ML19249B034

Transcript of TMI-2 Investigation Interview W/B Zewe

ML19249B034
Person / Time
Site:	Crane
Issue date:	07/06/1979
From:	NRC OFFICE OF INSPECTION & ENFORCEMENT (IE)
To:
References
NUDOCS 7908290292
Download: ML19249B034 (49)

Text

.... - -

UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION 1

In the Matter of:

I 2

IE TMI INVESTIGATION INTERVIEW 3'

of Mr. Bill Zewe 4

Si Ei 7

8 Trailer #203 9

NRC Investigation Site TMI Nuclear Power Plant 10 Middletown, Pennsylvania 11l 12l (Date of Interview) l 13l July 6, 1979 (Date Transcript Typed) 14' 273 15 (Tape Number (s))

16i i

17 18 19 h

Z0 21; l

NRC PERSONNEL:

'22

""O 232 r

23 24 25

y e

1.

INTERVIEWER:

This is tape one, side one of the interview with Bill Zewe.

2 I'm sorry, say it again.

3 4

ZEWE:

Zewe.

5 INTERVIEWER:

Zewe.

6 7

ZEWE:

8 This I believe is the fourth time that I have been taped.

I had a g

tape there with GPU before with Long.

And also I was interviewed with the NRC individually and then we were interviewed as a group.

And the group 10 session, I thought. was much better than the rest of them because we were g

all doing different things and looking at things a little bit differently at the time and that interview, more than anything else, helped us, you know, to say, to bring it more into a time perspective because personally I I

was very far off in the times.

Times were much longer and I felt they were

,St much shorter.

As it ended up, we sort of got a better perspective for what I

17l times the things happened, more so than what we had at the beginning, even f

individually, we were all fairly close to the times that we thought that we 18 had and then once we got together we seemed to think times were quite a bit different.

And judged on a few of the graphs and things that we had seen, we had noticed that the times there were a lot different than what we had 21 f

previously imagined.

22' 23 INTERVIEWER:

We were finding it out also in reviewing the strip charts and 24 what not, it seemed like looking at your original interviews that the times 25 didn't jive with what the real world was doing.

"O 233 r

I

s 2

1..

ZEWE:

Probably the time that we compressed the most as a group, were the 2

times we were in the emergency feed actuation.

I could have sworn that we 3

had emergency feed going within two minutes.

And we actually had it going, 4

but everything that we see was more like eight minute period had elapsed.

5 And I still can't believe that, but it is true because it has to be true.

6 And then plus other people entered the control room at about the six minute 7

point and they remember they we didn't have it on at that point.

So that was true.

That was one area in particular, that it just seemed like, you 8

know, that it couldn't have been that long.

How else... do you want to go g

10, n and just ask questions on it and I'll just try to answer the questions or what.

g 12:

INTERVIEWER:

Can you tell me what time to s ur recollection the atmos-pheric dump valves were opened?

l 15i 16li I purposely had the atmospheric valves opened up, but it was quite ZEWE:

some time.

It was -- I am trying to remember -- it was sometime just before we broke bacuum alright, I went down to two circ water pumps which automatically forces the turbine bypass valve controls to shift to the l

atmospheric dumps.

So I did that on purpose on only the A side, because I had what I felt at the time was a confirmed leak in the B steam generator, 21 primary to secondary side.

And I didn't want that activity to be dis-22 charged to the atmosphere, but we had lost steam from the aux boilers from 23 Unit 1 feeding the seals in Unit 2 turbine and I had a choice of either 24 dedregating that or goina to the atmospherics.

And I was reasonably sure 25 i

j "O

234 I

s 3

1 that the A generator was still intact.

I didn't have any reason to believe 2

any problem from the A.

So we did get on to two circ water pumps and then 3

go on the atmospheric dumps to the A valve only, until we regained the 4

boilers frcm Unit 1 and then we reestablished seals for the Unit 2 turbine 5

and then we went back on to the turbine bypass valves, but we did break 6

vacuum.

And then it took us a considerable amount of time to get steam 7

back from Unit 1 and then draw vacuum again, because we did have vacuum off 8

completely for quite some time.

And that would probably approach, you know, well over an hour or two hours.

But we did shut the turbine bypass g

valve alright, going to the atmospheric sometime before we reestablished 10 vacuum and came back to the condenser, but I am not sure of the time frame I

there.

13 INTERVIEWER:

You can't put it into perspective, with respect to running I

the 2B reactor coolant pump and getting some of the radiation alarms...

15l I

breaking vacuum and lifting the atmospheric?

1 61 17 ZEWE:

Prior til -- as close as I can remember -- somewccre around 6:30 or 18j l

so, we decided to feed the B steam generator and also to try to start a reactor coolant pump.

We did these rather closely together as I remember.

We fed the Bravo steam generator and then shortly thereafter, we tried to start any coolant pump that we could.

We first tried the 1A, 1B, 2A pumps 22 and we weren't able to get them to run.

So we finally got the 28, I believe, 23 to run but it only drew about 100 amps and we weren't sure if the pump was 24 actually running or not.

So we went and checked the amperage at the breaker 25 too and it read just under 100 amps what we read up in the control rocm.

"O 235 I

i 4

f 1

INTERVIEWER:

Was the B steam generator isolated when you ran the 8 coolant 2,

pump?

3 4

ZEWE:

Yes, the B steam generator, as far as I know, we had it isolated up 5

until that time, just before we started the coolant pumps alright, but we 6

unisolate 8 as far as feeding it goes, just before we started the coolant 7

pumps, of vhich 28 was the only one we were successful in getting to run.

8 Shortly ther eafter was when we really had the first signs that we had a.

problem with radiation levels.

The offgas condenser monitor came into g

alarm.

10 And then we started to see all the alarms for the auxiliary buildings 3

and the fuel handling buildings everywhere then they started to come in at once.

Prior to that the only real alarms that we had, that I can remember 7

and I looked at it pretty hard, was the intermediate letdown coolers.

They were in high alarm fairly early after the accident.

But they have a very 1 w setpoint.

And they are very susceptible to background.

And they are 15 right down by the R8 sump themselves.

So I didn't figure that was too much I

of a problem, because they are usually in alert, in high alert during 171 normal operations, just from background.

And we have proven that as we 18(

l escalate in power, background comes up on these monitors.

19!

t 20[

INTERVIEWER:

Did that alarm come in before you had stopped the... all four reactor coolant pumps?

22 23 ZEWE:

Oh yes, yes before that.

Those were in quite early and I am not...

24 oh yes, yes long before that.

25i

" O 236 i

i 5

1 INTERVIEWER:

Did you have your reactor coolant sump pump on at the time?

2 3

ZEWE: I uh.

4 5

INTERVIEWER:

This is around the time that you started the 2B pump, were 6

the reactor coolant sump pumps running?

7 ZEWE:

8 No, because there was approximately forty five minutes after we g

tripped, the operator informed me that the high sump alarms were in for the reactor building.

So I had the Control Room operator contact the aux 10 operator on the primary side to turn off the Reactor Building sump pumps.

So he turned them off and at that point, it was like probably quarter to five in the morning, somewhere before 5:00, but they had been on previous to that, because the alarm that we get up in the Control Room is that the 15l

" pump run" up on the computer.

j i

161 INTERVIEWER:

Let me just ask this question just to see how you answer it.

We know that you let down on the 2B side particular letdown (unintelligible) would that have given you the alarm, you know... would that give you the alarm that you saw in the auxiliary building?

It is hard to figure out why you got an alarm in the Auxiliary Building.

22,

[

ZEWE:

I don't know.

""0 237 23l t

24 25(

l l

l l

i 6

1-INTERVIEWER:

How did the radiation get out, saying, there has got to be a 2

path?

3 ZEWE:

That I don't know yet.

All right because we did confirm later on in 4

5 the next day and a couple of times, that the operator did turn off both 6

sump pumps.

So they weren't in automatic start so they were off alright.

7 I just have to assume that nobody returned them to normal or started them prior to that.

But I don't know that to be the case alright.

g 9

INTERVIEWER:

Do you think the computer would have printed it out if they 10, did?

12 ZEWE:

If the pumps were to come back on, the computer should have picked that up.

I am not sure if the status on the computer is true or not in that respect.

15j 16 G:TERVIEWER:

We lost about... the computer lost about an hour or so of time.

It was turned off or out of commissio.1 or something.

l 19!

i ZEWE:

I don't know that, alright.

21 INTERVIEWER:

After the're shutoff I can find no... no other, you know, 22 reactor building sump alarms all through the rest of the day, either on or 23 off.

It went off for 38 minutes.

I would imagine that...

24 25j "O 238 i

I t

7 1

ZEWE:

That is what I thought, because he said that it was approximately 45 2.

minutes after the trip that he turned it off.

So we said, you know, it was 3

probably quarter to five or 5:00, so 38 is probably in the ballpark area 4-you know.

5 INTERVIEWER:

What time did the auxiliary building high radiation alarm?

6 7

ZEWE:

8 Not until just before we declared the site emergency, ten to seven.

So I'm saying that they started to come in somewhere between like twenty to g

seven and ten to seven.

Long after we... and we thought maybe we had O

gotten water in somewhere from the letdown system.

The letdown relief valve goes into the miscellaneous waste holdup tank system.

And that tank, as far as I know, never got higher than 7.4 feet, so it never overflowed.

So I don't -- there are other reliefs like on the makeup tank.

That one f

goes into the bleed tank.

But there are some downstream pf the makeup 15t j

pumps themselves, which go directly to the floor.

One of those could have 16; been blowing or something else that I haven't determined yet, I don't know.

But I felt sure then and now until somebody could prove me that somebody i

turned on the sump pump, plus the sump pump should have been lined up to 19 I

the miscellaneous waste holdup tank, and we never overflowed it, to my 20' knowledge.

There - you can put it into the sump, but it should have been 21 lined up to the miscellaneous wastes holdup tank.

22 23 INTERVIEWER::What tank overflowed? Where did you get the water on the 24 floor?

25

"O 239 l

8 1

ZEWE:

Aux building sump and the aux building sump tank.

2 3

INTERVIEWER:

Where did that water come from?

4 ZEWE:

That is what I don't know.

5 6

INTERVIEWER:

Okay.

All right.

7 8

ZEWE:

I am not sure.

g 10 INTERVIEWER:

So where ever the water came from, that is the same place the activity -- the airborne activity -- came from and you are just not sure what the source is?

14l 15l ZEWE:

I don't not at this time.

And then the operator all right, after he i

turned off the sump pumps alright, he had been in the basement in the aux 16!

I building, after that and I am talking about probably forty-five minutes to 17l1 an hour after that time, he was still in the aux building and the and the 18j drains were not backed up to where we had found them after we had all the 19!

alarms.

20 21 INTERVIEWER:

This is when Terry Dorgey went down there?

22 23 ZEWE:

He was there and also Don Miller was there, because I sent him down 24 to the valve alley to check the pressurizer level there by MEV 17 and 155, 25

""O 240 i

t t

9 1

was behind the compensated pressurizer level indication.

And they're both 2

those valves are in the basement there and there was no water coming out of 3

the floor drains at that point and it looks like all the water was concen-4 trated around the floor drains and that the water just came from the floor 5

drains themselves, but where it was from, I don't at this time.

Maybe 6

somebody else does but I haven't found that out yet.

7 INTERVIEWER:

Where does the reactor drain tank pumps pump to?

g 9

ZEWE:

They pump to the aux building ch the...

10 lli INTERVIEWER:

Reactor drain tank -

quench tank I guess it is.

13 ZEWE:

Right.

We direct that alright to the bleed tanks.

15 INTEP. VIEWER:

Were they pumping in automatic?

16 17 ZEWE:

No, no.

We manually do that on high level, alright.

We manually throttle a valve and then put it into the bleed tanks and we were not doing 19 that to my knowledge at all, no.

20' 21 INTERVIEWER:

Were you letting down most of the time?

22 23 ZEWE:

Yes we were.

e were trying to determine why the pressurizer was 24 going up so rapidly, alright.

I had my shift foreman assigned to the 25; i

l

"'O 241

10 1

makeup system and the pressurizer level alright, and I let him and a control 2

room operator try to handle that, while I was trying to work with emergency 3

feed and trying to work with hotwell level and try to isolate a leak in the 4

turbine building, plus I had hauled over the other shift supervisor that 5

was onsite Ken Bryant to the control room.

So with my foreman there and a 6

control room operator plus with Ken and the other control room operator, I 7

was trying to get some of the other things wrapped up, alright because we 8

had lost the feed pumps and the one feed pump doesn't have a turning gear on it.

g And I diverted a lot of my attention to those items, while they were 1 king at primary plant.

I didn't know why the pressurizer level was 10, 3

indicating so high, nor why the pressure was holding low, but we seemed to be fairly stable in all other indications and I left the control room, y

probably I don't know, fifteen minutes after it happened, to go down and 13 try to recover the condensate to reject from the hotwell.

We were flooding out the hotwell and we had some leakage down there by the Condensate pumps and so forth and I had gone down there to try to get the bypass on the polishers open and find out some other details.

And then when I got back to the control room and I am not sure how long that I was gone, and we were i

18!

joined by George Kunder then, who we had called right after we tripped and I guess George was there then, within about 20 minutes after we trippedand 20l with George and Ken and Fred and Ed and Craig in there, I was still trying to wrap up some other things alright.

And then we just went through, trying to put our heads together, trying to come up with weird indications that we had alright.

The high level and it really didn't don on me or 24 anyone else at that point, you know, that we had really transferred that 25 i

" O 242 l

11 1

bubble.

And uh I don't know at that point, we were just trying to come up 2

with alternatives and the problems and checking out our pressurizer level 3-on the computers, compensated and uncompensated, local and remote, compen-4 sated levels and our reactor coolant system temperature then looked real 5

good to us, at that point alright.

We didn't really have a lot of problems 6

until we began to get some flow oscillations on the reactor coolant pumps 7

and then George and'I then agreed that we close to the net positive suction g

heads for the pumps plus the fuel pin compression...

nd then we secured

^

g the coolant pumps and then it was probably twenty minutes or so later that 10 we tried to go on natural circulation and secured the last two pumps and fed steam generator A up to about 85 or 90%.

11 12 INTERVIEWER:

13 When did you take the sample of the coolant system for boron did you check

...?

l 15jy ZEWE:

I am not sure of the exact time.

But as soon as we tripped we had informed the lab you know we had tripped and they would have to start to 17 take samples.

So the first sample result that I got back was about 700 ppm 18l l

boron and it didn't make any sense to me at all.

I said that it had to be 19{

l wrong because we had been on high pressure injection from BWST for a while 201 and the boron concentration had been over a thousand to begin with and so then they resampled and they came up with 400.

23 INTERVIEWER:

What time was that event involved related to (unintelligible) 400?

Can you think of any other things that were more than one around that l

time...

'"'O 243

1 12 1

ZEWE:

Not really, you know, I -- I'm not sure of their exact sample time 2

at the point and uh I...

3 4

INTERVIEWER:

Were all the coolant pumps off at the time?

5 6

ZEWE:

When I finally got those samples, yes they were alright, and then I 7

was joined in the control room, and I am not sure on the times, by several 8

other people, by Mike Ross and Brian Mehler, Joe Logan, Rick Faraduit and I g'

am not sure of exactly where we were when they got the sample, but I am 10 sure that are our pumps were off, yeah.

11 INTERVIEWER:

They were in the control room at that...

12 13 ZEWE:

Yes, George was - plus there were there other than the normal shift g

15 w s Ken Bryant was there and probably 5 or 6 minutes after it happened and i

then George was there within 20 minutes and then Mike and Brian Mehler were 6

there, I don't know, probably less than an hour at the most when they were there and then Joe Logan and several other people showed up at varying times and I am not sure.

i 20' INTERVIEWER:

Did you ever determine what was the reason for the bad boron analysis?

23 ZEWE:

I think that we have.. that we were actually flashing the letdown and it was evaporating the letdown sample and we leaving the boron behind.

I think that is what ultimately it was.

"O 244

{

13 1:

INTERVIEWER:

Do you know when you shut RC RV2?

2 3

ZEWE:

Alright, well I had isolated the B steam generator, first of all 4

because I had some indications that I had a steam leak off of it, so the 5

pressure in the B steam generator was about 300 pounds less than A and we 6

had a building pressure somewhere around 2 pounds alright, so when I isolated 7

the B steam generator and the pressure stopped going up and started to come 8

down on the RB pressur1, so I thought oh that was it, but then a short time g

later we shut RCV 2 the block for the electromatic and pressure took a 10 marked drop right there and that would be timed on the building pressure g

recorder, but I am not sure of the exact time, it was not until...

12, INTERVIEWER:

13

... you saw a unique change in the building recorder, in the reactor building recorder...

15l 16;l ZEWE:

Right away.

17 INTERVIEWER:

Right away.

l 19!

ZEWE:

Right.

I did see a change right away too when we isolated the B steam generator but not near of the magnitude as when we shut RCV 2.

22 INTERVIEWER:

Would that be about 6:15?

23

"'0 245 25i

14 1

ZEWE:

That's about when we were indicating.

Okay, that could have been 2,

it, yeah.

When was... when was the boron sample relative to that -- that 3

the...

4 ZEWE:

I'd say probably four, but not a great deal before that, but I am 5

6 just guessing here, see I don't know for sure.

7 INTERVIEWER:

So you got the results for... (unintelligible) 8 9

ZEWE:

Yeah.

10 11 INTERVIEWER:

That would tell us how... (unintelligible) 13 ZEWE:

I am not sure exactly.

We had checked the discharge relief valve temperature fairly early.

As soon as Ken Bryant got to Unit 1, I asked him to look over the computer alarm, you know, to help us out.

And then had 16 him checked the discharge temperatures of the relief valve.

And he said, you know, that RCRV 2's a little high, but you know, about 30 degrees high i

and then we knew that it had lifted, so we felt that was still cooling 191 I

down.

So that's why that I went off of the RCRV 2 problem and went back on 20!

something else, because I just felt from that it was alright and but it 21 ended up that I should have gone back to it, but I didn't until Brian 22 Mehler and Ross was there and then Brian was looking over the computer 23 again and then he said well it's a little high, well let's try to isolate 24.'

it, I don't think it is anything and that is what it was, but we didn't 25 have..

I

^^0 246 I

m 15 1

INTERVIEWER:

... a couple of scrap metal in there, it is not an RTD and 2

the highest it ever gets is...

l 3

4 ZEWE:

Well the -- all it was then was about 228 or something like that and 5

they had been running around 190 and then knowing that we had it lifted and 6

plus that we had higher than normal temperatures I didn't see that was much 7

of a problem.

So we didn't go back to it to that point really not knowing 8

for sure, but just saying let's isolate it you know and see and that was the problem.

g 10' INTERVIEWER:

What happened that could... did you put the high pressure injection pumps on after you isolated RCRV 2?

13 ZEWE:

We had partial high pressure injection most of the time alright,

]

because we had high pressure injection within, you know, a couple minutes of the problem and then we had complete actuation, we bypassed it and then I told my foreman then to try to establish, you know, normal pressurizer level.

So then he went and then he was throttling back on the 16 valves 18{

and then he turned up the makeup pumps and then it went on one makeup pump.

Then he tried to establish that.

That is when I left to go over to help with the emergency feed, because it was just about that time that the operator had said that we are still at ten inches and the -- well at first 22 he didn't have full open indication on the 11 valves -- BF 11's -- so I 23 told him, you know, to take manual control and then open them up all the 24 way,.because it looks like that we weren't feeding alright.

And so as soon 25!

""O 247

16 t

1l as he had full open indications on those, we looked and he still didn't 2,

seem like he was feeding, so then we checked real quick and found the 3

valves were shut alright, and then he's on them just as I walked over to 4

where they were controlling the pressurizer again and then he yelled that 5

12's were shut.

And I said why'd you shut the 12's.

And he said no they 6l were shut, they weren't opened up.

So then we opened them up right away 7

and then I was with him for awhile trying to establish a goed level with 8

about 30 inches, when we had considerable problems trying to maintain g

level, because bravo steam generator wanted to go high on us alright, it seemed harder to control it.

So somewhere down the line we ended up cycling 10 n

the EFV 5's -- the header isolation valves upstream to control flow to generators, because even with 118 and 128 shut the B steam generator was still increasing.

g 14!

INTERVIEWER:

That was from what about maybe a half an hour into it on that 16l you were cycling the five valves to maintain level?

i 17 ZEWE:

I would think so yes, as far as I can remember, yeah.

l 19 INTERVIEWER:

This... if I can ask William one more question, when you closed that block valve you could see in the one graph reading reactor coolant system pressure started coming up pretty good.

In a space of about 22 45 minutes it came up to 2100 or so pounds.

Did you throw makeup pumps on at that time, both four makeup tanks to get you pressure all the way down.

Is that when you put the makeup pumps on to deal with these valves?

25 I

l

'" O 248

17 1

ZEWE:

Yeah.

2 3

INTERVIEWER:

(Unintelligible) at that time?

4 5

ZEWE:

I think we only had on two at that point but I can't remember for 6

sure, I really don't recall.

7 INTERVIEWER:

Once you come up to pressure then were you starting to worry 8

that you were going to overpressurize?

g 10 ZEWE:

Yeah, well we were watching it closely at that point to make sure, g

you know, that did have some control of pressurize at that point.

But I really wasn't worried at that point we really weren't concerned about g

overpressurizing at that point I was just glad that we had some of the pressure back.

16' INTERVIEWER:

I was wondering if shortly after that it looked like you opened up the automatic (unintelligible)? What the reason was on that?

19 ZEWE:

We went through a couple of evolutions there to where one that we tried to come up in pressure to compress the system up alright, with the bubble and everything and then af ter that then we thought to come down ar.d 22 try to float the core flood tanks on the reactor coolant system, so that we 23 opened up and sprayed down, alright, all that we could to try to get down 24 to get the core flood tanks in and also to try to get down low enough in 25

""O 249 l

18 1.

pressure to go on decay heat removal, but we weren't able to get down any 2

less than about 410 or 420 pounds or so.

So we weren't able to do that, so 3

then we sat there for some time and then we decided to just inject again, 4

try to go back up and compress the bubble again and try to rerun a reactor 5

coolant pump at that point and now we are talking about early afternoon and 6

I w uld say, you know, 3:00...

7 INTERVIEWER:

Were you lifting atmospheric dumps into that time?

g 9

ZEWE:

What?

10 11 INTERVIEWER:

Were you lifting atmospheric dumps during that time?

13 ZEWE:

We had indication of somewhere around like 11:00, 12:30 during the day..

9:00 at night or something like that.

16 INTERVIEWER:

About 12:30 is when they stopped because that is when Gary Miller and all had to go to a meeting with the Governor see and they had them shut so they wouldn't dump anymore.

191 1

20 ZEWE Right.

We had it opened only when we broke vacuum and then for a short time after that, alright, and then we shut them, even though we didn't have vacuum until and then we didn't ' reopen them again until we 23 reestablished vacuum again to dump it to the condensers and again only the 24 A side was ever affected.

We never unisolated B again after we had uniso-25 I

I "O 250

19 1

lated it to feed it just before we started the pump and we had all the alarmsandIreisolated5'...anditstayedisolated,Ithinkuntilright 2 ;

3 now and it's still isolated just as we left it, the steam and feed both.

4 S

it was unisolated for raally a-fairly short period of time.

There 5

whenever we had the offgas monitor, we went and isolatec it and we didn't 6

use it since then, either the atr.;ospheric dump, nor the turbine bypass dump.

7 8

INTERVIEWER:

Okay the offgas moaitor came in before you ran the reactor g

1#Ut E"*P?

10 11 ZEWE:

No.

It scened Uke that we fed tiie generator and started that 28 pump and shortly t:1ereafter -- a*ter we did both of those, the alarms all came in, we had off the pump then, alright, the pump was only running for, I don't know, maybe 15 minutes et the most, alright, but as soon as ue got 16l the alarms alright, we went ar.d isolated the B steam generator again and we already had that ccolant pump off at that time.

That it was after we fed and started the pump that we had the problem.

19 INTERVIEWER:

Okay so you had B steam generator isolated for awhile then i

20 you unisolated it and fed it to and started the B pump and at that time 21 steam off the stean generator was,goliig to the condenser you still had 22 vacuum on the condeM er.

2

"O 251 24 25

20 1

ZEWE:

Yeah.

We still had vacuum till quite some time after that.

We were 2

able to run with one aux boiler, trying to fix the other one up, but then 3

finally we lost them both and then we had lost... our primary had been d

supplying our turbine with sealing steam but we were down cool enough now 5

to where we losing that because our pressure in the steam generators at 6

this point, we had A was on very low, Bravo was at about'350 pounds or so.

7 INTERVIEWER: ~ That is why I was wondering why you had the atmospheric open 8

g at the time when the A steam generator pressure low.

That's definitely a

~

manual route, you can operate it in manual.

10 11 ZEWE:

You can operate it from the control room at the Bailly station.

13 INTERVIEWER:

In the first hour or so was there any specific reason for not going on natural circulation?

16 ZEWE:

No, not really as long as we had coolant pump flow I really didn't think, you know, to go on natural circulation or why I should do so.

19 i

INTERVIEWER:

When did you notice the source range increase?

21 ZEWE:

I noticed that after all the pumps were off.

23 INTERVIEWER:

It was af ter all the pumps off...

24 I

25!

"O 252 i

I i

I 21 1l ZEVE:

As far as I can remember, it was after all the pumps were off as I gf recall and then as soon as we seen that, then we started to inject boric

{

3I acid into it and I believe then is when we started to get the sample back i

4j about the 700 ppm boron, because it confused me a bit at the time the we 5

had just done a shutdown aargin calculation and they said that we were 6%

6 shutdoan and the count began to come up and I think, John was there at that 7

point too.

And then the counts were coming up and the boron samples were g

cohing back low and I said it is real, it is actually... counts are increas-g ing boroa is low while shutdown margin isn't what it is, so we started to inject boric acid in there.

Well I think we secured the pumps somewhere 10, around quarter to six or so.

So S was somewnere around then at that point we had quite a few people in the control room there and I remember John was there at that point I believe.

14 9

15 l

was real at that time.

16!

l 17 ZEWE:

Okay.

That's right because I had asked you that, I remember that 18j but we had several people there, but I just didn't understand, you know, 19l 20 why the boron was so low, we had been on high pressure injection plus added some acid prior to this and we were 6% shutdown but then when John said the 21, explanation of it I guess that is probably what happened and it took the 22 flux from the core then from the under moderation.

23 24..

3 25j "O 253 l

i 1

4 22 1

INTERVIEWER:

I am just trying to establish whether that occurred before or 2'

after the pump stopped?

3 4

ZEWE:

I am almost positive that it was after the pump stopped.

5 6

INTERVIEWER:

The lapse of time after it took to record and log it...

7 ZEWE:

And also we noticed a rapid change in the source range and interme-8 g

diate range levels whenever we started that 2B pump.

The counts were going 10 up as soon as it ran, it came right down.

11 INTERVIEWER:

g That is when your call was flashing pumps... you've cut your pumps ff (unintelligible) you start your pump again, you pumped up water 13 your count rate will go down and you cut it off shortly thereafter they me up a little again.

1s 16 INTERVIEWER:

You say that is when after the pump shut off was that the first little rise that you saw or was that the one where it went up to the l

intermediate range?

19{

i 201 ZEWE:

They both had...

22 INTERVIEWER:

Well there was...

23 24 2s "O 254 1

\\

i f

23 lt ZEWE:

Well the first two were when the pumps were shut off.

There was l

2!

probably 20 minutes or so between whenever we went from four to two and 3

then from two to zero.

4 5

INTERVIEWER:

(Unintelligible) 6 7

ZEWE:

I can't -- I think we started to add the acid whenever we went down g

to two pumps.

That was the point.

I 9'

INTERVIEWER:

You were running one makeup pump and then for some period of 10 g

time and then kicked off two reactor coolant pumps and I am wondering at g

what point that you went back to full makeup flow and (unintelligible) makeup pumps relative to the coolant pump operation.

y3 14 ZEWE:

I think that we were in back to much more injection flow so that we g

secured the pumps.

17 INTERVIEWER:

And was that before or after you sample...

i 19l ZEWE:

I thi;ik that would be before the sample came out.

21 INTERVIEWER:

At 4:00... was there any switching going on, were they going 22 to try to shift the condensate cump or was the dispatcher -- does he normally do something at 4:00 -- it seemed strange that at exactly 4:00 this whole 24 mess starts and that you lose so many things in such a chort period of 25 l

""O 255 i

l t

24 it time, the condensate pump, the feed pump and the turbine and the differen-2 tial trip and everything all happens within a period of just seconds.

It 3

usually doesn't on a loss of feed pump, you know, loss of feed trip.

4 5

ZEWE:

I don't know because at that point as far as the electrical plant 6f goes is normal.

Everything was normal.

We were running'the normal running 7

equipment.

The only real problem that we had, you know, was the -- we had resinthatwasstuckinthelinebetweenthenumber7polisherandthe 8

g' receiving line and we had been trying to unclog that from about 4 p.m. the previous shift, so that was like 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of trying to free up the resin 10, froa that number 7 polisher.

g 12 INTERVIEWER:

Were you running on fewer than normal polishers than you 13 normally do?

l 15l 16;f ZEWE:

No.

I 17l 1

INTERVIEWER:

The same number.

18l 19 ZEWE:

Seven polishers with one standby and it was the standby 1 that we 20 were transferring out we already had a charge for it to put back into it.

But we couldn't get all the resin out of it, that the -- I don't know alright, because everything else, why we actually tripped I felt sure at 23 the time, after I tried to start a condensate pump from the control room 24 alright, I couldn't get one started, we reset the breakers.

I finally'got "O 256 t

)

25 l

1 one of the condensate pumps to start and then we couldn't get any suction 2

pressure aver to the booster pump, over the booster pump wouldn't start 3

alright, we had' good discharge pressurs from the condensate pumps, so then 4j I said, well the only thing between them is the polishers, so I tried to 5

open up the bypass valve from up in the control room C0P 12 from up there 6

but it wouldn't come open so that's when the operator called saying that we 7

had a bad leak on the suction of C0P 2A booster pump.

Well I went down 8

there then to help him isolate it and to try to open up C0P 12 or see what g

the problem was and why we weren't getting pressure pump from the polishers.

S 10 I went down and we isolated that leak and I looked at the polishers and g

all the polishers valves were failed shut, so then I went up and the hand-wheel was 'off of C0P 12, it had fallen off and was down behind the ventila-g tion ducts.

13 So I sent an operator for a wrench, so that we could open up 12 manually.

The hotwells were outasight high at this point and we couldn't reject the water.

So when he went for the wrench I -- I was looking around 15 for another handwheel to use and I found that one laying down behind the duct work.

So I crawled up there and started to open it up, got it open partially, called the control room and had them open it up all the way.

l And then we started to reject the hotwell back into storage tanks at the 19' time that I was out of the control room and then I went up in the control room.

At that time I felt that the problem was that the polishers had 21 isolated themselves and it wasn't til several hours later that we received 22 on the computer printout that the condensate pump had tripped first or was 23 the first real alarm that they had.

24 2s,

""O 257 b

l 26 lj INTERVIEWER:

The hotwell pump was on then?

I 2l l

St ZEWE:

Yes.

We were successful in rejecting the hotwell once we had the 4

12's open, we took manual control of the reject valve and rejected to the 5

hotwell -- or to the storage tank.

6 7

INTERVIEWER:

Okay, that was about twenty minutes.

8 ZEWE:

It seemed like two, but I'd say twenty, yeah.

g 10 INTERVIEWER:

After you left the control room about 15 minutes after it 11 3

started, you were in here for about 15 minutes...

13 ZEWE:

And up until Friday we didn't even overflow a bleed tank.

Friday we g

15 16i I

INTERVIEWER:

g Well somebody mentioned to us that -- that they -- that 8f W dnesday you lost the seal injection, feed water injection for all you radwaste pumps.

19l 20 ZEWE:

Yes.

21, 22 INTERVIEWER:

Would that have caused the release of all the water onto the 23 floor?

24 "O 258 25

27 1!

ZEWE:

When I came in, I guess, it was quarter to four Thursday morning, I

2!

alright.

I looked at the bleed tank and I noticed that the bleed tanks 3

were at a lower level than what they were when I had left at 8:00 the night before.

So then, I noticed that the bleed tanks were less than what they 4;

5 were before and, I said, they were still putting water into the sumps 6

somewhere that either one of the tanks are leaking and it looked like that all the bleed tanks were showing it, so it had to be something rather 7

common.

So then, in a couple of minutes, we sort of figured it out that we 8

g had lost bus 32A and 42A, probably 4:00 the afternoon before alright on the 10, the day it happened and they there was still trying to restore it and they had restored partial loads to it.

So then, my control room operator, 3

Ed Frederick, was checking through all the loads off of there and he said, p

you know, the seal water pump units are off of there too and I said och g

that's probably good that all the pumps are draining back through their seals becwse they don't have any sealing water.

So then, we had an operator go into the building, -- in the aux building and I had him close the electric breakers alright, and then we were able to restore the cooling water unit i

and things like the boric acid pumps were off, and the oil pumps, reactor coolant pumps, and the aux building sump pumps, the aux building sump pump tank was off and so forth, so that we restored that and then the levels were normal.

21 22 INTERVIEWER:

What time do you think it was that you started to get water coming up out of the drains in the auxiliary building in Unit 2?

25 "O 259 6

28 1

ZEWE_: When I first knew about it, it was from the alarms that i know of 2

anc that was, you know, twenty to seven or something.

Prior to that point, 3

they had seen a little bit of water collected around the drains, you know,

^

4 very minimal amount, but it was the rad chem tech that said that he had put 5

a radiac on it an_d that there was no ac'.ivity.

I think that may have been 6

boric acid from the boric acid pumps and that's a relief valve at that 7

point, and that was sometime before we actually found the great deal of water on the floor.

g 9

INTERVIEWER:

What time did you -- well I know you shut off the reactor 10lg coolant sump pumps at around 38 minutes and what was giving you indication that you needed to do that?

13 ZEWE:

Just that the sump only goes up to six feet, and then down at the

]

radwaste panel, whenever we talked to the operator before he secured it, we asked him wiist the level was, and it was offscale high down there at six 16j feet, and I knew that when we came over that, we really weren't in good shape as far as primary inventory goes, that our aux building sump and aux building sump tank were rather full to begin with, and I didn't want to 19l bring over anymore water and overflow in to the sump because I knew that we couldn't handle it, and plus what was in the building, I wanted to keep in the building.

I had so many other things I was looking at with the RC drain tank, you know, and so forth, and I knew at this point that we either had the relief valve was soaping on it, or the rupture disc had blown, but 24 at this point, the highest pressure that I had seen in the plant, which is 25

" O 260

(

29 I

1, from the corridor, the primary pressure corridor, showed about 2360 maybe.

2 ;

That is as high as I had seen on any instrumentation.

3' 4.

INTERVIEWER:

Did you say that both of the reactor building sump pumps were 5

running when you had this high level?

6 1

ZEWE:

They should have, yeah.

I don't recall asking if both of them were 7

8 r n t, but as soon as Dick, I, knew that the sump was full and the on g

pumps were on, pump or pumps were on, when I said let's go ahead and shut 10 them off.

I don't want them going into the aux building.

111 INTERVIEWER:

It would have to be a fairly substantial leak for the one pump not be able to handle it.

Apparently, one pump kicked on and the level continued to go up, right? What does one pump, pump about?

15!

ZEWE:

You are only talking alright, if I remember right, the aux building 16j i

sump from the pump down area is only like 117 gallons or something like 17!

I that, alright so, we are talking about a cr3at volume from when the pump 18; i

starts and stops on levels.

You are not talking about thousands of gallons.

19!

l You are talking about just over a hundred of gallons normally.

So to have 20!

l them both come on really isn't a great deal of volume.

21{

i 22f INTERVIEWER:

Doesn't that happen periodically?

24 as "0 261 I

i l

e

30 1.

ZEWE:

Oh yeah.

2\\

3 INTERVIEWER:

Oh I see, even under the normal operation?

4 5

ZEWE:

Yeah, because like we know that the RV sump or the RV core the 6

AHE-11 fans, alright, that they have a pencil stream of condensate water 7

coming out of their drains all the time alright, so that this, added into 8

any other leakage that we have in the building, and we have some from the seals for the letdown monitors on the intermediate side of the letdown.

g They leak quite a bit and they go over to the sump.

So all this draining 10 into the sump, periodically the RV sump pumps come on I'd say once a shift g

normally just to pump that water down, which is really non radioactive waste water but it still goes into the sump.

14 INTERVIEWER:

I was trying to make something of both pumps running at the same time.

It looks like it was a big leak there and one pump couldn't handle it, so the second one had to turn on, since you say that is routine.

18 ZEWE:

Well, normally only one would run but I knew that we had a problem with the RC drain tank in there, blowing the water out of the RC drain 20 tank, and normally we have about 75 inches of water in the RC drain tank.

21' 22 INTERVIEWER:

What volume is that equal to?

23 24' 25j "O 262 l

I f

t f

31 1

ZEWE:

(Unintelligible)...

Because it's a horizontal tank that lays on it 2

side and I don't recall the volume changing once your up past the point of, 3

I don't remember there is a chart that says some many inches and so much 4

level, but it isn't one inch per quarter or whatever or eight gallons per 5

quarter, anything like that.

6 INTERVIEWER:

Did the one panel of pumps (unintelligible).

Didn't they run 7

a lot longer than you expected?

8 9

10 11 INTERVIEWER:

Didn't they run a lot longer than you expected?

Did they run you that long to pump a 100 gallons out or how long did it take?

14 15ll ZEWE:

Oh, no.

They normally don't run long, but I really don't have a feel for how much, because it depends on the filter conditions that they 16!

pump through with everything else, but really I hadn't dwelled on the fact that the sump pumps were on for so many minutes.

I didn't even realize that they were on as long as they were.

I was just going by the RV sump levels full and I knew that they should have been transfering the water 20 over to the aux building and I just didn't want it, so I just had to turn 21; it off regardless of how long they were on or how much was in there until I 22 had room for the water and I knew exactly where it had been.

23 24 25

^0 263

32 11 INTERVIEWER:

Bill there was a (unintelligible), there was a remark made 2

about you had an intimated alarm of the electrical type coming in, and you 3

weren't certain whether it was spurious or not.

Aand there was a comment 4

made that they couldn't see, I think it was off the aux transformer, they 5

couldn't see the effect.

Does that ring any bells?

6 ZEWE:

7 I was unaware of any inte'mittent electrical alarm, no.

The only g

one that Craig Faust, the one control room operator that was taking readings g

at the time, said that when he had alarms, electrical alarms, he mainly 10' f cused on, you know, that the generator breakers had opened up and that to any and the alarns, that he was over on the computer side of the control room, so the electrical panel you know is 30 feet from him, so you can't read the alarms unless your very familiar with their position but he was mainly interested, you know, in seeing alarms and he'd look and the breakers were open on the turbine trip.

So I don't know of any intermittent alarm i

151 that we had that night.

At least, I don't remember anything.

16 17 INTERVIEWER:

When were you aware of lines pressure tap line breaking off a i

one of the waste gas decay tanks?

Did you hear any stories like that?

19]

20 ZEWE:

No.

No I didn't at all, no.

I knew that we did have considerable 21 t ansfer of waste gas directly to the aux building from some unknown source, 22 because everytime that we tried to vent the makeup tank we would have a 23 pretty good release into the aux building, meaning that either the line was 24 open or that a valve was leaking by, or one of the relief lines in the 251 I

"O 264 t

J f

33 I

waste gas header was open.

But are you asking, at this time did I know 2

that or what?

3 4

INTERVIEWER:

No.

That right now, you didn't know that.

5 6

ZEWE:

No I didn't.

I didn't know it.

7 INTERVIEWER:

Somebody mentioned that to me the other day.

As far as the 8

g makeup tank, was that being vented at all?

10' ZEWE:

The makeup tank was normal.

The makeup tank, normally, we hold 3

about 18 to 22 pounds of hydrogen overpressure in the tank.

We were main-g taining that alright.

We weren't venting it at the time.

The real problem 3

that we had with the makeup system, the makeup tank, was later on whenever I

we were trying to degas the reactor coolant system, then we couldn't control 15) the pressure in the makeup tank, and that was quite sometime after that w'lich resulted in the release that we had on Friday.

We actually opened up the relief valve between the makeup tank and the makeup pumps and overfilled 18{

two out of the three RC bleed tanks before we finally terminated that.

19f i

That happened about 7:00 in the morning, I guess it was, on Friday morning.

20' 21 INTERVIEWER:

When did you finally get the hotwell pump down when you were 22 working on that?

23 24 "O

265 2si r

I

34 1.

ZEWE:

Well, we were rejecting it pretty heavily before I came up with the 2

control room and then I assigned an operator down there you know to just, 3 :

reject as much as he had to to get level back in the sight glass.

4f 5

INTERVIEWER:

Okay.

And you eventually did that?

6 ZEWE:

He eventually did that yes.

7 8

INTERVIEWER:

Do you remember the approximate time that you lost your g

heaters and pressurizer?

10 11 ZEWE:

Well, we started to get heater ground going that was sometime later.

I w uld say approaching mid morning alright, but I am not sure.

Whenever 13 we first had the problem and the pressure was lower than what I thought it i

should be, I suspected then that we may have had some problems with our 15j j

pressurizer heater capacity because we had blown the reliefs and it has 16 been a problem in the past that whenever the M20 area or the control building area west gets warm that we have considerable tripping of the pressurizer heater breakers, and I and I felt during that time period that we had lost some heaters and I had asked an operator, you know, to be dispatcned down there to check out the heater breakers.

But, as far as tM grounds bit and everything goes, that was considerably later and I would ;ay mid morning to 22 the best recollection on that.

23 24 "O 266 25 l

i r

35 1

INTERVIEWER:

So then you dispatched the operator down there, you didn't 2

have any report back to the, that you had lost breakers, heaters at that 3

time?

4 5

ZEWE:

As a matter of fact, I don't remember him reporting back at all on 61 it alright, because at that point, I just didn't follow up any more on it.

7 I asked to have it checked and I really didn't ask the operator myself.

I g

asked the control room to and have it done as soon as he had somebody free, g

and I not sure at this point whether he did check it and reported to the 10 perator and the operator didn't tell me, or he did and I didn't hear him or what have you.

But I don't remember any feedback at all from it and I really didn't pursue it ar.f further.

I didn't think about it much after 12 that.

3 14 ay, when the initial condensate pump tripped, feed pumps 15 tripped, were you operating three booster pumps and the two condensate 16, pumps at that time prior to

...?

18l ZEWE:

Three booster pumps, no.

There was two condensate pumps and two booster pumps and two feed pumps.

21 INTERVIEWER:

Okay, when the one condensate pump tripped and the two feed pumps tripped, do you recall whether the booster pumps also tripped?

23 24

"^0 267 2s i

l

36 l

1:

ZEWE:

Well when I first seen that we had a problem and I came out of my l

2; office and I looked, I had seen the turbine tripped, then the reactor 3

tripped just as soon as I took another step.

I announced a trip to let 4

everybody know and to get my shift foreman back to the control room alright, 5

and then we went through the starting the makeup pump and everything and 6

just about that time that I realized what caused the turbine trip alright, 7

I didn't realize at that point that we had lost feed and I didn't get, as, g

it was my one control roo:. operator that said, you know, that he was trying g

to get emergency feed and I realized that the cause for the turbine trip 10 was loss of feed and when I looked, we didn't have anything running; no 11 condensate pump at all, either one of them; no booster pump or feed pump; everything was tripped.

g 13 INTERVIEWER:

00 you have an idea as to when that was?

I 15!

16 17 18l 19 i

ZEWE:

It should be a couple of minutes into it, I would say.

21 INTERVIEWER:

Two minutes or five or six?

22 23 ZEWE:

I don't know.

I'd say just a couple minutes alright.

I had -- I think that Craig yelled out that we had lost all feed, alright, but I i

"O 268 i

t i

37 1

hadn't looked at it for say maybe two or three minutes later to verify the 2

fact that everything was off.

3 4

INTERVIEWER:

That was then about half way to the time that you realized 5

they still weren't feeding? About half was...

6 ZEWE:

Yes.

Because one of the things I noticed that as I traced it that 7

g appeared that condensate booster pumps continued to run to the sump pump g

and at the discharge of feed pump, it came back to through the sump to the feed pump and then discharged to the booster pump and that is what caused 10 that bang and I didn't get any indication of the computer printout booster g

pump pressure until five minutes later.

13 INTERVIEWER:

Five minutes later.

15j

]

ZEWE:

Time right then, you know, I really don't know but I had assumed it had tripped when the con feed pumps had tripped on their low pressure trip.

17 But they had low suction pressure trips on both the booster pumps and the 18(

feed pumps, so I assumed that that would take it out.

20 INTERVIEWER:

What are the booster pumps low suction pressure trips set at?

21 22 ZEWE:

They're at about 20 pounds.

They are really low, because normally 23 we run with a booster pump suction pressure of 50-60 pounds.

And the 24 suction to the faed pumps at about 400 pounds.

And the feed pumps trip at 25 about 280. Suction pressure is staggered, 275 and 280.

"^0 269

38 l'

1NTERVIEWER:

What was that 6ill? At what point do they trip? What's the l

2 set point on the feedwater trip?

3 4j ZEWE:

They trip at about.280 suction pressure, but A and B aren't exactly 5

the same point.

They are separated by about 15 pounds, so that one trips 6{

sooner.

But I wouldn't know why the booster pump would run after the 7

condensate pump was off.

8 INTERVIEWER:

I think the suction pressures came up.

The suction pressures g

10 stayed up.

I had a reading of like 38 pounds on suction pressure.

11 ZEWE:

g That would have held it up probably yeah.

13 INTERVIEWER:

The feedwater pressure was low enough for the feedwater pumps to trip, and it looked ike they tripped and the booster pumps continued to 15 nn, and then when that pressure surge came back and discharged to the feed 16 pumps through the suctions of the feed pumps and hit the discharge booster g

pumps and that kept the pressure up also for awhile.

I don't know how they 18,1 I

kept the suction pressure booster pumps up, but they were high for awhile.

201 ZEWE:

That I don't know, that is brand new to me.

I had thought they all went together because... go ahead.

23 INTERVIEWER:

Did you ever have a bang like that on loss of feedwater flow?

25

" O 2 ~/ 0 i

39 1

ZEVE:

We had, alright, we have isolated the condensate before while running 2

one condensate pump and one booster pump on feedwater heating, where the f

3 p lishers had gotten water in them and they had isolated themselves.

And I 4

was on for one such incident where we did have a tremendous hammer when 5

they went shut, but the pumps were only running at about 2500 gallons then, 6

not at the flow rate that we had now of about 16,000 to 17,000.

So I am 7

sure that the surge was a lot worse now, but I really didn't hear the hammer either.

I don't recall hearing it.

I did hear release going on the 8

feedwater heaters, which I thought was release on the feedwater heaters.

g 10' INTERVIEWER:

Okay.

Back on the steam generator feedwater pumps, emergency pumps, okay, did you take the steam driven pump off line?

13 ZEWE:

After we had established approximately 30 inches in the steam 14 15l generators, then the operator did secure the steam driven feed pump and the one electric driven feed pump and we just were feeding on one electrical 16 feed pump at that point.

18l l

INTERVIEWER:

Okay.

You indicated that the steam generator atmospherics 19{

and turbine bypasses, they all opened when the initial spike went right?

20 You went through the bypass valves and the atmospherics, and assuming all 21 the safeties are set at the right points, you picked up about six or eight 22 safeties on each generator that's if the pressure spike (unintelligible),

23 but after you started coming back down, you dropped off the code safeties 24 and then came back on the atmospheric dumps, did anybody take control of it 25}

f "O

271 i

l

40 1

manually then?

It wasn't covered in any of the interviews, so I am just 2

asking you how you were controlling your bypass flow in total?

3 4

ZEWE:

At this point, we should have been just on the turbine bypass valves 5

to the condenser, alright, and then at this point we should have been in 6

automatic, and we should have been controlling that ten-ten alright, as far 7

as their setpoint goes.

8 g'

INTERVIEWER:

Right.

That's with the 125 pound bias after trip.

10 ZEWE:

Right.

g 12

1. Y'

• #* Y "

"9 Y *" * "" "'

13 that you are aware of to do other than...?

i 15' ZEWE:

No, no.

Right.

As far as I know, that was in automatic the whole 6

f time at this point, until we began to isolate the B steam generator and so forth is when we changed it.

19f INTERVIEWER:

When you isolated the 8, you changed it?

21 ZEWE:

We isolated the bravo term bypass valve also, part of the isolation 22 part.

23 2a o 9 777 25

\\

I l

41 i

1!

INTERVIEWER:

So you just wrapped up the B generator completely, split this 2

down, and were steaming the A?

3 ZEWE:

Exactly.

4 5

INTERVIEWER:

How many times when you got into establishing circulation, 61 7

natural circulation, attempting to establish it, how many times did you g

attempt to establish, I mean, did you feed up the generators and try to establish it?

g 10 ZEWE:

Well we, as soon as we tripped the pumps, then we began to feed up to 50 inches or 50% on the operating range, 21 feet, on the A side only because bravo was still isolated, and then we felt that probably wouldn't be enough so then we started to feed up A higher than that since we only had one generator that would be used.

Bravo was still holding level, 15; alright.

At some point and I'm not sure what alright, when, but the bravo 16!

steam generator kept on coming up and up and up and then finally it leveled off at some time and then I didn't even, I didn't notice when it began to '

18 not come up anymore, but then sometime later we noticed that the Bravo 19 steam generator had stopped a ramp up and that we were actually somewhat holding a level in the B steam generator.

But I am not sure exactly when 21 that occurred and I am sure it is on the chart and it's all timed, but I 22 don't recall how long a period it was.

23 24

"'O 273 25!

l 1

I

42 1!

INTERVIL'sER:

When you isolate that steam generator, what valves do you 2

close?

3 ZEWE:

Well, we isolated normal and emergency steam to it, alright, the 4

5 n rmal startup valves, the normal feed reg valves, the block valves, the 14 6

valves automatically go shut when the startup valves go shut, and we isolated 7

the main steam valves, 4's and 7's, isolated the turbine bypass valves.

8 INTERVIEWER:

g What do you attribute the problem to early, or do you have any ideas on it? You were having trouble feeding the B generator?

0 11 ZEWE:

Yes, well, even had some problem feeding the A from the standpoint as that we found it was a little hard to control, the desired level.

It seemed like the 11 valve, the automatic valves were controlling level what 15ll we thought they would be at 30 inches, so we did have it in hand and then we found out that we had to overfeed and underfeed to hold the desired 16j levels that we wanted to hold.

And after awhile, we ended up to where the levels were going high on us and we ended up shutting the 11's, shutting l

the 12's, and finally cycling the EFV-5A and B to try to control level.

19 And then sometime thereafter, then we lost the breaker for the EFV-5 bravo.

It must have tripped on thermal.

So then we lost that.

So then sometime 21; thereafter then we went to normal feed using the one condensate pump feeding 22 into the normal startup valve and tnat is how we spent most of the day past 23 mid morning was using normal feed.

24 25 "O 274 i

43 1

INTERVIEWER:

Okay.

Say ten or eleven, that is when you did that type of 2f shot?

l 3

4 ZEWE:

May even been a little earlier than that, but in that neighborhood, 5

give or take, I'm not sure, because the operator automatically did it.

He 6

informed me later on, you know, that he had problems that he isolated, that 7

he had lost that breaker and everything, that he was going back on normal 8

feed, and at that point, I didn't have any objections to it because he could control it better.

g 10' INTERVIEWER:

Right.

So then we got a lower discht.rge pressure that should 17 keep the valves open so that...

13 ZEWE:

At that point, we were down rather low on generator pressure anyway.

I 15' INTERVIEWER:

Is there a procedure, now I am not familiar with your plant, 16 so I am just asking this off the wall, is there a procedure for feeding a dry generator that is different from feeding a normal generator, you know, a generator with say, 30 inches or more in it?

20 ZEWE:

Yes.

Well the thing was alright, at the time, I never considered it to be a dry steam generator.

23 INTERVIEWER:

You hadn't any indication.

25 1

t f

44 ZEWE:

1{

Because everytime that I looked, I had greater than ten inches.

2.

Normally, we consider a dry steam generator eight inches or less alright, 3

and I always had feed.

If I had realized that it had taken us 8 minutes to 4

establish that feed, which I didn't, I'd have known that it had to been 5

dry, alright.

But even on the operating range, it indicated like 5% or 6%

6l n the operating range, and it showed greater than 10, 11, 12 inches on the 7

startup range, alright, so I didn't think that it was dry, that it was 8

still coming back, and that we were actually holding some level before we'd g

seen any significant increase.

But knowing that it was actually an 8 minute 10 peri d here that I was talking about, it would've gone dry.

11 INTERVIEWER:

The reactor coolant high radiation alarm (unintelligible)?

13 ZEWE:

Right around the same time that everything else happened.

That's when I n ticed the dome monitor was coming in hot.

15 16i 17l INTERVIEWER:

By everything else, you mean all the other radiation alarms?

18l ZEWE:

All the other alarms, right.

I really don't remember seeing very 19l much up until that point and then everything began to ccme in at once.

Really, the first one that I noticed was the 748, the offgas monitor.

That was the first one because I felt sure that we did have a problem with the B steam generator and then once we fed it, it came in, I was sort of keyed to 23 that one anyway.

24 2s "O 276 t

I

V 45 l

1 INTERVIEWER:

Does anybody else have any...

I've got a question; before 2

you ran the coolant pump or after?

3 4

ZEVE:

After.

5 6

INTERVIEWER:

All the radiation...

7 ZEWE:

It seeined to me, afterwards, right.

But we did one and then the 8

g other in fairly rapid succession, you know, just a few minutes apart, so 10 Fred started the pump and then it was x number of minutes after that that everything came in.

3 12 :

INTERVIEWER:

3 Do you know or could you differentiate between where along that line of pumps that it came in? Was it before you ran the last one that stayed on or was it somewhere in between?

15 16 ZEWE:

Before I ran the last one?

18I INTERVIEWER:

Well I am thinking back to one of the other...

19{

20 ZEWE:

We had started another pump, alright, sometime after this, alright, because we started that 2B pump and then we were concerned if it was really running or not, and we checked the amps and everything and we knocked it off.

And then we tried to start another pump which was a A side pump 24 sometime later, alright.

I am not sure of the time frame.

It may have 25

'"O 277

46 1.

been an hour later.

I don't know.

But then we tried to start another one 2

and that one acted normally at first, alright.

It started and we got the 3

amps were real high on the offscale amps.

Normally the amps, you know, are 4

offscale for about the first 10 seconds and then they come down to about 5

600 amps normally alright.

But the first pump that we had on, I never 6

really seen the amps, you know, the starting current come up and then come 7

down; but on that pump, the starting current came up; in ten seconds she 8

dropped back down but she kept on going down to about 100 amps.

So we felt g

sure that, alright, at that time, we weren't 100 percent sure that that 28 10 had been running actually before alright, but then the A pump that we ran 11 nce we had the starting current, she held and she came down and we were 12 sure that she actually had started.

13 INTERVIEWER:

And it stayed on, okay?

g 15l l

ZEWE:

Yeah.

16 17 INTERVIEWER:

And that meant that the other one had done the same thing.

i 19l ZEWE:

Yeah.

But then...

20 21.

INTERVIEWER:

Where did that alarm come in?

22 23 ZEWE:

I didn't have any flow indication, there.

25r i

"O 278 I

l

1 47 Il INTERVIEWER:

On ne!thar of the two runs?

t 2

3 ZEWE:

Neither of the two runs, no.

4 I

S INTERVIEWER:

Where did the alarm come in along that pattern of starting 6

pumps?

7 ZEWE:

Alarm of what?

8 9

INTERVIEWER:

I thought that's where you got one of your radiation, high 10, radiation..

17 12 ZEWE:

Sometime after we startad that 28 pump.

13 14 INTERVIEWER:

After the 2B pump.

15l I

16!

ZEWE:

That is as I remember it.

I 18!

INTERVIEWER:

The last cumo.

19j l

20 ZEWE:

Which was the first one.

21 22 INTERVIEWER:

Or the d est pump, okay.

S.; it was the first one of the 23 others.

24 2s

^^0 279

48 ZEWE:

1; Yes, that's as I remember, because there was sometime between the 2

first pump start and second pump start.

It was, I'd say, maybe an hour.

3l I don't...

4 e

INTERVIEWER:

Any other.

One thing, is there any power dispatcher records 6

that would show the spike maybe when these pumps started?

7 ZEWE:

The coolant pumps?

I don't remember if they would have recorders 8

that would see.

Well, there again, I don't remember but I don't think so.

g 10 But there is a possibility that they do.

Maybe the Lebanon dispatcher 11 might have that, but I don't remember, alright.

The only place that I've 2

ever been is at the Reading dispatcher and I don't recall seeing it up there.

3 But sinee the, I have never been to the Lebanon dispatching, I don't know.

A would think that if thay would, they would have gotten the 15j!

second pumps to start with the amps and I am not even sure that we just missed seeing the starting current on the 2B or not, but the operator remarked that he don't see it either, and that is one of the first things that he checks as he starts it and he times how high the amps are up for a 18{

l locked rotor indication alright, and he don't remember seeing it at all 19l i

either because that's when he closed it in and the breaker went right, red, 20!

night, meaning that it closed.

But he didn't have really any other indica-tion because 100 amps is just about the first mark on the amp gauge there.

22 It is really hard to tell, so we weren't sure at that point.

The second 23 l

one, I did see it come up and ccme down.

24j 2s "O

280 l

i f

r A 49 1.

INTERVIEWER:

Any cther questions? Anything else.

Okay that ought to do 2h i t.

3 4'

5 6

0 281 7

8 9I 10 11 12 13 14; 15; 16,!

l 171 18 19 20i 21 22 23 24 25t i

f,

,