Meeting of the Joint Subcommittee on Reliability and Probabilistic Risk Assessment - December 16, 1999
227
1 UNITED STATES OF AMERICA
2 NUCLEAR REGULATORY COMMISSION
3 ***
4 MEETING OF THE JOINT SUBCOMMITTEE ON
5 RELIABILITY AND PROBABILISTIC RISK ASSESSMENT
6 ***
7
8
9 U.S. Nuclear Regulatory Commission
10 11545 Rockville Pike
11 Room T-2B1
12 Rockville, Maryland
13
14 Thursday, December 16, 1999
15
16 The above-entitled proceedings commenced at 8:33
17 a.m., pursuant to notice, the Honorable Dr. George
18 Apostolakis, chairman, presiding.
19
20 MEMBERS PRESENT:
21 GEORGE APOSTOLAKIS, ACRS Chairman
22 MARIO V. BONACA, ACRS Member
23 ROBERT E. UHRIG, ACRS Member
24 ROBERT L. SEALE, ACRS Member
25
. 228
1 P R O C E E D I N G S
2 [8:33 a.m.]
3 DR. APOSTOLAKIS: The meeting will now come to
4 order. This is the second day of the meeting of the ACRS
5 Subcommittee on Reliability and Probabilistic Risk
6 Assessment. I am Dr. George Apostolakis, Chairman of the
7 Subcommittee.
8 ACRS members in attendance are Mario Bonaca,
9 Robert Seale, and Robert Uhrig.
10 The purpose of this meeting is to discuss NRC
11 Staff efforts in the area of Risk-Informed Technical
12 Specifications and associated industry initiatives proposed
13 by the Risk-Informed Technical Specification Task Force.
14 The Force is risk-informed?
15 SPEAKER: Risk-Informed Task Force.
16 The Subcommitte will gather information, analyze
17 relevant issues and facts and formulate proposed positions
18 and actions as appropriate for deliberation by the full
19 Committee.
20 Michael T. Markley is the Cognizant ACRS Staff
21 Engineer for this meeting.
22 The rules for participation in today's meeting
23 have been announced as part of the notice of this meeting
24 previously published in the Federal Register on December
25 1st, 1999.
. 229
1 A transcript of the meeting is being kept and will
2 be made available as stated in the Federal Register notice.
3 It is requested that speakers first identify themselves and
4 speak with sufficient clarity and volume so that they can be
5 readily heard.
6 We have received a request from Mr. Jim Riccio of
7 Public Citizen to enter a written statement into the public
8 record related to risk-informed technical specifications.
9 Mr. Riccio may offer oral statements later in the meeting as
10 time permits.
11 Mr. Riccio's statement is as follows:
12 "Statement of James P. Riccio, Staff Attorney,
13 Public Citizen's Critical Mass Energy Project, before the
14 U.S. Nuclear Regulatory Commission's Advisory Committee on
15 Reactor Safeguards, December 16, 1999.
16 Good morning. My name is James Riccio. I am the
17 Staff Attorney for Public Citizen's Critical Mass Energy
18 Project. It is a pleasure to once again present our views
19 to the Advisory Committee on Reactor Safeguards.
20 You are here this morning to discuss yet another
21 deregulatory effort focused on reducing the technical
22 specifications that govern the operation of the 103 nuclear
23 reactors still splitting atoms across the United States.
24 While I'm sure that NEI and the NRC have been working
25 hand-in-glove in order to deregulate what remains of the
. 230
1 technical specifications, we, the public, have been able to
2 find precious little information regarding this deregulatory
3 effort.
4 In fact, the only reference to technical
5 specifications I have been able to find in the deluge of
6 deregulatory paperwork being generated by the NRC actually
7 argues against a further diminution of technical
8 specifications. Public Citizen opposes any further
9 reduction in the technical specifications.
10 The NRC's new-and-improved technical
11 specifications were never intended to improve safety, only
12 the economic viability of the nuclear industry, by reducing
13 the limiting conditions of operation by 40 percent. After
14 NEI has successfully colluded with NRC to reduce LCOs by 40
15 percent. They now want to risk-base the remaining 60
16 percent. The argument that this will somehow improve safety
17 in the long-run is at best sophist.
18 NRC rightfully argues that the industry can expect
19 little in the way of burden reduction due to the previous
20 deregulation that occurred during the last technical
21 specification rewrite. However, since this effort is
22 apparently being driven by NEI, perhaps they will share
23 their deregulatory plans with the public. I am sure they
24 have already shared them with the NRC.
25 However, with less than a page of information
. 231
1 buried in the middle of a 90-page SECY paper, the public is
2 left with little to comment on.
3 If the NRC wants to garner any public confidence,
4 it is going to have to provide opportunities for more
5 meaningful participation. Allowing the public to comment at
6 ACRS hearings is virtually meaningless if we do not have
7 prior access to the information."
8 This is the end of the statement.
9 We will now proceed with the meeting, and I call
10 upon Mr. Beckner and Mr. Barrett of NRR to begin.
11 MR. BECKNER: Thank you. My name is Bill Beckner.
12 I am the Chief of the Technical Specifications Branch in
13 NRR. I would like to thank the subcommittee for having us
14 this morning. I am going to give a very, very brief
15 introduction by way of just introducing who the players are
16 here today and also indicate what we would like to try to
17 accomplish this morning.
18 First of all, Rich Barrett is here, Chief of the
19 Probabilistic Safety Assessment Branch. He will be making a
20 brief opening remark. Also Biff Bradley from NEI is here.
21 This is a presentation that we have coordinated with NEI.
22 It was discussed and planned at the public meeting in
23 October. I will let Biff introduce some of the industry
24 players.
25 Our main presenters today will be my Section
. 232
1 Chief, Bob Dennig and my staffer, Nan Gilles, here, and of
2 course Mark Reinhart, who is the Section Chief in the PSA
3 Branch and a former tech-spec'er, so he's been involved in
4 this effort for some period of time.
5 As far as what we would like to do today is I
6 understand it's been almost two years since we have talked
7 to you, and we have accomplished a lot. That has been
8 primarily in the area of risk-informed AOTs. That effort is
9 evolving now. We would like to tell you where we think we
10 are going, provide the ACRS with an opportunity to comment
11 early in the process.
12 We are asking for nothing specific at this point
13 in time other than to let you know what we are doing and to
14 give you a chance to provide any guidance or counsel that
15 you may have.
16 With that, let me go ahead and see if Rich wants
17 to make some comments.
18 MR. BARRETT: Thank you, Bill. I will be brief.
19 I just want to make a couple of points.
20 This is an initiative that the Staff is very
21 excited about. It is one that I think it's fair to
22 characterize as one that grew from the thinking that went on
23 within the NRC Staff, and I think basically where it comes
24 from is our thinking about what is it that controls risk.
25 We know that there are many different things that
. 233
1 control risk -- reliability and availability certainly being
2 one of them. We think that controlling the configuration of
3 the plant is something very, very important to risk --
4 controlling the availability of equipment one at a time but
5 also in combinations is very, very important to controlling
6 the risk of nuclear plant operation.
7 I think you are all aware that the new, revised
8 maintenance rule -- (a)(4) -- is an attempt to control the
9 risk associated with voluntary changes to the configuration
10 of the plant, and that change to the rule was made in the
11 context of the existence of the technical specifications,
12 and with the knowledge that there was always -- the
13 technical specifications were there as a backstop to any
14 changes that were made.
15 The effort that we are going into now is to try to
16 bring this risk-informed thinking into the technical
17 specifications today. What you are going to hear about
18 today is rather broad-ranging. You are going to hear about
19 seven specific initiatives which vary from some relatively
20 mundane proposals that are being looked at in the nearest
21 term to more sweeping proposals that would be phased in in
22 the somewhat longer term.
23 You will also be hearing about a vision that goes
24 more to the heart of the specifications and the philosophy
25 of the specifications themselves.
. 234
1 The idea of all of this is to present the NRC and
2 the industry with a range of options that will provide all
3 of us with the flexibility to go in small steps in the
4 direction of risk-informing the technical specification or,
5 for those that are prepared to and want to, to make a giant
6 leap toward risk-informing the technical specifications.
7 But in sum, we believe this is an effort than can enhance of
8 NRC's pillars in terms of maintaining safety, improving
9 public confidence, improving the effectiveness and
10 efficiency of regulation, and reducing unnecessary
11 regulatory burden. I think you are going to find it a very
12 interesting presentation today.
13 I would like to introduce Biff Bradley, who would
14 like to make a brief --
15 DR. APOSTOLAKIS: Of the seven initiatives that
16 you mentioned, a giant leap or something else is a giant
17 leap?
18 MR. BARRETT: Well, I think a plant that would
19 want to go all the way from (1) through (7) all at once, I
20 think you would call that a major leap, but the option is
21 there for plants to take smaller steps.
22 DR. APOSTOLAKIS: Okay. Thank you.
23 MR. BARRETT: Any other questions?
24 DR. SEALE: I have one after Biff gets through.
25 Go ahead.
. 235
1 MR. BARRETT: Biff, would you care to talk?
2 MR. BRADLEY: Thanks, Rich. I would just like to
3 reiterate Rich's remarks and we will talk about this when we
4 get a chance to make our presentation. We do believe the
5 congruence with the (a)(4) rulemaking is a timely
6 opportunity to make some fundamental changes to risk-inform
7 tech specs.
8 What I would primarily like to do now is just
9 introduce the industry people who will be here along with
10 myself today to discuss this. We do have an active industry
11 group that involves all four Owners Groups, EPRI, NEI, and
12 it is a cohesive effort to try to achieve this on an
13 industry-wide basis.
14 Today we have Tom Hook from Southern California
15 Edison, which as most of you know, is one of the plants that
16 are more further along with a lot of PRA applications;
17 Sharon Mahler, from Nebraska Public Power District; and
18 Wayne Harrison from South Texas, another of the plants that
19 is heavily involved in a lot of the risk-informed
20 activities. Thanks.
21 DR. SEALE: Rich, in the statement that we heard
22 from Mr. Riccio, there seemed to be considerable pain
23 expressed over the unavailability of information as to what
24 was going forward and so on. Now you told us that your
25 intent at this point is as a progress report, to get input
. 236
1 from us, and at some point down the road there will be a
2 more formal I guess presentation of what all of this implies
3 and so on.
4 Do you have any plans as a part of your progress
5 as you move forward to assure that the public as represented
6 by Mr. Riccio will have better or access to what your plans
7 are and be in a position to comment in an informed way?
8 I have a hard time being able to respond to
9 generalities in this area.
10 MR. BARRETT: Well, the answer to your question is
11 yes. I think that the NRC more and more has an emphasis on
12 making sure that we bring stakeholder input into the process
13 at every stage of the game, and I think to the extent that
14 there is a perception we haven't done that so much this far,
15 then I think that means that we should do even more in the
16 future, and so the answer is absolutely yes. There will be
17 multiple opportunities for the public and other stakeholders
18 to have input at various stages along the way and perhaps
19 Bob could give you more details about that.
20 MR. BECKNER: Yes -- Bill Beckner again.
21 We have had a number of internal concerns about
22 the standard tech specs in general about how the process to
23 changing the standard tech specs is really not a public
24 process in the sense that we have public meetings and we put
25 information in the public document room, but we do not
. 237
1 solicit comment in general.
2 We have made a decision at this point in time that
3 following issuance of the next formal -- issuance of the
4 standard tech specs, which will be within a few months, that
5 changes such as this will go out specifically for public
6 comment and again in our meeting we had in October, which
7 was a public meeting at the end of a long series of public
8 meetings, I think I made the statement that we would be
9 putting these seven items out specifically for public
10 comment, so hopefully there is a chance as we go along with
11 meetings like this and other public meetings for people to
12 know what we are doing, and then there should be a chance
13 for formal public comment at this point in time.
14 MR. MARKLEY: Bill, is your December 14th letter
15 to NEI in the PDR?
16 MR. BECKNER: What is the subject here?
17 MR. MARKLEY: The meeting summary.
18 MR. BECKNER: I don't know if it is today. It
19 should be. It will be.
20 MR. MARKLEY: Okay, fine. Would you have any
21 objection to us putting this on the back of the transcript?
22 MR. BECKNER: Absolutely no objection.
23 MR. NEWBERRY: Scott Newberry, NRC Staff. We
24 weren't aware of Mr. Riccio's concerns and I plan to take an
25 action item based on this letter to contact Mr. Riccio
. 238
1 personally to chat about this and inform him of the process
2 improvements and make sure he gets all the information that
3 we have available.
4 DR. SEALE: The Commission has also considerable
5 sensitivity towards this problem of garnering real public
6 comment and we have thought -- I think the subcommitte
7 chairmen will agree -- we have had the impression that these
8 meetings in fact provide some opportunity for public
9 awareness, not only of the written word but also of the
10 mindset that in many cases is behind the development of some
11 of these things.
12 So we would be interested and we are interested in
13 ways -- in what you are doing and we are interested in ways
14 that we can make our process as well more responsive to the
15 needs of the public.
16 MR. BARRETT: I guess I would like to add a word.
17 I think this is a case where a Staff initiative has met with
18 such enthusiasm and progress has been made so fast that I
19 think we would have to go back and look at the record of our
20 public workshops and other meetings to see if we have
21 touched all the bases in terms of stakeholders and make a
22 commitment to be certain we do so in the future.
23 I think as an example we would rarely have gotten
24 to this stage in any program without having had a full
25 discussion with the ACRS, so I think that is evidence to the
. 239
1 fact that this is a program that has kind of tumbled forward
2 with a great deal of enthusiasm, but we are totally
3 committed to having input from all the stakeholders
4 including the public interest groups.
5 MR. DENNIG: Just one last word. With regard to
6 the subject matter of this particular meeting, the Staff did
7 meet with the Risk Informed Task Force on November 10th and
8 basically what we are going to talk about today is a reprise
9 of the material that was discussed at that meeting. That
10 meeting was a publicly-noticed public meeting that would
11 have afforded the opportunity to see all of the
12 decision-making and the input and the thoughts and the
13 process over a number of hours.
14 MR. REINHART: If I could just add one other
15 thing -- this is Mark Reinhart in the Risk Assessment
16 Branch -- one of the initiatives that our branch is trying
17 to do, get underway, even as we speak is to put a website,
18 publicly available, that has not just this initiative but
19 all of the risk initiatives that are going forward along
20 with the status and whatever information we can provide.
21 It is readily obvious a person comes onto the NRC
22 site they can get to the risk site and just see really what
23 is going on there and have some contacts, people they can
24 get in touch with.
25 DR. APOSTOLAKIS: Okay. By the way, is the Staff
. 240
1 Requesting a letter?
2 MS. GILLES: No.
3 MR. BECKNER: This is Bill Beckner. We are
4 requesting any comments that you may want to provide.
5 MR. MARKLEY: We do have one logistic problem
6 here, and that is anyone who uses the speakers, you are
7 going to have to walk up to the table or go to the head of
8 the table with the microphone.
9 MR. BECKNER: This is Bill Beckner. We are not
10 specifically asking for endorsement, but we would welcome
11 any comments that you might have in whatever form you want
12 to give them to us.
13 DR. APOSTOLAKIS: Okay.
14 MR. BECKNER: If that is a letter, we would
15 appreciate it.
16 DR. APOSTOLAKIS: Thank you.
17 MS. GILLES: Good morning. I'm Nannette Gilles
18 from the Technical Specifications Branch in NRR.
19 This morning, I'm going to try to help remind you
20 where we were when we were last before this Subcommittee.
21 That was two years ago.
22 And then I'll tell you what progress we have made
23 since then in the area of risk-informed technical
24 specifications.
25 Our last briefing before the PRA Subcommittee was
. 241
1 in February of 1998, and at that point in time, we were here
2 to present you with the final risk-informed Regulatory
3 Guides and Standard Review Plans.
4 And also, we were in the final stages of wrapping
5 up the risk-informed technical specification pilot
6 applications.
7 The risk-informed pilots were in their final
8 approval stages, and at that time, the pilots and the
9 guidance only addressed risk-informing technical
10 specification, allowed outage times, and surveillance test
11 intervals.
12 Also at that point in time, the industry
13 participation was limited. You really had one of the owners
14 groups as the main participant, and a few individual plants
15 applying for these risk-informed applications.
16 Since then, we have issued over 30 tech-spec
17 amendments to risk-informed allowed outage times, using the
18 guidance in Regulatory Guides 1.174 and 1.177, which is the
19 specific guide for technical specifications.
20 And we have 10 amendments currently under review,
21 so there has been wide industry interest in pursing these
22 risk-informed improvements.
23 In addition, the principles in the Reg Guides have
24 been applied to other types of amendments, some one-time
25 amendments, some amendments to actually remove some
. 242
1 requirements from tech specs.
2 Most of the allowed outage time improvements that
3 have been made have included the implementation of what has
4 been called the Configuration Risk Management Program.
5 And that is a PRA-based assessment to look at the
6 risk of being in multiple equipment outages at the same
7 time.
8 And you'll hear this term, this risk assessment
9 term, and risk management program used throughout today's
10 presentations, and that is essentially the equivalent of
11 what the Maintenance Rule A-4 now requires. So those are
12 really synonymous programs at this point in time.
13 DR. APOSTOLAKIS: Excuse me, the CRMP, is that the
14 predictive tool, or is it something like the risk monitor,
15 which gives the current state of the plant?
16 MS. GILLES: It is to look at the configuration
17 you intend to enter.
18 DR. APOSTOLAKIS: Okay.
19 MS. GILLES: So predictive in that sense, and also
20 if you are performing some maintenance and some other event
21 occurs, another piece of equipment becomes inoperable, it is
22 used at that time also to look at the immediate
23 configuration you are in.
24 DR. APOSTOLAKIS: So somebody sitting in an office
25 and does calculations, and says, if I take these things out,
. 243
1 then there are all these possibilities that are part of the
2 CRMP?
3 MS. GILLES: It varies from the range of an online
4 risk monitor to a matrix type of tool, to using the base PRA
5 to do calculations.
6 DR. APOSTOLAKIS: The online risk monitor would be
7 different, it seems to me. Because you know what is going
8 on at that instant. You're making projections. You have to
9 include the possibilities of human error and this and that.
10 If you are looking at the plant now, you know what
11 the status is, you know that people have not made any
12 mistakes, perhaps.
13 So they are two different things. So I was
14 wondering whether Configuration Risk Management Program, is
15 somebody going to talk about it later?
16 MR. REINHART: Maybe if I could just offer it, I
17 think the way the program was written originally, it was
18 written in Reg Guide 1.177, and it was not prescriptive.
19 There are some principles, and certain licensees,
20 as Nannette said, would implement it, very much real-time.
21 Others would pre-analyzing configurations they
22 intended to go into. So it wasn't prescriptive, per se. I
23 think as time goes on, and these applications get more and
24 more sophisticated, we'll see the licensees that have the
25 more flexible tool will be qualified for the more flexible
. 244
1 applications.
2 DR. APOSTOLAKIS: But is anyone going to talk
3 about it in more detail later, or this is something that
4 we're supposed to know?
5 MR. DENNIG: I think perhaps as far as how this
6 may be implemented on an individual plant basis, maybe we
7 have some people here from industry that have experience and
8 know how they're doing it at their facility. That might
9 help if they can get into that.
10 DR. APOSTOLAKIS: Okay, thank you.
11 MS. GILLES: If there are no questions, I'll turn
12 it over to --
13 DR. APOSTOLAKIS: Yes, there is a question.
14 MS. GILLES: Okay.
15 DR. APOSTOLAKIS: Can you explain the 19 and the
16 12 there? What, exactly do these mean?
17 MS. GILLES: That just means 12 plants, some of
18 them might be a unit plant, so that there were actually 19
19 units getting the change.
20 DR. APOSTOLAKIS: Oh, I see, okay.
21 MS. GILLES: Bob Dennig is now going to give
22 information about the current risk-informed tech specs.
23 MR. DENNIG: Thanks very much, Nan. I'm Bob
24 Dennig, a Section Chief in the Tech-Spec Branch.
25 We thought we'd start at the top. My purpose is
. 245
1 to outline for you, some of the principle elements that are
2 underpinning staff thinking as we further evolve and develop
3 in the area of risk-informed tech specs.
4 So we have sort of a vision statement or a vision
5 slide to begin with. The first major point is that what we
6 envision is an evolution to fully risk-informed standard
7 tech specs. What does that mean?
8 That means that in addition to using risk
9 technology, reliability analysis, to improve various
10 parameters that appear in tech specs such as allowed outage
11 times or completion times, and to optimize in some
12 instances, surveillance intervals and so on and so forth.
13 What we're talking about is taking the
14 configuration risk management concept that's present in the
15 AOT work that we've already done, and emerges again in the
16 revision to the Maintenance Rule, and explicitly marrying
17 that into the technical specifications and moving in the
18 direction of having risk-informed decisionmaking for the
19 actions taken within technical specifications. So that's
20 the real big picture that we have in mind, eventually going
21 to that kind of arrangement.
22 DR. APOSTOLAKIS: At some point, develop an actual
23 sentence that describes that vision, not today, perhaps, but
24 that would be very useful.
25 MR. DENNIG: I'll take that as an assignment.
. 246
1 Another part of the vision is that this is a
2 phased approach, just echoing Rich's comments earlier. It's
3 phased in time, phased in extent, in that we mentioned
4 already the seven initiatives.
5 Those are meant to be things that plants may pick
6 up all or some or none; that will be a function of their
7 situation, and their PRA capabilities.
8 And that ties into the notion of the degree of
9 reliance on PRA in these different initiatives.
10 There are varying degrees of reliance in
11 qualitative and quantitative ways on PRA technology
12 underpinning some of these initiatives. But our vision
13 accommodates that kind of a structure.
14 Finally, we're premising this on voluntary
15 adoption. We have an established process for making changes
16 to the standard technical specifications. That's the Tech
17 Spec Task Force process.
18 Mr. Beckner, Dr. Beckner, mentioned that earlier.
19 So we can feed the changes that we envision through that
20 process. We will be getting public comment on these
21 modifications to the standard.
22 And also we intend to focus on those areas where
23 there is interest in the industry in implementation. I
24 think we share a goal with industry of having things
25 actually eventually getting done and implemented in the
. 247
1 field, and that's where we're going to focus.
2 DR. APOSTOLAKIS: I must say that I'm confused,
3 maybe because I don't understand the concept of standard
4 technical specifications.
5 Are you preserving anything by moving from the
6 current system to a risk-informed system, and what is that
7 something, if you are preserving it? Is it risk or what?
8 I mean, what does it mean to have a fully
9 risk-informed standard technical specification?
10 How will that -- how will it be different from
11 what we have now?
12 MR. DENNIG: Again, current technical
13 specifications were written with a set of LCOs, Limited
14 Conditions for Operation.
15 DR. APOSTOLAKIS: Right.
16 MR. DENNIG: They have a fairly prescriptive
17 structure. If you find yourself in this configuration, then
18 you will take this action. If you find yourself in this
19 configuration, you will take this action; if you find --
20 with time limits and so on.
21 DR. APOSTOLAKIS: Right.
22 MR. DENNIG: Okay? That structure can't --
23 obviously does not anticipate all the configurations that
24 plants can be in. It was constructed, for the most part,
25 with the notion of one piece of equipment at a time, a
. 248
1 particular system performing a particular function being out
2 of service.
3 We're now in an environment where there are
4 multiple concurrent equipment outages for planned reasons
5 and unplanned reasons.
6 The prescriptive actions that are written in the
7 script that's in tech specs don't always put you in the
8 safest or least risky situation.
9 And the objective is to use what we've learned
10 since the concept of structured tech specs were first set
11 up, to improve the ability to assess plant state, take all
12 the factors into consideration, and go in the safe direction
13 where there is decisionmaking that's informed by risk
14 technology.
15 DR. APOSTOLAKIS: I understand that, but the
16 current system, the prescriptive system, implies a certain
17 core damage frequency, right? I mean, I can calculate that,
18 the PRAs, that's what they do, because they use the plant
19 and so on.
20 It also implies a certain unavailability for the
21 diesels, trains, for all kinds of things.
22 Now, as I move to the new system, am I going to
23 try to keep the core damage frequency the same, or increase
24 it a little bit according to 1.174? Am I going to try to
25 make sure that the unavailability of the trains are the same
. 249
1 or increased a little bit?
2 What is it that I'm preserving? That's not clear
3 to me.
4 MR. REINHART: I hope we're going to get into this
5 in a minute, but the thought is that if you look at a given
6 configuration the way the tech specs would put the plant
7 now, like I said, we could calculate a given core damage
8 frequency.
9 DR. APOSTOLAKIS: So it is the CDF?
10 MR. REINHART: That's a measure.
11 DR. APOSTOLAKIS: A measure. Okay.
12 MR. REINHART: You could also look at that
13 calculated core damage frequency in a state to which you
14 might go, compare those two, and you might also calculate
15 the risk of the transition, and you might put in some
16 compensating factors and try to overall determine what is
17 really the safest thing we could do with the plant right
18 now?
19 I think what the industry is proposing are some
20 steps that are leading us in that direction, and as we go
21 down the path and get a better feel, more practice, if you
22 will, we're going to be able to come up with a more refined
23 system.
24 DR. APOSTOLAKIS: Well, CDF is, of course, a major
25 measure in our metric in 1.174, 1.177, and so on, which use
. 250
1 that. But we can also think in terms of the cornerstones of
2 the new oversight process.
3 MR. REINHART: Yes.
4 DR. BONACA: I hope you are going to try and
5 preserve those, I mean, whatever it was before, we want it
6 to be the same in the future.
7 There are certain things that you leave unchanged.
8 You still have a protection system that is in tech specs.
9 MR. REINHART: Yes.
10 DR. BONACA: You have settings, you have a number
11 of things that define your level of action, okay, for
12 mitigating systems. So those are some of the things that
13 you are probably not going to touch.
14 MR. REINHART: I agree. I think right now --
15 DR. BONACA: Let me just finish. So what you're
16 going to talk about is the level of configuration that you
17 are going to allow by tech specs in relation to those
18 protection systems, okay, to allow you more latitude to do
19 certain things on a risk basis, knowing now that you are not
20 affecting, really, risk significantly by having a certain
21 component out of service for a longer time, in relation,
22 again, to something that doesn't change.
23 I mean, you're not planning to change set points
24 and things like that.
25 MR. REINHART: No, no. The set points would stay
. 251
1 the same. Really right now, the function, I think is where
2 you hit the configuration in which the plant should be or
3 should go to do the safest thing.
4 DR. BONACA: But I think the question that Dr.
5 Apostolakis raised is very good in the sense that it will be
6 certainly helpful to have almost like a conceptual model,
7 you know, presented on how you're going to move on this, and
8 understand what is not changing there.
9 You know, again, what is changing? Because
10 oftentimes when I look at these presentations, I'm left with
11 a lot of questions about what are they going to come with
12 next week to change? I would like to understand if it falls
13 within a certain pattern.
14 DR. APOSTOLAKIS: It may be obvious to you guys
15 because you're working with it, but to a third party, it
16 would be nice to have these things up front.
17 MR. REINHART: Maybe as the industry goes through,
18 they're going to lay out these seven pieces in a time line,
19 and I think that will put into perspective, at least the
20 thinking today.
21 Like Dr. Barrett mentioned at the outset, we'll
22 see the long range vision as it is today in one of the
23 pieces.
24 DR. BONACA: That kind of information, however, if
25 you made an effort to develop it conceptually, would go a
. 252
1 long way also for the concern we have for the public.
2 Because it would indicate what you're trying to
3 do. All the messages I get is that, for example, you have
4 the next slide -- I'm sorry.
5 DR. APOSTOLAKIS: Let's go to the next one.
6 DR. BONACA: Yes. You're talking about --here
7 you're talking about necessary -- the theme is always
8 producing a conservatism. It seems to me that you're really
9 making a better tech spec. That's really what I see coming
10 out of this.
11 I think that should be a better communication, a
12 better tech spec from a safety standpoint.
13 MR. REINHART: That's a good comment, and I
14 believe that's our concept.
15 DR. APOSTOLAKIS: Can you explain the words,
16 emphasizes the Commission's nuclear reactor safety
17 performance goals?
18 MS. GILLES: I think those are the next couple of
19 slides.
20 MR. DENNIG: Yes.
21 DR. APOSTOLAKIS: Okay, let's go to the next
22 couple slides.
23 MR. DENNIG: And there they are.
24 DR. APOSTOLAKIS: So I guess we've been discussing
25 the first bullet for the last few minutes; is that correct?
. 253
1 MR. DENNIG: Yes.
2 DR. APOSTOLAKIS: To maintain safety?
3 MR. DENNIG: Yes.
4 DR. APOSTOLAKIS: And how do we define that?
5 MR. DENNIG: Let's just go to the next slide.
6 DR. APOSTOLAKIS: Okay.
7 MR. DENNIG: We'll be getting into this in a
8 little more detail. It's a little more useful.
9 What we've done is take what we want to achieve
10 and parsed them amongst the pillars or the filters or the
11 safety program, reactor program safety goals.
12 We believe that the direction that we're headed in
13 will allow us to maintain or enhance safety by balancing
14 transition, shutdown risk with the risk continued degraded
15 operation.
16 There is a premise that was built into the
17 original construct of technical specifications that if
18 something is out of service to a certain degree, you begin
19 shutting down. You go to cold shutdown.
20 That was just the default. We think we know
21 better, we know how to do better. So we think that -- and
22 what we're trying to do is, we're going to be able to make
23 an improvement in that area.
24 Part of that notion should be going to a safer end
25 state, if you will. Why should we go to cold shutdown if
. 254
1 that's not what we know is the best place to be from our
2 experience over 30 years?
3 So we're reexamining the premise of where tech
4 specs arrive at, too. We've already touched a couple of
5 times on addressing multiple equipment unavailabilities.
6 DR. APOSTOLAKIS: This leaves something basic out,
7 in my view, because I don't understand. Let me tell you
8 what I would do with this, if this was the only thing:
9 It's clear to me that the risk is, let's say, CDF,
10 or even the two cornerstones or three cornerstones,
11 initiating events, the integrity of the primary systems, and
12 so on; is the NRC telling me that all they care about is
13 preserving, maintaining those that I already have calculated
14 in my PRA, say, the core damage frequency, for simplicity?
15 So I can go back and develop a revolutionary,
16 entirely different system that I will spend a lot of time
17 arguing that it preserves the core damage frequency, and now
18 we come here and upset everything you've been doing for the
19 last 40 years. You're not going to accept that, are you?
20 So there is much more to this than what you say
21 there. In other words, I'm coming back and I'm telling you,
22 look, I'm going to -- if I do a much better job on this
23 system, you know, I can do the probabilistic calculations,
24 you know.
25 Then I make sure that the CDF remains the same, so
. 255
1 I can neglect something else that for the last 20 years I've
2 been taking care of by following the allowed outage times
3 and so on. In other words, I create an imbalance within the
4 Level I PRA, whose results, however, are still the CDF that
5 I have.
6 I'm not sure you will accept that. You probably
7 want a gradual change. You probably have other goals that
8 you are not stating there.
9 And that's why I think a crisp statement of what
10 you want to do would go a long way towards explaining these
11 things to the public and maybe to this Committee as well. I
12 mean, what does balancing transitional shutdown risk mean?
13 What does addressing multiple equipment unavailabilities
14 mean?
15 These are words that take many -- can accept many
16 interpretations. This is so important that I think we need
17 that, we need a vision statement with goals, this is where
18 we're coming from, and this is what we want to do.
19 Is there somewhere where I can find it, or is that
20 something that you will have to develop? I don't think it
21 was in 1.177.
22 MR. DENNIG: No. Again, let me just go back to
23 Dr. Beckner's opening remarks.
24 The purpose of coming and talking today was to lay
25 out a concept that has a lot of work to do as far as the
. 256
1 details and implementation and the kinds of things that
2 you're talking about.
3 And the list of questions that you're giving us,
4 and the things that you're pointing to that cry out for
5 additional information, is very useful to us, and it's the
6 essence of why we're here, in a sense.
7 DR. APOSTOLAKIS: Good.
8 MR. DENNIG: So keep them coming. And where there
9 is something that just is like you guys are smoke and
10 mirrors, you know, we recognize that there are a lot of
11 boundary conditions on what we're doing as far as transition
12 from where we've been to where we want to get to. How do we
13 do -- the phrase that comes off of people's tongues very
14 easily, blending the deterministic and complementing it with
15 the risk and what does that mean, in essence, when you're
16 really doing something.
17 To a certain extent, that's going to emerge from
18 the doing part of this, but, again, we're getting good value
19 for our effort from your questions, so please continue to
20 ask.
21 DR. APOSTOLAKIS: So this is something that you
22 think would be useful?
23 DR. BONACA: Yes. It would be good if it started
24 by explaining what the tech specs are all about.
25 DR. APOSTOLAKIS: I was about to ask that
. 257
1 question, but I was --
2 MR. DENNIG: The purpose of the tech specs --
3 DR. APOSTOLAKIS: Can you tell me what the tech
4 specs --
5 MR. DENNIG: Funny thing.
6 DR. APOSTOLAKIS: Oh, you have it.
7 [Laughter.]
8 MR. DENNIG: This statement was the most succinct
9 that we could find. It comes from a statement of
10 consideration and states essentially what at the highest
11 level tech specs are supposed to accomplish.
12 MS. GILLES: It's the backup slide.
13 DR. APOSTOLAKIS: We don't have it?
14 DR. SEALE: The ghost of Tom Kress walks.
15 [Laughter.]
16 DR. APOSTOLAKIS: Oh, boy.
17 MR. DENNIG: So that is the mission statement, if
18 you will, for tech specs, and interestingly enough, if you
19 go back to the law itself it becomes even less specific as
20 far as what tech specs are supposed to do.
21 DR. BONACA: My suggestion was more to go into
22 more specifics. You know, you can come a step below and
23 explain how the specs are structured. You don't have to go
24 through -- you know what I'm saying? You don't have to do
25 it today. I am only saying that it would be satisfying to
. 258
1 say how it's structured, you know, there is a protection
2 system, a discussion of the conditions that you are allowing
3 the plant to have in configurations, and then you can go and
4 show some of the deficiencies tied to the deterministic
5 assumptions because we did not know really whether this --
6 you know, that the design showed the improvements of the
7 tech specs.
8 DR. APOSTOLAKIS: Can we have those slides, by the
9 way -- since we showed them now?
10 MR. DENNIG: Sure, absolutely. These are the
11 basic subsections of 50.36, which is an implementation of
12 the statement.
13 DR. APOSTOLAKIS: Again, coming to the previous
14 slide, in spite of calls from groups outside this agency, I
15 am not sure that adequate protection can be quantified.
16 That is a personal view, not the ACRS, so maybe you can
17 start with this and then in stating what you plan to do, I
18 guess you can say that the big change between the early days
19 and now is that part of the risk can be quantified, okay?
20 The challenge here is to use that quantification
21 in combination with other things that create adequate
22 protection for public health and safety to change the tech
23 specs so that the adequate protection to the public health
24 and safety is still there.
25 Now this is very high level. Then you have to go
. 259
1 down and start explaining what these things mean, but this
2 is really the major change, that now we can quantify part of
3 the uncertainty, right? There are still other
4 uncertainties, which is the same thing that we said about
5 defense-in-depth and everything else.
6 Anyway, I think we spent enough time on this and I
7 am glad to see that your guys agree. We have done this,
8 haven't we?
9 MR. DENNIG: Reducing unnecessary burden. We
10 think that this is unnecessary regulatory burden. We think
11 the congruence with the maintenance rule principles
12 50.65(a)(4) is a worthy goal. We would like to have, I
13 believe, utilities working off of one construct as far as
14 the configuration for their plant and we would like to at
15 least understand in a detailed way how the (a)(4) concept
16 works with tech specs.
17 It has been said that they are a backstop. In
18 some cases they would keep you from going someplace that
19 your configuration risk management program would let you go
20 and in other cases the tech specs would allow you to go
21 places that your configuration risk management program won't
22 let you go. We would like to not have two regulations out
23 there that are battling each other depending on what the
24 situation is.
25 DR. SEALE: There is a point here that you haven't
. 260
1 made yet and I think has considerable impact here, and that
2 is when the technical specifications in their traditional
3 statement evolved the assumption was that maintenance would
4 be done on a shut down plant. This in part was accepted
5 because of the generally held, if as it turns out possibly
6 incorrect assumption that a shut down plant is the safest
7 plant.
8 Now we find that we can do certain maintenance
9 under operating conditions and in fact because of the active
10 condition of certain systems and so on in that configuration
11 it is actually a preferable way to do it. The maintenance
12 rule is the articulation of the process by which you go
13 about doing that maintenance whether the plant is active or
14 shut down. It almost seems necessary the to re-examine the
15 tech specs to ask yourself if this new mode of maintenance,
16 that is, with an operating plant, will have impact on those
17 tech specs since they were in fact originally conceived
18 under the notion that the plant would be shut down during
19 maintenance.
20 So the tie-in here is almost more intimate than
21 you have inferred here.
22 MR. REINHART: You're right. It's been a driving
23 thing that -- an age driving force in this program.
24 DR. SEALE: Yes.
25 MR. REINHART: It's recent -- it's 1994 -- so a
. 261
1 tie with the maintenance at power versus shutdown and the
2 maintenance rule is certainly there.
3 DR. APOSTOLAKIS: I have another comment on this
4 slide. As you know, the agency has been emphasizing the
5 need to communicate with the public more and better, in
6 better ways, and it seems to me I know you didn't intend
7 that, but if you put bullets like that, you are really
8 separating the safety issues from the unnecessary burden,
9 and as you recall perhaps from Mr. Riccio's statement he
10 also believes that the NRC's new and improved technical
11 specifications were never intended to include safety, only
12 the economic viability of the nuclear industry, he says.
13 Now as I recall, when we were preparing or
14 reviewing the 1.174 regulatory guide there was an argument
15 that was made that reducing unnecessary burden does in fact
16 contribute to the enhancement of safety, especially the
17 last -- well, all three, but say providing operational
18 flexibility to the plant.
19 So perhaps when you prepare slides like this or
20 documents to be released to the public, these subtle points
21 need to be made, in my view. Now I am not saying that by
22 reducing any kind of burden you enhance safety, but the key
23 word is "unnecessary" -- then, you know, the utility can
24 utilize its resources in a more efficient way. Operational
25 flexibility is enhanced, therefore safety is enhanced.
. 262
1 One thing that sometimes is lost is that in 1.174
2 even when we calculate the change in CDF and LERF we can
3 still make a case that the overall change is in fact
4 negative, but there are qualitative factors that we cannot
5 quantify it at this time, or at least it is not as much as
6 we calculated, so if we calculate that three 10 to the minus
7 5 -- no, you can't have that -- ten to the minus 6
8 increasing CDF, I think it was our collective belief that
9 most of the time it is not going to be three 10 to the minus
10 6 because of these other nonquantifiable benefits you are
11 getting from this, so I think this is a key point, because
12 if we keep talking about reducing unnecessary burden without
13 making this connection, it is not -- we are not really
14 communicating very well.
15 MR. DENNIG: Your point is well-taken.
16 DR. APOSTOLAKIS: We are falling desperately
17 behind, so let's see -- I think maybe you can go to #9
18 unless you feel that you have something very important to
19 say on #8.
20 MR. DENNIG: I think your decision is a wise one.
21 The way that we are conducting this effort, the
22 vehicle underlying the effort is a joint NRC-industry effort
23 that originated in July '98 at a Licensing Managers'
24 Workshop.
25 The Risk Informed Tech Spec Task Force includes
. 263
1 representatives of all the NSSS owners, NEI, EPRI and
2 supporting contractors. We have had five meetings held to
3 date. Those have been public meetings. In one we plan a
4 trip and that lead plant is San Onofre. Let's go on to the
5 next slide.
6 By the way, we have a lead-in to the
7 presentation-discussion here from industry after the break.
8 The Risk Informed Tech Spec Task Force is currently pursuing
9 seven initiatives that were initially identified and we have
10 just binned them here into some -- throwing the seven into
11 some larger categories, and they basically break down to
12 risk-informed plant maneuvering, elimination of shutdowns
13 for low risk situations, and risk-informed surveillance
14 testing.
15 These initiatives take advantage of and
16 incorporate the concepts of configuration risk management
17 and consistent with the maintenance rule concept. Perhaps,
18 a very important point, we feel that we can make a great
19 deal of progress and largely get to where we envision going
20 without any changes needed to 50.36.
21 It is flexible enough that we can use the existing
22 structure.
23 That concludes my presentation. I would like to
24 turn it over to Mark Reinhart from Risk Management.
25 MR. REINHART: If I could start maybe to put this
. 264
1 in perspective, I'm Mark Reinhart from the NRR Probabilistic
2 Risk Assessment Branch, the Section Chief in that Branch.
3 Maybe to put it in perspective, at least to help
4 me interact, if I view tech specs, I look at a couple
5 pieces:
6 One piece I have is an operating envelope, you
7 might say, power, temperature, pressure flows, some limits,
8 some margins, in which we want to the plant to operate.
9 We have configurations of equipment that help it
10 stay in that envelope or bring it back, given some
11 initiating event or transient, bring it back where it is in
12 a safe envelope.
13 And then we have some administrative controls that
14 help balance that.
15 I think what we're doing is looking at that middle
16 piece, we're looking at what equipment should be operating,
17 how redundant it should be, how diverse it should be under
18 different operating conditions.
19 And when the industry proposes a change and we
20 review it, we're going to look at these five principles in a
21 risk-informed manner.
22 We're looking to make sure we comply with
23 regulations, we're looking for defense-in-depth, safety
24 margins, and then we get into the 1.174 piece where we're
25 looking for a risk decrease if we can do that.
. 265
1 Risk-neutral, okay, and if it is, the calculated
2 risk is an increase, we're going to look into Reg Guide
3 1.174 for some guidelines on how much to allow that
4 calculated increase, but at the same time -- and I'm going
5 to get to this in the next slide -- we're going to be
6 looking for some compensatory measures to account for that
7 delta.
8 And the fifth principle: We're going to be
9 looking for subsequent ways to monitor that plant's
10 performance. I think that if we see, starting with some
11 fundamental proposals, moving to the more sophisticated over
12 a time line, we're going to be looking not only to the
13 staff, but also to the industry to be able to articulate how
14 they're doing this, how we're doing this, and what are some
15 ways to monitor performance.
16 I'd like to comment about the cornerstones. One
17 thought is some sort of performance indicator that would
18 indicate where we are in that program.
19 DR. APOSTOLAKIS: It's much more involved than
20 that, though.
21 MR. REINHART: Yes.
22 DR. APOSTOLAKIS: Because, no, no.
23 [Laughter.]
24 DR. APOSTOLAKIS: The cornerstones, you see, our
25 statement of defense-in-depth, now if you go to 1.174, they
. 266
1 give a list of bullets there with what defense-in-depth
2 means. So there is a lot of flexibility here on how you
3 interpret these concepts, all right?
4 MR. REINHART: Right.
5 DR. APOSTOLAKIS: And I understand the staff is
6 working now on a new risk-informed Part 50 to try to see at
7 what level defense-in-depth should be recommended.
8 So, this is -- there is a lot of flexibility, a
9 lot of latitude here, how to do it.
10 Now, one question: Of the seven initiatives, when
11 we look at the seven initiatives, each one will be subjected
12 to the is process separately from the others?
13 MR. REINHART: Yes, as applicable. I think you
14 could say some are going to be, particularly in the fourth
15 bullet, more risk information will be available to be
16 brought to the table than in others.
17 In a couple, the industry is going to make their
18 proposal that some of their risk information is going to be
19 after the fact.
20 DR. APOSTOLAKIS: Again, going back to the days of
21 the development of the Regulatory Guide, as I recall, the
22 idea was that the licensee submits a request.
23 MR. REINHART: Yes.
24 DR. APOSTOLAKIS: The staff reviews it and so on,
25 and if it is approved, there will be the monitoring of the
. 267
1 subsequent performance.
2 MR. REINHART: Yes.
3 DR. APOSTOLAKIS: Now, in principle, the
4 Regulatory Guide does not forbid the submission of a new
5 request next Monday.
6 MR. REINHART: That's true.
7 DR. APOSTOLAKIS: So we don't have to wait to see
8 what the subsequent performance will be, due to the first
9 request, right? They can come back and request something
10 else.
11 MR. REINHART: They can make the request,
12 certainly.
13 DR. APOSTOLAKIS: So these seven then will be
14 submitted in seven consecutive Mondays or altogether?
15 MR. REINHART: They have a time line. It's over
16 several years.
17 DR. APOSTOLAKIS: Several years.
18 MR. REINHART: That they are proposing to make
19 those submittals. Some, I believe two. Are there two that
20 are inhouse?
21 MR. DENNIG: Two.
22 MR. REINHART: Two are inhouse, just recently, and
23 the others will be coming in over a period of time.
24 Again, they're proposing to tell you what that
25 time line is. I think we can all have a different view of
. 268
1 how ambitious that time line is.
2 DR. APOSTOLAKIS: Now where under these five you
3 would evaluate -- five bullets -- Dr. Seale's statement
4 earlier that perhaps our notion of what is a safe state now
5 is different than it was assumed to be before? Shutdown may
6 not be a safe state.
7 I mean, changing that falls where?
8 MR. REINHART: Could I go to the next slide and
9 show you that?
10 DR. APOSTOLAKIS: Of course you can.
11 MR. REINHART: Particularly under Bullet 4, and I
12 want to preface this. We're talking right now with
13 operating within the current rule that we have, 10 CFR
14 50.36(4), the tech specs. We're not looking at rule changes
15 to what we're looking at.
16 But, if you will, define an integrated minimum
17 risk locus, which would include the risk of the at-power
18 configuration, versus the risk of that configuration in some
19 other mode, plus considering the transition risk to get from
20 one mode to the other, valve lineups, et cetera, that a
21 plant would have to go through.
22 At the time, placing on top of those, whatever
23 appropriate compensatory actions we come up with, say what
24 is the safest thing to do with the plant, given the
25 situation? That's the goal.
. 269
1 Some of it is going to be quantitative. Some of
2 it's going to be qualitative, some of it's going to have to
3 account for uncertainty.
4 Some of it's going to be single component, some of
5 it's going to be multiple components. But what we're trying
6 to do is, rather that second-guess today, where that plant
7 is going to be tomorrow, once the plant gets there, what's
8 the safest thing to do to get to where we want to go, and
9 can we identify some particular areas that we want to avoid?
10 If we can do that, we do have a safe plant. We
11 likely have a safer plant than we have today.
12 We've allowed flexibility to the operator, and we
13 fully intend to be public, and I think you've made comments
14 about we need to be crisp and clear in what we're saying so
15 that people catch on to where we're going.
16 DR. APOSTOLAKIS: I think what you just said makes
17 perfect sense.
18 Now in terms of the availability of tools in order
19 to do your third bullet there. Will the licensees be
20 required to have decent PRAs for power transition, all
21 specific?
22 MR. REINHART: Outstanding question, and the way
23 we're viewing that today, is there are some plants that
24 could probably approach that right now. They have very
25 nice, high-quality PRAs.
. 270
1 To give a plant that flexibility, we would be
2 looking for that type of a PRA. We're looking at a
3 voluntary program.
4 If some plant doesn't have the PRA, doesn't want
5 to play in this arena, they're not going to be forced to,
6 but the plant that does want to play in the arena, we need
7 to make sure they have a quality PRA that justifies what
8 they're doing.
9 And I think the ones that can go there today are
10 going to be able to demonstrate to the industry that this is
11 a safe, cost-effective, reasonable way to operate a nuclear
12 power plant.
13 DR. APOSTOLAKIS: Okay.
14 MR. REINHART: That's really my presentation, just
15 those two slides.
16 DR. APOSTOLAKIS: So what's next?
17 MR. REINHART: We'd propose we take a break, and
18 then the industry will come and present to you, their view
19 of risk-informed standard tech specs.
20 DR. APOSTOLAKIS: So let's take a break. Till
21 when? 9:55?
22 DR. BONACA: Why not?
23 [Recess.]
24 DR. APOSTOLAKIS: Okay, we're back in session.
25 Mr. Bradley?
. 271
1 MR. BRADLEY: Thanks. I'm Biff Bradley from NEI.
2 I wanted to proceed with the industry presentation today.
3 As you know, there are a lot of risk-informed
4 initiatives underway now. As a matter of fact, the industry
5 is nearly deluged with trying to make a lot of changes in a
6 lot of areas at the same time: Oversight process, the
7 regulations themselves.
8 And the tech specs are a high priority for us, for
9 a number of reasons. I just wanted to mention some of those
10 here today. I know the ACRS hasn't been hearing a lot about
11 this up till now, and I did want to mention the priority.
12 As we talked about, I think a lot of these things
13 are a rehash of what the staff already discussed, but there
14 are significant plant operational benefits.
15 The tech specs are really an artifact of a
16 previous era in the history of the plants, and there is a
17 much greater state of knowledge now and additional things
18 going on with risk-informed, particularly the Maintenance
19 Rule that we'll talk about, that really allow greater
20 operational flexibility of the plant and the ability to use
21 risk insights to get the plant into the safest state, and to
22 make decisions on configuration control and mode changes,
23 using a lot of insights that the tech specs didn't really
24 bring to bear when they were created back at the inception
25 of the plants.
. 272
1 Another nice thing about this activity is that it
2 doesn't require rulemaking. And because of that --
3 rulemaking even with the best of intentions, is always a
4 multiyear effort, and the ability to make some changes, some
5 really significant changes that impact the way the plants
6 are operated without having to wait out a rulemaking effort,
7 is a great benefit.
8 At the executive level, when we discuss the whole
9 set of risk-informed initiatives with the industry, there is
10 a lot of interest in this because it's something the plants
11 believe they can achieve and get into place in the next year
12 or two, a lot of these changes, and we'll see that in the
13 schedule.
14 Finally, those of you that are familiar with all
15 the discussions we're having on Option 2 and Option 3 of
16 Part 50 reform, it's a very ambitious effort. There are
17 some very thorny issues that we have to overcome.
18 I'm confident that we will, but in some ways, the
19 tech specs are more directly amenable to be able to use the
20 risk tools we have now to get some changes without having to
21 spend years trying to, for instance, determine what the
22 impact of special treatment is on equipment reliability and
23 availability.
24 Tech specs typically are changing things that are
25 like equipment out of service times or mode changes or
. 273
1 whatever, where there are generally ways to address that in
2 a risk-informed way, without hypothesizing what kind of
3 impacts there are of these softer regulatory programs.
4 Next slide.
5 As has been discussed already today, I just want
6 to reiterate this; this is a timely effort, because in the
7 middle part of next year, the rulemaking to (a)(4) of the
8 Maintenance Rule, which is the configuration control
9 provision, will be implemented at every plant in the
10 country.
11 The Committee has already reviewed the
12 implementation guidance for (a)(4) a number of times, and I
13 think, George, you were asking about the CRMP earlier. In
14 my view, the (a)(4) implementation guidance and the CRMP are
15 essentially one in the same, so you can go back and look at
16 what you've reviewed previously on (a)(4).
17 I'll just mention that (a)(4) essentially does
18 what a large part of the tech specs do. Now, the tech specs
19 do lay out safety limits and some other things, but in terms
20 of the configuration control, which is a bit chunk of the
21 tech specs, that is essentially being duplicated by the
22 (a)(4) regulation.
23 And it's doing it in a risk-informed way; the
24 scope of equipment involved for (a)(4) is the scope of the
25 PRA. It's the large of scope of equipment in the tech
. 274
1 specs, and the general concept of (a)(4) is that, as we
2 talked about today, is to manage the risk around your
3 baseline value.
4 And that's what we have. As you know from
5 reviewing the guideline, configurations of specific risk,
6 the temporary risk increases, cumulative risk, all those
7 issues are addressed in the (a)(4) guidance.
8 Basically it's using the tools that we have in the
9 smart way to manage configuration control. When you look at
10 what tech specs does, being something that was developed
11 prior to the risk-informed era, it's more aligned to the
12 deterministic licensing basis of the plants.
13 So what we're doing in (a)(4) is looking at larger
14 set of components and systems in the plant, and making
15 configuration control decisions based on that larger scope.
16 Just a note that in our comments on the (a)(4)
17 rulemaking, industry did support that rulemaking, even
18 though it is a fairly significant new requirement, because,
19 as we talked about, tech specs weren't really designed to
20 address multiple equipment out of service conditions.
21 However, we did note that ultimately the tech
22 specs needed to be reconciled with that. As was mentioned
23 today, in a lot of cases, tech specs will constrain you.
24 The (a)(4) evaluation may tell you that one in-state or mode
25 change or what have you is the right way to go, but tech
. 275
1 specs will say, no, can't do that.
2 Of course, the tech specs are the mantra for the
3 operators. It's part of the licensing basis of the plant,
4 and you can't, unless you go in for an enforcement
5 discretion or -- I mean, whatever the provision is to get
6 temporary relief, you're pretty much locked into what your
7 tech specs are.
8 And just to conclude on (a)(4), much of what is in
9 (a)(4), Rich talked about voluntary configuration changes,
10 and (a)(4) addresses more than just that. It addresses
11 emergent conditions, corrective maintenance, surveillance
12 testing.
13 If you look at the list of things in the (a)(4)
14 rule, it's very duplicative of what tech specs are there to
15 address now. And it's important for the industry that this
16 activity go forward, and that we can achieve some synergy
17 between this and (a)(4).
18 The timing is perfect. If we can meet the
19 schedule for this activity, it will fit perfectly with the
20 (a)(4) implementation schedule.
21 And the (a)(4) is not an insignificant rulemaking
22 for the industry. We did issue -- PRA quality came up, and
23 (a)(4) does include specific discussions of those areas of
24 your PRA that have to be addressed, references the industry
25 peer review process as a means to address areas such as
. 276
1 treatment of dependencies and support systems that are
2 important for these types of considerations.
3 Clearly, as was suggested here, PRA quality is an
4 element of being able to make these tech specs changes, and
5 being able to implement (a)(4), so everyone is going to take
6 advantage of the limitation of the scope of (a)(4), using a
7 risk-informed process as the rule allows. It's going to
8 have to address those PRA quality issues.
9 So at this point, I'd like to turn it over to
10 Sharon -- I'm sorry, to Tom, Tom Hook, who will proceed with
11 our presentation.
12 MR. HOOK: Good morning. My name is Tom Hook, and
13 I'm the manager of nuclear safety oversight at San Onofre
14 Nuclear Generating Station, which is owned and operated by
15 Southern California Edison, primarily.
16 I'm going to speak briefly about the risk-informed
17 tech spec goals. These are goals that were established by
18 the industry/NRC task force that's been developing the
19 proposed changes to the technical specifications.
20 These are also goals that were established by the
21 Combustion Engineering Owners Group as a part of their
22 risk-informed license changes that have been developed over
23 the past several years.
24 The first major goal is to establish a framework
25 for this particular project for decisionmaking that utilizes
. 277
1 risk-informed tools, as well as traditional engineering.
2 That framework would be different than the
3 framework that we have right now at plants; that is,
4 deterministic-driven and rule-driven in the present tech
5 specs.
6 This framework would rely more heavily on a risk
7 management, a configuration risk management program, and the
8 (a)(4) for providing the boundary conditions for
9 decisionmaking.
10 It would integrate the Maintenance Rule (a)(4)
11 decisionmaking, the revised reactor oversight process, the
12 significance determination process, as a part of the NRC's
13 and the industry's toolbox for assessing the significance of
14 deviating from the technical specifications, missing
15 surveillance.
16 It would also provide a means to look at the
17 existing technical specification allowed outage times, the
18 end-states, and how the Maintenance Rule decision criteria
19 in the NEI Guide would allow us to place the plant in the
20 safest operational state.
21 Moving to the next slide, in terms of goals, the
22 additional goals were enhancing plant safety and reducing
23 unnecessary burden. And I acknowledge the earlier comment
24 from the Committee that a blending of these and
25 acknowledging how reducing unnecessary burden also enhances
. 278
1 plant safety is an appropriate comment.
2 And these are the similar types of bullets that
3 were in the staff slides in the earlier presentation:
4 Eliminating unnecessary mode transitions; and determining
5 the appropriate end-state when equipment is inoperable;
6 integrating risk information into the maintenance and
7 operation of the plant more fully, and for some plants such
8 as San Onofre, we feel we have gone to great lengths to do
9 that right now.
10 And this project here is an ability for us to move
11 that decisionmaking process into the technical specification
12 area more fully, and take advantage of it in terms in terms
13 of maneuvering the plant.
14 Monitoring and controlling plant risk to
15 acceptable levels, and those levels are defined in the
16 (a)(4) NEI guidance, Chapter 11. And there are goals there
17 for work.
18 There are acceptance levels for cumulative risk
19 and core damage frequency, and also in Reg Guide 1.174, for
20 increases in core damage frequency and large early release;
21 and, lastly, selecting the appropriate actions when
22 equipment is inoperable in terms of maneuvering the plant,
23 taking compensatory measures, providing a framework for
24 making those decisions in a predictable, repeatable, and
25 appropriate program.
. 279
1 And lastly, reducing unnecessary burden, reducing
2 occupational exposure. We're going to a lower mode to a
3 cold shut down that may create the potential for unnecessary
4 occupational exposure due to in-containment or radiation
5 activities.
6 There is the reducing the costs of operating the
7 unit by increasing the availability of the unit, and adding
8 resource allocation flexibility to promote the attention to
9 the equipment that is most important, most significant to
10 safety where currently that is not possible due to limited
11 resources.
12 I also wanted to speak specifically relating to
13 the configuration risk management program that we have at
14 San Onofre, based upon an earlier question from the
15 Committee, and describe how that program is evolving into
16 (a)(4) and the rigors of that program at San Onofre.
17 We have a safety monitor at San Onofre. It's the
18 Scientech Safety Monitor that is utilized to predict,
19 calculate the risk of future plant configurations, and that
20 tool is actually in the hands of the NRC. It was purchased
21 by the NRC and the NRC has our safety monitor model, as of a
22 number of months ago, that included at-power and shutdown
23 risk calculation capability.
24 Since then, we've added transition risk and
25 additional interim mode calculational capability, as well as
. 280
1 the external event fire and seismic initiator models.
2 DR. APOSTOLAKIS: Has the staff reviewed the
3 models that are used to develop the risk?
4 MR. BECKNER: We've looked at it.
5 DR. APOSTOLAKIS: Here. You can speak, whatever
6 you want.
7 MR. BECKNER: We've looked at it, but we haven't
8 done a formal review, but we've been at San Onofre, we've
9 worked with it there, we've had lots of discussions, and
10 we've looked at the software capabilities.
11 But as of yet, we've not sat down and done what
12 you would call a review that we would write a safety
13 evaluation on.
14 DR. APOSTOLAKIS: But you plan to do this, to
15 review it?
16 MR. BECKNER: That's a good questions, and I think
17 one of the open questions for us is just to what extent does
18 the staff go when a licensee provides a PSA, they do their
19 cross-comparisons, their independent reviews, what part do
20 we have to do and how deep should it be?
21 DR. APOSTOLAKIS: I looked a little bit at the
22 concept of this monitoring, and it seems to me because its
23 intent is different from the PRA, you can't just take the
24 plant's PRA and computerize it.
25 MR. BECKNER: I agree.
. 281
1 DR. APOSTOLAKIS: So I think, and I don't recall
2 ever discussing these issues with the staff. Maybe we ought
3 to have a supplemental meeting to discuss this.
4 For example, if a piece of equipment is out, what
5 do you do to the PRA, just set the unavailability over to
6 one? Or are there other things you have to do to reflect
7 the fact that this equipment is out of service?
8 In fact, I would say that there are several terms
9 that are affected by that fact, so if the risk monitor is to
10 play such an important role in configuration risk
11 management, it seems to me that we ought to do this, and
12 make sure that we understand what the models behind the
13 pictures we see are.
14 MR. BECKNER: We agree. As we said, as the
15 industry starts to develop this capability and wants this
16 flexibility, we have to move with them. We're talking about
17 it, and I think a discussion with the Committee would
18 provide valuable insights and help in that arena.
19 DR. APOSTOLAKIS: Okay, that's something that I'm
20 very much interested in. We already have a comment that --
21 not directly on the monitor, but in the context of
22 risk-informed in Part 50 on importance measures in our
23 letter of -- do you remember when it was? Last June.
24 MR. MARKLEY: No, it was this Fall, it was like
25 October.
. 282
1 DR. APOSTOLAKIS: So a lot of the comments we made
2 there, I think would apply to a risk monitor, because there
3 the issue was what happens to the risk importance measures
4 when you change configuration? Right?
5 DR. SEALE: Do you really have the detail to
6 capture all of the changes that are implicit in certain
7 equipment being out of service in your risk, so that you can
8 really tell what the change is?
9 DR. APOSTOLAKIS: That's right. So I think we
10 should do that work at some point.
11 MR. HOOK: And as most nuclear stations, we have a
12 proceduralized control process for modifying the safety
13 monitor of the PRA model that is in the risk tool. We have
14 certification efforts underway in the owners group as the
15 other owners groups have to review all the living PRAs at
16 nuclear stations.
17 We also have internal controls to make sure that
18 the plant staff understands the limitations of the tool. We
19 have external peer reviews that have been conducted, and
20 we've submitted a significant amount of detail and results
21 from our PRA as a part of the prior risk-informed allowed
22 outage time changes, including cut sets and such, to the
23 staff for their review.
24 So we feel that we have a very high quality PRA.
25 We're also embarking upon additional activities that are not
. 283
1 specifically a part of this.
2 That includes developing a plant risk trip model
3 that would be incorporated into the safety monitor as a few
4 other nuclear stations have developed, and we're developing
5 that for economic reasons, as well as safety reasons, to
6 address the cornerstone regarding initiating events, as well
7 as looking at the mitigating systems.
8 I don't have anything else on the San Onofre, so
9 that concludes my part of the presentation.
10 MR. HARRISON: Good morning. My name is Wayne
11 Harrison. I am from STP Nuclear Operating Company and I am
12 also wearing two hats today. The other hat -- I am going to
13 primarily be representing the Westinghouse Owners Group in
14 my discussion.
15 Before I go into this slide, I'd just a comment, a
16 follow-on to what Tom was saying on configuration risk
17 management programs. South Texas also has a configuration
18 risk management program that based on our probabilistic risk
19 assessment that is done using risk man. -- and we have
20 quantified several, a number of thousand, of plant states
21 into a risk assessment calculator that we call RASCAL that
22 we can readily quantify plant configurations and the risk
23 associated with plant configurations. We use it on weekly
24 maintenance and assessing plant configurations.
25 This slide kind of puts into pictorial format the
. 284
1 vision, and what we are really looking for here, to try to
2 put this somewhat into words, is to develop a set of better
3 engineered risk-informed technical specifications that
4 facilitate safe and effective plant operations. We want to
5 base this on early successes and a phased approach and have
6 a set of products that can be applied across a wide range of
7 PRA capabilities.
8 I agree with Dr. Apostolakis in that we really
9 need to refine that statement and get it focused, because I
10 think a good vision statement really does keep a task in
11 focus.
12 I think this does complement the NRC vision that
13 was articulated earlier in Bob Dennig's presentation.
14 What you see across the top is an estimated
15 timeline. This is certainly not cast in concrete. This is
16 just our best guess of what the timeline would be for these
17 set of initiatives and to the left of the little vertical
18 dotted line you see our current initiatives and Sharon is
19 going to talk about those a little bit later as we go into
20 the follow-on slides. I am not going to really go into
21 those in any degree of detail.
22 The next three boxes are our future initiatives,
23 which are the risk-informed surveillance requirements, the
24 risk-informed surveillance test intervals, risk-informed
25 allowed outage times with risk-informed actions, and then
. 285
1 the fully risk-informed technical specifications that after
2 Sharon talks about the seven existing initiatives I will
3 talk a little bit more about those, which are very
4 conceptual at this point in time.
5 You can see what these do as you go across from
6 left to right on this slide is the degree of complexity and
7 the demands on the PRA increase. What we would like to be
8 able to do -- I think we have alluded to this earlier -- is
9 that as you go across these set of initiatives, it gives the
10 people who are implementing them or the Licensees who are
11 implementing them a cafeteria, if you will, of options.
12 People who may not have quite as sophisticated a
13 probabilistic risk assessment or some of the older plants or
14 what have you may implement those line item changes that you
15 see there on the left.
16 As people have more sophisticated PRAs you can see
17 that they will be able to implement the things further to
18 the right on the chart.
19 What we are looking for here in the items, these
20 first six or seven items is some early successes that we can
21 apply across the board. We feel that is important for the
22 continued support of the industry.
23 For instance, I will just speak in terms of the
24 Westinghouse Owners Group. The Westinghouse Owners Group
25 has been very supportive of risk-informed regulation. We
. 286
1 have had a number of initiatives that Sharon will talk about
2 in 4(a) for line item improvements that the WOG has funded
3 and supported and they have been very supportive of that.
4 We need to show them that these things are going
5 to progress, that they will continue to be supportive of
6 these risk-informed initiatives.
7 We are looking for a phased approach here that
8 builds one items on the other, and I think you can see that
9 is what we have here. We are talking about the consistency
10 with the overall regulatory approach and Biff talked a
11 little about that with 4(a) of 50.65, the maintenance rule.
12 We think this fits in very well with 4(a) of 50.65 of the
13 maintenance rule. It fits in with the new inspection
14 process and we feel that it fits in very well with
15 risk-informing Part 50, where we could really focus on those
16 things which are really important to our plants.
17 The allowed outage times and risk-informed actions
18 but not the exceed times are really the first, I would say
19 the first complex step in this, where I think, I would say
20 that you really get to the new generation of technical
21 specifications and certainly the risk-informed, fully
22 risk-informed technical specifications represents a very
23 significant change in technical specifications, and we need
24 to assess as we go through this do you get 90 percent of the
25 benefit at the risk-informed allowed outage times and
. 287
1 risk-informed actions and is it fully risk-informed with the
2 rule changes and so forth -- is that going to be worth it to
3 us for that incremental increase in effort and cost?
4 I guess the only other thing I would point out
5 here, just as sort of a final statement, is we are working
6 within the existing process for generic technical
7 specification changes. We used the technical specification
8 task force to process these changes, so we are working with
9 a framework both the industry and the Nuclear Regulatory
10 Commission is familiar with.
11 That basically concludes my discussion of the
12 overall vision.
13 DR. APOSTOLAKIS: I have two questions.
14 MR. HARRISON: Sure.
15 DR. APOSTOLAKIS: We keep talking about burden
16 reduction. Can you give us an idea what kind of burden we
17 are talking about here that would be eliminated?
18 MR. HARRISON: I think one example I like to use
19 on burden reduction is primarily with regard to the
20 surveillance requirements and the surveillance test
21 intervals.
22 I think that the surveillance requirements and the
23 surveillance test intervals were basically put into the
24 technical specifications based on the best judgment, best
25 engineering judgment at the time, and the surveillance test
. 288
1 intervals may or may not have, and I think in many cases do
2 not for instance have a statistical basis for how frequently
3 you do them.
4 I think if we can identify those surveillance
5 requirements that really are important to the safe operation
6 of the plant, and separate them from the ones that we use
7 for routine assurance of reliability, and then relocate the
8 surveillance test intervals to a Licensee-controlled
9 program, that allows a Licensee to have realistic
10 surveillance test intervals. It allows a Licensee to
11 determine what resources they need to put on surveillance
12 test intervals and that can translate to actual burden
13 reduction for each licensee.
14 It adds to safety, as we discussed before, because
15 you very well may be adding to the availability and
16 reliability of that particular component, so that is I think
17 a good example of where you have a regulatory burden
18 reduction and an increase in safety.
19 DR. APOSTOLAKIS: But how much would the burden
20 reduction be in terms of dollars?
21 MR. HARRISON: That's very hard to quantify.
22 Sharon thinks that she can talk to that.
23 DR. SEALE: Could I ask a question in that regard?
24 The kind of test intervals you are talking about
25 would -- let's say in the initial establishment of what the
. 289
1 test interval would be, you might take into account the kind
2 of equipment that were involved, whether it was analog or
3 digital or so forth. Let me ask the Staff or whoever, if
4 you make a change to a different technology equipment, is
5 that question of test interval open for redefinition or has
6 it remained at the old frequency?
7 MR. BECKNER: This is Bill Beckner again.
8 That is I think one of the reasons we have the
9 relocation of intervals up here. A tech spec number is a
10 license requirement that must be met --
11 DR. SEALE: Yes.
12 MR. BECKNER: -- unless they come in for a tech
13 spec change, and if Licensees change their hardware of
14 course they could come in with a license amendment request
15 and convince the Staff it is the appropriate thing to do.
16 DR. SEALE: They could do that?
17 MR. BECKNER: Correct. Another way to do it, if
18 we can convince ourselves that within the rule this can be
19 the type of information that can be put in the Licensee
20 control document, controlled under 50.59, that provides the
21 Licensee with the flexibility again, but with hopefully no
22 reduction in safety given that 50.59 is an adequate process.
23 DR. SEALE: Well, in general I think the more
24 modern technology is less susceptible to the kinds of things
25 that your surveillance program is designed to capture, so
. 290
1 you need that flexibility whether it has been there in the
2 past or not.
3 MR. BECKNER: I think we are in agreement.
4 DR. BONACA: But on the issue of cost savings, if
5 I look at the current initiatives alone, I mean there are
6 opportunities for millions of dollars of savings on
7 individual decisions.
8 MR. HARRISON: That's correct.
9 DR. BONACA: I mean you have conditions where you
10 may not shut down where today you would shut down.
11 MR. HARRISON: That is correct.
12 DR. BONACA: There are conditions where you stay
13 in hot standby where today you have to go to cold standby.
14 I think those are big --
15 MR. HARRISON: That is correct. You are
16 absolutely correct.
17 I would also point out on the surveillance test
18 intervals there is a certain amount of precedent that we
19 have already begun with that, with the inservice testing for
20 the test intervals are now being placed under Licensee
21 control and risk-informed inservice testing.
22 DR. APOSTOLAKIS: Now two more questions. Are all
23 these changes quantifiable in the PRA? Let me explain that.
24 We have found that the requests that have come in
25 for the changes, you know, when the Regulatory Guides were
. 291
1 developed, fall into one of two classes.
2 In one class you can actually quantify the change
3 in CDF and LERF -- for example, if you change the test
4 intervals. We have equations for those things. You go back
5 there and you get a number. That is one class.
6 The second class you can't quantify the change and
7 a good example is quality assurance. In that case we went
8 around it and said, well, we will develop categories for the
9 components and systems and structures that will be based on
10 risk measures, importance measures, and then we will change
11 the QA requirements to the low risk significant category or
12 the non-risk significant category. In fact, your
13 organization has done a lot of work in that area. 14
Then the assumption is that it really doesn't
15 matter if you remove those requirements or relax them
16 because we can quantify the impact but it stands to reason
17 that because of the relative unimportance of these SSCs it
18 is a good thing to do.
19 Now are we going to have any situation like that
20 here with the seven initiatives or all of them can be
21 quantified and give a delta CDF and delta LERF?
22 MR. HARRISON: It varies. Some of these don't
23 very well lend themselves to quantification. We'll see that
24 as we go through --
25 DR. APOSTOLAKIS: Okay.
. 292
1 MR. HARRISON: -- the individuals and some do. I
2 think in most cases you can at least do a risk assessment of
3 a confirmatory sense to at least say yeah, from what we know
4 and are looking at we can satisfy ourselves that it is not
5 risk significant.
6 DR. APOSTOLAKIS: So we will have to go back to
7 these categories of SSCs in those cases and make a judgment
8 there, okay.
9 The last question has to do with a timeline you
10 have there. You are starting in January of '99 -- right?
11 MR. HARRISON: No?
12 DR. APOSTOLAKIS: That's what it is.
13 MS. MAHLER: That is intended to be 1999 through
14 2001.
15 DR. APOSTOLAKIS: Oh.
16 MS. MAHLER: 2000 through 2001.
17 DR. APOSTOLAKIS: Okay. Shall we move on now?
18 MS. MAHLER: My name is Sharon Mahler, and I work
19 with NPPD. My intent is basically to run through the
20 concepts of the seven initiatives that the industry has
21 undertaken. We are trying to stay at the conceptual level.
22 Some of these have been more fleshed out than others of the
23 issues, but since this is your first introduction we wanted
24 to sort of lay them all out for you.
25 The first initiative --
. 293
1 DR. APOSTOLAKIS: Don't assume we know much.
2 MS. MAHLER: The first initiative is called end
3 states. As we spoke earlier, one of the fundamental
4 presumptions in current tech spec space was that it defined
5 cold shutdown as your safest end state or for what we call
6 the default action, to go down to cold shutdown.
7 What this particular initiative is doing is really
8 looking at that presumption with the new risk tools that we
9 have available. Let me back up.
10 What we are finding and basically even before we
11 started using the risk tools one of the things we found,
12 discussing it from a traditional engineering standpoint, is
13 actually hot shutdown is a more resource-risk state.
14 For example, we have RHR and aux feed available
15 for PWRs versus just RHR in cold shutdown -- those kinds of
16 things -- so that is sort of what led us to, hey, we need to
17 be looking at this.
18 SONGS has drafted, has a draft analysis, which
19 supports that indeed hot shutdown may be the less risky
20 place to be, so that is basically what we are looking for in
21 this initiative, what we are looking at.
22 DR. APOSTOLAKIS: Can you explain the last
23 statement there -- being hot standby in six hours or would
24 go to hot standby in six hours?
25 MS. MAHLER: That is how the standard tech specs
. 294
1 are set up. It basically says that if your required action
2 and associated completion time is not met be in -- they use
3 "modes" -- but we tried to use normal terminology, if you
4 will -- be in hot standby within six hours and be in cold
5 shutdown within 36. That is what the current specs say.
6 What we are trying to show you is how it would be
7 revised.
8 DR. APOSTOLAKIS: Oh.
9 MS. MAHLER: It would now say be in hot standby in
10 six hours or be in hot shutdown in 12 hours.
11 DR. SEALE: And how long could you stay in hot
12 shutdown?
13 MS. MAHLER: Indefinitely.
14 DR. SEALE: And that is PTS is one of the things
15 you avoid if you stay in hot shutdown.
16 MS. MAHLER: Yes.
17 DR. BONACA: So you feel this may be one situation
18 where it is a win/win situation?
19 MS. MAHLER: It basically represents a risk
20 decrease.
21 DR. BONACA: You may actually have a reduction in
22 risk but you also have the benefit?
23 MS. MAHLER: It does save time and cost relative
24 to the additional testing that comes into play when you go
25 into cold shutdown as well as once you get the thing fixed,
. 295
1 time to come back up.
2 DR. APOSTOLAKIS: So you would apply Regulatory
3 Guide 1.174 to this? How would you review it? Or is it a
4 matter of just agreeing?
5 MR. REINHART: Again, we would go back to those
6 five principles in the review, have an integrated review,
7 but in doing this look at a comparison between the risk in
8 the operating mode versus risk in shutdown or maybe risk in
9 cold shutdown versus risk in hot shutdown, and there would
10 be quantitative as well as qualitative arguments.
11 San Onofre has already submitted and had approved
12 one change. It was on the diesel -- where they made a case
13 for it was safer to stay in a mode other than cold
14 shutdown -- by the Staff.
15 DR. APOSTOLAKIS: Do you plan to come before this
16 committee as you review these initiatives or after you
17 approve them? Are we going to see the process at work?
18 MR. REINHART: My expectation is that as the
19 improved standard risk-informed tech spec work goes on that
20 there would be periodic visits to the ACRS.
21 Let me as, Tech Spec Branch.
22 MR. BARRETT: Always.
23 MR. REINHART: Yes. That's the expectation.
24 MR. BECKNER: I think that is a good question. I
25 said earlier that we would definitely put this out for
. 296
1 public comment, and it is a good question.
2 Does the committee want to also -- you obviously
3 can see them, but do you want to be explicitly asked to
4 review them? I think we would like to hear --
5 MR. MARKLEY: I think the committee needs to
6 decide that today. That's part of the mission of this
7 subcommittee is to, you know, sort it through whether they
8 want to evaluate the risk assumptions and analysis that goes
9 with each one of these proposals and which ones of the
10 meetings they want to set up down the road and whether it
11 even goes to the full committee or whether they just keep it
12 in the subcommittee for awhile.
13 DR. APOSTOLAKIS: So at the end of today's
14 meeting we will discuss it with our members what the
15 recommendation to the full committee will be. Okay.
16 But the Staff is willing to come and talk to us?
17 What did you say?
18 MR. BARRETT: Always.
19 MR. MARKLEY: Barrett is the ultimate politician.
20 [Laughter.]
21 MS. MAHLER: The next initiative is associated
22 with what you do when you find a missed surveillance.
23 Today in current tech specs when a Licensee finds
24 that they have missed a surveillance, they have 24 hours to
25 perform the surveillance and in that 24 hours, depending on
. 297
1 what the surveillance is and that kind of stuff, you may
2 have to write special procedures. You may have to call in
3 your maintenance people, get your plan together, get it
4 approved, and also perform the surveillance.
5 If you don't get it within that 24 hours and/or in
6 some case try and get an NOED, if you don't do it within
7 that 24 hours then you have to declare the LCO not met,
8 enter the actions, and proceed down that way.
9 Generic Letter 8709 indicated in the vast majority
10 when you do find a missed surveillance that when you finally
11 do perform it that the surveillance is passed, and in
12 looking in current Licensee performance that still holds the
13 case.
14 The concept is to allow the Licensee to have time
15 if they find a missed surveillance to systematically look at
16 when is the first reasonable opportunity, and that
17 determination of reasonable opportunity includes the
18 consideration of risk if it is a surveillance that you would
19 actually have to change mode to perform and that kind of
20 stuff, is that the appropriate thing to do balancing the
21 importance of that particular surveillance.
22 One thing I do want to emphasize is that missing
23 these surveillance and that kind of stuff does not happen
24 very often. Typically what it is when you are going through
25 the paperwork and comparing it in detail to the specs, you
. 298
1 may have had to look at five contacts and you found in this
2 particular case that you only looked at four, and those
3 kinds of things. It is not just that you did not perform
4 the surveillance. That is Concept or Initiative 2.
5 The next initiative hast to do with mode restraint
6 flexibility. Again, current tech specs in some cases does
7 not allow you to go into a mode of applicability without
8 having the LCO met, but if you were at power and all of a
9 sudden a particular component went inoperable, you are
10 allowed to stay in that mode for a certain period of time
11 defined by the actions, so what we are looking at is to
12 allow a Licensee to value the risk of going up in that
13 particular mode and if it is acceptable enter the mode and
14 go ahead and enter the associated action.
15 You were talking about examples of cost savings.
16 In the November meeting with the Staff we provided a couple
17 of examples, and in this particular case there was a CE
18 plant that had containment spray pump that was inoperable.
19 Everything else was going along fine. Everybody was working
20 to get the plant online. The containment spray pump -- they
21 were doing some repacking on it, which it was the best place
22 to do that was in Mode 4 before they proceeded up to Mode 3.
23 However, there were a few difficulties with that, and in the
24 November meeting we showed that we potentially could have
25 saved 48 hours worth of critical path time on the outage,
. 299
1 which equated to $2 million, so there are a couple of
2 examples in the November meeting minutes on cost saving
3 relative to both Initiative 2 and Initiative 3.
4 DR. SEALE: If I could make a comment on the
5 Initiative 2. You earlier stated that this wasn't a regular
6 occurrence. It was relatively rare. I might suggest that
7 the reason it is relatively rare is that no one wants to go
8 through the black and blue spots associated with declaring
9 that a limiting condition of operation has not been met, and
10 so you make sure you get these things done, and the question
11 then is if you make this modification are you getting rid of
12 the goad that pushes you to do these things at the time and
13 I don't expect you to say anything about it right now.
14 It is just that that is the kind of question that
15 might come up.
16 MS. MAHLER: Obviously when you miss a
17 surveillance with the new reactor oversight program and all
18 that stuff and even before then we would be entering that in
19 our corrective action program, and if trends were showing up
20 it would manifest itself in there.
21 MR. HARRISON: I would add that you are still
22 going to have to some extra planning and work that you
23 weren't otherwise going to have to do.
24 We do these things in a very structured manner.
25 DR. SEALE: Sure.
. 300
1 MR. HARRISON: And I really think that will be the
2 case.
3 MS. MAHLER: Initiative 4 has to do with
4 risk-informed allowed outage times. The first portion of
5 that the industry has been and will continue to submit
6 individual AOT extensions utilizing the risk Reg. Guides.
7 Successes have been diesel generator allowed outage times,
8 LPSI pumps, safety injection tanks, and where possible, why
9 those typically come in on an Owners Group basis, individual
10 industry Owners Groups are sharing that information to
11 maximize our benefit there.
12 The second portion, the concept is that the
13 allowed outage times in the specifications would be
14 dependent upon information from the risk management program
15 with a not to exceed time limit. This is certainly one
16 place where (a)(4) and the tech specs, we are looking at
17 bringing the things closer together there.
18 DR. APOSTOLAKIS: So what exactly does 4b mean, it
19 would be dependent on information from the risk management
20 program?
21 MS. MAHLER: To give you an example, if your
22 condition required action, then completion time would look
23 something like this. Pump A is inoperable, your required
24 action would be restore pump A to operable status. Your
25 completion time would be something of the nature of as
. 301
1 determined by the risk program and 30 days from discovery of
2 failure to meet the LCO.
3 So you would basically have, if your risk
4 management program said, for the configuration that the
5 plant is in at that particular time, you needed to restore
6 it within seven days, then you would need to restore it
7 within seven days. If your configuration risk management
8 program said you could potentially have that out for 60 days
9 with the configuration that you are in, then the not to
10 exceed limit of 30 days would say you still had to have it
11 restored within 30 days.
12 DR. APOSTOLAKIS: And how would the not to exceed
13 limit be determined?
14 MS. MAHLER: That would be by the risk management
15 program, similar to the (a)(4) and the configuration --
16 DR. APOSTOLAKIS: No, but I mean you just said
17 that the risk management program might say 60 days is okay,
18 but the not to exceed limit might be 30 days. So I was
19 wondering where the 30 days came from.
20 MR. DENNIG: One of the things that Mark Reinhart
21 mentioned earlier in a balanced review of these things, I
22 think he touched on the need for the time. Part of what
23 goes into this is a justification. If previous experience,
24 repair experience with a particular piece of equipment would
25 show that you can do a total repair/replace, whatever, the
. 302
1 most it has ever taken is 15 days or 20 days or whatever,
2 that there is no practical reason to push it out
3 significantly beyond what experience has shown is the needed
4 envelope to get the equipment restored.
5 So as far as factoring in where does that 30 days
6 backstop come from, I think operating experience and the
7 actual evolution that is being envisioned, given the state
8 of the plant, of the equipment, circumstance comes into that
9 determination.
10 DR. APOSTOLAKIS: Why would this be so important
11 in this case? I mean if the risk management program says
12 that 60 days can be allowed without increasing risk, I mean
13 you are bringing now into the argument a separate kind of
14 thinking. You are saying, well, you know, we really don't
15 want it to go beyond 30 days, and we have observed in the
16 past that within that time period, all these activities have
17 been completed, but they have been completed under
18 assumptions on the part of the licensee, which now we are
19 changing. I thought you would give us a defense-in-depth
20 kind of argument.
21 MR. HARRISON: I can give you a little bit of a
22 risk insight kind of argument. Your risk management program
23 is going to be looking at the configuration of the plant at
24 that moment in time, which may be at that moment in time
25 very -- maybe pristine, that you have completed all your
. 303
1 maintenance, everything is 100 percent operable and so
2 forth. So, based on what your risk management program looks
3 at that point in time, it says 60 days is fine. Well, over
4 the course of that 60 days, the configuration of that plant
5 may change, something else may change state or
6 what-have-you, so that is another thought of looking at a
7 shorter backstop time, it kind of, in a sense, takes into
8 account the changes in state that may occur.
9 DR. APOSTOLAKIS: But it seems to me that it would
10 be a good to determine these limits also using risk
11 arguments.
12 MR. BRADLEY: Yeah, just to reinforce your point,
13 I mean the point Wayne made, if you have an emergent
14 condition of a change in mode, or what-have-you, that
15 changes the risk evaluation you previously made, the (a)(4)
16 guideline is going to require you to reassess that. So that
17 is -- in my mind, I mean would tend to agree with you that
18 this would seem to be arbitrary. I mean the (a)(4) guidance
19 we developed certainly doesn't include imposing some
20 additional time limit beyond what all those insights are
21 telling you. I think this is something worth further
22 discussion.
23 DR. APOSTOLAKIS: Yeah. I mean if we are to
24 change things, I think we should be careful not to inject
25 into the new process bad elements from the past. So if
. 304
1 there is a risk argument to be made, or as Mr. Harrison just
2 said, other considerations, I think there should be some
3 reason why a particular limit was chosen. And, in fact, I
4 mean based on risk arguments, you may not even have a limit
5 for every single AOT.
6 MR. BRADLEY: This is also an example I think
7 where the (a)(4) type approach, the risk management approach
8 is a much smarter way to address AOTs than tech specs,
9 because tech specs looks at AOTs individually, it is not
10 looking at cumulative risk. The (a)(4) evaluation of the
11 configuration risk management program is looking at it in a
12 much more integrated fashion and there is really, as you
13 know, nothing in tech specs to preclude you from repeatedly
14 entering an AOT, whereas, under the (a)(4) program, you do
15 have to consider that impact. So it is just a smarter way
16 to go.
17 MS. MAHLER: As we talked earlier, Initiative 5,
18 the concept here is basically to optimize the surveillance
19 requirements that are in the tech specs, and that is broken
20 into two parts. The first part includes relocating out of
21 the tech spec some of the surveillance requirements, and
22 then relocating the surveillance test intervals out of the
23 tech specs to a licensee controlled program such that the
24 surveillance test intervals could be modified based on risk.
25 And this is very similar to what has been done in
. 305
1 risk-informing the IST intervals.
2 DR. APOSTOLAKIS: Excuse me, let me go back to
3 Initiative 3. Would that be a candidate for the second
4 class that I mentioned earlier, where we cannot really
5 quantify the impact to risk -- on risk?
6 MS. MAHLER: In some cases, you may be able to
7 quantify it, in other cases, you may not. It all depends on
8 what the nature of the component is that is out. In the
9 particular example we showed in the November meeting for
10 containment spray, the people were able to, for the purposes
11 of the meeting, run some risk numbers and show that it was
12 extremely small. But there may be other cases for
13 components in the tech specs that you wouldn't be able to
14 quantify them.
15 DR. APOSTOLAKIS: But in all other initiatives, it
16 seems that we can quantify it, is that correct?
17 MS. MAHLER: You may have some difficulty on the
18 missed surveillance one.
19 DR. APOSTOLAKIS: Yeah.
20 MS. MAHLER: Issue 2 and --
21 DR. APOSTOLAKIS: Well, okay, probably not, but --
22 MR. HARRISON: The missed surveillance was similar
23 to what Mark Reinhart was talking about earlier where some
24 of the initiatives, you can sort of quantify on the front
25 end, but others, it is really you do the quantification or
. 306
1 the judgment perhaps on the back end after implementation.
2 MS. MAHLER: Initiative 6 is similar to Initiative
3 1 in that the current technical specifications have buried
4 within, if you will, the original presumption that if you
5 came up on a condition not covered by the actions, or you
6 lost function, or as you were directed by the spec, that the
7 safest thing was to commence a shutdown within an hour. The
8 concept of this initiative is basically to allow time to
9 initiate and develop a risk-informed course of action if you
10 find yourself in a situation where you have lost function or
11 you don't have a condition in the specifications already
12 specified. And this is where potentially you are talking
13 about a risk decrease again in implementing this concept.
14 And Initiative 7 is basically to provide
15 appropriate actions for equipment that is inoperable but
16 still functional.
17 DR. APOSTOLAKIS: Can you elaborate a little?
18 MS. MAHLER: In some cases you may find where
19 there is what we refer to as a paper problem and/or it may
20 have missed, its surveillance acceptance criteria was 61.1
21 and you found yourself at 61.1. -- you know, 61.15 kind of
22 thing, where there are still conservatisms in that set point
23 or in that particular acceptance criteria, and so you know
24 you are still functional, but from the way the tech specs
25 are set up, you are necessarily fully operable.
. 307
1 DR. APOSTOLAKIS: And how will these appropriate
2 actions be determined?
3 MS. MAHLER: This is one that is still in the very
4 conceptual stage.
5 And now I will let Wayne talk about the fully
6 risk-informed specifications.
7 DR. APOSTOLAKIS: So long-term here means beyond
8 '03, 2003, is that what it means?
9 MR. HARRISON: Yes, sir.
10 DR. APOSTOLAKIS: Okay.
11 MR. HARRISON: Sharon talked about conceptual.
12 Certainly, the long-term intiative the fully risk-informed
13 technical specifications is very conceptual at this point.
14 We have just barely started to put pen to paper on it. And
15 it essentially builds on the 4b initiative which is the one
16 we just discussed that had the backstops, if you will. In
17 the case, the limiting conditions for operation would be
18 dictated by the plant specific PRA safety functions. This
19 is a situation where I think we would envision this as
20 probably not having a backstop implemented or involved.
21 I think the key point to the LCOs being dictated
22 by the plant-specific PRA safety functions is that they are
23 not necessarily driven by the design basis accident as we
24 know it now, it is really to look at the risk significance
25 of that particular state.
. 308
1 The plant's risk management program would control
2 the allowed outage times at the system or train level, and I
3 think that, again, is consistent with the implementation of
4 (a)(4) of the maintenance rule. And Dr. Apostolakis asked,
5 well, how is this different or how is from our existing tech
6 specifications? And I think it preserves or improves our
7 existing level of safety. But it is going to be defined, as
8 we mentioned earlier, or these states would be defined by
9 the core damage probability, and that might be defined by
10 the core damage probability on a yearly basis or on a weekly
11 basis, where you would quantify the plant risk.
12 Our actions would be specified based on a risk
13 criterion in traditional engineering. Certainly, it is
14 going to be -- there would be some things that we are still
15 trying to identify as to how we would handle them under risk
16 and maybe they won't be, for instance, the safety limits of
17 the technical specifications, where you specify, well, you
18 will never have your reactor coolant system go over 27 -- I
19 think it is 2735 PSI. And I think the other things that we
20 keep within the -- we would probably keep the limiting
21 safety system settings like the reactor trip set points and
22 so forth. And perhaps we would look at keeping the activity
23 within the reactor coolant system at a certain level,
24 because these sort of form a foundation or an underpinning
25 of your basic assumptions that are not necessarily
. 309
1 risk-informed, although I may stand to be corrected as we go
2 through and actually do this evaluation.
3 DR. APOSTOLAKIS: A question. Can you give us an
4 example of something that you will be able to do when you
5 have fully risk-informed specifications that you will not be
6 able to do, assuming that all seven initiatives are
7 approved?
8 MR. HARRISON: Well, that is a good question,
9 because if you said except for --
10 DR. APOSTOLAKIS: The only ones we ask you.
11 MR. HARRISON: Well, that is why we are here, so
12 we can get these good questions. If you could say the first
13 six initiatives, I can answer that question.
14 DR. APOSTOLAKIS: Oh, this is the seventh, I'm
15 sorry, the first six.
16 MR. HARRISON: The first, well, this is -- no,
17 this is Number 8.
18 DR. APOSTOLAKIS: Or 8.
19 MR. HARRISON: And there may be some -- and that
20 will, that was what I mentioned earlier, that we need to
21 look at the incremental improvement over the 4b initiative
22 on the other risk-informed allowed outage times. This, how
23 much, what can you do with this that you wouldn't be able to
24 do with that? I think the answer to that is, well, you are
25 not going to have a backstop that you have to contend with
. 310
1 probably, but I don't know how of benefit that would be.
2 DR. APOSTOLAKIS: That limit we discussed.
3 MR. HARRISON: Right. Or the limit we discussed.
4 DR. APOSTOLAKIS: Okay. That doesn't sound like a
5 major change.
6 MR. HARRISON: Right. But what this will allow
7 you to do over the other seven is that it really lets the
8 licensee manage the configuration of the plant. With this,
9 I think it is really hard for me to envision a situation
10 where we would ever have to go back to the Nuclear
11 Regulatory Commission and request enforcement discretion for
12 technical specifications, if the licensee were to implement
13 this, because the risk management program basically tells
14 him where he needs to be.
15 DR. BONACA: But still, I mean you would have --
16 you are assuming a world where you still have your accident
17 analysis in the FSAR, that represents the phases, that comes
18 in the tech specs as initial conditions. You have your
19 safety system settings. I mean it is in EPRI.
20 MR. HARRISON: Well, it goes beyond -- but it goes
21 beyond that in this case.
22 DR. BONACA: But I say there is a hybrid that you
23 are looking at, right?
24 MR. BRADLEY: I think this is somewhat analogous
25 to the issue, but less the difference in Option 2 and Option
. 311
1 3 of Part 50. I think what we are looking for here is more
2 -- one reason this is long-term is you are probably talking
3 about rulemaking to 50.36, to look at those four criteria.
4 Here you may even be able to show that some of the things
5 that have current, you know, requirements coming out of your
6 deterministic design basis wouldn't need to be in the tech
7 specs at all, whereas, there are other things that maybe
8 aren't there that would be. It is basically just bringing
9 the tech specs fully into the realm of risk management, you
10 know, whereas, I think with the other stuff, you would have
11 some vestigial framework of what you have now.
12 And you can't really change that until you change
13 50.36 because you are required by the first three criteria
14 to have specs for all your design basis considerations.
15 DR. APOSTOLAKIS: Do the first seven initiatives,
16 do they affect safety margins at all? In terms, you know,
17 of the safety limit.
18 MR. HARRISON: I think we would say, no, they
19 don't.
20 DR. APOSTOLAKIS: No. Would the fully
21 risk-informed specifications do that? You will start
22 attacking now other areas?
23 MR. HARRISON: Yes.
24 DR. APOSTOLAKIS: So that would be a major change.
25 DR. BONACA: It gets into Part 50.
. 312
1 MR. HARRISON: It would get into Part 50. In
2 other words, as we say the DBA, or the design basis
3 accident, for instance, a double-ended break would no longer
4 be the defining event for a number of these technical
5 specifications. So would you affect the margins to that
6 particular event? Yes, but there would be certainly
7 improvements in other areas.
8 DR. BONACA: But then in the case -- it is not
9 really a changing tech spec, it is changing your FSAR. You
10 are changing your accident analysis, you are changing then
11 -- and then, because of that, then you have a step to this.
12 So that is really a jump.
13 DR. APOSTOLAKIS: This would have the fully -- the
14 long-term initiative would have to be combined with other
15 long-term initiatives like Option 3 that Biff mentioned.
16 MR. BRADLEY: Specifically, if you look at
17 SECY-264, which is the staff SECY on Option 3, there is a
18 discussion in there of Option 3 changes to special treatment
19 regulations. And I think you consider to be 50.36 to be one
20 of those. And it basically talks about just generally the
21 philosophy that things like Maintenance Rule 84 tends to
22 overlap tech specs.
23 You can change some of that in Option 2, but to go
24 to Option 3, you can fully realize that, which I think is
25 what this is all about. And there are other issues, too,
. 313
1 you know, where as like the monitoring programs of the
2 Maintenance Rule can be viewed to overlap QA, where you are
3 building in the same assurance on the front end or the back
4 end, and it may be possible to integrate those two.
5 But the SECY-264, in Option 3, I think discusses
6 the type of change to tech specs that we are talking about
7 here.
8 MR. GRANTOM: George, if I could. Leading
9 directly into --
10 DR. APOSTOLAKIS: You have to come to the
11 microphone. This is the microphone, right?
12 MR. GRANTOM: Well, leading into what Biff --
13 DR. APOSTOLAKIS: Who are you? Who are you?
14 MR. GRANTOM: Rick Grantom. What -- this fully
15 risk-informed thing pulls in a number of regulations, and
16 Biff is talking about, into one integrated set. And what it
17 does is it allows the utility the flexibility to work within
18 a threshold, a risk threshold and that can be tied to safety
19 goals. So all the equipment, all the unavailabilities, all
20 the performance aspects are rolled into an integrated
21 pictorial view that the utility operates under a given
22 threshold.
23 Everything that you have seen on the lefthand side
24 is, in a sense, subsumed by going to this fully
25 risk-informed method, and it puts it into a context and a
. 314
1 much abbreviated and shorter version. I mean we have talked
2 about these conceptual one page tech specs, but it is very
3 much abbreviated. At that point in time still conserves all
4 the defense-in-depth aspects, but now provides a pictorial
5 and actual measure of risk, quantifiable measures of risk
6 that can be used to operate a weekly threshold and,
7 cumulatively, under a yearly threshold.
8 So at this point in time, then you are getting to
9 the situation to where now you can actually start to risk
10 manage a nuclear power plant. The regulator can risk manage
11 a fleet of plants at that point in time, and this is the
12 segue that allows a fully risk-informed approach here.
13 DR. APOSTOLAKIS: So we are talking really about a
14 truly risk-informed regulatory system.
15 MR. GRANTOM: Exactly. At that point in time,
16 combined with some of the other deterministic things we
17 talked about.
18 DR. APOSTOLAKIS: Very good. Thank you.
19 MR. HARRISON: There would be very few limiting
20 conditions for operations that would be left. We mentioned
21 the safety limits. We still have to figure out what are we
22 going to do with the fuel limits. There may still be
23 limiting conditions for operations for those kind of things.
24 DR. APOSTOLAKIS: Are there any documents that
25 describe these initiatives in more detail, like you gave us
. 315
1 a few example and so on, that we --
2 MR. HARRISON: I think we had given some examples
3 of that in the handouts from the November meeting. I can
4 show you a couple of things where we basically corrected
5 some typos in those, but those are fundamentally where that
6 is at this point.
7 MR. REINHART: The two submittals.
8 DR. APOSTOLAKIS: We have not seen those, right.
9 We have not seem them.
10 MS. GILLES: Right. The industry has submitted
11 Initiative 2 and 3 to the staff for review and we are
12 working on the other ones. So, as we progress, we will be
13 submitting to the staff under the TSTF process.
14 MR. GRANTOM: Rick Grantom again. But we do have
15 conceptual slides of how this would work. It hasn't been
16 endorsed yet on the list, but at South Texas, we have
17 actually looked at some of these things with actual risk
18 numbers where one can actually postulate and show that. And
19 there is information out there that can be looked at along
20 that line.
21 MR. REINHART: Would you like copies of the
22 submittals?
23 DR. APOSTOLAKIS: Let's wait and see what the
24 subcommittee will decide to do, and at that point we will
25 decide that.
. 316
1 Now, you were going back to the time here.
2 MR. HARRISON: I was going back to the timeline
3 basically to sum up, to recap. What we have gone through
4 here is a phased approach. We have seen the increasing
5 demand on the probabilistic assessment as the initiatives
6 progress. And as Biff has pointed out as we have gone
7 through here, it is consistent with the overall
8 risk-informed strategy.
9 DR. APOSTOLAKIS: Do you think we have the
10 infrastructure to implement this?
11 MR. HARRISON: I think that is part of what we are
12 doing as we go through here, Dr. Apostolakis. We are
13 developing the infrastructure. I think the infrastructure
14 exists to varying degrees at certain plants, plants like
15 South Texas and San Onofre. I think we could probably go
16 straight to Initiative 4 Bravo or even the fully
17 risk-informed technical specification. We have the
18 technology. I think the staff probably has the technology
19 to do that review.
20 Do we have the regulatory infrastructure within
21 the rules as they exist right now? No, we do not. I think
22 we can probably get to 4 Bravo within the existing
23 regulatory infrastructure.
24 DR. SEALE: I wrote myself a note here that said
25 that Biff used the word, deluge, and it was -- and my
. 317
1 comment was that one man's deluge was another man's April
2 shower.
3 [Laughter.]
4 DR. SEALE: And I think that's really true. Some
5 utilities have the kind of PRA organization that could field
6 a large part of this effort rather rapidly, whereas other
7 utilities are essentially off the chart. I mean, they
8 really haven't started looking at this with any kind of
9 self-imposition of these things at all.
10 MR. HARRISON: Personally, I'm kind of speaking
11 for the rest now, since I think that some of the owners
12 groups are of the same opinion. We may be able to use
13 owners groups to address that to a certain extent, and take
14 advantage of what some plants are developing, some generic
15 applications that the others can take advantage of.
16 DR. SEALE: I have another question, and it's to
17 the staff: I notice here that there are a lot of things
18 which are being transferred to the licensee control program,
19 things that no longer would be within the tech specs or
20 whatever normally, but would rather be a surveillance
21 interval program or whatever it would be, up to the licensee
22 to control according to the ground rules.
23 That means then that those activities come under
24 the purview of the inspection program. Are you keeping a
25 list of all of the things you're dumping on the inspectors
. 318
1 as you go through this?
2 Have you done a job task analysis in the old sense
3 that we did when we looked at what the requirements were for
4 a licensed operator, or an inspector that would have to live
5 in this environment?
6 And have you decided whether or not your
7 inspectors are, a) have enough time in the day, and b) have
8 the skills, personally, at the level they need in order to
9 field your half of this program?
10 MR. BARRETT: Well, I think there are a couple of
11 different answers to that question. That's one of the
12 biggest questions right now related to the entire oversight
13 process.
14 But the answer to your question is yes; as we go
15 into more and more applications regarding risk-informing the
16 regulations, the regulatory process -- I'll take this as an
17 example: Let me give you an example of decommissioning.
18 In the area of decommissioning, as we are defining
19 what we think will be the new regulatory requirements there,
20 we're trying to define specific performance measures that
21 are risk-informed, that can be monitored by the licensees
22 themselves, and can be reported to the NRC so that hopefully
23 it won't be necessarily an increased inspection burden, as
24 much as an increase -- as an effort to audit the licensee's
25 oversight and monitoring, their own monitoring and reporting
. 319
1 requirements.
2 DR. SEALE: But the virtuosity of your inspector
3 has got to increase as you dump these different kinds of
4 things in his lap.
5 MR. BARRETT: That's certainly true. As I look at
6 this, for instance, we're talking about taking specific
7 surveillance test intervals. There can be a quantitative
8 basis for surveillance test intervals, and so that's
9 something that could be monitored quantitatively, possibly,
10 as Biff said earlier, through compliance with the
11 Maintenance Rule, reliability and availability.
12 There you're basically making a tradeoff between
13 reliability and availability. So it may not be as much of
14 an inspection burden as a new quantitative -- I want to use
15 the word, performance indicator, and perhaps that's too
16 grand a term to use.
17 But your concern is well understood, and you're
18 absolutely right; the skill set of the inspectors and the
19 burden on the inspectors is a very important consideration
20 in all of these areas.
21 DR. APOSTOLAKIS: Are you done, Mr. Harrison?
22 MR. HARRISON: Yes, sir, I am.
23 DR. APOSTOLAKIS: Okay, I have a few questions on
24 the meeting summaries that the staff developed, and then we
25 have to discuss the future, so I propose we take a break now
. 320
1 and then come back and do it, and we'll try to finish
2 everything before 12:00. Is that all right with the
3 members?
4 How about 15 minutes? Okay.
5 [Recess.]
6 DR. APOSTOLAKIS: We are back in session. Before
7 we discuss the future course of action, I will ask whether
8 the staff has any comments to make. Mr. Barrett obviously
9 does.
10 MR. BARRETT: Mr. Chairman, we thought it would be
11 useful if we took a minute and summarized some of the major
12 messages we heard today, and make sure that we've heard what
13 you think is most important for us to know.
14 This is the first chance we've really had to
15 discuss this initiative with you, and so a lot of the issues
16 that you brought up are -- or that have been brought up, are
17 very high-level issues, and I just wanted to summarize them
18 and make sure we've covered the key ones.
19 Going back to the beginning of the meeting, we
20 heard a very strong message about the need to turn up the
21 gain on interaction with the public, interaction with
22 stakeholders.
23 We're certainly going to take a look back and see
24 whether we've done a good enough job of that up to this
25 point. But more importantly, Scott and I and others, all of
. 321
1 us on the NRC management team, are committed to do that.
2 We will make the commitment now to make sure that
3 as we go forward, there is complete opportunity for all of
4 the stakeholders, including the public and public interest
5 groups, to participate at all levels in this effort.
6 The second message we got very early on in the
7 meeting was the need for us to better articulate the vision
8 that we have for this effort, and how that vision relates to
9 the strategic goals of maintaining safety, reducing
10 unnecessary burden, efficiency and effectiveness, and public
11 confidence.
12 And related to that also is to give some thought
13 to how we will measure success, what measures we will use,
14 and possibly even what criteria we would use for measuring
15 success as we move forward in this thing, to be more
16 specific, so people can monitor how well we're doing.
17 And a corollary to that was to have maybe a little
18 more discussion about the purpose and the structure of the
19 technical specifications. I don't know if the Subcommittee
20 was satisfied with that discussion today.
21 Another issue that came up and I don't think was
22 fully examined, but was something that was going through my
23 mind as I listened to some of the discussion was that there
24 -- we're pursuing a lot of risk-informed initiatives in
25 parallel. Now, this shows promise of growing into a
. 322
1 relatively strong initiative.
2 It's running in parallel with the initiative to
3 risk-informed Part 50. You heard of a number of places
4 today where those paths might cross, would definitely cross,
5 and I think that we need to do a good job as we go forward
6 in articulating better and better, how these various
7 initiatives relate to each other, and tying them together.
8 And then there are a couple of issues that relate
9 to -- well, first of all, the quality of PRA, there was a
10 lot of discussion today about how PRA quality, the quality
11 of licensees' PRAs, and, I would say, the quality of
12 licensees' capabilities, because as the Chairman pointed
13 out, it's not just the PRA, it's the risk monitor. And it's
14 also the people running the risk monitor.
15 So how that relates to each of the levels we're
16 going to step up into here, even the final level, Level 8 or
17 Option 8, Step 8, or whatever it's called, where you go to a
18 fully risk-informed framework for technical specifications.
19 It's not only important that the licensees have
20 the capability, but it's important for the staff to be able
21 to put a stamp on that and say that we have confidence in
22 the capability and we can assure the public that we have
23 confidence in the capability.
24 That's the challenge that we have in front of us.
25 Finally, the last message that I have written down here is
. 323
1 the question related to infrastructure, the NRC's
2 infrastructure, and that is what capabilities, what
3 challenges does all of this represent for the staff in terms
4 of our methods, our staff, our training, our inspectors, and
5 that we need to keep that in the forefront of our minds and
6 make sure that that doesn't become a limiting issue for this
7 issue and other risk-informed issues.
8 Those were the main messages I think I heard
9 today, and have been discussing with some of the other NRC
10 participants that we felt we heard. Are there any comments
11 or questions or additions?
12 DR. APOSTOLAKIS: Good summary. The industry
13 group, Mr. Bradley, Ms. Mahler, Mr. Harrison, Mr. Hook, do
14 you have any final comments? You don't have to have some,
15 but if you do, we'll hear them.
16 [No response.]
17 DR. APOSTOLAKIS: No? Any members of the audience
18 who would like to say anything?
19 [No response.]
20 DR. APOSTOLAKIS: Members of the public?
21 [No response.]
22 DR. SEALE: Can I ask a question?
23 DR. APOSTOLAKIS: And now we're coming to you.
24 [Laughter.]
25 DR. SEALE: How can we anticipate how big or how
. 324
1 intensive this effort is going to be over, say, the next
2 year or so, as we try to figure out how we integrate our
3 interests and suggestions into this program? We need to
4 know what kind of effort we're talking about. We need to
5 know how many Subcommittee meetings we're going to hang on
6 Sam here.
7 DR. APOSTOLAKIS: This is definitely something we
8 want to follow, right?
9 DR. SEALE: Right, I think there's no question
10 about that.
11 DR. APOSTOLAKIS: And it's very, very important.
12 Now, there are already two initiatives that have been
13 submitted to the staff.
14 I think that related to Dr. Seale's question is a
15 question of how many Subcommittee meetings we want to have,
16 at which point do we go to the full Committee, at which
17 point do we write the letter?
18 Do we want to write an interim letter at some
19 point, raising perhaps a few issues or praising people. I
20 mean, we can write letters that say a job well done.
21 Let's see, in terms of the effort, what can you
22 tell us today about how fast do these things move?
23 You have those two initiatives, and what are the
24 plans?
25 MS. GILLES: Nannette Gilles from the Tech Spec
. 325
1 Branch. I think if you looked at the industry's vision
2 slide, you see that over the next two years, they're
3 planning to pursue basically the first four initiatives.
4 So I think that those would be our area of
5 concentration to our inhouse and, I think, the next two will
6 be coming in probably sometime next year, closer to the
7 middle of the year.
8 And I think our review process, we are, as we
9 mentioned earlier, planning to implement a new process for
10 review of changes to the standard tech specs that actually
11 would involve a public comment period and things like that.
12 So we are kind of working on the timeline for that
13 process now.
14 DR. APOSTOLAKIS: Now, you have already two
15 initiatives, right?
16 MS. GILLES: Yes.
17 DR. UHRIG: Are these specifically individual
18 utilities, or are these owners groups?
19 MS. GILLES: These are changes to the standard
20 technical specifications for all owners.
21 DR. UHRIG: All owners?
22 MS. GILLES: Yes.
23 DR. SEALE: Do you expect that there might be a
24 trailblazer effort by someone in the same vein as we had the
25 pilots in some of the other areas? Would you expect a pilot
. 326
1 or a group to move forward with, say, five or something like
2 that?
3 MS. GILLES: San Onofre has identified that they
4 are willing to be a pilot for many of the initiatives, and
5 I'll let either Tom or Sharon address that -- six of the
6 seven, so San Onofre has identified them.
7 DR. SEALE: And that would be earlier than this --
8 it would still be within this two-year time period that
9 she's talking about?
10 MR. HARRISON: That's correct. This is Wayne
11 Harrison, South Texas Project. In response to Dr. Seale's
12 question, we would -- STP will expect to be submitting some
13 or our risk-informed technical specification changes. We
14 had not yet submitted our improved technical specifications,
15 and I think we're going to look at going to something along
16 the lines of the four Bravo initiatives or perhaps a little
17 step closer toward the risk-informed to see how far we can
18 go without challenging the -- having to request an
19 exemption.
20 Maybe we'll request an exemption. We've been
21 rumored to have done that before.
22 DR. SEALE: Well, the pilots are certainly have
23 proven, I think, to be an effective way to look at these
24 things. And I would like to see us go that way.
25 MR. HOOK: Tom Hook from San Onofre. I have a
. 327
1 match with executive level commitment to pursue as a pilot,
2 the first six initiatives. However, to pursue those in
3 concert with the joint NRC/Industry Task Force, and just to
4 process.
5 So we'll be following with plant-specific
6 submittals as soon as the process works for those particular
7 six initiatives.
8 DR. APOSTOLAKIS: Okay, let me ask the staff, when
9 do you think would be a good time for you to meet with the
10 Subcommittee again regarding the Initiatives 2 and 3, and
11 then go to the full Committee? --
12 MR. DENNIG: Okay, we've got 2 and 3 inhouse.
13 We're basically doing the turnover to the technical review
14 apparatus, which is fundamentally the Risk Assessment
15 Branch. We are expecting additional materials from the
16 industry in support of those submittals.
17 The reason for that is because we've stepped this
18 up a notch during the public comment. We expect to write a
19 safety evaluation. Historically in this process, we haven't
20 done all those things.
21 So, how quickly our review moves is somewhat
22 dependent on that additional information, and perhaps
23 contingent on that, someone from Risk Branch would speculate
24 on the range of times that we might be looking at to make
25 some progress on those submittals.
. 328
1 MR. REINHART: Not having received the submittals
2 yet and having the opportunity to look at them, I wouldn't
3 -- I don't know exactly, but I'm going to throw out -- and
4 shoot me if you want -- that we don't have to be through
5 them before we come here; we can just get familiarized with
6 them, and come and talk.
7 So probably, would early February be a reasonable
8 timeframe? Or is that too soon?
9 MS. GILLES: Six months.
10 MR. REINHART: Do you want to make it six months?
11 DR. APOSTOLAKIS: Well, if you want us to come
12 into a work-in-progress, so you will have a chance to adjust
13 the way you're going based on our comments, then we should
14 do it before you finish your SER.
15 MR. MARKLEY: The ideal time, I think, is at the
16 draft SER stage where you haven't finalized it but you're
17 willing to share what you plan to do with it.
18 MR. REINHART: Let's go for April.
19 DR. APOSTOLAKIS: Okay, and then perhaps we can
20 schedule or try to schedule a meeting with the full
21 Committee at the May meeting.
22 MR. MARKLEY: Right.
23 DR. APOSTOLAKIS: Okay.
24 MR. MARKLEY: Just to ask also, this is going to
25 compete with other activities, and Part 50 is just one of
. 329
1 them, and license renewal and lots of other things, so if
2 there is some way we can get a more integrated schedule than
3 this block of a two-year time period or whatever it is
4 you're dealing with, or one year, how each one of these is
5 probably going to sequence in, and when it might be best to
6 fit it into what we have would really help a lot.
7 MS. GILLES: That's something we're going to have
8 to come up with.
9 MR. REINHART: Before we leave this first
10 submittal, I want to really caveat that. I'm proposing that
11 we'll show up in April where we are.
12 We might be at the draft stage, we might be
13 getting more information.
14 DR. APOSTOLAKIS: That's fine with us.
15 MR. REINHART: We want your input.
16 DR. APOSTOLAKIS: Okay, now, something related to
17 an earlier discussion: Since San Onofre would be a pilot
18 for six of these initiatives, when do you plan to look at
19 their risk monitor and the models and make a judgment?
20 That's not going to be a trivial thing.
21 MR. REINHART: I think that's a good question, and
22 I think we need to give an answer to it, but I don't know
23 what we're ready today.
24 DR. APOSTOLAKIS: Is it part of reviewing
25 Initiatives 2 and 3, or you're going to have a separate task
. 330
1 for that?
2 MR. BECKNER: This is Bill Beckner. I don't think
3 we would do that as part of these first two initiatives. I
4 think we would try to approve these based on generically
5 coming to risk conclusions, along with whatever
6 plant-specific things we would have to with a plant-specific
7 amendment.
8 DR. APOSTOLAKIS: If it's plant-specific, though,
9 and they rely on their risk monitor --
10 MR. BECKNER: No, these first two, I don't think
11 --
12 DR. APOSTOLAKIS: They don't?
13 MR. BECKNER: No, they would not rely on it.
14 DR. APOSTOLAKIS: What are they again?
15 MR. BECKNER: These would have to rely on insights
16 where we could say they're generically applicable.
17 DR. APOSTOLAKIS: That's fine.
18 DR. SEALE: Do we want to see their risk monitors?
19 DR. APOSTOLAKIS: I don't know. I mean, that's
20 looking at the screen.
21 DR. UHRIG: The alternative is to have some sort
22 of presentation, either by one of the utilities or perhaps
23 the vendors.
24 DR. APOSTOLAKIS: That's good, that 2 and 3 do no
25 rely on an extensive review of the risk monitor, but it
. 331
1 seems to me you ought to start planning for it now, so when
2 the other initiatives come, you will be in a better position
3 to judge.
4 Now, Mr. Hook told us that the PRA, which I assume
5 is a basis for the monitor, has been reviewed. Has the
6 models of the monitor itself been reviewed by peers?
7 MR. HOOK: The answer is -- this is Tom Hook --
8 the answer is portions of the model have been reviewed by
9 peers. The risk monitor model that we have is based upon
10 our IPE and IPEEE with enhancements to make it more useful
11 and more specific to risk monitoring.
12 And those models are going to be undergoing
13 certification by the CE Owners Group early next year, as
14 well as those models are being utilized by our contractor as
15 the basis for other risk monitor models of other similar
16 types of plants, and are being reviewed by those other
17 plants.
18 And the Combustion Engineer Owners Group, we have
19 a cross-comparison activity where we compare initiating
20 events and major contributors to risk from all the
21 Combustion Engineering PRAs as a means to ensure consistency
22 and validity, and that there are no major mistakes in any of
23 the PRAs that would produce invalid results.
24 So it's been subjected to that type of review at a
25 relatively detailed level over a number of years.
. 332
1 DR. APOSTOLAKIS: Okay, that's good to know, but
2 still the staff has to somehow convince itself that the
3 result of this process is acceptable for the intended
4 purpose.
5 So it seems to me you ought to plan for something
6 like this in addition to the six or seven initiatives you're
7 going to get. And the Committee is certainly interested in
8 this.
9 DR. SEALE: In line with Mr. Markley's remarks,
10 too, about this not being the only thing that's going on,
11 this question of keeping sort of a cumulative score of the
12 kinds of things that will impact on the inspection and plant
13 evaluation process, so that we don't suddenly wind up one
14 day realizing that we've got a problem that we hadn't
15 anticipated, I think would be something that we'd like to
16 hear about occasionally, that, as of now it looks like these
17 are the things that have changed, and these are the things
18 that are going to be necessary in order to keep track it and
19 that sort of thing.
20 MR. REINHART: If I could address two pieces that
21 you brought up there, the second one, programmatically,
22 licensees that have risk-informed improvements are supposed
23 to be required to keep a cumulative track of changes to
24 their risk.
25 DR. SEALE: Okay.
. 333
1 MR. REINHART: So you have to be able to see what
2 it is to tie it in. At the same time, we're developing a
3 database to do that for ourselves so that that should
4 provide a basis for that information.
5 The second piece: I was just talking with Rich
6 Barrett. We recognize we need to come up with an approach
7 and a structure to review PRAs. And so we need to, without
8 making a commitment as to exactly how we're going to do
9 that, we need to pursue that.
10 Likely, in the sense SONGS is going to be a pilot
11 and we're probably going to learn --
12 DR. APOSTOLAKIS: There are two things you have to
13 review, the PRA and the monitor. There are differences and
14 there are sometimes important ones.
15 MR. REINHART: I see what you're saying.
16 DR. APOSTOLAKIS: If they propose to use the risk
17 monitor, then you really have to review the risk monitor,
18 not the PRA.
19 MR. REINHART: I think if I understand Tom, the
20 PRA is the basis for the monitor, and then the monitor,
21 i.e., the process, the program for evaluating risk.
22 DR. APOSTOLAKIS: Well, there are non-trivial
23 changes that they want us to make to the PRA in order to
24 make it real-time risk monitoring, so I think the staff
25 should be aware of it.
. 334
1 MR. REINHART: Okay, good point.
2 MR. LEVISON: Stanley Levison. In talking about
3 whether to or how to evaluate the PRA, it seems that the
4 staff has two things to review: One is the actual process.
5 Regardless of what the PRA looks like, you're going to be
6 looking at the process for evaluating these things.
7 Irrespective of whether you have the mechanism to
8 look at the PRA or not, you can endorse through an SER, a
9 process of how the utility should submit something without
10 even looking at the quality of the PRA and set up the
11 framework for that.
12 Then, of course -- and I'm not saying that it
13 doesn't need to be done -- then separate SER document
14 process for how you are going to deal with the PRA quality.
15 So if you haven't finished how you're going to do the PRA
16 quality issue, that wouldn't impact the ability to review
17 and look at how the tech spec changes are done.
18 And Dr. Seale, concerning the issue about
19 cumulative changes in risk and the certification process and
20 the ASME standard, both address those items in requiring the
21 utilities -- said the NRC requires them, but the utilities
22 themselves are aware that there are, you know, cumulative
23 changes as they do risk-informed activities, and as they
24 change their PRA, some of them actually have procedures in
25 place that require them to go back and reevaluate the
. 335
1 applications that they've done to make sure that changes in
2 the plant that might change the PRA doesn't change, the
3 results that they go when they used that to support an
4 application. Thank you.
5 DR. SEALE: Thank you. I want to make a note that
6 I'm also concerned about the cumulative changes in the
7 regulatory process that accrue to these changes.
8 MR. BARRETT: If I could just comment on that and
9 raise a question? I believe there was talk about meeting
10 with the Subcommittee on the subject of possible barriers.
11 I forgot what the word was.
12 DR. APOSTOLAKIS: Impediments.
13 MR. BARRETT: Impediments to the implementation of
14 risk-informed strategies, and that that might be scheduled
15 in February.
16 MR. REINHART: It is.
17 MR. BARRETT: I think that some of the issues
18 you've brought up today about the impact on staff resources
19 and training, questions related to staff assurance of PRA
20 quality and quality of PRA methods, might be good topics to
21 discuss at that meeting.
22 DR. APOSTOLAKIS: Dr. Uhrig, do you have any
23 comments to make now?
24 DR. UHRIG: Well, I have a continuing interest in
25 the predictor, the PRA meter. As I told you at the break,
. 336
1 15 years ago, we played around in Jerry Fussell's simulator
2 of a PRA on an IBM XT as I recall, or AT. And the results
3 were rather fascinating, although probably not very valid.
4 But I would be very interesting in getting up to
5 date on what's going on and finding out more about it. I
6 actually corresponded with Mr. Hook on this on a couple of
7 occasions.
8 It seems to me, though, that there are, what,
9 three vendors that provide these now? Scientech, PLG, and
10 SAIC. It might be useful somewhere along the line to get a
11 picture of how they operate, how they differ, what the
12 features of these are. That might ge a subject for a
13 Subcommittee meeting at some time in the not-too-distant
14 future.
15 And it might be joint with the INC part of the
16 ACRS. So maybe we'll think of that as a joint Subcommittee
17 meeting.
18 DR. APOSTOLAKIS: Before I forget, Mark, I said
19 earlier that we have to review both risk monitor and the
20 PRA. Several of our comments that would apply to the
21 changes from the PRA to the risk monitor or PRA in the
22 October letter on risk-informed Part 50 in connection with
23 the risk achievement worth measure, because there you change
24 a component. You assume it's down, right? So it's a very
25 similar situation with a configuration change.
. 337
1 And we raised a few concerns, gave a few
2 references, so that would be a good place to start to
3 understand where I'm coming from.
4 MR. REINHART: The ACRS letter?
5 DR. APOSTOLAKIS: The ACRS letter to the Chairman
6 on risk-informing Part 50.
7 MR. REINHART: Right.
8 DR. APOSTOLAKIS: Okay?
9 MR. REINHART: Okay.
10 DR. APOSTOLAKIS: Dr. Bonaca?
11 DR. BONACA: First of all, in general, I think
12 that this is a great initiative, just 20 years late.
13 [Laughter.]
14 DR. BONACA: But I think it's great that it's
15 taking place. Now, there are a number of points: One is
16 that in many cases, really these are improvements to the
17 current tech specs. They're not really evolutionary in any
18 way, just because current tech specs have a lot of -- for
19 example, AOTs, they have no basis oftentimes. And so
20 finally we have a technology that we have had for awhile,
21 and we can use it to put a reasonable time into tech specs,
22 and that's just an improvement.
23 I wouldn't sell it is as risk initiative, it's
24 just an improvement that is there.
25 In some other cases, like for example, Initiative
. 338
1 Number 7, I would like to hear more in the next presentation
2 about how you go to talk about functional. I don't
3 understand that.
4 I mean, functional versus operable seems to me
5 that we're talking about functional and you're still going
6 in a deterministic way, and maybe you have some ideas about
7 how probability can come in, and I would like to hear about
8 that.
9 That's a fundamental issue that it's so critical
10 we'd like to hear about what the concepts are. And maybe
11 that's too far ahead of time, because you have a couple of
12 years before you're getting there, I think, two or three
13 years.
14 The third point that I would like to make is that
15 when you talk about a one page tech spec, tech specs have
16 always been seen by the overwriters in the control room as
17 the contract, okay? -- and now there is no more contract, so
18 it would be just one sheet of paper, and I would like to
19 understand how this new world is evolving to that because I
20 think it has an impact in the control room.
21 All those commitments that the NRC has required
22 were very much taken care of -- would create attention.
23 Those that were not required, well, it was always, you know,
24 we are sorry but, well, corrective action will take place,
25 so I would like to understand how this new world is coming
. 339
1 and the point the Dr. Seale brought up, that the whole
2 burden now is placed on inspection and inspection really has
3 a lifetime. I mean there is a lifetime between -- so that
4 is the third point I would like to make, that we have to
5 reflect on that. Again, I am extremely supportive of what I
6 see.
7 DR. SEALE: In that regard we have to recognize
8 and maybe even marvel at the fact that the tradition of an
9 NRC-imposed and conducted individual licensing examination
10 for licensed operators and the tech specs, which was largely
11 the basis for that examination, at least for the
12 operational, some of the operational parts of it, related to
13 an idea that the contract was personal between the
14 individual operator and the NRC -- and that was a personal
15 challenge. It wasn't a company challenge. We have to be,
16 you know -- that is an important point and we want to be
17 sure we don't lose that personal commitment that is
18 implicit.
19 DR. APOSTOLAKIS: So what you are saying it may be
20 an impact on the safety culture.
21 DR. SEALE: It could, and so we have to ask
22 ourselves what do we do to preserve that?
23 DR. BONACA: Or what do we substitute for it?
24 DR. SEALE: Yes, right.
25 DR. BONACA: Maybe commitments which are internal
. 340
1 but are shared with the NRC that, you know, maintain that
2 kind of level of personal relationship -- with regulation.
3 DR. UHRIG: But I think what you are doing here in
4 many respects is shifting the guidance here over to
5 technology and the risk meter is probably a much better
6 judge of some of these outage times and other things that
7 have been in tech specs for years and years, because there
8 were some gross inconsistencies in tech specs.
9 Very often, certainly in the early days, it was
10 whatever you could negotiate that you get wound up in the
11 tech specs -- having been on the other side of the table.
12 MR. MARKLEY: That is still true.
13 DR. UHRIG: Having been on the other side of the
14 table for a number of years, the tech specs were always the
15 hardest part of getting a plant online as far as I was
16 concerned. Once you have essentially got it finished, then
17 negotiating the tech specs were a big item.
18 The guidance now is going over to a rational
19 basis, whereas before it was judgment --
20 DR. SEALE: Irrational basis.
21 DR. UHRIG: -- and sometimes irrational decisions.
22 DR. APOSTOLAKIS: Okay. The members have finished
23 their comments?
24 DR. UHRIG: Yes.
25 DR. APOSTOLAKIS: I have a couple comments on the
. 341
1 summary, for Mr. Beckner -- and Mr. Hooks.
2 It says here the industry stated during the
3 meeting of -- I don't know -- the industry stated it was
4 their general philosophy to use qualitative risk assessments
5 where they believed the benefits of a proposed change were
6 obvious and to use quantitative assessments where the
7 outcome was not obvious.
8 The Staff reserved judgment of such an approach,
9 noting that benefits that are obvious to one person or group
10 might not be obvious to another, so what does qualitative
11 risk assessment mean? Can someone from the Risk Informed
12 Technical Specifications Task Force enlighten us?
13 MR. REINHART: This is Mark Reinhart. I guess my
14 comment on that was really the thought of what might be
15 obvious to one is not obvious to the other and I would like
16 to see some proof that what is qualitative is really
17 qualitative and maybe it does need some quantitative
18 support. I don't know.
19 DR. APOSTOLAKIS: That's why I am asking the
20 industry what they mean by qualitative risk assessments for
21 cases where the benefits of a proposed change are "obvious."
22 We have a willing speaker.
23 MS. MAHLER: Willing -- maybe. In that meeting it
24 was given in the context of I believe we were talking about
25 Initiative 1, where we are talking about having a more
. 342
1 resource-rich state, okay? -- for where mode hot shutdown is
2 a more resource, qualitatively -- just think about it -- hot
3 shutdown is a more resource-rich state than cold shutdown,
4 and that was the context in which we were talking about
5 that, where we have for Initiative 1 the tack that is being
6 taken right now is SONGS in their draft analysis has looked
7 at their end states and has quantified things.
8 The question, again in the context of that
9 meeting, was okay, do all of the other CE plants have to
10 quantify their transition risk, or can they qualitatively
11 assess the risk and compare it to SONGS in that manner, so
12 that was the context in which those statements were made.
13 Does that answer your question?
14 DR. APOSTOLAKIS: Well, yeah. That context was
15 not given here. It says that the context here was a
16 discussion of the meaning of the term "risk-informed" as it
17 related to regulatory applications, which is much broader
18 than what you just said, so in that context, in the context
19 that is in the writeup, it created questions in my mind,
20 although what you just said, it can be quantified, but what
21 you are saying is it is so obvious I mean let's not spend
22 time to do it.
23 DR. SEALE: Except that then, you know, there is
24 no question I don't believe but what the hot shutdown
25 situation is the more resource-intensive, as you have
. 343
1 characterized it, but are the resources the same, and are
2 the resources that may not be common to both sets ones that
3 are crucial to the capabilities that you need in the event
4 of a problem. Are there things that are available with cold
5 shutdown that you don't have with hot shutdown that really
6 go to the heart of a difficulty. That is why you have to be
7 quantitative to a degree anyway.
8 DR. BONACA: That is why I think at times it is
9 hard to come from a specific insight to a generic
10 application to everybody that says, oh, sure, because I mean
11 conditions may be different, and the other thing, there is a
12 benefit in the process of quantification. The utility that
13 goes through the process of quantification understands why,
14 and typically also it reinforces those things which are
15 considered important for the quantification. It puts them
16 into guidance. It puts them into procedure.
17 For a utility to simply say all right, yes, since
18 you gave it to Plant X, this truth is applicable to all the
19 other plants -- that is not necessarily true. In some cases
20 it may be and may be obvious maybe, but --
21 DR. APOSTOLAKIS: Okay, thank you. There is one
22 paragraph here that I think needs some attention --
23 representatives from the South Texas Project presented a
24 concept of a fully risk-informed set of technical
25 specification that essentially relies on a configuration
. 344
1 risk management program as the backbone of the technical
2 specifications. They likened the proposed risk limits in
3 their concept of radiation protection limits -- that is
4 ALARA limits.
5 The NRC regulates ALARA limits at a high level and
6 Licensees control limits at a lower level administratively.
7 The group also discussed whether there was a need for an
8 instantaneous risk cap for technical specifications. The
9 Task Force stated that the major question is the cost
10 benefit of going to this extreme. For example, if plant
11 PRAs essentially become the technical specifications, then
12 Licensees would have to control changes to the PRA model to
13 the same degree as technical specification changes are
14 currently controlled.
15 The question is why is that bad?
16 MS. GILLES: Nanette Gilles from the Tech Spec
17 Branch.
18 I don't think -- the implication was not
19 intended -- I am Nan Gilles from the Tech Spec Branch. I
20 don't think that -- the implication was not intended that
21 that was bad. The only thought there was that once you did
22 the first seven initiatives that the Task Force in general
23 had a question about whether going the next step to the
24 fully risk-informed tech specs was going to be cost
25 beneficial if you already had the benefits of the first
. 345
1 seven initiatives. That was the thought that was trying to
2 be conveyed.
3 But then the example is that PRAs would have to be
4 controlled for changes to the PRA model to the same degree
5 as technical specification changes.
6 MS. MAHLER: This is Sharon Mahler again. In that
7 context the concern was you had developed your PRA, you were
8 controlling your plant configuration in accordance with your
9 PRA, and you did some other modification where you wanted to
10 enhance your PRA model. Would that then involve you going
11 back to the NRC with your full-blown model to again get
12 approval of your PRA model?
13 DR. APOSTOLAKIS: Your PRA model is the main tool
14 that allows you to reap all the benefits of going to a fully
15 risk-informed set of technical specifications. Wouldn't
16 that be a reasonable price to pay?
17 I see Mr. Grantom there willing to make a few
18 comments.
19 MR. GRANTOM: Yes. Rick Grantom, South Texas.
20 In the process of approving these types of tech
21 specs that go to that level, the process by which the PRA is
22 updated is also going to be included in that the change
23 control process has to be there, so in that regard what I
24 would anticipate that we would do is that the PRA
25 configuration control process is a part of the approval that
. 346
1 happens when you go to a fully risk-informed regime.
2 Therefore, the Staff has already approved the mechanisms by
3 which we evaluate changes to the station, by which we
4 implement those changes, and the mechanisms by which we
5 implement changes.
6 I want to emphasize that rolling out a PRA model
7 is not an insignificant effort. It has to be controlled.
8 There's other risk-informed applications that are going to
9 be affected that people have to get together and meet about
10 assessing any changes, so it will be in my personal view --
11 it should be included as part of the approval that the PRA
12 change control process is part of that approval, and it
13 would be handled in that regard.
14 DR. APOSTOLAKIS: So you are willing then to
15 accept this?
16 MR. GRANTOM: Yes -- the process, yes. I probably
17 would be hesitant to say I am going to roll out a new PRA
18 every day because somebody --
19 DR. APOSTOLAKIS: No.
20 MR. GRANTOM: -- changed the procedure, but I
21 would be not opposed at all to be saying here is the
22 configuration control process. Every 18 months or some
23 period of time we collect data, we look at procedure, we
24 roll out a new model from that. We have in a sense an
25 effective model that is in effect right now versus a working
. 347
1 model that we may be collecting changes. That process I
2 think is workable in that regard and if it is approved as
3 part of this tech spec effort there, I wouldn't be opposed
4 to that personally.
5 From a PRA manager's point of view and being able
6 to control my process it is not really any different than
7 what I am doing currently.
8 DR. APOSTOLAKIS: And that would include changes
9 in the model that are being done because the
10 state-of-the-art advances. Let's say the NRC now is
11 launching or has already launched a research project on fire
12 risk assessments, so if this comes up with something better
13 than what we have now, you have to change your model and
14 that kind of change would have to be controlled under these
15 conditions.
16 MR. GRANTOM: There's two processes involved.
17 There's internal changes, and you can liken that to
18 procedures, design modifications -- those types of things
19 that happen internal within the plant itself, and then there
20 may be external changes that occur where someone says, gee,
21 we'll handle fires differently than that.
22 As long as the process allows us to properly
23 schedule those changes and to do that, then I think it can
24 work, but there are those two facets that have to be
25 compared in.
. 348
1 DR. BONACA: But you have also a significant plant
2 change process that takes place, right?
3 MR. GRANTOM: Oh, yes. That's why I said there is
4 the internal plant changes that occur. We do a
5 modification. That is all part of --
6 DR. BONACA: Okay -- I thought you were talking
7 about the PRA --
8 MR. GRANTOM: -- of the internal. What Dr.
9 Apostolakis was talking about is the methodology for
10 evaluating some phenomena changes. That is a little bit
11 outside --
12 DR. APOSTOLAKIS: So they are both -- yes.
13 Am I to assume that the line that is forming here
14 is --
15 [Laughter.]
16 MR. FRANK: If I might just comment on this?
17 DR. APOSTOLAKIS: Sure. Would you please go to
18 the microphone, Frank?
19 Tell us who you are again.
20 MR. RAHN: Well, for the first time, this is Frank
21 Rahn from EPRI.
22 DR. APOSTOLAKIS: See, that is the importance of
23 language -- "again" --
24 MR. RAHN: I wanted to make a brief comment in
25 terms of the ASME and the PRA standard which is in the
. 349
1 process of being developed and configuration control will be
2 an integral part of that PRA standard I hope that will be
3 adopted by the NRC once it is complete. Thank you.
4 MS. MAHLER: I wanted to help perhaps put it in
5 perspective. I relate it very much to the similar process
6 for the Kohler, where the Staff has approved a methodology
7 and an individual plant change you can go and revise your
8 curves that are in your Kohler and not come back to the
9 Staff for approval of those changes in those curves, but if
10 you change your methodology then you do need to come back to
11 the Staff, so that was more along the lines of what we were
12 originally thinking.
13 DR. APOSTOLAKIS: Also, as a side remark, would
14 you please stop using the word "instantaneous" risk? It is
15 conditional on that particular configuration. There is
16 nothing instantaneous about it. It is a temporary
17 situation. All risk estimates are conditional. This
18 happens to be conditional in a particular configuration but
19 it is not of long duration, because instantaneous confuses
20 people.
21 DR. SEALE: George, is it fair to say based on the
22 remarks that we have just heard that if we are going to have
23 a risk cap, we ought to get on with it and let people know
24 what it is and --
25 DR. APOSTOLAKIS: I think we will have to face
. 350
1 that, yes. Yes.
2 DR. SEALE: And we ought to do that sooner than
3 later because it has a front-end impact.
4 DR. APOSTOLAKIS: Now risk cap -- what, on CDF?
5 On what?
6 DR. SEALE: That is part of the question.
7 DR. APOSTOLAKIS: Is it a goal? I mean the Staff
8 is already working on a set of goals and I understand in
9 fact there are some exchanges with the Commission's
10 assistance yesterday.
11 DR. SEALE: Yes.
12 DR. APOSTOLAKIS: So certainly that issue will
13 come up when we discuss this, because it is a broader
14 issue -- the safety goals of the Commission.
15 Are there any other comments from anyone?
16 [No response.]
17 DR. APOSTOLAKIS: Well, silence tells me that this
18 is the end of this subcommittee meeting. Thank you all for
19 coming. It was very useful to us, and we will see you again
20 some time in April -- preferably late April.
21 [Laughter.]
22 [Whereupon, at 12:08 p.m., the meeting was
23 concluded.]
24
25
Page Last Reviewed/Updated Tuesday, July 12, 2016