481st Meeting - April 6, 2001
Official Transcript of Proceedings
NUCLEAR REGULATORY COMMISSION
Title: Advisory Committee on Reactor Safeguards
481st Meeting
Docket Number: (not applicable)
Location: Rockville, Maryland
Date: Friday, April 6, 2001
Work Order No.: NRC-147 Pages 233-337
NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers
1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005
(202) 234-4433. UNITED STATES OF AMERICA
NUCLEAR REGULATORY COMMISSION
+ + + + +
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
481ST MEETING
+ + + + +
FRIDAY,
APRIL 6, 2001
+ + + + +
ROCKVILLE, MARYLAND
+ + + + +
The Committee met at the Nuclear Regulatory Commission,
Two White Flint North, Room T2B3, 11545 Rockville Pike, at 8:30 a.m., Dr.
George E. Apostolakis, Chairman, presiding.
COMMITTEE MEMBERS PRESENT:
GEORGE E. APOSTOLAKIS Chairman
MARIO V. BONACA Member
F. PETER FORD Member
THOMAS S. KRESS Member
GRAHAM M. LEITCH Member
DANA A. POWERS Member
COMMITTEE MEMBERS PRESENT: (cont'd)
WILLIAM J. SHACK Member
JOHN D. SIEBER Member
ROBERT E. UHRIG Member
GRAHAM B. WALLIS Member
INVITED EXPERT:
STEPHEN L. ROSEN
ACRS STAFF PRESENT:
SAM DURAISWAMY
CAROL A. HARRIS
JOHN T. LARKINS
JAMES E. LYONS
MAGGALEAN W. WESTON
ALSO PRESENT:
PETER BALMAIN
TOM BERGMAN
RALPH CHACKAL
NANCY CHAPMAN
DAVID FISCHER
BOB GRAMM
AL GUTTERMAN
SCOTT HEAD
ALSO PRESENT: (cont'd)
KAMAL MANOLY
EILEEN McKENNA
MATHEW A. MITCHELL
A.C. MOLDENHAUER
JOHN A. NAKOSKI
BOB PALLA
STUART RICHARDS
GABU SABO
THOMAS SCARBROUGH
G.E. SCHINZEL
JACK STROSNIDER
DAVID TERAO
DALE THATCHER
HAROLD VANDERMOLEN
I N D E X
AGENDA ITEM PAGE
Opening Remarks. . . . . . . . . . . . . . . . . 237
Draft Final Safety Evaluation Report for the . . 241
South Texas Project Nuclear Operating
Company (STPNOC) Exemption Request
P-R-O-C-E-E-D-I-N-G-S
(8:30 a.m.)
CHAIRMAN APOSTOLAKIS: The meeting will now come
to order. This is the second day of the 481st meeting of the Advisory
Committee on Reactor Safeguards.
During today's meeting, the committee will consider the
following: the South Texas Project Nuclear Operating Company's exemption
request; closure of Generic Safety Issue 170, reactivity transients and fuel
damage criteria for high burnup fuel; report of the Materials and Metallurgy
Subcommittee regarding risk-informing, 10 CFR 50.46; future ACRS
activities; report of the Planning and Procedures Subcommittee; reconciliation
of ACRS comments and recommendations; and proposed ACRS reports.
This meeting is being conducted in accordance with the
provisions of the Federal Advisory Committee Act. Mr. Sam Duraiswamy is
the designated federal official for the initial portion of this meeting.
We have received no written comments or requests for time
to make oral statements from members of the public regarding today's
sessions. A transcript of portions of the meeting is being kept, and it is
requested that the speakers use one of the microphones, identify themselves,
and speak with sufficient clarity and volume so that they can be readily heard.
I would urge the members to read Item 13 that was handed
out this morning, reconciliation of ACRS comments and recommendations,
because we will have to visit this issue later today.
Now, it's my pleasure to announce to the committee that,
acting upon the recommendation of Chairman Meserve, the Commission has
decided to give the NRC's Honorary Distinguished Service Award to Dr.
Powers for his leadership of this committee during 1999 and 2000.
(Applause.)
Dr. Powers will be recognized at the 24th Annual Awards
Ceremony on Thursday, June 7th, at 2:00.
I'm also pleased to announce that Dr. Peter Ford is now an
official member of the ACRS, with all the rights, privileges, and
responsibilities --
(Laughter.)
-- such as they are today.
And Mr. Stephen Rosen is an invited expert to this ACRS
meeting, and the paperwork is being processed for him to become a full
member, subject to reconciliation of any conflict of interest matters.
Finally, I would like to introduce Mr. Rob Elliott. Rob,
where are you? From NRR, who will be working with the ACRS staff for a
few months, principally in the area of license renewal.
And with that, I will turn the meeting over to Mr. Sieber,
who will guide us through the first session of the day on the draft final safety
evaluation report for the South Texas Project Nuclear Operating Company,
STPNOC. Any way to pronounce this?
(Laughter.)
Exemption request. Jack?
MEMBER SIEBER: Thank you, Mr. Chairman. Before we
start, I will mention that I brought with me and provided copies of the INEEL
technical letter report on commercial practice. And this is a draft report. I
don't think it's been issued as a final report, and it might make some
interesting reading. And at least it will provide ballast for your suitcase when
you travel back home.
This is one of a series of meetings on the South Texas
Project exemption request. The initial exemption request was filed July of
1999. The SER was issued in November. We have already had three
meetings with South Texas, the last one of which involved the categorization
process. In today's discussion, we'll describe issues related to and the process
of what really is special treatment, what does it mean, and/or commercial
treatment.
And with that, we have a mixture of NRC and STP folks at
the table, and they will jointly give the presentation as they go through it.
And I'd like to introduce from NRR the Senior Project Manager for this
project, John Nakoski.
John, would you like to go ahead?
MR. NAKOSKI: Yes. Thank you.
My name is John Nakoski. I'm a Senior Project Manager
for the South Texas Project, specifically for the exemption request. Here with
me from the NRC is Jack Strosnider, Division Director for the Division of
Engineering in NRR, and to my right is Scott Head. And I'd ask that he
introduce the folks that he has up here.
MR. HEAD: My name is Scott Head. I'm the Manager of
Licensing at the South Texas Project. Joining us today is Alan Moldenhauer,
one of our PRA staff members. We have Glen Schinzel, who will be doing
most of our presentation. He is the Project Manager of our Option 2 effort,
our exemption request effort.
Joining us also is Ralph Chackal from our staff, and one of
our legal representatives, Al Gutterman.
MR. NAKOSKI: And with ACRS' indulgence, I would like
to modify the agenda as outlined in the meeting notice. As was indicated, we
would like to do a joint presentation to the ACRS for efficiency and using the
two hours that we have allotted.
And with that, if we could go to the second slide, I'd like to
just give you some perspective where we are in this process and what level of
effort has gone into it. As was mentioned, the submittal -- original submittal
was submitted in July of 1999. We've had a number of meetings with South
Texas, and we have met with ACRS on a number of occasions.
In December of last year, we met with the ACRS on the
draft safety evaluation that was issued in November of last year. We also met
with ACRS in February, a subcommittee of ACRS in February on
categorization, and today we're going to be talking about treatment issues that
were identified in the draft safety evaluation.
Going forward, you'll see there's a TBD -- to be determined
-- bullet in there. We are still working closely with South Texas to resolve the
remaining open items. All of them have not been resolved yet. When those
are resolved, we will be able to follow a schedule similar to what's laid out
here as far as intervals.
If we can get the open items all resolved by the end of this
month, I still think we're in a pretty good position to meet with the
Commission in early June to discuss the issuance of the safety evaluation and
the granting of the exemptions.
To move forward, in the draft safety evaluation there were
16 open and two confirmatory items. Six of these items have been closed
without any further effort required on the part of either the staff or South
Texas. Six of these open items can be closed based on an agreement on the
details contained in the South Texas FSAR section that describes the
categorization, treatment, and management and oversight processes.
MEMBER SHACK: Is there a difference between a finger
and a triangle?
MR. NAKOSKI: Yes. A finger indicates that that's a
treatment open item. And we will discuss those in a little more detail later on
during our presentation.
Two of the open items on categorization -- we have a success
path agreement in principle with South Texas on how to close those items. It
will require South Texas to provide us additional insights before the --
additional submittal before those can be closed. There are two open items that
are treatment open items on repair and replacement and in-service inspection
of ASME code components that South Texas owes as a revised response to
that open item.
There is really only one issue that remains, and that's one of
their proposed alternatives for repair and replacement. And we're working
with them to resolve that issue.
Open Item 8.1 is an item related to environmental
qualification, and the staff has the lead for action in this area. We need to
internally resolve our position on what level of detail would we expect to be
described in the treatment process, primarily in the area of procurement in
their FSAR.
And, similarly, Open Item 18.1 that deals with seismic
design requirements or seismic qualification requirements -- we -- the staff has
come to a position, shared that with South Texas. However, as you'll see
later in the presentation, I think it requires some further discussion between
the staff and South Texas to consider this item resolved.
If we could go to slide 4 in the handout. I think it's
important to let the ACRS understand what the process was or -- for the staff's
review of the treatment area. The first thing that's important to recognize is
that under Option 2, in which South Texas is a proof of concept for, the
design basis would not change.
Second, is that even low safety significant but safety-related
components -- structures, systems, and components would remain functional.
They still need to be able to do those safety-related functions that they are
credited for in the design basis.
What we would expect to be in the FSAR is a high-level
description of what the elements of the treatment processes in South Texas'
balance of plant or commercial practices are. We're not asking how those
processes would be implemented. We're relying on South Texas' engineering
judgment to determine how to implement those elements of the treatment
processes.
And it's also important to recognize that in order to grant the
exemption we need a licensing basis, and that basis is provided in the
description in the FSAR.
MEMBER LEITCH: So the key word there is "low safety
significance."
MR. NAKOSKI: That's correct.
MEMBER LEITCH: But systems -- what is it -- not related
to safety are not included in that.
MR. NAKOSKI: Non-safety-related LSS -- to use the South
Texas vernacular, LSS, low safety significant or non-risk-significant are not
subject to NRC regulatory treatment requirements. And in most cases, they
would not have been before.
MEMBER LEITCH: Okay. Thank you.
MEMBER BONACA: Multiple trains of very high safety
significant systems can be categorized as lower. Like for example, medium
or low in the methodology that South Texas has used, because of the number
of redundancies.
MR. NAKOSKI: They can factor in -- and I'll give an
answer and let --
MEMBER BONACA: Well, the question I have is that, are
there any high safety significant SSCs or systems that have low safety
significant trains supporting them?
MR. NAKOSKI: I'll let South Texas field that question.
Pull the mike to you, Glen.
MR. SCHINZEL: Do you want to address that, Ralph?
MR. CHACKAL: Yes. We can have -- this is Ralph
Chackal with South Texas. We can have a medium or high safety significant
system function with a level of redundancy. And diversity provided by the
components allows us to rank some of those components as low.
MEMBER BONACA: So how do you guarantee that the --
the lower level of confidence, then, for high safety significant systems? You
see, I'm taking exception with the statement there. It says, "The functional
capability of low safety significant systems would be maintained, although at
the lower level of confidence."
Now, if I have four trains of a high safety significant system,
and for those now I maintain the functional capability at the lower level than
for high safety significant. Then, the overall safety significant system -- I
believe the functional capability of that is at the lower level of confidence.
That's the outcome of that, right?
CHAIRMAN APOSTOLAKIS: But isn't that the whole idea
of using the PRA, that it's built into all of these importance measures, so you
don't have redundancy and all that. You will get a low Fossil-Vesely and a
low raw -- I mean, all of these questions are built into the calculation. You
don't have to revisit them afterwards.
MEMBER BONACA: You're talking calculation. I'm
talking about the programs that you are applying to each individual train,
okay, that you're saying will provide a lower level of confidence on each
individual train. That's what I read there.
MR. SCHINZEL: This is Glen Schinzel from South Texas.
Just a little clarification. On the safety significant systems, generally there
will be a certain number of components. Maybe it will be the pump. It will
be the main valves. It will still be safety significant, either high or medium.
But within that train there may be a lot of auxiliary components; you'll fall
into a lower classification.
So those vent valves, drain valves, instrumentation, we may
be able to treat those with a lower level of assurance. But for the pump, for
the main valves, those will still fall within the full treatment requirements.
MEMBER BONACA: Okay. I understand what you are
saying. Again, however, I think it's a bit obscure how you apply the rule that
you showed us before in the presentation. That is, that the number of
redundancies on trains may justify a lower level of significance for individual
train than the system they support. I still have some --
CHAIRMAN APOSTOLAKIS: It seems to me that the
words "lower level of confidence" are unfortunate. Really, I'm not sure it's
a lower level. It is a lower level, but it may be lower by 10 to the minutes
whatever -- n. It's not that, you know, you are reducing the confidence of --
MEMBER BONACA: Here we're questioning, then, and
we're discussing the effectiveness of the QA program. I cannot give you
estimates of probability or confidence based on judgment. I'm only saying
that if we do believe that implementation of a quality assurance program or
certain components provides for a higher level of confidence, I don't know.
MR. SCHINZEL: This is Glen Schinzel from South Texas
again. We essentially begin with our base level of confidence, if you will, to
be what's currently required by the regulations. So that gives us a degree of
assurance that these components, through various testing, through
qualifications, will perform their functions.
And what we're saying for -- through the categorization
process and recognition of low safety significant/non-risk significant
components, these components don't have to have the same level of assurance.
MEMBER BONACA: So let me ask just one question. Will
the quality, or, let's say, the confidence in the overall system be maintained?
MR. SCHINZEL: Yes, it would. You know, we intend on
using commercial-type practices. Our balance of plant is running extremely
well. We have a very high level of reliability, and we fully expect these low
safety significant/non-risk significant components to fully function.
MEMBER BONACA: I understand. Fully function.
CHAIRMAN APOSTOLAKIS: When you say "low safety
significant," I remember the four categories. Which ones are these, risks
three and four?
MR. SCHINZEL: These are Box 3. It would be --
CHAIRMAN APOSTOLAKIS: Box 3 only?
MR. SCHINZEL: -- Box 3. yes.
CHAIRMAN APOSTOLAKIS: Four is what?
MR. SCHINZEL: Box 4 is non-safety-related.
CHAIRMAN APOSTOLAKIS: We don't care about that.
MR. SCHINZEL: That's correct.
CHAIRMAN APOSTOLAKIS: From what you just told us.
MR. SCHINZEL: That's correct.
CHAIRMAN APOSTOLAKIS: So it's Box 3, which is low
Fossil-Vesely, low raw, but they are safety-related.
MR. SCHINZEL: That's correct. That is correct.
CHAIRMAN APOSTOLAKIS: So that's where the battle
is, Box 3.
MR. SCHINZEL: That's correct.
CHAIRMAN APOSTOLAKIS: Any questions about Box 2?
Any issues on Box 2?
MR. NAKOSKI: We have found that their description in the
FSAR of the treatment that they would apply to non-safety-related, high safety
significant, or medium safety significant, which is the Option 2, Box 2
category, to be acceptable, because they will -- they have described a process
that would target enhanced treatment.
CHAIRMAN APOSTOLAKIS: Oh, that's the target.
MR. NAKOSKI: That's correct.
CHAIRMAN APOSTOLAKIS: Okay. So 3 is the
commercial treatment. Now, when you did the sensitivity analysis where you
increased the failure rates by 10, did you do it to the components in Box 3?
MR. MOLDENHAUER: This is Alan Moldenhauer from
South Texas. When we did the sensitivity study where we increased the
failure rates by a factor of 10, we only applied it to Box 3 and Box 4.
CHAIRMAN APOSTOLAKIS: So it was both boxes.
MR. MOLDENHAUER: Yes, both boxes.
MR. NAKOSKI: This is John Nakoski. Just to clarify -- to
the extent that they were modeled in the PRA, right?
MR. MOLDENHAUER: That's correct.
CHAIRMAN APOSTOLAKIS: Well, and there is a reason
why they are not, if they are not. Let's not forget that.
MR. MOLDENHAUER: Right.
CHAIRMAN APOSTOLAKIS: If you leave it at that --
MEMBER BONACA: I'm raising this issue -- I raised it
before -- because I think it has generic implications to whatever is going to
happen in the future with this. And I still have some conceptual concern about
the inconsistency between a system which is in Box 1 and supported by
components that are in Box 3. That's all.
MR. NAKOSKI: This is John Nakoski with the NRC. I
think where the staff is at is that the categorization process will identify
components, or could identify components, in an HSS system that are not
significant to the capability of the system to fulfill a high safety significant
function.
I think we acknowledge that that can happen, and we would
expect it to happen. And we would expect that the treatment on those could
be reduced.
MR. STROSNIDER: This is Jack Strosnider. I'd just add,
I guess, the -- I think this discussion is good because we need to look at an
integrated view of the program. However, in the context of today's discussion
where we talk about the reduced level or lower level of confidence, we're
really focusing on the treatment programs.
The expectation, as is pointed out here, is that the design
basis and the functionality will be maintained. But the challenge we have is
to figure out what relaxation from the current requirements can be granted and
still maintain that functionality but define that as a lower level confidence.
You said something interesting earlier when you talk about
a QA program or one of these programs where -- how do you measure or
quantify a reduction in confidence. And that's a very difficult thing, because
a lot of it is judgment. And that's part of the challenge that we have in
reaching agreement on what the appropriate level is here.
CHAIRMAN APOSTOLAKIS: Now, South Texas also
identified something like 600 non-safety-related components that actually were
risk significant in some sense. You have moved them up to some category
where some extra treatment would be required for those?
MR. NAKOSKI: This is John Nakoski. The Risk 2 box or
the HSS/MSS non-safety-related we would expect South Texas -- and as
described in their FSAR, we would expect that they would implement targeted
or enhanced treatment to the attributes or characteristics of the SSC that would
cause it to be high or medium safety significant.
CHAIRMAN APOSTOLAKIS: Yes. I didn't expect you
to object to that, obviously.
MR. NAKOSKI: Well, I think it's stronger than that. I
think it's we expect that that -- that's an expectation that the staff has.
CHAIRMAN APOSTOLAKIS: Right. But you didn't know
before they came to you that some of the non-safety-related SSCs, in fact,
were of high significant. Right? You didn't know that?
MR. NAKOSKI: Not the specifics. But I would -- systems
like maybe aux feedwater that maybe aren't safety-related or some other parts
of main -- we understand the significance of those and are --
CHAIRMAN APOSTOLAKIS: No. It's not a matter of you
not understanding. My point is this: that if we look at the whole project or
the whole system as we are about, you know, to do, the fact that the
methodology that was used identified those SSCs and moved them up to Box
2 in fact should add to our level of confidence that the overall final result
would be pretty good.
MEMBER SIEBER: It's better than it otherwise would have
been.
CHAIRMAN APOSTOLAKIS: Yes. If we focus only on
3 -- and, again, I don't like the words "lower level of confidence," but let's
say we accept them for a moment. I mean, you know, it tells me that the
methodology has some merit, that it's not only something that tries to make
the life of the utility easier. In some sense, it went the other way.
MEMBER BONACA: I'm not at all arguing that. I'm
suggesting there is an inconsistency in the process. And it is important at the
site, because if you have a system with four trains, typically the expectation
of the operators, etcetera, has always been that you treat it in a special way,
because you have four trains not because you have four trains now you can
treat them individually less importantly.
Therefore, trains, because it supports an extremely important
system, you have that message. Now we are saying, well, but there are four
of those. Therefore, each one of them has less importance. And it may make
sense in a probabilistic sense. I'm only trying to understand what it does, the
message that it conveys, first.
And, second, the inconsistency between a high level system,
high safety significant -- however, the elements are less safety significant --
because each one of them is redundant to the other one. It's just -- it's a
personal concern I have with that, and I wanted to --
MR. STROSNIDER: I don't know if this helps at all to
answer your question, but I guess I would point out that the proposal from the
licensee is that these low safety significant components will maintain their
functionality. And also, the input we get from the PRA Branch was, yes, you
want to maintain the functionality of these.
And then, that's what we're looking at as one of the goals
of the treatment to be applied. So to whatever extent that influences your
concern there, the idea is that the functionality and design basis would be
maintained, and you might require a lower level of treatment to do that. But
it would still function.
MR. SCHINZEL: This is Glen Schinzel, South Texas.
Maybe to help put your mind at ease a little bit, as we go through the
categorization process, we do begin with system functions. And we look at
the functions to determine how significant those functions are. And then we
start mapping those functions against components.
Now, if we have a high safety significant function, the
component initially is going to be categorized high safety significant. Only
unless there is -- if we feel that there is multiple trains, diversity in satisfying
that function, we may be able to downgrade that component to a medium
categorization.
Generally, it's not going to fall more than one level. If it
goes from a high to a medium, it's still viewed as safety significant, and it will
still receive all of the special treatment requirements.
MR. CHACKAL: We should also clarify that the vast
majority of components remain at the risk of the system function that they
serve. It's the exception rather than the rule that a component's risk is
lowered. And, really, to clarity it further, as far as redundancy and diversity,
really, we don't -- we don't go across that much across trains.
CHAIRMAN APOSTOLAKIS: Would that be consistent
with your finding that only about six percent of the safety-related components
are really of high risk significance? How can that be consistent with what you
just said? I'm missing something.
MR. CHACKAL: Well, because not all system functions,
obviously, are high risk. There's a high, medium, low and not risk
significant. Most of the functions are not -- system functions are not high
risk. I don't know what the exact percentage is between medium and high.
But when you count up the total number of components,
percentage-wise the major components -- pumps and major valves, etcetera --
constitute a relatively small percentage of the overall population because in
that population you have a tremendous number of vents and drain valves,
instruments, things of that nature.
CHAIRMAN APOSTOLAKIS: I think the concern that Dr.
Bonaca has raised has to be looked at in the context of the fact that there is no
evidence that I know of that tells us that if we apply those special treatment
requirements the failure rates are different. I think that's a very significant
observation here.
It's not that, you know, we are applying -- we are less
confident, and we do less things. No. We are moving from one state to
another, and there is really no evidence, at least that I know of. It's a
presumption on our part that all of these things really improve the failure
rates. This gentleman said earlier that in the balance of plants they haven't
seen any problems. Of course, you haven't had any severe accidents.
MEMBER BONACA: I could supplement that comment to
say that, then, that puts into question at all the value of the quality assurance
program. I mean --
CHAIRMAN APOSTOLAKIS: And you think that's going
to hurt my feelings?
MEMBER BONACA: No. Maybe not. But I think it has
to be a coherent -- let me just express my concern again -- express about the
generic implication of this assumption. Here we are looking at it as an
assumption made, and the staff has a means of verifying that and checking that
-- I'm concerned about the generic implications.
For something of this nature, there had to be a position to
establish that says we will look at something and -- and something that gives
more comfort, that there isn't here a downgrading of the actual effectiveness
of a system by default, just because of the implementation of this.
CHAIRMAN APOSTOLAKIS: And what I'm saying is that
this degradation is the maximum in whatever scale you want. I mean, that's
an important observation. It's small.
MEMBER BONACA: Let me just make --
CHAIRMAN APOSTOLAKIS: It's small. I mean, I can't
quantify it, but I think it's small.
MEMBER BONACA: So now you have an RPS with four
trains. You could possibly conclude that since there are four of those each one
of the trains is not any more as safety significant as the RPS is. So you could
say, well, then, we don't have to have anymore the testing of this or the
checking of that or the verification or the quality of certain components. You
could say that.
For electrical components, I mean, we know there is a
dependency on quality of components and the reliability of their systems. So
I don't want to belabor it. I think I expressed my thoughts and my view, and
I think that there has to be some -- certainly on a generic basis, when this
approach is being used on a larger scale, there had to be a way to address this
issue.
MR. SCHINZEL: And, again, just to say one more time --
as we did the categorization process, we had safety significant functions.
Every system that was necessary or every component that was necessary to
satisfy that function received a high safety significant categorization initially.
And only unless we saw specific justification on why it
should be lowered, it was left high safety significant. And so just because
there are three trains or four trains, the exception is that there would be some
lowering. The rule is that those components remain high safety significant.
CHAIRMAN APOSTOLAKIS: Do you think two hours will
be enough?
(Laughter.)
MR. NAKOSKI: Yes.
CHAIRMAN APOSTOLAKIS: Can we move on?
MR. NAKOSKI: Yes, I think we should -- this is John
Nakoski. I agree, we should move on.
I'd just like to emphasize what the staff finding -- the
direction that the staff finding is heading in. And it's really we're focusing on
that treatment processes have the necessary elements, that if South Texas
effectively implements these elements they can conclude that they have
confidence the SSCs will be capable of performing their safety-related
functions.
CHAIRMAN APOSTOLAKIS: Again, I mean, for the big
picture, it seems to me the fact that there will be a monitoring program is of
extreme importance, is it not?
MR. STROSNIDER: And, actually, I was going to address
that on the next viewgraph with regard to the performance-based aspects of the
program. But I could talk about it now.
CHAIRMAN APOSTOLAKIS: Why don't we do it?
MR. NAKOSKI: We can go to the next slide.
MR. STROSNIDER: Actually, before you do that, leave
this one up here with the finding, because I think -- as John pointed out, the
finding that the staff is working toward at this point is focused on a very high
level in terms of the elements of the program and the idea that those elements
will be sufficient for the licensee to maintain functionality of the components,
without getting into how to do that. Okay? Some high level elements and
guidance.
In my mind, to understand that it's probably easier -- or it's
helpful to say what the finding is that we're not making. All right? We're not
making a finding that anyone who picks up the program as defined in the
FSAR and implements that program step by step will necessarily get
functionality or maintain the design basis when they're completed.
It's very high level. It's the elements. These are ways that
you can accomplish that. But there's a large degree of reliance on engineering
judgment of the licensee who is implementing it, and it's squarely in their
court, which it always has been. All right? And I want to point that out.
It's always been the licensee's responsibility to implement
their programs, but we say that on here to emphasize the point that the staff
is not going to make a finding that anybody could pick this up and implement
it and get those results.
That's a change in the finding that we were working toward
when we first started this. Okay? Where, you know, we were looking at
attributes at a lower level. All right? In terms of, as I say, the point that
somebody could pick it up and follow it and accomplish this.
Now, our basis for doing that is the finding that the PRA and
categorization process, that the components that are identified as low safety
significant, contribute small amount to risk. Right? And so this is the
reduced competence that we're talking about in terms of treatment, that the
staff doesn't need to know the hows. Okay?
But, nonetheless, there has to be some outcomes established
and some elements to make sure you can get there.
Now, in the ideal world --
MEMBER SHACK: Before you move that, Jack, let me --
I'm trying to read that last sentence and make sense out of it. If I take the
"whether" out, should I read this as the staff's finding regarding treatment is
that the licensee's treatment processes include the necessary elements, if
effectively implemented?
MR. STROSNIDER: The finding we want to make is that
it does include the necessary elements.
CHAIRMAN APOSTOLAKIS: Okay. And you're not
ready to make that yet?
MR. STROSNIDER: Well, actually -- I think actually we
have concluded that the elements are there, and we're talking about what level
of guidance might be appropriate within those elements. And this is where we
get into issues of --
MEMBER SHACK: So the "whether" is a yes.
CHAIRMAN APOSTOLAKIS: It's a yes.
MEMBER WALLIS: It's not really a finding. It's a
criterion for --
CHAIRMAN APOSTOLAKIS: It's a criterion. It's not a
finding.
MEMBER SHACK: Well, I think he's got a criterion up
there, and they actually have a finding. I think that the --
CHAIRMAN APOSTOLAKIS: To guarantee functionality.
MEMBER SHACK: Well, I think the criterion is whether
the treatment will include the necessary elements, if implemented, to assure
this. And the finding seems to be yes, it does. So as soon as we tweak the
commitment --
MR. NAKOSKI: Yes.
CHAIRMAN APOSTOLAKIS: I really think that sentence,
"although at a lower level of confidence," is misleading. I understand what
you are trying to say, but I think it's misleading. It's out of context.
MR. NAKOSKI: This is John Nakoski. I think our intent
is with sufficient confidence.
CHAIRMAN APOSTOLAKIS: Would be maintained.
MR. NAKOSKI: But it's --
CHAIRMAN APOSTOLAKIS: Why didn't you just stop
after "design basis conditions"? The function capability will be maintained.
MEMBER SHACK: Although with a lower level of special
treatment.
CHAIRMAN APOSTOLAKIS: So what? It will be
maintained.
MEMBER SHACK: It's an absolutely correct statement.
CHAIRMAN APOSTOLAKIS: Well, this is not absolutely
correct. What you just said may be absolutely correct.
MEMBER SHACK: There's little doubt that if I have less
treatment I have a lower level of confidence. Now, it may be only epsilon
lower, but --
CHAIRMAN APOSTOLAKIS: It's a matter of --
MR. STROSNIDER: It may not be necessary to support the
finding, but I think, you know, the reason we've had it in there going back to
the early SECYs is so that it's clear to people that this is a lesser level of
treatment, and that what goes with that lesser level of treatment is some
reduction in confidence.
If I can move on to the next part of this -- and, Dr.
Apostolakis --
CHAIRMAN APOSTOLAKIS: What you really mean is a
level of treatment --
MEMBER POWERS: Let me understand just a little bit of
-- why is it you think it's misleading, George?
CHAIRMAN APOSTOLAKIS: Because I don't think that
the level of confidence is high enough anyway. The lowering of it is really
minuscule. So, you know, to emphasize that I'm going to -- at the lower
level, I think you are sending the wrong message.
You have, then, to sit down and explain to people what you
mean by that. It's not an accurate statement. I mean, lower can be epsilon,
strictly speaking, is correct. But are you really conveying the reality by
saying that?
MR. STROSNIDER: Yes. There's a lower level of
confidence associated with this program in the staff's mind than there is with
the special treatment programs as exist in the current regulations.
MEMBER BONACA: If that level of treatment doesn't
make a difference, why are we here doing this?
MR. STROSNIDER: Right.
MEMBER BONACA: If level of treatment doesn't make
any difference, why are we here doing this? I mean, just -- let's just --
MR. STROSNIDER: Let me say something that might help
shed a little light on this discussion, too, with regard to the performance-based
aspects. And you pointed out the importance of a monitoring program.
So we indicate the finding that we want to get to is that the
licensee has the elements, and the licensee is going to maintain design basis
and functionality. If we look at the sort of guidance that's provided in Reg.
Guide 1.174 for risk-informed activities, and if we look at the move toward
performance-based activities, in the ideal world what we'd have, then, is some
sort of feedback loop to say, yes, in fact, these things are being accomplished.
All right. That would be a truly performance-based approach.
And one of the points we wanted to make with regard to
performance-based is that the degree to which the treatment process is
performance-based varies. And let me give a few examples. And let me start
off with perhaps the easiest -- maybe some of the easiest examples which are
environmental qualification and seismic qualification.
And you mentioned the fact that there's a lot of testing,
there's a lot of data with regard to the secondary side of the plant, and its
reliability. That's all true. But you have to remember that the purpose of
these special treatment rules is to determine whether these components and
systems will function during a seismic event in a harsh environment.
All right. So unless you're performing tests where you're
shaking the plant at a couple tenths of a g, or you filled the space with
radiation and steam, you're not really getting data back as to whether the
seismic functionality and the environmental functionality is maintained or not.
All right. And that's an important point to make.
Now, it shouldn't be a surprise, because under the existing
special treatment rules you don't get that information either. All right? But
the work that went into the rules was to provide some level of confidence that
without having that feedback, you know, you do the testing, you do whatever
you need to do to maintain this, such that if an accident occurs the equipment
will respond as designed. All right.
But, so -- and it wasn't there in the original special treatment
requirements, and it's not there in this program either, although I have to point
out, yes, if you're doing surveillance under normal operating conditions, for
example, and the component fails, that probably does tell you something about
what it might do under a more challenging environment. So, yes, you're
getting some feedback. But you really don't get feedback on those two cases
as to whether it will perform its function or not based on the sort of
surveillance testing that's done.
Now, on another case, though, if you look at in-service
testing, this is an area where, in fact, you can provide some element of
performance base. If you do testing and you look at the characteristics of the
valve function, you don't test it under design basis conditions necessarily, but
you can get information -- this has been developed over the last 10 to 15 years
-- that will give you some idea that -- or confidence that, yes, it will function
under the design basis.
So you can do some of that in this program. All right?
And, you know, the staff has maintained that -- some level of that targeted
grouping of components, less frequent than you might normally do, etcetera.
But you can build in some performance-based aspect, and we think that you
should where that opportunity exists.
So I think the point here is is that you need to recognize that
from a performance-based and from the monitoring that you get, there is some
limitations, and in some areas some fairly significant limitations, with regard
to showing that you're maintaining functionality. So you have to have some
level of confidence, albeit lower, but some level of confidence in the program
that's being presented.
So I think --
CHAIRMAN APOSTOLAKIS: I think that's a very valid
point, what you just said. It's very valid.
But I -- to take it a little further, though, and I think that will
address a little bit Dr. Bonaca's concern, I think the focus really should be on
the impact of the relaxation of the treatment requirements, the impact on the
possible dependent failure of these redundant elements. Because, yes, it
should make a difference whether I have two trains or three trains or four
trains. I mean, that's why I go through the expense, right?
But it's really the -- the dependent failure that I should worry
about, and whether I can see that in an accident condition -- you know, all of
them failing -- before I have some warning signals through tests or
inspections, and so on.
And are you focusing on that, the possibility of a common
cause failure? It is one thing to talk in the abstract about lowering our level
of confidence and quite another to make it specific and say, yes, they have
three trains. But if you don't do this, all three might fail. Now, then, you
have a very strong argument.
MR. STROSNIDER: We've had some discussions --
CHAIRMAN APOSTOLAKIS: Yes. And that relates to my
question earlier regarding the sensitivity study. I really want to understand it
a little better what you did there. If you have two components, two trains, and
you say the common cause failure -- let's say, in the beta factor model -- is
beta times Q -- Q is the failure probability of one train -- you increased Q by
10, did you do anything to beta? Did you ever change that?
MR. MOLDENHAUER: No, we didn't make any changes
to the beta.
CHAIRMAN APOSTOLAKIS: But wouldn't the relaxation
of special treatment requirements affect the value of the beta?
MR. STROSNIDER: Let me comment on that, if I could,
because we've had discussions with the PRA Branch. And we're trying to
make the treatment fit the categorization process. And I think, you know, as
I understand it, these different parameters you can characterize in terms of the
intersystem common mode or intrasystem common --
CHAIRMAN APOSTOLAKIS: It's the conditional
probability of failure of the second train, given the first one has failed. It's
very simple to think of it that way.
MR. STROSNIDER: Right.
CHAIRMAN APOSTOLAKIS: Okay. Is that conditional
probability changing as a result of the relaxation?
MR. STROSNIDER: Well, the input that we get from the
PRA folks when we talk to them is, you need to maintain functionality. All
right. Otherwise, it could impact the resulting risk.
A couple of comments on the sensitivity study, and South
Texas might want to expand on this. But I think one of the issues that comes
up in our mind is you talked about the factor of 10. I've heard reference to,
well, that represents a factor of three on the distribution of failure frequency,
etcetera.
I at least put the question out, is that the right question to
ask? Because I think what we really are interested in is when you change the
treatment, does that statistical distribution change? Does its mean change?
And you're not working any longer with the same distribution.
So to say you're going to three sigma on a distribution that's
contingent upon the existing special treatment rules, all right, that factor of 10
may not be as significant as it sounds. If you ask yourself the question, could
this treatment result in the mean value of a valve change failure from 10-3 to
10-2, then you might have a different perspective on it.
And the other thing that came up is exactly I think the issue
that you just raised with regard to when you do the sensitivity study, is it
looking just within a system, or is it looking across systems? My
understanding is that it does not really look across systems in terms of if you
have similar and you change treatment form.
Now, how that's addressed in the categorization I'm not
certain. But the bottom line of all that, that the engineering staff, the input
we're getting from the risk assessment reviewers is that you do need to
maintain functionality because of those issues.
CHAIRMAN APOSTOLAKIS: Well, functionality I
understand that you do have to do that. But, you know, I think, at least in my
mind, the most important argument in favor of the categorization that the
licensee has done is a sensitivity study. Fossil-Vesely and all of that, you
know, we can argue forever.
But if they say, look, an angel came down and told us that
these components are not safety significant. We're going to raise their failure
rate, and we'll show you that delta CDF and delta LERF is negligible.
According to 1.174, that should be enough. I don't care how they came up
with a set of components, right? Delta CDF and delta LERF is acceptable.
So I'm willing to forget about Fossil-Vesely and the criteria
and all of that. I mean, the sensitivity study they did is the key. So we really
have to scrutinize it.
MR. STROSNIDER: And you have to have confidence that
you've increased by the right amount --
CHAIRMAN APOSTOLAKIS: By the right amount, that
we did the right thing across systems, the common cause failures, and so on.
So it seems to me that's a very important thing that the licensee did, and we
really have to understand what that means.
Now, increasing the mean value of a distribution by a factor
of 10 is incredible, because that means the whole thing is shifting up, you
know, en masse. So that doesn't really bother me when it comes to individual
components, but it's not beta that bothers me, especially since they have
higher redundancy, so the -- the beta, gamma, you know, the Greek stuff.
Do you touch those when you relax the requirements? Is
beta now, instead of .1, maybe .8 or .9? Is gamma something else? And
given the limitations of the multiple Greek letter model, it deals only with
similar components in one system, not across systems.
MR. STROSNIDER: And I think this is where you get into
-- yes, and this is where you get into the challenge of understanding the
program that's being proposed, no longer the special treatment program, but
this program and how does it influence that analysis.
CHAIRMAN APOSTOLAKIS: Exactly.
MR. STROSNIDER: And that's a tough challenge, because
as you point out, and as I think we'd agree, when you start off with special
treatment rules, and you say, what level of confidence does that provide, and
how much do you reduce it when you go to this new program, that's a difficult
thing to get to, because as I said you really don't have performance data to
calibrate that.
CHAIRMAN APOSTOLAKIS: Yes.
MR. STROSNIDER: I'd like to come back to just one final
comment on this. I think we agree with your observations here, but in terms
of the big picture -- and it was pointed out that this has implications how we
move forward in risk-informing regulation.
I want to reemphasize the point that the approach we're
taking here is that staff is not going into a level of detail with regard to this
treatment program, because of the low risk significance of these components,
to say, yes, if you do it just like this you're going to get -- you know, you're
going to get functionality and you're going to get the -- maintain the design
basis. Right?
We're saying that it's just the right elements in a program,
that if a licensee does it properly you'll get that. In our mind, that's a reduced
level of confidence because we're not going to that level of review. All right?
And that's important when we look to going forward in the big picture.
What level of scrutiny is necessary? We've concluded that
we can use this approach.
CHAIRMAN APOSTOLAKIS: And this will guarantee
functionality, if the program is in place, is that what you're saying?
MR. STROSNIDER: That's why I wanted to go back over
this.
MR. NAKOSKI: If I could say -- this is John Nakoski. I'd
like to characterize it a little differently. Historically, the staff -- the NRC
staff -- has shared a certain amount of responsibility with the licensee for the
ability of their treatment programs to provide confidence that those
components would be functional, would be capable of performing their safety-
related functions. Because we would have gone out and reviewed the details
of how they would be implementing those programs, we would be looking at
the effectiveness of the implementation of those programs.
Where we are after the exemption, what this finding says is,
we are no longer sharing the responsibility with the licensee for assuring that
these components will function. We've recognized or acknowledged that for
this class of component, these LSS, low safety significant, and non-risk
significant components, the burden of providing confidence of functionality
rests with the licensee.
We need to look at the elements of those treatment processes
and make a decision as to whether or not if the licensee effectively implements
them, they could conclude that the components would remain functional. The
burden is on them. They are responsible for concluding --
CHAIRMAN APOSTOLAKIS: Let me understand this.
This is the only thing you are willing to grant them? I'm not --
MR. NAKOSKI: We are not going to say that their
programs will assure functionality.
MR. STROSNIDER: Let me answer. This is a lot -- this
is a lot to grant them, because it provides them the flexibility of how they
want to do this. What we're looking for in the FSAR are some high-level
expected outcomes, and there is some high-level guidance. But how they
actually accomplish that is up to them, and we're not going into that level of
detail. We're trying to grant as much flexibility and latitude as we can.
CHAIRMAN APOSTOLAKIS: So for Risk 3 components,
they will decide what the appropriate -- I mean, they will have the element,
but then how it's implemented they will decide how to do that.
MR. NAKOSKI: Yes, sir.
MR. STROSNIDER: Yes.
CHAIRMAN APOSTOLAKIS: So you're not going to
decide now that for environmental qualification, for example, you do this and
this and that. They will decide.
MR. NAKOSKI: That's correct.
CHAIRMAN APOSTOLAKIS: You will just have an
element there, environmental qualification.
MR. NAKOSKI: It would be in applying the treatment you
maintain the environmental design, environmental conditions. Let me -- the
capability of the component, the function under the design environmental
conditions.
MR. STROSNIDER: We would expect some high-level
guidance, as John just said, to maintain the design basis through the expected
life of the component to capture those kind of things. Right. And that's some
of the discussion we're having with South Texas now in terms of what level
should that be at.
But as I say, as we move forward, we take this approach.
We really are saying the staff is not going into any real level of detail with
regard to how these programs are implemented, and the other thing that -- the
reason I wanted to talk about the performance-based aspect of this is that you
also have to recognize that the feedback you get when this program is
implemented is limited depending upon the area you're talking about.
All right. But the basis for it is that the PRA and
categorization process provides confidence that these are truly low-risk
components, and that you don't need that additional level of --
CHAIRMAN APOSTOLAKIS: Jack, you are not, then,
distinguishing between a program that applies to Risk 1 or Risk 2 or 3. They
decide that.
MR. STROSNIDER: No.
CHAIRMAN APOSTOLAKIS: See, I'm confused. Is it
clear to everybody else?
MR. NAKOSKI: This -- our finding in this area, if we
could go to the previous slide -- that's not it. It's that one. Right there. Yes,
right. This finding really is for LSS, low safety significant and non-risk
significant.
CHAIRMAN APOSTOLAKIS: Three. Risk 3.
MR. NAKOSKI: That's correct. Risk 3 components.
CHAIRMAN APOSTOLAKIS: So for the others you are
still --
MR. NAKOSKI: For the other ones, those are still -- for
Risk 1 components, they are still subject to all the requirements and they are
--
CHAIRMAN APOSTOLAKIS: Using your words of a few
minutes ago, you still share responsibility with the licensee.
MR. NAKOSKI: That's correct.
CHAIRMAN APOSTOLAKIS: Okay.
MEMBER SIEBER: Let me ask a question to help clarify
some of this for myself. Let's say we have a component that's low safety
significance, but it's a Q component and it's -- environmental qualification is
required. And sometime during the life of the plant this component, the
physical component, wears out or dies, and then you have to go and replace
it.
The big expense for replacing it is to buy a -- a like
component that satisfies the original design specification, do the EQ test, and
meet the 18 criteria in Appendix B. That adds cost about 10 times what the
-- a non-Q component.
From this, I take it that all of that will still be required
because that's part of the design basis of the plant regardless of its risk
significance. Is that correct or not?
MR. NAKOSKI: No, I -- I would characterize it that when
they go to procure a replacement component, they need to look at the design
requirements for that component. What are the conditions, environmental
conditions that it would need to function under? They identified five methods
by which they would be able to procure a replacement component.
For example, they could go to a vendor catalog. If the
vendor catalog provided sufficient detail on the capability of that component
to operate under harsh environmental conditions, that could be sufficient for
them to conclude that it would be capable of performing the safety-related
function.
They would not be required to apply the Appendix B
provisions. They would not be required to test the component. They would
not be required to do an engineering evaluation that says -- or analysis that
says, "This component can function."
MR. SCHINZEL: This is Glen Schinzel. Let me carry your
example a little bit further with some specifics. If that EQ component -- let's
say it was designed to withstand 250 degrees and 100 percent humidity.
MEMBER SIEBER: And a bunch of radiation and -- okay.
MR. SCHINZEL: Well, let's stop with the first two.
MEMBER SIEBER: All right.
(Laughter.)
MR. SCHINZEL: If we -- if that component failed and we
originally had a safety-related EQ-qualified component installed, we're saying
if that is low safety significant we could, based on the temperature
requirements and the humidity requirements for the equipment qualification,
we could buy a commercial component that satisfies those environmental
requirements and have confidence that when that component is installed it will
perform and satisfy the designed functional requirements.
We don't need it to be safety-related, and we don't need it
to be qualified.
MEMBER SIEBER: On the other hand, when you go to the
vendor's catalog, what you read in there is a vendor's claim. That somehow
or other should be substantiated by some kind of a test that the vendor did on
a prototype, for example.
Most of the vendor catalogs that I've seen don't give you that
kind of detail. They'll say it's explosion-proof and has a NEMA housing and
this kind of thing, and it's up to the eye of the beholder to imagine the degree
to which it can withstand a harsh environment.
MR. SCHINZEL: Well, again, we would expect that we
would go to reliable vendors that we have confidence in. And, again, we'll
take a look at the component upon receipt. And if it doesn't appear to be of
the quality that we would expect to satisfy those requirements, we would look
for a different type of component.
But as far as requiring a test from the vendor to validate or
to qualify that this component, beyond a shadow of a doubt, will satisfy these
requirements, we don't feel we need that for the LSS or the NRS components.
MEMBER SIEBER: The harsh environment qualification
is part of the design basis of the plant, right? For various components.
MR. STROSNIDER: Well, a couple of comments on this
example. I think it's a good one. With regard to the procurement -- and
there's five other approaches that are listed in these programs that could be
used. But with regard to procurement as an example, what the staff has
suggested is that if the vendor provides a statement that this satisfies the design
conditions that you've put in your purchase spec, that's acceptable. All right?
So, you know, without saying, "Well, the vendor has to have
test records, or the licensee has to have test records," if you put this in the
purchase spec and the vendor comes back and says, "Yes, this satisfies your
spec," we've said that that looks acceptable to us.
To back up and look at this thing from a bigger picture -- EQ
just, again, as an example -- when you look at the EQ rule, there's 10
different things that you have to deal with in there. And they talk about -- it
talks about things like radiation, humidity, submergence.
MR. NAKOSKI: We're going to get to this later.
MR. STROSNIDER: Oh. This is going to be covered later.
So maybe we ought to just move on.
Except let me just summarize this. We went through those,
and we tried to identify which ones were design basis and said, "Well, you're
going to maintain the design basis, and seven of those 10 fit into that
category." However, with regard to documentation and margins, there's room
for relaxation, and that's what we can talk about. But I guess John has
actually got that --
MEMBER SIEBER: I have one -- one corollary question.
If you buy an item commercial grade, even though it matches your spec and
you wrote the spec properly, which is a difficult job, does that relieve the
vendor of the obligation under Part 21?
MR. NAKOSKI: Included within the scope of the exemption
request is relief.
MEMBER SIEBER: Relief.
MR. NAKOSKI: So, no, it --
MEMBER SIEBER: So that would occur in category 3
components.
MR. NAKOSKI: That's correct.
MEMBER SIEBER: No more Part 21.
MR. NAKOSKI: That's correct.
MEMBER SIEBER: So the staff, through the equivalent of
AEOD, what used to be AEOD, would not have a way to know if there is
some defective component out there in the industry floating around, nor would
other utilities or -- right?
MR. STROSNIDER: Only if that same component had a
problem in a safety-related Box 1 application.
MEMBER SIEBER: Caused an event, right.
MR. STROSNIDER: Because those would still be
reportable, but, you're right, it reduces the population of that component from
which you might get information.
MEMBER SIEBER: We have covered two out of 11 slides
in just one hour, so we should finish sometime around midnight.
MR. NAKOSKI: Okay. Well, let's see if we can do better
than that.
MEMBER SIEBER: All right.
MR. NAKOSKI: And if we could, I'd like to go to the first
slide, that's slide number 6, that talks about some of the open items that we
have.
Okay. Open Item 4.1 -- and we talked about this earlier --
treatment for the high safety significant, medium safety significant
components, we've reached agreement in principle on what level of detail
needs to be in the FSAR. South Texas just needs to finalize that in the final
submittal.
We are pretty much comfortable that they're going to target
the treatment to the reason that these components are HSS or -- and -- let me
say and continue to apply the current special treatment requirements to the
extent they apply to both safety-related and non-safety-related.
MEMBER LEITCH: So this basically represents no change,
then, right? This is --
MR. NAKOSKI: Well, the change is enhanced treatment.
MEMBER SHACK: For the Risk 2 component.
MR. NAKOSKI: For the Risk 2 components, and actually
for elements of the Risk 1 components that are beyond design basis, if they
should take credit for those.
MEMBER LEITCH: Okay.
CHAIRMAN APOSTOLAKIS: So this is categories one and
--
MR. NAKOSKI: One and two. That's correct. If I can
move on to Open Item 4.2 --
CHAIRMAN APOSTOLAKIS: The notation is so
confusing.
MR. NAKOSKI: Well, I'm trying to be consistent with
South Texas' vernacular here.
Open Item 4.2 -- we are still working with South Texas to
resolve some of the level of detail in here. There's two sections that are
underlined under the first check mark, procurement process, inspection test
and surveillance process.
I think we still need to work with South Texas to come to an
agreement in principle on what the words in the FSAR should be. And at this
time, I'd like to let South Texas give some of their thoughts on where they see
we're at with that.
MR. SCHINZEL: Okay. Glen Schinzel, South Texas.
Again, from the standpoint of the FSAR, the FSAR right now is including
about 14 pages of additional detail that's going to be included. That's to
address some of the whats, but we also consider that it's still including a
number of the hows to detail a bit on the methodology that does need to be
pursued.
With the procurement process, you know, what we anticipate
is that we're going to use a program very similar to our commercial grade
program, and that follows the guidelines of SECY-98-300 with the intent for
Option 2. As we started into our exemption submittal and some of the follow-
on questions and RAIs, the staff was looking at trying to make a finding based
on functionality. That did require South Texas to submit an extensive amount
of detail in order to ensure that the staff was satisfied with our processes and
approaches.
As the staff is now moving to thinking toward just wanting
to understand the whats and not the hows, a lot of that detail that was
originally provided has remained. So we're still looking, like John says,
specifically in the areas of the procurement process, inspection, testing, it does
include some of the maintenance and the management and oversight processes,
identifying what we consider to be an appropriate amount of detail to be
included in the FSAR. We do feel that 14 pages is a bit much right now.
MR. NAKOSKI: Well, to a certain extent, I feel I have to
clarify. There are a lot -- a lot of those 14 pages go into the description of the
categorization process. The categorization process is the cornerstone for the
staff to move forward with granting the exemption.
I think there are probably four or five pages in the FSAR
section that discuss the elements of the treatment process. So just to be very
clear on what is provided in those 14 pages, it's important that you recognize
the bulk of that relates to the categorization.
Okay. Are we ready to move on to the next slide?
One of the areas that we had an open item in that's kind of
related to treatment is the Open Item 7.1 that dealt with the quality assurance
exemptions that South Texas had looked for. They've agreed that they should
submit an update to the Quality Assurance Program. We've looked at that and
have concluded that it's acceptable, and we'll note that in our safety
evaluation. And that closes this open item.
Open Item 8.1, that's the open item that asks South Texas,
why do you need the exemption to 50.49? They've told us why they need the
exemption. We accepted that. But in doing so, it opened questions in our
mind about the elements of the treatment process that need to be described in
the FSAR related to environmental design conditions.
We see that in 50.49 there are seven what we consider
design requirements -- temperature and pressure, humidity, chemical effects,
radiation, aging, submergence, and synergistic effects. We see those as design
requirements that need to be considered when procuring or maintaining or
implementing the treatment processes that need to be maintained.
We see that in the rule we can grant relief documentation.
50.49 is fairly prescriptive on the documentation that's required for these low-
risk components, Box 3 or Risk 3 components. We agree that you don't need
to meet the prescriptive requirements of documentation in 50.49.
We agree from these low-risk or Risk 3 components there
can be a relaxation of the margins when you do the testing or when you do --
well, basically, it applies to testing.
And then, thirdly, the methods by which you can provide
confidence that these components will be able to perform their functions under
design basis conditions. 50.49 relies very heavily on testing and test data.
We are trying to allow methods other than that to procure essentially --
primarily in the area of procurement, to allow them to procure replacement
components without the need to do testing.
This is an area where the staff still needs to come to internal
consensus on what we would expect the element of the treatment would do,
what it would accomplish. And we are not yet at a point where we can fully
discuss with South Texas where we are with that.
If I can, I'd like to move on to 10.1 and 10.2, or if South
Texas has some comments they would like to make.
Okay. Open Items 10.1 and 10.2, just briefly, it deals with
repair and replacement and ISI requirements. South Texas has proposed the
treatment for that, proposed a response for that. There is one area that they've
proposed that we still need to assess and that they are going to be providing
us additional insights on, and that's one of their alternatives -- they proposed
three.
One of their alternatives for repair/replacement would allow
them to -- I'll shorthand it -- mix and match requirements from various codes.
And we're working with them to bring that issue to closure.
If I can, I'd like to go on to the next slide.
One of the open items we identified related to the need for
an exemption to IEEE 279 as imposed through 10 CFR 50.55(a)(h). They
provided an adequate reason why they needed the exemption. We recognize
it's an independent regulation that imposes qualification and quality
requirements. In the draft SE, we accepted the quality provisions that they've
described in their processes. What remained was the qualification provisions.
That's really embedded in the resolution of Open Items 8.1
and 18.1, which deal with environmental qualification under 50.49 and seismic
qualification under 10 CFR Part 100, Appendix A. When we close those two
open items, that will close this open item.
Another treatment item was, what was the scope of the
exemption they were seeking from the maintenance rule? And they provided
clarification in their submittal that they aren't seeking an exemption to
50.65(a)(4), and that's acceptable to the staff.
Open Item 18.1 on the seismic qualification exemption --
they have -- again, the open item is, why did South Texas need the exemption?
They've explained that to us. We accept their explanation, and, again,
resolution of the treatment issues is necessary to establish the elements in the
FSAR so that we can bring this item to closure.
We see the requirement -- the design requirement that needs
to be retained is that these SSCs are designed for the earthquake motion
described in their design basis, which is the seismic inputs and design load
combinations. But the methods for confirming the capability, we think if
they're consistent with the elements of the treatment process that's sufficient.
Again, primarily, this relates to the area of procurement.
They procure a replacement component. They need to make -- consistent with
the description of the elements in their treatment process effectively, that
should provide us confidence that the -- should allow them to conclude they
have confidence that these components will remain functional.
Is there anything you wanted to add at this point?
MR. SCHINZEL: Not at this point.
MR. NAKOSKI: Okay. Just for completeness, there were
a couple of open items that weren't necessarily treatment or categorization.
I thought at this point it would be good to go over those, just to let you know
what the status of those are. And that's the two confirmatory items.
The first one was areas of inconsistency in their South Texas
submittal that we identified in our draft safety evaluation. South Texas has
provided us with their resolution to these open items, and our plan is to
discuss that resolution in our final safety evaluation and allow -- and we would
consider this item closed.
The other confirmatory item is within the scope of the 50.59
exemption that South Texas requested, they asked the staff to allow them to
remove all of the commitments to these -- this Risk 3 group of components.
And our response back to them was, "There's a process that exists that the
industry put forth, and that the NRC has endorsed in NEI 99-04. We would
ask that you would follow that guidance."
And the reason we asked them to do that is there -- to give
such a blanket exemption to 50.59 we didn't think was prudent. There could
be changes to treatment that are beyond the scope of the exemptions that they
requested, and we needed to have a clear understanding of what that scope was
before we could grant it. And we recognize that there is a process out there
that would allow them to change commitments.
The final open item that still was out there that didn't deal
with either treatment or categorization was, how do you control changes to the
processes upon which the staff will base its exemption, because this is kind of
a unique situation where an exemption is based -- but we're approving
processes as the basis for granting the exemption. It's not something we
would do normally. So how do you make sure that the basis for the
exemption remains valid?
The first thing is we need to have a licensing basis
description of these processes. That's provided in the FSAR section. Second,
is if they are going to change that process, they can't really reduce the -- either
the effectiveness provided in the categorization process or the assurance in the
treatment process that these components will remain -- these Risk 3
components now primarily will remain capable of performing their function,
and that they need to continue to do some sort of assessment, evaluation, and
oversight.
So, really, we're not looking for allowing them to reduce the
effectiveness or assurance provided those. They make enhancements that,
again, relies on their judgment to determine. Even when they do make
changes that don't reduce, we'd like to know about it.
And, finally, if they do make a change that reduces it, before
they implement it it has to be approved because, in essence, it changes the
basis upon which we granted the exemption.
And South Texas and the staff have come to consensus on
-- or agreement in principle on the language in the FSAR, and we're working
with finalizing the specific wording on the change control process to be
incorporated in the FSAR. And also, that will be used as a condition of the
exemptions that are granted.
With that, that's all I had to discuss on the treatment open
items for South Texas, and I'll turn the floor over to South Texas and let them
proceed with their presentation.
MR. HEAD: Are there any questions before we start?
MEMBER LEITCH: I just had one question. I'm looking
at the minutes of a mid-February meeting that occurred. And in those minutes
it -- these are NRC minutes, but it attributes to South Texas the statement that
South Texas believes the NRC staff positions on assurances of functionality
for LSS and NRS safety-related SSCs has paralyzed the effort.
I guess my question is, I think we heard presented here a --
what seemed to be a fairly manageable list of open items well on their way to
resolution. Has there been that much progress since the February meeting, or
is this still a fair characterization? I mean, this makes it sound like you're
miles apart, and I --
MR. NAKOSKI: I'll share my perspective on that. I think
that NRC has moved substantially towards allowing South Texas to implement
their commercial treatment practices. I will say, the fundamental change that
occurred is that the staff is relying on South Texas' engineering judgment to
effectively implement the elements of this treatment process that they've
described.
We are no longer asking them how they are going to
implement these programs. Maybe there is some disagreement that in some
instances South Texas may think we are, but I think we are trying to stay at,
what is the outcome that we expect from the elements of these treatment
processes? So I think we have moved.
MEMBER LEITCH: And that kind of changes since the
mid-February timeframe.
MR. NAKOSKI: That's correct.
MR. HEAD: This is Scott Head. Our presentation I think
will speak to that question.
MEMBER LEITCH: Okay.
MR. HEAD: We would agree there's been movement, and
we -- so we applaud the effort that the staff is continuing to see here. But, like
I say, our presentation will give our perspective based on what we're seeing
right now.
MEMBER LEITCH: Okay. I'll defer that until I hear your
presentation.
MR. SCHINZEL: Okay. Again, I'm Glen Schinzel from
South Texas. As far as the treatment status, just a few bullets of overview.
With SECY-98-300, that was really the initiating document that allowed South
Texas to start pursuing a risk-informed initiative with the LSS and NRS
components.
And, really, from a SECY-98-300 standpoint, the approach
was for the LSS/NRS components that commercial practices would be
sufficient and appropriate for the treatment for these components. As we've
gone through a couple of years of working with the staff on resolving this,
initially the staff was working toward making a finding based on functionality.
And, again, that required an extensive amount of detail that
South Texas provide, and the staff recently, since the February meeting, has
made some movement in trying to focus more on the whats, rather than getting
into the details of the how. And so we do see that as a positive step. But like
John did mention, there is some disagreement on the level of what constitutes
a what and what constitutes a how.
Now, you hear of wolves in sheep's clothing. We think that
there are some hows in what clothing.
(Laughter.)
So we're still trying to work through those details, and that
is, you know, some of the detail that's currently in the 14 pages of the FSAR
that we feel is going a little bit beyond what it needs to right now with the new
approach.
Based on the current status, we do have several concerns.
In the area of seismic, it was mentioned that that's an area that we're still
working with the staff on, and there are some issues here that we need to work
through. We don't see immediately a real clear pathway for success for
implementation.
One of the design inputs for safety-related, seismically-
qualified components is that the component has to satisfy five OBEs --
operational bases earthquakes -- followed by one SSE. And for us to go out
and try to commercially procure and be able to demonstrate that this one
component is, in essence, going to satisfy five OBEs and five SSEs,
essentially, there is not a clear pathway on us ever being able to do that unless
we have an extremely detailed analysis or we physically go out and test the
component.
So as far as the benefit to be seen with seismic, right now we
don't see a large benefit that we'll be able to implement and receive.
In the area of equipment qualification, John has already
covered this in some detail. And it is an issue that is still with the staff. The
staff is still internally discussing that, and we're anticipating some additional
discussion on this short term.
For the areas where the regulations currently require for
safety-related components some type of testing -- this is whether it's in-service
testing type testing and maybe inspections. One of the feedbacks that we're
getting from the staff is, again, if it's removed from the regulations, if the
exemption is granted, the expectation is that there is still some type of testing,
inspection, surveillance program that would be retained on these components.
We agree that from the standpoint of a commercial
application that we want to ensure the functionality of these components. The
disagreement really is, does this need to be bound in our FSAR as an
obligation that testing, surveillance, inspection, will be occurring? And we
feel that the answer there is no, we don't need to have that detail in the FSAR.
With the detail in the current FSAR, you know, I mentioned
that there are 14 pages. Five of those pages are dealing with categorization
and the details there. The others are dealing with treatment. And based on
the meetings yesterday and the day before with the staff, there still is some
additional detail that we are looking at needing to add into the FSAR to satisfy
some of the staff's needs. So, again, we feel that that's going beyond what is
necessary.
Some of the specific areas with the procurement area --
we've talked to this a little bit, but we do get into the FSAR talking through
five methodologies of how we can go through -- and I'll emphasize the how
we can go through and do the procurement.
It also covers the documentation, the handling requirements.
The maintenance process gets into discussions of the design life,
considerations for corrective/preventative/predicative maintenance. The
management and oversight process gets into the
qualification/training/certification of personnel, how you'll handle measuring
and testing equipment.
So, again, we see that as some details that are getting into
still the hows as opposed to an expectation that the design functional
requirements will be satisfied, and commercial practices will pick up, and
those same good commercial practices that are present today that are working
effectively for us on the balance of plant side will still be used.
South Texas has seen progress in the staff's approach. We
haven't seen as significant progression as what we would have liked to have
seen. We would have liked to have been at a point where we could have
considered all of the open items by this time. But, again, we do have some
items that are still open, and we'll continue to work with the staff on these.
The actual timing for the exemption -- John mentioned that
there was a target schedule. Again, that schedule is somewhat fluid and in
doubt right now as to exactly when the exemption will get granted. So that's
still a concern to South Texas.
And, again, just to have the committee understand that South
Texas is a prototype pilot. There are others in the industry that are watching
the South Texas effort very closely. And I know that there is some
apprehension among some of the industry players on whether this is an
appropriate process to step into. And so the future of Option 2 is -- is
somewhat in question right now.
So that details my comments, and I'll address any questions.
MEMBER SIEBER: Does anyone have any questions?
MEMBER POWERS: One I just can't resist asking. Why
should I be concerned about the future of Option 2?
MR. SCHINZEL: Well, from a South Texas perspective --
MEMBER POWERS: Put that aside. You said some of the
industry were looking at this with trepidation about going into it. I guess I
would look at, with trepidation, going into it as well.
MR. SCHINZEL: Well, with -- one of the efforts that the
Commission is looking at pursuing is risk-informed regulations or a risk-
informed approach. Option 2 is a very important cog in moving forward with
the risk-informed applications.
If there are no other players other than South Texas who are
willing to embark upon an Option 2 program, the Commission's goals of
moving forward with the risk-informed approach could be at risk.
MR. STROSNIDER: This is Jack Strosnider. I just want
to point out that from NRC's perspective risk-informed regulation does
leverage our four outcome goals. We talk about maintaining safety,
efficiency, and effectiveness, reducing unnecessary burden, and public
confidence. And we focus on the high safety significant areas -- maintain
safety, pointed out some -- some components might get additional treatment
or different treatment under this than they were before.
So risk-informed regulation, Option 2 and Option 3, they go
toward those goals. I think -- nobody said it would necessarily be easy,
though. We need to make sure that we do it technically correct, so that we
really come out on target with those goals. But it is part of our policy and
part of what we're working to implement.
CHAIRMAN APOSTOLAKIS: If the categorization process
were perfect, would we have all this debate regarding treatment? If you were
really convinced that these things were of low safety significance, Category
3, why would you -- would you still care that much about the treatment?
MR. NAKOSKI: This is John Nakoski. Under the
conditions that were established for Option 2, we need to have some or
sufficient confidence these components will remain functional. They will be
able to do their job.
CHAIRMAN APOSTOLAKIS: Is that question addressed
in the PRA, like the seismic example that Jack mentioned earlier? When you
did the sensitivity study and the importance measures, you included seismic
events?
MR. MOLDENHAUER: Yes, we included all the external
events in that. And we went and did another -- a further sensitivity study at
the behest of the NRC to look at just seismic -- if we had a seismic event,
would that change any of the categorization. And we found no changes in
categorization for just seismic events.
CHAIRMAN APOSTOLAKIS: Is it the credibility of the
seismic analysis, then, that is in doubt in your mind?
MR. STROSNIDER: Well, I think -- and I don't know if
any of the staff has been involved directly in that and maybe can comment on
the seismic PRA sensitivity study. But I think one of the challenges you
always have when you go into these sensitivity studies we discussed earlier is,
what's the starting point for the inputs to the PRA in terms of what's the
conditional probability of failure given a seismic event, and how much does
it change when you change the treatment?
And I'm not -- that's a difficult thing to get a handle on. So,
you know, the approach when we talked about the treatment process here is
focused more on what elements and what things do you need to consider in
order to maintain that functionality. And it's perhaps a more qualitative
approach than what you see in a PRA.
But I think with all the sensitivity studies, as I mentioned
earlier, there is always difficulty I think in establishing the input distributions
for these hypothetical design basis situations, which, you know, there is really
no empirical data out there, other than perhaps what you get -- when you
actually go out and you do seismic qualification tests and when you do
environmental qualification tests under the existing treatment rules, you do
see, you know, some of the components fail. So you can get some insights
from that, and it can help you to say what's important in terms of the
treatment process.
MR. NAKOSKI: The other thing I'd like to add is even
when you do the sensitivity studies you don't assume that these components
fail. You expect that these components are going to be able to do their
function with some reduced availability. So it's not -- it's expected that these
components will be able to do their function. Whether that's an assumption
going into it or an expectation, we could argue the words. But the bottom line
is we expect that these components will be able to do their function.
So how -- what gives you confidence that these components
will be able to do their function? Some method for assessing that, and we call
that treatment, special treatment under the current regulations, commercial
treatment as for South Texas, we're trying -- we're working to get described
in their FSAR.
CHAIRMAN APOSTOLAKIS: Well, that's true, then. If
PRA does assume that functionality is there, then it takes it from there.
There's a probability they will fail.
MR. NAKOSKI: That's correct.
CHAIRMAN APOSTOLAKIS: So you're saying that the
fundamental assumption of the PRA, that the thing is capable, in fact, of
carrying out this function -- comes into question under certain questions. And
the thing that gives you confidence that this is true -- this will be through
special treatment.
MR. NAKOSKI: Or some level -- method of treating.
MEMBER POWERS: Just an element of curiosity. When
you do your seismic-only calculation, do you consider seismically-induced
fires?
MR. MOLDENHAUER: No, we didn't include that.
MEMBER POWERS: But seismic events can induce fires.
MR. MOLDENHAUER: Yes. We do have a fire PRA, but
we don't have any of the seismic-induced fires.
MEMBER SIEBER: Have you ever run your PRA model
with the Risk 3 components failed, instead of just changing the failure rate?
CHAIRMAN APOSTOLAKIS: Well, I'm sure they will see
an impact.
(Laughter.)
MEMBER SIEBER: I would think so. But it would be
interesting to know what it is, and that tells you something about the validity
of the Risk 3 categorization.
MR. SCHINZEL: Well, I guess one clarification I want to
bring up is, you know, having something seismically qualified and having
something that's not seismically qualified doesn't necessarily mean that that
component that's not seismically qualified is automatically going to fail in a
seismic event.
MEMBER SIEBER: I understand that.
CHAIRMAN APOSTOLAKIS: We have evidence of that.
MR. SCHINZEL: Absolutely. There is evidence in
California, in Japan, where there are non-qualified components that operate
just fine, and they perform their function well when demanded. And there is
evidence where there are some qualified components that, when demanded,
will fail. So --
MEMBER SIEBER: So the answer to my question is no,
you never run that kind of a study?
MR. SCHINZEL: Well, again, you're going to see an
impact because you're guaranteeing failure of the majority of the components
that are modeled in the PRA.
MEMBER SIEBER: That's right. And its only value is a
value of curiosity, that it puts a bound and gives you some kind of assurance
as to the --
MR. STROSNIDER: Part of the discussion we've had with
the PRA reviewers -- it doesn't go that far to say fail them all at once, but I
guess the categorization process has a raw value of two. Right? And there is
at least the indication that if you pick the wrong combination of these low
safety significance, you could exceed that.
I mean, that doesn't mean you have to fail all of them at
once. So, and that's part of when we have these discussions, saying, "Well,
yes, we really want these things to function because of that -- that
consideration."
CHAIRMAN APOSTOLAKIS: It seems to me any
sensitivity study that is being done should be done in the context of the
decision we have to make.
MR. STROSNIDER: That's true.
CHAIRMAN APOSTOLAKIS: And I think this would be
an extreme case, Jack. But even in what Dr. Powers said, the question is,
okay, the PRA doesn't have fires induced by earthquakes. But the real
question in front of us is, if we relax the treatment, is that going to be a more
likely event? Are we doing anything now that's related to that issue, that if
we relax it, then, my God, you know, next time you start shaking you are
going to start seeing fires all over the place.
That's really the decision we are facing, and let's not forget
that. I mean, otherwise, we can start talking about the limitations of the PRA
in the abstract, and that will not take us very far. It's the decision that
matters, and I think it's relevant to what Dr. Bonaca was saying earlier, what
Dr. Powers said, and what Mr. Sieber just said.
We're not talking about wholesale change of the regulations,
I don't think. No. That may be -- we are not talking about fundamental
changes in -- major changes in the failure rates, except I am willing to accept
Jack's example with the earthquake. I mean, you really don't know that. You
rely on Jack's -- the conditional probabilities of failure, and so on, is
something that's largely a matter of judgment.
MR. STROSNIDER: And I think for environmental
conditions, too. If you look at the testing, you know, under harsh
environments you can get -- get interesting results.
CHAIRMAN APOSTOLAKIS: It's really -- I think the --
eventually, we're going to think -- be thinking about some rule or something
to -- you know, for Option 2. This is just proof of concept.
MR. STROSNIDER: Yes. But I -- I mean, I -- in my mind,
of course, we're learning experience as we go through this. And some of the
-- if you'll look at the -- what's been characterized as the whats and
expectations of what a program would accomplish, those sort of things, in my
mind, at least would find their way into a rule at some point.
So, I mean, we --
CHAIRMAN APOSTOLAKIS: Now, there you would have
to revisit the whole issue of categorization. You know, I --
MR. STROSNIDER: Well, they have to complement each
other in the decision.
MEMBER SIEBER: Maybe I could expand on that question
a little bit, because I've thought about it. Let's say that South Texas and the
staff reach agreement on everything, the exemptions are granted, and there is
all of these other licensees sitting out there with their papers in hand to come
in and ask for Option 2 exemption requests.
It may look something like South Texas' original request or
as it has been modified, but perhaps the plants are different. Would you treat
things like categorization differently because the risk profile of a plant was
substantially different than South Texas?
MR. NAKOSKI: This is John Nakoski. I think we would
have to look at the quality of the PRA or assess the capability of the PRA for
the specific plant to support a robust categorization process.
For South Texas, I think we have looked at their PRA with
sufficient detail to come to the conclusion that the PRA and the risk profile
derived from that is sufficient to support moving forward with an exemption.
For other plants, we would have to look at the PRA to some level to be able
to conclude that it is sufficient to move forward in exemption space. And I
think that's consistent with the guidance in Reg. Guide 1.174.
MEMBER KRESS: Yes, but that's really not the question.
Of course you look at the quality of the PRA. You have to do that. But the
question is, assuming you do that and you -- your assessment is that the PRA
is of sufficiently good quality for this, but the PRA tells you that this
particular plant has a risk status that's significantly worse than South Texas --
MR. STROSNIDER: This is --
MEMBER KRESS: -- now, are you going to treat the
categorization process any differently?
MR. STROSNIDER: This is Jack Strosnider. I would be
much more comfortable having some of our PRA staff answer that question,
quite frankly. But the one comment I would make on it is that one thing we
know is that we have to have confidence in that categorization process. That's
a cornerstone of this whole approach.
And so I think it's a very valid question, and I think it's
something we certainly would need to take into consideration. But I think we
probably ought to have some of our PRA people --
CHAIRMAN APOSTOLAKIS: A related question is both
Fossil-Vesely and raw deal with fractional changes of risk.
MEMBER KRESS: Yes, that's true. That's the source of
the --
CHAIRMAN APOSTOLAKIS: In South Texas, it's low
because of the redundancy, and so on. They are nice people.
If you get another plant where the core damage frequency is
A10-5, you still go with the fractions. Fossil-Vesely still is .005.
MR. NAKOSKI: This is John Nakoski. I don't think we
have the right people here to answer that question. I think it's a valid
question.
CHAIRMAN APOSTOLAKIS: But we have it in the
transcript.
MR. NAKOSKI: I'm sorry?
CHAIRMAN APOSTOLAKIS: We have a transcription,
and these questions are being transcribed.
MR. NAKOSKI: And we appreciate the question, and I --
CHAIRMAN APOSTOLAKIS: But this is -- these are the
things that concern some members of the committee. And one step further,
why should the agency have four cornerstones when it comes to the oversight
process, but when it comes to these things it deals only with CDF and LERF?
Why don't you guys care about initiating events in this case?
In the oversight process, we do care about them, and we
don't want to have too many of those. But then where we're talking about
relaxing special treatment requirements, we just focus on CDF.
MR. NAKOSKI: Well --
MEMBER SIEBER: The problem with that, just to expand
your question a little bit, is when you license the plant you licensed it to
50.34, which had some performance-based requirements in it, and Appendix
A, which lays out another bunch of performance-based requirements -- 50.49.
And if you lay all of those out, do they reflect themselves properly in using
CDF and LERF as surrogates for all of these possibilities?
MR. STROSNIDER: Yes. I think those are all fair
comments and questions. And I guess the one thing I just want to come back
to, one of the objectives we had today, NRC staff, in terms of the treatment
process here is to try to explain clearly what the approach is that we're taking.
MEMBER SIEBER: Right.
MR. STROSNIDER: And then what it does and what it
doesn't, and there is -- you know, there's issues involved there that we have
to be comfortable with as we move forward. So I hope that we've at least laid
that out, and I think that --
CHAIRMAN APOSTOLAKIS: I wonder whether in your
deliberations regarding treatment you really used in the back of your mind the
four cornerstones of the oversight process, not just the Fossil-Vesely and raw
--
MR. STROSNIDER: Well, I think --
CHAIRMAN APOSTOLAKIS: You don't want to see any
small LOCAs or --
MR. STROSNIDER: Sure. In general, this whole process
-- yes. In general, I think we need to back up and look at the integrated
program here. And, in fact, in our safety evaluation, we're looking at the
Reg. Guide 1.174 sort of approach which says, you know, you need to take
into account defense in depth and margins, and that sort of thing. So we
haven't lost sight of those, certainly.
CHAIRMAN APOSTOLAKIS: So the question is whether
part of the reason why there is this extended debate is that the categorization
process uses one set of metrics, but then when you guys evaluate the treatment
you use another set. I think that --
MR. STROSNIDER: I think there's a real challenge if you
try to correlate the two in some quantitative way.
CHAIRMAN APOSTOLAKIS: I mean, and I think the
oversight process has made progress there, by identifying the four
cornerstones and saying explicitly, "We do care about the frequency of
initiating events." The agency does care. It's not just CDF and LERF.
MR. STROSNIDER: I agree with all that. But I do come
back to stress again that when you talk about special treatment rules, that, you
know, they were written largely to mitigate -- to deal with components,
functions, and mitigating accidents, when you look at EQ and seismic. And,
I mean, we've been over this several times now, but, you know, you don't get
much data on that sort of thing.
MEMBER SIEBER: Another question is, you used an
operating phase PRA as the basic structure of the risk analysis to support
categorization. Have you looked at -- and you also discussed the fact that you
used an IPEEE type analysis. Have you done shutdown PRAs or operating
transient PRAs? Because that's part of the risk -- overall risk profile. And if
you have or haven't, do you think that the -- the risk analysis by components
would be different, depending on what phase of the operation you're in?
MR. MOLDENHAUER: From a PRA perspective, we
haven't included the shutdown PRA analysis or any transient analysis.
However, from the deterministic process that we include into the risk ranking,
we do ask those questions, and we are confident that we would get the same
results if we look from a PRA perspective, because our operational answers
match closely to the PRA results. So we would feel that from a shutdown and
transient mode PRAs that we would also match closely.
MEMBER SIEBER: Okay. Now, did your expert panel
have the opportunity to look at the PRA results at the same time as they did
their own categorization, or was that some kind of a blind test that we could
draw some confidence from?
MR. MOLDENHAUER: No. At the beginning of each of
the sections, each of the systems that we went through to analyze, each
member is responsible for bringing his perspective of that system to the
working group. For example, Licensing brings licensing commitments to the
working group.
Me, as the PRA expert, I bring the PRA stuff. I discuss the
assumptions and the limitations that go into the PRA, and all the analysis that
goes into the PRA, with the working groups before we start addressing any of
the categorization processes.
MEMBER SIEBER: So they are pretty much aware of what
SSCs the PRA has identified as risk important before they make their
independent decision.
CHAIRMAN APOSTOLAKIS: But the vast majority of
SSCs, though, that were categorized were not even in the PRA.
MEMBER SIEBER: That's right.
CHAIRMAN APOSTOLAKIS: So the --
MEMBER KRESS: He's addressing those that overlapped.
CHAIRMAN APOSTOLAKIS: Sorry?
MEMBER KRESS: He's addressing those that overlapped
as a validation of the process.
MEMBER SIEBER: Yes. But if they already knew what
the PRA results were, then it's -- it doesn't validate anything other than it
reinforces what it is they're saying.
MR. MOLDENHAUER: But they don't know the results.
I don't give them the PRA results. We don't give them the risk ranking and
the PRA results until we actually do the components. When we do the system
functions and that, the working group doesn't have those results.
MEMBER SIEBER: Now, another question that sort of
addresses Dr. Bonaca's question, let's say that you had a component, and your
plant is very robust because you have three trains of almost everything, which
gives it a pretty good risk profile, in my opinion.
But if you had, for example, an intermediate head safety
injection pump, and it was risk significant, you know, RSC-1, and it was
supported by a raw water pump that cooled the lubricating oil, that raw water
pump would also be an RSC-1. It would not -- even if you had three or four
trains, it would not end up as a three?
MR. SCHINZEL: What we look at, again, would be the
function that needs to be satisfied. If that function that the safety injection
pump has to satisfy is high safety significant, and if the raw water pump that
cools that system -- if its failure would cause loss of that safety injection
function --
MEMBER SIEBER: Sooner or later it would.
MR. SCHINZEL: -- then it would also be classified as a
safety significant type --
MEMBER SIEBER: Regardless of how many trains you
had.
MR. SCHINZEL: Well, we would look to see, are there
independent and diverse means of satisfying that cooling? And if there are,
there may be a potential of, again, lowering that categorization down by one
level.
MEMBER SIEBER: By one level. Okay.
MR. MOLDENHAUER: And just to add on to that, the
PRA does include those dependencies in that. So if there isn't any cooling
water for the safety injection pump, well, then, the safety injection pump
wouldn't work.
MEMBER SIEBER: Yes. You actually don't have -- you
have three trains, but the three trains are -- they are directly three trains. You
know, you have a raw water pump that supports train A, and everything in it,
including the diesel, right?
MR. MOLDENHAUER: Yes, for the most part.
MEMBER SIEBER: And you do have some cross-ties as I
recall.
MR. MOLDENHAUER: Yes.
MEMBER SIEBER: But those cross-ties are manual, right,
on things like cooling water? The operator has to do something?
MR. MOLDENHAUER: No. For the charging pumps, for
example, the cross-tie data, there is nothing for the operators to do.
MEMBER SIEBER: Oh. Check valves, okay.
MR. MOLDENHAUER: Yes.
MEMBER SIEBER: All right.
MEMBER BONACA: Thank you. Just to specify a little
bit, my concern before was where there is multiple redundancies of a system.
It was to do mostly with the level of testing that you do, etcetera, to that
system if it is a lower category, the reason being that -- my thought process
was you are vulnerable to common mode failures, because many of them are
driven by maintenance practices.
The experience often times is that you do something to a
pump, you put in some seal that is different from the previous one, or the seat
is different, and then you discover through the testing that you have a problem
with one. You go back and check the others, and you find a similar problem.
It is -- this is a common experience there has been.
And that's why I was asking those questions. I wasn't
looking only about the materials. I was thinking about the testing that you
perform for the systems and trains, and that's why I raised the question.
MR. HEAD: And we recognize -- this is Scott Head. We
recognize the limitations, especially with seismic and environmental aspects.
But we do have what we believe is a robust feedback program that involves
the corrective action program and involves periodic assessments of the system
health. It involves the system engineer, the people that are closest to the
system, and we ask them specific questions. What has this process done to
your system?
And right now we don't expect many answers of a positive
nature, because we haven't -- it's mainly in a graded QA form, not with
respect to what we're doing here. But that's a process that will continue
throughout this, you know, that -- it goes with the exemption request, and so
we expect that sort of feedback, and at least opportunities for common load
issues like that to be identified and for us to take appropriate action.
We agree that in seismic and EQ areas they are probably
never going to manifest themselves. But, I mean, if we have a vendor that
ultimately ends up starting to provide us stuff that's not of appropriate quality,
that process could reveal itself there also.
MR. SCHINZEL: And one additional depth of defense
generally is we take trains out of service. You know, we'll perform the
maintenance or the repair activity, replace something. We don't do that to all
trains at once, at the same week. So there is a series of time between the time
you implement -- install a part on one train, before you install it on another
train.
So if it is some time dependency on that failure, we would
expect that feedback process to give us that insight.
MR. CHACKAL: This is Ralph Chackal. We should also
mention that the maintenance rule trends failures at the system and function
level. So if there was a failure of one or more components that failed a
medium or high safety significant function, that would be required under our
program to be identified, trended, and corrective action determined by the
maintenance rule program.
MR. HEAD: I'd like to make one last point that it's a
critical, I guess, safety aspect. Our ECCS trains are not headered or cross-
connected. There is -- aux feedwater is ultimately cross-connected with
operator action, with instrument error.
MEMBER SIEBER: Your auxiliary systems.
MR. HEAD: Yes, sir.
MEMBER SIEBER: Right.
CHAIRMAN APOSTOLAKIS: The issue of functionality
that the staff raised is an interesting one. Using different words, are you
concerned about design and construction errors?
Do these special treatment requirements give you higher
confidence that the components will not have those that will manifest
themselves in a very harsh environment that we're not going to see -- it's
likely we'll never see? Is that really the issue here, design and construction
errors, which PRA doesn't handle at all? It assumes that the component is
capable of performing its function. Or am I using words that should not be
mentioned here?
(Laughter.)
MR. STROSNIDER: I think the answer is yes. I'm not sure
that it's the way I'd characterize it exactly. But if you take the situation where
they would want to replace an existing component, the staff has an expectation
that that replacement component would perform its function throughout its life
in the plant.
And the programs of procurement and maintenance and
surveiling and testing -- and I think there's eight elements in this program --
all of those elements contribute in some way to maintaining that functionality
under design basis conditions. So that is the expectation.
CHAIRMAN APOSTOLAKIS: Which in different words
is they reduce the probability that there is some flaw there that will not allow
the component to function, which is a --
MR. STROSNIDER: Well, okay. But it's more than
looking for -- for random or perhaps isolated fabrication defects. The
situation when you go out to buy a commercial component -- a valve or a
pump or something to replace the existing one -- yes, you can buy commercial
products that have perhaps comparable materials, etcetera, to what's in the
safety-related.
You can buy commercial products that will perform the
normal operating basis that have plastic -- maybe that's an extreme example,
but they may have different materials in them. All right?
So the point is when you go out to procure that sort of
component, that you want to make sure that it's -- that you've laid out exactly
what the design conditions are, that it's going to have perhaps, under an
accident condition, to work in an environment where there's high radiation
and temperature, and that the materials and the design would accommodate
that.
CHAIRMAN APOSTOLAKIS: So it --
MR. STROSNIDER: But it's more than a random -- there's
a defect here or there. It's a more general --
CHAIRMAN APOSTOLAKIS: I didn't say random. I said
design.
MR. STROSNIDER: Okay.
CHAIRMAN APOSTOLAKIS: In other words, if you find
after a strong earthquake that the wrong material had been used, then you
would say it was a design error, wouldn't you?
MR. STROSNIDER: Well, I think, you know, for the
seismic examples what might come more into play is the geometries involved,
and perhaps materials in terms of masses, etcetera, when you get into dynamic
sorts of analyses and whether the internals are going to operate under those
sort of whatever the acceleration forces are.
And then, that's part of why -- well, you know, what we say
is when you go to procure that or to "qualify" it internally, you need to
consider all those inputs. A lot of times the vendor catalogs may say, well,
you know, this is good up to some g level. Is that appropriate for the
application, you know, or depending upon what the seismic consideration is
at the plant, which I think happens to be low at South Texas.
But, so does that answer your question?
CHAIRMAN APOSTOLAKIS: Yes.
MR. SCHINZEL: And, Dr. Apostolakis, I'll point out that,
you know, from the standpoint of our commercial program, we have the same
expectations on the commercial side of the plant. Once we install something,
we're going to post-maintenance test that, and we're going to validate that it's
performing its function.
So, again, we have assurance that that component is doing
what it's intended to do, and we expect that those same practices could be used
on low safety significant and non-risk significant components and give the
same degree of assurance.
MEMBER BONACA: We are asking questions, however,
because on the secondary side of the plant everything runs. And if something
doesn't work, it manifests itself. And most of the safety systems don't run;
they sit there. And so the only way you are going to find out whether or not
it will work, it's really often times an indirect kind of test, a verification, and
you have to rely on those.
There is a difference there. You have to -- I just want to
point out --
CHAIRMAN APOSTOLAKIS: What I'm trying to do is
trying to understand or trying to translate the concerns that staff has into PRA
language. And I think this comment earlier that you really are concerned
about functionality -- that's a very important comment, because PRA assumes
that the thing will start doing its job, and then there is a probability it will fail.
That's a very important point. That's why I raised the issue
of design and construction errors. Now, is that really what you are talking
about?
And a related question is, all of these -- I mean, your
sensitivity analysis was really done on a point estimate basis. You took the
mean values and multiplied by 10, or something like that.
MR. MOLDENHAUER: Yes.
CHAIRMAN APOSTOLAKIS: You didn't look at the
whole distribution. Right. And, I mean, the importance measures are also
derived on the basis of point values, mean values of failure rates, and so on.
And I wonder, when you go to those external events like earthquakes, where
the uncertainties really become very large, where they're deriving importance
-- the importance of the components using just the mean value is good enough.
You see, and the real engineer then worries about it from a
different -- doesn't express it that way. But I think that has a lot to do with
it, and the uncertainty, the spread of the distribution probably plays a role.
Jack, do we have anything else? I think we're running out
of time now.
MR. STROSNIDER: Any questions? No?
MR. HEAD: A closing comment -- it might also answer a
question. The effort that STP is expending on this -- and I'd say also that the
Nuclear Regulatory Commission has been expending -- is extensive. We
embarked upon that a couple of years ago because we felt like that there were
two substantial benefits from this process.
One was we believed that by identifying the Risk 2
components that we would ultimately positively impact plant safety, or, as a
minimum, be neutral between the tradeoff, between Risk 2 and Risk 3. But
we felt like the positive benefits we get from that, from being able to focus not
only on the safety-related components but the non-safety-related components
that were important from a risk perspective, was an overall important thing for
the station and for the public health and safety.
Clearly, the second reason we embarked upon this extensive,
expensive effort is we expected some financial benefit. And that's a -- that's
a driver for this -- you know, for our entire effort. We remain committed to
pursue this because we believe the financial benefit is there. We know we're
taking -- we are getting the advantage of the safety benefit.
And so the question -- the answer to the question about why
pursue Option 2 is we -- there are other plants out there that that's probably
the same case for. There are clearly some plants out there, maybe of a certain
vintage or certain sophistication from a PRA standpoint, that the initial upfront
cost just would not justify this process.
There are other plants that we believe -- and they're out
there we think, and they know they're there -- that this process would provide
those benefits. So I'll use your seismic example. We've never at STP really
used the seismic aspect of STP as important for this process because we want
it to be able to be used across the industry, because we think that as part of
risk-informing the regulations it's important for this -- not only for STP for
this to succeed but for the industry.
And, as I say, we remain committed to pursuing this. We
believe the benefits are tangible and meaningful, so we certainly appreciate the
Nuclear Regulatory Commission's continued involvement in this process.
CHAIRMAN APOSTOLAKIS: I got the impression that the
staff was a little bit more optimistic than you that you will reach consensus
soon. Is that the wrong impression? False impression?
MR. HEAD: I guess what I'm saying is I -- clearly, there's
an exemption request approval coming. We believe that there is one that's
coming this summer. I think what we are wrestling with is, will the financial
or the benefits that we can see in the commercial procurement area, will they
manifest themselves?
Or will the expectations that we believe are there on the part
of the staff -- can we fulfill them without actually building something -- you
know, on one end you have commercial grade, and the other end you have
safety-related. That if you end up spending too much money on this
commercial grade component, at some point in time it makes no sense to
pursue commercial grade, and you'll be left with basically a safety-related
component, which is okay, but it's certainly not what we thought SECY-98-
300 would have said is the way we're going to end up.
So our perspective is that there is an exemption request in
here that is certainly approvable. Whether it's what we ultimately can use to
the full benefit, as we expected, that's still where -- I think what we're still
wrestling with. And there are some -- you know, between the maintenance
rule and a number of other aspects in there, there is clearly some benefits
we're getting, but it's -- there are certain areas we still need to work through.
CHAIRMAN APOSTOLAKIS: Okay.
MEMBER SIEBER: Any other comments?
MR. NAKOSKI: This is John Nakoski. I'd just like to
thank the ACRS for the opportunity to provide you with our insights on the
treatment.
MEMBER SIEBER: Thank you.
Mr. Chairman?
CHAIRMAN APOSTOLAKIS: When is your report coming
to us, the safety evaluation report?
MR. NAKOSKI: To a large extent, it's being driven by
when we can resolve the open items. I think if you look at the first slide, we
would expect to provide the ACRS with the safety evaluation at about the
same time that we provide it to the EDO, which is currently scheduled for
early May.
CHAIRMAN APOSTOLAKIS: And then they will request
a letter when?
MR. NAKOSKI: Well --
CHAIRMAN APOSTOLAKIS: Not in May.
MR. NAKOSKI: -- for categorization, I was hoping we
would have had a letter already, but --
CHAIRMAN APOSTOLAKIS: We were hoping, too.
MR. NAKOSKI: Okay.
CHAIRMAN APOSTOLAKIS: But hopes --
MR. NAKOSKI: We would -- we're planning to talk to the
Commission in early June, and we would hope to have your insights around
that time.
CHAIRMAN APOSTOLAKIS: Around that time. So we'll
have roughly about a month.
MR. NAKOSKI: We're trying to give you that time, yes.
CHAIRMAN APOSTOLAKIS: That would be really
helpful, because there is a lot of material here.
Well, thank you, gentlemen, very much. We appreciate
your coming here and talking to us.
And we will recess until 10:55.
(Whereupon, at 10:38 a.m., the proceedings in the
foregoing matter went off the record.)
Page Last Reviewed/Updated Monday, August 15, 2016