ML20034B047

Transcript of 900413 Briefing in Rockville,Md Re Risk Based Tech Spec Program.Pp 1-56.Related Info Encl

ML20034B047
Person / Time
Issue date:	04/13/1990
From:	NRC COMMISSION (OCM)
To:
References
REF-10CFR9.7 NUDOCS 9004250262
Download: ML20034B047 (68)

Text

..,

it 8: w

(

UNITED STATES-OF AMERICA' dN NUCLEAR. REGULATORY-COMMI S S IONe N.;,.3c.

,s 4

Y r

$k6*.

BRIEFING ON RISK-BASED. TECHNICAL SPECIFICATION PROGRAM '

LOCatiOD ROCKVILLE. MARYLAND DdkO*.

APRIL 13. 1990 bd903 56 PAGES EAL R. GROSS AND CO., INC.

COU3T REPORTERS AND TRANSCRISERS 1323 ?.hode Island Avenue, Northwest Washington, D'.C.

20005 (202) 234-4433 9004250262 900413

~

PDR 10CFR PT9.7 PDC j.

l

L

~

a J.

L DISCLAIMER l=

l This is an unofficial transcript of a meeting'of the United-States Nuclear Regulatory Commission held on April 13, 1990, in the Commission's of fice at One White Flint - North, Rockville, Maryland'.-

The meeting.was open to public attendance and observation.

This transcript

~

has not been reviewed, corrected or edited, and it may i

contain inaccuracies.

The transcript is intended solely for general

'l 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 determination or beliefs.

No pleading or other paper may be filed with l-_

the Commission in any proceeding as the result of, or addressed to, any statement or argument contained herein, except as the commission may authorize.

HEAL R. GRO$$

CoUtf AttoAftt$ AND TRAN$CRl0ER$

1323 rho 0611 LAND AVfMUt, H.W.

(202) 234-4433 WA$HlHOToN, D.C.

20003

.(202) 232 6600

i'. -

1 a

4 - -

UNITED CTATES'CF, AMERICA

~

NUCLEAR REGULATORY COMMISSION-L:

1 l

l-BRIEFING ON RISK-BASED TECHNICAL

-l j

SPECIFICATION PROGRAM i

PUBLIC' MEETING.

5 Nuclear Regulatory Commission One White Flint, North Rockville,' Maryland Friday,-April 13, 1990 The Commission met in open sessio~n,-

pursuant.to notice, at 10:00-a.m., Kenneth M.

Carr, Chairman, presiding.

COMMISSIONERS PRESENT:

KENNETH M. CARR, Chairman of.the Commission THOMAS M. ROBERTS, Commissioner KENNETH C. ROGERS, Commissioner JAMES R. CURTISS, Commissioner FOR' REST J. REMICK, Commissioner i

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODC ISLAND AVENUE N W.

(202) N WASHINGTON. D.C. 20005 (202) 232 4600 1

- f-i'. -

.2.

i STAFF AND PRESENTERS SEATED AT THE COMMISSION TABLE:

t

(

WILLIAM RUSSELL, Nuclear Reactor Regulation.

DOCTOR THOMAS MURLEY,. Nuclear Reactor-Regulation

~

JAMES M. TAYLOR,.' Nuclear Reactor Regulation CHARLES E. ROSSI, Nuclear Reactor Regulation RICHARD LOBEL, Nuclear' Reactor. Regulation SAMUEL J.

CHILK, Secretary WILLIAM C.

PARLER, General. Counsel I

1 l

l l

t.

l 1

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHOOE ISLAND AVENUE. N W.

(202) 2344t33 WASHINGTON. O.C. 20006

- (202) 232 4600

4

)'

~

1 P-R-0-C-E-E-D-I-N-G-S s

2 10:00 a.m'

-(

- i 3

CHAIRMAN CARR:

Good' Jnorning, ladies - and-1 4-gentlemen.

The purpose of today's meeting is for the j

5 staff to brief the Commission on the-status of 1

6 efforts in developing risk-based technical 7

specifications for nuclear power-plants.

This; f

8 activity-is part of the-ongoing Technical" 9

Specification Improvement Program.

'The Commission 10 was last briefed on the Technical -Specification 1 11 Improvement Program in June of 1989.-

At that' 12 briefing the staff offered to provide this separate t

13 briefing on risk-based on technical specifications.-

i 14 I understand that copies of the briefing -

l 15 slides are available at the entrance to the meeting l

i 16 room.

l.

17 Do my fellow Commissioners-have any opening.

18 comments?

If not, Mr. Taylor, please proceed.

19 COMMISSIONER ROBERTS:

I'm, going out of 20 town for the weekend, and I'm leaving at-11:00.

And 21 I'll read the transcript for the ' balance of the 22 meeting, but I may not be here for the.end of.the 23 meeting.

24 CHAIRMAN CARR:

They'll be quick today.

25 MR. TAYLOR:

We promise.

We promise, t

l NEAL R. GROSS l

COURT REPORTERS AND TRANSCRIBERS l

1323 RHODE ISLAND AVENUE N W, (202) N WASHINGTON O C 20005 (202) 232 4 600 '

4

~

1-Q:od corning.

I-would liko to noto in 2

introducing the staff, J tint on this topic-the staf f j

1 3

is very much at the evaluative and exploratory stage

]

'4 of studying and utilizing-risk-based technical 5

-specifications.

To this stage, as you'll hear in 6

briefing, we are ~ seeing benefits and insights into 7

current technical specifications in~ areas such as-8 surveillance intervals and permissable or allowed 9

outage time of-equipment.

So,-we're in the early~

a 10 stages.

11 And with that thought, I will introduce 12 those at the table.

Tom Murley,. Bill Russell..To my i

13 left, Ernie Rossi and Rich Lobel, all from the Office

[

14 of Nuclear Reactor ' Regulation.

Bill Russell will' 15 commence the details of the briefing.

16 MR.

RUSSELL:

I'd like to emphasize ~ that-17 this is an area that we are giving-. senior management 18 attention to.

Doctor Murley has asked me.to pull a 19 number of activities together that relate to this.

20 We are, in fact, planning a visit, that is Tom and'I, 7

21 to Heysham in June to get firsthand information.

And

~

22 we also are looking into ' related activities.

For 5

23

example, the issues which are coming out. of. the 24 Vogtle IIT with respect to activities during 25 shutdown.

As Mr. Taylor mentioned, issuec related to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W (202) 231 4 33 WASHINGTON.DC 20005 (n2) 232-6600 i.

.I

A 5

1-clicwnblo -cuttg3 tieca cnd - h w thsco Lc03 cro uccd.-

2-We're pulling information. together.

We expect to be f.

3 in. a position to' discuss this at the senior

,jt 4

management. meeting in June and would - be prepared to 5

brief the commission on these activities at the time 6

we do an' update briefing on the Tech Spec Improvement f

-7 Program, which would - be after the' senior management '

8-meeting.

9 With that brief discussion, I'd like to 10 turn it to Rich Lobel who will' walk you through the' i

11 presentation.

12' MR.

LOBEL:

Good morning.

My name is-13 Richard Lobel and I'm in the Technical Specifications 14 Branch.

I've been asked to come.here today-'to brief j.

o 15 you on the status of the staff's work on risk-based 16 technical specifications.

17 We're in the early stages of' evaluating the

[

18 benefits that may be 'possible using-technical I

l 19 specifications that are - based on risk calculations.

l-20 At the'present time, the completion of the review of 21 the new standard technical specifications and - the l

l 22 lead plant technical specifications is the highest 23 priority effort in the Technical Specifications l

24 Branch.

Some studies have been done, however, 'and 25 I'll discuss those later exploring the use of risk-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBER $

1323 RHODE ISLAND AVENUE, N W.

(202) N WASHINGTON. D C. 20005 (202) 232-6600 I'

I t

6~

?

1-bocod tochnical opocificationo.

And cono-2 applications of risk to technical specifications have 3

already been implemented by.the staff.

4 For example, the staff has approved nine 5

topical reports on changes to the time that safety 6

related instrumentation is allowed to be inoperable 7

and to surveillance frequencies for this 8

instrumentation.

The changes proposed in these

~

9 reports were based on risk calculations.

The changes 10 reviewed and approved in these reports are available 11 to licensees as line. item improvements; that is they 12 can propose to make these changes independent of the 13 new standard technical specifications..

They're also 14 included in the new standard technical specifications 15 that are underway that are being worked on.-

16 In - addition, the staff has reviewed and 17 approved individual changes over the last several 18 years to technical' specifications in which the basis j

19 for the approval was a reduction in risk.

20 Next

slide, please.

This figure 21 illustrates the concept of controlling plant risk by 22 adjusting the amount of time that a

train or 23 component of a system can be inoperable, which is the 24 basic concept of risk-based technical specifications.

25 The amount of time that a piece of equipment is NEAL R. GROSS j

COURT REPORTERS AND TRANSCRIBERS 1323 RHOOE ISLAND AVENUE, N W (202) 234 4 433 WASHINGTON. O C. 20005 (202) 232 6

[

~7 1_

permitted to bo inspercblo beftro the eparctor hn3 to 2-take a remedial action is called the allowed - outage i

t 3

time or AOT.

The figure demonstrates __ risk as 4

function of time that the component of a system-is 5

inoperabl'a for two cases.

In the first case _on the 6

left,- the' increase in risk due to the - inoperability

- t 1

7 of the' piece of equipment is large, relatively:large.

8 There aren't any numbers on the slide.

In order to E

9 limit the total integrated risk, which is'the area-of 10 the rectangle, the allowed outage time'~ is short so 1

11 that the condition isn't' allowed to exist for a very-12 long time.

13 In the second case,.the increased risk due 14 to inoperability of the equipment is relatively small 15 so that the piece of equipment may'be-inoperabl'e for 16 a longer time and still result in the same total l

17 risk.

18 This exampic holds equally-well 1f instead 19 of talking about one piece of-equipment, the 20 simultaneous outage of several pieces of equipment 21 causes the increase in risk.

The concepts,-

22 therefore, are not limited to the inoperability of a 23 single piece of equipment as in the current technical 24 specifications.

This provides the operator with' a 25 means to assess the overall risk of plant operation j

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODEISLAND AVENUE,NW (202) 234 4433 WASHINGTON. D.C. 20005 (202) 232 C400

8 1

rathOr th2n focucing on ono cyct m et o tim 3, i

2 There are several features-of ' the risk-3 based technical specifications which make them 4

attractive for further study that are illustrated on I

5 this feat;ure.

First, risk-based calculations provide 6

a rational consistent bases for defining reaction 7

times of technical specifications depending on the 8

significance of the component.

9

Second, the operational flexibility is 10 balanced against safety in an explicit clear way 11 where safety allows the operational flexibility, 12 which is reflected in the allowed outage time, the 13 time allowed to restore an inoperable component can 14 be increased.

15 And third, as I just mentioned, the concept 16 can easily be applied to more than one component

~

17 being out'of service simultaneously.

18 And I'd like to illustrate these examples 19 further. with the next slide, please.

This figure 20 shows an example of a calculation of allowed outage 21 times using risk for four pieces of safety related 22 equipment.

Remember.the allowed outage time is-the 23 amount of time that the equipment can be inoperable.

q 24 before the operator's required to take.an action.

25 The allowed outage times calculated based on risk are

~

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W.

(202) 234-4433 WASHINGTON, D.C. 20005 (202)232 4600 j

9 I

1:

chewn in tho'figuro oc vortical baro.

Alongsida occh 2

of these is another vertical bar which represents.the 3

allowed outage time of the current. technical 4

specifications for that piece of equipment.

The i

5 risk-based allowed outage times are in black and the l

6

' strip ed bars are the current allowed outage times.

7-The_ figure compares the allowed outage 8

times calculated with risk to.those in the current 9

technical' specifications.

10 You can see in three cases the risk allowed 11 outage time is greater than the value in the current-l 12 technical specifications.

And in one case, it's 13 less.

In the current technical specifications the 14 allowed outage times of the 'three safety related 15 pumps; the motor driven emergency feed water pump, 16 turbine driven emergency feed water pump and-the low 17 pressure injection pump, all have an allowed outage I

j 18 time according to the current tech specs of 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.

19 And in the risk-based

approach, the safety 20 significance of the - pumps can be differentiated and 21 the allowed outage times can be different.

22 Two of the pumps are auxiliary feed ' water 23 pumps which provide water to the reactor to remove 24 heat when the regular source of water isn't 4

25 available.

One pump is driven by an electric motor NEAL R. GROSS j

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W 1

(202) 234-4433 WASHINGTON. D.C. 20005 (202) 232-6600

~

10 ~

+

1 cnd tho othsr 10 drivan by a turbin3.

Th3'.courco of 2

power for the pumps makes a big difference for some 3

accidents, and you can see this in the difference in 4

the allowed outage times for the two pumps in the-5 risk calculations.

6 These examples-illustrate how risk-based 7

technical specifications are capable of providing-an 8

explicit rational approach to selecting allowed

[

9 outage times.

The approach works just as well in.

10 selecting surveillance intervals, surveillance 9

11 frequency times, for a system or component.

12-COMMISSIONER REMICK:

Quastion.

On that 13 comparison, your reference to risk, was this ~ core 14 damage frequency we're using as the comparison?

15 MR. LOBEL:

Yes.

Yes.

16 The risk-based ' approach -isn't suited for 17 all technical specifications.

For ' example, some-18 process variables like pressure and temperatures are 19 not suitable to the risk approach..

Some technical 20 specifications would, therefore, remain -the same.

l I

21 They remain in the present form even if risk-based 22 technical specifications were adopted.

L 23 Next slide, please.

In 1987 NWR initiated 24 a project as part of the Technical Specification 25 l

Improvement Program to study the application of risk l

NEAL R. GROSS l

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE N W.

(202) 234-4433 WASHINGTON, D.C. 20005 (202) 232-6600

d

~ '..

11

-1_

-cenc:pto to technien1 cp cificatieno..

Tho-two 2

purposes for this program were to explore..the.

l 3

effectiveness of the current technical specifications

[

I 4

for limiting a

plant's operational risk and to 5-identify and develop alternative risk-based 6

approaches that would more effectively control plant l

7 risk.

The goal was to select one or more alternative 8

approaches for potential trial application.

This 9

study was: completed in 1988 and recommended a real 10 time risk-based approach.

11 The real time risk-based approach utilizes 12 a risk model for. the plant, which is programmed.in 13 such a way that it's interactive with'the user.

That 14 is, a person can input a plant specific equipment 15 configuration and in a matter of only a few minutes 16 he can obtain an estimate of the changing risk-over l.

1 l.

17 the baseline risk value.

Because the. system's 18 interactive in real time, it can-be placed. in the 19 control room to directly advise operators of the 20 plant risk of operation in the current configuration.

21 The real time risk model can use the. most.

22 comprehensive risk model available.

It's not 23 necessary to use a smaller, less complete version.

r 24 Increases in plant risk are directly linked 25 to the unavailability of plant equipment.

The effect NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE N.W l

(202) 2M WASHINGTON. D.C. 20006 (202) 232 4600 r

12

/

1 cn rick of nobilo ' cquipr nt cutzg33 czn b3 oscily 2-determined.

g l

^

3 It's important to note that the user

)

4 doesn't need to have any knowledge of risk 5

assessment.

The instructions issued to the user from-6 the computer will be in terms of specific actions for 7

specific

systems, not core melt frequencies 'or 8

r.imilar numbers..

9 COMMISSIONER. REMICK:

The risk model that I

10 you would use in this, would'this be a level one PRA 11 or is there another model that you would use?

12 MR. LOBEL:

It. would be based on a level 13 one PRA with some revisions to the software to make 14 it work in. real ' time, make it interactive in real 15 time.

But it would be based on a level one PRA.

16

-Now, this is preliminary.

We haven't made 17 any decision that it would be based on'one level or 18 another.

But that's the-thinking right now.

19 COMMISSIONER REMICK:

Right.

But your 20 reference to risk model, you were using level one?

,21 MR. LOBEL:

That's right.

If you limit it l

22 to a

level one,.you might limit the amount of 23 technical specifications that it could be applied to.

24 COMMISSIONER REMICK:

That was -another 25 question I had, what about containment systems.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) 234 4 WASHINGTON, D C. 20005 (202) 232-(400 ~

l l

.I 7

.13 i-l' MR. ICBELif Right.

Right..

2 The 1987 study', identified. a primary. issue

,i 3

for further investigation as' the practicality. of=

l 4'

-implementation.

A study ~ was initiated -in 1988 to 5

continue and expand this workfand in particular'to-3 6

address the implementation" issues.

It was to study.

I t

7 the feasibility of a pilot program in which the real'

-i 1

8 time system would be set up at-a. nuclear power plant 9

to be run; in parallel. with the existing - technical.

10 specifications.

The existing

. technical i

11 specifications would, of course, still'be governing 12 but.the real time system would be used to find' 13 situations in which even though the technical i

14 specifications were -satisfied, the risk 'was.much 15 higher than the ' baseline value.

And - also to find 16 situations in which plant action was' required.by~the 17 technical specifications even though the effect on 18 risk might be small.

19 The characteristics'that such:a system must 20 have would be identified and the effectiveness of-1

[

21 such a system would be-studied by-looking-for cases 22 where the technical specifications would give actions 23 that would be different than the results of this risk 24 calculation.

25 And a

. final goal of this

effort, an-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l

1323 RHOOE ISLAND AVENUE, N W (202) N WASHINGTON. D C. 20005 (202) 232 4600

j' 14 1

inp::rtant goal, wno to obtcin. industry participntien i-2 and' input.

.A few members of the utility industry who 3

we knew were interested in the concept became-members.

{

4.

of a working group which was established to study the

[

1 5

issue.

The utility members used their plant risk.

j 6

models ~1n-'this program: to evaluate actual-plant 7

situations for which the real time system would'have 8

been helpful.

There may very well be other' utilities

];

9 who have similar programs and a-future goal of our 10 work is to-try to identify and broaden the industry 11 involvement.

12 The 1988 feasibility study is now complete.

1 13 The-results of this study show. that it. would be 14 feasible to install - a pilot risk-based system at a 15 commercial nuclear power plant.

No: technical ~,

16 economic or institutional; issues were ' identified =

17 which would preclude the installation of the pilot:

l l-18 real time system.

The system would enable the NRC l

l 19 and the industry to assess - the chara'cteristics and i

20 effectiveness of the risk-based approach to technical 21 specifications on a pilot basis.

22 Next-slide, please.

Now. that I've 23 described a little bit of the work the NRC staff has 24 done,.I'd like to talk about and describe some of'the 25 work that the U.S.

industry has -done-and then some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W.

(202) 234 4 433 WASHINGTON, D C. 20005 '

~(202) 232 6600

.e

?.

15 1

fCreign offcrt3 in thio OrO3.

2 So far I've discussed applying risk to 3

technical specifications requirements.

Even with the 4

current technical specifications, the risk concept 5

can be applied to the management of equipment 6

outages.

For example, for performing testing and 7

maintenance in a way that minimizes risk.

Both 8

Philadelphia Electric and Southern California Edison 9

have programs that use risk analysis as a guidance in 10 the scheduling of equipment outages so that the 11 increase in risk is minimized during maintenance and 12 testing.

These are voluntarily efforts by these 13 utilities to operate in the manner that minimizes 14 risk.

15 In both

cases, risk engineers at the 16 utility have input into the scheduling of equipment 17 outages.

The technical specifications are always 18 governing and are only supplemented by these 19 programs.

20 Pacific Gas and Electric Company is the 21 program manager for an EPRI program, which is a three 22 tiered effort.

The purpose of this program is to 23 develop an approach, methods and a computer based.

24 system for implementing risk-based technical 25 specifications.

The first two tiers of the program NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS i

1323 RHODE ISLAND AVENUE.N W.

i (202) 2344 433 WASHINGTON.D C 2M W2) 2324600

16 1

would involvo chcngos to oxicting tochnicci 2

specifications which can be justified in terms of 3

risk.

The third tier would be equivalent to what I've been calling a real time risk-based system.

In 4

5 the EPRI pro]ect it's called the interactive risk 6

advisor.

It would be used by the operator to limit 7

the risk of operation with compenents inoperable, f

8 The description of this process by the.

l 9

participants stresses that " integral to this process 10 is the interaction with the efforts of the NRC 11 staff," and that also, " considerable interaction is 3

12 expected with the NRC staff."

13 We hope to use this interaction as a method 14 of focusing on detailed implementation' issues related 15 to the acceptability of risk-based tech specs from 16 the licensing or regulatory viewpoint.

Since this is 17 an industry

effort, our participation depends 18 entirely on the progress made on the project by the 19 industry group and the extent that they invite our 20 participation.

21 Next side, please.

Let me now discuss two 22 significant foreign efforts related to risk-based 23 technical specifications.

We've already inentioned 24 Heysham. Heysham is an advanced gas cooled reactor in 25 England.

It uses a computerized risk-based system in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHDDE ISLAND AVENVE. N W

(?O2) 234 4433 WASHINGTON, D C 20005 (202) 232 0600 P

]

l 1

17 1

picc3 of tr:diticnal t:chnical cpe3ifientirn3 fcr i

2 many of the heat removal and essential electrical 3

systems.

The system is called the Essential Systems 4

Status Monitor or ESSM.

f 5

The British recognize that their l

6 deterministic technical specifications with written 7

fixed rules were very conservative and were limiting 8

the availability of British gas cooled reactors.

So 9

work was begun in the early 1980s on a risk-based 10 system to replace the traditional technical 11 specifications.

The system that was developed was 12 installed at Heysham 2 in 1987.

The results of the 13 ESSM probabilistic assessments are displayed-to the 14 operators as so-called maintenance categories.

Each 15 maintenance category corresponds to a

range of l

t 16 probabilities of core damage.

The operator's actions 17 depend on which maintenance category the plan is in.

18 In addition, there are deterministic rules 19 that they call " backstop rules" which must always be 20 satisfied.

If they're not satisfied, the reactor 21 must shutdown.

22 off-line modules of the Heysham 2 ESSM'can 23 be accessed outside the central control room by.

24 personnel, planning maintenance and testing.

This 25 allows maintenance personnel to assess the effect of NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS 1323 RHODE ISLAND AVENUE. N W.

(202) 234 4 433 WASHINGTON. O C P3005 (202) 232 4 600

l L

19 1

regultt:ry iccues which h;va be:n id:ntified' fcr 2

risk-based technical specifications.

I've listed 3

them on the viewgraph and I'd like to go through them 4

briefly.

5 We looked at the regulation that governs i

6 technical specifications to see if the use of risk-7 based technical specifications would be consistent 8

with the regulation.

Section 50.36 of the Code of 9

Federal Regulations requires each licensee to have 10 technical specifications for the facility and 11 specifies the contents of the technical 12 specifications.

It also specifies that the technical 13 specifications will be derived from analyses included l

14 in the safety analysis report.

These are design i

15 basis calculations.

16 The risk-based approach may not meet this 17 requirement since risk calculations go beyond design 18 basis calculations.

And we have to investigate this 19 further.

i 20 The current safety analyses, which are the-l

.21 basis of the current technical specifications, use 22 criteria such as fuel rod temperatures and reactor i

23 coolant system pressures for success / failure 24 criteria.

These criteria generally contain a large 25 margin of safety.

Risk calculations are, in general, NEAL R. GROSS

. COURT REPORTER $ AND TRANSCRIBERS 1

1323 RHODE ISLAND AVENUE, N W (202) 234 4433 WASHINGTON.DC 20005 (202)232 6600

20 1

C ro be;t CCtic ta.

Th rofcra, th3 new rick rolCt;d 1

2 criteria suitable for risk analyses must be developed 3

and used for risk-based technical specifications.

l 4

These would probably be similar to criteria t

5 already used in probabilistic risk analyses and so 6

this shouldn't be a major effort.

t 7

There should be regulatory requirements for 8

the criteria for operator action.

In the British f

i 9

ESSM, the actions an operator takes are, in a very 10 general

way, determined by the magnitude of the l

4 11 increase in risk.

If the risk increased due to a 12 change in the status of a system is small enough, the 13 operator may not have to take any actions at all.

On 14 the other hand, for a situation that has a relatively 15 high risk associated with it, the operator has to 16 take prompt action.

The NRC would have to develop or 17 approve similar criteria for use in a U.S.

system.

l

\\

18 From a regulatory viewpoint, configuration l

i 19 control is important.

A risk calculational model 20 used for technical specifications would probably not 21 be allowed to be changed without NRC approval.

On 22 the other hand, the models must be updated as the 23 design of the plan is modified to assure that the i

24 risk-based system accurately models the plan.

25 Therefore, a system of controls to satisfy these NEAL R. GROSS COVGT REPORTERS ANDTRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

l (202) 234 4 33 WASHINGTON. D C 20005 (M2) 232-6600

l t

l l

18 1

plCnned cuttgC3 en th3 plCnt rick.

Wh0t-if Cc:n2riC3 1

2 can be run and optimum maintenance activities can be I

3 planned.

1 4

ESSM also gives the operator a choice of

(

I 5

what actions to take to reduce risk in a given 6

situation by identifying options that the operator l

7 can take to restore which piece of equipment he 8

should restore to operable status first.

9 The Nordic countries also have a program to j

10 apply risk techniques to technical specifications

.f 11 issues.

One aspect of their work, which is of

[

12 particular interest to us, is a study they're doing 13 of the risk of shutting down a reactor.

Because so 14 many systems have to work properly to shutdown a

~

15 reactor; to stop the fission reaction and reduce the 16 temperature and pressure, shutting down a reactor 17 inherently has some risk associated with it.

In 18 order to effectively optimize the risk reduction, the 19 risk of continued operation in a given configuration 20 must be balanced against. the risk of shutting the 21 reactor down.

That is, the Nordic countries are 22 asking is it safer to let the reactor run or to shut l

23 it down with certain equipment inoperable?

We're 24 following this work closely.

25 Next slide, please.

There are several 4

NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS 1323 RHODE ISLAND AVENUE N W (202) 234 4433 WASHINGTON. D C 20006 (202) 232-t.600

't

f L

21

. C:mewh;t c nflicting 90010 h33 to be d;v01cped.

i 2

Access to the computer model used by the 3

utility must be controlled by the utility in order to 4

assure that the risk model approved by the-NRC cannot 4

5 be changed except in an authorized way.

The British i

6 go to some trouble to assure this with ESSM.

7 The technical adequacy of the risk analyses 8

is another important issue and the staff will want to 9

have some input.

The risk model itself

must, j

10 obviously, be accurate and complete.

As many systems 11 interactions as possible should be modeled.

Human 12 factors considerations must also be modeled.

The 13 failure data that are used must be applicable to the 14 specific plant. And uncertainty in the risk analysis 15 must also be included.

l 16 A final thought which I know is not just a l

17 regulatory issue but also a concern to utilities, is 18 that the risk advisor must not be used by the 19 operators as a black box.-

That is, it must not 1

20 exclude operator judgment and assessment of the plant j.

21 status.

No computer program ever models reality t

l 22 perfectly, so the operator's judgment is always 23 necessary.

Any system of the type which is actually 24 used in place of written technical specifications has 25 to take this into account.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) D4 4433 WASHINGTON, D C 20005 (202) 232 4 600 i

i

L 22 l

1 N xt clida, plc303.

Fin 311y, W3 C00 th3 2

role for the NRC staff in risk-based technical 3

specifications in the near future as rather limited j

4 due to our present available resources and priority i

5 of other technical specifications work.

We will, 6

essentially, monitor industry programs and foreign 7

efforts.

We will participate in discussions with the l

8 industry to provide a

regulatory prospective to e

9 industry work on risk-based technical specifications.

e 10 I believe the industry expects and welcomes this.

l 11 Most present industry effort' appears to be 12 directed to outage and maintenance planning, which 13 doesn't require direct NRC involvement.

That is, i

14 prior review and approval.

Therefore, no NRC action 15 in the near future is necessary.

16 It appears that our efforts to date have i

17 been successful in encouraging the ' industry in the 18 direction of risk-based operating rules and we hope r

19 to continue to encourage this work in the future, 20 since it appears to hold great promise for increasing 21 safety and operational flexibility.

22 That ends my presentation.

Are there any 23 questions?

24 MR. TAYLOR:

Tom Murley, I believe, would 25 like to --

NEAL R. GROSS COURT REPORTERS ANDTRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W.

(202) 234 4433 WASHINGTON, D C 20005 (202) 232-6600

23 l

[

1 DOCTOR MURLEY:

YOo.

I w;nt to cdd a bit.

l 2

It had. occurred to me during Richard's presentation I

3 that there is an application for future reactoro, and

}

4 I think Commissioner Rogers asked this at an earlier 5

briefing, where we will be more proactive.

Although

{

6 we can use this risk-based procedures for operating i

7

reactors, I think it's best done in the initial j

8 design stage when we know what's in the designer's 9

mind; why he designed the system the way he did.

And 10 so for that reason I

think this will have its 11 probably near term application in the U.S.,

at least, 12 in the evolutionary reactor designs.

13 We have notified the designers, in this it would be GE and Combustion Engineering, that 14 case

+

15 we expect them to develop a reliability assurance 16 program, that we haven't defined it in detail other 17 than a fairly highly level of extraction which means I

+

18 that we expect the safety that's designed into the l

19 plant at day one to be maintained for the full life i

20 of the plant.

And we're going to do that through the l

21 use of tech specs, through the use of surveillance i

t j

22 procedures and in-service inspection, in-service i

23 testing and maintenance program, and that sort of 24 thing.

We haven't got into the details of specifying 25 what we want because we're kind of in an iterative NEAL R. GROSS COURT REPORTERS ANDTRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W (202)234 4433 WASHINGTON DC 20005 (202) 232 4600

l 24

_1 Ctiga now with th3 d:Oi'~ ntra.

But I think prob bly

!l i

2 the best application in the near term will be for

)

3 these evolutionary plants.

l 4

I just wanted to mention that because it 5

dawned on me during the presentation that we hadn't j]

6 mentioned that in the briefing.

7 CHAIRMAN CARRt Commissioner?

i 8

COMMISSIONER REMICK

Tom, you said 9

evolutionary, but I assume you're not ruling out a

10 advanced.

11 DOCTOR MURLEY:

No.

But we're further away 12 on the passive and advanced, I think, than we are in i

13 the evolutionary.

14 COMMISSIONER REMICK2 How about the case of 15 the U.K., their ESSM, are they attempting to model 16 human factor considerations into their risk model?

f-i 17 MR. LOBEL:

They are to a certain extent, 18 yes.

19 COMMISSIONER REMICK To the extent we do 20 or do you know?

j

.21 MR.

LOBEL:

I'm not familiar with the i

22 details.

I can't answer that.

23 COMMISSIONER REMICK As I understand, you 24 look at the change in the risk or core damage 25 frequency, in this case.

What do-they use as the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBER $

1323 RHODE ISLAND AVENUE, N W (202) 734 4 33 WA$ <lNGTON, O C. 20006

(?p2) 232-(400

I.

25 l

l 1

baco, tho tetuni calculcticn bared en tho 1cv01 cn3 i

l l

2 PRA as a reference or base point?

j g

i l

3 MR. LOBEL:

Yes, they use a base model that l

\\

4 considers all the equipment to be available.

The I

j l

5 difference would be from a model that would be done i

6 for predictive calculations that they assume all the 7

equipment to be available for their baseline value.

8 And then as equipment is taken out, that gives them 9

their increase in risk.

4 10 COMMISSIONER REMICK:

Well, that leads into 11 the next question I had; available for what?

Based i

12 on the example given in the report, it looks like 13 you're using a full year, something like 8,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> 1

14 or something like that.

Is that what they do?

They 15 assume that in the ideal case the equipment is i

16 available the entire year?

Is it 365 days or they 17 assume some kind of a availability factor?

J 18 MR. LOBEL:

Well, there's two things that 19 they do for their basic model, and that would just be 20 an instantaneous model.

But then they also have a 21 goal that they

meet, I

think they call it a

22 cumulative outage factor, which they trend equipment j

i 23 outages and the increase in risk and they compare 24 that with a number that was the basis for their PRA.

25 And that number, theoretically, should come out close NEAL R. GROSS COURT REPORTERS AND TRANSCR$ER$

1323 RHODE ISLAND AVENUE N W (202) tMe WASHINGTON. O C 20005 (202) 232 6000 i

l

~ !,

26 1

to th3 number th-t w33 cccused in th3ir PRA.

'And 2

when we visited them and talked about this, their l

3 number was fairly close for the first year of i

4 operation.

l 5

So in their base model they don't look at 6

cumulative, they look at instantaneous risk 7

calculations.

But then they have a way of trending l

8 so they can look back and see how well they did.

And i

i 9

they can do this over a quarter or over a year.

I 10 believe the data they showed us was for both year and 11 a quarter.

12 COMMISSIONER REMICKt You may have answered i

13 my question.

Maybe I didn't understand, but 1A you were basically saying that the area of the l

15 rectangle, the change in risk kinds of time must be 16 the same.

17 MR. LOBEL:

Yes.

18 COMMISSIONER REMICK:

And in the base case 19 what time are they assuming?

20 MR. LOBEL:

In the base case you don't have i

21 any equipment out of service.

22 COMMISSIONER REMICK:

Right.

23 MR. 14 BEL:

So you don't have any. allowed 24 outage times.

All the equipment is assumed to be 25 available and that gives you the lowest risk number

.NEAL R. GROSS COURT REPORTERS AND TRANSCR$ER$

l 1%*3 FNODE ISLAND AVENU$. N W.

3' (202) 2 % 4433 WASM1NG10N. D C. P3305 (202) 232 4 600 l

1 bec;uno 011 y ur cquipment 10 th:ro to v000to Cny 2

event.

l f

3 COMMISSIONER REMICK:

Right.

4 MR. LOBEL:

Then when equipment is taken i

5 out of service, the operator enters that into his 6

module in the control room and the answer he gets 7

back is a maintenance category which has done that 8

calculation.

My example is much simplified, but it's l

9 essentially calculated the area of the rectangle for 10 him.

And the number he sees is a maintenance 11

category, normal maintenance, urgent maintenance, 12 whatever the category.

l 13 COMMISSIONER-REMICK I see.

I see.

14 What kind of reaction have you.gotten so

]

15 far from

industry, the ones who volunteered to 16 participate?

Have you had enough experience to get 17 any kind of reaction yet?

j 18 MR. LOBEL:

I think they're interested.

I l

19 don't want to overstate their ' interest.

I think j

20 they're interested, they want to proceed.

Like I l

21 said in the talk, some of them are using risk in l

22 other ways other than this risk calculator in the i

l 23 control room.

l=

24 I

think they're understandably cautious 25 about making something like this available to their i

NEAL R. GROSS 1

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE N W.

J (202> 234 4433 WASHINGTON. O c 20005 (2 23 232.ss00 e

a

i I,

28 1

cperatero.

I'c n:t cura thSt I'vc h00rd a wholo 1ct l'

2-of talk that they're enthusiastic about doing that at 3

this point.

That would be one of the evolutionary 4

l 4

things that would have to come from this.

But I t

5 think they're definitely interested.

6 They're interested in ways that they can 7

use this as guidance for minimizing risk, whether-8 it's in the control room or not in the control room..

9 Most of the advantages of this can be had without 10 putting it in the control room because it's not an 11 emergency thing the operator needs to respond to an 12 emergency with.

He can do this in some reasonable i

13 time.

He can go to another room or call another 14 group and ask them, you know,

say, "I'm in this 15 configuration.

What should I do about it?"

You 16

know, "What's the advice I should get from this 17 system?"

18 COMMISSIONER CURTISS:

You said they were 19 reluctant to give it to their operators.

Actually, 20 one of the questions that I've had, this system gives 21 the operators a great deal more latitude and judgment 22 in exercising their day-to-day decisions.

Have you 23 gotten a feel from the pilot plants that pursued it 24 whether the operators liked the system, liked the 25 judgment that they get or prefer the black and white NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1.T3 RHODE ISLAND AVENUE, N W (202) 234 4433 WA$HINGTON, D C 20005 (202) 232 6600

E 29 1

n turo of tho currcnt tcch cpeca"/

+

2 MR. LOBEL:

I haven't talked to operations 3

people.

Really, I think it's other people at the 4

utilities that we've been talking to.

I'm not sure I 5

could give a real good answer to what the operators

{

l 6

themselves think about it.

i 7

CHAIRMAN CARR:

I doesn't sound like it's i

8 that far a3cng.

I don't think it is.

I 9

MR. LOBEL:

No.

10 DOCTOR MURLEY:

My understanding from my 11 staff who have been to Heysham, and maybe some of you 12 have heard the same thing, that talking with the 13 plant manager and the operators there, they do find i

14 it useful and they understand it, and they dc 15 appreciate the latitude that it gives th'em.

It gives 16 them better understanding of outage times.

17 Bill mentioned that he and I and a couple l

18 of other on our staff will be there in June and we l

19 intend to pursue just these kinds of things.

20 MR. LOBEL:

When we talked to the British 21 when I was there, they were very enthusiastic about l.

22 it and their operators did like it and glad to have l.

l 23 the flexibility.

The planning people, also, were 24 very happy to have this system.

It helped them to do 25 their planning in a more rational, thought.out way.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1

1323 RHODE ISLAND AVENUE, N W.

(202) 234 4433 WASHINGTON, D C. 20005 (202) 232 6600 L

I

(

30 1

Th y c:uld plcn cut gno in a w:y whcro th y c uld I

I-2 still get all their work done, but minimize the risk 1

t 3

to the plant and public.

4 COMMISSIONER REMICK:

I assume that since 5

tech specs are a part of design certification that I

6 none of the evolutionary plant design certification 7

applications in-house is proposing anything like this i

8 yet?

9 DOCTOR MURLEY:

No, they're not.

But we 3

10 haven't gotten to the stage of talking with them what 11 we want and what they're going to propose in terms of i

12 this reliability assurance program.

We've told them 13 that we do want to see the tech specs as part of the 14 certification, but also the maintenance program, the 15 surveillance program, the in-service inspection and 16 testing.

t 17 I view that whole collection of programs as 18 essentially a reliability assurance program.

And so 19 risk-based tech specs may very well be the most 20 logical way to go.

I think we're going to promote

,21 that.

The only trouble is I don't know how advanced 22 and how comfortable the designers feel with that 23 approach.

The logic is inescapable, it seems to me.

24 COMMISSIONER REMICK:

I agree.

I did not 25 know about this program, but I found it extremely NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W (N2) 234-4433 WASHINGTON, D.C. 200%

(202) 232- % 00

~

=n t

i l,

31 1

intCrOOting to rOnd cbrut.

i 2

That's all, Mr. Chairman.

3 CHAIRMAN CARR:

Commissioner Rogers?

{

4 COMMISSIONER ROGERS:

Well, I think it's a

[

5 very interesting program.

I was at Heyshem.

I could t

6 just say that I looked at some of my notes from that 7

visit and, just very briefly without extending. this l

8 meeting, that model gives operators a very rapid set 9

of answers and they're told whether a deterministic 10 criteria would require immediate remedial action or 11 whether a

probabilistic criteria would require 12 immediate remedial action, so it does both.

And 13 applying probabilistic criteria whether there would 14 be a plant reconfiguration warranted within 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> 15 or whether no short term action is required, so they 16 get out of that.

17 I know they find it very useful and I think i

18 it's used ea a day-by-day basis in running Heysham as j

l l.

19 part of their whole style of operation.

l 20 I had a couple of questions.

How does the 21 accident sequence precursor study, how would that be 22 used in this pilot program?

Do you know in detail 23 how that would be used?

1 24 MR. LOBEL:

I don't think we've gotten to j

25 the point where anything has been thought out as to NEAL R. GROSS COURT REPORTERS AND TRANSORIBERS

)

1323 RHODE ISLAND AVENUE, N W l

(202) 2344433 WASHINGTON. D C 20005 (202) 232-6600 1

i 32 i

1 d;toilo ct thic time.

l l

2 COMMISSIONER ROGERS:

Well, I was just 3

looking at the report, this report that just came out 4

on March 29th from Science Applications, 91033, and i

5 they mentioned in there that that's part of the 6

process analysis, and I just was a little curious as i

7 to how that worked.

l 8

MR. LOBEL:

Well, one thing you could do 9

with this before you ever got to a risk-based system 10 is you can get a lot of good information as a plant 11 operator on how best to operate the minimized risk 12 without having any system like this based on PRA 13 studies that are being done now.

And I think that 14 would factor into operations at this point.

i 15 The system that I was talking about was 16 kind of the ultimate system.

It would be an 17 interactive system right in the control room and, 18 like I said before, you really don't need to go that 19 far to get a lot of benefits from this.

And you 20 could back off even from having a computer system and-21 just have a set of rules that were based on your 22 knowledge of the results of your plan t PRA.

23 COMMISSIONER ROGERS:

Well, is there any f

24 thinking here in using this approach other than 25 ordinary operations?

For example, we know that the NEAL R. GROSS COURT REPORTERS AND TRANSOR$ERS 1323 RHODE ISLAND AVENUE. N W (202) g34 4433 WASHINGTON.DC 20005 (202) 232-6600

I k

33 1

outogo ticoo oro onoo thot hovo cortoin f

{

2 vulnerabilities.

We're just discovering that,.you j

3 know, we see it keep popping up at Calvert Cliffs and.

4 Vogtle and things that one has to be worried about l

5 during mid-loop activities.

I wonder if this might l

6 not be a good place to think about applying this, l

7 because that's something that people seem to have, i

i 8

more or less, relaxed on that when the reactor's 9

down, it's safe.

Well, it isn't necessarily safe.

I i

10 mean, it depends on what you do.

And this is one 11 area of application, I think, that might be well 12 worth looking at because it seems as if it's an area 13 ripe for fresh attention, let's put it that way.

14 MR. LOBEL:

Well, we're trying to include 15 the lessons we learned from these things in the new 1

16 standard technical specifications.

And to the extent 17 that lessons come up that are important

enough, 18 they'll be factored into guidance to operating i

19

plants, too.

But we're trying to take the 20 information that we get.

Our probability Risk

]

21 Branch has been looking at things related to 1

22 technical specifications and we talked with them and 23 try to factor all this into the new standard 24 technical specifications.

I l

25 DOCTOR MURLEY:

There's something to add to j

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W (202) 234 4433 WASHINGTON. O C 20005 -

(202) 232W40

34 1

th:t.

Th3 CCae tOchniqu 3 th3t yCu w;uld u23 to l

i 2

display risk in real time, say, to an operator are i

t

)

t 3

the same techniques that we could use to develop j

4 allowable outage times and surveillance intervals and j

5 that sort of thing.

In fact, that has been done to

}

l 6

some extent in the past.

And we will be using those.

l 7

Now, you're quite right that we still find l

8 things, I guess, we didn't fully appreciate it in the i

9 tech specs.

There is this fooling that - once you're 10 in mode 5 that things are all right and there is a 11 relaxation at the plant.

That's become probably.my 12 number one or two concern these days.

13 You recall in mid-loop operation we sent 14 out not only a generic letter to each licensee, but I 15 personally sent a copy to each operator and told them 16 of the concerns of shutdown operation.

17 Bill Russell is going to be looking at this 18 area.

He's going to pull together all these issues, 19 because Vogtle has taught us some things, other 20 operations, other events have taught us some things.

21 And we're going to be looking at this whole question 22 of shutdown safety and whether we need to look at our 23 tech specs.

i 24 You also mentioned the accident sequence 25 precursor program.

That generally is a backward NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS I

1323 RHODE ISLAND AVENUE, N W.

(202) 2MM33 WASHtNGTON. D C. 20005 (202) 2324f 00

35 e

1 looking progrO3.

Th:t 10, it look3 Ct th3 Cxpericnce 2

in the past and then asks what was the risk at the 3

time.

It's an important program and I've asked the i

4 staff with AEOD to pull together a briefing.

I think 5

we're still probably a month or two away from that, l

t 6

but we'll come down to the Commission with a briefing 7

on that because it's a very important program.

8 COMMISSIONER ROGERS:

Mell, I understand' 9

from the contractor's report that this is an j

10 important part of the pilot study, that part of the 11 actual conduct of that _ pilot program would involve 12 comparisons using that information the --

13 DOCTOR MURLEY:

Yes.

They could use actual 14 data to test how --

15 COMMISSIONER ROGERS:

Right.

Right.

16 DOCTOR MURLEY:

I see.

Yes.

17 COMMISSIONER ROGERS:

The statement here is 18 that you look at-the plant configuration change-data l

l 19 and these are then used by each utility in their 1

20 plant specific PRA to calculate the corresponding 21 changes in plant core melt frequency.

This-information, combined with an analysis of the results 22 L

23 of the accident sequence precursor study form the 24 basis for some insights about the effected _ current 25 tech specs on plant operational risk.

So we've got l

NEAL R. GROSS I

COURT REPORTERS AND TRANSCRIBERS l

1323 RHDDE ISLAND AVENUE N W tre r3m33 wASHinctoN.oe ras (m> me,m

. - - =

j 36 j

t 1

it integrated in.th ro cnd I wao ju;t curicu3 oc to i

i 2

how they were doing that.

3 I would think that this multiple component 4

out of service thing is very irportant and how one 5

chooses those.

You know, how many layers down?

Two, t

6 three?

How pany components would it consider might

[

7 be out at any one time in one of these studies.

And 8

the application of it to the outage periods would 9

seem to me would be well worth putting some real 10 effort into it.

11 The costs that are suggested in this 12 contractor study are not negligible, but they are not 13 ovetvhelming if those estimates are correct.

I was 14 wondering if you'd had any thought about whether the 15 cost estimates of implementing this, both by the 16 licensees and the NRC particularly using the real 17 time approach, are at all realistic?

18 MR. LOBEL:

Well, we've just gotten that 19 report and, hopefully, we'll get some comments from 20 the people in the industry and they're the ones that 21 can best tell us how realistic those costs are.

22 COMMISSIONER ROGERS:

Oh, for them.

But i

23 for also for NRC?

1 24 MR. LOBEL:

For us?

i 25 COMMISSIONER ROGERS:

Yes, also for you?

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 133 RHODE ISLAND AVENUE. N W, (M2) 2344433 WASHINGTON.DC 2000$

(202) M2 M00

I, 37 1

-MR.

14 BEL:

My opinicn, I think th0y'ra i

2 probably very good estimates as far as the NRC costs.

[

1 3

COMMISSIONER ROGERS:

Well, I'd just like j

f 4

to say that, you know, my own personal view is that 5

this is a very important program and it's well worth 6

looking at very hard.

It is an approach that there i

e j

7 are a number of questions about and, you know, we 8

have to probe those.

But at first blush it looks 9

very promising.

And certainly the experience of the 10 British and others suggest that it is a practical 11 tool that can be used.

12 CHAIRMAN CARR:

Commissioner Curtiss?

13 COMMISSIONER CURTISS:

Just let me add one 14

voice, I

guess, to what was said earlier by 15 Commissioner Rogers and Commissioner Remick.

There 16 are a number of things about this program that are 17 attractive to me personally, including the focus on 18 risk and the integration of the PRA into a utility's i

19 operation where the actual results of the PRA get 20 plugged into the day-to-day operation of a facility.

l L

21 Let me focus just for a minute on a subject

)

l 22 not within the four corners of risk-based tech specs.

-l l

\\

l 23 But in reading the material for this briefing and in 1

24 talking to the folks at Southern Cal Edison earlier

]

25 this year when Commissioner Rogers and I had a chance NEAL R. GROSS i

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W (202) 234 4 433 WASHINGTON. D C 20005 (202) 232 6600

i

\\

38 1

to 90 cut t100t, cno cf th3 thinga that occurred to me j

i 2

is that a real advantage of this program is that it 3

tells you in terms of timing when it's best to 4

conduct the surveillances and when it's best not to, I

5 and when it's best to conduct your maintenance and 6

when it's best not to.

P 7

Question:

Has any thought been given to 8

whether this basic approach would provide an avenue 9

for us to focus on a broader application of the risk-10 based principle in the maintenance context generally 11 so that rather than focusing here just on when you 12 should and shouldn't conduct the surveillance, when 13 your systems are in the right alignment that would 14 permit you to. conduct the surveillance with the i

15 lowest possible risk, or the maintenance?

Couple 16 that principle with a risk-based approach that would 17 focus on frequency; not just when it's appropriate to 18 but how of ten and how frequently you should conduct 19 your surveillances and your maintenance.

Is what t

20 we're learning here presenting us with an opportunity-l i

21 to infuse this principle of risk-based focus in the 22 maintenance of the plant in a broader way in the i

23 maintenance area generally?

Can you speak to that?

24 MR. LOBEL:

Yes.

I hate to keep going back I

25 to it, but this is still very preliminary.

NEAL R. GROSS COURT REPORTERS ANDTRANSORIBERS 1323 RHODE ISLAND AVENUE, N W, (202) 234g WASHINGTON.OC 20005 (202)232 4600 l

A 39 1

CONNISSICNER CURTISS I und0rctsnd th0t'.

2 MR. LOBEL:

But we've talked about that.

j 3

We've talked about using it and, in fact, as I 4

mentioned in the talk a little, there are other uses.

L i

5 without getting in the technical specifications, and i

6 that would be an ideal one.

And we've talked about 7

the concept of doing that.

And that could even be a i

8 first application.

9 We haven't talked to the extent that the 10 NRC would get involved in encouraging.it.

Utilities 11 can use that approach without approval from. us as 12 long they're following their technical 13 specifications.

Within a surveillance interval they 14 can choose, you know, when is the best time or how 15 many to do simultaneously and that kind of thing.

16 As far as the effectiveness of maintenance, 17 if I was understanding what you were saying, I'm not 18 sure that it would be effective for that.

This is 19 just my own opinion.

But I'm not sure it would be 20 that effective in telling you the effectiveness of f

21 maintenance because it would be hard to get enough 22 data to feed that back into the risk model to see the f

23 improvement in there.

I think you'd probably have to 24 depend on other measures of the effectiveness of the 25 maintenance.

This would just tell you in terms' of NEAL R. GROSS COURT REFDRTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

e (202) 234 4 433 WASHINGTON. D.C 20005 (202) 232 6000

i 40 1

rick wh;n it wruld be best to do it.

I 2

COMMISSIONER ROGERS:

Well, I don't know.

3 Why couldn't you couple the reliability data, actual j

4 hard reliability data, with this and --

5 MR.

LOBEL:

Theoretically you could.

6 There's no reason why you couldn't.

7 COMMISSIONER ROGERS:

Holds it together and 4

8 give you a running risk element.

9 MR. ICBEL:

The-problem is getting enough e

10 plant specific data that you could feed into the 11 model.

But theoretically, there's no reason why you 12 couldn't do it.

13 MR. RUSSELL:

Let me comment, just for a 14 minute, because that's in essence the approach that's 15 taken if you look at, as Heysham does, how close do 16 they come at the end of the year or the end of the l

17 quarter to the model that was built into the PRA.

18 And so while you're looking instantaneously to help i

l 19 you make decisions now as to when's the best time to i

20 perform the surveillance or is it appropriate to have 21 a

longer schedule or a

shorter schedule for 4

22 performing a

particular maintenance activity, you 23 collect that data and you look'at that at the end of 24 the year to see how close you came to the model.

So 25 there clearly is a relationship between what I would NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1?23 RHODE ISLAND AVENUE. N W (202) 234 4433 WASHINGTON. D C. 20005 (202) 232-6600

~.

's 41 1

charact rizo c3 o rolicbility c:ntored maintOnnnes 2

2 program where your objective is to maintain an l

l 3

availability or a reliability of a system because 4

that's indeed what you've modeled.

And so you have, l

{

5 if you're comparing what your experience was to where e

6 you are in that nodel on an integrated basis,, you 7

are, in fact, judging how well you have - done in 8

meeting that.

And so then you get to things like 9

time to repair the availability of your spare parts 10 and components.

If you did it in a shorter time, 11 that clearly effects availability.

And you're i

12 actually keeping book of what your history has been _

13 and how close you are to the model.

14 So in that context the two are related.

15 But we're very early in the discussions, and-this is 16 one issue that I wish to explore with the British 17 while we're there.

We need to discuss it ourselves, 18 but clearly they are related activities.

19 COMMISSIONER ROGERS:

Well, the whole 20 technique and the technology of doing this, it seems 21 to me, has various ways in which you can shade it and-I 22 use it.

And while it happens to be coming now to us 23 under the rubric of risk-based tech specs,_ what's 24 supporting that whole thing has many, many different 25 ways of being used.

It's a tool.

The tool can be J

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE t$ LAND AVENUE, N W.

(202) 234 4433 WASHINGTON. D C 20005 (202) 232 4 600

t 42 1

u;Cd in a varicty cf w2ya.

I think th;t w] cught to 2

be looking at those possible applications, not 3

necessarily to engage in them ourselves, but we ought 4

to be well aware of that and not simply look to see 5

how this can be used on tech specs.

That's one i

6 possible way.

I think the relationship with 7

maintenance, I

think, is one that's well worth 8

looking at.

Well worth looking at.

3 i

9 DOCTOR MURLEY:

There's a far broader, I 10 think -- what we're getting at here ultimately is the i

11 pushing out of the risk-based insights that we've.

12 gotten down to the operating staff in a plant.

13 Initially, if a utility did a PRA at all, it was at 14 their engineering headquarters and they'd do it and 15 it would sit on the shelf and gather dust, generally.

16 We've been pushing it, as you know, and now we're 17 getting more and more of the staffs involved.

But 18 what we're talking about here is a far -broader r

19 application of PRA.

l 20 COMMISSIONER ROGERS:

Right.

,21 DOCTOR MURLEY:

And quite frankly, I think 22 that's where the agency ought to be putting its 23 effort in the future.

I think we've gone about as 24 far as we can go in refining our models and cutting, 25 you know, whether such-and-such a risk number is i

NEAL R. GROSS

(.

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W (202) 234 4433 WASHINGTON. D.C. 20005 (202) 232-6800

T 43 1-oxtetly right.-

I think thoro's onough insights now 2

that we ought to be pushing this emphasis.

3.

COMMISSIONER ROGERS:

And, in a way, it's

.i 4-looking at a way of testing engineering judgments.

5 DOCTOR MURLEY:

Yes.

6 COMMISSIONER ROGERS:

I mean, engineering 7

judgments are the best that can be done at the-time.-

l 8

And then the question is,

well, you know, let's

]

+

9 review those issues in some ways because we know that 10 traditionally engineering judgments have.been very j

11 conservative and you do the best-job you can on a j

l 12 design and then you multiple everything by a factor l

1 13 of 10.

I think it's, you know, you're covered.

And 14 I think that one could very well look at some of 15 those things and visit them and this is the tool to 16 do it with.

Not a perfect tool.

I mean, I'm not 17 unaware of its limitations, but it's still a~ new tool, t

18 MR.

TAYLOR:

And it increases the 19 attractiveness to the utilities themselves.to try to 20 look at this as we proceed here, not just in NRC, but 21 to the utility to --

22 COMMISSIONER CURTISS:

It does seem to-me i

23 that I do think the utilities, particularly from the 1

i 24 standpoint of their pursuant of the risk-based tech 25 spec program for the existing reactors are probably, i

NEAL R. GROSS i

COURT REPORTERS AND TRANSCRIBERS 1323 RHOOC ISLAND AVENUE. N W

.(202) 234 4433 WASHINGTON. D.C,20005 (202) 232 4600 I

C

.44 1

frco what ycu'vo daccribed-hcro, going to do thio cn i

~

2 their own motion if they're going to do it.

But it p

3 does seem to me that one 'of the' immediate 4

opportunities that we have to do the kind of things J,.

5 that Doctor Murley is talking about is before us

- i 6

right now with the maintenance. initiative where we've

)

i. !

7 gotten now considerable information, at least we'.ve I'

i 8

had.a lot of reactor years.

Whether the utilities 9

have actually.been collecting the data on systems or-.

L 10 components remains to be seen.

But perhaps with some,

[

11 encouragement from the Commission

---it seems to me-12 the corrolary here is that what you're saying-in the-13 tech spec area.is that we have a set of tech specs.

14 that are based, by in large, on engineering judgment.

'I 15 And as we take the risk prospective and focus on the 16 tech specs, we're learning things like this system is'-

l 17 permitted to be out of operation too long from a' risk:

18 standpoint.

The LCO is 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> where a system that 19 is much more important may be permitted to be out-of f

20 operation for a much ;onger period of time.

21 I'd like to take a look here'in more detail a

22 at what you have.

I'll be anxious to see what you 23 come back with when you go to the Heysham plant.in 24 Great Britain.

But it does seem to me that this is 25 an opportunity to take a look at what, not only this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE N W.

(202) 234-4433 WASHINGTON. D C. 20005 (202) 232 4 600

I 45' li indu3try but ethOrc lik's tho cirlino industry h2v3 2'

done - in the risk area and to begin focusing in the 3

PRA context with a

particular-initiative like 4

maintenance where, I think what we're saying here for l

l 5

toch specs is,.we're interested inlthe result, in the p

6 risk-based result and it's'- apparent now from the ;

4

.t 7

standpoint of the tech specs that some cases we may:

[

1-f 8-not-be achieving the result that_ risk'would drive _us

t 9

in the direction of saying we'd like to see.

In a 1

10 very similar way, I guess-I just toss out for your-11 thought from here on out, the notion of - extending 12 this principle now and using. this ~ as a point' of 13 departure for potential application in. the

~

14 maintenance area.

I don't have any, firm suggestions i

15 to make at this point,' but it's an intriguing notion-j-

16 to me.

17 Let me ask you a

couple of s,pecific 18 questions.

On the = international - front, I recall at 19 the last meeting-the Rumanians were looking 'at 20 applying this principle. to their next generation of 21 reactors.

Is there any advancement on that front or 22 is that still on the fire?

23 MR. LOBEL:

I'm not familiar with that.

f.

24 And with conditions the way they _are now, I'm not 2

25 really sure that that would be continuing now.

I'm NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) 234g WASHINGTON DC.20005 (202) 232-6600

~

i 1

~46' 1

not rcolly curo what th3 Ruminien3 cra doing.

{

2 COMMISSIONER CURTISS:

Okay.

3 Tom, on the timing of the implementation of-4 this concept for the advance reactors, you focused on 5

the evolutionary reactors and GE~ and CE in that i

6 order, can you give me a feel, and I realize you're 7

going through the preliminary discussions here on

^

. - i 8

what a reliability program would look like in areas

-t 9

beyond just this particular one, in vie'w of the 10 schedule that we're talking about here,-can you give 4

11 me a feel for. when you would expect to reach-some 12 kind of view on whether risk-based tech specs is 13 something that we'd like to not only encourage-but, 14 perhaps, beyond that we would like to see as part of.

15 the application and submit it-for design 16 certification?

f 17 DOCTOR MURLEY:

We don't have firm times 1

18 when we would~ be dealing with this.

We've

)

19 highlighted this-whole issue of the reliability i

20 assurance program to the applicants.

I think it was l

21 about a year ago.

Sometime last spring of

'89.

At i

2 l

22 the time, we didn't have any details and we still l

23 don't, as a matter of fact, on what we want to see.

24 But we thought we'd get them thinking because they're 25 the ones that they know what we want, generally, and l-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) 234-4433 WASHINGTON, D C. 20005 (202) 232-6600 -

3 a

47-c)

.1

'it'o th0t.cb2trcet thsught.thtt I caid that a progran 2

~ of maintenance,- surveillance, tech. specs and in-~

3 service testing'and so forth that makes'sure that, as i

l t

4 near as we can, the safety-of the plant doesn't l

l 1

5 degrade with time.

Because we're going to license--

6 we could be certifying many of these plants. But in l

I I

7 any case, it's a 60 year design and there could be 8

many plants.

So the idea was to build'in as much as j

9 we'could of assurance that the safety doesn't degrade

-[ ]

i 10 once the plant starts.

~

f 11 And I think risk-based tech' specs is' going 12 to be a part of that thinking.

So, I would. hope that 13 probably sometime this summer we would ask. both GE 14 and Combustion what thinking they've done along these 15 lines and we'd sit down and talk about their whole 16 reliability assurance program.

17 COMMISSIONER CURTISS:

Okay.

That's all I 4

18 have.

19 CHAIRMAN CARR:

Well, at the risk of being 20 a

wet

blanket, speaking of

risk, I

think we're I

l 21 spending a lot of time talking about apples and we're L

22 discussing apple blossoms.

This program is just-23 barely getting underway and the way I read it,'the 24 real concern I guess I've got with it is kind of 1

l 25 hidden in the words there on page 29 that says, "The l

i NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 132$ RHODE ISLAND AVENUE, N W.

(202) 234 4 433 WASHINGTON O C.20005 (202) 2XM 4

- f, 48 1

0031 cot wSy. to do thio 10 for tho NRC to d;valcp

~

2 specific guidance for-different types of human error

]

j 3

analysis and recovery action. "

And we have a hard l

i 4

time doing that, and - with the amount of human error i

1 5

we've got in the program now, I don't'know how much q

6 anybody's going to depend on these level one PRAs to 1

7 take that into account.

That's one problem I got.

8 The second one is, I. don't know how much 9

confidence we got in the. current set of level one 10 PRAs that'we wouldn't have to go back and look at.

I 11

.think your set of issues you've got have got 12 tremendous number of problems in there when you start 13 to flush them out.

14 Having said that, let me ask you a couple 15 of questions.

The cost estimates, I notice, are.only 16 for NRC and utility efforts.

l There's no way in the 17 world, in my opinion, SAIC-is going to come in with 18 that nice a

study without. a lot-_of contractor 19 manhours involved.

And before they get through with i

20 this program, is somebody factoring in all those

.21 contractor hours?

22 MR. LOBEL:

No, not at this point because 1

23 we really haven't decided how to proceed with this.

24 CHAIRMAN CARR:

Do you really have a feel 25 for if we decide to use these level one PRAs, how NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS 1323 RHODE ISLAND AVENUE. N W.

(202) 234 4 133 WASHINGTON. D C 20005.

(202) 232-6600 i

49.

F :l 1

cccurato cnd htw valid wa'ra going to rcquiro th^.a to 4

2 be and then how much control we're going to hold over i

3 that continually updated PRA?

4 MR.

LOBEL:

No.

The issue's just been i

5 identified as one that needs to be looked into.

When.

~

6 I first started working on this, that was one of my

)

7 first questions about this issue was the accuracy of l

i 8

the-PRA.

And there's another issue, too, and that's l y 9

the validity of the PRA.

Even if it's doing "the.

10 calculation perfectly, how do you know that ---you 11 know, it isn't like doing an analytical calculation i

.l 12 where you have a reduction in flow and you go out and-e

]

13 stop pumps'and measure against your calculation.

I 14 think there are a lot of issues that need to be 15 addressed.

It may be, though, that you'can use.this i

L 16 in a relative way where you don't have to have exact 17 numbers, where you can just do comparisons.

18 CHAIRMAN CARR:

My impression is while 19 we're currently using it that way if we get a 20 justification for continued operation beyond the t

l 21 limiting condition, everybody looks at it immediately l-22 and says, "Okay, what's the risk" and letting them go 23 beyond what the tech specs say.

And I. assume we.'re 24 doing that now when we approve those things, aren't 25 we?

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N.W.

(202) 234-4433 WASHINGTON. D.C. 20005 (202)232 4 600

~-.

1 50 1

DOCTOR-MURLEY:

W3 bhco 03:L3 of cur 2

thinking and our decisions on our insights from PRAs, 3

yes.

4 CHAIRMAN CARR:

Yes.-

l s

5 DOCTOR MURLEY:

Let me - just comment, Mr.

6 Chairman, 'the PRAs themselves don't have to be l

7 absolutely-accurate in terms of ' their bottom line 1

8 numbers.

We can still take the difference between ai i

9 diesel generator-being out'for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> and 16-hours i

10 and get an increment in the sensitivity of that to

?

11 the risk.

And that is a fairly good number. -

That.

12 is, the differences from one ~ set to another set we-l 13 find to be fairly reliable.

It's where one tries to 14 take a bottom line, kind of a risk number, and use it 15 that we get very nervous, i

l 16 CHAIRMAN CARR:

That's not what worries me.

l l

17 What's worrying me is that human factor piece of the i

18 problem.

l 19 DOCTOR MURLEY:

Yes.

I think, by in large',

1 20 the models have been developed to be fairly good over 21 the years.

The models of, like the sensitivity of a' l-22 piece of a equipment or a system ' to _overall. risk. -

23 I'm pretty confident that'we have a handle on that.

t 24 CHAIRMAN CARR:

Are you confident of each 25 plant's PRA that they've done.

Some are older than NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l

1323 RHODE ISL.AND AVENUE N W.

(202) N WASHINGTON. O C. 20005 (202) 232 6600

?,

y

51 1

OthOra?

1 2

DOCTOR MURLEY:' No.

Nc..

g 3

CHAIRMAN CARR:

No.

That'sfmy concern on l'

l 1

4 the PRA.

{

]

5 DOCTOR MURLEY:

But the state-of-art is j,]

1 6

good enough, but.not all plants are up to.the' state-'

7 of-the-art..

l 8

CHAIRMAN CARR:

Yes.

9 MR.

TAYLOR:

You raised the same-

-t 10 reservations I think about the ~ PRAs, and that is l

i 11 we're much better at being - able to work with the.

12 equipment. issues than the human. issues that sometimes 13 enter into the risk.

i 14 COMMISSIONER ROGERS:

.If I

could just 15 pursue this a little bit.

And that is, I'm sort of l

l 16 the impression that we don't understand the human 1

i l.

17 factor elements in these PRAs.

And maybe we take the 18 point of view that there are --

19 CHAIRMAN CARR:

They're not in there,'are i

20 they?

21 COMMISSIONER ROGERS:

Well, some..

-22 But that the human factors effects are not very 23 different maybe I'm putting it incorrectly but 24 let's just put it this way anyhow that lack of 25 knowledge tends to have one take the point of view NEAL R. GROSS l

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE. N W.

(202) N WASHINGTON D C :tX)o5~

(202) 2324 600.

+-

a, 52 1-that 'oinco wo don't know 'vsry nuch ab:;ut tho hu an -

2 factors effects that they're more or less equivalent 3

in all situations and it would. seem to me that,

-l 4

again, the. same kind of analysis might reveal that

{

L 5

human factors are much more important in certain-j-

~

6 situations than - in others, as we know is, the case.

f

+

7 And that somehow one might be able to do. some i

1 8

relative importam guesstimates using this - approach,,

[

9 folding in' some human factors an'alysis.

Again,~you-r' 10 know, you're still not going to be 1 coking at' the -

i 11 bottom line, but you're going' to look at relative f ',7 12 importance of human factors in certain kinds. of 13 situations.

14

Now, we know how very important the 15 operators-are and things like that, but I'm thinking 16 about other areas where we might get some' insights on s

17 relative importance of human factors for different 18 equipment and things, 19 DOCTOR MURLEY:

Some studies, in fact, our 20 research program as far back as ten years.ago looked j

21 at the impact of various maintenance activities as i

22 well as control room activities on sequences of 23 accidents within a PRA.

This happened ' to be the 24 Surry plant.

From that, one can get sensitivity.

25 curves of risk.

In this case it was core damage NEAL R. GROSS

^

COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) 234-4433 WASHINGTON. D C. 20005 -

. (202) 232 4 600 [

i

53 e

I

• ~

1 frcquancy cc o 'functien of hu;On crror rota fcr i

2 various sequences.-

And one could then get. the l

g 1

like, for, example, it's fairly 3

sensitivity of 4

' obvious, but a transient sequence is more important,.

j 2

-d 5

more sensitive.to human' errors than one where -

like-j 6

-pressure vessel failure where it's just

a. single

]

7 component has to fail..

8 So there are' areas like the.t that have'been' 9

done.

But I think we have quite a bit more to do to a

10 really understand.

And maybe it's not even possible, 11 but to understand the kind of cognitive errors that i

12 operators can make.

And by that-I mean they have an 13 idea in their head that. something's 1 wrong with the 14 plant and it's the wrong mental picture of the plant.

~

15 And they go out and they ignore everything else.

And 16 I don't know how you model.something like that.

17 If you read again the Chernobyl event, I 18 mean they clearly blocked out a lot of'information 19 that they shouldn't have.

And how could anyone 20-predict a sequence like that?

21 So I think there are limits to how far we q

22 can go.

I'm not sure that was the question you were 23 getting at, but we have to keep probing.

And'I think 24 that's where our research program is heading.

25 MR. RUSSELL:

Another piece related to your I

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

. (202) 234 4433 WASHINGTON. D C 20005 (202) 232 4 600 f

.\\

t 54' t

1 cram:nt 10 that.tho GOnnitivity ctudico ycu do in 2

human reliability analysis are very model dependent.

3 And they all seem to build on the same data that is i

4 somewhat dated that was gathered back in the late

,y i

5'

'50s.

And so what you do by Way of validating'the j

6 model is a difficult question.

And I'm not aware.of 1-7 activities going on to collect data to attempt to do-8 model validation in the human error analysis areas.

9 So you get some insights for a particular model.

y' {

10 COMMISSIONER ROGERS:

Well, I

thought j

e 11 that's what we were raising questions about when we 12 heard about our human factors program, that there was 13 a need to collect some more data.

I mean, that that l

14 was one of the things where work had to be done.

15 MR. RUSSELL:

That may very well be.

I'm 16 commenting based upon as I knew the program three and i

17 a half years ago when I was heavily involved in it.

l 18 And I'm not aware of any work that's been done in 19 that intervening time.

Yes, the validation of the 20 models and the data is a key issue if you're going to

.21 use it to make significant regulatory decisions.7 j

22 COMMISSIONER REMICK:

I will say as an 23 engineer I would welcome the added insight.from risk-24 based models realizing they're imperfect'in making a 25 judgment than not having that information available NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N.W.

(202) 234-4433 WAShMGTON. D.C. 20005 (202) 232 4600 o

f 55-1 to to in E2 king th0t judgaOnt, j

i 2

CHAIRMAN CARR:

There's no doubt about.

3 that. -We all appreciate all the help we can get.-

{

i 4

COMMISSIONER REMICK:

Right.

Sure do.

5 CHAIRMAN CARR:

Has any licensee, even-one 1

6 of the three that was involved in this working effort 7

so far, expressed an-interest in volunteering for the 8

pilot study?

9 MR.

LOBEL:

There's 'been some interest, i

i

.i 10 yes.

Nobody has volunteered per se, but people have i

11 expressed interest in it.

12 CHAIRMAN CARR:

Well, one thing I know they 13 like is stability in the. regulatory process.

They g

14 got a system there they know how,to --work with and i

l 15 they know what to do now, and trading thatfin for i

16' something that's kind of-indefinite,.it seems-to me, 17 would undoubtedly be one of their concerns..

But, 18 personally, I think -. the thing.'s - got promise.

You 19 know, all those great things that I've worked:towards j

20' ever since I've been; around have always been just 21 around the corner and it's just a great' idea. -

I 22 think that it's certainly worth pursuing.

i 23 Well, I.thank the staff for updating the l

24 Commission on the status of the program.

The concept 25 and its proposed implementation show promise in terms NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVENUE, N W.

(202) 2344 433

' WASHINGTON. D C. 20005

{202) 232-6600 l

56' i-1 of b3ttOr dofining occ ptOblo'cquip'nnt, outtga tits 3 4

2 and surveillance test intervals from a

risk r

3 prospective.:

I do note, ' however, that staff and 4

industry resource-expenditures to develop and i

5 implement the program described seem considerable. I e

6 wguld caution the staff to carefully. determine the 7

potential benefits that may derive from this concept-a 8

as the program evaluation continues.

9-Any of my fellow Commissioners have - any 10 additional comments? 'Thank you very much.

~

11 We stand adjourned.

12 (Whereupon, at.11:07 a.m.

the briefing was 13 adjourned.)

14 15 16 17 18 t

19 20 l

21 22 23 24-25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE IGLAND AVENUE. N W.

(202) Wg WASHINGTON, D C. 20006 (202) 232-6600 '

r

q L

l CERTIFICATE OF TRANSCRIBER I

l

^

This is to certify that the attached. events of a meeting of the United States Nuclear Regulatory Commission entitledt TITLE OF.MEETINGt BRIEFING ON RISK-BASED RECHNICAL SPECIFICATION PROGR d PLACE OF MEETINGt ROCKVILLE, MARYLAND DATE OF MEETING APRIL 13, 1990 1

were transcribed'by me. I further certify that said transcription:

j

~

1

.is accurate and complete, to the best of my' ability,~and that the 1

transcript is a true and accurate record of'the foregoing evento.

j M74.

vV

'X Reporter's namet Miles Anderson j

e HEAL R. GROS $

COUtf ht90RTit$ AND TRANSCRittR$

1323 RNOct l$tAND AYINUf. H.W.

(202) 234-4433 WASHtK3?ON, D.C.- 20005 (202) 232 6600

I j

U i

5 t

4

-i l

RISK-BASED TECHNICAL SPECIFICATIONS L

APRIL 13,1990 4

L

?

RICHARD LOBEL i

E 1

I l'

l.-

k.

V.

4 4

w

-.___-_______________.______.._._..___________.__,____.__.+-_..n m

's:

_f

-6 g,

4-j

~%

i N

[

t ge N

g E

a O

b U-O C

r m:

L C

BBBM,

gi l L

e a

v z

a w

o

-g l

4 7

n

=

g e.

i o

r=

a l,

R#asv4Tib U

k w

M i

Q R

C L

Z r

4 l

C'

\\

U:

W vs E

o l

I

3, or -

Risk-Based Technical Specifications 3

Allowed Outage Times (Hrs) j i

l A

g f

l I

I

. M reiskeased i

1 l

0 Cummt j

W

-l e

l d

l O

l u

'I l

t 92 a

i 9.

72 72 72 l

' O

/

/

/

/

T M

~

i i

/

/

i Accumulator MD EFW Pump TD EFW Pump LPI Pump MD EFW-Motor Driven Emergency Feed

- TD EFW - Turbine Driven Emergency Feed LPl-Low Pressureinjection 1

i

~._

'y w

• Ap'Q p

s s,ny

,emwtg4 m

y,y gq yyy_m

,4, 4 M.

n wsm %l, T

y

• r ypp

NRC EFFORTS ON RISK-BASED TECHNICAL SPECIFICATIONS l

o NRC Initiated Two Studies Of Risk-Based Technical Specifications t

Best A' proach (1987) p l

Feasibility Of A Pilot Program (1988) o Real Time Approach Selected l

Most Comprehensive And Rapid Assessment OfInstantaneous Plant Risk i

s i

i o

A Pilot Study Should Be Initiated At A Volunteer Utility To Address j

l And ' Resolve Remaining Technical Issues j

i Gain Confidence About Ability To Minimize Plant l

OperationalRisk i^

l

s-

'h i

i WORKING GROUP UTILITY MEMBER EFFORTS i

i o

Philadelphia Electric Company Uses The Limerick PRA To Enhance

~

t The Safety Of Plant' Operation 1

4 4

o Southern California Edison Does Quarterly Monitoring Of j

Risk Profile l

i, o

EPRI/ Westinghouse / Pacific Gas And Electric: Interactive Advisor

)

i J

.1 I

5 4

4 i

s::

~ > -

. - -. =

.. m

6*.

9.-

1

)

q FOREIGN EFFORTS l

J U.K. - Essential Systems Status Monitor ESSM)

=

J o

Heysham 2 Was Designed And Is Operated With Real Time Risk Analyzer (ESSM) i L

Nordic Countries (Finland. Sweden) l l

L o

Reconsideration Of Plant Shutdown Requirements When Failures l

l Occur In Specific Safety Systems 4

I i

L l

l

7.

?

I i

O i

f REGULATORYISSUES 4

o New Interpretation Of10 CFR 50.36 2

o Risk-Related Criteria Supplements Current Licensing Criteria o

Criteria For Operator Action In Terms Of Risk ~

i o

Configuration Control Of The Risk Model o.

Controlled Access To Risk Model l

o Adequacy of Base PRA L

I i-l l

l l.

~=, _ _ _ _ ' __

a"g--'

-&m fq

(

X 4 9 er* *.

h' v'

A-m

• V'k--

yrf%.g 4 wg4 w

%-m4m'-

n-'

4%

w

.r'"'

-+m-*

4"me 9q-w.';"w me T w+

,.~

• 13-q

. j i

a NRC ROLE.

I o Monitor And/Or Evaluate Industry Efforts o Provide Guidance, As Requested, To Industry Programs o Monitor Foreign Efforts o Take Part in Foreign Information Exchange Activities (e.g.,IAEA),

h' l

l I

T i

I

'}

L l

l i

E-l:

l,?

,._,,.,.m-.

e

._,.r.,

gy..,,,,.,,.

I "M%%Ed&6%W6WWWWd%%%WWW6WWM(444Wdgfggg TRANSP.1TTAL TO:

Document Control Desk, 016 Phillips ADVANCED COPY TO:

The Public Document Room

/ 9/io DATE:

FROM:

SECY Correspondence & Records Branch 5'

i Attached are copies of a Comission meeting transcript and related meeting ij.

document (s). They are being forwarded for entry.on the Daily Accession List and lj placement in the. Public Document Room. No other distribution is-requested or

.l required.

n i

f Meeting

Title:

O_UL A M_d bMM MPAel

[

&_;s J.

y y w;?

l -

l Meeting Date:

M//5/9o Open X

. Closed j.

ii j

ltem Oescription*:

Copies Advanced DCS

8 to POR C3 i

f~

l

1. TRANSCRIPT 1

1 d

o,/Muh gl g

.:a 2-ll j

v z:

L.,

2 3-3:l 3

3 3j 3;

1 ll 4.

f: -

aal

?l i l 5-ll a

l E.

3 3 !

6.

g ;

g ';;

2:

2al

• POR is advanced one copy of each document, two of each SECY paper.

.p 3'

C&R Branch files the original transcript, with attachments, withcut SECY ll gt papar$ -

1 h

A/AA Y

_