l
MR. STIER:
That's our test 17.
I 24 JUDGE BRIGHT:
Question 409:
Mr. Stier has no 25 discussion of any feed and biced operation in aannection with
,.(l) 1 ACE-FEDERAL REPORTERS, INC.
202-347-3M)
Nationaide Coverase 80tMJ66646
.... - ~.
4
-62124.6 ERT 1824 4
I i
~ /
t 1
NRR test 141.
On what basis do you conclude that NRR test 2
141 was begun during a. feed and bleed operation?
r 3
I MR. CAPRA:
During the test there is a logged feed 4
'and bleed operation of 200 gallons that was logged in the control room log four minutes before the beginning of the 5
i I
6 test.
7 As can be seen on the strip chart, approximately 8
100 gal 10ns of DI water was added initially to the makeup and 9
the remainder fed aid bled over the next 10 minutes at a I
10
{ fairly constant rate.
i 11 NRR believes this feed and bleed operation 12 accounts for the flattening of the traca during the initial
()
13 part of the test.
We do believe that this is, in fact, one 14 of the tests that fits into the pattern of-manipulated tests, 15 however, based on the water addition made by shift D during 16 the last three minutes of the test.
17 JUDGE BRIGHT:
Mr. Stier?
18 NR. STIER:
Yes.
We did not consider the feed and 19 bleed, if it did occur, to be significant.
As you can see, f it did not leave a very distinctive trace on the strip chart 20 l if, in fact, it happened.
21 We did use this test, however, to infer that a 22 23 water addition had been made at the end of the test with the 24 intent to mainipulate the test result, and I think that the 25 evidence of that is very graphic on this strip chart.
(1) l l
ACE-FEDERAL REPORTERS, $NC.
L_
202-347 37J)
Nationwide Coverage _
80433M646
82124.0 BRT-1825 1
MR. MC BRIDE:
Judge Bright, we would withdraw 2
question 410.
3 JUDGE BRIGHT:
Withdraw 410?
Very well.
4 Then, NRR test 147.
5 MR. STIER:
That's our test 11.
6 JUDGE BRIGHT:
Question 411:
Does an unlogged 7
water addition toward the end of a leak rate test necessarily 8
evince deliberate manipulation?
9 MR. CAPRA:
Not necessarily in this test.
The 10 operator conducting this test raade two errors in that he 11 included the 500-gallon additions that he made during the 12 test as an operator change to the reactor coolant drain tank-()
13 and did not include the quantity of water pumped from the 14 reactor coolant drain tank in the calculation.
15 These data entry errors do not necessarily 16 indicate manipulation.
However, the results of the test 17 indicating minus.678 gallons --
~
18 MR. RUSSELL:
6.78.
19 MR. CAPRA:
I'm sorry -- minus 6.78 gallons per I
! minute gross leakage is impossible.
The fact that these 20 i
21 i obvious errors were accepted and approved confirms that 22 operators were keeping tests with less than 1 gallon per 23 minute unidentified leakage without any substantive review.
JUDGE BRIGHT:
Mr Stier?
24 25 MR. STIER:
I would agree with that.
We used it (J
l i
ACE-FEDERAL REPORTERS, INC.
202-347-r00 Nationwide Coverage 800-33MM6
82124.0 BRT 1826
(^/')
1 for the same purpose.
2 JUDGE BRIGHT:
Well, that concludes the questions 3
l to Mr. Russell and Capra.
On the next page -- we discussed I
4 these yesterday.
5 JUDGE KELLEY:
Off the record for a moment.
6 (Discussion off the record.)
7 JUDGE KELLEY:
Why don't we take a coffee break 8
and have an off-the-record discussion of where we are.
9 (Recess. )
10 JUDGE KELLEY:
We are back on the record.
Just to 11 state for the record the questions we are going to turn.to 12 now, we have been discussing questions about particular f')T 13 tests, addressed to the NRR witnesses.
Mr. Stier has, as we x
14 have gone along, commented on those same questions.
There is 15 a section of questions addressed to Mr. Stier that have not 16 yet been discussed.and they range from Stier question 328 to 17 i Stier question 342.
Our intention is to follow the same 18 format, namely putting the question to Mr. Stier and then 19 giving the Staff the opportunity to comment on these tests.
20 The questions to Stier in the prefiled questions 21 following 342 relate to tests that have already been 22-discussed, so we don't propose to ask those questions.
I'll 23 start with Stier test 328, and could the Staff state what 24 test that is?
l 25 MR. CAPRA:
Yes, test NRR 26.
l (1)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-33MM6
1 1
82124.0 BRT 1827 O,
l 1
JUDGE KELLEY:
328:
What evidence do you have 2
that the change in pressurizer level during Stier test 132 3
invalidated that leak rate test?
4 MR. STIER:
Judge, I could be wrong but I don't 5
believe that we said that the change in pressurizer level 6
invalidated this particular test.
I believe that there was a 7
substantisi pressurizer level change, a difference of 7.338 8
inches.
We noted that in our description of that test in 9
table 1, in volume III-A, but I did not draw an inference 10 that this was a test t. hat was run during steady state 11 conditions.
Maybe I should describe what I mean by that.
I 12 only made a finding that a particular test was run on the
()
13 basis of unreliable data when I believed that the 14 unreliability of the data that went into that test was 15 obvious and should have been known to anybody running the 16 test.
17 In this particular test, although there is a 18 rather substantial change in pressurizer level, there is 19 nothing else about the test that I considered so obviously 20 wrong with it that the test should not have been filed by the 21 operator, and for that reason I drew no inferences.
So 22 characterizing it as " invalidated" I don't believe is 23 accurate.
24 JUDGE KELLEY:
Okay.
We'll go to the next 25 question.
329:
Is it likely that the "large change" in ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coscrage 8@33M646
)
82124.0 BRT 1828
~
.r t
1 pressurizer level that occurred during Stier test 132 was the 2
result of deliberate action by the TMI-2 operator?
3 MR. STIER:
I don't know what caused the large 4
change in pressurizer level.
We don't have log entries, to 4
5 my knowledge, that indicate what caused it, so I have no way 6
of knowing whether it was done as the result of some action 7
taken by the operator or was caused by something over which 8
he did not have direct control-.
9 JUDGE KELLEY:
Staff comment?
10 MR. CAPRA:
Yes, sir.
The change in pressurizer 11 level during this particular test was actually 5.3 inches 12 instead of the 7.3 inches.
That may be as a result of us
()
13 having a clearer copy of the strip chart, in this particular 1 44 case.
15 JUDGE KELLEY:
The 7.3 was something -- that was a 16 mistake of yours in reading the chart?
17 MR. CAPRA:
It was not a mistake of ours; no, 18 sir.
19 MR. RUSSELL:
The cover sheet for the test that 20 was performed by MPR indicated the change in level was 7.3 21 inches.
The change in level from our copy of the 22 surveillance cover sheet indicates it's only 5.3 inches.
23 JUDGE KELLEY:
Thank you.
24 MR. CAPRA:
We also did not conclude that this I
25 test was performed during unstable plant conditions, although ACE-FEDERAL REPORTERS, INC.
i 2C2-347-3700 Nationwide Coverage 804336 6M6
82124.0 BRT 1829 (O) 1 those are, regardless of whether it is 5 inches or 7 inches, 2
they are fairly large changes in makeup level and pressurizer i'
3 level that took place during this test.
However, as you can 4
see from observing the trace of the strip chart, it is a very 5
smooth transient that's taking place such that while these 6
large changes were going on, I believe the plant was in 7
essentially equilibrium throughout.
In other words it-was 8
not a fast-moving transient.
9 one of the reasons the procedure cautions 10 operators from varying pressurizer level and T av and rod 11 position during these periods is you are trying to establish 12 equilibrium conditions within the primary.
We believe that
()
13 while these changes may have affected the test, there is 14 nothing to make us believe it was conducted during unstable 15 plant conditions.
16 JUDGE KELLEY:
Thank you.
17 The next test --
18 JUDGE CARPENTER:
Just to be clear, was this 19 change in the pressurizer level accounted for in the 20 calculation of test results?
21 MR. CAPRA:
Yes, it is.
22 MR. STIER:
Judge?
23 JUDGE KELLEY:
Mr. Stier?
24 MR. STIER:
I'm not quite sure what the question 25 is directed to.
It would have been compensated for, in this O
ACE-FEDERAL REPORTERS, INC.
__ Nationwide Coserage __,
800-3{6 6646 202-347-3700
82124.0 BRT 1830 0
1 particularly, by the offsetting changes in makeup level.
I 2
think what Mr. Capra was talking about before is that if you 3
have a transient that occurs rapidly, you may not have a 4
proper offset in plant parameters which would smooth out the 5
re= ult.
In this case, because this took place over time, the 6
reactor coolant system itself made the appropriate 7
compensations and so the readings during the course of the 8
test were not seriously af fected by what occurred --- by the 9
change in pressurizer level.
10 JUDGE CARPENTER:
Thank you.
11 MR. CAPRA:
During this particular test it was a l
12 very small change in T av, of only.05 degrees.
The 5-inch
()
13.
increase in pressurizer level has to come from somewhere, so 14 it comes out of the makeup in this particular case.
15 Normally, the pressurizer level control is in 16 automatic and it is trying to maintain a level of 220 17 inches.
In this particular case there is no log entry to i
l 18 indicate what happened, but it may be that the pressurizer I
19 level control was placed in manual for some particular reason 20 and that's where the difference in water comes from -- the j
21 difference in levels comes from.
I 22 JUDGE KELLEY:
Okay.
Stier. test 90.
Question 23 330:
Please reconcile in your conclusion that the " source of 24 values on manual calculation form DS-1 is not known," with 25 your statement that " manual makeup lvl readings apparently ACE-FEDERAL REPORTERS, INC.
Nationwide Coverage. __ _.. _ _.,.. _ __ _ _ _ _ __ _-3 36-6M6 _ _. _ _ _ _
800 l
._..__ _ _... _.- _.. _. _-3700_, _. _ _ _ _.. _.
l22 347
82124.0 BRT 1831 f3 l
r 1
taken from bad X-mitter" -- meaning transmitter, I take it --
2 "on computer."
3 MR. STIER:
This was a hand calculation that was 4
signed by Mr. Morck.
It is, as the next question indicates, 5
the only reported test.
The question is, is it the only test 6
-- it's the only filed test signed by Morck.
7 Apparently the data was taken on one day and the 8
calculation was done on another day.
There are a lot of 9
uncertainties about this particular test.
We did not know 10 what the source of the values in the hand calculation were.
11 However, our best judgment was that it was taken from the bad 12 level transmitter because we couldn't match up the numbers
()
13 with the strip chart that was -- couldn't match up the 14 numbers with the strip chart, so we inferred that the 15 readings were taken from the bad level transmitter.
But 16 there is, as I say, a lot of uncertainty about where these 17 numbers came from because it was a hand calculation.
18 JUDGE KELLEY:
Okay.
19 MR. CAPRA:
This is NRR test 68.
Initially, when 20 we went through this particular test we did not note the 21 source of the level transmitter information, whether it was 22 from LT-1 or LT-2.
23 After going back and looking at this test we agree 24 with Mr. Stier, there is some question about where the data l
l 25 came from, since there is a difference in days between when CE)
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
82124.0 BRT 1832 O
1 this calculation was made, which was on the 19th, and the 2
data, which came from a period of time on the 18th.
This was 3
during the period where level transmitter number 1 was tagged 4
"out of service."
5 The level transmitter itself that was providing 6
input to the strip chart during this test was LT-2, so that 7
if the data was obtained from the computer, it is highly 8
likely that it was obtained using LT-1.
9 Based on the hand calculation sheet that was used, 10 the fact that it is data sheet number 1, which is for use
~1 when the computer is available -- okay -- we believe that the 12 data may have been taken from LT-1, the unstable one that was
()
13 tagged "out of service."
14 JUDGE KELLEY:
Mr. Morck has been referred to 15 before and he is not a party in this proceeding and, as I 16 understand it -- is that right?
17 MR. MC BRIDS:
That's correc*., Judge Kelley.
18 JUDGE KELLEY:
Nor has he been subpoenaed.
This 19 may be in the record already, but could someone remind me 20 about Mr. Morck's position at this time?
l 21 MR. STIER:
He was, I believe, an engineer 22 assigned to work with the operations department in the 23 control room.
He, to the best of our knowledge, was involved 24 with the decision to correct the calculation error on March 25 16th.
We believe the evidence points to Mr. Morck as the ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
82124.0 BRT 1833
( -)
1 individual who actually made the judgments that ultimately 2
went into that temporary change notice which corrected half 3
of the calculation error and not the other half.
That's one 4
of the things we want to talk to him about.
5 In addition, we see Mr. Morck involved in 6
performing this hand calculation in December and it suggestss 7
to us that he may have had some involvement in examining 8
problems with leak rate testing during the period of TMI-2 9
operation.
That's another thing we wanted to ask him about:
10 Were these problems referred to him?
Has he examined the 11 test?
Was he performing a hand calculation to try to find 12 problems with it?
These are all of the uncertainties about
()
13 him and his role at TMI-2 that we wanted to.tsolve.
14 His serving an engineering function tends to 15 suggest to me that perhaps some of this responsibility fell 16 into his area of jurisdiction.
17 MR. MC BRIDE:
Judge Kelley, Mr. Morck is a client 18 of ours, and to the best of our knowledge and recollection he 19 was not assigned to the operations department, although he 20 may have assisted them in the performance of functions from l
21 time to time on a detailed basis, but rather worked for the 22 Unit 2 superintendent -- technical support, prior to December 23 1,
1978 or thereabouts was Mr. Seelinger; after that date was 24 Mr. Kunder.
25 JUDGE KELLEY:
Seelinger and Kunder are both to be
(^.
U ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6
82124.0
.BRT 1834 f3
\\/
1 l witnesses?
I 2
MR. MC BRIDE:
That's right.
3 JUDGE KELLEY:
I ask because in the course of 4
subpoenaing several people in the late summer that was an 5
outgrowth of our further review of the record and our 6
interest in getting people in who could contribute to j
7 significant issues.
We thought from the record that it 8
appeared, as you all know, that people like Seelinger and 9
Kunder -- I think he was in initially -- but a couple of 10 other people we added in, sort of upper supervisory 11 positions, if you will, that seemed to have pertinent 12 information.
()
13 I guess the question arises now, in view of what 14 Mr. Stier har said, whether we ought not to invite Mr. Morck 15 to come in as well.
16 Let me just ask the Staff if they have comments, 17 if they have knowledge about Mr. Morck, whether they think he 18 might have information on topics of interest?
19 MR. RUSSELL:
I just wanted to make one minor 20 correction.
In our makeup strip chart extract for leak rate 21 test number 69 -- 68, we indicated that Mr. Morck was an 22 off-shift SRO.
That should be an off-shift STE, shift test 23 engineer, not a SRO.
24 The engineers, to the best of my knowledge, were 25 assigned to the technical function as a part of plant staff O
l l
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6 4 6
82124.0-BRT 1835 O
(_/
1 but were not, per se, in the operations department under 2
Mr. Floyd.
But they did assist.
3 JUDGE KELLEY:
Is Mr. Morck a current employee of 4
GPU?
5 MR. MC BRIDE:
No, your Honor, he has not been so 6
employed for, I believe, approximately seven years, give or 7
take a year.
Six years --
4 8
JUDGE KELLEY:
Did you want to make a further 9
comment?
l'0 MR. MC BRIDE:
Yes, I do.
11 JUDGE KELLEY:
I don't mean to be suggesting any 12 conclusion on my part or the Board's part.
I just wanted to 13 get some information.
14 MR. MC BRIDE:
I understand that and I just wanted 15 to say three things.
16 First of all, this is, as Mr. Stier has testified, 17 the only test performed by Mr. Morck, which would eliminate 18 the necessity for question 331, and there are, as the I
19 witnesses have said, a number of uncertainties about it.
20 So far as I know -- they can correct me if I'm 21 wrong -- it is not the basis for any adverse findings about i
22 Mr. Morck or perhaps anyone else, with the possible exception 5
23 i of Mr. Mehler, who came in much later.
1 24 The TCN, the temporary change notice to the 25 procedure in March of 1979 which the record shows Mr. Morck O
1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4336-6646
. _., _ _ _ _. _. ~ ~
M 82124.0 BRT 1836
()
1 was involved in, I don't think is really a major part of the in'uiry in this proceeding.
In other words -- and again, the 2
q 3
investigators can correct me if I'm wrong -- but I don't 4
think that they have formed any conclusion that there was any 5
sort of manipulation of leak rate test results that was 6
intentional or some conduct approaching that level, based on 7
the fact that the TCN was adopted.
8 To be completely candid with you, I want to make 9
sure the Board knows that when the Commission issued a notice 10 of violation to the company in October of 1979, there was a 11 failure -- what was noted was a failure to also correct the 12 procedure for the density differential for the makeup.
So 13 the TCN played a role in that, but I don't think it played a 0
14 role in leak rate falsification.
15 Mr. Morck did not wish to be interviewed by Mr.
16 Stier because, so far as I recall, it has just been a long 17 time, he's not involved in this industry, he's not in the 18 nuclear industry, and these are matters that are just not of 19 any great moment to him any longer.
I really believe based 20 on these very isolated events in which the record shows that 21 he was involved, it really isn't necessary to hear from him, 22 in our judgment.
But of course the Board has to make its own 23 determination about that.
24 JUDGE KELLEY:
Morck worked for Kunder?
l 25 MR. MC BRIDE:
He worked for Seelinger, I believe, O
1 ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800-336-6646
~ ' T J --
1 *'.
82124.0 BRT 1837
()
1 and then Mr. Kunder.
2 JUDGE KELLEY:
Presumably we can get more 3
information, if need be, about Morck's role, in talking with 4
the two of them, who will be witnesses in the case.
5 MR. MC BRIDE:
Yes.
6 JUDGE KELLEY:
I think we can leave it now with 7
the observation that there is some question in our mind about 8
whether we ought to call him.
But we might hear from them 9
and then make a judgment as to whether that le necessary.
10 I would just want to note that it is being raised 11 now.
If we defer any final decision whether to call Morck or 12 not until later in the proceeding, then as his counsel, you 13 should know we have beeh thinking about it.
The deferral O
14 would be without prejudice to our right to call him later and 15 without our being that concerned about a late summons.
16 MR. MC BRIDE:
I understand that, Judge Kelley.
I 17 don't take issue with that in any respect.
I think that's 18 entirely fair.
But let me just observe one other thing for 19 the Board's perspective.
20 Mr. Seelinger was subpoenaed and Mr. Seelinger 21 will be here, despite some personal inconvenience, and 22 professional inconvenience.
Because neither he nor we would 23 assert that he didn't have a significant number of things to 24 contribute to your record.
And Mr. Kunder, of course, is a 25 party, so that's not even an issue.
But with respect to O
ACE-FEDERAL REPORTERS, INC.
N2-347-3700 Nationwide Coverage 800 336-6686
~'
~
~..
82124.0 BRT 1838 (G"T 1
Mr. Morck, I really do believe objectively that a different 2
conclusion is appropriate.
3 JUDGE KELLEY:
You may well be right.
4 MR. STIER:
Judge, if I can correct a 5
misimpression I might have left in what I said before.
I 6
didn't mean to suggest that Mr.-Morck was employed by the 7
operations department.
What I meant to say was that in our 8
judgment he was an important link between the operations 9
department and another level of management.
And because of 10 his connection with leak rate testing in the two respects 11 that I mentioned before we thought it was important to talk 12 to him and see what he knew about it.
13 JUDGE KELLEY:
Fine.
We'll bear that in mind.
O 14 I don't know if we've reached the thi.rd question, 15 332.
I can read it:
Is it possible that Mr. Morck was 16 unfamiliar with leak rate testing and thus could have 17 unwittingly used an unstable level transmitter for Stier test 18 90?
19 MR. STIER:
The answer is yes.
20 JUDGE KELLEY:
Staff?
Comments?
21 MR. CAPRA:
We did not talk with Mr. Morck, but it 22 certainly seems reasonable.
23 JUDGE KELLEY:
Stier test 40, question 333:
NRR 24 used the " peak-to-peak" method to calculate the amount of the 25 unlogged water addition during Stier test 40.
Why did you C)
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800-336-6646
. - ~ -. - -.. _, -
82124.0 BRT 1839
()
1 use the " slope offset" method to make that calculation?
2 MR. STIER:
If I might explain the difference 3
between these two methods, it may become useful.
4 MPR used the slope offset method each time they 5
wanted to estimate the amount of water that had been added to 6-the makeup, based on the strip chart trace.
7 If I can, perhaps, do another drawing, it might 8
help ycu understand what the difference is between these two 9
methods.
10 JUDGE KELLEY:
Okay.
11 MR. STIER:
What I have tried to depict here is a 12 trace on the strip chart.
(Dra wing. )
f 13 I have shown the normal downtrend of the makeup O
14 level with a water addition, which is generally reflected as 15 a sharp straight rise and then a resumption of the normal 16 downtrend of makeup level.
17 What MPR did was, in effect, to extend this line 18 and this line; that is, the angle of' decline in makeup level 19
-- strike that -- the rate of decline at the same level that 20 it had been moving previous to the water addition and the 21 rate of decline and the position of that trace on the strip 22 chart after the water addition.
23 By measuring the vertical distance between these 24 two lines we get the slope offset method of measuring the 25 amount of leakage -- strike that -- the amount of water that O
i
.f 1
ACE-FEDERAL REPORTERS, INC.
1 N2-347-3700 Nationwide Coverage 800 336 6646
~.
82124.0 BRT 1840
()
I was added to the makeup tank during the period in which the 2
sharp rise is indicated on the strip chart.
3 The reason that MPR used this method is because 4
this method takes into account the leakage that is occurring 5
while the water addition is being made.
That would not be 6
shown in this rise, because as the strip chart trace is being 7
drawn there's a certain amount of leakage -- a certain amount 8
of water leaving the makeup.
I don't want to attribute it to 9
leakage.
There is water leaving the makeup and by the time 10 the water addition has ended, the full amount of water that 11 has been added would not be clearly depicted if you simply 12 measured the line indicated by that water addition.
13 So you've got to do the offset in water to 14 determine how much water was really added.
15 The peak-to-peak method --
16 JUDGE KELLEY:
So that the amount is the vertical 17 line?
18 MR. STIER:
The amount is the vertical line.
19 Right.
If you look at the number of inches that is covered 20 by that vertical line you could estimate the amount of water 21 that was added during that time.
22 The peak-to-peak method is a matter of drawing a 23 horizontal line at the beginning of the water addition, a 24 horizontal line at the end of the water addition, and 25 measuring the vertical distance between those two lines.
O ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646 l
___7_.
_ _7
..._,-y_..-
82124.0 BRT 1841
().
1 As you can see, there may be a difference between 2
the two.
That difference is the amount of water that has 3
leaked out of the reactor coolant system -- strike that --
4 that water, that distance is the amount of water that has 5
left the makeup during the course of the water addition.
6 I think with all of those corrections I have 7
described what the difference is between those two methods.
8 The reason that MPR used the method that they did, the slope 9
offset method, was, as I have indicated, to try to get the 10 most accurate measure of the full amount of water that was 11 added during the course of the water addition.
12 JUDGE KELLEY:
Does Staff want to comment on the e
- 13 method?
S s/
14 MR. RUSSELL:
The Staff agrees that the slope 15 offset method as described is technically more accurate.
The 16 reason we used the peak-to-peak was based upon the feedback 17 from operators as to how they used it in reaching Judgments.
18 It became important on one shift, shift A, based upon 19 Mr. Faust's testimony to the Staff, that he would look at the 20 indication from the makeup strip chart and use the change in 21 level times 30 gallons per inch; and in looking at those, 22 compared to what the batch totalizer was showing for the 23 tests involved, they were relatively close to each other.
24 This was during a period of high leakage.
So that the 25 difference that was off, which may have been a result of the O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 4646
_2. _
82124.0 BRT 1842 1
understatement of the effect of a water addition from the 2
problems with the level transmitter, were, in fact, offset.
3 We gave that shift the benefit of the doubt based 4
upon operator testimony, where he said I added water and if 5
the batch totalizer indicated 300 gallons, and he used the 6
peak-to-peak method and he saw it was about 300 gallons, 7
that's what he entered into the addition.
We found that was 8
iconsistent.
Whereas if you used the slope offset method you 9
may show that the amount of the addition was 340 ' gallons, 10 because of the behavior of the makeup level transmitter 11 during that period of time.
12 We don't disagree that the slope offset is more 13 correct.
We just used the method that was available to the operators.
Operators don't check changes in level by using a 14 15 slope offset method.
They just count the number of vertical 16 lines of change and we used that method because that was the 17 practice the operators used.
18 MR. CAPRA:
Let me clarify one thing.
Mr. Russell 19 said we gave the operators on shift A the benefit of the 20 doubt.
We did not treat, in our analysis, the operators on 1
21 shift A any differently than we did for the other operators.
22 We used peak-to-peak in all cases.
It was just that we had 23 some testimony oa shift A that confirmed what we were seeing.
24 JUDGE KELLEY:
Is the ball in your court, 25 Mr. Stier, or not?
(')
c; ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-3364M6
82124.0 BRT 1843
()
1 MR. STIER:
Well, I don't know if there's a 2
pending question.
3 JUDGE KELLEY:
All right.
334.
I'll take that.
4 Is it a fact that, subject to a small error, the 5
amount of the unlogged water addition made during Stier test 6
40 was in fact included in the leak rate test calculation by 7
the TMI-2 operator?
8 MR. STIER:
Yes.
This is a test that was 9
performed at a time when we had some difference in the way 10 the makeup level was reading from the batched water 11 additions.
Therefore, you could infer that a loop seal was 12 in existence during that period of time.
13 In this case, the amount of the addition that was 14 recorded was 200 gallons.
The makeup strip chart' trace 15 measured by the slope offset method shows 231 gallons.
- And, 16 of course, the test would have been affected by what was 17 reflected in the makeup.
This is if the loop seal affected 18 both level transmitters the same, the difference between what 19 was put into the makeup and reported as having been put in 20 the makeup, and having been added to the leakage in the test 21 calculation, and what was reflected in the final reading of 22 the makeup, which would have been affected by the loop seal, 23 that would have given you more of a chance of a better leak 24 rate result.
That is, one under 1 gpm.
25 JUDGE KELLEY:
Staff?
O ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800 336-6646
82124.0 BRT 1844 l
rnj 1
MR. CAPRA:
For the record, I don't know if I 2
identified this test.
The NRR test number is 118, l
Using our peak-to-peak method which we described 3
4 earlier, the water addition appears on the strip chart to be 5
210 gallons.
The amount used in the calculation was 206 6
gallons.
That accounts for why we did not classify this as a 7
manipulated test.
8 MR. MC BRIDE:
Judge Kelley, I wonder if I could 9
request that Mr. Stier and his associates provide a copy for 10
- the record of the essence of the chart that he has just drawn 11 so that the record will be clear what he was saying as he was 12 drawing lines that were depicted visually for the Board.
13 JUDGE KELLEY:
It might be helpful.
Any.
7,
\\
14 objection?
15 MR. GEPHART:
Perhaps he can label it, one line 16 being peak-to-peak and the other one --
17 JUDGE KELLEY:
You could add a label.
I'm sure 18 that --
19 MR. MC BRIDE:
It's a representative thing.
20 Everyone understands he's not depicting actually data, but if 21 it could be labeled " peak-to-peak" or " slope offset" or 22 something of the sort --
23 l
JUDGE KELLEY:
Are you proposing this as a formal 24 exhibit or just a lay-in or what?
25 MR. MC BRIDE:
I'm proposing it as a formal i
e/
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646
82124.0 RT 1845 I
1
. exhibit to accompany his testimony today.
2 JUDGE KELLEY:
Proposed exhibit wherever we are.
3 14?
15?
4 MR. STIER:
It doesn't have to be done by me 5
personally, does it?
6 MR. GEPHART:
No.
I'd prefer not.
7 (Laughter.)
8 MR. BLAKE:
13.
9 JUDGE KELLEY:
13.
There will be an Exhibit 13, 10 which will reproduce essentially what Mr. Stier did on the 11 easel here this morning and there is no objection to that; it 12 will be admitted to illustrate his testimony, essentially.
13 (Exhibit 13 received.)
14 JUDGE KELLEY:
Stier test 4, number 335:
NRR 15 concluded that it was not appropriate to use the " slope 16 offset" method to determine the amount of water added during 17 Stier test 4.
Why did you use that method?
It may be 18 anticipated by your earlier answer but --
19 MR. STIER:
Yes.
It was used because we felt this 20 was a more accurate method of determining the amount of water 21 that was actually added during the course of the test.
22 JUDGE KELLEY:
Follow-on question, 336:
Is it a 23 fact that the two water additions of 100 gallons each were in 24 fact acknowledged in the control room operator's log to have 25 been made during Stier test 4?
O
\\
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336M86
82124.0 i
BRT l
1846 l
I }
l 1
MR. STIER:
Yes.
I think it was actually -- it 2
!was 100, plus two 50-gallon water additions.
I think that's l
3 i what was indicated.
4 JUDGE KELLEY:
Okay.
And 337 on the same test:
5 Is it possible that makeup level instrument errors, of which 6
the TMI-2 operator for Stier test 4 was unaware, contributed 7
to the unidentified leak rate you determined?
8 MR. STIER:
Yes.
That is possible.
9 JUDGE KELLEY:
Staff comnent on that?
10 MR. CAPRA:
This is leak rate test 154 for the 11 Staff.
Perhaps the choice of words of using " slope to slope" 12 was inappropriate.
It's not quite correct.
I think we have 13 indicated by our testimony that the slope offset method is
,7 sx
(
/
14 more technically correct and I think we explained why we used 15 peak-to-peak.
16 In this particular case, there were two additions 17 of approximately 100 gallons.
If you look at the time of the 18 addition, they are fairly quick additions, as opposed to, 19 let's say a 300-gallon addition, which takes a longer period 20 of time.
21 The amount of leakage out during a 100-gallon 22 addition is not as significant.
In other words, the error is 23 not as significant during a 100-gallon addition as it is 24 during a larger addition.
But we do not disagree with 25 Mr. Stier's analysis.
/'3 l
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 853M-44
82124.0 BRT 1847 1
MR. MC BRIDE:
Judge Kelley, excuse me, I haven't 2
had time to write this follow-up question but I wonder if the 3
witnesses would indicate whether they agree that this test 4
was performed during a period of high identified leakage and 5
whether they agree that it might have been necessary for the 6
operator to be adding water during this test.
7 JUDGE KELLEY:
All right.
8 MR. STIER:
Yes.
Maybe I should comment on that a 9
little bit further.
10 During the period of mid-February through the end 11 of March, as we have already testified at some length, there 12 was a very high level of collection in the reactor coolant i
13 drain tank.
That rate of collection -- collection in the 7~s
~'
14 reactor coolant drain tank from leaks from valves at the top 15 of the pressurizer increased as the month of March 16 progressed.
17 It became increasingly necessary for operators 18 during that period of time to add water to the makeup in 19 order to maintain makeup level at the low point.
In fact, as 20 we approached March 28th, water additions were being made 21 very frequently, every hour or perhaps even more frequently 22 than that.
I'd have to look at the strip charts in order to 23 tell you with some precision how frequently they were being 24 made.
25 We examined water additions that were made during o
C
}
,_ /
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coserage 80ME6446
82124.0 BRT 1848
(~)
1
,that time and examined the behavior of the makeup level N_s 2
instrumentation and determined that there was an offset.
3 That is, that this loop seal condition we have been talking 4
about was present during that entire period of time.
5 Therefore, each time a water addition was made, the makeup 6
l fin &1 reading for the leak rate test showed a higher level 7
than it should have shown, therefore reducing unidentified I
8 leakage in the calculation.
9 During the period of late February through 10 mid-March, we believe that there were a whole series of tests 11 in which accounted-for water additions were made with the 12 intention of taking advantage of that loop seal condition and 13 manipulating tests.
14 We have testimony to that effect and we have a 15 clear pattern of water additions that had the effect that I 16 just described.
As you get into the later part of March, 17 although you have continuous water additions being made and I
18 although those water additions had the same effect on the 19 1 outcome of the test -- that is, artificially reducing I
20 unidentified leakage -- the water was being added with such 21 frequency, and there was such a degree of necessity to add 22 that water during that latter part of March, that we did not 23 infer that those water additions in that period of time were 24 definitely made with the intention of manipulating the test.
25 That is, those water additions could have been made for O
ACE-FEDERAL REPORTERS, INC.
202-347 37(Y)
Nationwide Coserage WB)36-6646
^?l24.0
.T j
1849 4
l 1
l legitimate reasons to maintain the level in the makeup and to l
2 l offset the amount of leakage that was collecting in the 1
3 l reactor coolant drain tank.
4 l
Because of this change in necessity to add water S
and because of the fact that those who administered leak rate 6
l tests during the latter part of March were not the same 7
people who were engaged in the test manipulation in February 8
and mid-March, we didn't feel that the evidence was strong 9
enough to characterize these last group of tests as 10 definitely manipulated.
11 That is not to say that we found that they weren't 12 manipulated or the operators didn't know what was going to 13 happen.
We just didn't feel it was strong enough to say with 14 any real degree of certainty that that was the intention 15 behind those water additions.
16 MR. RUSSELL:
I think there's another factor that 17
! is potentially even more significant, and that is, the effect l
18 l of the correction that was made through the temporary change I
19 l notice on the leakage collected in the drain tank.
I 20 i
The leakage to the drain tank is quite large i
21
' during this period, resulting in understating the amount of l 1eakage at reactor coolant conditions by as much as a gallon 22 23
! and a half per minute.
This is the 40 percent error due to l
24 i that effect.
l 25 That was taken out on March 16th, and they made a ACE-FEDERAL REPORTERS, INC.
I 202-347 3700 Nationwide Coverage PJk33(A 646 c-
.82124.0 BRT 1850 l
([ )
1
- correction for it.
So that if one added a large addition of 2
water during a leak rate test after the 16th of March for the 3
purpose of affecting the test result, you would actually 4
likely come out with a large negative number because of the 5
fact that he had to make the correction for the leakage 6
collected in the drain tank; that error was taken out, but 7
you still had the error for the undervaluation of water 8
added.
So that the two were no longer offsetting each other 9
as they were during the period prior to the 15th -- 16th of 10 March.
11 It is also characteristic that there's not a lot 12 of oscillation in the plant parameters.
Basically at 98 13,
power, steady state, you don't see the oscillations in the 14
' makeup strip chart so that that error.that we were talking l
15 about before, of plus or minus
.7 gallons per minute is not 16 apparent from the strip chart during this period of time.
17 So you don't have the oscillation effect.
The 18 error for collection of the drain tank has been collected.
19
' And the only two ef fects you now have, which are opposite 20 each other -- one is the understatement mathematically for i
21 l the value of water added, and the other is the effect of 22 hydrogen, which tends to increase it as a result of i
23 pressurizing the tank -- those two tend to offset each other 24
-- the loop seal effect.
25 MR. STIER:
I would like to add something to that, O
I ACE-FEDERAL REPORTERS, INC.
l 202-347-1700 Nationwide Coverage 800 336-6646 Q --
3-8212460 BRT 1851 1
though, which I think may argue the cther way., slightly, 2
)
We may have gotten into such an obscure issue that Y
3
- it may not be making any sense to anybody at this point but 4
let me add someth;ing for the record in case somebody tries to 5
figure it out.
6 L'te loop seal condition did exaggerate the amount t
7
/ of decline in raakeup level, and if you hadn't added water, 8
that difference -- that is, 20 to 30 percent during that 9
period af time -- in the amount of water that was indicated I
10 as having left the nakeup, would have given you an error that 11 had to be corgensated for somehow.
So that is another factor 22 that would have to be considered ip determining the need to 13 I nanipulate tests in that latter period of time.
14 I think that the need was still there.
I don't 15 know whether the water additioi;* were made, hownver, with the 16 I intent tc manipulate the testr It may have been just that 17 the eufect -- they got the benefit of the effect but they 18
- inter.ded simply to nake up' water that was being lost to the i
13 Ireactor coolant drain tank.
k 2G JUDGE KELLEY:
Okay.
j 21 JUDGE CARPENTER:
Mr. Russell, you indicated an 22 operator might have added water, being aware that the 23 conversion of volume to mass was customarily improperly made 24 Iat THI-2.
Do you have any testimony from any operators that 25 they were aware that -the error was there?
,Nv h
ACE-FEDERAL REPORTERS, INC.
202-347 37(A)
NatNstwide CovcTage MD4M46 i
I lF 82124.0 ERT 1852
( )'
1 MR. RUSSELL:
Ves.
With respect to the mass lost 2
to the drain tank with the TCN, they were aware, and the 3
reason for making the hand correction to account for the
!densitydifferencewasknowntotheoperatorsafterthatwas i
4 5
put into place and the operators did in fact do the hand 6
, calculation and it is contained on the surveillance test 7
records that were maintained after the 16th of March.
So 8
they were aware of this densite difference as it related to 9
the leakage to the drain tank.
10 I don't have any testimony that anybody was aware 11 of making a similar correction for additions to the makeup 12 after the 16th of March.
That error remained and it's still 13 a mystery as to why it wasn't discovered.
,-s k_)
14 JUDGE CARPENTER:
I thought that was your 15 testimony, that the operator might have made the addition of 16 water because he anticipated that erroneous density 17 correction would help the leak rate test result?
18 MR. RUSSELL:
No.
I don't believe that was my 19 testimony.
I didn't intend to give that implication.
20 JUDGE CARPENTER:
Perhaps I misunderstood.
21 MR. CAPRA:
We do have testimony from some l
22 l operators which indicates they were aware of the i
23
{ instrumentation inaccuracies, with respect to adding a i
24 I certain amount of water and seeing a different value 25 indicated on the nakeup strip chart.
In other words, if you O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-lhMM
82124.0 BRT 1853
)
1 added 100 gallons'of water, some operators were aware that 2
the makeup level indicator may have indicated 120 gallons or 3
130 gallons.
4 JUDGE CARPENTER:
That's an indication error, not i
5 a computation error.
6 MR. CAPRA:
Yes, sir.
7 MR. RUSSELL:
They were not aware of the 40 8
percent error for the density correction.
9 MR. STIER:
I think, in order to make the record 10 complete, I ought.to refer back to something that I said in 11 the very early part of my testimony when this same question 12 was raised at that time; that is, whether anybody was aware 13 of the temperature correction error on the makeup side of the g3
\\ '/
14 calculation.
I said that we found one note, one reference in 15 an attorney's notes to an interview of Mr. Coleman that 16 indicated that he had been aware of that error.
In 17 interviewing Mr. Coleman he denied it.
And it may just have 18 been confusion on his part or on the part of the person who 19 wrote the note as to what his knowledge was at the time of 20 TMI-2 operation; I just don't know how that note -- how that 21 reference was included in the note.
But you have that in our 22 appendix and it is available to be resolved if you feel 23 that's an issue.
24 MR. MC BRIDE:
Judge Kelley, on that very point I 25 wonder if we could ask Mr. Stier whether the interview to O
ACE-FEDERAL REPORTERS, INC.
202 347 37fX)
Nationwide Coverage 800-339M6 Y
82124.0 BRT 1854
()
I which he just referred of Mr. Coleman was before or after the 1
2
' Commission's notice of violation and the company's procedural 3
change to account for this density correction?
4 JUDGE KELLEY:
All right.
5 MR. STIER:
I don't know.
6 MR. MC BRIDE:
May I make a representation based 7
on the official files of the Commission?
8 JUDGE KELLEY:
Yes.
Go ahead.
9 MR. MC BRIDE:
The Commission's notice of I
10 violation to the company was dated October 25, 1979, and 11 signed by Mr. Stello.
I believe the record indicates the 12 procedural change -to the leak rate procedure at Unit 2 was 13 made in either very late 1979 or early January of 1980.
O 14 Having said that, I wonder if the question can be-15 re-posed to Mr. Stier if he knows the date of the interview.
16 JUDGE KELLEY:
Do you need to look that up?
17 Well, if you can look into that over lunch, 18 perhaps we could button it up.
19 That brings us, I believe, to Stier test number 20 3.
21 MR. GOLDBERG:
Judge Kelley, there have been a 22 number of references to that notice of violation.
If the 23 Board wishes we can provide a copy to the Board and the 24 parties.
25 JUDGE KELLEY:
Fine.
Why don't you at least O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646
' ~ ~ *^ ~~'
--'.,J_
,J.,- i.__. ---..,,-.- -..~_,._:::
~~L. X 'L T.'":.
~~.:: --.::~:~
82124.0 BRT 1855
(
1 provide one.
We'll. decide whether to put it in or not.
2 We'll take a look at it.
3 MR. CAPRA:
Stier test 3 is test 155 for the 4
Staff.
5 JUDGE KELLEY:
Thank you.
Question 338:
In 6
connection with Stier test 3, is it a fact that the control 7
room operator's log. reflected a 300-gallon water addition, 8
and that the leak rate test calculation performed by the 9
TMI-2 operator included a water addition of 301 gallons?
10 MR. STIER:
Yes.
11 JUDGE KELLEY:
339:
In light of the facts set out 12 in question 338, does it seem likely that there was a t'%
13 deliberate manipulation by the TMI-2 operato'r of Stier test
\\-)
14 3?
15 MR. STIER:
Well, just on the basis of those facts 16 you can't reach such a conclusion.
You need some additional 17 information, and the additional information is whether that 18 particular operator recognized that there was an inaccuracy 19 in makeup level indication.
If he recognized the effect of 20 the loop seal condition on the water addition -- that is, the 21 final reading in the test would be exaggerated by something 22 wrong with the makeup instrumentation -- then you might infer 23 that it was intended as a manipulation.
If he didn't, you 24 couldn't.
That's the key fact, his knowledge of the effect 25 of the makeup level instrumentation.
C)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 j
_ - ~. -.,, _. _ _ _. - _. _ -. _.. _, -. _ _.. _ _. _.. _ _ _.. ~ -. _ _ _. - _.. -. _ _ _ _ - - - _ _ _ _ _ - -
82124.0 BRT 1856
()
1 JUDGE KELLEY:
Okay.
340:
Is it possible that 2
makeup level instrument errors, of which the TMI-2 operator 3
for Stier test 3 was unaware, contributed to the unidentified 4
leak rate that you determined?
5 MR. STIER:
I think I answered that.
6 JUDGE KELLEY:
Yes.
7 Staff comment on those questions on Stier test 3?
l l
8 MR. RUSSELL:
This is shift A.
The panel operator 9
was Mr. Faust and the individual who signed the surveillance 10 test record was Mr. Frederick.
We indicated earlier that we 11 had testimony from shift A that when they made an addition, 12 they would look at the amount of water that went in from the 13 batch controller and they would count,'using the peak-to-peak 14 method, the level change in the makeup.
In this case, using 15 that method, we see a water addition of 300 gallons at 0608.
16 The log in the control room also states 300 gallons.
And the 17 calculation included 301.
18 We believe the 301 was taken from the digital 19 batch controller and those indications would be consistent 20 with each other, indicating that the operators were not aware 21 of this effect.
That would be consistent with their 22 testimony.
And we did not conclude that this test was 23 manipulated as a result.
24 MR. MC BRIDE:
Judge Kelley, Mr. Stier may have 25 missed the thrust of the questions about these last two tests O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646
82124.0 BRT 1857
()
1 because they are performed by different people.
But his 2
answer to question 340 was that his answer would be the 3
same.
.It would seem to me it would clarify the record if he 4
would indicate whether he believed that the people who 5
performed either of these tests were or were not aware of 6
this phenomenon as he testified about it.
7 JUDGE KELLEY:
Can you comment on that, 8
Mr._Stier?
9 MR. STIER:
I think that's a somewhat different 10 question than I was asked here.
I'd have to take a look.
We 11 did not find evidence from which we inferred that these 12 individuals were aware of the effect of the loop seal.
13 JUDGE KELLEY:
That brings us to Stier test 1, 14 341.
That's Staff test --
15 MR. CAPRA:
156.
16 JUDGE KELLEY:
341:
In connection with Stier test 17-1, is it a fact that the control room operator's log and the 18 leak rate test calculation each reflected a water addition of 19 303 gallons while the makeup level rise indicated a 20 320-gallon water addition?
21 MR. STIER:
We estimated that the makeup level 5
22 indication was a 354-gallon rise, not a 320-gallon rise.
But 23 in other respects the answer is yes.
24 JUDGE KELLEY:
And the next question, 342:
In 25 light of the facts set out in question 341, does it seem i ()
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coserage 800 336 6646
82124.0 BRT 1858
(}
1 likely that there was a deliberate manipulation by the TMI-2 2
operator of Stier test I?
3 MR. STIER:
Again, you can't draw an inference 4
simply from those facts.
You have to examine other evidence 5
relating to the knowledge on the part of the people who ran 6
the test, of the loop seal condition.
7 In this particular case we did not find evidence 8
from which we inferred that they were aware of the 9
malfunctioning makeup level indicator.
10 JUDGE KELLEY:
Staff?
11 MR. RUSSELL:
The only slight change in response 12 to this question from the previous one is that we would note 13 that Mr. Frederick was the individual who was on the panel.
14 He testified that he used the batch controller because he 15 felt it was more accurate.
It was Mr. Faust who testified he 16 counted the number of divisions of change and used a 17 30-gallon effect.
18 We don't see a significant differences and we 19 don't believe it would have been apparent to him without 20 going back and reconstructing, to use a slope offset method.
21 We did not conclude that this test was manipulated.
22 JUDGE KELLEY:
Now, 342, if I'm not mistaken, 23 covers the statement we were talking about?
24 MR. MC BRIDE:
Yes, your Honor, that's correct.
25 JUDGE KELLEY:
Do we have any additional O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8043364M6
82124.0 1859 BRT l
i
,.(j 1
follow-ups on the questions we went over?
2 MR. MC BRIDE:
Not from us, your Honor.
3 JUDGE KELLEY:
Okay.
Anyone else?
Mr. Blake?
4 MR. BLAKE:
No, sir.
Wait a minute, Dr. Carpenter 5
has something.
6 JUDGE CARPENTER:
I would like to ask Mr. Russell 7
and Mr. Capra.
Mr. Capra, in the course of your evaluation 8
of these test results, if you had used the correct estimation 9
of the amount of water added in terms of the slope offset 10 method, would that have changed your conclusions about any of 11 the tests?
What I'm trying to get at is, does evaluation of 12 the test results depend critically on, particularly the 13 numerical value of the water added or not?
[t
' ~'
14 MR. RUSSELL:
With respect to water additions 15 occurring during leak rate tests, and using a slope offset 16 method in lieu of the peak-to-peak, that information -- where 17 that was significant, is summarized in table 10 of our July l 30, 1985 memorandum under the heading " water additions (only 18 19 partially included)."
20 By using the peak-to-peak, we think we have been l
21
' conservative.
That is, we understated the amount of the 22 difference.
We reached our conclusions even considering that i
23
' conservatism, s,,
24 We did go back and in fact, we had lengthy 25 discussions on which is the better method to use,
(~1 v
l ACE-FEDERAL REPORTERS, INC.
20L347 3700 Nationwide Coverage 804336-6M6 l
82124.0 BRT 1860
(
't 1
particularly as it related to the later test just prior to 2
the accident; the tests that were done by shift A and shift 3
F.
We felt that it did not make a difference and that's why 4
we did not revise our analysis.
5 So, I believe that we attempted to give the 6
benefit of the doubt to the operator and do it in a method 7
that was available to him to easily use in the control room.
8 That is, you make a water addition, you see what the batch 9
controller tells you as to how much went in and you can count 10 the number of divisions that it changes, multiply by 30 11 gallons per inch and if they come out close to each other 12 that indicstes you are not getting as much of an effect.
13 It's only critical during this latter period where you have 7-14 very high leakage where you are not accounting for the amount 15 of water that is lost from the system during the period of 16 time of the addition.
17 JUDGE CARPENTER:
You spoke to the specific one, 18 if there was a bias it was not a bias against your evaluation 19 of the operators?
20 MR. RUSSELL:
That's correct.
It would be a bias 21 in favor of the operators.
22 JUDGE KELLEY:
Okay.
It's quarter after 12:00.
23 We are going to have, after the lunch break, some discussion, 24 probably off the record, about just how we proceed and who 25 comes next, as I have indicated before.
(O/
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 8 @ 336- % 86
I l
I 82124.0 ERT 1861 73
()
1 It might be useful if I could just state my 2
understanding of the situation.
You correct me if I'm 3
wrong.
Then what I'm going to suggest is that, if counsel 4
haven't already done this, we might be able to allow a little 5
extra time for the lunch break today, perhaps you can get 6
together and hopefully agree on the sequence.
7 We can go off record, I guess.
8 (Whereupon, at 12:15 p.m.,
the deposition was 9
recessed, to be reconvened at 1:45 p.m. this same day.)
10 11 12 13 g
14 15 16 17 18 19 20 21 22 23 24 2S (3
'ss' ACE-FEDERAL REPORTERS, INC.
202 347 1700 Nationwide Coverage 80433W
82124.0 BRT 1862 l
()
1 AFTERNOON SESSION (1:35 p.m.)
2 JUDGE KELLEY:
We'll go back on the record.
We 3
have a ruling to announce.
We have a pending question about 4
whether the strip charts which in the view of MPR showed 5
hydrogen additions, which, I take it, occurred outside of 6
test times and therefore are not in your study, the question 7
was whether we would like to have those.
Is it two strip 8
charts, is that what it comes to?
9 MR. STIER:
Yes.
10 JUDGE KELLEY:
With a covering letter as an 11 exhibit.
The Board is inclined to that idea.
Is there any 12 objection to those two charts as an exhibit?
13 MR. GEPHART:
We have no objection.
(3) 14 MR. MC BRIDE:
No.
15 JUDGE KELLEY:
Without objection, does that bring 16 us to Exhibit 14, then?
Board Exhibit 14 will consist of the 17 covering letter and two strip charts as we've discussed.
18 (Exhibit 14 received.)
19 JUDGE KELLEY:
We would like to announce our 20 rulings on the questions that we spoke to the day before 21 yesterday, I believe.
The broad question might be termed 22 whether there should be Staff input or comment on the 23 employee testimony and whether there should be any such, and 24 if so, how.
In that connection we talked about three 25 different possibilities.
One might have been for Staff oO ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 804336-6M6
82124.0 BRT 1863 I
witnesses, or Mr. Stier, to simply come back and testify 2
after the employees had appeared, and give their views on the 3
employee prefiled and whatever they said on the record.
4 The second possibility was that of having the 5
Staff propose questions in advance of the employee witnesses' 6
appearance, based on the prefiled.
And the third 7
possibility, discussed at somewhat greater length, was 8
whether Staff counsel should be entitled to propose follow-up 9
questions.
We considered these options and the arguments of 10 the parties.
11 As to the first approach, of having additional 12 testimony by the witnesses following the employees, there was 13 no support for that proposition from the Board or the 7s
(
I 14 parties.
The Board simply put it on the table as something 1S to look at.
We do not intend to do that.
We want to, 16 however, drop one caveat there.
As Mr. Blake suggested, we 17 might want to call a particular witness back on a particular 18 technical point, say for purposes of clarification.
We might 19 do something like that.
What we are not going to do is have 20 the experts come back and testify generally about whatever 21 the employees said.
That we don't propose to do.
22 option two, the idea of having prior questions 23 from Staff based on prefiled, we are also rejecting that 24 option.
The Staff itself didn't support that option.
In 2S light of the discussion, the Board doesn't see any clear C) o ACE-FEDERAL REPORTERS, INC.
202 347 37(1)
Nationwide Coverage 80k336 6M6
82124.0 BRT 1864 l
l
-s 1
- advantage of doing that, so we are not going to do that 2
either.
3 j
The third possibility was that of having Staff 4
counsel suggest follow-up questions, and on that point the 5
numerous employees' counsel were opposed to that and the 6
other parties favored it, or at least did not oppose it.
I 7
think favored is a fair characterization.
8 Let me note that we reserved on this question last 9
week.
We had a somewhat different question, whether there 10 should be follow-ups, in the context of this panel, and we 11 said the Staff can propose follow-ups to its own witnesses 12 ibut not to the other witnesses.
There it seemed to us that 13 there was a perfectly adequate means o'f Staff input through 73 14 the interchange between the witnesses themselves, and so that 15 was a substantial reason for that ruling that we have been 16 operating under ever since.
But we did, on the broader 17 question of Staff follow-ups generally, when that came up we 18 didn't reach a decision on that and that's what we are 19 reaching today.
20 The Board, for the reasons that I'll outline, has 21 decided to allow follow-up questions by the Staff.
Looking 22 first to the Commission's order, if the order had spoken to 23 this explicitly one way or the other of course that would be 24 dispositive, but it doesn't.
The order in fact, addresses 25 l relatively few procedural questions of this nature.
What it r~1 m
ACE-FEDERAL REPORTERS, INC.
m.3-m
~. m. c _,.
82124.0 BRT 1865
()
1 does is give the Board some broad outlines and leaves to the 2
Board's discretion within those guidelines to determine 3
questions of this nature.
4 There are, as was pointed out in the discussion, 5
two relevant provisions.
One provision says that the Staff 6
is not a party and that, to be sure, can carry with it a 7
negative implication on this question because one is 8
accustomed to hearing parties ask questions in evidentiary 9
proceedings.
10 Another provision, though, carries somewhat more i
11 weight in our minds, and that is that Staff is explicitly 12 assigned the responsibility to assist the Board in developing I
13
- a record.
This, to us, strongly suggests that there has to (q-i
)
14 be some Staff participation in the hearing.
If they can't 15 even ask follow-up questions, it's hard for us to see how 16 they can do their job.
So we believe that that provision 17 rather strongly supports the notion of follow-up questions i.
18 which are, in comparison to some other approaches, is a 19 rather -- I'll put it differently.
If the Staff were 20 testifying directly or suggesting questions on prefiled, that 21 would be, in our mind, a more major Staff mode of operation.
22 This follow-up question role seems to us to be sort of a 1
23 minimal role, consistent with, again, their helping to
]
24 develop the record.
l 25 We do want to make clear our belief that the Staff CE) i ACE-FEDERAL REPORTERS, INC.
202 147 1700 Nationwkle Coverage 800-)JMM6
a 82124.0 4
BRT 1866 j ()
1 could be very~ helpful in suggesting follow-up questions, as j
2 the case develops, as the various witnesses appear.
So it is
)
3 our desire in our trying to do our job to have that source of 4
assistance.
1 5
There was some concern expressed about the extent 6
of Staff participation by several parties; concern from
}
7 Mr. Voigt that the Staff had adopted an adversary role, 8
essentially, and would, in effect, carry too much influence 9
if given too many procedural rights in the case.
It made you l
10 look too much like what he viewed as a party.
Mr. Blake L
11 expressed some concern along those lines also.
12 It does seem to us that the degree of Staff 13 participation through follow-ups can be and will be l
14 controlled by the Board.
We think that we have complete
~
15 discretion to ask or not ask any question proposed by the 1
1 16 Staff, partly because they aren't a party.
I 17 Let us add in that connection that we think the 18 situation for parties is somewhat different; that if a party i
j 19 proposes a question and the Doard declines to ask it, that a j
I 20 party has a right to request an explanation and the Board, 21 upon such a request, is required to give it.
We had not been j
22 into evidentiary rulings on particular questions to date.
\\
l 23 One reason is very, very few questions have been rejected.
24 I'll note, however, we rejected a small number of questions i
25 proposed by the numerous employees.
They were listed by l ()
i ACE-FEDERAL REPORTERS, INC.
j 202 147 )?00 Nanonwde Coverage 80063)CM6
-.~, _.-,._. _ _... _ - _... _ __._.._ _.,
82124.0 BRT 1867
()
1 number in the transmittal memorandum to the witnesses.
We 2
did not give explicit question-by-question reasons for that 3
in the order.
It seemed to us that the reasons were, at 4
least in most cases, fairly obvious.
5 For example, a number of questions we rejected had 6
to do with Unit 1.
We have addressed the role of Unit 1 or 7
lack of it in this case in some detail.
So it seemed to us 8
that our rejecting a qtestion about some particular aspect of 9
Unit 1 is something that ought to be pretty" clear to 10 everybody.
But if, upon conclusion or at some appropriate 11 time, counsel for the employees or other counsel wants to 12 press a question and wants a Board ruling and wants a reason, 13 we'll provide one.
14 We do not think we need to do that in the case of 15 Staff suggestions.
We don't believe the course we are 16 adopting of allowing Staff to put follow-ups is unfair or 17 inconsistent with the scheme established by the commission's 18 order.
It is true th&t the Staff isn't neutral on the 19 issues, at least on most of them.
They have taken positions 20 on those issues.
They are set forth in a report, taken 21 positions here and testified.
22 It is, however, in judging this question, 23 significant to us that the Staff has not functioned as a 24 prosecutor.
We think overall the Staff has attempted to 25 present the facts and their analysis in a fair matter and we (1)
ACE-FEDERAL REPORTERS, INC.
M 147)?00 Nmon*kie Coserage an))o6644
82124.0 BRT 1868 l
f 1
ecall to attention that they have called to our attention l
2 lexculpatory matter; and this is also part of our decision in 1
3 allowing Staff to make follow-ups.
All those factors 4
considered, that's where we come out.
5 I would like to note just a couple of other 6
related matters.
Mr. Voigt suggested in his argument the 7
other day that we were late in spelling out the role of the 8
l Staff; that the role of the Staff was unclear and confusing.
9 We do not agree with that.
We said early on and we've said 10 more than once that we are working our way through a case in 11 which the guidelines are not previously laid out, either by 12 the Commission or the Commission's rules of practice and that 13 we intend to address issues as they arise and as they need to 14 be addressed.
Indeed, I think it would have been a mistake 15 for us to sit back here, last March, and try to figure out 16 whether the Staff could ask follow-up questions.
It seems to 17 us that it is better to wait and see what the concrete 18 situation presents so long as nobody is prejudiced by waiting 19 for the situation to arise.
20 I might acknowledge I don't recall any party 21 suggesting until the other day what the rule ought to be on 22 this point, and I dare say, had it been suggested to us last 23 March, we would have declined to pass on it.
But it's just 24 that no one has been knocking on the door all along, trying 25 to get a resolution of this question.
It came up here and
(~)
m ACE-FEDERAL REPORTERS, INC.
202 34%)?m Nanonwide Comage WIk 3)MM6
82124.0 BRT 1869
()
I now we've decided it and we'll move ahead.
2 One other point that might bear mention, too.
t 3
There has been some reference to the fact that the Staff 4
doesn't have party status so that the Staff will be able to 5
confer freely with the Board.
And, indeed, the Board and the 6
Staff have conferred from time to time about how to put the c
7 case together on a number of procedural points and that has 8
been extremely helpful.
9 What I do want'to say for the record is that we l
l 10 have never conferred with the Staff on the merits of any 11 issue in this case nor do we intend to.
It may be that if 12 the ex parte rule is technically inapplicable, that we at 13 least are not barred by that rule from doing so.
But it 14 seems to us that we are the ones that are supposed to decide l
15 the case based on the record being developed here.
That's 16 what we intend to do.
So we, to repeat, have not viewed it l
17 appropriate nor has the Staff ever indicated any desire to 18 discuss with us, issues of individual responsibility.
19 That concludes our comments on that subject.
I 20 MR. GEPHART:
Judge Kelley, may I assume that, 21 since we have noted our opposition earlier and Mr. Voigt put l
22 the reasons therefore on the record --
23 JUDGE KELLEY:
Yes.
24 MR. GEPHART:
-- that at this juncture it wouldn't 25 be necessary for us to take an exception to this ruling?
l (Z)
ACE-FEDERAL REPORTERS, INC.
202147.)hv)
NationwtJe Coverage sm))MM6
82124.0 BT.;T 1870
()
1 JUDGE KELLEY:
I think that's fair.
Let's borrow 2
from the rules of practice which say that once you have made 3
your position known, if I rule against you on it then you've 4
got an exception.
The appellate rules in this proceeding 5
will turn out to be whatever they are.
6 (Laughter.)
7 But they are not going to be our problem.
I 8
believe there's an explicit statement that the appeal board 9
-- there's no appeal board involved here, but beyond that, 10 what is going to happen, should any party be dissatisfied whenthismattergoesbefore$hecommission,isreallygoing 11 12 to be between the parties and the commission.
But at least 13 from our standpoint we don't look for a restatement of CE) 1 14 position.
You have an automatic exception as far as we are 15 concerned.
16 MR. MC BRIDE:
And that will be generally l
17 applicable, I take it, throughout the remainder of the 18 proceeding?
19 JUDGE KELLEY:
Yes.
Right.
20 Could we go off the record, then, for a discussion 21 of the order of proceeding?
22 MR. MC BRIDE:
Yes.
I think we should be off the 23 record.
24 JUDGE KELLEY:
All right.
Off the record.
25 (Discussion off the record.)
O ACE-FEDERAL REPORTERS, INC.
202 147 1700 Nationwde Coverage RG))66M4
... ~..,, - +,.. +
82124.0 BRT 1871
()
1 JUDGE KELLEY:
We are going to shift, now, to 2
several questions from Mrs. Aamodt.
3 MS. HENSLEY:
Before we start in with the i
4 questions, Judge Kelley, can I mention one point?
5 JUDGE KELLEY:
Sure.
i 6
MS. HENSLEY:
You asked me on Tuesday if I minded 7
serving the parties with copies of the questions that Gary 8
Miller had for Mr. Stier, and I said I didn't mind.
I still 9
don't mind doing that.
I was under the vague impression that 10 Mrs. Aamodt might be here today, so I did not -- I have her j
11 questions with me.
I hand-served all the other parties and 12 have not mailed hers out to her.
l 13 JUDGE KELLEY:
Your questions to Mr. Stier; is 14 that right?
l 4
15 MS. HENSLEY:
Right.
Should I just mail those to 16 her?
Hang onto them?
17 JUDGE KELLEY:
I think it's fair to say that we've 18 followed a procedure whereby proposed questions from parties 19 go to the Board and not to the other parties.
When we are 20 all here in a hearing, as a point of convenience when nobody i
21 really cares any more, we hand them out so people can look at 22 a piece of paper instead of listening.
The distinction is 23 not a great one; that is to say, we have not formally served l
24 these questions on other parties.
We have just, as a matter 25 of convenience and simplicity, handed them out.
O l
l ACE-FEDERAL REPORTERS, INC.
j 202 347.)?00 Nation *ide Coverage 804))HM6
82124.0 BRT 1872
()
1 If you wish to mail, fine.
I don't discourage you 2
but I don't think it's necessary.
It is going to be in the 3
transcript.
Mrs. Aamodt will be picking up ith us again 4
soon, I'm sure, and I really don't see that she is in any way 5
prejudiced, whether you mail the questions or not.
6 Any comment from counsel on that?
7 MR. FLYNN:
Judge, I have the same question 8
regarding Mr. Herbein's questions for Mr. Stier.
I also have 9
a copy here for Mrs. Aamodt.
10 JUDGE KELLEY:
I think you can hold them.
That's 11 reasonable.
12 MR. GOLDBERG:
If we are finished with that, 13 before you get into the Aamodt questions, I would suggest O
14 that you allow the Staff.to answer your question to them 15 about Mr. Morck.
16 JUDGE KELLEY:
Oh, yes.
We talked about Mr. Morcl:
17 and his role.
Mr. Stier has a point?
18 MR. STIER:
There is just one other pending matter 19 from this morning and that is I had been asked for the date 20 of a certain document.
I have that and I can give that to 21 you whenever you need it.
22 JUDGE KELLEY:
Go ahead.
23 MR. STIER:
The document -- the attorney's notes 24 are dated April 10, 1980 and they are found in tab 52 of 25 volume V-D of our report.
O ACE-FEDERAL REPORTERS, INC.
M 147.)?00 Natmnnkle Coverage ko))MM6
' ~~ '~"2? ':
2 J ~ " * ? I
~
83124.0 BRT l
1873
/%()
1 MR. MC BRIDE:
And that, for clarity of the 2
record, is the date of notes of an interview with 3
Mr. Coleman.
4 MR. STIER:
Mr. Coleman, yes.
And the 5
interviewers are Kirkpatrick and Campbell.
6 MR. RUSSELL:
Gamble.
7 MR. STIER:
Well, it says Campbell here.
It's 8
very cryptic notes.
9 JUDGE KELLEY:
Staff had some comments about 10 Mr. Morck's role?
11 MR. RUSSELL:
Staff did not investigate matters 12 which we felt had been resolved by the criminal proceeding 13 and, in fact, was directed by the Commission not to duplicate 7-)
14 matters which had been resolved through that process.
We 15 did, however, review materials in 680, supplement 5, which is 16 the restart safety evaluation issue by the Staff in the TMI 17 restart proceeding.
18 Based upon my understanding of Mr. Morck's role, I 19 believe it would be necessary to gather information from him 20 on his involvement in the procedure change and, in 21 particular, discuss his general knowledge of problems with 22 the leak rate tests; whether there was activity going on to 23 try and correct it and what his role was.
And I believe that 24 is within the scope of the commission's order.
Were I doing 25 it, having been directed by the Commission, I certainly would O
ACE-FEDERAL REPORTERS, INC.
202 4 47 4 700 Nanon* Lie Coverage WM4M
l 82124.0 BRT 1874 i[
1 l interview Mr. Morck.
l
~-
2 JUDGE KELLEY:
Could you elaborate a little bit 3
more about what you think he would know about the problems, 4
generally?
5 MR. RUSSELL:
It is my understanding that the 6
engineering support from the group that was headed up by 7
first Mr. Seelinger and then Mr. Kunder, that they provided 8
technical support to operations,and the conduct of operations 9
of the plant.
While Mr. Seelinger and Kunder were the 10 managers of that activity, the individuals responsible, 11 actually implementing, doing the engineering analysis, doing 12 the review, reported to them.
13 It is my understanding that Mr. Morck, as the i
14 shift test engineer that supported the change to the TCN 15 which was issued on the 16th of March was the individual 16 involved in how leak rate tests were calculated.
17 He may have been involved earlier in the issues 18 associated with roundoff.
I don't have documentary evidence 19 that I can point to, but based upon discussions I have had 20
,and information that I've come upon in the course of doing 21 this work, that is my belief.
22 It is somewhat speculative as to the earlier role, 23 during the time of the LER and the roundoff issue.
I believe 24 it is correct that he was involved as the individual who did 25 the technical engineering work associated with the TCH.
(v)
ACE-FEDERAL REPORTERS, INC.
202147.)*m Nanonwale Coverage
- nl%M46
. ~...
82124.0 BRT 1875 ex k_)
1 JUDGE KELLEY:
Thank you.
2 MR. MC BRIDE:
Does the Board wish to hear 3
argument on that?
4 JUDGE KELLEY:
Well, I think we understood your 5
point before when the issue arose.
Do you have anything 6
different to add to it?
7 JUDGE KELLEY:
I just want to say this.
I'm not 8
aware of any evidence that he was involved in the roundoff 9
issue, but I think you can confirm that when you hear from 10 Messrs. Havercamp and Seelinger.
Secondly, as to his role, I 11 think you can ask Seelinger and Kunder.
And thirdly, as to 12 the TCN, it seems to me there are two theoretical b'~T 13 possibilities.
It never occurred to him that the same 14 density correction should be made, in which case'you call him 15 down here and he's a witness and he doesn't contribute 16 anything to the record.
Or he was; if he was, there's no 17 showing that anybody else ever thought about that or had 18 anything to do with an operator falsifying a test or anything 19 of that sort, so you still end up with a big nothing, but I 20 think you can defer this until you hear from the other 21 witnesses.
22 JUDGE KELLEY:
We will consider.
Thank you.
Let 23 me ask Mr. Russell a question I'm hesitant to ask, but in 24 this same vein, you have seen this proceeding develop.
I 25 don't know whether or not you have noticed we subpoenaed some O
ACE-FEDERAL REPORTERS, INC.
202 147.)?00 Natsonwide Cogetage a@))NM4
82124.0 BRT 1876
()
1 additional people as time has gone on.
2 Just in view of the basic issues in front of us 3
and what we are trying to find out, do you think there is 4
some other -- you had indicated maybe we should call 5
Mr. Morck.
Are there other people we are missing so far, in 6
your view?
What we've done is go over the reports and infer
~
7 from what is in there who seems to know about things.
8 MR. RUSSELL:
I believe that the best approach 9
would be -- I'm not aware of anyone that you have left out, 10 based upon either the individual being a party or having been 11 subpoenaed.
12 There is some confusion in my mind as to the 13 actual sequences around the licensee event report and the 14 facts as characterized in that licensee event report from the i
15 facts as we know them from being able to reconstruct the 16 record.
In my mind they paint two different pictures.
17 I think it would be very important to establish 18 who was involved in developing that licensee event report, to 19 what extent there was review of it prior to issuing, and I 20 don't know who-all the individuals were that were involved.
21 I believe that that would be appropriate to 22 address to the individuals that signed it, find out who 23 assisted them, et cetera, and work backwards.
It would be 24 speculative on my part to name names as to who was involved 25 in that now, just as it was speculative with respect to the (2)
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nat6onwide Coverage 800 33M646 r-
82124.0 BRT 1877 (O
_)
I roundoff.
And I don't think that really helps at this 2
point.
3 JUDGE KELLEY:
But that's an area you would point 4
to where there may be other people with knowledge of relevant 5
facts, I take it?
6 MR. RUSSELL:
That particular event was very 7
significant, in that practices didn't change after it.
If 8
there was a conscious effort to play down the importance of 9
that, and that that contributed in some way to the practices 10 continuing, that I think could be rather significant.
11 JUDGE KELLEY:
Thank.you.
Do you have any 12 comment, Mr. Stier, in the same vein?
13 MR. STIER:
Yes.
I have been thinking about the 14 same issue and 1 think, with hr. Morck, you would have the 15 most -- the people that I would call -- with two -- I have 16 two questions.
I'm not sure whether these people are coming 17 or have been subpoenaed or are parties.
One of them is 18 Kidwell.
19 MR. MC BRIDE:
He has been subpoenaed.
20 JUDGE KELLEY:
He has been subpoenaed.
21 MR. STIER:
And the other is Philljppe.
22 JUDGE KELLEY:
Well, we decided, it's not un 23 irrevocable decision, but we thought abfut Ph1111pp3 24 Phillippe is, I think, the only person ch the staff's list on 25 table 1, wh11h lists all the people who were then shift O
Ae-FEDERAL REPORTERS, INC.
202 347 3700 Natior. aide Coverage Im)366M6
82124.0 BRT 1878
)
1 supervisors, foremen and CRos, and everyone is coming with 2
the exception of Phillippe.
We thought he was a trainee, he 3
didn't respond to the letter, we don't know where he is, and 4
the number of questionable tests in which he was involved was 5
fairly small, so we decided to pass on Phillippe.
But that 6
may be a mistake.
7 MR. STIER:
The reason I raise his name is he was 8
on shift with Congdon and Cooper.
There is a significant 9
issue as to whether the shift supervisor, Mehler, or the 10 shift foreman, Adams, had knowledge of test manipulation, and 11 if so, to what extent.
He is a person who was on that 12 rhift.
Had we been able to gain access to him I would have 13 wanted to, talk to him, a'nd that's why I raise his name.
14 It is not a question of specific tests that he may 15 have been involved in.
It's a question of what he may have 16 heard from other members of his shift, conversations with a 17 shift supervisor or shift foreman, that I would think would 18 be helpful.
19 JUDGE KELLEY:
You don't happen to know where he 20 is, do you?
21 MR. STIER:
No, I don't.
22 JUDGE KELLEY:
Okay.
Thank you.
23 MR. RUSSELL:
I believe he's at Waterford 3, 24 Louisiana Power & Light.
25 JUDGE KELLEY:
Thank you.
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 33MM6
-~
82124.0 BRT 1879 tQ
(_j 1
MR. CAPRA:
He's not an operator there.
2 JUDGE KELLEY:
That brings us to the questions 3
from Mrs. Aamodt to Russell and Capra, in the Board's 4
understanding.
5 (Recess.)
6 JUDGE KELLEY:
For the record, Exhibit l'2 and 13, 7
the exhibits we accepted earlier are to be supplied later so 8
they won't becoming in until later.
9 Our next order of business is to ask some 10 questions from Mrs. Aamodt to Mr. Russell and Mr. Capra.
4 11 I'll make a couple of observations.
There are just three 12 questions for Mr. Russell, three for Mr. Capra.
13 These questions are, I think, kind of a mix of 14 technical and individual responsibility.
I'll just note that 15 we see them that way but I think we'd like to go ahead and 16 put them.
17 Our plan, generally, is to have Mr. Christopher 18 appear with Russell and Capra for individual responsibility 19 questions, but since these are a mix de'11 go ahead this 20 afternoon; Mr. Christopher can look over the transcript or if 21 there's some thought he should look at these, too, we can 22 take that into account.
23 Having said that, Aamodt question number 1 we 24 disallowed because we didn't think it was an understandable 25 question.
O ACE-FEDERAL REPORTERS, INC.
M.347 3700 Nationwide Coverage 8 6 3346646
82124.0 l
BRT l
1880
(
1 Number 2, and I'll have a comment on that as 2
well.
You conclude, page 12 of your report -- that some i
3 operators were motivated to manipulate or falsify leak rate 4
test results.
Can you explain what you mean?
What was the 5
basis for each operator's motivation that you include in this 6
generalization?
7 That question, especially in the light of what's 8
been covered in the J ast 10 days is too broad and we 9
recognize that and we don't expect you to give chapter and 10 verse on every test and response to it.
But if you could 11 respond as you think appropriate.
12 MR. RUSSELL:
My conclusion is contained in the 13 NRR report and it is based upon the fact that some operators 7,! '/
'~
14 were motivated to manipulate or falsify leak rate tests.
15 That is based upon the fact that five operators, t
16 Mr. Blessing, Mr. Coleman, Mr. Congdon, Mr. Cooper and Mr.
l 17 Hartman, have admitted making hydrogen and/or water additions 18 during leak rate tests and admitted prior knowledge that 19 these additions could, and some indicated they hoped would, 20 influence the test result.
21 I also believe that the weight of evidence is 22 sufficient to support a conclusion that at least three other 23 operators, mr. Olson, Mr. Booher and Mr. Wright, made 24 hydrogen and/or water additions during leak rate tests with 25 prior knowledge and the expectation that their actions would
')
(G ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80k336-6M6
,-, y e..
82124.0 BRT 1881 j
1 favorably influence the leak rate test result.
Thus, I 2
l believe that at least these eight operators attempted to 3
manipulate and thus falsify leak rate test results.
4 Some of these operators have argued that hydrogen 5
and/or water additions were not precluded by the procedure 6
and, thus, their actions did not constitute falsification.
7 Using this rationalization to get a good leak rate test in 8
order to keep the plant operating is contrary to reactor 9
safety and the terms and conditions of their licenses.
Their 10 motivation was clearly to satisfy what they believed was a 11 paperwork requirement to allow continued operation of the 12 plant.
13 This attitude on the part of operators and the
,s
(
)
14 management within the plant operations department that a leak 15 rate test was only a paperwork requirement and the operators' 16 belief that the real leak rate did not pose a safety problem 17 are factors which motivated them to attempt to manipulate 18 what they believe was a meaningless test.
19 I believe that the TMI-2 operations department 20 l management bears primary responsibility for the action of 21 these operators because of their failure to resolve problems 22 with the leak rate test, to instill high standards for procedural compliance, their failure to correct the 23 4
l 24 misunderstanding of requirements associated with reactor 25 coolant system leakage limiting conditions for operation, o
U l
ACE-FEDERAL REPORTERS, INC.
202 347 3700 N u nonwide Coserage 800-33 9 686
82124.0 BRT 1882
()
I their failure to resolve hardware problems associated with 2
the makeup level instrumentation, their failure to correct 3
pressurizer leakage problems which made plant operation more 4
difficult for operators, and their failure to develop and 5
implement a correct and meaningful reactor coolant system 6
leak rate procedure.
7 JUDGE KELLEY:
Okay.
8 Mr. Stier, I think your answer is over the many 9
days, you have spoken in part to this also.
Do you want to 10 comment on Mr. Russell's comments?
11 MR. STIER:
I think my conclusions are essentially 12 in agreement with Mr. Russell.
I think we may have 13 structured the issues that we looked at somewhat differently 18 and maybe I ought to take just a minute to tell you what the 15 is' sues were that I looked at, how I defined the 16 responsibilities of TMI-2 employees involved in leak rate 17 testing.
18-It seemed to me in examining the technical 19 specifications and procedures which were, as far as I was 20 concerned, the law by which these individuals had to operate, 21 they created two basic sets of responsibilities:
one was to 22 document leakage; that is, to actually measure and record the 23 amount of reactor coolant system leakage that was in 24 existence at any given time, by means of leak rate test.
So 25 the test had to be used for that purpose; they had to have a O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6646 2..
__-._m
c--
c.-. -
82124.0 l
BRT 1883 O)
(_
1 sense that, when they recorded a leak rate test, that it had 2
some connection with reactor coolant system leakage.
The 3
second duty was really independent, in my judgment, of the 4
leak rate test itself.
But it was a duty to respond to 5
circumstances where they really believed or had strong reason 6
to believe that leakage from the reactor coolant system 7
exceeded a limiting condition for operation.
8 The leak rate test may very well have indicated to 9
them that they were beyond the leakage limits allowable under 10 technical specifications, but if they had information from 11 some other source that led them to believe that leakage 12 exceeded 1 gpm, the technical specifications required a 13 response to that.
They couldn't keep the plant operating if
~
14 the leakage exceeded 1 gpm, beyond the time periods allowed 15 by the technical specifications.
16 I observed a number of procedural violations that 17 were pervasive in connection with leak rate testing, such as 18 failure to make log entries and a variety of other procedural 19 violations associated with leak rate testing.
As far as I 20 was concerned, the bottom line of the procedures, the purpose t
l 21 of the technical specifications and procedures, were as I l
22 have just described them:
To measure leakage and you had to 23 respond when you believed leakage exceeded the limiting 24 condition for operation.
25 Our report examines the behavior of plant I
C:)
ACE-FEDERAL REPORTERS, INC.
l l
202-347-3700 Nationwide Coverage 800 336-6646
82124.0 BRT 1884 1
personnel in response to both of those duties and my 2
conclusions with respect to their attitudes and motivations 3
differed somewhat with regard to those two duties; that is, 4
my conclusion was that leak rate test, in the minds of 5
members of the operations department, including supervisors 6
within the operations department, leak rate test was 7
meaningless as a measurement of leakage.
When they filed a 8
leak rate test result, they didn't evaluate it to determine 9
whether it represented a measurement of actual leakage.
When 10 they discarded a leak rate test, they didn't evaluate it to 11 determine whether leakage might have exceeded 1 gpm.
There
~
12 seemed to be virtually no connection in the minds of the 13 people running the tests or supervising the test, between the O
14 results that they were getting and actual leakage.
- And, 15 therefore, I believe that the members of the operations 16 department who performed tests and supervised those tests, 17 failed in that duty.
18 I believe that the failure to document tests, to 19 document leakage, to use the test to document leakage and the 20 continued utilization of a test that they viewed as 21 meaningless throughout the entire period of TMI-2 operation, 22 that that motivation stemmed from the policies and attitudes 23 of people in the operations depa'tment up through the r
24 supervisor of operations, the top level of the operations 25 department.
As a matter of fact, the reason why we.went all O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 80 4 336-6646 7
82124.0 BRT 1885
()
I the way back to the beginning of TMI-2 operation and examined 2
tests was to try to trace the source of that policy, and I 3
think it is quite clear from what we observed in those early 4
tests, that from the outset of TMI-2 operation, there was 5
virtually no regard for the validity of test results.
Some 6
of the earliest tests were the ones that had the most absurd 7'
results.
And yet they were still filed.
8 Even the hand calculations -- and we may get into 9
this in response to some other questions later on, but I 10 think it's appropriate to kind of pull it together in answer 11 to this question.
Even the hand calculations that were done 12 in that early period contain evidence that there was no
'13 regard for the test validity.
Some of the hand calculations 14 contained, for example,~ voltage readings instead of level 15 readings.
Temperatures were recorded, in hand calculations, 16 temperatures were recorded at levels tnat couldn't possibly 17 have existed at the time the leak rate test was run because 18 the plant was in a mode where we know the temperature was far 19 below that level and the person who ran that test must have 20 known that.
Yet he still performed the test and filed the 21 result.
22 So I think that the evidence is clear that the 23 motivation for regarding the test in that way came from the 24 top level within the operations department.
25 Motivation to manipulate tests, which is a O
ACE-FEDERAL REPORTERS, INC.
~"
~ ~ "*' ~ "*
)'
82124.0 BRT 1886
(~h 1
subpart, in my judgment of the overall attitude toward the s_)
2 validity of leak rate testing, the motivation there, I 3
believe, stemmed from the operators who, in response to 4
demands from their operations department supervisors to get a 5
good leak rate, found various means to achieve acceptable 6
results -- that is a result under 1 gpm -- including 7
manipulating the results by adding hydrogen or adding water 8
and -- or by, perhaps,. accident, by using a level indicator 9
that was malfunctioning, thereby increasing the chances of 10 obtaining a leak rate under 1 gpm, even though actual leakage 11 exceeded that.
12 Motivation for failing to respond to indications l
13 of leakage over 1 gpm was somewhat different.
And in my 14 report you will find that responsibility for failing to 15 respond to leakage over 1 gpm goes somewhat higher within the 16 TMI-2 organization, outside the operations department up 17 through the level af superintendent for technical support, at l
18 least.
And I believe that there was, based on the evidence 19 that we gathered, that there was an attitude within the TMI-2 20 organization that the 1 gpm limit for unidentified leakage 21 was inappropriate, unrealistic, and unrelated to safe 22 operation of the plant.
23 Now, to balance that, I believe that there is l
24 clear evidence that people who were operating the plant 25 wanted it to run safely, were concerned about leakage, looked O
i ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 800-336-6646
82124.0 BRT 1887
(}
1 for leaks, tried to repair leaks.
But the evidence, I 2
believe, is clear that they did not apply a 1 gpm limit to 3
What the specific source of motivation 4
was for that attitude, I'm not really sure.
5 JUDGE KELLEY:
I would like to pursue with you, 6
some, your reference to top levels of the operations 7
department and tech support, in terms of actual people.
But 8
not right now.
We have to quit at 3:30.
We would like to 9
finish up with Russell /Capra.
Maybe'what we'll do is just 10 put these few questions to them and then take such time as we 11 have left this afternoon we can give back to you.
12 Mr. Russell, Aamodt question 3.
You indicate in 13 your conclusions concerning most operators that physical O
14 evidence was insufficient to conclude that these operators 15 had willfully manipulated leak rate tests.
However, NRR 16 only considered the evidence for the final six, month period 17 of operation of Unit 2.
How does NRR rationalize not taking 18 into account the evidence of leak rate manipulation during 19 the first months of TMI-2's operation or by these individuals 20 at Unit I?
Did NRR make an evaluation of this prior 21 involvement for any individual?
22 I note simply that we've ruled out Unit 1 issues 23 except in a very narrow sense.
Go ahead.
24 MR. RUSSELL:
My decision to limit NRR's 25 investigation to the six-month period prior to the accident C)
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336-6M6 w
---,-n
,w-,.
g w
-*--igr---
~r
82124.0 BRT 1888
()
1 was based upon a review of materials in OI'9 Region 1 office 2~
in early 1985; a review of the U.S.
Attorney's statement of 3
fact and the significance of the October 1978 NRC 4
inspection.
5 As a matter of fact, I was more concerned with the 6
periods of high leakage during the last 2-1/2 months of 7
operation' because it appeared from Hartman's testimony and 8
the limited analysis available at that time that this was the 9
period during which it was difficult to obtain good leak rate 10 tests.
11 Based upon reviewing the final results of our i
12 analysis in July 1985, I saw no need to go back to review 13 leak. rate tests pri'or to September 30, 1978.
I believe 14 Mr. Stier's analysis for the period prior to September 30, 15
'78 confirms my earlier conclusion to limit NRR's analysis to 16 the last six months prior to the accident.
17 JUDGE KELLEY:
The next is 4:
How does NRR 18 justify limiting its investigation to presently licensed 19 operators?
How will NRR respond to the application for 1
20 licensing of those likely involved but not inves;tigated?
How 21 does NRR rationalize not having investigated those likely 22 involved who presently hold responsible positions at TMI, in s
23 other parts of the nuclear industry, or who may return to 24 responsible positions in either?
25 JUDGE KELLEY:
NRR limited it's investigation to O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 800-336 6646
82124.0 BRT 1889 I
1 10 operators.
My reason for recommending that scope be
\\s) 2 limited to 10 operators are as follows:
First, these 3
individuals were directly involved in activities which had I
4 the potential to affect public health and safety; that is, 5
the operation of San onofre 2, 3, Waterford 3, and cleanup at 6
TMI-2.
7 Second, the action could be legally taken to 8
remove these individuals from licensed duties, if needed.
I 9
also believed that the five-year statute of limitation had 10 l expired for enforcement actions against other individuals.
11 The scope of this hearing is not limited to 12 currently licensed operators, and pursuant to the 13 Commission's order the Staff will, following this hearing,
(_
14 make recommendations on appropriate actions, if any.
- Thus, 15 questions related to actions against individuals outside the 16 scope of NRR's initial investigation are premature.
17 Three questions from Mrs. Aamodt to Mr. Capra.
18 1:
How does NRR's identification of invalid tests 19 compare with those of Dr. Chung for the same period?
20 MR. CAPRA:
When NRR began its investigation of 21 individual operators, the results of Dr. Chung's analysis 22 were used in questioning the operators.
The analysis 23 provided to me on September 18, 1984, by Dr. Chung, included 24 his summary table of individual leak rate tests.
I did not 25 learn until recently that the version provided to me on O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Coverage 804336-6M6
^
82124.0 BRT 1890
/ 'l 1
September 18th was a slightly different and updated version
(._/
2 of table provided to representatives of the law firm Kaye, 3
Scholer and MPR by the Department of Justice on February 15, 4
1984.
5 The latter version appears in the Stier report as 6
Exhibit 2 in volume IV-L.
7 The summary table of Dr. Chung's analysis focused 8
principally on the identification of leak rate tests where 9
water additions and/or hydrogen additions took place during 10 the tests.
Except for some preliminary interviews conducted 11 by the NRC, in March 1980, Dr. Chung's analysis was performed 12 without the benefit of detailed interviews of the operators 13 involved in conducting the leak rate tests.
Dr. Chung was 7S V
14 not personally involved in those interviews.
15 Following our interviews with the first 13 16 operators, we found it necessary to go back and do a complete 17 reevaluation of each test, to factor in the information we 18 learned throughout the interview process.
19 The information developed that we factored into 20 our reevaluation, included the following:
Which operators 21 admitted manipulating tests; the significance of the period 22 of time LT-1 was out of service; what indicators operators 23 relied upon to quantify the amount of water added to the 24 makeup; how shift operations were conducted; the interface 25 I and communications between individual operators, shift O
ACE-FEDERAL REPORTERS, INC.
202-347-3700 Nationwide Cosetage 80 4 336-6646
.. ~. -. -..
82124.0 BRT 1891
()
1 foremen and shift supervisors; what information operators 2
believed their shiftmates and supervisors knew and what they 3
believed operators on other shifts knew; the difficulty of 4
getting good leak rate test results during periods of high 5
leakage without test manipulation; the consistency of logged 6
events with operators testimony; which shifts were able to 7
get " good leak rate tests" when other shifts couldn't; and 8
patterns associated with leak rate testing.
9 Our reevaluation included a review of all 10 testimony, the control room log, the shift foreman log and 11 the plotting of all leak rate tests and logged entries on the 12 makeup strip chart roles.
Prior to this time only the makeup 13 strip chart excerpts provided by Dr. Chung were used.
14 This reevaluation was completed in May 1985.
The 15 analysis was updated again at the completion of the interview 16 process.
The final NRR analysis, dated July 30, 1985, is the I
17 one contained in the NRR report.
18 Because Dr. Chung's summary table focused on 19 hydrogen additions and water additions during leak rate test 20 period -- during the leak rate test period, it is difficult i
to make a one-to-one comparison with test conclusions to the 21 22 NRR analysis, which includes nine separate categories.
23 However, if a comparison is made by classifying tests only as 24
" valid" or " invalid," the NRR analysis agrees with 25 Dr. Chung's on 119 of the 161 tests; that is, 44 valid tests, O
ACE-FEDERAL REPORTERS, INC.
202 347-3700 Nationwide Coverage 80 4 336-6646 u--- ---.----.--.: :
82124.0 BRT 1892
()
1 75 invalid tests.
We did not reach the same conclusions on 2
42 tests.
3 of these 42 tests, NRR believes that 35 are 4
invalid based upon our reanalysis.
The vast majority of 5
these tests, 25, were classified as invalid by NRR as a 6
result of level transmitter problems associated with the 7
makeup.
8 There are seven tests classified as invalid by 9
Dr. Chung that NRR has classified as valid.
Six of these 10 tests involve possible small underrecorded water additions 11 that could not be confirmed during our reanalysis.
12 In summary, Dr. Chung's analysis contains 82 13 invalid. tests and the NRR analysis contains 110 invalid
()'
i 14 tests.
15 I have to caution about my response there, in 4
16 trying to be responsive to Mrs. Aamodt.
I ran the same 17 problem -- ran into the same problem in trying to do a 18 one-to-one comparison, as I did doing it with Mr. Stier's 19 analysis.
l 20 Dr. Chung's table does not classify tests as 21
" valid" or " invalid."
So I may have misrepresented some of 4
22 Dr. Chung's conclusions.
l 23 JUDGE KELLEY:
Thank you.
That's a very helpful 24 answer.
i 25 Question number 2:
What evidence does NRR possess i
l CE)
ACE-FEDERAL REPORTERS, INC.
1 202 347-3700 Nationwide Coserage 800 336 4646
82124.0 BRT 1893 1
that, " actual and unidentified leakage did not exceed 1 gpm" 2
(NRR, Stello memo, page 3) since 69 percent of the leak rate C
3 tests reported in the final six months' operation of TMI-2 4
were deemed invalid and the leak rate test was the only 5
measure of unidentified leakage?
6 MR. CAPRA:
The statement " actual unidentified 7
leakage did not exceed 1 gallon per minute" in context 8
describes the fact that many times operators obtained Jeak 9
rate test results greater than 1 gallon per minute when the 10 plant was not leaking at a rate greater than 1 gallon per 11 minute.
NRR did not perform an independent -- did not 12 perform independent calculations to determine the periods of 13 actual unidentified leakage greater than 1 gallon per
[}
14 minute.
However, we did have access to and reviewed the 15 Faegre, Benson analysis, which identified only three periods 16 where underidentified leakage could have exceeded 1 gallon 17 per minute.
Those periods are December 16 through December 18 20, 1978; January 6 through-15, 1979; March 27 and 28, 1979.
19 Also, we met on September 20, 1984, with Mr. Stier 20 and MPR Associates.
At that meeting we discussed the MPR 21 analysis for periods where unidentified leakage could have 22 exceeded 1 gallon per minute.
A transcript of this meeting 23 was taken and provided to the TMI-1 Hearing Board as Board 24 Notification 84-166.
25 Analysis included in the Stier report, volume O
}
l' M-347-3700 Nationwide Coverage 80433M686 ACE-FEDERAL REPORTERS, INC.
o 82124.0 BRT.
1894 I) 1 IV-A, appears to be a later version of this analysis that was 2
reviewed during the September meeting.
3 All three analyses -- that is, the Faegre, Benson 4
analysis, the Stier MPR September '84 version, and the Stier 5
MPR June '85 version -- support the conclusion that for many 6
leak rate tests with results greater than 1 gallon per 7
minute, the. actual or real unidentified plant leakage was not 8
greater than 1 gallon per minute.
9 It is also true that for some periods the plant 10 did have unidentified leakage greater than 1 gallon per 11 minute, and also had leak rate test results which were 4
12 greater than 1 gallon per minute.
13 JUD'GE KELLEY:
Thank you.
The last question for i
14 Mr. Capra from Mrs. Aamodt:
Identify all individuals who 15 participated in the identification of invalid leak rate l
16 reports and provide curriculum vitae for each.
l 17 I take it that refers to the NRR report.
Other 18 than yourself and Mr. Russell, were there others 19 significantly involved?
20 MR. CAPRA:
Myself and Mr. Russell were the two 21 individuals who did the entire NRR analysis.
We did consult 22 with Dr. Chung, as I mentioned in my previous answer, and 23 also Mr. Kirkpatrick during the course of the analysis.
Our 24 technical qualifications for all four of -- well, for myself I
25 and Mr. Russell and Mr. Kirkpatrick, are part of the record.
j CE) i l
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80433MM6
82124.0 BRT 1895
("T
(_/
1 JUDGE KELLEY:
Right.
Thank you.
I don't know 2
that we need a curriculum vitae, necessarily, for Dr. Chung.
3 We do have a pending request to call him as a witness which 4
Mrs. Aamodt, who is the one suggesting this, may wish to 5
pursue later.
Right now that's fine.
6 Mr. Stier, I'm sorry.
Can we backfit?
Some of 7
these questions, I guess, point pretty much to NRR, but in 8
any case, do you have comments on them?
9 MR. STIER:
No.
I don't have comments on the last 10 questions.
I would like to add something to what I said 11 before, but if you are going to ask me more questions about 12 it, maybe it will be developed in those questions.
13 JUDGE KELLEY:
Well, the thing that caught my
{}
14 attention in your last comments, you were talking about sort 15 of general attitudes and why things seemed to happen the way 16 they did and you made reference to people in the, I think 17 your reference was " top levels of the operations department" 18 was one phrase; and another was a reference to the tech 19 support department and their role in all of this.
20 We have talked pretty much about CROs and foremen, 21 and not about senior supervisory people such as Mr. Seelinger, 22 and others.
In your reference -- well, could you expand on 23 what you referred to?
We had your report.
I know you 24 discuss these people in your report, and it may not be 25 anything more than what is in the report, but, since we O
ACE-FEDERAL REPORTERS, INC.
E-347 3700 Nationwide Coverage 80433MM6
. _ _ _ _ _. - ~._. _ _ _--__ - _ _ _,_____ _ _._
82124.0 BRT 1896
()
1 really haven't heard from you on very much about those 2
individuals, could you say more specifically who you had in 3
mind?
4 MR. STIER:
Yes.
First of all, one note of 5
caution.
I'm here speaking extemporaneous 1y and this report 6
was written, I think, as carefully as I can write a 7
document.
If I say something here that is inconsistent with 8
what I said in the report, please believe what is in the 9
report because I still believe it is completely accurate and 10 valid and I don't intend to amend or modify the report in any 11 way by my answers.
12 I want to be fair, of course, to the individuals 13 that I'm talking about.
And so I think I should qualify to 14 some extent, or at least describe to some limited extent, the 15 kind of evidence that I was talking about when I identify l
t 16 them, because I don't mean by what I'm saying to be a blanket 17 indictment of these people.
And I think the evidence varies 18 in its strength with respect to each of them.
19 The individual I was talking about in the 20 operations department is James Floyd.
He headed the 21 operations department at TMI-2, from the beginning.
He had 22 headed the operations department in Unit 1.
And to the 23 extent that we have been permitted to talk about Unit 1, I 24 think we have indicated that the source of practices in Unit 25 2 originated in his administration of the Unit 1 operations O
i ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800 33MM6
82124.0 BRT 1897 t'
(_)
1 department.
And I think that we've made that point in our 2
report.
3 I think the policy of discarding tests originated 4
with Mr. Floyd, and I think the evidence indicates, and we --
S I described specifically the evidence in our report -- the 6
evidence indicates Mr. Floyd's attitude toward the filing of 7
or discarding of leak rate tests, regardless of the validity 8
of those test results.
And I think that it was his attitude 9
and his policies that permeated the operations department.
10 We did not find evidence, however, that knowledge 11 of test manipulation, which was the methods by which the 12 operators carried out those policies, rose to the level of e
13 Mr. Floyd.
14 The superintendent of technical support at TMI-2 15 was Mr. Seelinger for a period of time up through, I believe, 16 sometime in December of 1978.
And then he was replaced by 17 Mr. Kunder.
18 I think that the evidence relating to 19 Mr. Seelinger's knowledge and attitude focuses on the LER and 20 on the roundoff procedure.
What I had in mind was 21 Mr. Seelinger's awareness of the difficulty of obtaining leak 22 rate tests, tests results under 1 gpm, his awareness that 23 that problem was being addressed by rounding off the results 24 and thereby giving a broader range of results.
25 JUDGE KELLEY:
It gave you 1.49; right?
O ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 800-3 % 6646
= _ _
4
.82124.0 BRT 1898
()
1 MR. STIER:
Well, it may have given you more than 2
that.
When you look at the roundoff of results, some of them j
3 appear to have been truncated, which means eliminating 4
everything to the right of the decimal point; and some seem 5
to be rounded off.
6 JUDGE KELLEY:
You mean 1.9 could become 1?
7 MR. STIER:
Well, we are not sure.
If you look at 8
the results during that period of time, for example, you find 9
that a result of.55 gpm was rounded to zero; a result of.57 10 was rounded to zero.
But then you find a result of.94 11 rounded to 1.
We don't know how the computer was programmed, l
.1:2 so we don't know, but assuming that it gave you at least up 4
13 to 1.49 --
l 14 JUDGE KELLEYr Oh, the computer was doing this?
15 MR. STIER:
The computer was programmed to round l
16 off.
There were come original results that were rounded off j
17 by hand but then on the 19th, I believe it was, of October, i
l 18 the computer was programmed to round off or to do something 4
19 with the results.
We don't know quite how it was programmed 20 because it seemed to do different things.
21 Seelinger was aware that the rounding off was 22 being done.
The evidence indicates that Seelinger knew of 23 difficulty in obtaining leak rate test results consistently l
24-under 1 gpm.
He knew that thhy had to stop rounding off.
25 He, therefore -- we believe that there was evidence available O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 800 336-6646
.-----,=m.
--. --x
82124.0 BRT 1899 1
to Mr. Seelinger that might have suggested that'they would 2
continue to have difficulty obtaining leak rate test results 3
under 1 gpm.
However, he was transferred to Unit 1 in 4
December and we don't know what, if anything, he did in i
5 response to that knowledge after that.
6 The only additional evidence I would refer you to 7
is Mr. Seelinger's involvement in the preparation of the LER, 8
which we believe contained inaccurate information in'its 4
9 description of the events that had led to the filing of the i
i 10 LER.
The LER described conditions that might have suggested l
11 to anybody reading it that were temporary in duration and 12 were not inherent in the leak rate test itself.
13 He was succeeded by Mr. Kunder.
We looked at l
14 Mr.'Kunder's conduct most closely in connection with the l
15 period of high leakage, high actual unidentified leakage in 16 January.
17 During that period, from about the 4th through the 18 15th, actual unidentified leakage was over 1 gpm.
During 19 that period of time, Mr. Kunder became involved in examining 20 the leakage and we believe, based on some documents --
1
'21 although the evidence is not strong; there is little evidence i
22
-- but we do believe there is documentary evidence to 23 indicate that he participated in a decision that kept the i
24 plant operating at a time when we believe it should have been i
25 shut down as a result of the actual leakage.
- O ACE-FEDERAL REPORTERS, INC.
M347-3700 Nationwide Coverage 800-336-66 %
'82124.0 BRT 1900 1
How he participated, what he said, who he talked 2
to are difficult questions for me to answer because 3
Mr. Kunder's recollection is very, very dim.
But there is 4
some documentation indicating his participation in some 5
evaluation of leakage during that period of time, and that 6
evaluation of leakage, in my judgment, did not conform to the 7
requirements of technical specifications or procedures.
8 That is the evidence that I made reference to and' 9
I hope that my characterizations of knowledge on the part of 10 management earlier were not so overly broad as to give you an 11 impression that the evidence was greater than described it.
12 JUDGE KELLEY:
In this context -- and again, this 13 may well be in your report,.but there was a committee called 14 the PORC, or P-0-R-C committee whose responsibility was to 15 take on serious problems, see that they were worked out.
16 That's a very rough paraphrase, 17 Did your study indicate whether that committee had 18 any occasion to focus on leak rates as a problem?
19 MR. STIER:
Yes.
We examined the minutes of PORC 20 to try to determine if this issue had been raised.
We found 21 it was raised only in the context of the LER.
22 One of PORC's responsibilities was to oversee the 23 implementation of corrective action that was called for by a 24 LER.
25 Incidentally, the chairman of PORC was O
ACE-FEDERAL REPORTERS, INC.
202-347 3700 Nationwide Coverage 80433 M 646
.r.,
82124.0 BRT 1901 1
1 Mr. Kunder.
So he held both positions.
2 When you read the technical specifications 3
establishing and defining the responsibilities of PORC, and 4
you review Mr. Kunder's job description, it is clear that in 5
Mr. Kunder, both as chairman of PORC and as superintendent of 6
technical support, resided a great deal of responsibility for 7
assuring compliance with technical specifications.
8 We examined as best we could, and reconstructed 9
from the sparse evidence that was available, what Mr. Kunder 10 did as PORC chairman in response to the LER, and what PORC l
11 did.
12 We got the evidence in the report.
Suffice it to 13 say, I think at this point, unless you have some specific
()
14 questions, that whatever action they took was not effective 15 in changing the behavior of anybody in the operations 16 department in administering leak rate tests.
All the 17 testimony indicates that nothing happened as a result of the 18 filing of the LER, except that, perhaps, there was more 19 consciousness of the need to gets rid of test results over 1 20 gpm, than there had been before the filing of the LER.
But 21 certainly the administration of leak rate tests did not 22 improve at all as a result of the LER.
23 I hope I've answered your question.
24 JUDGE KELLEY:
Thank you.
We are coming up on our 25
! time here, now.
We did want to finish, save for those O
t i
ACE-FEDERAL REPORTERS, INC.
E3471700 Nationwide Coverage 800 336-6646
82124.0 BRT 1902 l'
questions that you.were going to respond to with 2
Christopher.
I think we have done that, have we not?
3 MR. CAPRA:
Yes.
4 MR. MC BRIDE:
The Russell-Capra questions, I 5
think you are correct.
6 JUDGE KELLEY:
Are there follow-ups?
7 MR. MC BRIDE:
I have one.
J 8
MR. RUSSELL:
I have one comment.
I referred to 9
" operations department management."
By that I meant the 10 manager of the operations department and some of the shift 11 supervisors within the operations department and some of the 12 shift foremen who were specifically responsible for conduct 13 of operations on Unit 2.
Not all shift supervisors or all
}
14 shift foremen; some of those.
And up to and_ including 15 Mr. Floyd.
16 JUDGE KELLEY:
For Mr. Stier:
Do you believe that 4
17 Mr. Havercamp or the NRC generally or both bear some 18 responsibility for problems with leak rate test performance 19 and procedural compliance at TMI-2?
20 MR. GOLDBERG:
I object to the question, Judge 21 Kelley.
Even though I know I'm not permitted to object to 22 questions to other witnesses, other than the Staff, this is 23 beyond the scope of the proceeding and the Commission's order 24 makes it clear that the Board is not to deviate from the
)
25 issues which were identified in that order.
CE) i I
i ACE-FEDERAL REPORTERS, INC.
)
3c-347 370)
Nat.onwide Coverage 80 4 336 6646 s
82124.0 BRT-1903 1
The conduct of the Staff is not at issue in this 2
proceeding and I object to testimony of any witness going to 3
the propriety or impropriety of Staff inspections or conduct 4
or actions or what have you.
5 JUDGE KELLEY:
Well, let me be clear.
Either you 6
have a right to object to questions put to other witnesses or 7
you don't.
This is a question put to another witness; 8
right?
Is there some reason for us to deviate from that?
9 MR. GOLDBERG:
I think there is.
There is a clear 10 direction to not deviate beyond the scope of the issues which 11 were identified in the commission's order and I believe it is 12 incumbent upon the Board to follow that directive and not 13 permit testimony :uch as this which is not material to any
{}
14 issue before the Itoard.
And I'm pointing that out because it 15 is taking us f ar a field and I'm concerned that if we are 16 going to have testimony from witnesses on whether the Staff 17 performed its job adequately, that in order for the record to 1
18 be complete on that, that there will have to be significant i
19 additional testimony by the Staff as to exactly what its 20 inspection practices are, what the IEE manual provides, and I 21 think it is prohibited by the commission's order.
22 JUDGE KELLEY:
Well, it's an objection, though.
23 We've already considered whether the Staff can object to l
l 24 questions to other witnesses, we said no, and now you are 25 making an objection.
It's a scope objection, I understand O
ACE-FEDERAL REPORTERS, INC.
202 347 3700 Nationwide Coverage 2 336-6646 I
82124.0 BRT 1904
()
I that, but it's still.an objection and we've already said that 2
you don't have standing to do that.
So we'll put the 3
question to Mr. Stier.
4 MR. STIER:
I have an answer.
It may be lengthy 5
and I don't want to tax the patience of the Board.
6 JUDGE KELLEY:
Could you give it tomorrow 7
morning?
8 MR. STIER:
I could give it tomorrow morning, 9
yes.
10 JUDGE KELLEY:
Let's do that.
It's 3:30.
We have 11 to stop.
Let's resume tomorrow morning with Mr. Stier at 12 9:30, and we'll be seeing you gentlemen then, probably a week 13 from Tuesday I gather; correct?
Thank you.
14 MR. MC BRIDE:
Judge Kelley, did you say 9:30 and 15 mean 8:30?
c 16 JUDGE KELLEY:
Yes.
8:30 tomorrow morning.
I 17 always mean 8:30, no matter what I say.
18 (Whereupon, at 3 :30 p.m., the hearing was 19 adjourned, to reconvene at 8:30 a.m.,
on September 19, 1986.)
20 21 22 23 24 25 i
(2) 1 ACE-FEDERAL REPORTERS, INC.
2 @ 347 3700 Niionwide Caerage 8 % 33 4 686
a-CERTIFICATE OF OFFICIAL REkORTER
- O This is to certify tnat the attached proceedings before the UNITED STATES 13UCLEAR REGULATORY COMMISSION in
'he matter of:
NAME OF PROCFEDING:
INQUIRY INTO THREE MIT.E ISLAND UNIT 2 - LEAK FATE DATA FALSIFICATION a
DOCKET NO.:
LRP PLACE:
BETHESDA, MARYLitNP O
tg DATE:
THURSDAY, EEPT$dBER 18, 1986 were held as herein appears, and that this is the briginal transcript thereof for the file of the Ur.ited Stetes Nuclear i
Regulatory Commiscion.
k f, r (sigt)
/
(TYPED /
JOEL BREITNER Official Reporter
[ekrterYAfEdatYori I
i
' O i
1
+
i 4
,,,,-e
,a-
---n---,.-a
--w---
--v-,,---,,n,-.,,,,m,-
,,,,-,---.,-r
--,n
,-,n.-m,.
n
,,e-, - - -
.