ML19249B040
| ML19249B040 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/20/1979 |
| From: | Flint J BABCOCK & WILCOX CO. |
| To: | |
| References | |
| NUDOCS 7908290308 | |
| Download: ML19249B040 (14) | |
Text
_.. _. _.
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 1.
In the Matter of:
l 2'
IE TMI INVESTIGATION INTERVIEW 3
of 4
John Flint Physics Test Coordinator, Unit 2 5
6' 7
l 8
Trailer #203 9
NRC Investigation Site TMI Nuclear Power Plant 10, Middletown, Pennsylvania lli Aoril 20, 1979 12 (Date of Interview) 13 July 5,1979 (Date Transcript Typed) 14 275 15!
(Tape Number (s))
i 16' 17 p
18 f
19 20i l
21' NRC PERSONNELi 22 o"9 346 ed 0,Connor 23 24 25j; i
1 0'CONNOR:
This is an interview with John Flint of B&W being interviewed 2
by Ed OConnor.
The date is April 20, 1979.
The time is approximately 3
15:50.
John, would you briefly state when you came on site, what your 4
role is at TMI, and pretty much what you saw as you entered the control 5
room the mcrning of the 28th of March.
6 7
FLINT:
My function onsite was to be physics test coordinator the Unit 8
2 startup.
Previous to this I had been working on writing the Unit 2 9
startup report.
At approximately 08:30 on the 28th of March, I arrived 10 at the north gate to the Island only to be stopped due to security 11 personnel restricting anyone from coming on the Island.
During approxi-12 mately the next 20 minutes to a half hour I attempted to have one of the security personnel contact the control room, and find out if my 14 services were required.
This is based on fact that I am familiar with 15l the core physics of the Unit.
At approximately 0900 permission was 16 received for me to go on the Island.
I immediately drove to Unit 2, 17 entered through the Unit 2 search trailer and went directly to the 18I control room.
Upon entering the control room, I noticed the normal 19 alarms were lit.
The was printing ou', is normal after a turbine reactor 20 trip.
I also noticed that the emergency team or the radiation type of 21 emergency was in attendance in the control room.
In talks with the 22 control room personnel, primarily, Bill Zewe, the shift supervisor, Ed 23 Frederrick, control room operator, and Lee Rogers, B&W site representative, 24 quickly established that conditions were abnormal for this type of 25 nng i
t
2 I
1 transient.
In talking with these personnel and looking at the console 2
indication in the computer print-out noted that the hot leg temperature 3
for the primary system were in excess of 620 degrees, that the cold leg temperatures significantly varied from this, they were quite a little
~
5 bit cooler.
The pressure 'vas low in the reactor coolant system.
All 6
the controi rod were on the bottom.
The indication for the source and 7
intermediate range appeared to be normal for a shutdown condition. I 8
did, however noticed the two blips on the recorders for the source of 9
intermediate range.
Ed Frederick informed me that they had thought at I
the time that they were going critical and that they had added boron.
115 At this time I informed him that in all probabilities that was the not 12 case that there had been a change in the leakage... Flux path from the 13 reactor core to the detectives and it was not in fact a case of the 14j reactor going critical again.
Looking at the recorder that prints out 15 the OTSG and re&ctor coolant temperatures er wide range 0 to 800 degrees, 16 there were several temperatures that were reading from 770 to 800 17 degrees in the hot leg of the reactor coolant system.
Since these 18 thermocouples are not normally calibrated in that range, knew the 19f temperature were high, but did not totally believe the indication.
20 Ivan Porter the Met Ed I&C engineer showed me a set up were he was 21 reading approximately 243 ohms, which converts to approximately 725 F 22 on RTD that he set up specially opposite side of the control room.
The 23 talks with various personnel in the control room, Gary Miller, and the 24 rest of the operations personnel informed me of the sequence of eventc 25 that had led up to this position.
At that time in talk with Ed Fredericks,
" 9 348
?
l k
[
3 1
1, we were convinced that night that we in fact had a solid steam bubble 2
in both loops of the hot legs.
At this time I attempted to initiate 3
the filling of steam generators to induce natural circulation or at 4
least remove enough heat that would collapse the bubble sufficiently to 5
run the reactor coolant pur.,ps.
6 7
0' CONN 0R:
John, let me ask you a question at this point.
You said 8
that some time after you got there which was 9:00 you started to sense 9
natural circulation.
From that can I infer that the control room staff 10 u to that time was not trying to establish natural circulation?
11 12' FLINT:
There was some doubit at that time as to whether they were 13 trying to establish natural circulation or not.
They were filling the 14 generation to some extent, but it wasn't clear that they were going to I
15l try to establish natural circulation.
16l 17 0'CONNOR:
That was not their goal?
18 19 FLINT:
I didn't infer that at the time.
I talked to several different 20 people and I can't remembsr exactly who said what now.
Some of them 21 said they merely filling the generators to obtain a level in them and 22 remove some heat.
Others said they were doing -- try to induce, natural 23 circulation, and some said that they were doing it in order to just 24 "9 349 2s
4 1
cool er.ough to try to run a full pump.
I think most people this time 2,
did not believe that there was in fact super heated steam bubble in the 3
hot legs.
4, 5
0'CONNOR:
And this is some time after 9:00.
6 7
FLINT:
Sometime between 9:00 between approximately 9:00 and 9:30.
8 9
0'CONNOR:
What was the control room staff trying to do at that time 10 to re establish some flow in the reactor coolant system.
11 12 FLINT:
The operators reported to me that they had attempted to run 13 the reactor coolant pumps, had seen a low current indication on the 14 pump, had not observed any flow indication of any significant degree, 15j and therefore had concluded that they were totally steam valve, or that 16l the pumps were just sitting there and cavitating and not operating 17 corectly.
18f 19' 0'CONNOR:
Were they trying to it; crease plant pressure to prevent 20 function cavity.
21 22 FLINT: Shortly after a period of time, there was an attempt to pressurize 23 the system and to collapse the bubble that way.
I explained to them at 24 the time that if the temperature we were seeing were anywhere near true 25 we could not in fact collapse the bubble.
It would be in excess of
{
" 9 350 k
i 5
1 allowable pressures even if we gave the code safetys.
The code safetys 2
would lift before we could reach this pressure and the system itself 3
was not designed for this pressure.
The 4
5 0'CONNOR:
That was because the steam bubble was super heated?
6 7
FLINT:
That is correct.
8 9
0'CONNOR: So in summary you are saying that the temperature of the 10 steam bubble was so hot that its saturation pressure was greater than 11 set points for the code safetys and even if they charged the system up 12 the code safetys would have lifted before they could have suppressed t
13 the bubole.
141 1
15!
FLINT: That is correct, a rough approximation would be say in excess of 16 3 thousand lbs.
And a code safety is set for less than 2400 lbs - 2500 17 right in that range.
So there would be no way that we could have done 18 it. About that time for aporoximately an hour and a half they were 19l dumping out of the atmospheric dump on the A generator.
We then received 20 a call that stated the Governor requested that that valve be closed 21{
immediately the request was in the form that made it a demand.
Gary 22 Miller complied and was at that time was preparing to go talk with the 23 Governor.
" 9 351 24 25 i
i l
l I
6 e
1 0'CONNOR: Do you recall why you were lifting the atmospheric dumps?
2 3l l
FLINT: We were using atmospheric dumps to try to steam the A generator 4
remove enough heat that we could use it as a heat sink and collapse the 5
bubble that way.
Induce natural circu' tion.
6 7
0'CONNOR: The vacuum in the condenser was not available at that time?
8 9
FLINT: No.
We had lost the auxiliary boiler and they were attempting 10 to regain it I believe they gained one boiler porportioned time lost it 11 and was quite some time before we got both boilers back.
12 13 0'CONNOR: What did you use as a the heat sink once the plant staff 14 st.opped dumping steam through the atmospheric dumps at the request of 15l the Governor?
16!
17 FLINT: Some of the heat removal was dumping through the electromatic 18{
relief valve into the building, some of it was taken away with the 19 filling of the generators.
About thic time the level in the B generator 20I was also increased.
They did dump some down into the condenser on the 21 bypass valves but that was not, you know in that portion of time it is 22 a little difficult to remember exactly which sequence.
That was their 23 only method of heat removal at that time.
Firnlly, verified that in 24 fact that we could not collapse the bubble.
The pressure was then 25 reduced again.
At that point in time it was tried to start the reduce i
onn 7r JJ2 i
[
7 1
the pressure to the point where we could get down on the core flood 2;
l tank.
There was some concern that the core was not in fact covered.
I 3
felt that it was since the leakage flux as seen by the source range 4
channels appeared to be in a normal range for a shutdown condition 5
considering the length of time we had been shut down.
However as a 6
precaution, the pressure was reduced the core flood tanks then slowly 7
came into the system and since we did not see a significant pressure 8
change or rapid dropping of the core flood tank levels where they feed 9
directly into the downcommer annulus or the reactor vessel itself, from 10 that we presumed that in fact the core was covered and that we were 11 essentially feeding in an either bypassing the core or just merely 12 coming out of the core coming back through the legs and dumping out the 13 pressurizer for cooling water.
We were not getting anything up through 14 the loops themselves.
We did have a high pressure indication in the 15 reactor building somewhere along this length of time.
We had the alarm i
16' for containment isolation about this time there was a double thud at 17 the time was not thougnt to be from the building.
At least, I personally 18 did not think it was from the containment building. I thought that it 19 was the ventilation damper cycling it.
It was very close to that 20 sound.
And since we had been in and out of respirators due to levels 21 in the building, I just thought somebody had cycled the ventilation 22 dampers again and related it to that.
The reactor building spray pumps 23 came on a normal actuation sequence for high pressure in the building.
24 They were bypassed and later shut down.
During this portion of time we 25l finally managed to get the A leg temperature decreased to where we
)
i OD9 353
?
I t
8 1
could see the hot leg temperature on scale we the operator then elected 2
to try to go over to anoter leg and collapse the B side and in doing so 3
he lost the gain we had made on the A side.
The A side hot leg again 4'
went off scale.
Somewhere along in here we finally did see some changes 5
of decrease where the hot leg was coming down.
The cold leg tempera-6 tures were going up indicating that we were getting some type of heat 7
transfer across there.
Later in the afternoon we were requested to 8
once again try to collapse the bubbles by increasing the pressure.
9' Once again we tried to explain that we had tried that that morning gone 10 to approximately 2,000 lbs and it had not work.
That the temperatures 11 at this time were still not significantly different from those earlier 12 in the day but we could not persuade them that this could be done.
13 Being unable to convince offsite personnel that the pressure increase 14 would not collapse the bubble, we went ahead and increased pressure of 15 approximately 2300 lbs charging with the makeup pumps.
We gradually 16[
throttled back on one pump until we achieved this and held it for a 17 considerable period of time.
This did not collapse the bubble.
- However, 18!
by this time due to early indications of the whole leg temperature 1bl increasing and the hot leg temperature coming down the A loop, we felt 20 that we had enough water around the pump at this time that we could run 21 one of the A loop pumps.
The A loop pumps were chosen for two reasons.
22 One we had a better chance of establishing circulation there since that 23 generator was not bogged up and No. 2 the pressurizer spray line comes 24 off there which would give us mixing in the loop and help to bring the 25 temperature down a little more rapidly plus allow us to get better l
" 9 354 i
9 l'
boron samples if we had to take them out of the pressurizer system.
We 2
bumped the reactor coolant pump 1A after trying to start the 2A and it 3
did not start.
We went to IA ran it for approximately 10 seconds this 4l was long enough to indicate to us that the current appeared to be in 5
the normal range and that flow was in fact verified to be in the loop.
6 We then shut the pump off, let it sit for approximately 15 minutes.
7 The 15 minutes was based on the fact that during normal startup due to 8
surge currents you do not want to start and stop the pump anymore 9
frequently than this.
10 11 0'CONNOR: Excuse me John, why was the pump stopped after just a few 12 seconds of running?
13 14 FLINT: We were not certain that what would happen was we would pull the 15{j steam bubble down into the pump and cause severe cavitation or possibly 16i with a rapid pressure change and of mass density change fail the seals 17:
in the pump.
We therefere just wanted to verify the moving water get l
18 some mixing once again turn that water over collapse the bubble a 19l little further and then run the pump once we were verified that we felt 20 we could do without any significant problems.
21 22 0'CONNOR: If you didn't have the, I withdraw that question.
23 24 o c,-)
7r-25 J 3 ')
I
10 1
FLINT: Since the pump had appeared to be normal.
The pump was restarted 2;
and we saw the expected pressure decrease and flow rate that we expected.
3 The pump current was normal at approximately 600 amps.
The pressure 4
decreased was due to the collapsing of the steam bubble and we saw the 5
A loop temperatures on the hot leg come down and the cold leg temperatures 6
going up indicating we did have flow and mixing across there.
The 7
smooth trace on the flow plus the pump current indicated we did not 8
have any serious steam bubble problems in fact we were pumping primarily 9
water if not entirely all water.
That period of time the pump was left 10 running to mix the system up.
In this period of time we started drawing 11 a vacuum on the condenser and were preparing to get the aux boilers 12 back on line so we could start feeding back into the condenser normal 13 method for cooldown.
After a period of time we, fairly late in the 14' evening by now, established the bubble in the pressurizer, we had 15 brought the temperatures in the B loop down on scale and we were getting 16' some small amount of circulation through the B side as well.
One 17 point that I did forget to mention earlier was that shortly after I 18 came in we also started calling up on the computer the incore thermocouples 19 attempting to establish what had happened in the core.
Many of them i
20l were indicated question marks which indicated they were greater than i
21 their 700 degree range only one or two seemed to indicate there were in 22 fact bad.
These temperatures were monitored for the rest of the day 1
23 during this portion of time to follow what was happening with the core.
24 9 356 25 I
r f
11 1
0'CONNOR: John, does B&W, were you functioning as a B&W employee during 2
this event, or were you just there providing advice.
3 4'
FLINT: I was providing primarily there providing advice not as a B&W 5
because I was still on master service's to GPU and felt that I was 6
filling the role of advice and consultation to the customer from that 7
standpoint.
I was however interacting with Lee Rogers providing him with a physics parameters and...
9 10 0'CONNOR:
Where was Lee?
11 12 FLINT: Lee was in the control room during all this portion of the time 13 frequently he would be in the meetings with George Kunder, Gary Miller, 14 Mike Ross, Bill Zewe and the rest of them establishing what they were 15 going to do next what the conditions were.
Well I was primarily working 16l with the operators themselves monitoring parameters and helping them 17 try to assess what the actual situation in the core in the primary 18f system was.
19 20 0'CONNOR: I see.
Was B&W in Lynchburg communicating with Lee Rogers?
21 l
""9 357 22 23 24 25 i
I l
f'
12 1,
FLINT: There was an open line to B&W Lynchburg on a continuous basis as 2
well as an open line for the Nuclear Regulatory Commission to communicate 3
with their personnel and open lines direct over to the observation 4i center and to other personnel on the outside for advice from the GPU 5
and Met Ed facilities.
O h
7 0'CONNOR: Were you in a position to observe how the emergency plan was 8
being carried out?
9 10' FLINT: During the portion of time that I was there it seemed to be 11 extremely well organized, personnel knew what they were required to do, 12 there was excellent communication between Unit 1 control room personnel 13 which was set up as an emergency control center, the Unit 2 Control 14 Room, the offsite personnel and I would say that all aux operators, NRC 15 personnel, HP that were associated with this, were almost as if this 16!
was a drill rather than the real thing since they were functioning 17 extremely smoothly during this course of time.
18 19l O' CONNOR: How would you describe the atmosphere in the control room 20 during the period of time that you were there?
21 22 FLINT: I would say that the personnel were extremely calm and well 23 organized.
Late in the afternoon it had a tendency to get a little 24l noisy in there when the immediate problems were over we have a large 25 I
i i
nr9 358 i,
't3 1
nJmber of personnel irr the control room by this time we had a large 2f number of NRC personnel, we had at least a shift and a half of personnel 3
plus additional personnel from Unit 1 and so forth.
It got a little 4
noisy from that standpoint but it was still very well controlled and 5
well organized.
6 7
0'CONNOR: People were not running around not knowing what to do.
Or panicing?
9 10 FLINT: No, no panic, especially since frequently announcements were 11 made over the page system to let all personnel know what was going on.
12 Unnecessary personnel were released either to their homes or were 13 mobilized over at the obsarvation center for assistance if necessary.
14 The only personnel that remained onsite were those that were absolutely 15l required and since they were well informed of what was occurring and 16l since most people were in either one of the two control rooms, there 17 was no problem of people appearing not to know what they were doing.
18{
19 0'CONNOR: Are there any other observations that you would like to get 20 into the record?
21 22 ILINT: No.
I do not believe I have any other observations or remarks 23 concerning the incident.
24 "9 359 3
l l
l 1
i;
14 1
1 0'CONNOR: Well, thank you very much John.
Appreciate it.
2 3
4 5
6 7
'"9 360 8
9 10 11 12(
13 14, i
15!
16i 17 1
18' 19l l
20 21 22 23 24 25 I
i t