ML20129D376
| ML20129D376 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 06/17/1985 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| Shared Package | |
| ML20129D374 | List: |
| References | |
| NUDOCS 8507300011 | |
| Download: ML20129D376 (59) | |
Text
,_
3 % 4 lL' L & ute p
- s ', ).
fbb
- )g o
~
UNITED STATES OF AMERICA
[
WUCLEAR REGULATORY COMMISSION
.,y...
'In tlse matter of MEETING BETWEEN THE NRC STAFF AND TOLEDO-EDISON COMPANY CONCERNING j
AFW SYSTEMS l
~
l fJ-bdN A
Docket No.
I I
m Location: Bethesda, Maryland Date: Monday, June 17, 1985 Pages:
1 - 50 8507300011 050619 DR ADOCK 05000346 PDH E PILEY & ASSOCI.%TES Court Reporters
(
1578i I fit,
wN Suite 921 Washington, D.C. 20006 (202) 293-3950 S
1
! s i
8 l
1 UNITED STATES OF AMERICH 2
NUCLEAR REGULATORY COMMISSION l
3 4
MEETING BETWEEN THE NRC STAFF AND TOLEDO-5 EDISON COMPANY CONCERNING RFW SYSTEMS 6
j l
8 Room p-110 9
phillips Building 10 7920 Norfolk Avenue 11 Bethesda, Maryland 12 Monday, June 17, 198b 18 14 The meeting in the above-entitled matter convened, 15 pursuant to notice, at 10:30 o' clock, a.m.,
Mr. Harold Denton, 16 presiding.
e 17 NRC ATTENDEES:
10 H.
Denton NRR 19 B.
Sheron DSI 20 O.
Parr DSI 21 W.
Houston DSl 22 J
9tnfz NRR 23 F.
Rowsome OST 24 A.
DeAgazlo NRR 25 D.
Eisenhut NRR
(
i f\\
t i
2 1
NRC ATTENDEES (continued).
2 A.
Thadani Safety Technology 3
D.
Wessman Division of Licensing 4
A.
El-Bassioni PRAB 5
B.
Newlin public Affairs 6
G.
Lainas NRR 7
F.
Miraglia NRR 8
9 LICENSEE ATTENDEES:
l 10 J.
Silberg, Esquire l
i 11 B.
Fink bi
(,j 12 R.
Gradomski 13 B.
Peters 14 T.
Myers I
15 S.
Jain 16 M. O'Reilly 37 18 ALSO PRESENT:
\\
19 L.
Connor Doc-Search Associates 20 R.
Borsum Babcock & Wilcox 1
21 J
Nurmi Engineering planning & Mgmt 22 M.
Ryan Nucleonics Week
'? t 24 25
s a
vs 1.
3 1
p ROC EED 1 N GS 2
(10:30 a.m.)
3 MR. DENTON:
Good morning.
My name is Harold 4
Denton.
I want to thank you for coming in on such short 5
notice.
6 The purpose of this meeting es to develop a 7
chronology of actions that have been taken on the feedwater
~~
8 system at Davis-Besse and the studies which have been done 9
leading to those actions.
10 I realize I didn't give you much lead time to 11 prepare for this n.aeting, and all this information was not yet i
12 available.
But I have had the Staff go through our files on 13 what has transpired between us with the steamline accident, 14 and I wanted to be sure we had a complete chronology developed
~
15 on that point.
16 I think there have been some questions asked since
~~
17 the incident at Davis-Besse as to what the status of 18 operations is and when st would be completed and this mort of 19 thing, and I would like to get a clear understanding of what 20 you have done, what studies you've done, and we'll chip in 21 with our knowledge of the precess where we can.
d2 Bet' ore we oog e r),
I thenk it might be useful to go verd
+b7 c..c+
r>f o a '- h y.rson itse*Ify t h eem ' se
?. n d their 24 responsibilities for the record.
25 Why den't we start, AI, with you.
l n
l
. =
f I
8 4
I am Al DeAgazio, Davis-Besse n
2 Project Manager in the Division of Licensing.
S MR. STOLZ:
I'm John Stolz, Branch Chief of the 4
branch to whsch Davis-Besse is assigned.
g i
I 5
MR. RDWSDME:
I'm Frank Rowsome, Division of Safety I
l 6
Technology.
7 MR. EISENHUT:
Darrell Eisenhut, Deputy Director of B
NRR.
k 9
MR. DENTON:
Harold Denton, Director of NRR.
10 MR. SHERON:
I'm Brian Sheron, Chief of Reactor
-6 11 Systems Branch, DSl
. n l
12 MR. PARR:
Dlan Parr, Chief, Auxiliary Systems-e l
}
13 Branch.
14 MR. HDUSTDN:
Wayne Houston, Assistant Director for 15 Reactor Safety and today Acting Division Director, DSI
~
e 16 MR. THADANI.
Ashok Thadani, Chief, Reliability and 17 Assessment Branch.
te MR. WESSMAN:
l*m Dick Wessman, representing the 19 Branch Chief of the Division of Licensing.
20 MR. FINK:
Bob Fink, MPR Associates, consultant to 21 Toledo Edison.
22 MR. GRADOMOKI Rich Gradomski, Engineer.
??
Mn PETERS-Rnb Peters, l.ir,ensing Manager.
24 MR. SILBERG:
Jay Silberg, Shaw, Pittman, potts &
25 Trowbridge, Attorneys.
1
(
a e
I, a
5 1
1 MR. JAIN:
Sushil Jain, I'm with Tole,do Edison.
2 MR. MYERS:
Director, Core Safety and Licensing.
3 Toledo Edison.
4 MS. BRYAN:
Margaret Bryan with " Nucleonics Week,"
5 MS. NURMI.
Joy Nurmi with Engineering Planning &
6 Management.
\\
7 MR. NEWLIN:
Bob Newlin, Public Affairs, NRC.
8 MS. O'REILLY:
Mary O'Reilly, Toledo Edison.
l 1
9 MR. DENTON:
We have two more seats for anyone who 10 would like to move forward.
)
11 MR. BORSUM:
Babcock & Wilcox.
12 MS. CONNOR:
Lynn Connor, Doc-Search Associates.
13 MR. MIRAGLIA:
Frank Miraglia. Deputy Director of 14 Licensing, NRR.
15 MR. LAINAS:
Gus Lainas, Assistant Director, 16 Operating Reactors, Division of Licensing.
.~
17 MR. EL-BASSION1-Adel El-Bassioni.
I'm with pRAB.
18 MR DENTON For the benefit of t i,e pecple who aunt 19 arrived, what we're doing today is developing a chronology of 20 actions that have been taken at Davis-Besse regarding the 21 feedwater system and the auxiliary feedwater system since the di TML accident We do not intend to get into any discussions of M
sh a i &sppened a c a r. t ' y
- t D 4 v i 0 !1a... a D.; '. ihi; i.s e
back 24 to some of the earliest actions that were taken in the sumnie r 25 of '79 and to bring those forward.
And I would like to go
s i.
6 1
forward right up until your plans for late this, year regarding l
2 completing the operating of the auxiliary feedwater' system, 3
including the larger pump associated with the electric drive l
t 4
system for the auxiliary feedwater.
)
5 So at this point, Al, unless there are some other 6
preliminaries to get into, I propose that we let Toledo Edison 7
have the floor and give us your understanding of the actions 8
that have transpired.
9 MR. DeAGAZIO:
I think we can probably do that.
10 MR. MYERS:
I am here on behalf of Dick Crouse, who
. A11 apologizes for not being able to come down here.
We have 12 brought several of our key individuals from the engineering 13 and licensing area, who have been involved with the auxiliary
)
14 feedwater system over the years.
It was our intent to 15 exchange information today at your request from Saturday, 4
e 16 much as we could, and if anything was left unvalidated or
~
17 prelim 6 nary, like much of our information is that we developed 18 yesterday and thin morning, we would ves ify that in time to 19 support any further use of that information that we've been 20 involved with that perhaps people have not been aware of over 21 the years.
22 MR. DENTON:
I think what might be useful would be
.M f ur-us to 4 va je 4 p=s ; ud ai t 5 6.e io ruv i a.s (1.3 tranac.ipt 24 and then provide any additional clarifying material that you 25 think might be appropriate.
s s
e e
7 1
MR. MYERS:
We appreciate that 2
The intention we have this morning was to provide a 3
short description of our auxiliary feedwater system, 4
recognizing several people may be new to it or have not been 5
involved recently in it to bring the system back into focus.
6 1 think most of the indsviduals or many of you here have b e e r, 7
involved over the years since TMI with Davis-Besse, but I will e
give you a short review of its operation and its basic design 9
and then go into both the procedural and the modification 10 activity that we've been involved with since TMI.
11 Again, much of that has been exchanged with you.
12 Some of that has not specifically been identified to you.
The 13 intent today is to come up with a complete list.
Hopefully, 14 that would give you both of those packages.
15 As I said, we have some of our people here, but some 16 of our key people are also back supporting the fact-finding 17 team.
So if we need to validate any information or if there 19 are burning questions, wa cari probabt, still dc that teday by 19 phone.
But to at least give you an overview of what those 1
20 changes have been, we should be able to supply that here.
21 We are working again back at Toledo to actually 22 identify actual i n:s t a l l a t i on and modification dates for you to ER qive wea r en accaJr=+o chronoloDie of 9xactiv *when + b e c. c I
I 24
' modifications were in place.
We can give you estimates today, i
1 25 to the best of our knowledge and what we were able to c:ome up
e i
I e
8 1
with yesterday.
But we wi l l be validating, an& you should 2
recognize this is preliminary information, so we can append 3
the transcript, possibly, with any changes that we would find 1
4 in the interim to validate that information for you.
5 I would like to introduce now Mr. Jain, who will be 6
describing in general the system function.
I had expected a 7
small,or meeting, so unfortunately our overhead doesn't seem to 8
focus large enough.
9 We have a system layout drawing that we will pass 10 around that will help in the discussion.
11 MR. DENTON:
Do you have sufficient copies?
If not, 12 we would make some.
13 MR. MYERS:
I think we do, hopefully.
14 This is a simplified drawir.g and is not the actual, 15 exact detail that's in the field.
Any differences will be 16 pointed out as we go through, but I think it will offer a 17 basis for discussion.
18 MR. JAIN:
Davis-besse right now has two 19 independent, redundant trains of auxiliary feedwater systems.
20 The mode of power provided right now is a steam turbine, and 21 each train, which drives one full hundred percent capacity aux 22 feedwater pump, normally is taking suction from the condensate s t ne m gr+
tank f ne Fe. sm i r.
I su9 plies and the ca raty grade-24 supplies the service water system, for which we have an 25 automatic switchover from the condensate to the service water
9 1
in the event of a LOCA condetion from the condensate storage 2
tank.
3 On the discharge side, we have each aux feed pump 4
discharging to its respective steam generator with provisions 5
for crossties, so that each aux feed pump can feed either 6
steam generator at a given time.
We have similar crossties on 7
the steam inlet side on the aux feed pump turbine.
Normally S
the No.
I steam generator would be providing steam to the 9
No. I aux feed pump turbine, and the No. 2 steam generator 10 would be doing that for the No. 2 aux feed pump turbine.
We 11 have similar crossovers where any steam generator can provide l
C 12 steam to either aux feed pump turbine.
13 The initiating system for the aux feedwater system 14 is what we call a steam and feedwater rupture control system.
15 The SFRCS, as we call it, is actuated on four different 16 parameters ranging from steam at low-level for e
17 loss-of-feedwater condition, a loss at the coolant pumps to 18 promote natural circulation, and then a steamline break, which 19 is a low steam pressure condition.
20 The alignment of valves on the discharge of the 21 pumps, as well as the steam inlet to the turbines, is 22 dependent upon which steam generator is good, meaning which
- ?
h a s ey.ed precsure
= S ' e ? "- qart'c"
- 1 1 In " 3
'3 1
24 considered to be a bad generator, and auxiliary feedwater is 25 isolated from that generator, and so is the steam coming from
________----J
r 10 1
that steam generator.
2 The system that we have right now is essentially all 3
safety grade components, both on the steam inlet as well as on 4
the discharge side, with all the motor-operated valves 5
supplied fece either the diesel, in the case of loss of 6
offsete power, or from the DC power supply from one train of 7
the aux feedwater system.
8 Tgat is the basic overview of the system, very 9
briefly.
10 MR. DENTON:
What is the function of the system 11 labeled " steam generator wet layup generation and
-C
,_ x, 12 recirculation pumps?"
13 V9. JAIN:
This drawing includes two systems.
The 14 basic intent here is to look at the aux feed pumps.
The 15 recirc pumps are used when you have flooded the steam 16 generator during a shutdown condition, and you have the steam
)
l
~~
17 generators in a wet layup condition, and then you want to 18 insert to maintain chemistry, and you have steam generators 19 flooded up to the top nozzle.
20 MR. DENTON:
So that's merely a recirculation 21 system?
22 MR. JAIN:
Correct.
That is merely relevant to what
'?
v?**o
? m f f=
.}
.s 5 -*
- today 24 MR. DENTON:
Let's stop and see if the Staff has any 25 questions about this system.
i.
11 1
MR. STOLZ:
Were you also going to discuss the 2
startup pump and its relationship to supporting the aux feed 3
system that you show here?
4 MR. MYERS:
Verbally describe that 5
MR. EISENHUT:
Before you do, though, if we could 6
maybe on the aux feed system, with what you were just 7
desceabing, the normal mode, then, on any event such as a low 8
steam generator level would be that both aux feed pumps get a 9
signal to turn on?
10 MR. JAIN:
Correct.
11 MR. EISENHUT:
And they actually get a signal to i
12 turn on, and are there any valves that have to open?
Where 18 would the normal water supply be coming from in such an 14 event?
Could you just describe the scenarlo of what would 15 have to happen to turn on the auxiliary feedwater?
16 MR. JAIN:
For a steam generator low-level 17 condition, there is one valve that has to open to supply steam 18 to the aux feed pump turbine, and those are shown on the left 19 top of this drawing.
But for just a low-level condition, the 20 respective valve from the No.
I generator would open the 21 No.
I aux feed pump turbine.
With respect to the valve for 22 the No. 2 generator, it would open the No. E aux feed pump l
dd turbene.
i i
24 MR. EISENHUT:
Could you tell rae which those are on 25 this diagram?
l
12 1
MR. JAIN:
100 and 107 is for the No.,2 aux feed 2
pump turbine.
3 On the discharge side, I will tell you what the 4
configuration is today, because we have changed it over the 5
last few years.
In the configuration today, there will be one 6
discharge valve that would have to open on a low-level signal, a
7 and that would be 3870 and 3872 for the other generatcc.
8 MRb EISENHUT:
So given an event of a low-level or 9
dry generator, for whatever reason, upon the loss of or 10 low-level upon the loss of the main feed system, those four
-g
. A 11 valves would open, and that would normally turn on the a,u x
)
12 feed system, and it would deliver flow to the steam generator?
18 MR. JAIN Correct.
The normal water supply in this
- 4 case would be the condensate storage tank.
15 e
16 17 18 19 20 21 22 24 25
13 0
1 MR. ROWSOME:
Would not AF-300 also have t o oper.?
I 2
MR. JAIN:
This is an older drawing.
AF'300 and 3
388, whicP, were also on the discharge of the pump, they used l
4 to be speed control valves.
It used to open on an rpm of 5
2800.
We have since deleted that interlock in order to 6
improve the reliability, it's one of the chareges l*ll be 7
talking about later as to what has transpired.
8 MR. ROUSOME:
Okay.
They're normally open, then?
9 MR. JAIN:
Yes.
10 MR. DENTON:
Could you start your description of the 11 changes, beginning about the time of the TMl accident?
{
i
'12 MR. MYERS:
Excuse me.
We had one other question 18 that John wanted, having to do with the current use of the 14 startup feed pump and how that would be utilized.
We can do 15 that before we get into.that.
16 MR. JAIN:
The startup feed pump, as it is
)
\\
17 configured right now, delivers water to the main feedwater 18 nozzien into the steam geriorators.
It takes.uct. ore or at
]
19 could also take suction from the condensate storage tank.
j 20 That's the normal alignment for the startup feed pump.
We 21 could go into detail as to what actions are required for it to 22 be put in service, given a enmplete Inss of feedwater -- main 28 and auxiliary feedwater 24 MR. DENTON:
What is its normal function?
25 MR. JAIN:
The startup feed pump is basically there
l 14 l
1 to support low-power operation and for feedwater cleanup for i
l 1
2 the steam generator chemistry again.
3 MR. DENTON:
And the capacity of that system again?
'1 4
MR. JAIN:
It's 300 gpm.
It's not enough to remove 5
decay heat by itself.
6 MR. LAINHS; Was that about half the capacity?
7 MR. JAIN:
Yes.
~
8 MR. MYERS:
Any other questions as far as the basic 9
operation currently?
10 MR. THADANl That startup pump, what is the mode of 11 power for the pump and the associated valves?
12 MR. JAIN:
It's supplied from a nonessential 4160 13 bus, but we have provisions for it to be backfit from an 14 essential 4160 bus.
15 MR. MYERS:
Those are also discussed.
That activity 16 is also discussed in the changes we've provided over the
~
17 years.
18 Anything else?
19 Okay, we have Mr. Jain will also continue, then, 20 if you want to talk about the equipment and associated 21 procedural changes.
Again, some of these -- the exact times 22 whmn t reey ' r e implemented, we're in the process of valadating.
?
- Ja *, e w 1
.. " p ! s. em 1: en - P.
t ;c e s,. t! (Im
.* r 1j te 3eo 24 but we would jockey their positions, based on the actual times 25 we would hopefully get back to you today or tomorrow or in the
s 15 1
transcript here.
2 MR. JAIN:
Originally the Davis-Besse aum'iliary 3
feedwater system motor-operated valves were all AC power 4
operated.
We made changes so that the No.
I train over the 5
auxiliary feedwater system has valves which are all DC power 6
operated, so one t r a i r.
of valves is entirely independent of HC 7
power.
This change, we think, was done sometime in 1980 in 8
order to reduce the probability of human errors on manual 9
valve mispositioning in the flow path of both the aux 10 feedwater system as well as the startup feedwater system.
So 11 we have put padlocks on the local handwheels.
We have also 12 put padlocks on local handwheels, as well as the pushbutton 1E stations for vital motor-operated valves in the field.
So 14 nobody in the field could inadvertently misposition those 15 valves.
16 There is an administrative procedure in place which 17 is a control on the positioning of both the manual valves, as 18 well as the motor-operated valves.
19 We have made provisions for the startup feedwater 20 control valve and the feedwater block valve to be controlled 21 from the control room by the operator.
These are the two 1
22 valses that get 12c ated on a steam and feedwater rupture
??
ea-trn!
-vit=*
vhich in tha leitl=*'nq
- y4 tem (n*
tha 24 auxiliary feedwater system.
So for a loss-of-feedwater 25 condition, these valves will get isolated.
However, they will 1
16 1
have to be reopened to provide startup feedwater to the steam 2
generators for a complete loss-of-feedwater condition.
3 We have provided the capability, so that the SFRCS 4
signal to these valves can be blocked from the control room, 5
and these valves can be reopened in order to expeditiously 6
provide feedwater to the steam generator and the startup feed 7
pump.
~.
8 Mg EISENHUT:
Can you point out where they are on 9
this?
10 MR. JAIN:
I'm afraid this startup feed pump is not M
11 on this drawing.
2,
)
/
12 MR. EISENHUT:
But the valves also, they're not 13 on here, obviously.
Okay, thank you.
14 MR. JAIN:
Okay.
After the TMI accident, we were to 15 demonstrate that the aux feed pump turbine could be driven on 16 a dry steam generator.
We actually conducted a test where we 17 ran the steam generators dry and started the aux feed pump 18 turbine with a dry steam generator.
The test was successful.
19 The SFRCS --
20 MR. DENTON:
Let's stop there.
Can you describe a 21 bit more about the test?
Is that because when it's dry, it's 22 not really dry?
99 Me
.antN It's dry in
+ke Snrse
+b+*
b=-
b't*'td 24 up steam in it with all the isolation on the feedwater as well I
(
25 as the steam side.
a e
17 1
MR. DENTON:
When you ran the test, d,o you recall 2
what the pressure was?
3 MR. JAIN:
I don't recall personally.
We would have 4
to look at the test results.
5 MR. EISENHUT:
But the reason you ran the test was 6
the concern that the pumps would not have any inlet flows 7
sufficient inlet flows so that they would overstrip and
-~
8 overspeed?
9 MR. JAIN:
The reason to do that test was to 10 demonstrate that you had enough power, if you will, in the 11 steam generator that could initially roll the turbine, and 12 once the turbine was rolled, you had enough water going into 18 the steam generator to start the cycle, the production of 14 steam, and then running the turbine again.
15 MR. EISENHUT:
But yet enough to actually start the 16 turbine flow?
17 MR. MYERS:
There was a question -- the question 18 we're trying to answer was, in all the lines and everything, 19 once the system was bottled up, would we lose the motor force i
20 prior to getting water backflashing to raise pressure to 21 recover the motor force.
1 j
00 MH.
THADHNI Can you tell me two things?
Number I
[
ao l
on.
s c.y tenr fia, ao t&'
"'r' 1
1 24 Number two, is there any implication in terms of i
25 controls?
Steam inlet valves, for example, and so on.
The 1
18 the speed of the turbine, what happens?
1 speed of the pump 2
MR. JAIN:
I'm afraid I cannot answer that. right S
now.
I don't have firsthand knowledge of that.
4 MR. MYERS:
The test engineer is not here.
We can 5
get that information in the detailed test.
As a matter of 6
fact, I believe that was even submitted.
We can dig those 7
details up.
~
8 MR. DENTON:
Do you recall if this is one that we k
9 asked you to perform, or you performed it on your own 10 initiative?
"i 11 MR. MYERS:
1*m pretty sure the details were g
. +
/
12 provided.
I don't know whether it was one that was asked for 13 or was part of our program anyway.
I'd have to -- we changed 14 Project Managers and a lot of people since then.
I'm not sure 15 that we can dig that up,*too.
16 MR. DENTON:
I think we've had a turnover of e
17 personnel also in the intervening five or six years, and it's 18 difficult for us to reconstruct all the details.
19 Let me ask just about dryout When you use the term i
20 "dryout," what do you really mean with regard to the water 21 level in the steam generator?
What does B&W normally mean 22 when they talk about dryout conditions?
Is that no liquid 23 remaining or 24 MR. JAIN:
Well, the procedure that the operators i
again, you have to realize what's 25 are told to use is i
I
19 1
happening to the level instrumentation, because the level 2
instrumentation doesn't read the actual level it'gives you 3
the indicated level, based on the conditions it was calibrated 4
to and so on.
5 The procedure they have been given is eight inches G
in the steam generator indicated, eight inches endicated et 7
could be anything actual, depending on what the steam 8
generator pressure is.
9 MR. DENTON:
I was asking in connection with that 10 test you performed, whether there was water still in the 11 bottom of the steam generator, or whether it was 12 MR. JAIN:
I think the assumption was that there was J
13 no water, but we will have to dig further into that.
14 MR. EISENHUT:
A correlated followup question, if 15 you could look at it at the same time.
The B&W eight inches 16 left in the bottom of the vessel, is that the definition of a e
17
" dry generator?"
What's the instrument error?
18 I think eight inches i s B& tJ --
19 MR. MYERS:
It's a determination by the ATOG 20 guidelines, which utilizes an either/or level, eight inches or 21 below, or pressure below, and I believe that's 960 pounds.
22 IJhe t h e r there is water er not, you should assume that that 23 generator sa dry 24 MR. DENTON:
Why don't you go ahead, then, with your l
25 chronology?
i
i 1
20 1
MR. JAIN:
The steam and feedwater rupture control 2
system, as well as the auxiliary feedwater system. 'u t i l izes 3
several pressure switches, either for actuation or for 4
protection of the pump or the steamlines.
Initially, for a 5
few years after we started up, we had recurring problems with 6
the pressure sw i f e.hes They were failing at a very faat rate 7
because of corrosion problems in the diaphragm.
8 We have since gone to a modified design for the 9
pressure switches, and since then, the failure rates of these 10 pressure switches have gone down considerably, which 11 essentially improves the reliabity, because it eliminates 12 the potential failure mechanism for valves going inadvertently 13 closed or so on.
14 As part of the HUREG-0737 requirements, we installed 15 a control-grade flow indicator in each train of the auxiliary 16 feedwater system, and we also installed a safety grade flow a
17 indicator on each train.
These indicators have since been 10 tech spec'ed 19 MR. MYERS:
Those were part of the 0737 20 requirements.
Specifically the control grade, I believe, was 21 part of the original startup after TMI, and 0737 required a 22 safety-grade upgrade.
23 MR. STOLZ:
Ted, as you go along, can ynu indicate 24 the dates when these were implemented?
25 MR. MYERS:
We're going to try to do that wherever
.l
21 1
we cans however, the concern is that the best d,ates that we 2
have right now are off the top of a couple of engineers' 3
heads.
We will get those dates specifically down, so we can 4
have a chronology.
Where we have them, we'll provide them.
5 The control grade was prior to restart after TMI, 6
and the safety grade was sometime in 1982, I think it was, 7
1982.
~~
8 MR. JAIN:
We have also modified the control room 9
annunciator in order for the operator to be better able to 10 diagnose what has caused an aux feedwater pump or train 11 inoperability, given that the pump has been called upon to 12 actuate during an accident or a transient.
18 Using that revised annunciator window and the 14 computer printout, we can tell what exactly has gone wrong 15 with the aux feedwater train and take remedial action to 16 correct it.
~
17 MR. LAINUS:
These are additional annunciators?
18 MR. JulH:
This was an addstional annunciator, for 19 which we revised the logic, so it's for him to tell exactly 20 what went wrong.
l 21 As I mentioned earlier while describing the system, 22 we deleted the speed switch interlock from the aux Isary
( e 4 +.,.3 t e r n'.wr d I q e ' t + q-
., f is s -
r.4
'nf*
'.2e va r --
- v.,
. v 24 open wLth the local handwheel and the pushbutton stations l
25 locked to minimize the number of valves that had to open on
l 22 i
1 the discharge side.
2 In terms of the analyses we have done a f t.e r TMi, the S
startup and makeup pump pORU analysis, which talked about 4
using the startup feed pump, if you lost both trains of 5
auxiliary feedwater, there was also a B&W reliability analysis 6
done as part of NUREG-0737 requirements, which was done on the 7
auxiliary feedwater system, and then in December '81 we 8
submI'tted a detailed PRA analysis of the Davis-Besse aux k
9 feedwater system and the several improvements tha' ;.. w ue 10 made to it to improve the reliability overall.
M 11 MR. DENTON:
Is that the one that was done by EDS7 s.
.I 12 MR. JAIN:
Correct.
13 MR. DENTON:
Was there also a Bechtel study?
Was 14 there a Bechtel study of the aux feedwater system reliability.
15 in addition to the EDS report?
16 MR. JAIN:
I don't think there ever was a Bechtel e
17 reliability.
There was a B&W aux feedwater reliability 18 analysis done and submitted to the NRC.
19 MR. DENTON:
Just to be sure I understand, you have 20 mentioned so far three reports:
the original report done in 21 response to bulletins and orders, and then what was the second j
22 report that you mentioned, the one done by B&W7 23 MR. JAIN:
Right.
24 MR. DENTON:
That was the reliability analysis of 25 the auxiliary feedwater system.
And then the third report was
23 1
the reliability analysis of the auxiliary feedwater system 2
also, but done by EDS?
3 MR. JAIN:
Correct To a different scope and to a 4
different detail 5
MR. EISENHUT:
Let's see, in fact, the last one that 6
EOO did was really thw repve t submitted under NUREG-075/ ot 7
the pRA study?
8 MR. JAIN:
No.
We never claimed that.
The B&W 9
analysis was part of the NUREG-0737 requirements.
The EDS was 10 sobmitted basically to assess what we in the company could do 11 to improve the overall reliability of the aux feedwater 12 svstem, what options.
13 MR. EISENHUT:
Was it the study that concludes that 14 a third train of aux feedwater is not required?
15 MR. JAIN:
Correct 16 MR. MYERS:
The conclusion was that there was equal e
17 improvement in reliabilities that could be made in utilizing 13 an installed pump and other pr wwww s.1 activsties am the th e r d 19 pump, part of that third capacity full pump 20 21 22 29 24 25
24 1
MR. JAIN:
There have been some procedural changes 2
-- excuse me.
I'm sorry.
We have also undertaken a governor improvement program, 3
comprehensive government 4
and Rick Gradomski is going to talk about that 5
MR. DENTON:
We could also probably use an overall 6
government improvement program, too.
7 CLaughter]
8 MR. SRADOMSKI.
I am Rick Gradomski from Toledo 9
Edison.
In % peaking to the governors that are installed on 10 the Terry Turbine / Byron Jackson pump auxillary feedwater pump 11 turbine system, the Woodward governors that were originally "5
A-12 installed on this system were Woodward pGpL pneumatic type 13 governors that were modified with the addition of a Bodie 14 motor attached to the hand control knob to accept electric 15 pulse signals from the steam generator level control 16 There are two major points to this program other 17 than the fact that we continue to try to adjust and make
~
IS repairs as necessary to the modified pGpL system.
19 MR. DENTON:
When did this system go into operation?
20 MR. GRADOMSKI.
This was September 1977 or July 1977 21 when we started it up.
22 MR. MYERS:
It was the original operational system.
23 MR. GRADOMSKI From roughly startup until May of 1097 wo es 3 e n.m.t te w..s r : a m r*
- +.-'se p r e
') ;.- ;
.Dr i'M.
25 system until it became apparent to us in roughly late 1981,
25 1
early 1982 that some serious work and evaluation had to be 2
done on this in order to improve the reliability of..this 3
system in general 4
in May of 1982 the Woodward Governor Company began 5
conducting an evaluation of the Woodward governors supplied by 6
the Terry Turbine Company and modified by the Terry Turbine 7
Company for Davis Besse.
~ '
e in approximately September of 1983, we had completed 9
a major modification to eliminate speed setting problems 10 associated with, adaptation of the electric motor drive to the 11 manua! 3,eno control knob.
At that point in time thero was a sn
!k,
t 12 1983 refueling outage, and we made changes to both tho 13 installed governors and the spare governor that we had, so all 14 three of them were now modified as a result of the exhaustive 15 testing program that both Woodward and Toledo Edison had 16 conducted in order to solve a lot of the recurring mechanical e
17 problems that we were seeing.
18 in October of 1983, we began a program for the 19 qualification and installation of a replacement governor.
The 20 fixes that we had instituted in September of 1963 were at best 21 considered short term.
We know that at this point in time we 22 couldn't continue to operate with the system as it was.
We 23 felt confident that the changes we had made had drastically 24 improved the reliability of the governors and that had been 25 proved out in testing, that the recurring problems that we had
26 1
seen had, in fact, been solved.
2 Since that time, I don't recall any of the recurring 3
problems with either jamming on the high and low speed stops 4
nor the problems of the slip clutch recurring since that point 5
in time.
6 MR. MYERS:
I think en.esponse to one of the early 7
bulletins and orders, we went back through the failures and
~
B described those that had happened over the period of time.
9 Many of those were associated with valve operation as well as 10 governor speed settings.
This is the main thrust, that the 11 failures had been in that area.
We didn't mention that.
I y
12 just wanted to bring that in.
18 So go ahead.
14 MR. GRADOMSKl:
All right.
15 Again, in October 1983, we began a program for 16 qualification and installation to improve the reliability of 17 the aux feedwater pump speed control.
As a result of that, in nu Novembwr 1934 during our last refueling outage, we leis t a. l uJ a 19 Woodward model pGG governor on auxiliary feedwater pump No.
2, 20 and we plan to install at the 1986 refueling outage, depending l
21 upon the continued evaluation of the performance of the new i
22 modei pOG governor, we intend to install a new model pOG M
e;~.seme-er t b ra No 1 v.<liIery f s c r8.*>1 t or r:.mc 24 MR. PETERS:
I'm Bob poters.
I'm going to discuss 25 the procedure generation and modification that we went through 1
27 w
1 to support the use of the electric-driven startup feed pump.
2 prior to our restart from the TMl accident approximately May 3
1979, we generated a procedure that allowed the operators to 4
utilize the electric-driven startup feed pump to feed one 5
steam generator at a time.
Although we are not sure of the 6
exact dates, early on we provided the capability of powering 7
the electric-driven startup feed pump from one emergency 8
diesil generator, and subsequently we modified the plant to 9
allow the startup feed pump to be powered from either 10 emergency diesel generator.
11 As Mr. Jaim mentioned earlier, in September of 1980, 12 we modified the procedure again to reflect the plant 13 modification that allowed the reset of the SFRCS trip on the 14 main feedwater block valves from the main control room.
This 15 allowed the control room operator to regain control of the 16 main feedwater block valve to utillze the startup feed pump 17 feed the selected steam generator.
18 Thor, in July of 1981, we again modified the 19 procedure to reflect the modification to the plant to allow 20 use of the startup feedwater valve I'm sorry. Regain 21 control of the startup feedwater valve from the control 22 room. That, again, would be to reset the steam rupture control 23 system trip from the control room.
24 And then finally in January of this year, January 25 1985, this procedure was incorporated into our abnormal
28 1
transient operating guidelines, the symptom-based operating deve' loped 2
procedures that we have utilized that were as a 3
result of the TMI accident.
And that's it.
4 MR. MYERS:
That brings you up to current changes.
5 There are several additional changes that are being ple.nned 6
that are down in the areas of lower contributors but are still 7
in our plans in addition to the mode of the startup feed pump.
8 Sushut, would you like to cover those?
Mk.DENTON:
9 I would like for you to be sure and 10 cover the status of the latest amendment.
I guess it was
~
11 Amendment SS that requires installation of a new startup
/
12 feedwater pump prfor to starting cycle No.
6.
13 MR. JAIN:
As a result of the pRA study that we i
14 submitted in 1981, we had identified there that or.e of the 15 most dominant contributdra to the aux feedwater en eliability 16 was the failure of the motor-operated valves.
To that end we e
17 designed and engineered several changes so as to reduce the 18 number of valves that have to open on demand for the aux 19 feedwater system.
20 We are planning to leave the discharge valve on this 21 drawing 3870 and 72, leaving them normally open and l o cl< e d 22 open, so that there wouldn't be any valve in the discharge of 23 the pump which will need to be opened to provide water to the 24 steam generator.
25 The other change that we are planning to make is the c.
\\
l i
29 1
change to the logic of the steam and feedwater rupture control 2
system, as I had mentioned earlier.
On low level or loss of S
feedwater condition, the respective steam generator provides 4
steam to the corresponding aux feed pump turbine through its 5
normal path, MS-106 or 107 in this path.
6 The logic is to be changed such that when steam 7
generator can provide steam to both aux feed compartments 8
in o't'her words, two valves will open, providing two paths, two 9
redundant paths for a given aux feed pump turbine, so if one to path or one valve falls to open, the other path can still 11 provide steam to the aux feed pump turbine.
12 So essentially, eacn aux feed pump turbine has got 18 two paths of steam for it to be run.
14 We are proceeding on designing changes to improve 15 the operation of the steam inlet valves.
The steam inlet 16 valves have several interlocks.
One interlock is the one that a
17 closes these valves if you had a break in the steam inlet l i n' 18 to the turbine itself We hava proposed a tech spec change to 19 the NRC to delete that interlock so that the closure of the
)
20 valve is eliminated and thereby eliminating a potential 21 failure by that valve.
22 These valves are also interlocked with the decay 23 heat drop line valves, and the interlock has been there 1
l l
24 because once you go to the decay heat mode, you trip all your l
l 25 reactor coolant pumps, and tripping all four reactor coolant
e 30 1
pumps starts the auxiliary feedwater system at Davis Besse. so 2
the interlock was provided to prevent inadvertent start when 3
you are going to the decay heat mode.
4 We are proceeding to delete that interlock and 5
taking manual action to de-energize the valve, which we 6
already have been doing for the last six or seven years anyway 7
to remove power from those valves so they don't come on open 8
to start thgg aux feed pump turbine.
9 These two modifications essentially reduce all the 10 control systems or the interlock failures that could og
, x,11 potentially contribute for these valves to fall to come,up.
12 As part of the control room design review, we have 13 Identified several changes to the steam generator system. as 14 well as the SFRCS.
We are going to be providing a redundant 15 steam geneerator level and pressure Indication in the control
\\
16 room so the operator has better knowledge of the steam 17 generator status as far as what the level is and what the 18 pressure in as far as its umability f or sua feedwater s y n t en..
19 We are also going to be relocating some of the SFRCS 20 manual trip switches to enhance human engineering.
21 The normal suction supply from the condensate 22 storage tank has a valve that is normally open to provide M
Sur: t f or
+e the aux f e+*f pump turb'na, i n. t i t.<
- g?n, c ? c. _ c.!
on r
24 a low suction pressure to transfer to service water, which is 25 a seismic suction for the pump.
. s I
31 1
One potential failure mechanism would be that the 2
valve in the suction from the CST could spuriously.go closed, 3
thereby robbing the aux feed pump of water for a few seconds 4
before it is called upon to actuate, and then it transfers to 5
We are proposing to delete or remove power 6
from that valve so that particular failure mechanism is again 7
eliminated.
~
8 And finally, we are putting in a new startup feed g
pump which is of a higher capacity, which has a capability of 10 feeding water into the steam generator, both in the main 11 feedwater nozzles as well as the aux feedwater nozzles.
We 12 were planning to implement that in the next refueling outage.
13 MR. DENTON:
What would the capacity of that pump 14 be?
15 MR. JAIN:
As I recall, 600 gpm, but we could l
16 provide the exact number, e
(
17 MR. DENTON:
So that would be an estimate of the I
18 equivalent capacity l
19 MR. JAIN:
It is equivalent to 100 percent capacity 20 decay heat removal aux feedwater pump.
21 MR. DENTON:
Have you had any studies done of 22 reliability of this system since that one that you mentioned 28 done by EDS?
24 MR. JAIN:
What we have is we have internalized most 25 of the risk assessment work. We have fault trees and other
e s
j 32 l
1 models made up for the system, but we are unable to t,upport at 2
the present time any numbers in order to compare whgt the 3
numbers were when we submitted the EDS PRA study.
4 MR. DENTDN:
Could you summarize what the EDS study 5
stated?
6 MR. JAIN-The premise of the EDS study was to 7
evaluate what we should be doing as far as spending our money 8
on installing either a third aux feed pump or otherwise 9
improving the existing aux feedwater system to include the 10 reliability, and the analysis-based configuration in that 11 report was the one that was aimed at addressing each of the 12 most dominant contributors to the unreliability and making 13 changes to the system to eliminate those dominant 14 contributors.
15 There was a third change analyzed in the report, 16 which analyzed a third aux feed pump, if you will, and the e
17 analysis showed that you have a bigger improvement in 18 reliability for the analysis-based confinorst c,
for a lesser 19 expenditure, and you had lesser improvement in reliability for 20 the third aux feed pump with a much greater expenditure.
21 MR. MYERS:
I think the third the moving of the 22 startup feed pump was initiated and actually har several other 28 functions in addition to providing 100 percen+ auxillary 24 feedwater backup to our concept, it eliminates some of our 25 operational concerns about limited startup flow, it gives us
O e
33 1
improved operability in the normal function of the s.y s t em, and
\\
l 2
also helps separate some of our fire hazards as under the 3
Appendix R activities for control circuits and shutdown panels 4
and that sort of thing.
5 So there are quite a few issues that moving the 6
startup feed pump into a different area and increasing its 7
size actually addresses.
8 MR. DENTON:
If your configuration on the auxiliary 9
feed pump system is different than other B&W plants, as I 10 recall, do you happen to remember what the decision process 11 was leading to that decision?
(~ s) 12 MR. MYERS:
Maybe there is some here that can 13 help. In reconstructing and that's as good as I can do, 14 trying to put it together I think if we remember 15 timeframe-wise, Davis Besse was late in the licensing period 16 to where safety grade auxillary feedwater systems were being 17 discussed as requirements, and Davis Besse perhaps was the 10 first to actwally entor d e '. i g n a required ssfoty geade cy tem, 19 safety grade initial operation and control of the B&W units, 20 the 177 B&W units.
21 At that time the
--~l'm not sure exactly how this 22 particular configuration came out, but it was driven a lot on R3
- his feidwater ruptiere control system, 9nd the dedic tad 24 discharge you notice that is fairly unique pump to 25 generator rather than two pumps to header arrangement.
One
1 e r
- i l
34 i
1 pump
.o one generator, and if you need cross-connectors, you 2
dedicate cross-connect, so that original thought pr.ocess in 1
3 the instrumentation and in the original system seems to have 4
been generated in those days of converting from a normal 5
control grade for older plant's aux feedwater system into a 6
fully safety grade aux feedwater seismic qualified 7
control-started system.
e MR. DENTON:
la that normally under the jurisdiction b
9 of the vendor or the architect angineer?
10 MR. MYERS:
Architect engineer. The basic 11 requirements for steam generator cooling are provided by the
(*
. a e
12 vendor s however, the particular design is architect engineer 13 specific, and safety-grade systems being now, probably was 14 fairly unique, I would imagine, during that timeframe.
I'm 15 not sure if maybe someone here can help illuminate, who was in 16 a position at the time who could help.
This was the '74 e
17 through '76 timeframe.
18 MR. DENTON:
Well, I had heard at least I had a 19 memory, a recall that you had a study done, a Bechtel study 20 done, but you are saying you don't think Bechtel was ever 21 involved.
22 MR. MYERS:
In the original design?
1 28 MR. DENTON:
I'm not sure.
I thought there was 1
24 Bechtel involvement.
25 MR. MYERS:
Absolutely.
I'm sure there was Bechtel
\\
s
]
1 35 1
in the original design of the system, and Consolidated 2
Controls was our vendor for the steam and feeddater rupture 8
control system. So in the original design, that definitely was 4
the case.
5 MR. JAIN:
The effort that you may be thinkeng about 6
would be maybe a study that we had Bechtel do trying to define 7
different options for a third aux feed pump, and they did a O
cos t -ana l ys i s as to which alternative was going to cost how 9
much, and that's the one we were quoting when we met here four 10 or five years ago with the NRC would cost $6 million or $7 11 million.
That's the only involvement that I can think of in 12 those timeframes.
18 MR. DENTON:
That's probably the one I remember.
14 MR. ROWSOME:
I can clarify the record a little 15 bit.
There was a system reliability analysis done at Bechtel 16 by me.
I was at Bechtel at the time, e
17
[ Laughter.3 18 19 20 21 22 28 i
24 1
25
i
. s I
36 1
-- in the 1974-75 period.
It was not at the request of 2
Toledo Edison.
It was submitted to the project design team 3
snd I presume passed on to you all But it was not at your 4
request.
5 it did not result, to my know l edge, in any design 6
changes.
7 MR. MYERS:
Do you recall in this timeframe of the 8
original design, where you were off project?
A 9
MR. ROWSOME:
I was off project.
It was supported 10 under overhead as an exercise in developing Bechtel's 9
11 capability in using fault tree analysis and system reliability 12 techniques, and it was not requested by the project, 13 MR. MYERS:
So you weren't really disowning the 14 actual design?
15 MR. ROWSOME:
No, it wasn't part of the design 16 effort.
e 17 MR. MYERS:
Good timeframe, though.
18 MR. DENTON:
Well, does that complete your 19 presentation?
20 MR. MYERS:
That completes our presentation.
- But, 21 if there are any questions, again as I said, we are in the 22 process now, still now, trying to actually get dates for 23 chronologies here.
24 Any discussion or questions concerning either the 25 studies or
+
e s
j 37 1
MR. DENTON:
What I wanted was to be sure that we i
2 had as comple*e a record of the studies and the phys' i c a l 3
changes to the plant that we could amass.
4 Are there other studies you have now underway in the 5
system, or is there anything more you are doing with it, or do 6
yc.u see the completion of the effort just described as 7
providing a satisfactory feedwater system absent some further 8
problems?
9 MR. MYERS:
Let me cover the philosophy which 10 wasn't, obviously, in the discussion of the individual items.
11 in our early review of the system it was felt that C'1 12 having the installed system being somewhat unique and somewhat 13 inflexible to change, particularly without having a common 14 header on the discharge like many of the newer designs, would 15 have to allow options of dropping pumps onto the header and 16 that sort of thing, that we work on the major contributors to i
~
17 unreliability of the installed septic rate aux feedwater 1 73
- system, Decause, as we saw t h etu, they were quite e dera t i f i ab t o 19 and quite attackable.
The governor valve systems and 20 elimination of potential for new signals coming into isolate.
21 in addition we would provide procedurally a backup 22 to the condition of loss of aff main feed and auxiliary feed.
P9 and nee"8de that te 'b9 operater, gnd a l so e,on t inosa l f y 8 m p r o*r e
[
24 that for reliability.
So, just bringing an additional power 25 supply capability to it, and then from either side of the r
l l
o 44 1
not pursued the evaluation of that type of addi,tional backup 2
to date.
We felt much more reliable feed pump pressurewese 3
and availability and power supply-wise 4
MR. DENTON:
Well, then, a normal unplanned trip 5
what role does the main feed pump play?
6 MR. MYERS.
In just a normal plant trip without a 7
steam and feedwater rupture control signal?
~
8 MR. DENTON:
Whatever would normally happen, 9
assuming no equipment failures.
10 MR. JAIN:
Steam and feedwater control system, the 11 ICS will take the integrated control system will take the
.J 12 plant to a steam generator low level condition where the main la feed pumps will be used to maintain level at 35 inches in the 14 steam generators and removing decay heat thereby.
15 Once the pressure in the steam lines has gone down 16 significantly, you can either go on the auxiliary boiler to 17 run your main feed pump turbines or you can go to the decay i
10 h%t mede.
Tr. * ' ' s the f ong <... - i down p r o c e... w3th:ul 16.c 19 MR. DENTON:
What's the auxiliary boiler?
20 MR. JAIN:
It's an oil-fired boiler that we use for 21 station heating, for example, when the plant is not running.
l 22 or we also use it for initial warming up of the plant's r l' i ! ! +1 eteam fer ' 'n e t i.or b ? n 7 It !s
.!u t r* r w -.
).
! no,, '..
'9t 24 operation.
25 MR. DENTON:
If you had a loss of all feedwater, is
< r 45 1
that the system that you try to bring into operation?
2 MR. JAIN:
We normally don't 3
MR. MYERS:
I believe that the start-up time on the 4
auxiliary boiler is quite significant and I'm not sure if we 5
actually consider it able to be brought on 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> a day.
6 It's able to be started, but I believe it taken a while to get 7
that system going, and it would actually be -- l don't want to 8
say a detraction from the main recovery mode, but it would be 9
much less reliable and swift to recover feedwater than the 9
10 start-up pumps or recovering the aux turbines.
11 MR. DENTON:
Do you happen to remember the
\\.
12 conclusion of the reliability of the aux feed system that EDS 13 reached?
14 MR. JAIN:
As to what we should be doing?
~
15 MR. DENTON:
No, how reliable, say, the system was 16 likely to be, either in this state or upgraded.
17 MR. JAIN:
As far as the numbers?
I 10 MR DENTOtt
'/e s 19 MR. JAIN:
I don't remember it, but the report has 20 the numbers in there.
I don *t remember it right now.
21 MR. MYERS:
That was submitted.
22 MR DENTON-I'll look at that later.
Let's son if R?
there are any other questions T e^
- ba staff or other 24 parties.
And if not, I propose we take a caucus for the staff 25 and mull over what we have heard and see if we have some more
e s..
46 1
questions.
But before we break, we'll see if a,nyone has any 2
further questions.
3 is there any other information you want to tell us 4
about the system that you've got or actually plan to take?
f 5
MR. WESSMAN:
One quick question.
Maybe it's 6
outside the scope of what you-all wanted, but cara you 7
summarize some of the changes that are going on with the main 8
feed system?A 9
1 understand you have had some difficulty with 10 governors on main fewd pump turbines.
d 11 MR. MYERS:
We can probably give you SO seconds.
We 12 converted the complete main feedwater pump control system in 13 this last outage.
It's a General Electric system 14 modification.
We have had some problems, as a matter of fact, 15 prior to last Sunday's event.
We even did special 16 instrumentation of this system specifically to try to nail
~
17 down if any control failures occurred, and that's hopefully 18 when we get intu with the fact-finding team, ento the 19 machinery, that instrumentration is going to give us quite a 20 bit of input.
But it's a complete change-out.
Our main 21 feedwater pump control guru is right now working with our i
l 22 team, the fact-finding team, back there.
We can respored to l
R e.
jw
..e
- a *he f t 4."
s >.
v'4 h.a so ig'-
6 ( i.- ' 4'..*, t..: e n s w
e 24 can probably get specific answers today for you on that.
But 25 the staff here that information is not available.
t I
a.,.
l t
47 1
MR. DENTON:
Okay.
Why don't we break untsi say a 2,
quarter after 12:00.
That wiIi give us time to see if there 3
is other information or questions we might have on what you 4
have told us.
5
[ Recess.3 6
4 7
O M
j e
9 i
10 11 S
12 13 14 15 e
16 37 is 19 20 21 22 i
i i
24 l
25
< n
- 48 1
MR. DENTON:
Let's resume the meeting.
We don't 2
have any additional questions to ask.
What you t o,l d us this 3
morning has been very useful to help us reconstruct events and 4
actions that you have taken.
5 What we weil do is coordinate with our team at the 6
site to be sure that any follow-up questions that we ask of 7
you are coordinated with them.
I think some of the types of ques [lons e
that come naturally to mind are whether the EDS A
g study that was done for you sometime ago used what I would 10 call generic failure rates as opposed to plant-specific 9
11 failure rates, and whether if you were using plant-specific C
12 performance data over the past couple of years or you had 13 gotten a different result, and the question of whether it is 14 safe to resume operation with the type of failure rates that 15 we are experiencing is t,h e type of question that leaps to 16
- mind, e
17 But i don't have any specific requests to make for 18 additional information of you at this time What I will do is 19 coordinate with Mr. Rossi and others to be sure that we're not 20 having two different arms of the NRC asking for information, 21 so l'Il just mention these items as one that I think followed 22 from the discussion we hads namely, you had a reliability 23 study done.
it looks like the system was getting upgraded 4 24 you thought it was being upgraded, you worked on the valves 25 and control systems and those kind of things.
And then we had
< n.
r 49 4
1 both systems that didn't work, and that naturaljy leads to the 2
question of why the difference between what you had expected 8
from a previous pRA analysis and real life, and that's a 4
scenario that we will be looking into back here and with 5
Mr. Rossi.
6 MR. MYERS:
Har o l d, I would like to make one comment 7
t h a,',' think you can feed back to your staff now, and if the S
report doesn't stress it we should note that in the EDS PR4 9
that was done, plant-specific failure rates were used, as a 10 matter of fact, and because of that we did see specific 11 differences between generic and industry, and that started 12 some of our follow-on work with Limotorque operators and some 13 other consultants to actually -- in that comparison of our own 14 failure rates individually to industry's.
And we took some 15 action there.
16 So I believe that maybe doesn't stand out in the
~
17 report but it should be in there somewhere, the specifics.
15 Mn. GENTOI4 Do you ha,e a feel for wnethe,- yvur 19 plant-specific experience up to a few weeks ago would give the 20 same estimate that was used in the EDS reports whethur 21 patterns of changes in the reliability of 22 MR. MYERS:
I think the patterns that changed we saw f
'O
.w o, a '.wp r.w w....,1 n L!ke l 1,. 6 d, t h e int....
e, i t.. " * >.' r..< es. J 24 quite well defined and, in our corrective actions, were 25 noticeably turned, both in valve operation and in governor
l e n.
)
50 i
i performance, which were the main contributors to system l
2 unavailability.
8 in those particular areas I think we can say that in 4
going back and revisiting the data now, I'm sure we will see 5
past performance had reflected that.
Of course, the exact 6
cause of the two turbines being unavailable is a major focus 7
of our investigation, and with the factfinding team activity
~
8 at the site, both of us are anxious to get in there and 9
validate our findings.
And I think hopefully, that will shed 10 a lot of light back on the analysis to reflect whether it was 11 something that should have been foremeen or something outside 12 the procedures of PRA.
18 MR. DENTON:
That is something we'll have to await 14 for respective efforts to be completed.
15 Hll right.
If there are no more questions or i
16 comments, I appreciate your coming in on such short notice.
17 Thank you.
18 CUher wugon, at )E.40 p.m.,
thu meeting was adjvurneu.)
19 20 21 22
,, a 24 25
c.,
a e
e 1
CERTIFICATE OF OFFICluL REPORTER 2
3 o
4 5
This is to certify that the attached proceedings 6
before the Un6ted States Nuclear Regulatory Commissime in the 7
matter of.
9 Name of proceeding
- Meeting Between the NRC Staff and Toledo-2 Edison Company Concerning AFW Systems 10 11 Docket No.
l) 12 place:
Bethesda, Maryland is cate:
Monday, June 17, 1985 14 15 were held as herein appears and that this is the original 1
16 transcript thereof for the file of toe United States Nuclear 17 Regulatory Commission.
19
(
9""
39
[ &ws
__ ec/
(Typed Name of Reporter)
And Riley 20 21 22 23 Ann Riley & Associates, Ltd.
24 25
i
- e 4,
l E
5 it l
.[
~
s' a
t I "4 8"
I CJ.
j
=
e[= =d p,[1 7
i
" 'I di G'!
! l im en ilhi Iw
_o j
i
,, = =
0 i
@ ' !ff' Ih!
1 i, !
l_.tNm i
se i
i Lp ht;:j
=
r gg g g
s m
i g
Ipk 5
3, p
.NL% g!
a
$ Mi I
.A!
i r
i 1
M,e g
.d 4
4 Ti lif! idipf%
2 m
in i?.
<# p i o
(.
f,=(
sg tg, H5 m
g11y o
'g '! R 1
=
r' ri lii '- M t "E e
F.
l
_ i\\y' y
Yf p
es n I
>i i
11 e
li ai g
g g%g n
g i
A.2%
e u 'i.
i-
-h h\\
e e
,,u z
e
.e i
38 1
1 plant, as we saw the opportunity to actually move the startup 2
feed pump to not only increase the size, but also increase the 3
capability to deliver directly into the feedwater nozzles, 4
which is highly desirable on the B&W generic design.
We 5
certainly did not let that opportunity go by at all 6
The intentier is to continue to improve the 7
as-installed system while we are providing an upgrade also in 8
the startup date of system supports.
9 MR. DENTON:
Will this new system you are describing 10 actually replace the previous system?
11 Or, will you keep it in addition to the system you
.)
12 are talking about, talking about the upgraded pump and motor?
13 MR. MYERS:
The upgraded pump and motor will be 14 normally use to replace the startup feed pump.
It is in place 15 now.
16 MR. DENTON:
it will go in the same place in the e
17 plant?
18 Mr MYE9e Ch, no Plc We Tre me.ing t te an 19 entirely different location in the plant The size and piping 20 limitations are one of the reasons for moving it The 21 piping energy line break considerations, and also fire 22 considerations for moving it out of fire areas to give us 23 additional support so it will be moved into a different 24 portion of the plant, the turbine area where there is more i
25 room, a capability for a different pump activity and cross
= w.
',a 39 1
connects for water supply and discharge lines.
2 MR. EISENHUT:
When is Cycle 6 scheduled?
3 MR. MYERS:
It was scheduled for spring of 1986.
4 The long lead time components, as I believe Dick Crouse had 5
mentioned to you, they are already on order.
6 Again, we are trying tc get the details of wher, they 7
were put on order to provide to you.
But that design process
~~
8 la well underway and procurement is underway for that.
9 MR. EISENHUT:
How much time is programmed on your 10 schedule in Cycle 6 to actually do the installation, do you 11 know?
Is It a three month job, six month job?
I 12 MR. MYERS:
l*ll have to find that out.
Actually i 13 can give you the philosophy also of the location.
It is 14 relatively free of activity currently.
I don't want to say it 15 is a dead spot in the plant where the square footage is doing 16 no t h i ra g.
But, a lot of the work was planned to do pre-outtage 17 so that we would not impact in working around the area with ou as u vpor 4 t ; eeg, but we tv u ! c bo 10 more impcrtant pumpa wh.!c 19 able to do a lot of that construction independent of our 20 operation.
It would not hamper that.
21 I can certainly find out what the construction 22 module says new for actual outago time That is a dedicated
'3 plo* In o' e r pisnning 24 MR. DENTON:
Any other questions?
25 MR. THADAN1 Do you have procedures today,
, w.
+
. e 40 1
recognizing the low capacity of the start-up feed pump, in the 2
event of extended loss of main and aux feed?
Can the 3
operators perform combinations of actions, the start-up feed 4
pump as well as the other mechanisms for removing decay heat?
5 Do you have procedures in place 6
First of all, is it feasible' 7
Second', if it is feasible, do you have procedures
.~
8 today in place to be able to remove decay heat by a multiple A
9 source of actions?
The start-up feed pumps, maybe open up the 10 PORU, try to get the pressure down, that kind of stuff?
11 MR. MYERS:
Yes.
The procedure that Bob peters 12 mentioned earlier, the one that we originally developed was 13 called an abnormal procedure, total loss of main and auxiliary 14 feedwater, and specifically did give instructions to utilize 15 and make up the high head capability of the makeup system 16 utilizing the pilot operated relief valve and the start-up i
17 feed pump to provide full decay heat capability, and that's 18 our pra:edural.ctior, and :a, beers in place since that 19 procedure was initially approved.
20 Since then, this last cycle, we converted to the 21 ATOG, the symptom-oriented guidelines and incorporated that 22 whole thing into our current proDram, current procedesres 23 program and that, along with, of e.nvese, the recoverv of 24 auxiliary feedwater, is stressed in the procedure and actually 25 was the procedure we followed, the process we followed, last
41 1
summer and it is a result within, I believe, two minutes of 2
when we lined up the feed pump and the auxiliary pumps were 8
recovered also.
4 So, yes, that is proceduralized and is an integral 5
part of our ATOG program 6
MR. THADANI in ref er er.co to some analysis that was 7
done by you and supported this procedure, I assume?
.~
8 MR. MYERS:
We have done analysis to support it.
I 9
don't know what the ATOG references are.
We can find those 10 out.
They are probably in Volume 1 of the book, of the ATOG 11 guidelines for Davis-Besse.
We can dig out exactly what
~
12 the actual emergency procedure in the plant probably does not 18 have that as "this references such and such a*
14 MR. THADAN1*
No, I understand that, but the 15 procedures were developed on the basis of whatever analyses 16 were done.
17 CMr. Myers nodding.3 18 Mii. htJtJH3ME Twv t u r. l... i c. a t ques t e ve,.
ori what 19 Mr. Jain told us.
I didn't quite follow what you said about 20 dropping the auto start on the trip of the reactor coolant 21 pumps.
How is that being implemented?
22 MR. JAIN These valves get an open signal f r on, the 2'
s t 9 mir 1r ef #a?dwsta* r on her e *:nn t r ? ' sy9 ct 6
3.. -
p-h,..,, '.. : t l
24 all four reactor coolant pamps.
I l
25 Now we are proposing to and we have been doing this
(
~.. _ _ -,.
I 1
- ~.
1 l
s.. 1 42 1
ever since we started up, is when we go into the decay heat 2
mode, we take the power off of the steam inlet valve so the 8
interlock really never has been used, per se.
4 So we are deleting that interlock so that you are in 5
mode 1 operation, nothing can go wrong in the interlock to 6
cause the valve to se closed, you know, a spurious failure 7
MR. ROWSOME:
Instead of manually depowering the 8
valves, you gre going to change the logic so that they are 9
automatically cut out of the logic when you are in mode 57 10 MR. JAIN:
We will be depowering it still, but we 11 don't need the interlock any more.
I guess we never needed C:
4 12 the interlock, but it was put in there for that purpose, so 13 15at you don't start the aux feed pump turbine.
14 Maybe I am not very clear.
15 MR. ROW 50ME:
I don't quite follow, but I am 16 reassured.
We will find out in due time.
17 You are saying that you would lock open the disul.argo va !. w. 3370 arid 30727 19 MR. JAIN:
Yes.
20 MR. ROWSOME:
I think that gives you trouble with 21 isolating a main steamline break, does it not?
Do you depend 22 on manual action to isolate the affected steam generator' MR
. 'f4 f N -
This is t no+rt operatad
'm 'n'+
24 locking it open in the field, the hand wheel and the local 25 push button station doesn't disable the control from steam and
i b me *
- o 'n g 43 1
feedwater rupture control system.
2 MR. ROWSOME:
Oh, it's just the local control and 3
hand wheel All right 4
MR. JAIN:
So most of your SFRCS actuations are on 5
the low level.
6 MR ROWSOME:
So it will still be normally closed, 7
it will just be that the 8
MR. MYERS:
An operator in the plant could not 9
misposition the valve because it's locked.
He would have to 10 have there's a rigorous control for him to change the 11 position of that valve.
12 MR. ROWSOME:
But it remains normally closed?
13 MR. JAIN:
Open.
We are proposing it t o lie open.
14 MR. EISENHUT:
Open, and the logic would close it.
~
15 MR. ROWCOME:
- And you are counting on check valves 16 to constitute the pressure boundary?
17 MR. JAIN:
Right.
10 MC DCrTON-Let me ask 4L. et the main feod pun.p4 19 Once the main steam line isolation valve closes, are you able 20 to bring them back to service again following a reactor trep?
21 What prohibits those from being used as " auxiliary" feedwatar?
22 MR. MYERG The steam supply essentially would not b.* t c.* 2 6 t n % f a t*
s' e r ' 'hnm boraus
- \\..'s e sees Iar2s s t a a.s c P! J:;
24 1 understand, Ilke in Combustion Engineering plants, they talk 25 about condensate booster pumps and things like that, We have