ML19263E646
| ML19263E646 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 05/18/1979 |
| From: | NRC OFFICE OF THE EXECUTIVE LEGAL DIRECTOR (OELD) |
| To: | |
| References | |
| NUDOCS 7906200560 | |
| Download: ML19263E646 (89) | |
Text
3-Records facilities Branch (jd 016 Phil NUCLEAR REGULATORY COMMISSION
/
IN THE MATTER OF:
PUBLIC MEETING BRIEFING SY STAFF ON OCONEE ORDER Place.
Washington, D. C.
Date -
Friday, 18 May 1979 Pages 1 - 88 2287 156 Te4ecnone:
(202) 347 37CC ACE -FEDERAL REPORTERS, LNC.
Offic:alReponers i1.1 North C::itol Street Wcshingten. C.C. 20001 NATICNWIDE COVERAGE DAILY 79062005@ j
or 1
CR4893 DISCLAIMER This is an unofficial transcript of a meeting of the United States Nuclear Regulatory Co=nission held on Friday, 18 May 1979 in the Commissions's offices at 1717 H Street, N. W., Washington, D. C.
The meeting was open to public attendance and observation.
This transcript has not been reviewed, corrected, or edited, and it may contain inaccuracies.
The transcript is intended solely for general informational purposes.
As provided by 10 CFR 9.103, it is not part of the formal or informal record of decision of the matters discussed.
Expressions of opinion in this transcript do not necessarily reflect final determinations or beliefs.
No pleading or other paper may be filed with the Cormission in any proceeding as the result of or addressed to any statement or argument contained herein, except as the Commission may authori=e.
2287 157
u 2
CR4893 j.
UNITED STATES OF AMERICA
.v?' TIER /=m l
2l NUCLEAR REGULATORY COMMISSION I
31 4
PUBLIC MIETING 5:
6!
i 7
BRIEFING BY STAFF ON OCONEE ORDER j
8 i
~_ _
i 9
i 10 Room 1130 i
1717 H Street, N.W.
11 Washington, D.C.
l i
12 Friday, 18 May 1979 l
l 13
+
Hearing in the above-entitled matter was convened, 14 i
pursuant to notice, at 9:50 a.m.,
JOSEPH M.
HENDRIE, Chairman, 15 i
presiding.
16 i l PRESENT:
17 l JOSEPH M.
HENDRIE, Chairman 18 VICTOR GILINSKY, Commissioner 19 l l
RICHARD KENNEDY, Commissioner 20 j F.ETER 3RADFORD, Commissioner 21 '
I j
JOHN AHEARNE, Commissioner 22 i l
Messrs. Denton, Ross, Cunningham, Gossick, Chilk, 23 !
Case, Cottle, Israel, Novak, Wilson and Thatcher.
- rei A.=nen, inc. >
2-s g
p 25 '
L 3
i
s I
3 I
I
-7
~
D e
Il s _R O _C ?z z _D _I 'i _G _S L
t i.
9' CEAIRMAN HENDRII:
Good morning.
-l 3,
The Commission meets this morning for a briefing by i
3l the Staf f on the Oconee order.
Lee, please go ahead.
l t
I 5
MR.GOSSICK: Thank you, Mr. Chairman.
i i
6' Mr. Denton will introduce the speakers.
I 7l MR. DENTON:
As you recall, the Commission issued l
4 gi orders to the Duke Power Company regarding certain actions to I
- 9) improve the reliability of the aux feed systems and its response i
i 10 to feedwater transients.
11 In response to that order two of their units are l
b 12 ' presently shut down. The order would require that the third unit' i
13, be shut down tomorrow unless they have completed certain actions!
i l
14 Today I want to describe for the Commission the actions 15 that the Duke Power Company has taken, our review of these acticns
! and our review of some additional matters that we have considered 16 i
17 ; in that connection.
As a result of our review, we have concluded 18 that they have satisf actorily met the conditions of the order 19 and I would propose, with your concurrence, to make that finding 20 later today.
21 ;
Denny Ross is with me to describe in some detail, i
22 " the items in the order and our review of those items,
i 23 Denny?
24,
MR. ROSS:
3efore we have the first slide, I would co.
rsi Aecorters, Inc. ;
25 like to identify the principal workers hat have worked on the 2287 159 1
s 4
I i
mm 1
FSER for the last two weeks.
Under the general cognizance of 2 l Tem Novak, there are two project teams. The team that worked i
3l on the Oconee order, Sandy Israel was the team leader, comprised, i
4l of Chu Liang, Dale Thatcher, Bruce Wilson and Mike Wilber for I
I 5
the procedures part.
l 6'
These people are present if we need any detailed i
7 discussion.
j e
And, also Dr. Zoltan Ros:toczy, branch chief of the i
9 Analysis Branch -- and many members of the Analysis Branch 10 contributed to the review of the genetic B&W methodology.
11 So, if we get into details, these people are available.
l t
i l
12 CEAIRMAN HENDRIE:
Good.
i 13 MR. ROSS:
The letter -- we have a draft letter i
14 that would be signed by your order and sent to Duke Power.
i i
15 In the letter it cites the provision of the order, and states l
16 l that Duke Power has satisfied.
Again, this is draft.
l 17 It also mentions the need to do a three-unit stabilityi 18 test which we will cover later on; the need to provide tech 19 spec changes on a
fairly short time schedule and requires i
l 20 ;
Duke Power to give us a schedule for the long-term items of 1
21,
the order.
22 l We also enclosed in a draft form, a Federal Recister 23 ;
notice, which I don't think we need to get into.
24.
The structure of the SER is aligned the same way as 3.s.cu m a.oon m,inc.
25 the order, so there is Parts A through E.
k 2287 160 n
'.l I
v
.s 5
l 1
Part A has to do with the mechanical changes to 1
- 2) the emergency feedwater system, and if you have your draft i
1 l
3; SER in front of you, we will be starting bout page 3.
4' If you look at the first slide, please, it has the l
5l
- 1ve ingredients of the crder.
1 i
6l (Slide)
I I.
7 And, if we proceed to the second slide we will be talkhr i
8l about the changes in the emergency feedwater system.
l.
t 9
(Slide) t 10 MR. DENTON:
I would like to mention that Denny is 11 in charge of the review of all of the responses to all B&W j
i l
12 orders and all responses to the bulletins in NRR, and is i
13 chairman of this ad hoc group.
I i
14 MR. ROSS:
I will go back and forth between the SER i
i
' 15 '
and the slide. But the principal features of the changes in 16 the emergency feedwater system are two.
17 ;
First, the system was reconfigured such that when 18 any one of the Oconee units requires emergency feedwater, all i
19 j three pumps start.
Previously, only the emergency feedwater i
20 l pump for the unit that needed it started automatically.
Now 21,
they all three start.
22 l.
Mechanically the pumps were already piped into a i
I 23 common header. Now the valves between the separate units on i
24 this common header have been opened and locked open.
This
- . s.e.,.i m.oo m,s.inc.
25 gives a measure of redundancy when one unit or more than one g
0 2287 161
,t
'l
6 r
t i
i i
i 1l unit needs feedwater.
nm i
2l1 The second principal mechanical change has to do 3l' with the emission valves for feedwater. Both steam generators l
i 4l of the unit that needs the feedwater have been reconfigured i
'l i
l 5
such that the control valve opens automatically when that unit l
6 senses a loss of main feedwater.
l l
7I The operator then subsequently controls this valve l
8:
to achieve the desired steam generator level, depending on I
I i
9 wheth er the reactor filter pumps are running or not.
t 10 COMMISSIONER BRADFORD: Why would the system not have Il l
been built like this to start with?
i 12 l Is there a down side to d *'
I 13 i
MR. ROSS: Well, the operator now has to control it l
I4 manually.
i 15 COMMISSIONER BRADFORD: I understand.
When the plants' I
t 16 were first being built, why not have done it this way?
i I7 MR. ROSS:
I think that considerable reliance was i
t la ;
given to the integrated control system which served this I
"l same function automatically.
It was presumed that the integrated i
20, control system would achieve this function at a satisfactorv, 21 :
degree of reliance.
22 !
We concluded the opposite a few weeks ago.
i 23 'j l
I think the safety-grade control system, integrated il 24 U control system not only would do this job, but would do it better.
..s.o.,e neoonen.inc.
- S It was the presumed reliabiliev of the IC
' hat 'e to the i
s 2 ff/
Ib 1
l 7
i I
i i
conclusion to design it the way it was, l
i 2
This system now is completely independent of the ICS.
i 3)
As you recall, we did require a detailed failure i
i 4 I modes and effects analysis be done en it.
l l
i 5'
MR. CENTON: This part of the plan is outside the I
a 6!
scope of supply of B&W, so it reflects -- and it is outside l
i 7-the scope of B&W in all the other plants, so it reflects the I
a utilities' preferences, and that's why the auxiliary feedwater 9l systems differ among all the plants.
10 COMMISSIONER GILINSKY: Let's see, Denny.
Are you t
11 saying that the ICS would have made the other pumps available?
j 12 MR. ROSS:
Not shown on this diagram are other I
l 13 valves that control the level in the steam generator if you i
14 have a feedwater transient.
If the main feedwater pumps both I
15 are still operable, then you won't call on this system and level!
16 will be maintained as usual by the integrated control system.
17l But if it fails, as indicated by loss of turbines or feedwater i
1 18 l or loss of pressure, then this system gets called on.
1 l
19 h COMMISSIONER GILINSKY: Okay.
I':I i
20 o MR. ROSS: The plant could well run another 40 years l
21 and never call on this system.
i 22.
Okay.
Within Part A there were five, if you wan:
1 23 l to refer to them as subparts, and these, if you leaf through l
I 24 :; your SIR, carries through pages 3 through S.
6
- e.5eceral Reporters, Inc. 3 25
The first part is to autostart the -, _. _e __.._:s a nc, J
y 0
2287 163 a
t
8 l
l mm 1;
interconnect them.
As I have described on this chart, this has 2,'
been done.
I i
3 The second part of the order has to do with the i
4j stability test. And this was done for two units. The third unit I
3 was down for refueling.
In order to do the three-unit test 6
you have to have all three units steaming.
7' The three-unit test will be done when the third i
l i
8, unit is available -- pardon me, when three units are available.
i 9i Our concern was that if they had a general power outage i
10 ! where nore than one unit was calling on feedwater from a common l l
l 11 !
header with manual ocerator action, then the notential at i
t I
12 '
least for -e dynamic instability existed, and it should be l
i 13 '
tested.
14 It was tested for two units by manually initiating 15 a sudden and complete loss -- shutting of one of these valves.
I i
16 i And then the operators responded both by the other valve on i
l 17 !
that unit, as well as the other two valves on the other unit 18 l under test.
19 !
And the results of the tests were acceptable. We 20 witnessed the test.
21 i When all three unius are available, they will have i
I 22 o to repeat this, and we will witness it again.
t 23 l COMMISSICNER AHEARNE: Could I ask a question?
24 MR. ROSS: Yes, surely.
vJocef al Reoorters. Inc.,
25 N COMMISSIONER AREARNE:
Are you leaving par A?
l 2287 164 3
a
I 9
e
=m 1
MR. ROSS:
No, I have three more parts of part A.
2 !. The third portion of the order on part A had to do with the i
3l local operator backing up the automatic start.
i 4i There is a local operator full time at each of these n
5 feedwater, umps. He has training in procedures, he has i
6' communications to the control room, and he has local meters to i
e 7
observe that the feedwater pump is delivering flow.
8l COMMISSIONER KENNEDY:
One operator or three i
i 9, operators?
10 MR. ROSS:
One per unit. Three operators around the 11 clock.
12 !
COMMISSIONER GILINSKY: Now what is it the coerator l
13 j is making up for?
l 14 MR. ROSS: Well, in our SER we mention the fact that i
15 l the system is not comoletelv immune to single failure. That 1
16 ' there is one relay that starts this pump automatically.
7' I,
17 '
that relay fails then the operator locally will have to start 1
(
la '
the unit.
t l
19 :
And he also serves as a communication device to tell I.
4 20,
the control room operator what the flow rate is.
21,
MR. DENTON: We looked at this area rather carefully 22 '
this week when we were down there. They have at least two, i
23 ' maybe more ways to communicate with the operator from the 24,
control room.
me.eere seconers. inc.
25 I talked to the operators and viewed their I
2287 165
.1
e 10 t
l mm ii procedures.
I think, after looking at it, it is much simpler to 2;
I 3' start the pump from a local station than I had realized.
I i
4; think even I could start the pump, but I wasn't able to get i
5 cleared to do it.
i 6i It is essentially three manual operations.
And 1
7-from discussions with these operators I am finding that they do l
i g
understand their jobs. I am confident that given instruction i
l 9
from the control room, there is a high probability that they I
10 will succeed in starting the pump.
i I
i i
11 '
COMMISSIOMER AHEARNE: It mentions in the writeuo I
t I,
12 lthat a direct communication link was required and is there?
I 13 What kind of direct communication?
I i
14 MR.DENTON: They have two,and maybe another one.
15 The one I personally tested was a telephone from the control 16 rocm to each operator who sits by the pump.
i 17 And then there is a headset, a second system entirely, la ' that communicates to him.
19,
And then I think there are other horns and alarms I
20 ! that alert him.
But, maybe one of the other members of the t
21 Staff can give you more detail.
22 COMMISSIONER AHEARNE: When you say there is a 23 o telephone, you mean the operator has a telephone number that he
!.I 24 j dials to get the guy at the pump?
c.4.ceni neeor.n. ene. l 2287 166 25 j MR. DENTON: 'le s.
0
I I
11 1
And there is a headset that is plugged in, that is mm 2!
somehow right there in the center of the control panel, and i
3 the other end of the headset is right there by the operator, and A
they have some way of flashiag a horn or an alarm to h?/ e him i
5, put it on.
i 6
CCMMISSIONER KENNEDY:
What are the duties of this l
7 j man, specifically, this guy at the pump?
i 8i What does he do?
i 9
MR. DENTON:
He has about two pages of procedures, i
10 and his duty is that if the pump doesn't start, he is called upon II to start the pump.
i j
12 1 COMMISSIONER KENNEDY: Those two pages of duties is i
l 13 in the event the pump is called upon to start and doesn't, what l Id does he do during that 40 years in which the system may never 15 be called upon?
I0 MR. DENTON:
Oh, he was there originally because 3,' ! they had not interconnected the systems and had not made the 18 changes in the A throug.. E.
"l He will remain there until they install the d
I 20 4 redundant electric start pumps and the foundations for these
.)i 4 '
pumps have been poured.
The pumps, I believe, are on site, and i
nn ' progress is being made towards installing, I think two -- two 1
23 ! electric-driven auxiliary feedwater pumps per unit, which will 24 l supplement.
So when they install those pumps, then the
.c..s.c.,.i n.oo,:.ri.inc.
,c i
~; operator would no longer be recuired as a backup for the l
2287 167
s I
i 12 jl mm system.
h 2l CHAIRMAN HENDR!E: My understanding is that the 3l presence of the operator at the auxiliary feedwater pumps is i
3 a backup, and further assurance that they will be available i,
more speedily when needed.
5, 6,
Presumably if the operator wasn't there if you 7f called for auxiliary feedwater, as I understand it, and the l
gl single relay -- or for whatever reason didn't start -- that an 9'
operator would be dispatched there and then start the pump.
10 I guess that, maybe, has happened before.
And this simply saves, r
i 11 i the time that would be required to recogni=e the condition and 12 ;
get the person down there.
I i
13 !
But it is not contemplated as a permanent operating 14 requirement.
15 COMMISSIONER GILINSKY: As I understand what you are I
i
! saying, youarecompensatingforvulnerabilityintheelectricalj 16 I
17 '
circuit.
i l
18 MR. DENTON: Yes.
19 j CHAIRMAN HENDIRE: Until that can be dealt with on a i
20 l long-term basis.
21 1 COMMISSIONER GILINSKY:
Let's see, one of our reports i
22 j says that the time when those additional pu=ps will be in N will be about three, four months.
Is that the schedule now?
,3 'l I
24 MR. ROSS: That's the best schedule we have right new.
= s,anei Aeoonm. inc. l 25 ] We are asking for a betterdefined schedule.
But we saw the e
j 2287 148
t 13 l
mm I
hardware there, so they are moving.
i 2;
We have not sat down and discussed the detailed 3l schedule with them.
4 CHAIRMAN HENDRIE: Let's see, the electrics do what?
i 5
Go in in parallel to --
6:
MR. ROSS: Yes.
l t
7 CHAIRMAN HENDRIE: Let's see, and there were what, I
8 two for each unit?
9 MR. ROSS:
Two, I believe, 50 percent capacity.
But.
10 I don't know how many gpm. I imagine that's about --
i 11 MR. DENTON:
Each of these is 150 percent capacity.
l 12 So two of these could supply all three units.
l 13 And then there will be --
i 14 MR. ROSS:
I think two 50 percent capacity.
15 CEAIRMAN HENDRIE: For each of the units.
i 16 MR. ROSS:
For each of the units.
17 CEAIRMAN HENDRIE: And they will do what?
Prts2= ably 18 meader with each of the present pumps?
19 MR. ROSS:
Yes.
20 I suspect it would be a good idea to isolate the i
21 header again then and make one per tra.'1,
.ut ye haven't gone 22 - into that.
That would give immuc.ir; co,.,sive failures.
i, 23 i CEAIRMAN HENDRIE: Yes.
l 24 l COMMISSIONER GILINSRY:
And you regard rhis as an
- ..;.eerai a econ.ri, inc.
25 acceptable arrangement for three or four months?
2287 169
14 mm 1
MR. DINTON:
The operations are sufficiently simple 2
that he has to do; namely pull a lever and if nothing is amiss 3, and trips, that actually starts the pumps.
3 The other -- there are only two other things he t
5 has to do if it doesn't start that way -- is crank down a valve 6-and reset and open it again.
i 7
So the operating instructions -- the operations are l
8 quite simple from the pump.
And all the operators have started 9
a pump at least once from my discussions with them. They all 10 understand completely what their job is.
11 We had talked earlier about the salaries of some of l
4 12 the people.
I understand these people make S6 an hour.
4 13 COMMISSIONER GILINSKY: What does the man do --
i 14 MR. DENTON:
He spends four hours the2e waiting for t
13 a call from the control room. And then is relieved and performs,
16 other duties in the control room. And they alternate this 17 task among some of the miscellaneous operators.
Is COMMISSIONER AHEARNE:
If I could just make sure 19 I understand that call fromthe control room.
I am still not 20 ;' completely clear.
i 21 i.
The mechanism they use is, someone in the centrol 22 room when they recognize -- if they were to recognize a relay 23 l failed and the feed pu=ps didn't Teme on,they vould then call i
24 l each of those operators at each of chose three stations?
- eJederal Reoorters, Inc. f 2287 170 25 ll MR. ROSS:
Yes.
a d
.i
r e
I 15 i
j l
nm 1
We would want all three purc. started, even though 1
2 it just -- presumably we are only talking about one that didn' t,
i 3'
start.
But whichever unit didn't start, you would have to get l
4 that one started manually.
i 5
COMMISSIONER AHEARNE: So he has three telephone 6:
numbers on the little flag on the telephone, and he dials I
1 7'
tho se. three?
8 MR. DENTON:
I don't know all the details.
I 9j Let's ask if anyone here does.
i i
f 10 MR. COTTLE:
Bill Cottle from I&E.
I i
l 11 The headsets that have been referred to are sound-i 12 !
powered phone headsets.
All the operator has to do in the l
i i
13 control room is push a button which buzzes the guy instantaneously i
14 He picks up the headset, the corresponding headset on the phone l 15 in that control room, and it is direct communication.
16 COMMISSIONER AHEARNE: I see.
17 Thank you.
18 MR. COTTLE:
That is the third line.
19 j One other point is, the operator can manually stop i
20 the pump from the control board. as well in the main control 21 :
room.
It would take two relay failures before he would 22 i actually need the individual at the pump.
23 !
MR. ROSS:
The last cart of Item A that we haven't i
24 I talked about is level control.
a ral At:rfters, Inc. j s.
25 ]
The order says we should assure the role of control 4
2287 171
16 l
mm 1
room operator to maintcin level after the system has come on.
2 We did note the valves opened automatically, and t
3 then later the control room operator adjusts the position of the i
I i
4; valve based on the desired steam generator water level, i
i 5
And then we did verify that the control room l
6l operator is properly trained and has the procedure.
7 And that's the essence of part A.
l 8
MR. DENTON:
Denny, would you mention the flow 9
indicators that are available.
l i
10 MR. ROSS:
Yes, at the local station.
Yes, we 11 covered that.
12 COMMISSIONER KENNEDY:
What is the nature of those i
13 indicators.
~
I 14 MR. ROSS:
It is a dial calibrated in gpm.
15 COMMISSIONER AHEARNE:
On page 4 you have got, l
intalkingabouttheminimumflowwithtwopumpsoperating,youl 16 17 say:
18
.the licensee stated that the results 19 l of his analysis show that a minimum of 720 gpm EFW l
20 ;
flow will be delivered.
i 21 Have you checked that analysis?
22 MR. ROSS:
Let's see.
Sandy?
Or, Tom?
23 ;
MR. ISRAEL: No, we have not checked the analysis 24 p of, the 720 gym that they predicted for one steam generator.
- e. Federal AeDorters. Inc. '
25 ;
However, ' ased on this flow test result that we did c
2287 172 1
+
17 i
1 l
mm 1
see-- here again these tests did not try to maximize the flow l
2 to one steam generator -- we estimated they would have at 3
least 500 gym going into the unit. And if they only had a'
one steam generator operating, that would be a single failure in-5 effect for the unit.
l a
6!
And so if one considers tat if one brings on the i
emergency feedwater at 40 seconds or one minute, if you will, i
7 I
sj into the event, that's our single failure.
Wewouldautomaticayly i
9 bring on the high pressure injection pumps prior to when they 10 would have been turned on if we had waited for 20 minutes to i
11 turn on the EFW and had 720 gpm.
12 COMMISSIONER AHEARNE: So you feel that your tests i
13 at the 500 satisfies any recuirements you might have?
14 MR. IS RAEL: Yes, I do.
I 15 COMMISSIONER GILINSKY:
About emergency feedwater, 16 in reading Bob Tedesco's report discussing the various I
i 17' emergency feedwater systems in 3&W reactors, I notice that some 18 of them draw water from sources which are designed to Seismic 19 Category 1 Standards, and others do not, Oconee being one.
i 1
20 l What significance do you attach to this?
21,
MR. ROSS: Well obviously, standards today are I
22 l different than they were ten years ago.
I 23 l As far as the significance of a seismic event, let's 0
24 ;
sa'y arbitrary disabling all of the aux feed systems.
And, if
..F.e., i neoorters. inc.
25 you follow this -- and also the main feed system that led to a q
l l
2287 173
D 4
18 l
=m I
loss of feedwater transient -- then the core coolability I think !
2 we discuss back in Part B, the ECC systems should be able to i
3l cool th'e core if you make the assumption that this feedwater i
i 4
transient led to lifting of a relief valve.
l
,t 5
The decision as to whether the Oconee feedwater i
i 6:
system main auxiliary should be upgraded, it was not covered 7
in our response to this order.
I'm just not prepared to say l
8 whether it should be or will be at this time.
9 MR. DENTON:
I think the main difference is if you i
10 then have loss of feedwater flow, then you are relying on the I
11 safety system to cool the core.
Whereas if that is available, 12 that would be the first line of defense to turn to aux feed.
I l
13 We changed our requirements in this area some i
l i
14 years ago in Oconee and a number of plants, and they are not 5
15 required to have seismically resistant feedwater systems.
l 1
16 COMMISSIONER GILINSKY: So it is the newer plants i
17 that have the --
18 MR. ROSS:
Yes.
I 19 COMMISSIONER GILINSKY:
-- seismically qualified --
i I
20 COMMISSIONER 3RADFORD: The sentence at the top of page i
21 !
4:
I i
22 l
.no active failure to a mechanical componen:
23 such as a pump or valve would preclude obtaining ETW 24 j flow in any unit.a
>;.e.r.in. con.n.inc.i 25 ]
What is the distinction between an active and a u
k'i 2287 174 a
19 t
mm 1
passive failure?
i 2
And is that relevant?
3 MR. ROSS: Well, the things that have to start; in l
i 4
this case a pump has to come on and a valve has to come open.
I 5 !
If neither of those things happen, it would be an active failure'.
I 1
l 6i Af the pipe broke, the auxiliary reecwater -- the i
i 7
pressure in the pipes caused then that would i
i a
be a passive failure.
9 Certain electrical failures would be either active 10 or passive.
l i
11 MR. CASE:
The theory is that an active failure is l
12 a higher probability event than a passive f ailure.
13 CEAIRMAN HENDRIE: It is a separation between i
14 classes of events.
i i
15 If things have to function, valves have to open, i
16 pumps have to start, electrical circuitry has to -- relays i
17 have to act and so on, there is -- and the experience is pretty la clearcut on the matter -- that the possibility of failures in 19 those required actions are more likely than just something 20 : like a well-designed piece of pipe gratuitously bursting on you.
i 21 '
- "e passive failures would be typically things like l
i I
22 pipe failures.
23 j MR. ROSS:
Are we ready for Part 3?
24 CEAIRMAN HENDRIE: Go ahead and try it.
3.s.e.m a oonen.rne.
25 ]
MR. ROSS:
Next slide then, please?
2287 175 s
20
="
(Slide) i 2l We are on page 9 of the SER.
We ordered that they I
I 3
should be able to initiate and control feedwater independent i
3lI of the integrated control system.
We talked about this already.
i l
I 5!
The emergency feedwater is told to start in when
~
i 6
the main feedwater pumps, both main feedwater pumps of one l
7 unit -- or if a main feedwater discharge pressure is low --
i 8
then if the systems are not bypassed, then two things happen:
1 9
The emergency feedwater pumps for all three units 1
10 start. This is the lower right-hand block. And the emergency 1
ij feedwater valves open, which is the lower left-hand block.
l 12 I think we have talked about these already. This is i
13 independent of the integrated control system.
14 Other features in Part 3 is they didn't have to f
15 modify the procedures.
We did, in the latter part of the review.
16 discuss how to restore the equipment if it is in a surveillance '
17 '
of test maintenance mode, if there is a need.
Thisisdiscussed) 18 on page 10.
19 And we also audited the procedures for emergency 1
20 feedwater, and we discussed what the operators -- how they would' 21,
use these procedures if they had to.
22 MR. DENTON:
You probably know that divorcing aux 23 feed performance from the ICS is simpler in this plant than i:
24 '
is in some others, so this is a somewhat unique solution, and
- sJoceral Aeoorters, Iric. !
25 probably will not be quite as simple a solution in future plants.
0 1
k 2287 176 oit
21 I
COMMISSIONER AHEARNE: Denny, can I ask you a question; mm about page 10?
2, i
I 3'
The first paragraph on page 10, you say that you 4
had to consider readings of certain parameters by using I
alternate instrumentation.
5 I
6 Cbuld you expand on that?
What parameters, and what :
7 instruments?
MR. ROSS: Right.
Let me ask Tom, Bruce or Bill Cottle?
8l i
f 9
MR. NOVAK:
Let me first start, and then I will ask
{
l 10 Bruce to supplement as is necessary.
11 l In part of the review of the procedures, we felt it I
12 was key to the operator, after the immediate actions had taken l
i 13 place, for him in what they have called " subsequent actions,"
i 14 to verify key parameters.
By that we meant, select a different 15 instrument, whatever was appropriate for him in his judgment --
i 16 and we have had discussions with the licensee on this -- to l
17 verify the accuracy of a specific reading.
18 We felt the doublecheck concept was beneficial to 19 him in making sure that he had as complete an understanding of 20 the system as was possible.
So this is one of the things that i
1 21 we insisted on in the development of these revised procedures.
I 22 l Some of the specifics -- Bruce, do you have an i
l 23 l example of a -- Mike, I'm sorry.
Mike Wilber and Bruce Wilson --
I.
24 I I sort of get them mixed up at times.
l
..~.,.,._,..'"!
2287 177 25 (Laughter.)
l
22 mm 1
For example, if you wanted to verify a hocleg 2
temperature, for example, the opportunity to verify the core 3
exit water thermocouple readings to give you an indication that the system _. was well behaved in the sense that you read a hotleg!
4 S
temperature, and then you verify that the core exit water i
6l temperatures were related in the sense they were approximately 7
the same -- they might be a little higher, but it was giving-8 you some accuracy that there wasn't a wide variation --
1 9
COMMISSIONER AHEARNE: All three of these units have i
10 got incore thermocouples?
l 11 MR. NOVAK: To my knowledge, yes.
4 f
12 COMMISSIONER AHEARNE:
I am not sure if that answer i
13 would be yes or no.
i 14 MR. NOVAK: Well, we concentrated on Unit 1, obviously'.
15 We looked at the plant that was operating.
The answer is yes, f
i 16 I have been told, all three do have, yes.
17 COMMISSIONER AH EARNE:
Now if you have been using 18 ! the thermoccuple measurement as an additional instrumentation, 19 how does the operator get that information of thermocouple 20 !
readings?
i 21 MR. NOVAK:
.He has a computer readout available to 22 him, and they did it for me as part of our inspection.
23 1 COMMISSIONER AHEARNE: All three plants?
i 24 i
~
MR. NOVAK: All three plants.
e.Jede,al Reporters. Inc.
25 !
COMMISSIONER AREARNE: Had that concuter readout?
I k
2287 17; I
23 mm 1
MR. NOVAK: Each one has its own.
2 COMMISSIONER AHEARNE: Good.
3 Now, is that the principal key parameter alternate --
4:
MR. NOVAK: No, that was an example, Commissioner.
f 5
Depending on the parameter and depending on the kind of l
~
i 1
6:
instrumentation, you could select another one. The pressurizer I
7 1 water level, for example.
S, COMMISSIONER AHEARNE: That's a good one.
i 9
MR. NOVAK:
All right.
10 He could change by just simply turning a dial and l
Il l read another sensor through that same chain of information, i
I 12 but select one of three. There are three sensors, any one of 13 which would tell him level.
I 14 l COMMISSIONER AHEARNE: Right.
I 15,
But recently there was a question as to -- they all l
16 '
three would have been reading accurately, but giving a mis-17 !
reading interpretation.
What is his alternate instrumentation la that he turns to in that case?
19 ;
MR. DENTON:
Here they have flow instrumentation that l
20 shows that water actually is going in the line, that wasn't i
21 '
available at Three Mile, for example.
22 l MR. NOVAK:
No, we were not able to identify a 1
23 l backup way of verifying level as part of this procedure.
In i
24 '
fact, we are standng a little bit from the emergency feedwater
- ..s.o.,e a eoon.n. inc.
25 ] procedure, but this is a good illustration where we sought to 4
2287 179 t
s I
24 l
1 I
nm I
find indepe ndent ways, where feasible, to measure a paramet&r 2
using a different instrument or a different concept. You can't j
f 3
guarantee that you are going to find the backup that you are a
looking for in every case.
i i
s 5
COMMISSIONER AHEARNE: Harold, your point is --
l 1
6 MR. DENTON: They have positive indication of l
7 auxiliary feedwater flow which was not available.
8 COMMISSIONER AHEARNE: And that flow reading is at l
9 the pump or in the control room?
10 MR. ROSS:
It is locally at the pump now.
l I
i 11 We will be discussing where it should be on our i
i 1
12 long-term discussions with Duke Power.
l 13 COMMISSIONER AHEARNE: So as long as you have the 14,
operator stationed there at the pump --
I 15 MR.ROSS: That's right.
i 16 COMMISSIONER AHEARNE:
I guess you will be addressing' 17 that kind of a question at the same time you go through rhe 18 steps of putting in those additional pumps which we'll end i
19 up reading, to remove the operator?
20 {
MR. ROSS:
That's right.
21 l MR. ISRAEL:
If I can interrupt for one minute, there 22 l are indications in the control room also of ficw on the aux 23 l feedwater lines.
I 24 '
MR. ROSS:
I think those indications are rurbine
- s Federat Aecorters, Inc..
t 25 speed and discharge pressure, are they not?
j 2287 180 m
25 I
i 1
MR. ISRAEL: Flow indications.
2l f
COMMISSIONER KENNEDY:
Flow indications?
3 MR. WILSON:
Yes, they have flow indications for 1
4l Unit. 1 and Unit. 2 that I know of.
I am not sure of Unit 3.
1 i
I S1 They are in the control room near the controls for the j
i i
i 6i admission valves 315 and 316.
i l
7 COMMISSIONER AHEARNE: When you say " indication,"
8 what do you mean?
9l MR. WILSON: Flow gauges that read out in gallons l
i t
10 per minute.
l 11 MR. DENTON:
These have been installed since the 12 shutdown, haven't they' I
I 13 MR. WILSON: Yes, sir, j
i 14 MR. DENTON:
So they are in the process --
i 15 COMMISSIONER KENNEDY:
So they are duplicated, one i
16 down locally near the pump and the other in the control room?
17 MR. DENTON: Yes.
18 COMMISSIONER AHEARNE:
Good.
19 MR. ROSS: The next feature of the order is discussed 20 on the next slide.
Part C.
21 (Slide)
I 22 And we are now on page 11.
i 23 ll The order says:
n 24 j;I
" Implement a hardwired control-grade reactor trip
- .4.e.ru a.coners, inc.
25 '
on loss of main feedwater and/or turbine trip."
i e
4 j
2287 181
i 26 I
mm 1
At the tcp of this slide you will notice loss of 1
2 main feedwater or turbine trip has been added through new l
3l circuitry to provide, as you go through the electrical funda-4 mentals, it leads at the bottom to a reactor trip.
i i
5' This is, I think, a relatively simple portion of the ;
I 6;
five steps. The order said, put it in, they put it in.
So the 7
unit would now trip on either -- the reactor would now trip i
i 8
either on turbine trip or loss of main feedwater.
l r
l 9i The loss of main feedwater is indicated by ',ither I
i 10 the --
1 11 COMMISSIONER KENNEDY: Before this change, where wouldj 12 the circuit have run on this diagram?
13 MR. ROSS:
These inouts did not trio the reactor.
l 14 COMMISSIONER KENNEDY:
Neither?
l l
15 MR. ROSS:
Neither.
16 There were no secondary side trips.
i 17 1 MR. DENTON:
They would have had to await a high-18 pressure trip.
19 MR. ROSS:
If the steam admission valve for the main 20,
feedwater pumps, the turbine driver for the main feedwater I
21 l pumps closes, or if you get low feedwater pressure, this is how 22 vou decide that vou have lost main feedwater.
They picked up a 23 1 spare relay on the hydraulic control system for the turbine to 1
i h
24 indicate turbine trip.
2 Eederal Rtoorters, IrIC. l 25 !
COMMISSIONER 3RADFORD: What is the significance of l
2287 182
.I
27 6
j this being nonsafety-graded equipment?
2 MR. ROSS:
It is not single failureproof, and it is 3j rot designed -- it is not run in safety-grade trays with fire l
4' protection, it is not designed to IEEE.
i 9
5, MR. DENTON:
As an order we consider control grade 6
was adequate, and they were required as a Category I safety i
e i
7 grade as a long-term thing.
l 6
COMMISSIONER AHEARNE:
Do you have any estimate on l
8 9
what type of time that long-term fix would take?
l thisisthescheduleweareaskingforl 10 MR. ROSS:
No, 1
11 in the letter.
l 1
l 12 I don't know if they can get this in a few months i
l 13 or not.
I i
i 14 COMMISSIONER KENNEDY:
Is that simply because the equipment itself isnl't available?
15 l i
16 l MR. ROSS: Qualified equipment.
l 17 MR. DENTON:
Plus our review of safety grade i,
13 instrumentation changes takes longer than those for control 19 1 grade changes.
20 But my understanding, it was on about the same 21 '
timescale as the electric turbine pumps.
22 MR. ROSS: That's what we hope for.
i COMMISSIONER BRADFORD:
Is there a way to sta e
- n. l l
24 i the dif ference in your expectation between a peformance of
- ..s.eer.i seco,w s.ene.
25,
control grade equipment in general, and the performance of i
2287 183
28 i
l safety garde?
mm28 i
i 2
MR. ROSS:
Not in a numerical sense, no.
3 MR. DENTON:
Let's see if anyone on the Staff would a
like to comment on that, on the difference.
j 5,
MR. ROSS: The review team included one instrumencation i
6 control system member, Da'le Thatcher.
7 MR. THATCEER:
I guess I don't understand the i
8 question, exactly.
9i COMMISSIONER 3RADFORD: Well, speaking to someone who i
10 is not versed in the technical terminclogy, and pretend you 1
11 ! were explaining to your 14-year-old son, the difference between !
12 control grade and safety grade equipment in terms of what you 13 expect of it.
14 MR. TEATCHER: Well, I don't want to deal in numbers, i
15 but we consider order of magnitudes better for safety grade j
16,
in reliability.
17 CHAIRMAN HENDRIE: Typically, in safety grade 18 instrum entation, there are the components that will have had 19 l to have had seismic qualification; the circuitry will have to 20 ;
be single failureproof --
21 !
MR. THATCEER:
Multiple channels.
i i
22 CHAIRMAN HENDRII:
-- with multiple channels.
23 MT. TEATCEER: Special indication.
24 l MR. DENTON:
Even cont:01 grade has to have certain
- e Federal Aeoonen, Inc. 3 253 -='d=bility.
With fossil-fired plants you would say con:rci t
2287 184
/
29 j
grade. You would strive for some reliability of performance.
mm 2
COMMISSIONER 3RADFORD: But you can't quantify what 3
the reliability difference is in a meaningful sense?
I a
MR. DENTON:
I don't know really, if measures are 5
available or not.
6 MR. THATCHER: There are some numbers that have been i
7 quoted as far as the safety grade system, four-channel system i
I g
in various reports, on the order of -- I hate to give the i
9 numbers because we always get into an argument about t
10 reliability and unreliability.
But 10-3 to 10-6 range.
11 COMMISSIONER BRADFORD: That is for protection for 12 safety grade.
l i
13 If I understand you, you said there are broad I
i 14 uncertainties attached to that.
15 MR. DENTON: The order just required control grade.
i 16 COMMISSIONER BRADFORD: Right.
l e
17 The last question then is, why is that satis f actory 18 here?
19 MR. DENTON:
Because at least in my view, control 20 grade is not something that is going to fail on every demand.
i 21 ;
It has a relatively low probability of failing.
Safety grade 22 is much lower than control grade.
23 I think there is a high probability of getting turbine 24 i trips or loss of feedwater pumps that is going to result in a 2.s.aerm anemes. inc. I 25 0 trip even with control grade.
But it is not immune to seismic 2287 185
I 30 t
mm i
events, or less reliable.
2 COMMISSIONER GILINSKY: But you did think it ought to I 3,
get shifted over to safety grade after some reasonable period a
of time?
l 5
MR. DENTON: Yes.
6' CEAIRMAN HENDRIE: Let's see, we also ought to note 7
that -- wasn't there either in one of the bulletins -- memory l
l 8
fails at this point -- butteither in one of the bulletins, or I
i I
9 at least contemplated for one of the bulletins, there was a l
10 manual scram requirement in which you would make sure that l
11 the operators understood that if they got a secondary side i
l 1
12 trip which was going to go in that direction, that they would i
13 hit the scram button.
i l
14 MR. ROSS:
It was in Bulletin 053.
l 15 CH AIRMAN HENDRIE:
05B.
i i
16 And then, here the concern was, well, are we getting,l 17 you know, too many new procedures for operators,and the step 18 here to hardwire was to provide an automatic tripping which I
!9 would supplement, probably relieve the operator of the need i
20 ;
to take action.
He is still available to take that action, t
I 21 and in terms of speed of getting it in, you just can't ge I
l 22 i the qualified components and everything you use, shaken down l
23 !
on sa ety grade circuitry that rapidly. So that the thought i
24 l was a con:rol grade ~ automatic trip would be a considerable 3 7.e., i A.:er e,s. ine. '
25 i improvement over having to have the operator having to punch l
7287 186
31 l
mm I
the button, so it is in that sense that the measure was I
2l implemented for the interim operating period as I recall it.
I 3
MR. ROSS:
That's correct.
4 One other feature on secondary trips, is we did l
S.
review it to assure that since this is safety grade,that 6
this installation did not degrade the existing reactor protection 7
system, and we noted that near the bottom on page 13.
8 COMMISSIONER KENNEDY:
On page 13, Denny, you l
9 indicate:
i 10 i "The licensee has committed to perform a monthly II test on the added circuitry in order to demonstrate 12 its ability to cpen the ac circuitbreakers."
[
i 13 Is this in the interim period until the longer-term i
14 fix with safety-grade equipment is installed?
Or, is this a l
i 15 i
permanent arrangement?
r 16 l i
MR. ROSS:
I think the tech s. cec rec.uirement will 17 probably overtake this one.
18 I th,nk we are going to reconsider whether a monthly 19 test, when we d. tech specs, is enough.
But this_certainly 1
20 ) would be it for a short term.
i 2I '
COMMISSIONER AHEARNE:
On that same sentence, just 22 '
to make sure I understand, you are basing this on the i
23 licensee's design of the circuitry?
24 l You have gone through it, analyzed it yourself to
- ,.e,ae,e neconm. sne. '
e 25, assure yourself that it will perform as designed, is that n
j 2287 i87
.i
I 32 i
l I
=m I!
correct?
2 MR. ROSS:
Yes, sir. From both Bethesda and from l
3 Clemson; or whatever the nearest town is, Seneca, Dale Thatcher, a
who you just listened to, did the review from the NRR end, 5'
and at the site the resident confirmed the checkout tests.
i 6
And we have a detail of line diagrams on how it was done.
7 I think the most detailed part of the order --
l l
8' COMMISSIONER AHEARNE: Are you finishing C?
Are i
9 you through with C?
10 MR. ROSS:
Yes, sir, through with C.
l COMMISSIONER AHEARNE:
Let me ask one other question!
Il 12 on C then.
t 13 On page 11.
At the bottom, you say:
l 14 "The licansee has estimated that the new 15 anticipatory trip will result in a reactor trip 3 to 16 10 seconds earlier.
17 And my question would be, is that time important?
18 And if it is, do you agree with the licensee estimate?
19,
MR DENTON:
It is definitely important.
That i,
20 l is one of the reasons we requested these.
I'll let Denny 21 l answer about the amount of time.
22 MR. ROSS: Well, I agree.
If you looked at the one 23 I data point we have from the point of TMI 2 SCRAM on high 24 '
prsssure, that is about the right time.
c.
.e.,.i m.oonen. i nc.
25 i COMMISSIONER AHEARNE: Yes, I thought it was important.,
l l
2287 188
33 i
mm j
too.
The question really was, do you agree with the estimate?
2 MR. ROSS: What it does is, it gives more inventory in:
3 the steam generator until the auxiliary feedwater system is t
4 on.
5 COMMISSIONER AHEARNE: I am just asking, do you agree !
6' with the phrasing of the licensee?
l i
i 7'
(Simultaneous discussion)
I 3'
MR. ROSS:
Let me see.
I can only -- let me break 9
that into three parts.
10 If the event was a turbine trip, then you are going i
11 to get a reactor trip very quickly, within 10 seconds, probably.
12 If you get a loss of both main feedwater pumps due i
i 13 to the valves going shut, then I can see that happening very 1
14 quickly.
i 15 But, if the thing that senses loss of main feedwater l 16 is pressure decause the pumps are either coasting down, or the 17 valve ic slowly going shut, then the 3 seconds might be more 18,
appropriate.
That's the one I would think would be more 1
19 l sluggish, but it's got the compliance of the hydraulic system E
i 20 ; where the other two are binary electrical signals, either on or i
21 l of f.
i 22 l So I would say 3 to 10 seconds is probably -- is l
23 i that the question you asked?
i 24 COMMISSICNER AHEARNE: Well, what I In really asking
- Jeeersi aeooners. inc. j 251 is char I would rather have the Staf f's estimates of times, or il N
i 1
2287 189 m
s 34 i
mm 1
the Staff saying, yes they have confirmed or agree with. And i
2 that is what I was looking for.
3 MR. ROSS: Oh,I understand the distinction, we a
didn't write that in.
I agree those are reasonable times, that 5
if they can do trips, that would be the time that I would have 6;
written it, if I had written it.
7 COMMISSIONER AHEARNE: That you would have written
--l i
i 8
(Simultaneous discussion) i 9
MR. DENTON:
I think we were just bein'g descriptive i
10 there.
i 11 MR. ISRAEL:
I am the one that wrote that sentence.
1 12 We did not check to see if the times were apprcpriate!
i 13 or not. That was just to give some sort of indication to the i
14 reader of what the estimate would be for --
i 15 COMMISSIONER AHEARNE:
I wasn't even asking whether 16 they are appropriate.
17 What I was really asking is when the estimate of 3 18 to 10 seconds is made, I recognize the licensee has made that 19 estimate, dc you agree that that's about the right kind of time 1
20 l that would result, as a result of this anticipatory trip 21 circuit?
22 MR. ISRAEL:
I will put it negatively.
23 ;
I have no basis for disagreeing with it. The times 24 l vary from 3 to 10 seconds based on the initial pcwer level.
Of 3 5ederal Reoorters. Inc.
25 h course lower power levels give you longer cimes, higher pcwer il 1
2287 190 v
11 i
I j
35 I
i mm j
levels give you shorter times.
2 And I guess at 3 seconds or 2 seconds or 4 seconds, I
3 becomest sort of arbitrary in my mind.
a MR. DENTON:
I think I would like to summarize by 5
saying that the rationale for requiring it was chat in time l
6 these numbers do correspond with my understanding of the amount i
7, of time for the range of transients, power levels and conditions, f
I g
that there might be some remote transients that can go by you 9
a lot o f times, there might be others that --
10 COMMISSIONER AHEARNE:
Yes.
11 MR. ROSS: The lower high pressure reactor coolant I
12 trip set point is now so close to the operating point that it t
13 well, indeed, might be a race as to which gets to you first.
i 14 And in that case the number might be a draw.
i 15,
COMMISSIONER AHEARNE:
Yes.
I 16 MR. ROSS: But Ed reminds me, it did take 10 or 12 17 seconds for TMI.
We did have one data point on Rancho Seco t
18 that had a feedwater transient after the set points were altered'.
I 19 l I don't remember how long it took, but we can find out.
20 CHAIRMAN HENDRIE:
But really the point of the 21 :
automatic trip is to snap it down and get that full power heat i
22 l;generation out of the machine as soon as you have any indica-23 l tion that there is something going on on the secondary side, 1
24 f where you maybe would like not to have all that heat generation.
c.Jeeerei neoomn. inc.,
25 ;
And in particular to cover transients, perhaps it would run ha d
2287 191
36 i
in the TMI style, where you would continue to get full power mm 2
generation for some seconds into the loss of secondary side 3
coolant. capacity.
a So I expect, as Denny says, on some transients, I
S why perhaps a number of the lowered reactor pressure set points,,
4 6
will shut the machine down pretty oromotly. And then it may be i
l 7
a race which side you get.
i i'
8 But this ought to cover some otler transients 9' where it would go on for a few more seconds.
10 COMMISSIONER AHEARNE: Well, I was just trying to makel 11 the point that if we have -- if we are reaching a decision, or 12 the Staff is reaching a decision based upon some numerical i
13 estimates, it is fine to have the licensee do some numerical i
14 estimates.
i 15 But if it is important to our result, if those l
t i
16 numerical estimates are important, then I think the Staff has 17 to agree with those estimates.
That was really the point I 18 was trying to make.
19 MR. ROSS: This will come out when we discuss the 20 analyses about no feedwater.
I Let me move then to Part D, if I may, 21 i
22 The order for Part D was very short, butthis has 23 been where we spent most of our time:
24 l
" Complete analyses for potential small breaks 1
. 7.wm m.oon.n. inc.,
25 ;
and implement operating instructions to define L
?>87 19?
a
37 I
operator action."
2 This is pages 14 through 28 of our SER.
3 Briefly, B&W furnished the analyses.
As described I
4 in here, we did a couple of Staff audit calculations.
I 5
Leei s move to the next slide.
6 (Slide) 7 A little bit of the su= mary of calculations is shown j 8!
here.
These are B&W calculations. The first column is the i,
9 break area and the break location.
l 10 The second column is the operability status, if any, l i
II of emergency feedwater.
i 12 The third column is the operability of the high-l t
13 power injection system.
l Id And the far right-hand column is the results.
I 1 *8 COMMISSIONER AHEARNE: What does ANS mean?
16 MR. ROSS:
The third row was a stuck-open pilot 17 operated relief valve. That was the appropriate area in.0073 IO square feet.
19 There were two assumptions made.
One is the decay I
20 heat. The 1.2 ANS resfers to the Appendix K requirement on decay 1
21 l heat.
22 MR. CASE: ANS is a standard.
2**
MR. ROSS:
American Nuclear Standard Society 5.3, 24 l something like that, 1973 version, I believe.
- e. Federal Reporters. Inc.
SC'
'i I think we reported tothe Commission last year that 2287 193 4
-38
~ n'
"~
j some of the Ccamission-sponsored research showed that data 2'
now could support,in terms of isolation, lower numbers.
Not 3
necessarily in context with other conservative assumptions.
4 But, if you were just looking for the best estimate of decay 5
heat, then even the nominal decay heat curve which we call i
i i
6:
1.0 s somewhat conservative, depen ding on the period of time.
7 Well, for row 3, sensitivity calculation was done 8
where the decay heat was the parameter.
And in one instance 9
you would uncover, I think the time was between 30 and 40 10 minutes, and you would have to -- it would start to uncover.
i 11 Something additional should be done there.
If you l
12 use slightly conservative decay heat for the 1.0 ANS, the core I
I 13 would not uncover.
14 MR. CASE:
Of course there are other things you can 15,
do,, start other HPI pumps.
l 16 MR. ROSS:
Right, or restore emergency feedwater.
17 And subject to check, I believe all these 18 calculations are for a 2772 megawatt plant.
Is that right, 19 Zoltan?
I i
20 DR. ROSZTOCZY:
Zoltan Ros:toczy, NRR.
(
21 l!
Calculations were performed in a generic sense 22 l to be able to cover more plants than just Oconee,and therefore 1
23 l the selected the plant parameters in such a way that it is 24 conservative for each of the plants.
And it represents the a-e cere Reconen. sne. ;
25 Oconee class of plants.
l 2287 194
39 mm 1
MR. ROSS:
The purpose of these calculations was a t
2 three-part purpose:
3i Do the calculations; from them develop -- and these l
4, were done by B&W, reviewed and approved by us, and in some j
l S' instances nullified -- develop operating guidelines which will i
6 be sent to the individual utilities; from the guidelines plant-1 7
specific procedures would be developed, new procedures for 1
8; coping with small breaks with different failure assumptions.
}
9l It is the guidelines and procedures that I think we i
10, actually looked at the most because the concern was more fo-i l
1 11 the role of the operator, than the ansivsis.
l 12 We did review in detail, both with the onsite team i
f i
13 and the review team back here Bethesda, these procedures. As I
14 pointed out in Part D, after these three-step processes, that i
15 is analysis, guidelines and procedures, we audited the j
i l
i 16 l operator's understanding -- let me give you the specific i
17 !
phrase because we made a statement about some " generic 18 ' deficiencies."
l 19 COMMISSIONER AHEARNE:
Page 21 is where you 20
" generic deficiences" are.
That's on E?
21 l:
MR. ROSS:
D and E kind of blended together, t
i 22 {
Additional training was done. In fact, it probably 1
23, would be just as useful for this discussion to lump D and E 1
24 l because our review, I think the only separation was the
- e Fecersi Aecorters, Inc..
25 [ arbitrary separation in the order.
l 2287 195 i,
40 i
i i
I mm I;
We did audit the understandinc. of rhe TMI 2 traininc.,.
i 2' As a result of the interviews that we had with the plant 3; operators, some additional action was taken by Duke Power, 4
documented on May 16.
l 5
The action --
i 6!
COMMISSIONER GILINSKY:
What was the problem?
7, MR. ROSS:
The problem was, through our initial 8
interview, was twofold:
First, that Duke had not done a 9
management audit of the training.
And just as a matter of policy 10 we felt it should.
11 Second was that discussions with the operators 12 : indicated some training, additional understanding of certain i
13 parts of the TMI would be useful, and some of our initial t
14 discussions, probably due to the timing they weren't as familiar 15 with some functions of the bypass switches as we would have 16 preferred.
17 Well, Duke did the order and gave a written examinatien.
18 Based on that it indicated for some people, additional training 19 would be needed.
And there were some arbitrary cutoffs 20 l established, and below this cutoff people Just wouldn't be d
i 21 ' used on shif t until the training had taken hold.
22 j COMMISSIONER GILINSKY:
Did we review the examination?
l 23 l MR. ROSS: We did.
24 i Mr. Cottle, I believe, did if you want any further
- ..;,eersi m.cor.,s. inc.
25 ! details.
g7 jgg
41 l
mm 1l MR. COTTLE: We reviewed, both myself and Mr. Bu:y, i
2l from OLD, reviewed the exam that was given. In addition, 1
3 Mr. Bu y did a grading correlation on the results, in that he 4j took the grades that Duke Power Company came out with as
{
l 5;
far es assigning the individual grade on the exam -- he took 6
ten of the examinations, did his own grading, and with very, i
7l Very close correlation as far as his numerical grades versus i
i 8l the grade assigned by the training program.
l i
9' I obtained copies of the examination papers for j
i 10 ' those individuals who scored better than 90 percent, and I 11 have rereviewed their papers, and I again concurred and find I
12 a close correlation between the way I graded their exams and 13 the way they were graded by the training staff at the Oconee 14 Nuclear Station.
15 COMMISSIONER AHEARNE:
A couple of questions on 1
i 16 that.
17 A.e you going to require the same thing, now that la you have seen that those are generic deficiencies on the way 19 j they are approaching it, are you going to apply the same thing 20 l to the oth er plants?
21 MR. DENTON: Yes, sir.
I intend to inform the other 22 ! 3&W plants of what we expect in the way of scope and content i
2IlIcf the training program, and evaluatic:. at various steps along 24 j the way.
2 Federal A eoorters, Inc.
25 I think our original concern with Duke training was i
2287 197 a
42 l
I mm 1 !
the lack of formalization at each stage.
And Duke has now 2
retained outside consultants for the people who don't pass 3-the 90 grade in their work.
4 So, they will follow, they will establish programs i
i s'
and audits for future people and we will require the other i
6:
B&W plants to do the same kind of training.
I 7
COMMISSIONER AHEARNE: Denny, you mentioned that there a4 was that one area in the bypass with the switches that they i
9 hadn't emphasized as much as you would like.
10 MR.ROSS:
What happens, is when you put in --
11 COMMISSIONER AHEARNE: Now my question is, are you I
I i
12 l going to call that also to the attention of the other plants?
i 13 MR. ROSS:
Yes.
14 COMMISSIONER AHEARNE: Or, is that a Duke training e
15 '
rather than B&W?
l I
16 l "R.
DENTON:
There were several areas where we l
17l felt that there were weaknesses in areas, and so we were I
18 concerned not only with the bottom line score of the test, 19 1 but whether or not there were certain technical gaps that l
20 i everyone had.
21 And so our resident inspectors and B&W licensing i
22 l people did go through this and found the people who are 23 qualified don't have these gaps, but the training didn't 240 sufficiently focus on certain areas so there were larger
- ....rm m.oonen. inc '
25 proportions of the students who didn't get those --
t 2287 198 a
I 43 mm I,
COMMISSIONER AHEARNE:
Is that training by Duke or i
2l B&W?
l 3
MR. DENTON: THis is training by Duke that we i
4' originally audited. They are now having training by two outside i
5 organizations.
i i
6' COMMISSIONER AHEARNE: A last question on that.
i 7-I notice that you have a 90 percent passing score.
I 3
Is that a policy that we will now be implementing,that 90 9
percer.t be passing scores?
Sounds good.
i 10 l MR. DENTON: We thought that was necessary in this Il case in view of the lack of formalization and verification 12 of test scores, correlation with knowledge that was required at ;
13 Duke.
And we concurred that everyone who has made above 90 I
I 14hisadaquatelytrained.
l i
15 l
- Now, whether we want to adopt the same thing, maybe i
16
! 'ists at other plants can be more stringent, we will have a 17 h different grade.
1 I8 COMMISSIONER GILINSKY:
Are the croblems we have i
i 19 been discussing --
d0' MR. DENTON:
Let me amplify that a little bit.
21 '
We didn't feel like the students who had passed 22 70 were adequately trained here, and so we looked for other I
23 benchmarks and came to 90 based on our own interviews with the 1
2 l people and our own look at the qui::es.
And those people in
- ...cer.i n.oor en. inc.1 ne '
~J that group are adequate.
J 2287 199 a
44 i
mm 1
MR. CASE:
It depends on the difficulty of the quiz.
2[
COMMISSIONER AHEARNE: Yes.
Of course.
i 3;
I was, nevertheless, wondering wheth e r this I
4i indicated a change in philosophy of general operator qualifica-1 I
i tions.
6 MR. ROSS: One thing we did do this week is that on 7
Tuesday, we had an information exchange here in Bethesda with I,
3l the residents for the other four stations, to have a cross-9, fertili:ation along these other items where the members of I
I this Oconee team discussed what they did in the way of analysis,'
10 i
11 l in which guidelines are perceived so the residents can trade I
i 12 i notes.
This is important, of course, i
MR. DENTON:
I think it is f air to say we. didn' t 13 i
t 14 find " business as usual" adequate. We were expecting a higher 15, degree of scope and content in the training and better retention!
l, 16 ' than perhaps is the usual standards.
17 l COMMISSIONER GILINSKY:
Are the things we have been i
18 ldiscussing what you had in mind when you referred to generic l
19 [ deficiencies?
i 1
20 MR. ROSS: Yes, yes. Right.
l 21,
(Slide) 22 i This is the same subject on small break procedures i
23 and training.
Some of the elements we emphasized to recognize 24 ' a small break. They could vary in si:e.
And what you do a
ere seconers. sec. '
25 depends on whether the pressure goes down fast, slow or maybe 9
2287 200 1
il j'
45 i
i mm 1! not at all.
And the importance of the high-pressure injection, 2l turning it on and leaving it on, turning it off, as well as 3
simulations at B&W where you follow TMI and then you go back l
4i through it again on the simulator the way it would have been t
5 a lot better to have done it, rather than the way it was done.
I 6
I think that concludes the discussions both on D and i
7 E.
l 8i COMMISSIONER AHEARNE:
Can I ask some questions?
9 I noticed that on pages 15, 16 and 17, the focus I
i 10 ! is on, in one of your charts,the loss of feedwater for the i
i i
11 ! first 20 minutes.
I, i
12 '
And then on 17 it mentions that you don't get core i
13 l uncovered provided you get initiation within 20 minutes.
I I
I 14 What led you to 20 minutes as the criterion?
j I
I 15 '
MR.ROSS:
Zoltan?
l 16 l t
DR. ROSZTOC Y:
The 20 minutes is not a criteria.
17, Licensing andB&W has been asked to perform an analysis and i
18 ! establish approximate time available before the operator i
19[would have to take an action.
20 They have done some calculations for this, and the i
21 conclusion is for any of the plants, at least 20 minutes is it 22, available for the first cossible break si:e.
If the break size 1
22 ' is somewhat smaller, somewhat more time is available.
24 Also, there are some plants that have a little bit
&.w.i m.oonm. inc. a i
25 l more than 30 minutes.
0 1.
2287 201
46 l
a 1l The purpose of this was to compare this to the mm 2
discussions which were presented under Item A of how fast do 3
we expect that an operator would act.
t I
4; And I believe the conclusion from there is that the I
I 5
operator is expected to act significantly faster than the 20 i
I 6' minutes. So on this basis the analysis shows there is enough 7l time for operator action.
gi MR. ROSS: Thatconclusion is in the middle of page i
9 15.
l 10 COM'4ISSIONER AHEARNE: Yes.
11 !
In that sama sort of set of issues, on page 16, you i
i 1
12, say that Staff is determined that -- and you go on to give l
13 I the conditions.
I 14 l Is one of the conditions also that the operators 15 do those actions that they have committed to memory, as listed 16 on page 17?
MR. ROSS:
Let me ask Bruce, the immediate action 17 li I l 18 Umemori:ation?
19 Did you get the question?
I 20,
MR.
WILSON: No, I am not entirely sure I got the whole 21 question.
22 COMMISSIONER AHEARNE: The quesion is, on page 17 you i
22 s mention immediate operator action requires certain actions that il 24 are committed to memory by the operators.
ics r.deral Reoorters. Inc. l 25 MR. WILSON: Yes, sir.
2287 202 3
'l..
47 i
i mm l'
COMMISSIONER AHEARNE: In this small break analvses 2
that you are considering, in which you have this "no core i
3 uncoveries" or " things are done within 20 minutes," is this i
i al predicated upon the operator taking those actions that they i
5 have committed to memory?
i i
6' MR. CASE: Before 20 minutes are uo?
i
\\
\\
7, COMMISSIONER AHEARNE:
Very quickly.
8, MR. WILSON:
Well, the si rgle item that he has to i
i 9
do -- and this could go on for longer than 20 minutes -- is to i
l 10 < maintain his high pressure injection at full flow. Trat is a i
11 single item.
12 l There are other items and immediate actions that are i
13.lrequired, but if this action is taken then the 20 minutes, they l
14 ! can go on indefinitely. And on the basis of our discussions with c
15 l the operators down there, they are all aware of this criteria.
I i
16 l COMMISSIONER AHEARNE: That is really the only critical l
r 17' thing.
18 MR. WILSON: That's the only critical thing.
i 19 !
CHAIRMAN HENDRIE: Let me ask a cuestion about the t
20 li analysis, and about that. In the analysis do we count on reactor 21 pressure coming down and hitting the high pressure injection trip i
22 point?
Or, is this an operator institution at an early point?
1 23 ;
DR. ROSZTOCIY:
One of the main points that we had i
24 l, three weeks ago when we were discussing this with 3&W and with te um moomn. inc.
25 the Commission was that the small break can show different kind i
2287 203 I
I 1
i 48 I
mm 1
of behaviors.
The type of behavior which is a continuous i
1 2ldepressurization, the presssure is continuously dropping in the 3
varioustECC5 cystems -- first high pressure then safety injection i
l 4
tanks and then the low-pressure tanks coming into play, has i
5
, been evaluated in the past for small breaks.
i s'
CEAIRMAN HENDRII:
T hat is sort of what we used to i
i 7'
regard as the traditional classic small break analysis.
S i But generally those small breaks were larger than i
f 91 these small breaks.
10 DR. ROSZTOCZY: Yes.
i 11 So the new portion of this analysis was to extend i
I 12,
l this further and see what other t.v.oe of behavior is possible.
i 13 Doing that, B&W found there are three other possibili-14 I ties. One of them is when you depressurize tothe secondary I
i i
15 i existing pressure and maintain -- the primary system is settled 16 just above the secondary system pressure and stayed there for 17 a while.
18 f The second one is when they break even smaller, then i
19 3 there will be a repressuri:ation to some level, and then when P
20 ' natural circulation again established, it breaks down.
21 l And the fourth one is when you really don't create 22 4 voids, where the break is so small, so you stay in a subcode h,,
[3, state and go through on this.
4 2# h The operating procedures, the guidelines for the ce ral Reporters. Inc. -t 25 ! operating procedures, the ones which were done by B&W were very 2287 204 d
i I
49 I
I!
carefully done to cover each of these four possibilities.
am I
2l So independently, which one happens to be the actual case, i
3 there is guidelines given tothe operator how to follow.
4 The guidelines do differ somewhat depending on what i
I 5
does he see; whether he sees a repressurization or whether he I
6 sees continued depressurization, his actions are somewhat i
7' different. The differences are not very large, but there are 8, some differences,and these have all been factored very carefully I,
9i into the guidelines and I assume into the emergency procedure.
l 10 !
COMMISSIONER AHEARNE: Zoltan, can I ask you another t
11 ! question.
I 12 Sometime ago, I guess about a month ago, you had j
i 13 done a generic review of transients and B&W reactors. And,vou I
14 had mentioned in there the transient analysis B&W does, usually 15 limited to the first few minutes, and that you have to really l
16 lcarry out the stabilization through the whole time period of the i
17 l event.
i la !
And you went on to point out in that case sometimes.
19' operating procedures tnat are appropriate in one, may not be
!I 4
20 g accrocriate for another.
21 From that perspective, are you satisfied with the 1
22 l orocedures that have been develooed for Oconee?
Il 23i!
OR. ROSZTOC"Y:
For the case of the small break 1.,
24 j class of coolant accident, or which results in a small break y.i m.oon n. inc. jii a
25 l loss of coolant, then the main problem is to handle this.
I f'
2287 205
i 50 j; believe this is what has been done during the past month mm i
I 2
and it has been done in a satisfactory manner.
i 3;
I do see, however, a need to follow through in other 4f events, including transient and other accidents, for example, I
5 steam generator tube rupture, and do a similar type of thing i
I 6
for what has been done for small break during the past months.
i
! To do this systematically and to cover all the various events 7
l l
gl that are postulated in Chapter 15 of the SERs will take some i
i 9,
time.
I i
10 ;
And I think this is not something that we expect fromi i
11 the licensee or from B&W to be done within a very short time.
i 12 j But I think it ought to be done and this is part of the i
13 recommendation of NUREG0560, which is called sometimes the 14 Tedesco Report.
I I
i 15 ;
MR. CASE: That is covered on page 16, and further i
l 16 l analysis we will require Oconee to do.
f h
l 17 l CEAIRMAN HENDRIE: And everybody else.
i I
MR. CASE: And everybody else.
18 l 19 )
MR. ROSS: There are some additional issues that we ir end 7 20 ' wanted to discuss beyond the scope of the report.
21 1 22 l 2287 206 23 h il 4.....,1.,..m.j]
24 25 :d.
o 1
1
I
- R 4393 51 l
r.: ste 1 I
- - 8 I
MR. DENTON:
These are issues that the staff 2
considered in the course of the review.
There are some that i
3 we hadn't thought of when the order was written, and we want 4!
te just run through them so you are aware they're issues, 5
so ewhat in the main frame of the plant from A to I, although i
1 6,
somewhat related.
i 7
MR. ROSS:
Let's go to the next slide, then.
l l
BI (Slide.)
l 9'
CCMMISSIONER GILINSKY:
Are you finished --
I 10 '
MR. ROSS:
With the main frame of the SIR, yes, sir.
II l COMMISSIONER GILIMSKY:
I've got a question on that, I2 on these examinations of the operators.
In your SER, you said 13 l,Duke has committed to a written examination and individuals Id must receive a grade of 90, and so on.
I3 l Now, are these examinations to be given?
I under-16 f stood they had been given.
I 17 MR. DENTON:
A written examination has been given, i
13 l and it was on the basis of our review of that examination and 19 [' the scores of people who made various scores, that we wanted i
20 h a different examination established.
21 1 Let me ask the II inspector to discuss the future i
a2 l acticas.
We have now agreed on criteria and a program to
,3., assure tn.at only -- :n.at no one can pass it who is not by
'l 2#
definition trained.
And so, as pecple pass that, they become a
frei neooners, inc.
2"C l,' defined as trained and they qualify for a service in the con:rcl tl 1
2287 207
te 2 I
52 l
l 1l room to meet the requirements of the tech specs.
If they have i
2 enough operators above that to operate one unit, they can 3
operate one unit.
When they get enough above that to go to 4
two or three, then they can operate those units.
But it 5
doesn't mean that they necessarily have today enough operators 6
to operate all three units.
7 COMMISSIONER AHEARNE:
How about two units?
8 MR. DENTON:
Let me ask the I&E inspector.
9 COMMISSIONER GILINSKY:
How many individuals have 10 taken the exam and how many have scored over 90?
l l
11 l MR. COTTLE:
I'm not sure of the total number who i
12 have scored over 90.
I have only reviewed the papers for those 13,
people who are committed to the operating unit, Unit 1.
There i
14 were eight individuals who are committed solely to being I
I I either at the controls or a senior reactor operator for Unit 1. '
15 I
i 16 l Those are the people whose papers we reviewed and verified i
17 l that they made above 90.
i 18 I know that there are several others -- and I don't 19 h really have a number for how many -- who have made more than i
20 ;
that.
I believe it's on the order of maybe four other individu--
21 l als at this time.
i 22 !
MR. DENTON:
They are reconstituting shifts in I
23:l order to comply with the tech specs for qual _fied operators, i
1 24 ' that they must be composed of individuals who have passed this
.rai seconen inc.
25 test, and so we find this criteria inadequate.
And as they i
'l 2287 208
i.te 3 53 1
qualify people, they will be able to operate units.
2 COMMISSIONER GILINSKY:
What happens if a man fails
!3' to scor.e this grade?
Does he just take the test again or l
4 what?
5 MR. DENTON:
I'll defer to Bill or paul for the i.
6' details.
7 MR. COTTLE:
They have two audit examinations that l
aj have been prepared by the training services group, both of I
9 which have been reviewed by myself and Mr. Buzy.
There have 10 been a number of individuals who, on making 90 on the first 11,
examination, are being retrained to the extent that their l
12 l supervisor has sat down, has gone over each of the aspects that i
I i
13 they missed, has provided additional instruction.
And the i
14 individuals, by observation, have invested time in self-study.
I 15 Some of these individuals have been allowed to take 16 a second audit exam that we have reviewed and in fact have f
17 made higher than 90.
These individuals we feel are qualified.
18 There will be no further audit exams as far as, if 19.
you flunk the first two you take a third one or you take a 20 l 1
fourth one.
There will be no other audit exams.
If you make 21 ;
less than 90 on these exams that we have icoked at, you will 22 in fact go into the normal requalification, ongoing operator i
23 l training program, and have to satisfactorily complete that i
24 program and demonstrate your knowledge on a completely different
, i m.oon.n. inc.
25 :l exam that will be routinely administered prior to being 6
228.7 209
tte 4 54 l
i 1
allowed to go on shift on an operating unit.
2l CCMMISSIONER AHEARNE:
In that reexamination l
3 requailification, at the =cment that doesn't have the emphasis i
l 4l upon the TMI sequence, does it?
1 i
i 5
MR. COTTLE:
Say that again, sir?
i 4
6j COMMISSIONER AHEARNE:
The requalification cycle --
I 7
MR. DENTON:
Yes, it does.
They are putting a heavy g'
focus on --
t 9{
MR. COTTLE:
They have a B&W consultant on site.
10 COMMISSIONER AHEARNE:
I was just wondering whether i
11 the potential existed to fail twice on this TMI-oriented cycle, i
12 go back for requalification and pass the requalification without 1 i
13 ever having had to pass the knowledge of the TMI sequence.
14 MR. CCTTLE:
No, sir.
That potential does not exist.
15 We reviewed the initial plans for a new segment for the requali-16 fication program, and we will review the final segment of the 17 requalification program.
la MR. ROSS:
Okay.
These additional issues were 19 l brought up during our review by one or nore people.
i c
20,
(Slide.)
l 21 There was a suggestion that each facility -- this i
l 22 is the first item -- ought to be analyzed for natural circula-i 23 t tion performance, including for one loop once-through steam 24 generator operation is taking place.
The S&W analysis document
- erneaeoorters.ine.l'
.c 25 did provide a summary of all the natural circulation tests and 2287 210
s.te 5 55 l
l l
I believe it did an analysic for one steam generator also.
2 Is that not true, Zoltan?
i 3l DR. RCSZTOCZY:
Analysis has been performed for i
I 4l close cases or cases when both steam generators work, and also i
5 for cases when one steam generator wcrks.
However, the physical I
6-behavior with the core question is I
7 kind of a new question for us, and we have not seen that a
describing the physical behavior, and we don't have verifica-9 tion that the code correctly predicts all of this.
So we are 10 looking for a long-term type of confirmation of this, which I 11 believe is discussed under some of the other items.
12 If that is done, as I assume it's planned to be done, 13 it probably will take care of this concern.
14 MR. ROSS:
We thought this concern was satisfactorily 15 addressed, at least for the short term, when you consider the 16 requirements of NUREG 0560.
17 MR. DENTON:
All of these items I consider appropriate 18 to be included in our short-term interim look at what we ought i
19 I to be doing differently.
But they are issues which were 20 ~~ generic for either all B&W plants or all PWRs, and seemed to 21 he ones that ought to be considered outside of the main frame I
22 ] of the Oconee order.
23l' MR. RCSS:
There was a reminder that we should 24 ! make sure that whatever analysis was done was consistent with agrei seconen. ine. l 25 ;
the ECCS and the auxiliary feedwater system in their as-buil:
2287 211
- e6 56 1
1' configuration.
Or, looking at it the other way, to make sure, 2
if you assumed 800 gallons per minute, that is in fact what i3' was delivered.
Certainly that's a conclusion that anyone would I.
I 4,
agree with.
The plant has got to run the way you assumed it i
5l would.
\\
6i COMMISSIONER AHEARNE:
Now you're saying tne analysis i
7' was on the as-built.
t 8
MR. ROSS:
The concern was not that something was 9
wrong; just something that ought to be checked.
10 COMMISSIONER AHEARNE:
Right.
And are you saying Il that it was checked?
i i
12 :
'4R. ROSS:
Well, the loss of feedwater transients i
13 that were done, that says nothing delivered for 20 minutes and 14 then feedwater came on --
15 MR. DENTON:
We have not rechecked, as part of this 16 exercise, the LOCA analysis against the original design of the 17 ECCS.
Is that --
18,
MR. ROSS:
Right, not in this exercise.
We didn't 19hgoback, as part of this exercise, and say that the pump head t
20 ;
curve is, for example, is what was claimed.
This is the kind 21 of thing that would ordinarily be done during start-up esting i
22 l and documented when you start it up.
i 23 i CCMMISSIONER BRADFORD:
Do you know that it was done?
I 24 l MR. ROSS:
No, I don't know.
I can't describe
- h,v seconen. inc. i 25 depth to which the as-buil configuration for Oconee was t
7787 212
- te 7 57 1,l checked against the assumptions.
i 2
MR. DENTON:
Certainly the staff at the time it l
3, licensed the plant did what it was doing at that time.
- Now, 4
we don't know to what extent the audit was complete, and so 5
this suggestion is that at seme time there should be a complete 6'
recheck.
I think this kind of idea is probably a good one for i
7l our lessons learned group to think about, the need to reconfirm 1
8!
these things.
l 9!
But there is no -- we have no basis for concluding i
i 10 that it was not done.
11 MR. ROSS:
I believe chapter 14 of the SRP discusses 12 start-up test requirements on this subject, the specifics of 13 what we did in 1973 --
14 COMMISSIONER BRADFORD:
The SRP wasn't in effect 15 l then.
16 MR. ROSS:
No, but the start-up tests were.
17 COMMISSIONER 3RADFORD:
And they would have been the 18 same?
j I
19 '
MR. ROSS:
Well, I don't know.
The SRP should have i
I 20 documented what we were doing anyway.
But I don't have a line 21 '
by line comparison.
We can provide t.iat information.
It exists.
I 22 i I just don't have it with me.
I I
23 (
MR. DENTON:
I think it's fair to say we probably 24 l' took more on faith and did less complete audits at that time r i neoorters. ine.;l 25 than we do today.
.n 2287 213 a
o e
te 9 58 I
l l
l' MR. ROSS:
Just in general, the third bullet, more 2
reliable and more autcmatic feed system.
I believe the Oconee 1
3f aux fee'd system is more reliable than it was a month ago.
I l
4' was automatic then.
But to the extent you can get it, it's 5
more automatic now, if that's possible.
61 The fourth bullet for Oconee, we did review the i
7l procedures and various directives for running the plant.
There i
8i had been a concern that maybe we were too specific on the matter-l 9 l' of running the HPI in terms of arbitrariness and we were too i
10 specific with respect to keeping the pumps running no matter I
I 11 what.
12 On both of these items we have been in a continuing 13 dialogue with the PWR vendors, not just B&W, to currently 14 develop a more meaningful technical requirement than originally 15 came out in the bulletin.
And these dialogues are taking place 16,
in the :entext of review of bulletins for individual plants.
l l
17 i And we also meet generically with Westinghouse.
18 We have somewhat mcdified the original requirements I
19 and put in words like "if the pump is not pumping, delivering 20 force flow" as indicated by whatever, then it's not necessary 21 b to run it.
Certainly for a large LOCA, there's no necessity l
22 l to run the pump.
I i
23 !
And these criteria have been incorporated -- different a
24 j criteria have been incorporated in the new emergency procedures i
.'erei nemnen, inc. 3
-8 25 than were described in the original bulletins.
o d
1 2287 214 a.
r.te 9 59 i
11 CHAIRMAN HENDRIE:
Well, there's been a substantial b
2, mmount of analysis since that time that considerably amplifies l
3 the level of one's understanding of some of these transients.
i 4l MR. ROSS:
Right.
And things like the operator is i
5' provided with a saturation curve or steam table at the console, J
6 in effect, with a simple line, and if he's on the wrong side 7
of the line he's saturated and if he's on the right side he 's sub-8 cool, and he knows how to index the two ingredients of this i
9j curve -- pressure and temperature.
10 COMMISSIONER AHEARNE:
These are really following some 11 of the ACRS codes.
12 !
CEAIRMAN HENDRIE:
I was just going to say, these i
13 '
reflect some of those --
14 '
MR. ROSS:
In time, a subcool meter might be developed'.
15 But right now, at Oconee there's a table, a graph.
It is part 16.
of the procedure and it's taped to the control board, and it's 17!
simple to interpret.
18 (Slide.)
19 If we go to the next slide, there are some additional h
o 20 thoughts or issues, matters that could have been put in the 4
21 ;
staff SER.
Dr. Ros: toc y mentioned these a little bit ago.
i 22 These were that we could recuire some additional analysis i
23 l methods to be developed and applied for small breaks.
I've go:
24 ;
a detailed chart on that next.
~
.erei aeroners. inc. ':
25 j And we could also -- an alternative could be to just I.
i i
j 2287 215
.te 10 l'
60 1
not permit the pilot-crerated relief valve, at least for a
)
2{
short period of time, to be operable.
This can be done in 3;
either one of two ways:
you can either block an upstream valve, l
4' just shut an upstream valve, or you could raise the actuation I
i 5
pressure above the safety valve set point :ar enough so that l
6 would mean it wouldn' t work, in effect.
A little bit of detail I
7 on each of these two points will be given in the next slide.
8j (Slide.)
i I
9 Running down, the first three items discuss -- the 10 first three or four items, how we usually do business.
11 Appendix K for small breaks says develop methods, do sensitivity' I
12 !
studies, comparisens of data and so on.
Then, when you think 12 you've got an approvable evaluation model, file it with the g
I 14 staff, usually with a topical report.
15 Some utilities have done this directly.
But most 16 likely it's a vendor topical report.
We reviow the method, 17 !
issue a safety evaluation report saying that.'.t complies with l
18 ! Appendix K.
1 l
19 ;
In most instances, we seek the advice at least of the I
20 4 ACRS Subcommittee on ECCS.
Then, with an approved method, each l
21 utility would apply this to --
i 22 l CCMMISSIONER AHEARNE:
What does the third bullet 23 ;
mean?
24 '
MR. ROSS:
Sir?
These are done cenerically, not with a
trat Reoorters, Inc. !
25 g the --
4 4
2287 216 a
I ite 11 I
61 i
1 COMMISSIONER AHEARNE:
Finish two items?
1 2l MR. ROSS:
It should be first two items.
It's a I
3i beauty defect, I think.
i 4:
Then usually the utility would apply this approved t
5' method to the spectrum of breaks.
Well, the material in part, 6
D&E on small breaks, did not follow these steps.
We didn't get i
7 a detailed topical report with all sensitivity studies.
We t
3 didn' t interactover a period of months with B&W.
We didn't i
I 9, issue an SER that says it meets Appendix K.
And perhaps even 10 l if we had done all that, we would have decided we wanted more 11 smaller breaks at dif ferent locations with different hardware 12,
assumptions.
13 The choice was, should we require, as a long-term i
14,
commitmen t, this procedure to be done in order to demonstrate 15 rigorous compliance with Appendix K; should we make that 16 j requirement now.
17 l And what we did, the choice we made, starts on the i
i 18 middle of page 16, and it refers to page 8-12 and following on 1
i 19 h the recently issued staff report, NUREG-0560, and there are d
1 20 '
requirements in there on 8.4.1 on transients and 8.4.2 on 21 small breaks.
What we decided to do was to in general advise I
22 Duke Power that these two sections will be applicable to their o
1 23 station, and that we will be letting them know in the very near A
24 future exactly what this =eans.
x r.i a.oon.n. inc.
25 Now, the title of this 0560 refers to 3&W, but in I
1 2287 217
.te 12 62 l
1i fact the staff is preparing similar reports for the other PWR 2'. vendors and the same veneral concern would a Civ. to all three.
I 3
So I th' ink what we will be doing is trying to, in effect, get 1
4l our act together and say what degraded asst =ption we want en mul-l 5l tiple f ailures. or operator misfeasance and malfeasance and 6 i nonfeasance.
When we have that ready in the near future, we 7i will apply it as appropriate.
3; So that's som. of the choices we had on further l
9I analysis and explains what was the decision we made and why we 10,
made it.
I e
Il The other subject, the isolation of PORV, starts on i
I2 the next slide.
13 (Slide.)
14 There are two rather detailed slides.
We don't 15 discuss the pilot-operated relief valve role in terms of 16 reliability to any significant degree in the report, in the 17 SER.
NUREG-0560, I just reference Section 8.4.6, does make 18 reference to the valves in general and the PORV in particular l
Mhll that are connectable to the reacter coolant system.
The issue o
20 ] is, that we have looked at, in effect, could transients not I
21,feedwater transients, other pressurization transients, lead il 22 to openin_ of this valve, with a distressing high prehability ilo 23 !! such that we didn' t like the small LCCA probability.
2 '*
New, we had already done the consequences part, as
'e,as Aeoorters, Inc. '
2 c 'Iq described in the SER.
And we said the consequences, at least t
- l b
j 2287 218
.te 13 l
63 i
t i
I
-BU I!
in LOCA space, are acceptable.
But they might not be acceptable 2p for the transient that's going to occur three, four times a year.
i 3
Even though the core is un covered, you know you have had release i
l 4i of ori.Tary coolant systen -- the core remains covered, right.
i 5
But if it were a transient, every three months we 6
might want two valves to submit single-failure and high capacity; I
7; and so on.
I I
at We did require the SER, at top of page 17, that we i
l 9'
get some additional information on the mechanical reliability i
10 of relief and safety valves, some of the pros and cons that we i
l 11 l culled out on these two pages, and a little bit of supplemental.
I t
12,
information, also. If you did lift the PORV, as we saw on t
13 Three Mile, you're going to get a loss of inventory and the 14 !
existing makeup capacity would not be adequate that is in the 15 makeup mode.
And this is the classical definition of a LOCA --
16 i leakage you can't keen u.o with with your makene system.
17 l The concern of opening of the PORV and the relief 1
0 18 i valve is not unique to B&W, and some designs have larger PORVs i
19 than B&W and the makeup deficit would even be larger.
l 20 !
There have been three functional changes on the S&W 21.
machines:
lowering the reactor coolant system pressure trip, l
22 ll raising the PORV trip, and adding two secondary traps.
These
- i d
22 H should reduce the lif t probability of the PORV.
However, it i
24 :i,,still mav lift and we note this in our SER for transients that aeconen. ine. q o
4.i 25 ;I we usually analyze in the safety analysis, chapter 15.
.J
- 1 2287 219 i
i te 14 l
64 l
1, The PORV can open either by electrical f ailure or oper-i 2l ator error. There 's a switch on the console.
If you turn it to i
3i the wrong position, the valve will open.
If the valve opens 4!
it might stick open.
And we can't tell today the difference in i
3 probability between the PORV and the safety valve.
One may be i
6' much less likely to' stick open.
But we don't have that data.
I 7
We have analy:ed -- S&W has analy:ed, as described in i
8, the SER, what happens if it does stick open.
And you can i
9:
isolate it if it does.
10 l COMMISSIONER AEEARNE:
You can isolate the FORV?
Il MR. ROSS:
Yes.
The last bullet on this slide is, 12 the safety valve could stick open.
If it did fully, this would 13 be a larger break by about a factor of three or four, I believe,I 14 than the PORV, and there's no isolation capability.
The larger 15 '
break, if it's full throat open, might be better in the perfor-16 '
mance standpoint, because it would go ahead and depressurize 17 '
down and turn on the ECCS automatically.
l 13 ;
(Slide.)
19 On the last slide, there are no regulatory require-20 ments in general on the PORVs on PWRs across the board.
We 21,
don't require them to be open by tech specs; we don't require i
22 '
them to ce closed, h
23 l On those specifically for Westinghouse, the safety 24 ar.alysis is done, the chapter 15 safety analysis is done rei aeconm. ine..
25l,ignoringthePORV.
I; assumes it's not there.
If the valve jl!
2287 220 4
te 15 65 i
1 1
develops a leak, it's usually isolated.
I believe at Ocenee 2l they are isolated now.
There's a leak in at least one of the i
3, PORVs and it is isolated.
i 4l A crude estimate of the probability to stick in the 5
open position is.02 per transient.
We have had about 150 i
6 transients and we'vc had either three or four instances of it i
7' sticking open.
So that's why I say a rough estimate of.02.
1 8
B&W plants, based on data, somewhere between 30 and 9
35 reactor years, have had about 150 pressure transients, 10 transients where the pressure went on for different reasons, II including feedwater transients; averages out to about four per I
12 reactor year or it night be a little higher.
1 13 The last two are some things we had some information Id ! just in the last night.
I caution you, the last two things 15 l i
have been -- there's beert little staff probing and evaluation.
I 16 I So this is items we wouli like to investigate more.
i 17; Of these 146 oressure transients for the collection l
~
13 of the nine B&W plants, there is relatively good data on 42 of
'l 19 them.
It was prescrved, meters were running and so on.
3&W 20 looked at these 42 and said that the pressure overshcot was 1
21 " somewhere between zero and 55 pcunds.
Now, oversheet is 22 '
defined as the pressure in excess of the reactor coolant system 22 high-pressure trip set point.
That's what pressure 1
24 J overshcot is defined as.
It was 100 psi en one event.
erei mearte,s. ine..I n
2 * ]"
What we don' t know, for example, is en these 42 events i
2287 221 I
I te 16 66 l
Il which of them was the PORV open, which was it closed, and what's i
2:
the technical significance of the difference.
I think we would I
3 be more' interested in the subset where the PORV was closed, if i
4I there are any.
My speculation is we'll probably find a subset 5
where some were closed and we'll look at them harder, for this 6'
reason.
i 7l We would like to know what the probability is of 8l lifting the FORV at present.
Now, the high-pressure trips up on, l
9 is 2300 pounds.
If the overshoot can be shown to be reasonably 10 low with high probability, they calculate on some statistical i
l I
II !
basis the probability that it's 150 pounds, the overshoot, and i
12 ! that would mean on a typical expected, not a chapter 15, but an !
I3 expected transient, you can calculate the probability in the I#
PORV, comparing that with the failure probability and the I3 ;
transient arrival rate.
You've got a probability per reactor i
1 16 l year of a small LOCA due to PORV sticking open.
I7 COMMISSIONER AHEARNE:
What was the PORV set point I8 i before --
1 19 j MR. ROSS:
It was 2255, I believe.
The reactor was li 20 9 2355.
The PORV went up 200 pounds and the reactor came down 55.
'l MR. DENTON:
The cuestion we were asking ourselves, t
22 '
I think, comes down to this:
that there is a way available
,,i that we can almost guarantee that the PCRV will not be the 24 '
source of a sm'all LCCA, that is, by raising set points and v A.co m,s,inc.
2 *5 4 closing the by-valve.
However, when you do that there's a 4
i I
]
2287 222
I
.te 17 j
67 I
possibility that you will cause safety valves to 1 ft and i
2l maybe that lifting would have been prevented if the PORV had t
been ocerable.
I d
We discussed whether er not we cught to take a pcsi-i 5
tion for Duke at this time.
My position is that we should defer.
6 this further, because it does have wide applicability.
It's i
7f an interesting questien, but I don't feel like we have yet I
i 8 !
developed enough information on the pros and cens to take a I
t, i
9; firm regulatory position.
So we wanted you to knew that both i
10 l!, situations do exist, that some utilities have closed, others I
11 <
have them open, i
12,
ge s not required by the order to do it either way, I3 and I think it's an issue we will address in the near future.
I#
COMMISSIONER AHEARNE:
On that longer-term address, 15 l though, will you also examine whether it might be sensible 16 just to redesign the PORV?
1 I7 !
MR. DENTON:
That's another thought, and perhaps do I8l something different with the safety valves at the same time.
19 '
MR. RCSS:
That concludes our direct presentation.
20 i CEAIRMAN HENDRIE:
Other questions?
'l COMMISSIONER 3RADFORD:
Harold, in terms, if you i
1 3,
look ahead a year or two, if you're going to have to go through 23 l at least the different B&W plants and go through this sort of a.
J 24 a regimen of, in effect, individuali:ing even more than has
~
a wai neoonen.ine.:
2 *'.. been done by the criginal licensing process the conditions under tl 1
1 2287 223
te 13 63 t
I!j which they operate, how are you going to catch up?
That is, ll 2 l! if we reach the point where we have 70 or more plants around d
3 the country with scr:
of very individuali:ed requirements, is 4
that going to be manageable?
5 MR. DENTON:
At the moment, doing the B&W plancs and 6
the ones respending to the orders is very manpower-intensive.
7j Denny, how many people do you have working on the t
8 orders and bulletins?
9 MR. ROSS:
Abcut 35 or 40 engineers.
10 MR. DENTON:
Now, we see a big effort in the front 11 to go through the B&W orders.
We are also meeting with the 12 '
other PWR vendors, CE and Westinghouse.
I would think that i
i 13 ; before we launch into the widespread individual tailcring, we 14 '
should develop some o.csition on a c,eneric basis en items like I
I I S l' f
the PORV and small break LOCAs and see if we can' t find a i
16 simpler way to get them implemented than the sort of hand i
17 l: tailoring we are doing through the order method.
It certainly 18fisdivertingmanpowerfromcurnormalpriorities.
2-9 I9 h COMMISSIONER GILINSKY:
Harold, where does this leave J
ti 20 ' us with regard to the longer-term cuestions?
I mean, is the 21, licensee to present a schedule?
i 22 MR. DENTON:
My letter to the licensee requires thau h
!i 23 he do two things:
one is provide within the near future 241 changes to the tech specs to allow the changes that have been
- e.Jeral r"4ecorters. Inc. I 25 made to be easily enforced by I&E in terms of testing and j
2287 224 1
te 19 69 i
1' meeting the requirements.
l 7 li COMMISSIONER KENNEDY:
When you speak of near future, i
3 what are you talking about?
I i
MR. DENTON:
I think it says seven days.
Yes, it 4l 6
i 5'
requires in seven days that he propose tech specs that would 6
allow us to enforce the requirements of A through E that have r
7:
resulted, and in 30 days to provide a schedule for the long-i I
8; term modifications that he has committed to.
And I think the 9,
long term is in the order of three to four to six months --
10 CCMMISSIONER GILINSKY:
But you have some idea as to I
11 !
what a reasonable period is for those longer items?
i 12 I MR. DENTON:
Definitely, and there is evidence that 13 at Oconee that these longer-term are being aggressively pursued, 14 and the feedwater pumps are already on site in the station.
15 So I see progress even in the long-term fixes being made at 16 the plant.
17 j There is one individual who would like to speak and i
I 18 maybe ampl_fy on some of the things we have said here.
I've i
1,l, got a note.
7 l'
20 '
Zeltan, did you want to add something?
i 21 DR. ROSZTCC Y:
Yes.
The last items are items tha:
l 22 !i I had something to do with.
1 I
23 !
Denny, is it possible to get your slides brck, the 24 last two slides?
2 eral Reporters, Inc.
25 ll The last item was the question, how de we handle l
2287 225 a
.te 20 70 l
1l cases where we cannot say yet that the license is in compliance l
2 !
with Appendix K, but we have had enough in'Jormation to know 3!
that we' feel comfortable with the plant?
My role here was to 4l point out that in the past when we hsndled these cases we did 5
it differently than how it is being proposed for Oconee now.
I 6
In the past when this situation arose -- and we have quite a t
7l few examples last year in connection with the Westinghouse l
8 error on :irc-water reaction and those in connection with the 9j incompleteness with the small break spectrum last summer on I
10 the B&W plants --
II (Slide.)
12
-- then an order was issued for each of these plants 13 and they were required to submit revised analyses that shows 14 full compliance with Appendix K.
15 My understanding is this time it will be done different 16 for Oconee.
I have no problem with doing it differently.
As a 17 matter of fact, I welcome some change in this, because the old l
18 j
approach was somewhat cumbersome and maybe unduly legalistic (?).
I UbHowever, if there is a change in this, if we are instituting 20, a policy change in this, I think it would be important to spell 21 l out what is the new policy, so then we can enforce this evenly a2 j for every licensee, the same for Oconee as we are going to do I
23 i
it :or the others.
d !
But I consider this a relatively minor item.
It is d
c.- mean nosomn. Inc. ;
just more to straighten out how We do our business.
k 2287 226 n
if
tte 21 l
71 l
i I!
The second one is -- that is a much more important 2fone,thesecondone.
The basic question in the second case is 3
that du' ring this interim period, while we are resolving the question of whether it's safe to operate these plants with~~
4 5;
relief valve
, without block valve closed --
i 6!
(Slide.)
7
-- while we are evaluating that during this interim 8
period, should the valves be open or should the valves be i
9 closed, meaning the block valve behind the relief valve?
That i
10 is the basic question.
11 The staff SER that you have received does not make a 12 decision on this.
It leaves it to the licensee.
The licensee i
I3 can leave it open or.can leave it closed.
I#
After studying some of the items which were mentioned I
15 in Denny's presentation, I am convinced that the block valves 16 should be closed during operation of the plant.
And let me I7 l just emphasize some of those.
I agree with the list that I8 Denny has presented here.
I would like to emphasize only a i
19 liii few of them that I think play an important role.
20 The first point I would like to make is that most of 21 our work during the last month concentrated on how to handle a i
22 li, small break loss of coolant accident cace it already happen.
3 And if you read the staff SIR, almes: the entire, throughout
,a 24 !
i th'e entire SIR, that's the question.
A lo: less emphasis has ersimaners.inc.j
. c.
ec e: been made en the pein cf how do we prevent small breaks from g
9 U
2287 227 1
tte 22 72 1
happening.
2 The human factor plays a very important role here.
3 If we are going to see loss of coolant accidents every year 4
in the future, as we have seen them in the last 4 years, one I
i 5;
in every year in the S&W plants, sooner or later there will be i
6:
a case again when there will be a mishandled accident.
Studies h
7' that we have seen, some of them done by Pentagro(7)some of them i
8 by other people, show that human beings are expected not to do i
9 what they are supposed to do, even if they are fully aware of 10 what they are supposed to do, one out of four times, on the 11 average.
So if we see many accidents, there will be some that 12 will be mishandled.
13 We should not accept a large probability of this Is very important that we do this.
14 event.
c i
15 Now, in terms of the various general design criteria, 16 in the first item, I think there is one, number 14, that l
17 l requires special attention.
Number 14 is a very critical --
1 18 1 the whole thing is one sentence.
Basically what it says is I
i 19 l that the primary system pressure boundary should be designed
}
20 ;
to have an extremely low probability of failure.
The words l
21. " extremely low probability" are quoted from the design criteria.
i r
22 The relief valve is a portion of the primary system 23 pressure boundary.
This criteria is very important because i
24 l this could have played a very important role in the TMI case
..,.i a.conen. inc. d] and it 's very important c-25 for Oconee.
If the relief valve would ll
]
2287 228
te 23 73 l' have been reviewed as part of the TMI-2 review against this 2t criteria, then I am absolutely convinced and it's very clear l
3i that it would not have been designed as it was and TMI 2 would 4
mot have been accepted.
It clearly doesn't satisfy the S
criteria.
6!
During the review, this would have been pointed out, i
7 the reliability of the valve would have been evaluated.
It 1
I a
would not have met any standard, any possible standard.
9 Therefore, the changes we are talking about now would have been 10 made at the time that TMI-2 was licensed, rather than being 11 made today, and we would not have had a TMI-2 incident.
12 '
The same way, if we don't eliminate the relatively 13 high probability of occurrence from Oconee, then we are in the 14 same boat.
Now, certain improvements have been made and they 5
15 were corroborated by changing the set point and the new trips.
16 i Certain types of reactor transients will not lift the valve 17 any more, and that is a very significant improvement.
18 But there are many other types of failures which are 19 not addressed in the SER.
Those have to be addressed, and i
20 i the reliability of the fact, in view of all of these failure 1
21 ' modes, have to be evaluated.
If it meets the standard, whatever i
22 l it is -- something like 10 was mentioned from the instrumenta-
-3
-3 23 ! tion viewpoint, or 10
, what we use normally for small break i
24 h LOCAs, is a possible standard.
If it shows that it has been
- erei neporters. inc. il
.e 25 met, then the plants can operate with this valve on, because i
i
!a 0
2287 229
ate 24 74 1
the risk associated with it is acceptable.
2 If it cannot be shown that this valve meets this t
i 3'
requirement, then I think the block valve should be closed i
4' during operation and the plant is safe that way.
In that way, i
5 the failure of the break -- the failure of the relief valve i
6l will not produce a LOCA, and in that case the plant can operate.'
I I
i 7
So by closing the valve alone, independent of any other changes,!
i I
a; makes a huge difference, something like a factor of 100 I.
9' difference, in the probability of having a loss of coolant i
i 10 l accident on these plants.
11 And while we are evaluating the reliability of the 12 valve, and while we are specifying what is the NRC criteria 13 for the specific valve, I believe we would be a lot safer if i
14 the plant would operate with the valve closed.
15 As it turned out, most of the Oconee plants, I 16 believe two out of the three, have been operating that way 17 simply for another reason, because the valve was leaking and 18 they closed the block valve behind it.
19 'j So I do reccmmend to close this valve for the interi=
20 period, and then, when the reliability study of the valve is i
i 21 available and the judgment has been passed on it, then open it i
22 l up if it's appropriate.
i f
COMMISSIONER AEIARNE:
Do you see any arguments 23 j i
24[ against closing it?
oral Reporters, Inc. l ses 25 !
DR. ROSZTCCZY:
No, I don't see any convincing i
2287 230 1
.te 25 75 l
argument against closing it.
The valve has already been i
I 2l closed on many of these plants and it can be closed at any time.
I 3{
So whatever evaluation has been done, the plant has been 4
licensed with a closed valve operation.
5 We have also licensed plants which completely 1
6:
eliminated this valve.
They just simply don't have this valve 7.
on the system, and that is a fully accepted design.
I i
8 COMMISSIONER AHEARNE:
The valve not in the system l
t 9
or the valve not evaluated.
10 DR. ROSZTOCZY:
The valve doesn't exist.
There is l
11 no relief valve on the system.
12 MR. ROSS:
To be specific, Arkansas Nuclear 2.
13 DR. ROSZTOCZY:
Arkansas Nuclear plant is an operating i
14 plant like that, and there are approximately 20 plants in 15 various stages --
16 CHAIRMAN HENDRIE:
Is that a B&W?
l 17l DR. ROSZTOCZY:
The Combustion, standard Combustion i
18 l Engineering plant --
l 19 l
CEAIRMAN HENDRIE:
I think all the B&Ws do have the i
20 relief 7 2I '
DR. ROSZTOCZY:
All the Westinghouse and all the S&W i
22 i has it, and Combustion plants beyond a certain date do not have i.
23 ll the relief valve.
So the fact that the relief valve -- the 1
24 f 31ock valve closed behind the relief valve, or the f act tha:
- ..,.aerei n eooners. inc.
,ci there is no relief valve at all en the system, is not new to 1
2287 231 I
Ite 26 1
76 1
us.
This has been looked at many times during the past eight 2
years or more and there is no major problem.
3 Some of the items Denny mentioned here could turn out 4
to be a disadvantage or could turn out to be an advantage if i
5l the relief valve is closed.
L2t's take as an example the one ---
l 6i if you don't have the relief valve, you might -- you would open i
7 up the safety for some transients which go off on that high of I
I a
a pressure.
So the question there, is it a benefit or a penalty' l
9 of the safety opens up?
And I think the answer should come 10 from the reliability evaluation of both of those valves.
11 If the safety is the more reliable valve, then I 12 think just from that point alone it is better to close it.
If
[
l 13 it turns out to be that the relief valve is the more reliable i
14 valve than the safety, then it has a purpose.
Then it has 15 some public safety purpose.
16 Up to now it has been put there simply for convenience.
17 If the reliability shows it is better to have a relief valve i
18 l there because it's a more reliable valve, the safety is less 19 !, reliable, then I think there is a purpose for it, and in that t
i 20 : case it should be designed to the acceptable criteria and it i
21 should be in the open position.
22 The point is, we have not done the work.
We didn't 23 ; recuire this work from the applicants in the past.
We have not i
24 h done the work and we don't know yet which one is more reliable, c cus aeoonen. inc. h 25 f So my feeling is, for the time being just close it, because it l
2287 232 a
s d
I te 27 77 1
cakes a huge difference in terms of actually having a loss of 2
coolant accident.
3' CHAIRMAN HENDRIE:
Thank you, Zoltan.
f I
4 MR. DENTON:
There's a lot to be said for closing the 5l valve.
There also are some concerns that it might lead to I
6-other ones, and that's why I referred it to our lessons learned 7
group for prompt evaluation.
8 COMMISSIONER KENNEDY:
What are those concerns?
9 What do you see those concerns to be?
i 10 MR. CASE:
The concern would be the safety valve is 11 less reliable than the relief valve, and by blocking it off I.
12 you are, at least to some degree, increasing the likelihood of 13 opening the safety valve.
14 COMMISSIONER 3RADFORD:
Why should the safety valve, 15 with your safety grade equipment, be less reliable than a 16 relief valve that wasn't --
17 MR. CASE:
Your concern is reliability of receding, 18 and safety valves are designed reliably to open.
That's their 19 primary f unction, to open, and that's where they put their i
20 l money in getting reliability, f
i 21 !
COMMISSIONER 3RADFORD:
That seems, I take it, a 22 fair statement of what is the case.
But I assume that it turns t
23 !
out that receding is also an important safety feature.
!i 24 ll l
MR. DENTON:
Yes, that's right.
a
.r.i aeoorten. inc. a 25i But I don't propose -- I wouldn't, I guess, propose c,
j 2287 233
s te 28 78 I
li that the Commission try to make a decision on the basis of the i
2j preliminary information you have got today.
Maybe Zoltan would 3l advocate that you do.
To me, it is not such an urgent issue i
4l that we have to decide it in the case of Oconee, because it i
5!
has wide applicability to other PWRs.
I I
6; The staff is not of one mind about whether it's I
i 7'
really better to open or close.
We may well decide it's better l i
8 to close.
But I think we ought to make those decisions on a 9
documented review basis, and I don' t see that there 's any undue 10 risk with operating with the valve either way.
We have trained II '
the operators to respond if it does open and stays open, and i
12 it may well turn out on examination that to close it is exactly 13 the right thing to do.
But I don't think we have come to that l
14 with one mind yet, haven't had a chance to collect the data 15 and see what statistical differences it really makes.
16 CEAIR!iAN HENDRIZ:
Let's see.
There have been a 17 ll number of steps which certainly ought to reduce the rate of l
18 [ challenge to the relief valve, that is, the rate of calls for i
19 it to open, then have it close.
Now, those raising of the 1
20 ls1 set points and various other steps that have been taken 21 h do not, I recognize, eliminate all other circumstances in 22 l which the relief valve migh$ be closed upon opening (?).
v 23 But you have moved the set point up.
Remind me 1
2# L again.
Where is it?
ral Reoorters, Inc j
- e.
eS ifu MR. DENTON:
The intent was to move it up such chat t
0; 2287 234
te 29 79 i
i i
1 in the transients that occurred in the past it wouldn't be 2l tripped.
Now, we don't have, as Denny said, good data on all 3l those past transients.
I think for rhe data that you did manage 4
to get on the 40 or so cases, it would not have been tripped.
5 MR. ROSS:
The overshoot was not 15 0 po unds.
6 Unfortunately, we got the last bit of data about 6:30 this i
7 morning.
It wasn't divided into what the overshoot would have 8l been if the PORV had not been working, because the PORV would f
9 certainly trim some of the overshoot.
10 CHAIRMAN HENDRIE:
I recognize that.
11 MR. DENTON:
And if the valve did open, I think the l
i i
12 '
operators today in all these plants are sufficiently tuned i
13 to this problem they would close the block valve, and that's 14 the way to do it.
Now you can't be absolutely sure that they 15 do.
f 16 But until we can really be sure that we are adding 17 to safety and not detracting from it, I see it as a close lcallandonethat'sanincrementalchangeandnotanecessary 18 19 component of restart.
i 20 !
COMMISSIONER AEEARNE:
Harold, did you say the l
21 ;
operators in the plants today, if the PORV opened, they'd t
22 l use the block valve to close it?
I 23 '
MR. DENTON:
They'd certainly have that capability 1
24 l to close it with a block valve, I'm hoping contrary to the
- e e.ders: R econen, Inc.,
25 i safetv valve --
k 2287 235 o
e i
I 80
.te 30 1
COMMISSi.vdER AHEARNE:
It wouldn't automatically 2j close it?
l
+
s 3i MR. DENTON:
No, if they stick.
Whereas, the safety r
1 4!
valves, for some reason, if we haven't looked hard enough, if i
1 5
safety valves get open as a result of taking this out --
i 6!
COMMISSIONER KENNEDY:
You're assuming they know this 7
stuff.
8!
MR. ROSS:
This is all part of the procedures.
The l
9' new procedures cover this, tell them how to look for it, what 10 to do.
We did have one data point for Rancho Seco that was I
11 j 35 pounds overshoot, but that's just one transient.
12 MR. DENTON:
I think once we determine a firm posi-j 13 tion on that thing as a regulatory body, we can order it one f
14 way or the other, i
15 ll COMMISSIONER GILINSKY:
Zoltan is rising up behind I
16 1 you there.
17 (Laughter.)
18 You see, what I understand him to be saying is that, i
19! in the absence of data, the cautious thing to be doing is to --
20 j COMMISSIONER AHEARNE:
Well, you have some data, in 1
21 that you know the PORV is not as reliable as you would have 22 '
liked it.
i 23 COMMISSIONER GILINSKY:
Well, but the absence of 24 '
data that compares it to the safety valves and so on.
He
.,.i n conen.inc.
25 seems to he saying the cautious thing is to close them, since --
I 2287 236
I
.te 31 81 1
MR. CASE:
It's not clear to me that's correct, the 2
cautious thing is to close them.
l l
3 MR. ROSS:
That's what Zoltan was saying, that some 4
transients would lift the PORV first and no more, and if that i
5l stuck --
i 6
CHAIRMAN HENDRIE:
You've got a block valve.
i 7l MR. ROSS:
You've got a block valve.
So that's some I
I i
86 measure of comfort also.
9 MR. DENTON:
If there were block valves downstream i
10 of safety valves, for example, then you could argue why have i
11,
a PORV in there at all.
So it's the absence of the ability i
i i
12 !
to close a stuck-open relief valve, and the argument kinds of f
13 swings on.
What we need to do is look in more detail at the I
f 14 transients, collect reliability data, sort of~ statistical f
15 bounds of which way safety is best serviced.
I think people 16 come down on different sides of this issue.
17 It just isn't that clear which way is a preferable
-10 18 !
mode of operation.
t 19l COMMISSIONER KENNEDY:
Zoltan?
h 20 l CHAIRMAN HENDRIE:
Can you summarize, Zoltan?
Time 21 runs very fast.
I 22 l DR. ROSZTOCZY:
If I can cc==ent to Dr. Ahearne's 23 ! question about closing the valve, should there be a small 24 I break loss of coolant accident.
As i: turns out, these very a mai a.xn.n. inc.
I 25 ;
small breaks, like valve openings, provide similar system I
I ite 32 82 1
behavior than a reactor trip.
It's not easy to differentiate 2
between a small loss of coolant accident and any reactor trip.
3 So it will take a while before the operator can recognize that 4
he's dealing with a small break LOCA and not with another i
5 event.
i l
6j Obviously, later on he could close the valve or open 7
it up as part of the procedure, depending on how the system t
I 8'
behaved.
So it wouldn't be done right away and there would be 9
a loss of coolant accident, which produces a certain depressur-i 10 i:ation and so on, probably terminated after ten minutes or so. j i
11 :
In connection with the other question, which was, is 12 it safe to keep it closed, I feel rather strongly that it is i
13 safer to keep it closed.
If the valve is not closed, then the 14 probability of failure, even with the new design changes, even 15 if it's reduced, it still could be an order or more higher 16 than with the valve closed.
So I would strongly recommend 17 closing.
18 MR. NOVAK:
Just one point, Dr. Hendrie.
We have 19 had these discussions.
Just from the subject, you know there 20 i are a lot of things that would go into any final decision.
And 3
i 21 '
I think that's why I agree with Mr. Denton that we should 22 review them all.
23 I would like to give you j ust one example that I i
24 l considered last night, and I talked to the people from I&E.
.we neoomn. ine. ;
m 25 If we were to change our mind and suggest we should lock the l
2287 238 i
e tte 33 83 I
valve closed, I asked the I&E people, do you think this would 2
affect an instruction in the procedure to open the valve if 3j necessa'ry, and there are reasons during an event when you would l
4l want to cpen it.
And his answer was yes, that he thought an Sl, operator would be more reluctant to open the valve if he were I
i 6
previously instructed to always keep it closed.
7 So all I am saying is there are many facets to this I
I ai problem and I think they all have to be considered.
And here 9
is one that just came up in my mind last night.
I'm sure I t
10 could think of some more.
And I think our analysis supports l
Il the fact that the operator and the equipment now can handle an l
12 opening of a PORV without any core uncovery, and I think it i
13 merits the time to study the problem in more detail.
14 MR. DENTON:
I think what I'm saying is the correct I
i 15 answer isn't patently obvious, at least in the minds of the 16 staf f's experts on this one.
It came to the fore rather later i
17 in our review here of Oconee.
I would like to move promptly 18 and try to get a concerted position on it and go from there.
l9 COMMISSIONER KENNEDY:
But that meanwhile you are 20 l satisfied that the procedures that are now in place and the 1
21 understanding on the part of the operators provides a sufficien 22 margin of safety?
23 MR. DENTON:
That's right.
And it has much broader 24
.CS redff at RfCourt, Inc. lp implications than just Oconee.
I think it would bear on mos i
25 of the PWRs or all the PWRs that have this valve.
l 2287 239
Ite 34 I
84 1
CHAIRMAN HENDRIE:
I think we have had a morning of 2
ample discussion.
For myself, I end up concurring with 3!
Mr. Denton's proposed course of action, and would add to it I l
4! am sure an unnecessary admonition to move forward on the relief l t
S valve matter as well as the other, longer-term matters that are !
i l
6l at hand.
l 7
I look up and down the table.
John?
i i
8 COMMISSIONER AHEARNE:
Just for clarification, where 9'
did we end up standing on the UCS petition?
i 10 CHAIRMAN HENDRIE:
The majority of the Commission 11 ;
has agreed with the letter that you sought.
l i
12 COMMISSIONER AHEARNE:
So that petition is then I
13 denied, correct?
14 CHAIRMAN HENDRIE:
Yes, but I prefer to deal with 15 that in the context of that letter.
16 COMMISSIONER AHEARNE:
Yes.
Then I would go along 17 with Harold's position.
I 18 COMMISSIONER KENNEDY:
Yes.
19 COMMISSIONER BRADFORD:
A bit of the cart and horse i
20 l problem.
I think if one goes along with Harold's recommenda-l 21 tion, then I think the petition is -- (Inaudible.)
22 COMMISSIONER GILINSKY:
(Nods head in the affirmative.)
l l
23 I COMMISSIONER AHEARNE:
Also, since I go ahead with l
24 H
~
it, I'll just start with --
.c..s.e.r.i n. con n inc. I 25 i COMMISSIONER KENNEDY:
I proposed to deny the petition-2287 240 sl
s l
te 35 35 1
yesterday.
2 CEAIRMAN HENDRIE:
Vic?
3, COMMISSIONER GILINSKY:
Go with Harold's decision, I
4 yes.
5' CHAIRMAN HENDRIE:
Peter?
i l
l 6!
COMMISSIONER BRADFORD:
(Nods head in the affirmative.)
7 MR. BICKWIT:
May I ask a question?
l i
8 CHAIRMAN HENDRIE:
The Commission concurs.
9 MR. BICKWIT:
With relevance to how you concur, 10 will we need amendments to conform the license to the actions i
11 of the licensee which you have outlined?
12 MR. DENTON:
The amendments are not required to put 13 this into place.
The A through E can be accomplished without 14 violating any of the present technical specifications.
15 However,Ito make them enforceable in the future we will have 16 to add to the license requirements on testing and verifying 17 operability of these new pieces of equipment.
We will be 18 getting amendments to the license and we will be adding those 19h to the tech specs in the future.
Il i
20 lj MR. SICKWIT:
But until then, the licensee can 21,
operate within his license?
i 22 i MR. DENTON:
Yes.
i 23 !
FR. BICKWIT:
Let me raise some procedural options 24 l for you.
You can allow the short-term portion of the immediately 3
r3l AfDorters, Inc.
25 effective order to expire under its terms, which is really what 4
I 2287 241
.te 36 86 i
i the staff is proposing, leaving in place immediate effective-2 ness with respect to the long-term portion and leaving in 3i place the proceeding on both aspects, both the short-term and I
4 ~
long-term aspects.
That is essentially what I hear as the 5
staff recommendation and that would be my recommendation.
i i
6l But let me point out two other possible ways to 7
proceed.
8 COMMISSIONER KENNEDY:
Excuse me.
What does it say i
9 that you are leaving in place of the proceeding?
10 '
MR. BICKWIT:
When you issued your order, you i
11 initiated a proceeding, the most important element of which is i
I i
12 the opportunity for a hearing.
That opportunity still exists.
l 13 It would have -- it's hard to see the practical application 14 it would have with respect to-the short-tern portions of the I
15 order.
It's easy to see the application it would have with l
I 16 respect to the long-term portions of the order.
17 The second option would be to lift immediate effec-18 tiveness now with respect to the long-term portion of the 19 order.
The long-term portion is not the kind of thing that 20 !
you generally would impose effective ic=ediately.
- However, I
21 your actions in this entire B&W episode have been to conform 22 ! your orders to the commitments made by letter from the licensee.
23 So if you want to continue to adhere to that, you would leave i
24 i immediately effective your order with respect to the long-term
... a oen.n. me. i r
25 ;
portions.
And I would recommend that you do that.
2287 242 l
te 37 87 CEAIRMAN HINDRIE:
That simply reflects the staff --
Harold's letter.
21 3
MR. SICKWIT:
That's right.
4 The third option, which is really practically, as 5l far as I can see, equivalent to the first option, is to I
i 6{
terminate the proceeding with respect to the short-term portions I
i of the order.
The reason I see that as tantamount to the first 7
I 8
option is that it is hard to see any practical significance to I
9 the proceeding as it relates to the short-term portions of the 10 '
order at this time.
11 It seems to me that, given that those two options l
i 12 are probably the same, I would take the first option, concurring:
13 in the staff presentation from a legal standpoint.
I 14 CEAIRMAN HENDRIE:
Okay.
Will we need an order, I 15 presuae, to do that?
i 16 MR. BICKWIT:
No, you will not, because under the i
17 l terms of your order, the immediately effective portion would la expire with the staff's satisfaction --
19 CHAIRMAN HENDRIE:
Okay.
So that's simply going 20 ahead with the Commission's concurrence and the staff actions.
i 1 The order has been self-executing, as it were, and the longer-21 22 term portions simply stay in place.
23 {
MR. SICKNIT:
By simply being immediately effective.
24 i CEAIRMAN EENDRIE:
It seems to me that's a reasonable
-erai Aeeenen inc. i 25 !
configuration.
2287 243 i
i
- te 38 88 1
COMMISSIONER AHEARNE:
Yes.
2 CHAIRMAN HENDRIE:
It seems to me that's what we 3i intendsd to do here.
4 COMMISSIONER 3RADFORD:
Harold, what do you under-4 5!
stand to be the practical effect of this secuence?
Unit 1 i
6' stays on.
What about 2 and 3?
When do you anticipate --
7 MR. DENTON:
The tnit which is operating once we 8
formalize thic would be abit to continue to operate.
The l
9 second unit, which is down on_y because of the order, could 10 resume operation when they have sufficient number of operators 11 who meet the criterion.
And the third unit is down for 12 refueling and could resume operations when that is complete.
13 COMMISSIONER BRADFORD:
And when they have sufficient 14 operators.
i i
15 MR. DENTON:
And when they have sufficient operators.
16 CEAIRMAN HENDRIE:
Very good.
17 Let me sieze the moment of silence and thank you very 18 !
much.
l 1
-11 19 :
(Whereupon, at 12:00 noon, the hearing was adjourned.)
20 !,
i 21 ;
2287 244 22 -
l 23 l 24 l c....e r.i meconm. inc. I 25 l