ML19249B091
| ML19249B091 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/30/1979 |
| From: | Creswell J, Hunter D, Mark Resner NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III), NRC OFFICE OF INSPECTOR & AUDITOR (OIA) |
| To: | |
| References | |
| NUDOCS 7908290640 | |
| Download: ML19249B091 (33) | |
Text
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 6
l In the Matter oG 2
IE TMI INVESTIGATION INTERVIEW 3l of James R. Floyd, Supervisor of Operations l
4!
Si 6
7H 8i,'
i Trailer #203 9
NRC Investigation Site TMI Nuclear Power Piant 10 Middletown, Pennsylvania lli May 2, 1979 12 (Date of Interview) 13 June 30, 1979 (Date Transcript Typed) 14 91 15!
(Tape Number (s))
16l 17 18 I
JO e0h 20 NRC PERSONNEL:
~22l James S. Cresweli f
r Dorwin R. Hunter
/
1 Donald C. Kirkpatrick 23{
Mark E. Resner 24!
894 169 25)
{
i i
l
(
1 RE SNER:
The following interview is being conducted of Mr. James R.
2
-nyd.
Mr. Floyd is a Supervisor of Operations, Unit No. 2, Three 3
Mile Island Nuclear site.
He is employed by Metropolitan Edison.
The 4
present time is 4:04 Eastern Daylight Time.
Today's date is May 2, 1979.
This interview is being conducted in Trailer 203, which is 5
6 1 cated just south of the south gate to the Three Mile Island site.
7 Individuals present for this interview are Mr. James S. Creswell.
Mr.
8 Creswell is a Reactor Inspector of Region III of the Nuclear Regulatory g'
Commission.
Also present, Mr. Darwin R. Hunter.
Mr. Hunter is an Inspector Specialist with Region III of the US Nuclear Regulatory 10 Commission.
Also present, Mr. Donald C. Kirkpatrick.
Mr. Kirkpatrick is a Nuclear Engineer at IE Headquarters, Bethesda, Maryland. The moderator of this interview !, Mark E. Resner, and I am an investigator with the 07fice of Inspector and Auditor, the US Nuclear Regulatory Commissien Prior to this interview, Mr. Floyd was provided with a 151 16l copy of an advisement document, which addressed his rights concerning information to be obtained regarding the incident at the Three Mile Island.
This document apprised Mr. Floyd of the purpose of this 181 investigation, the scope of the investigation and the authority by 19{
l which Congress authorizes the Nuclear Regulatory Commission to conduct 20t I
this investigation.
The documcat consists of two pages, and on the 21l second page Mr. Floyd has indicated, he answered three questions.
For 22l tie record, these questions are (1) Do you understand the above? Mr 23 Floyd has indicated his answer to be in the affirmative.
Is that 24 correct Mr. Floyd?
25j 894 170 s
f
I 2
i t
ilt FLOYD:
Yes sir.
2, 3
RESNER:
Question No. 2.
Do we have your permission to tape the 4
interview? Mr. Floyd has indicated affirmative to this question.
Is 5
that correct Mr. Floyd?
61 7
FLOYD: Yes sir.
8 RESNER:
Question No. 3.
Do you want a copy of the tape or a transcript.
g ygj Mr. Floyd has indicated that he would like a copy of the transcript.
Is that correct Mr. Floyd?
11 l
12!
es sir.
13 141 RESNER:
Thank you.
If you will, Mr. Floyd, for the record, would you please give us your background in the nuclear industry.
17 FLOYD: In 1959, I went to Advanced Nuclear Power School in Mare Island, i
I was an enlisted man at the time, but I completed the 19j i
officer's program there.
To this day, I am probably the only white 20!
hat that's ever done that trick.
I followed that up with six months of training at the prototype at A1W.
Stayed on as a instructor in instrumentation and control for a little over two years, put in a year on the Enterprise before I got out of the Navy in January of 1964.
At 24 that time, I had completed three years of my chemical engineering bhk l[l I
i 4
i l
3 1,
degree so I went to Columbia University and completed that degree in i
2!
chemical engineering, being much too late to change to nuclear at that I
31 point in time.
While at Columbia, I was emp,.yed half time as an t
4 assistant to Dr. Marconian in the Nuclear Engineering Department, 5
doing basic research in the fission process.
I graduated from Columbia 6
in 1965, came to work for Met Ed in September of '65, out at Saxton.
7 I was there five months and acquired my senior operator's license at 8
Saxton Nuclear Experimental Reactor.
I became Supervisor of Operations g
at the same time, till late 1968, when I came to Three Mile Island.
I 10, was Supervisor of Operations on Unit 1, took it through its startup 11' and test program and first year of commercial operation, in that position.
At which time I came south to Unit 2 and I have done the g
13 same trick with Unit 2, except we didn't get through the first year of 14l commercial operation.
That probably brings us up to date.
i 15!
RESNER:
Thank you Mr. Floyd.
At this point I'll turn the questioning 16 g!
over to Mr. Creswell.
181 CREShELL:
,g; Jim, I'd like, if you could, to describe where you were on l
the day of March 28, when the event occurred.
20l 21.
FLOYD:
For better or for worse, I was in Lynchburg, Virginia, undergoing my annual requalification on the simulator.
24 894 172 25 t
4 1
1 1l CRESWELL: Okay, Jim.
How long had you been there at the simulator?
2b I
3 FLOYD: Since Sunday night.
4 Si CRESWELL: That would have been, what, four or five days prior?
'l 7l FLOYD: was Wednesday, so - Sunday, Monday, Tuesday, Wednesday - four 8
days.
9 CRESWELL:
Okay.
When did you first become aware that there was some 10 77l sort of abnormal condition existing at the plant?
12 F_LOYO:
At breakfast that morning, about 7:00 o' clock.
One of the 3
j gentlemen with me had been in communication with someone here at the i
site, who had told him that the safety valves on Unit 2 had blown for 15l two hours that morning from about 4 to 6.
So as soon as we finished 16f breakfast, we went over to B&W offices, and I got on the phone - it must have been about 7:30 or twenty of eight - to the island.
I 18f got through to the Unit 2 control room and I got a few sketchy pieces 191 i
of information from, I believe the first phone call was with Kenny 20l 21;l Bryan the shift supervisor.
I took the several pieces of information which I had and tried to reproduce the initial portion of the transient, through and including the introduction of emergency feedwater to the steam generators.
My information, at that timr, said they were about 24 ten minutes late bringing in the emergency feed, that they had the 25 d9i }[3 l
I
5 l
1 reactor coolant system pressure down around 1000 pounds or 1200 pounds, 2
and I tried to simulate that on the B&W simulator.
I called back into 3{
the plant later in the morning, it must have been in the area of 9:30.
4f I also found out at the earlier phone call that there was a number on 5
the reactor building gaseous activity... and I don't remember the 61 value of that number, but based on the value of that number, I estimated 7
at least a third of the cladding had failed, or an eighth of the 8
cladding had failed.
I called back in for more information about 9:30 g
in the morning and found out that the electromatic relief had been stuck coen.
So I then introduced that perameter into my simulation.
10 In n ne 11 these was I able to drag the simulator down to 1000 pounds or 1200 pound pressure.
At the time, I guess we were outside the simulation capability of the machine, and it would not recognize a 13 bubble in the head or in the hot legs and cause the pressurizer level to respond as it does respond.
So I spent most of the day, up until 15 4:00, working with the simulator, varying the possible time delay between the two feed pumps tripping and not knowing they had gone, effectively, simultaneously,.
sticking the electromatic relief valve 181 I
open for various lengths of time, bringing in emergency feed at different 19l flow rates at different points in time, trying to force the simulator down to its knees at a 1000 pounds.
I was never successful in doing i
that, however.
Those traces were preserved I think Bernie Smith 221 still has them.
23 894 174 24 25 i
i i
{
6 1,
CRESWELL:
Bernie Smith? Jim, who is Bernie...
i 2!
3 FLOYD: He is the shift supervisor in both units.
4 5
CRESWELL: Now you mentioned that you received a call while you were at I
6l breakfast.
7 8
FLOYD:
No, I received information from another gentleman in our g
group.
The group that was down there was Bernie's crew, consisting of 10 himself, Dick Hoyt, the shift foreman, and three control room operators, lli and one of those gentlemen brought the information to breakfast that f
12 the safety valves had been stuck open for two hours, which was enough to trigger me to call the plant.
13 14 CRESWELL: I see.
Do you recall who that gentleman was that had received 15 the information, that you had breakfast with?
16 l
17l FLOYD: I believe it was Dick Hoyt, but I wouldn't swear to that.
8 19l I
CRESWELL:
Does Dick work for Met Ed?
20' 21 FLOYD:
Yes.
23 CRESWELL: How many people did you have up at the simulator with you?
25 894 1/5 t
f I
l 7
1 FLOYD: I named them - myself, Bernie Smith, Dick Hoyt, Ray Brewer, 2.
Harold Hartman, and the CR0 trainee.
3 4,
CRESWELL:
I see. What was his name?
5 6
FLOYD:
I can find out if it is important to you.
7 8
CRESWELL:
Okay, we can check into t!.at later.
So the gentleman with g
you at breakfast had received a telephone call, and I believe you said 10 that a safety valve had hung open for two hours?
11 FLOYD:
12!
That is the word he got on the telephone from a non-technical person at this end.
13 1 I
14!
CRESWELL:
Do you know who that individual was on this end?
15 I
16 FLOYD:
No.
181 CRESWELL:
Okay, so then you proceeded over to, what, the office 19!
building, to make your call?
21, FLOYD:
Yes.
23 CRESWELL:
And I believe you named the gentleman that you talked to.
25 874 176 i
i
{
8 1
FLOYD:
Kenny Bryan.
2 3
CRESWELL:
He was the shift engineer?
4 5
FLOYD:
The shift supervisor I talked to him on one of the two calls.
6j I made two calls back to the site where I got into the Unit 2 control 7
room, one at 7:30, one at effectively 9:30 or 9:15, somewhere in that 8
region.
0o one of those two calls I talked to Kenny Bryan, I don't 9
know who I talked to on the other one any more.
10 CRESWELL:
Did you get the impression that Kenny was in the control 11 room?
12, 13 FLOYD:
Yes he was, because I dialed straight into the control room.
14 15 CRESWELL:
16l Okay, did you have any trouble getting in?
17 FLOYD:
No, not at that hour of the morning.
g i
19 CRESWELL:
And this would have been about time?'
20 21 FLOYD:
7:30, and between 9:15 and 9:30.
Now, I have some u.'isted numbers in my wallet that.
24 894 1/7 25 l
5
9 1
CRESWELL:
I see.
I 2'
3 FLOYD:
That aren't the public domain.
So later in the afternoon B&W 4l called me up to their board room to try to call some information into t
Sl the plant with those unlisted numbers.
At that time in the afternoon i
61
... I don't remember whether it was three or four or five... but 7
somewhere in that region, while I was in the board room, I could not g
get into the Unit 2 control room.
So the information that they wanted g
relayed I called into the Unit I control room to a Shift Supervisor 10 ver there by the name of Rich Hutchinson, and I told him who I was and it was very important to get this information to the Shift Supervisor 11l g
in Unit 2 and to send a courier over to do that, to hand carry the information.
So later in the day, it was impossible to get into the g
y Unit 2 control room, so I went in through the Unit 1 control room, Couriered to Unit 2 control room.
l 16i CRESWELL:
Okay.
When you went to the board room there at B&W, do you recall who the individuals from B&W were that you talked to?
18{
19l i
20t
~FLOYD:
There had to be 35 to 40 people in the room.
21 CRESWELL:
Was there one individual that appeared to be in charge?
23 FLOYD:
Yeah, there were probably a couple of vice presidents there.
I was probably introduced to them but I don't remember who they were.
l I can probably remember a couple of people in the room.
I 894 1/8 i
i
10 1
CRESWELL:
Well that's fine 2
3 FLOYD: Who could get back to tell you who was there.
Our previous 4
project manager for B&W, Grant Ward, was my primary contact with the 5
upper management there.
6 7
CRESWELL:
Okay.
8 g
FLOYD:
Grant brought a group of people down in the morning - the 10, 8:30, 9:00 o' clock tyoe time frame - and I sat them down in the classroom 11!
and I briefed them on what I knew and what I suspected, from what I g
knew had happened.
And then I heard no more of them until the afternoon 13 when they summoned me to the board room to try to get some information back into the Unit 2 control room.
p l
15i CRESWELL:
The briefing that you conducted, do you recall... and I 16 understand that you are getting preliminary information out of the g
site... but do you recall what the subjects were that were discussed, what you felt at that time, at that point in time?
g 20 FLOYD:
I knew that we had had a loss of both feed pumps.
22 CRESWELL:
That's the emergency feed pumps?
24 894 i79 25 l
I f
11 1
FLOYD:
No, the main feed pumps, which...
2' 3
CRESWELL:
Oh, the main feed pumps.
4 5
FLOYD: Which promulgated the whole trip, leading to the reactor trip, 61 high pressure injection, and high activities in the reactor building, 7
which of course, indicated failed cladding to ne, at least.
That I 8
knew.
I did not know the real reason the feed pumps tripped, that, gl that I was attempting to simulate the course of the transient from the i
10 feed pump trip to the failed fuel, and that is what I was unsuccessful 11 in d ing n the simulator.
12 CRESWELL:
I see.
13 14 HUNTER:
Jim, you mentioned about the auxiliary feed pumps being cut 15 16l in within ten minutes.
Were you aware that at that time, in this meeting, that the auxiliary feed pumps were in fact not cut in for ten 7
minutes?
g i
191 l
FLOYD:
Yes.
It was approximately ten minutes, was the number I was 20 21,l given on the phone. So on the simulator I varied from, like through six, seven, eight, nine, ten, eleven, twelve minutes on different runs, trying to see what difference it would make to the machine as one variable.
Then after I had gone through that evolution, I simultane-ously failed open the electromatic relief and went through the same i
e i
j 894 180 p
l
12 f
i 11 evolution, changed another one parameter at a time to see how it would 2
affect the plant.
But in no way was I able to bring reactor coolant 3
system on the simulator down to 1200 pounds.
The other thing which 6
4f was an anomaly - I knew the simulator would go up above 2500 pounds on 5l the reactor coolant system pressure, having run this same transient on l
6l it before.
Simultaneous loss of both feed pumps leads the simulator 7
to something greater than 2543 pounds, which was the max range of the 8
best instrument we had available to us there.
And it was above that g
point for about two seconds.
How high it got and came back down, I d n't know.
When I came back to the site, one of the first things I 10 11, was interested in was, were we above 2500 pounds? And I can find no evidence to that effect.
12 13 KIRKPATRICK:
Jim do you know if the simulator allows for water to be vented from the relief valve, or is the simulator relief valve value based only on steam?
17 FLOYD:
I do not know, but I never got the pressurizer solid on the simulator.
I understand since then, the simulator has been changed 19!
and now gives a very good representation of what went on here.
But of course it was programmed to give that representation because this is the way this B&W plant behaves. So they didn't model SMUD in this case, they modeled Three Mile Island 2.
And of course, the modelers can make it do anything they want it to do.
I have been very intimate 24j with the simulators since its initial inception.
I was in on the 25l l
8?4 181 l
I 13 1
design of the machine originally, before it was even a cardboard i
2j mockup.
I have more operating hours on the simulator than I have on
(
3j these two units together, actually turning controls, because it's i
4 something I very rarely do in the control room. I direct my control 5
room operators, but I rarely turn the controls myself.
And I have 6
been outside the simulation before, and I usually recognize when I am 7
utside of it.
So I spent the eight hours that I had rented the 8
machine for that day for my training, trying to simulate the accident, g
then turned it over to B&W, and proceeded to get some supper before coming home.
10 11lg CRESWELL:
Jim, how well... this is your opinion... how well was 13 the simult. tor in its modeling of TMI 2 for normal training purposes?
14!
15l FLOYD:
Beautiful.
i 16!
CRESWELL:
Okay.
You mentioned at one time, you had estimated a certain percentage of the clad had failed.
How did you arrive at that 181 type of conclusion?
20' FLOYD: That is a very good question and B&W had the same question in t
their minds.
22l Our emergency plan calls for us to declare a site emergency if we have
.. or a general emergency, we can declare a general emergency if we have a site emergency and 8 R per hour on the dome 24 monitor.
That 8 R per hour number is one I derived back in Unit 1, 25j 894 182 l
14 and we carried it over to Unit 2.
I derived that number based on gap lt' t
2 activity, as I recalled, at the simulator.
And I had a number... a 3
number was transmitted on the phone to me for the dome monitor, but I
4l based on the dome monitor, which the number was transmitted was somewhere 5
around 80 or 90, or 8 or 9 times 10 to the fifth counts per minute off I
6l the recorder, and until I throw the four inches of lead in front of 7
the detector, all of a sudden I have a number which is horrendously 8
high and I cannot believe, which is why I went to the only other g
activity number I had, was the gas monitor on the reactor building.
And I took a sensitivity for that gas monitor and I calculated it 10ly backwards to find out what was in the building, and based on that iy activity and what I remembered from calculating 8 R per hour, I don't know, five, six, eight years ago, I said at least an eighth of the 3
clad had failed at that point in time.
It was based solely on what I remembered from numbers, not on anything that I had with me to look 15j up.
16 17 CRESWELL:
Okay.
i 18(
19l l
FLOYD:
I could not have documented justification why it was one 201 eighth instead of two eighths.
21l 22 CRESWELL:
I am just trying to get a feeling.
23 24 25!
l 894 183
?
I l
15 FLOYD:
To me it was a large fraction of the core that the cladding 1,
2l had failed in, 3
4 CRESWELL:
Okay.
I would like to go back to the time that you were in 5
the board room there at B&W, when there.
I think you estimated 6
30, 35 people in there... and you gave them a restricted telephone 7
number to reach the site with.
Did they establish somewhat permanent 8
communications on that line?
9 FLOYD:
No, I did not give them the telephone number.
I dialed the 10 11 number for them on a squawk box, and I was unsuccessful in getting through to Unit 2.
So that is when I dialed into Unit 1, had the 12 messenger run over to Unit 2.
I had no more than rung off from that 13 phone call when our squawk box rang, and one of the B&W reps from the g
site, Greg Schaedel was, I believe, at the Visitors Center.
He was 5
not on the site, but he had just gotten information fr9m the site.
And he was of course desirous in getting this back to B&W's offices.
He came on the phone and'he told us the current plant status, as best 18l he understood it, and claimed that they thought they had collapsed the 19!
bubble in the hot leg, the A hot leg, in particular.
But the numbers he gave us on the phone are above the saturation curve.
I did not have a set of steam tables with me, but the temperature and the pressures 22 he gave me are just in the superheat region.
And so I lost my cool a little bit and ended up the conversation by telling him to look at 24 their steam tables because, if those temperature and pressures were 25!
(
l 891 184 1
f
16 1l true, they had not in fact collapsed the bubble in the A hot leg.
I 2
While the discussion was going on, one of the other B&W engineers down 31 near the end of the table apparently hsd a steam table with him and he 4
confirmed that we were still in a superheat region.
I passed to Greg 5
the information that I tried to send through the Unit 1 control room, 6
I asked him to try to get back in, and after getting off that phone 7
conversation I called back into the Unit 1 cor. trol room to Rich 8
Hutchinson again and reiterated the importance of getting my previous gl message into the Unit 2 control room... that message was get at 10 least four to five hundred gallons per minute of HPI water going into that vessel.
11 I
12, CRESWELL:
How did you arrive at the four to five hundred gallon 13 figures?
y l
15j FLOYD:
16 Straight B&W recommendation, which seemed reasonable to me, so I transmitted it.
g 18j CRESWELL:
Okay, do you recall who made the recommendation, frcm B&W?
20t l
FLOYD:
No, they had their best minds in Lynchburg at the table.
I 21' don't know who, what the man's name was that scratched the back of the i
envelope.
894 lg 24 2sj J
I l
l t
17 1!
CTtESWELL:
How long would you estimate that you and this group of 2
people sat at the table in communications with the control roum!
3 4
FLOYD:
We were never in communication with the Unit 2 control room.
5 I was only in communication with the Unit 1 contro'l room for two 6
periods of time long enough to transmit che information... a matter 7
f several minutes on each call.
We were in communication with Greg g
Schaedel at the Visitors Center, a B&W employee over there..
he is g
the second in command normally here at the site for B&W so I know Greg very well.
We were on that conversation for, oh, probably a half hour 10 or 45 minutes.
I was probably in the board room in the area of two yy Ig hours, total.
13 CRESWELL:
Did you have any indication that anybody from B&W at the g
site was in communications with the corporate offices, that is, from 15 the Unit 2 control room?
16; l
17l FLOYD:
No, they were not.
They wouldn't have been pinging on me if 181 they had an open line.
191 20 HUNTER: Right.
The people who were asking you for information, it was very clear then that they had no communication with the site at that time... Established at that time?
23 24 25 894 186 i
t
18 1
FLOYD:
No continuous communication.
2 3
HUNTER:
Okay.
At what time did the word get back to the site, approxi-4.
mately, that they should in fact get four to five hundred gallon per 5
minute high pressure injection to the flow? To the core?
61 7
FLOYD: Well, let me see if I can sort out the times in my mind.
8 Right after lunch, I went up to the second floor on another technical 9
problem which was a low priority item I found out when I got up there, 10.
at least I thought it was.
I come back down to the simulator, I 11 continued simulating.
It was probably after two o' clock when I went to the board room.
It was probably about four or five o' clock when I 12 left the board room.
13 14; HUNTER:
Okay and in your opinion... okay, we'll take a break.
15 l
16i RESNER:
Excuse me gentlemen.
At this time we'll take a break to Change the tape.
The time is now 4:32 Eastern Daylight Time.
19!
RESNER:
20 This is a continuation of the interview of Mr. James R.
Floyd.
The time is 4:35 Eastern Daylight Time.
I will now turn the interview over to Mr. James Creswell.
22 23 CRESWELL:
Okay, let's go back to the simulations that you tried on 24{
the simulator and the sequence that you went through.
I believe that 25l l
894 187 l
l t
I f
11 19 1!
you said that you tried delaying the emergency feedwater initiation, I
26 you tried failing the power operated relief valve open, you said that 3
the minimum pressure that you could get would be something like 1200 4l pounds.
5 6
FLOYD:
I never got that low.
7 8
CRESWELL:
Never got that low?
9 FLOYD:
In fact, I had trouble getting high pressure injection to come 10 I"'
11 12 CRESWELL:
Which would come in at 1600 pounds?
13 14, FLOYD:
Yes, well, in the simulator, the way it was set up there was 15 1500, but the same difference.
17 CRESWELL:
With the power operated relief valve...
18l.
19l FLOYD:
Even with it open.
21, CRESWELL:
You couldn't get down to, or had difficulty reaching the engineered safety features actually.
24 25i 894 Igg i
i i
i
20 1,
FLOYD:
And the reason, of course, was that I didn't leave it open for 2
two hours.
3 4
CRESWELL:
Okay.
5 6
FLOYD:
I was working in real time and I was just trying to simulate 7
up until the timer brought in emergency feedwater.
I thought it was a 8
combination of the electromatic relief and emergency feedwater with g
the elevated Tav sucking the bottom out of the pressurizer.
But in 10 fact, it was the electromatic relief just took all the hot water out 11 of the pressurizer eventually, and that is what led to the depressurization.
12 CRESWELL:
13 How long would you estimate you left the power operated relief valve open?
g l
15i n
ran any e ransients more man ab ut 16 twenty minutes, individual transients, and that wasn't enough time to drain all the hot water out of the pressurizer.
Plus, they may have l
had more heat in their pressurizer than we have in ours, I am not sure 191 of that.
20 21 CRESWELL:
Okay, how did the pressurizer level respond whenever you tock out the, or didn't have emergency feedwater initiation?
894 189 25j i
I
21 1!
FLOYD:
It just responded with a rapid rise in T average.
In the runs l
2 I made, Tav got up in the area of 605 to 615, by the time I bring in 3
emergency feed. And again, I am not sure that our cavitating venturies 4
are simulated on the simulator, so I may have been able to bring in 5
more emergency feedwater than would be possible in this plant, which 6l of course, would suck the pressure down that much faster.
I 7
8 CRESWELL:
The cavitating venturies are used on this plant to measure g
emergency feed water flow?
10' FLOYD:
No, to restrict emergency feedwater flow.
So that on the 11 y
steam break in the reactor building, you do not get too much inventory in the building too fast.
13 14 Okay.
15 i
16i FLOYD:
They are just downstream of the emergency feed pumps.
We do g
- ""* ** "9'"'Y 18 19l CRESWELL: About what time period in the morning were you doing these 20l I
types of simulations?
211 22 FLOYD:
As soon as I got off the phone with the site.
By 8:00 o' clock I was on the simulator and I carried it through into the afternoon.
So it was several period of time I left the simulator to talk to the i
871 190 I
22 1{
B&W group when they came down stairs, when I went upstairs right after 2
lunch, when I went up to the board room.
Bernie Smith continued on 3
with the aimulations.
We also varied the tripping of the two, the 4
time of tripping of the two main feed pumps.
I delayed them on fifteen 5
second intervals, tripping one and tripping the other one simultaneously, 6
then delaying the second one for fifteen, thirty, fortyfive seconds, 7-up to about two minutes because I wasn't sure thatboth feed pumps went 8
simultaneously here.
After I came back I find out that they did go g
simultaneously and I wouldn't have had to run a lot of those runs, but 10 my information was very limited at the simulator.
11' CRESWELL:
Understood.
Did you try tripping off any reactor coolant pumps?
13 14!
I FLOYD:
No.
151 16i CRESWELL:
Did you have any i.7 formation that the reactor coolant pumps had been tripped off at the plant?
t 181 19f FLOYD:
Not until I went to the board room in the afternoon.
20!
l 21l CRESWELL:
Okay.
At that period of time could you describe what the conversation was like in the board room?
23 24 25 b)i l9l r
t i
l l
23 1!
FLOYD: Well, the primary purpose of getting me there was to get the l
2 four to five hundred gallon a minute high pressure injection flow into 31 the core, to get that information into the Unit 2 control room.
So i
4 that was the initial; introduced to a half a dozen of the 30 or 40 5
people that came in; got on the phone, unsuccessful into Unit 2 control 6l room, got it into Unit 1, hung up from that; Greg Sheddel called, l
71 extended conversation.
As soon as I rang off from that, I went back 8
to the Unit 1 control room... well, I tried Unit 2 again, and gl couldn't get in on three or four different numbers.
So I went back to 10 Unit 1 control room and reiterated to Rich that that information had 11 to get over to Unit 2.
By that point in time, they probably had g
collapsed the bubble in the hot leg.
13 CRESWELL:
Okay.
14 i
15j FLOYD:
The other piece of information which was discussed at some 16 length in the board room was the hundred amps on the running pump.
It was very important to us there, whether it was 100 amps, 125 amps or 140 amps.
100 amps happens to be no load on the motor, which would indicate a sheared shaft; 125 is probably the seal, the extra current caused by the seals on the pump; and 140 or 145 would be indicative of the bearing in the pump itself, taking up that additional horsepower.
I And the number that came into us was 100 amps, and I know that came from an operator looking at some current gages up there, and I know those current meter well, and when it dropped down he knocked the PS l
l 894 192 l
l
I 24 i
i 1!
pump off and he said, "the needle was about here, that's 100 amps."
I i
2 was sure he hadn't looked at it in that detail.
So while I had Greg 3
Schaedel on the phone, I asked him to try to get back into the operator 4;
and find out, or if they ran another pump, to read the amp meter very 5
accurately.
I couldn't believe that we had two sheared shafts, but we 61 had two pumps that gave us 100 amps when they were running.
So we i
7 needed to know... from the amperage reading, we could trouble shoot 8
s me of the problems that it could have been.
It was, in fact, probably gj a 150 or 160 amps, but when its 33 amps per division on the meter, it i
10' is very easy to misread them.
11f CRESWELL:
I see.
Had you been in the control room during this period 12 13 f time, would you have gone through that type of evaluation?
14!
eg e, that was B&W inputed those horsepower numbers to me, 15 g
those amperage numbers.
17 CRESWELL:
Okay.
And this is in the afternoon.
Roughly..
18{
?
~
19i 20i j
FLOYD:
Yes, between two and four.
22 CRESWELL:
Did you have any indications that someone from the site, like, let's say, Lee Rogers, was in contact with B&W-Lynchburg?
25I 8?4 19f; t
I
,i
{
25 1
FLOYD:
Greg Schaedel is Lae's assistant and the gentlemen we talked 2j to finally.
I don't know whether it was said or not, but I had the 1
3' impression that Lee was in the plant.
Greg could not get into the 4
plant and so Greg wac trying to relay the information to B&W.
I think 5
I said on the earlier tape that I don't think there was an open line 6{
to B&W or they wouldn't have been calling me to the board room to try 7
to get into the control room with information.
If they had had an 8
pen line, they would have sent the information on their open line.
9 CRESWELL:
When was it in time again, that you realized that they had 10 significar.t fuel failures? That was around, what 8:00?
1 11 I
12{
FLOYD:
Yeah, that was in the morning.
Very early in the morning.
13 l
14i An as cally you had made dat decision on de informadon 15 you had about the dome radiation monitor?
16j 17 FLOYD:
No, on the gaseous activity in the reactor building.
19{
CRESWELL:
The gaseous activity.
Okay, could you go into that a little more?
21 i 22.
FLOYD: Well, as I said, I had a number for the dome monitor, which I could not believe and which I cannot believe the one that I still have 24 on it up there in the control room today.
If, in fact, I had 500 or 25 i
l 8?4 19.4 i.
i a
i 26 1
5000 R, whatever it translates out to' be, inside the reactor building 2
I would certainly see something outsida the concrete, which we do not.
3 So the only way that instrument, that dome monitor, can be seeing 4:
radioactivity is if I've breached the shield around it, and its 5
seeing a beta field.
Then that number of ten tc the fifth that it's 6
giving me nor could be accurate, and it could be a beta field that 7
high in there and I wouldn't see it outside the concrete. But it 8
certainly can't be a valid gamma number, so the four inches of lead g
must be breached and possibly even the shield around the detector 10 itself is breacned to let beta become, to let the detector become beta lij sensitive.
I recognized that the number that was transmitted to me at 12 7:30, quarter of eight, in the morning was too high to be a real gamma 13 number, ar, far as I was conce...ed.
So I based my feeling on core 14j damage on the number that was transmitted to me on the reacter building 15 atmospheric monitor, the gaseous channel, and by divif ng the sensitivity 15; f that instrument into the reading, I got back to some number of curles in the building.
Then, based on previous knowledge when I was g
calculating the 8R per hour for the dome monitor from the number of yg curies in the building, I thought it was about, representated about l/8 of the cladding
..a failed at least, and this was the numbe* that 201 was transmitted to me before 8:00 o' clock in.the morning.
21!
22 CRESWELL:
Okay. Were you aware of any other equipment problems,.
. let's see, we've mentioned,' what, the feedwater pumps...
.?
25 i
894 195 t
r
t
{
l 27 1
fLOYD:
The trip of the ferdwater pumps, which lead to the reactor l
2l trip and of course, it also tripped the turbine, but the.t wasn' t _too 3
surprising, the delay in tne opening of the emergency feed valves - I 4f was told about ten mirutes - and then later in the morning, the piece 5l f information I got that was helpful to me in the simulation was the 6
fact that the electromatic relief was leaking badly.
Based on that, I 7
had to go and simulate different reactor coolant system, different 8
size leaks and I finally got to the point where I had the valve, I g
just opened the valve on t,he trip or when it opened I'd fail it open 10 and leave it open, but what I had failed to do was to run the thing out for two or three cours to where I took all the heat out of the g
pressuri::er and alicwed the next hottest part of the water in the 12 reactor coolant system to flash.
I never ian out past, I don't think, 3
twenty minutes of any one simulation.
I 15j CRESWELL:
And those are basically... we already mentioned the reactor coolant pump.
18(
FLOYD:
That information I didn't get until the afternoon, when Greg 19l Schaedel called in to the plant.
20 21, CRESWELL:
In that context, that time period, were there any other 22 equipment problems that you became aware of?
23 24, b i
!hb 25
(
l I
l 1
l l
28 1!
FLOYD:
There may well have been, but none that I remember.
2!
I 3!
CRESWELL:
You mentioned that at that period of time, B&W was requesting i
4 at least 400 gpm high pressure injection flow.
At any time during the 5
conversations, did the history of high pressure injection flow come 61 into the power station?
7 FLOYD:
No.
8{
Si I
CRESWELL:
Dorwin, do you have any quest..' ors?
10 1 11 i
12{
HUNTER:
Jim, when you go back and look at the simulator and you talk 13 in terms of simultaneous feed pump trip, turbine trip, reactor trip on high pressure, what's the simulation for makaup flow during that 14, even Do you reca m 15 16l FLOYD: Well, on any reactor trip you get a contraction, and if you gj let the emergency feed come in normally, why the contraction will call for the MUV-17 to open and keep pressurizer level high, at set point.
19J 20l HUNTER:
Jim, could you go into MUV-17?
22 FLOYD:
Sorry about that.
MUV-17 is the automatic control valve that is positioned by pressurizer level to hold a constant setpoint in the 24i pressurizer, constant level setpoint in the pressurizer.
So that if 25l 871 19/
l t
i i
f
I l
29 It we cool down the reactor coolant system some ?S degrees on a reactor 2;
trip, the water will contract, pressurizer level will drop and MUV-17 3
will open to take water from the makeup tank and put it into the 4
pressurizer.
That's very normal on a reactor trip.
i 5
6i HUNTER:
In the simulator, Jim, would, in fact, the simulator also 7
require, or would it, in fact, include starting the second makeup pump 8
and taking the suction off the BWST?
i 9
FLOYO:
The simulator does not but we do it with our people as a 10 lit matter of course, when we are there. When you get the red light that g
says " reactor trip" at the simulator, why, the operator automatically 13 starts the second makeup pump and opens MUV-16 Bravo valve.
14!
15l HUNTER:
Okay, as another point in the procedures, in your procedures, for makeup system - upon a trip it's normal you start ona makeup pump g
and isolate the letdown, Okay, to try to keep the pressurizer level up.
Also another step in the procedure is to, if pressurizer level 18l goes below twenty inches, to start the third makeup pump.
Is this a 191 I
routine....?
20!
21; FLOYD: Well, normally the first two, you can prevent the third from happening.
24j b9i
}90 i
2Sj t
r 30 I
f 1{
HUNTER:
Have you seen the third come on, in your review of any transients 2
or trips?
3i 4
FLOYD:
We have had a dry pressurizer here on one or more trips.
5 6
CRESWELL:
You are going to have to clarify that a little bit, Jim, 7
what you mean by " dry pressurizer".., loss of indication or loss..
- ?
8 9
FLOYD:
Level indication went to zero.
101 11{
CRESWELL:
But you did not drain the pressurizer?
13, FLOYD:
Yes.
141 15j CRESWELL:
You drained the pressurizer?
16;:
l 17!
FLOYD:
Yes, oh yes.
It wasn't just indicating dry, it was dry.
18j 19l HUNTER:
201 And in that case the operator would nave started the third I
makeup pump, I presume?
21l i
22 FLOYD:
In that case, we'd probably have high pressure injection before that point and two makeup pumps, A and C, come on and 8 gets 24 kicked off, in that case.
25l 394
- 99
.i l
31 1;
HUNTER:
And would you describe the makeup flow path then, at that i
2j time on high pressure injection?
I 3
4 FLOYD:
Well, it's from the BWST through the DHV-5s, A and 8, MUV-148 5
through the two makeup pumps, through the MUV-16A, B, C, and D valves.
6l HUNTER:
Okay, you just described to me the two suction valves to the 7
8 split section header from the BWST to pump discharge valves, and then gj the loop discharge valves, which would be the 16A, 8, 0, and D.
Those 10 then would come open on the high pressure injection and supply a certain amount of flow through the loops.
12f FLOYD:
500 gallons per minute per pump, 250 gallons per loop.
14!
HUNTER:
Okay Jim.
1$j 16i CRESWELL:
This time where you voided the pressurizer, did you have automatic emergency feedwater level control?
i 19i FLOYD:
On some of those, I don't know that, it was imposed on us by 20l l
loss of feed, but if we had had a loss of feed then we yes, had automatic 21l 22l level control in the emergency feed system.
If the reactor trip is for some reason.other than loss of feed, of cour'se, the emergency feed 23 l
pumps do not get called on to operate.
2 41 I
25l 09 $
2[][)
I I
t
I 32 1
CRESWELL:
Okay so...
2.
3 FLOYO: We could have maintained steam generator levels of 30 inches 4j on the main feed system.
5 61 CRESWELL:
Do you have an automatic control, level control, for the 7
emergency feedwater?
8 gl FLOYD:
Oh yes.
10 i
CRESWELL:
And it is presently set at 30 inches?
11 12, FLOYD:
For the loss of two feed pumps, yes.
It goes up to 50 percent on the operating range for loss of four reactor coolant pumps.
It's a 15l fairly intelligent control system.
It looks at what's happening 1
f before it tells you what set point to go to.
16!
17 CRESWELL:
But the incident that you related earlier, where you voided l
the pressurizer, can you recall about when that happened?
191 20f l
FLOYD:
No, but our reactor trip reports will indicate it.
There were 21!
l at least two occasions that I know of when we ended up with dry pressurizers 22[
... the pressurizer dry, on Unit 2.
23 24 25l 4
i i
[
33 CRESWELL:
O k.
l 2l 3
HUNTER:
I don't he;se any further questions, Jim.
4 ci CRESWELL:
No, I think that that's all that I have.
l 6l HUNTER:
Don.
7 8
KiRKPATRICK:
No, I don't have any more questions.
g 10' RESNER:
All right.
Thank you very much for your time, Mr. Floyd.
This will conclude this particular interview.
It should be noted the interview was concluded ?.t 4:55 p.m. Eastern Daylight Time, May 2, 1979.
14) i 15j 6!
894 202 17 l
18(
19i 20!
21 e
22I 23 24 l
25{
l i
!